diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,131292 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 18750,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00016,
+      "grad_norm": 10.748890029011417,
+      "learning_rate": 3.552397868561279e-07,
+      "loss": 13.2267,
+      "step": 1
+    },
+    {
+      "epoch": 0.00032,
+      "grad_norm": 8.633134226392258,
+      "learning_rate": 7.104795737122559e-07,
+      "loss": 13.3945,
+      "step": 2
+    },
+    {
+      "epoch": 0.00048,
+      "grad_norm": 7.875647704370841,
+      "learning_rate": 1.0657193605683838e-06,
+      "loss": 13.0379,
+      "step": 3
+    },
+    {
+      "epoch": 0.00064,
+      "grad_norm": 10.984185433719045,
+      "learning_rate": 1.4209591474245117e-06,
+      "loss": 13.3064,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008,
+      "grad_norm": 8.509404941976166,
+      "learning_rate": 1.7761989342806394e-06,
+      "loss": 13.3321,
+      "step": 5
+    },
+    {
+      "epoch": 0.00096,
+      "grad_norm": 15.244023039252511,
+      "learning_rate": 2.1314387211367677e-06,
+      "loss": 13.1662,
+      "step": 6
+    },
+    {
+      "epoch": 0.00112,
+      "grad_norm": 10.070726446073905,
+      "learning_rate": 2.4866785079928953e-06,
+      "loss": 13.0837,
+      "step": 7
+    },
+    {
+      "epoch": 0.00128,
+      "grad_norm": 10.726694668860622,
+      "learning_rate": 2.8419182948490234e-06,
+      "loss": 13.3689,
+      "step": 8
+    },
+    {
+      "epoch": 0.00144,
+      "grad_norm": 16.757036419661855,
+      "learning_rate": 3.197158081705151e-06,
+      "loss": 13.3024,
+      "step": 9
+    },
+    {
+      "epoch": 0.0016,
+      "grad_norm": 13.00908055468233,
+      "learning_rate": 3.5523978685612787e-06,
+      "loss": 12.9443,
+      "step": 10
+    },
+    {
+      "epoch": 0.00176,
+      "grad_norm": 14.147419541641291,
+      "learning_rate": 3.907637655417407e-06,
+      "loss": 12.6594,
+      "step": 11
+    },
+    {
+      "epoch": 0.00192,
+      "grad_norm": 20.1111708834309,
+      "learning_rate": 4.262877442273535e-06,
+      "loss": 13.1741,
+      "step": 12
+    },
+    {
+      "epoch": 0.00208,
+      "grad_norm": 23.941213293631485,
+      "learning_rate": 4.6181172291296625e-06,
+      "loss": 12.9701,
+      "step": 13
+    },
+    {
+      "epoch": 0.00224,
+      "grad_norm": 23.8438984740048,
+      "learning_rate": 4.973357015985791e-06,
+      "loss": 12.5067,
+      "step": 14
+    },
+    {
+      "epoch": 0.0024,
+      "grad_norm": 31.450517327446203,
+      "learning_rate": 5.328596802841918e-06,
+      "loss": 11.2436,
+      "step": 15
+    },
+    {
+      "epoch": 0.00256,
+      "grad_norm": 31.450517327446203,
+      "learning_rate": 5.328596802841918e-06,
+      "loss": 11.505,
+      "step": 16
+    },
+    {
+      "epoch": 0.00272,
+      "grad_norm": 31.450517327446203,
+      "learning_rate": 5.328596802841918e-06,
+      "loss": 11.071,
+      "step": 17
+    },
+    {
+      "epoch": 0.00288,
+      "grad_norm": 36.79241604931043,
+      "learning_rate": 5.683836589698047e-06,
+      "loss": 11.5264,
+      "step": 18
+    },
+    {
+      "epoch": 0.00304,
+      "grad_norm": 27.443985972935277,
+      "learning_rate": 6.039076376554174e-06,
+      "loss": 10.7682,
+      "step": 19
+    },
+    {
+      "epoch": 0.0032,
+      "grad_norm": 39.10500136277021,
+      "learning_rate": 6.394316163410302e-06,
+      "loss": 10.9431,
+      "step": 20
+    },
+    {
+      "epoch": 0.00336,
+      "grad_norm": 34.441561500949625,
+      "learning_rate": 6.74955595026643e-06,
+      "loss": 10.107,
+      "step": 21
+    },
+    {
+      "epoch": 0.00352,
+      "grad_norm": 30.220200506781588,
+      "learning_rate": 7.1047957371225574e-06,
+      "loss": 9.4367,
+      "step": 22
+    },
+    {
+      "epoch": 0.00368,
+      "grad_norm": 24.834599291185945,
+      "learning_rate": 7.4600355239786855e-06,
+      "loss": 8.8942,
+      "step": 23
+    },
+    {
+      "epoch": 0.00384,
+      "grad_norm": 23.944832549644744,
+      "learning_rate": 7.815275310834814e-06,
+      "loss": 8.1854,
+      "step": 24
+    },
+    {
+      "epoch": 0.004,
+      "grad_norm": 17.28132583246269,
+      "learning_rate": 8.170515097690943e-06,
+      "loss": 8.3641,
+      "step": 25
+    },
+    {
+      "epoch": 0.00416,
+      "grad_norm": 13.921865583131904,
+      "learning_rate": 8.52575488454707e-06,
+      "loss": 7.7702,
+      "step": 26
+    },
+    {
+      "epoch": 0.00432,
+      "grad_norm": 8.765143728048887,
+      "learning_rate": 8.880994671403197e-06,
+      "loss": 7.6201,
+      "step": 27
+    },
+    {
+      "epoch": 0.00448,
+      "grad_norm": 8.049071217563789,
+      "learning_rate": 9.236234458259325e-06,
+      "loss": 7.407,
+      "step": 28
+    },
+    {
+      "epoch": 0.00464,
+      "grad_norm": 6.931427454250251,
+      "learning_rate": 9.591474245115453e-06,
+      "loss": 7.3922,
+      "step": 29
+    },
+    {
+      "epoch": 0.0048,
+      "grad_norm": 9.963291903477902,
+      "learning_rate": 9.946714031971581e-06,
+      "loss": 7.4138,
+      "step": 30
+    },
+    {
+      "epoch": 0.00496,
+      "grad_norm": 12.702392969060162,
+      "learning_rate": 1.030195381882771e-05,
+      "loss": 7.5477,
+      "step": 31
+    },
+    {
+      "epoch": 0.00512,
+      "grad_norm": 18.2656846775617,
+      "learning_rate": 1.0657193605683836e-05,
+      "loss": 7.6735,
+      "step": 32
+    },
+    {
+      "epoch": 0.00528,
+      "grad_norm": 11.443539706735015,
+      "learning_rate": 1.1012433392539966e-05,
+      "loss": 7.4873,
+      "step": 33
+    },
+    {
+      "epoch": 0.00544,
+      "grad_norm": 14.05097626563223,
+      "learning_rate": 1.1367673179396094e-05,
+      "loss": 7.6402,
+      "step": 34
+    },
+    {
+      "epoch": 0.0056,
+      "grad_norm": 17.792448791837504,
+      "learning_rate": 1.1722912966252222e-05,
+      "loss": 7.666,
+      "step": 35
+    },
+    {
+      "epoch": 0.00576,
+      "grad_norm": 18.32997347517834,
+      "learning_rate": 1.2078152753108348e-05,
+      "loss": 7.6959,
+      "step": 36
+    },
+    {
+      "epoch": 0.00592,
+      "grad_norm": 18.98379113590679,
+      "learning_rate": 1.2433392539964476e-05,
+      "loss": 7.7474,
+      "step": 37
+    },
+    {
+      "epoch": 0.00608,
+      "grad_norm": 13.876517754637133,
+      "learning_rate": 1.2788632326820604e-05,
+      "loss": 7.5892,
+      "step": 38
+    },
+    {
+      "epoch": 0.00624,
+      "grad_norm": 10.770414094848372,
+      "learning_rate": 1.3143872113676734e-05,
+      "loss": 7.3406,
+      "step": 39
+    },
+    {
+      "epoch": 0.0064,
+      "grad_norm": 7.7287366367664845,
+      "learning_rate": 1.349911190053286e-05,
+      "loss": 7.3276,
+      "step": 40
+    },
+    {
+      "epoch": 0.00656,
+      "grad_norm": 10.293870294038998,
+      "learning_rate": 1.3854351687388988e-05,
+      "loss": 7.2013,
+      "step": 41
+    },
+    {
+      "epoch": 0.00672,
+      "grad_norm": 4.294780674994706,
+      "learning_rate": 1.4209591474245115e-05,
+      "loss": 7.0896,
+      "step": 42
+    },
+    {
+      "epoch": 0.00688,
+      "grad_norm": 4.7363325423780465,
+      "learning_rate": 1.4564831261101245e-05,
+      "loss": 7.1163,
+      "step": 43
+    },
+    {
+      "epoch": 0.00704,
+      "grad_norm": 4.217791950532312,
+      "learning_rate": 1.4920071047957371e-05,
+      "loss": 7.0269,
+      "step": 44
+    },
+    {
+      "epoch": 0.0072,
+      "grad_norm": 6.053065565350551,
+      "learning_rate": 1.52753108348135e-05,
+      "loss": 7.0556,
+      "step": 45
+    },
+    {
+      "epoch": 0.00736,
+      "grad_norm": 8.587164373685457,
+      "learning_rate": 1.563055062166963e-05,
+      "loss": 7.1002,
+      "step": 46
+    },
+    {
+      "epoch": 0.00752,
+      "grad_norm": 5.768646462302799,
+      "learning_rate": 1.5985790408525757e-05,
+      "loss": 7.0773,
+      "step": 47
+    },
+    {
+      "epoch": 0.00768,
+      "grad_norm": 6.026656126273468,
+      "learning_rate": 1.6341030195381885e-05,
+      "loss": 7.0269,
+      "step": 48
+    },
+    {
+      "epoch": 0.00784,
+      "grad_norm": 5.484127640919594,
+      "learning_rate": 1.669626998223801e-05,
+      "loss": 7.1102,
+      "step": 49
+    },
+    {
+      "epoch": 0.008,
+      "grad_norm": 7.504649867138771,
+      "learning_rate": 1.705150976909414e-05,
+      "loss": 7.1157,
+      "step": 50
+    },
+    {
+      "epoch": 0.00816,
+      "grad_norm": 6.572186630817243,
+      "learning_rate": 1.7406749555950266e-05,
+      "loss": 7.1986,
+      "step": 51
+    },
+    {
+      "epoch": 0.00832,
+      "grad_norm": 7.5641245163631075,
+      "learning_rate": 1.7761989342806394e-05,
+      "loss": 7.018,
+      "step": 52
+    },
+    {
+      "epoch": 0.00848,
+      "grad_norm": 4.821251955598482,
+      "learning_rate": 1.8117229129662522e-05,
+      "loss": 6.967,
+      "step": 53
+    },
+    {
+      "epoch": 0.00864,
+      "grad_norm": 6.1137771051910015,
+      "learning_rate": 1.847246891651865e-05,
+      "loss": 7.0156,
+      "step": 54
+    },
+    {
+      "epoch": 0.0088,
+      "grad_norm": 7.681310106572676,
+      "learning_rate": 1.882770870337478e-05,
+      "loss": 6.9989,
+      "step": 55
+    },
+    {
+      "epoch": 0.00896,
+      "grad_norm": 3.6949094520150902,
+      "learning_rate": 1.9182948490230906e-05,
+      "loss": 6.8998,
+      "step": 56
+    },
+    {
+      "epoch": 0.00912,
+      "grad_norm": 3.8621587177741525,
+      "learning_rate": 1.9538188277087034e-05,
+      "loss": 6.8908,
+      "step": 57
+    },
+    {
+      "epoch": 0.00928,
+      "grad_norm": 4.348926673012381,
+      "learning_rate": 1.9893428063943163e-05,
+      "loss": 6.8632,
+      "step": 58
+    },
+    {
+      "epoch": 0.00944,
+      "grad_norm": 4.6669599382200655,
+      "learning_rate": 2.024866785079929e-05,
+      "loss": 6.8173,
+      "step": 59
+    },
+    {
+      "epoch": 0.0096,
+      "grad_norm": 4.065753177915794,
+      "learning_rate": 2.060390763765542e-05,
+      "loss": 6.8425,
+      "step": 60
+    },
+    {
+      "epoch": 0.00976,
+      "grad_norm": 6.743640210424077,
+      "learning_rate": 2.0959147424511547e-05,
+      "loss": 6.8788,
+      "step": 61
+    },
+    {
+      "epoch": 0.00992,
+      "grad_norm": 3.4332311194507157,
+      "learning_rate": 2.131438721136767e-05,
+      "loss": 6.7814,
+      "step": 62
+    },
+    {
+      "epoch": 0.01008,
+      "grad_norm": 5.312435765000791,
+      "learning_rate": 2.1669626998223803e-05,
+      "loss": 6.8385,
+      "step": 63
+    },
+    {
+      "epoch": 0.01024,
+      "grad_norm": 4.897691150210214,
+      "learning_rate": 2.202486678507993e-05,
+      "loss": 6.7889,
+      "step": 64
+    },
+    {
+      "epoch": 0.0104,
+      "grad_norm": 7.219656365769396,
+      "learning_rate": 2.238010657193606e-05,
+      "loss": 6.783,
+      "step": 65
+    },
+    {
+      "epoch": 0.01056,
+      "grad_norm": 5.264979423551402,
+      "learning_rate": 2.2735346358792187e-05,
+      "loss": 6.748,
+      "step": 66
+    },
+    {
+      "epoch": 0.01072,
+      "grad_norm": 4.186909646744688,
+      "learning_rate": 2.3090586145648312e-05,
+      "loss": 6.7143,
+      "step": 67
+    },
+    {
+      "epoch": 0.01088,
+      "grad_norm": 5.109656614479401,
+      "learning_rate": 2.3445825932504443e-05,
+      "loss": 6.7134,
+      "step": 68
+    },
+    {
+      "epoch": 0.01104,
+      "grad_norm": 3.4500539271326187,
+      "learning_rate": 2.3801065719360568e-05,
+      "loss": 6.6523,
+      "step": 69
+    },
+    {
+      "epoch": 0.0112,
+      "grad_norm": 3.4988547683930395,
+      "learning_rate": 2.4156305506216696e-05,
+      "loss": 6.634,
+      "step": 70
+    },
+    {
+      "epoch": 0.01136,
+      "grad_norm": 3.7057460914011093,
+      "learning_rate": 2.4511545293072824e-05,
+      "loss": 6.6448,
+      "step": 71
+    },
+    {
+      "epoch": 0.01152,
+      "grad_norm": 4.74460469334246,
+      "learning_rate": 2.4866785079928952e-05,
+      "loss": 6.661,
+      "step": 72
+    },
+    {
+      "epoch": 0.01168,
+      "grad_norm": 4.962898079680001,
+      "learning_rate": 2.5222024866785084e-05,
+      "loss": 6.6188,
+      "step": 73
+    },
+    {
+      "epoch": 0.01184,
+      "grad_norm": 3.640355688835305,
+      "learning_rate": 2.557726465364121e-05,
+      "loss": 6.6029,
+      "step": 74
+    },
+    {
+      "epoch": 0.012,
+      "grad_norm": 5.075124445947665,
+      "learning_rate": 2.5932504440497337e-05,
+      "loss": 6.5705,
+      "step": 75
+    },
+    {
+      "epoch": 0.01216,
+      "grad_norm": 5.041669806459398,
+      "learning_rate": 2.6287744227353468e-05,
+      "loss": 6.5401,
+      "step": 76
+    },
+    {
+      "epoch": 0.01232,
+      "grad_norm": 4.206575620884844,
+      "learning_rate": 2.6642984014209593e-05,
+      "loss": 6.5455,
+      "step": 77
+    },
+    {
+      "epoch": 0.01248,
+      "grad_norm": 2.4118992375262773,
+      "learning_rate": 2.699822380106572e-05,
+      "loss": 6.4816,
+      "step": 78
+    },
+    {
+      "epoch": 0.01264,
+      "grad_norm": 3.352669467495514,
+      "learning_rate": 2.7353463587921846e-05,
+      "loss": 6.5617,
+      "step": 79
+    },
+    {
+      "epoch": 0.0128,
+      "grad_norm": 3.259212074412052,
+      "learning_rate": 2.7708703374777977e-05,
+      "loss": 6.4684,
+      "step": 80
+    },
+    {
+      "epoch": 0.01296,
+      "grad_norm": 4.944124657347146,
+      "learning_rate": 2.8063943161634105e-05,
+      "loss": 6.4678,
+      "step": 81
+    },
+    {
+      "epoch": 0.01312,
+      "grad_norm": 5.4062336064583505,
+      "learning_rate": 2.841918294849023e-05,
+      "loss": 6.482,
+      "step": 82
+    },
+    {
+      "epoch": 0.01328,
+      "grad_norm": 3.506113879056529,
+      "learning_rate": 2.877442273534636e-05,
+      "loss": 6.4373,
+      "step": 83
+    },
+    {
+      "epoch": 0.01344,
+      "grad_norm": 2.768463691262186,
+      "learning_rate": 2.912966252220249e-05,
+      "loss": 6.3819,
+      "step": 84
+    },
+    {
+      "epoch": 0.0136,
+      "grad_norm": 3.2335027897775728,
+      "learning_rate": 2.9484902309058614e-05,
+      "loss": 6.4262,
+      "step": 85
+    },
+    {
+      "epoch": 0.01376,
+      "grad_norm": 3.257575581182437,
+      "learning_rate": 2.9840142095914742e-05,
+      "loss": 6.4341,
+      "step": 86
+    },
+    {
+      "epoch": 0.01392,
+      "grad_norm": 3.0905828897758916,
+      "learning_rate": 3.0195381882770874e-05,
+      "loss": 6.3704,
+      "step": 87
+    },
+    {
+      "epoch": 0.01408,
+      "grad_norm": 2.7606811728589964,
+      "learning_rate": 3.0550621669627e-05,
+      "loss": 6.3845,
+      "step": 88
+    },
+    {
+      "epoch": 0.01424,
+      "grad_norm": 4.857277981839788,
+      "learning_rate": 3.090586145648312e-05,
+      "loss": 6.4093,
+      "step": 89
+    },
+    {
+      "epoch": 0.0144,
+      "grad_norm": 1.8999210751257682,
+      "learning_rate": 3.126110124333926e-05,
+      "loss": 6.3365,
+      "step": 90
+    },
+    {
+      "epoch": 0.01456,
+      "grad_norm": 3.567282927553493,
+      "learning_rate": 3.1616341030195386e-05,
+      "loss": 6.3595,
+      "step": 91
+    },
+    {
+      "epoch": 0.01472,
+      "grad_norm": 2.4714613285668174,
+      "learning_rate": 3.1971580817051514e-05,
+      "loss": 6.3106,
+      "step": 92
+    },
+    {
+      "epoch": 0.01488,
+      "grad_norm": 3.574189647636856,
+      "learning_rate": 3.2326820603907635e-05,
+      "loss": 6.3457,
+      "step": 93
+    },
+    {
+      "epoch": 0.01504,
+      "grad_norm": 3.34142888213726,
+      "learning_rate": 3.268206039076377e-05,
+      "loss": 6.3475,
+      "step": 94
+    },
+    {
+      "epoch": 0.0152,
+      "grad_norm": 2.7292127489776505,
+      "learning_rate": 3.30373001776199e-05,
+      "loss": 6.3508,
+      "step": 95
+    },
+    {
+      "epoch": 0.01536,
+      "grad_norm": 3.0917905838231237,
+      "learning_rate": 3.339253996447602e-05,
+      "loss": 6.3443,
+      "step": 96
+    },
+    {
+      "epoch": 0.01552,
+      "grad_norm": 3.0640675951533822,
+      "learning_rate": 3.374777975133215e-05,
+      "loss": 6.314,
+      "step": 97
+    },
+    {
+      "epoch": 0.01568,
+      "grad_norm": 2.3659640255495447,
+      "learning_rate": 3.410301953818828e-05,
+      "loss": 6.3001,
+      "step": 98
+    },
+    {
+      "epoch": 0.01584,
+      "grad_norm": 3.2890578235583563,
+      "learning_rate": 3.4458259325044404e-05,
+      "loss": 6.3024,
+      "step": 99
+    },
+    {
+      "epoch": 0.016,
+      "grad_norm": 3.4095019878903132,
+      "learning_rate": 3.481349911190053e-05,
+      "loss": 6.365,
+      "step": 100
+    },
+    {
+      "epoch": 0.01616,
+      "grad_norm": 2.3058361277718222,
+      "learning_rate": 3.516873889875667e-05,
+      "loss": 6.2702,
+      "step": 101
+    },
+    {
+      "epoch": 0.01632,
+      "grad_norm": 4.222808087152799,
+      "learning_rate": 3.552397868561279e-05,
+      "loss": 6.2614,
+      "step": 102
+    },
+    {
+      "epoch": 0.01648,
+      "grad_norm": 4.8394217495391345,
+      "learning_rate": 3.5879218472468916e-05,
+      "loss": 6.2699,
+      "step": 103
+    },
+    {
+      "epoch": 0.01664,
+      "grad_norm": 4.317178309432362,
+      "learning_rate": 3.6234458259325044e-05,
+      "loss": 6.2838,
+      "step": 104
+    },
+    {
+      "epoch": 0.0168,
+      "grad_norm": 2.6853791877456037,
+      "learning_rate": 3.658969804618117e-05,
+      "loss": 6.2288,
+      "step": 105
+    },
+    {
+      "epoch": 0.01696,
+      "grad_norm": 2.6729278491718804,
+      "learning_rate": 3.69449378330373e-05,
+      "loss": 6.2534,
+      "step": 106
+    },
+    {
+      "epoch": 0.01712,
+      "grad_norm": 3.1287517705922636,
+      "learning_rate": 3.730017761989343e-05,
+      "loss": 6.2148,
+      "step": 107
+    },
+    {
+      "epoch": 0.01728,
+      "grad_norm": 5.6491645514730875,
+      "learning_rate": 3.765541740674956e-05,
+      "loss": 6.2394,
+      "step": 108
+    },
+    {
+      "epoch": 0.01744,
+      "grad_norm": 2.5065662633156975,
+      "learning_rate": 3.8010657193605685e-05,
+      "loss": 6.2125,
+      "step": 109
+    },
+    {
+      "epoch": 0.0176,
+      "grad_norm": 4.170158830008617,
+      "learning_rate": 3.836589698046181e-05,
+      "loss": 6.2352,
+      "step": 110
+    },
+    {
+      "epoch": 0.01776,
+      "grad_norm": 5.369928376523471,
+      "learning_rate": 3.872113676731794e-05,
+      "loss": 6.257,
+      "step": 111
+    },
+    {
+      "epoch": 0.01792,
+      "grad_norm": 2.739775329061468,
+      "learning_rate": 3.907637655417407e-05,
+      "loss": 6.2194,
+      "step": 112
+    },
+    {
+      "epoch": 0.01808,
+      "grad_norm": 3.528202158179349,
+      "learning_rate": 3.94316163410302e-05,
+      "loss": 6.2152,
+      "step": 113
+    },
+    {
+      "epoch": 0.01824,
+      "grad_norm": 5.066308770238294,
+      "learning_rate": 3.9786856127886325e-05,
+      "loss": 6.1918,
+      "step": 114
+    },
+    {
+      "epoch": 0.0184,
+      "grad_norm": 5.081114029780668,
+      "learning_rate": 4.014209591474245e-05,
+      "loss": 6.2043,
+      "step": 115
+    },
+    {
+      "epoch": 0.01856,
+      "grad_norm": 3.6172785399928884,
+      "learning_rate": 4.049733570159858e-05,
+      "loss": 6.193,
+      "step": 116
+    },
+    {
+      "epoch": 0.01872,
+      "grad_norm": 7.9045029176919765,
+      "learning_rate": 4.085257548845471e-05,
+      "loss": 6.2,
+      "step": 117
+    },
+    {
+      "epoch": 0.01888,
+      "grad_norm": 2.25587429176132,
+      "learning_rate": 4.120781527531084e-05,
+      "loss": 6.1479,
+      "step": 118
+    },
+    {
+      "epoch": 0.01904,
+      "grad_norm": 3.0742331853287426,
+      "learning_rate": 4.1563055062166965e-05,
+      "loss": 6.1601,
+      "step": 119
+    },
+    {
+      "epoch": 0.0192,
+      "grad_norm": 5.621901513007863,
+      "learning_rate": 4.1918294849023094e-05,
+      "loss": 6.1323,
+      "step": 120
+    },
+    {
+      "epoch": 0.01936,
+      "grad_norm": 4.619245330977712,
+      "learning_rate": 4.227353463587922e-05,
+      "loss": 6.125,
+      "step": 121
+    },
+    {
+      "epoch": 0.01952,
+      "grad_norm": 3.3718167771582865,
+      "learning_rate": 4.262877442273534e-05,
+      "loss": 6.1212,
+      "step": 122
+    },
+    {
+      "epoch": 0.01968,
+      "grad_norm": 3.6758392081241866,
+      "learning_rate": 4.298401420959148e-05,
+      "loss": 6.1248,
+      "step": 123
+    },
+    {
+      "epoch": 0.01984,
+      "grad_norm": 3.603169143099463,
+      "learning_rate": 4.3339253996447606e-05,
+      "loss": 6.141,
+      "step": 124
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 4.4328154500444565,
+      "learning_rate": 4.369449378330373e-05,
+      "loss": 6.1496,
+      "step": 125
+    },
+    {
+      "epoch": 0.02016,
+      "grad_norm": 4.307993055152668,
+      "learning_rate": 4.404973357015986e-05,
+      "loss": 6.1123,
+      "step": 126
+    },
+    {
+      "epoch": 0.02032,
+      "grad_norm": 2.4789972706369316,
+      "learning_rate": 4.440497335701599e-05,
+      "loss": 6.1147,
+      "step": 127
+    },
+    {
+      "epoch": 0.02048,
+      "grad_norm": 3.645105177750783,
+      "learning_rate": 4.476021314387212e-05,
+      "loss": 6.139,
+      "step": 128
+    },
+    {
+      "epoch": 0.02064,
+      "grad_norm": 3.6938849964920784,
+      "learning_rate": 4.511545293072824e-05,
+      "loss": 6.1368,
+      "step": 129
+    },
+    {
+      "epoch": 0.0208,
+      "grad_norm": 4.227815247551008,
+      "learning_rate": 4.5470692717584374e-05,
+      "loss": 6.0999,
+      "step": 130
+    },
+    {
+      "epoch": 0.02096,
+      "grad_norm": 2.4543446624121477,
+      "learning_rate": 4.58259325044405e-05,
+      "loss": 6.0854,
+      "step": 131
+    },
+    {
+      "epoch": 0.02112,
+      "grad_norm": 5.5826690078758965,
+      "learning_rate": 4.6181172291296624e-05,
+      "loss": 6.0424,
+      "step": 132
+    },
+    {
+      "epoch": 0.02128,
+      "grad_norm": 2.2812723925654295,
+      "learning_rate": 4.653641207815276e-05,
+      "loss": 6.0203,
+      "step": 133
+    },
+    {
+      "epoch": 0.02144,
+      "grad_norm": 4.112423023508773,
+      "learning_rate": 4.689165186500889e-05,
+      "loss": 6.0627,
+      "step": 134
+    },
+    {
+      "epoch": 0.0216,
+      "grad_norm": 2.5588165366842546,
+      "learning_rate": 4.724689165186501e-05,
+      "loss": 6.0574,
+      "step": 135
+    },
+    {
+      "epoch": 0.02176,
+      "grad_norm": 4.527803535087905,
+      "learning_rate": 4.7602131438721136e-05,
+      "loss": 6.0104,
+      "step": 136
+    },
+    {
+      "epoch": 0.02192,
+      "grad_norm": 3.4478123632482607,
+      "learning_rate": 4.795737122557727e-05,
+      "loss": 6.0214,
+      "step": 137
+    },
+    {
+      "epoch": 0.02208,
+      "grad_norm": 7.44164879120778,
+      "learning_rate": 4.831261101243339e-05,
+      "loss": 6.0564,
+      "step": 138
+    },
+    {
+      "epoch": 0.02224,
+      "grad_norm": 1.9569612729033925,
+      "learning_rate": 4.866785079928952e-05,
+      "loss": 6.0325,
+      "step": 139
+    },
+    {
+      "epoch": 0.0224,
+      "grad_norm": 6.38822253813897,
+      "learning_rate": 4.902309058614565e-05,
+      "loss": 6.0185,
+      "step": 140
+    },
+    {
+      "epoch": 0.02256,
+      "grad_norm": 7.218502117056281,
+      "learning_rate": 4.9378330373001777e-05,
+      "loss": 5.9876,
+      "step": 141
+    },
+    {
+      "epoch": 0.02272,
+      "grad_norm": 3.641526060870217,
+      "learning_rate": 4.9733570159857905e-05,
+      "loss": 5.9429,
+      "step": 142
+    },
+    {
+      "epoch": 0.02288,
+      "grad_norm": 7.8732038746418045,
+      "learning_rate": 5.008880994671403e-05,
+      "loss": 6.0414,
+      "step": 143
+    },
+    {
+      "epoch": 0.02304,
+      "grad_norm": 4.005914511212812,
+      "learning_rate": 5.044404973357017e-05,
+      "loss": 5.9619,
+      "step": 144
+    },
+    {
+      "epoch": 0.0232,
+      "grad_norm": 11.90103437163848,
+      "learning_rate": 5.079928952042628e-05,
+      "loss": 6.0042,
+      "step": 145
+    },
+    {
+      "epoch": 0.02336,
+      "grad_norm": 13.816149988297365,
+      "learning_rate": 5.115452930728242e-05,
+      "loss": 6.1429,
+      "step": 146
+    },
+    {
+      "epoch": 0.02352,
+      "grad_norm": 3.8472829790232437,
+      "learning_rate": 5.150976909413855e-05,
+      "loss": 5.9204,
+      "step": 147
+    },
+    {
+      "epoch": 0.02368,
+      "grad_norm": 19.526783410510966,
+      "learning_rate": 5.186500888099467e-05,
+      "loss": 6.129,
+      "step": 148
+    },
+    {
+      "epoch": 0.02384,
+      "grad_norm": 7.815186027063752,
+      "learning_rate": 5.22202486678508e-05,
+      "loss": 5.9644,
+      "step": 149
+    },
+    {
+      "epoch": 0.024,
+      "grad_norm": 16.443836123348067,
+      "learning_rate": 5.2575488454706936e-05,
+      "loss": 6.0365,
+      "step": 150
+    },
+    {
+      "epoch": 0.02416,
+      "grad_norm": 7.1128645312151,
+      "learning_rate": 5.293072824156306e-05,
+      "loss": 5.9482,
+      "step": 151
+    },
+    {
+      "epoch": 0.02432,
+      "grad_norm": 12.930314021619239,
+      "learning_rate": 5.3285968028419185e-05,
+      "loss": 5.9386,
+      "step": 152
+    },
+    {
+      "epoch": 0.02448,
+      "grad_norm": 3.7690274622134017,
+      "learning_rate": 5.364120781527532e-05,
+      "loss": 5.9137,
+      "step": 153
+    },
+    {
+      "epoch": 0.02464,
+      "grad_norm": 9.963505539653708,
+      "learning_rate": 5.399644760213144e-05,
+      "loss": 5.9821,
+      "step": 154
+    },
+    {
+      "epoch": 0.0248,
+      "grad_norm": 3.5218220880132955,
+      "learning_rate": 5.435168738898757e-05,
+      "loss": 5.8936,
+      "step": 155
+    },
+    {
+      "epoch": 0.02496,
+      "grad_norm": 6.5463808377913075,
+      "learning_rate": 5.470692717584369e-05,
+      "loss": 5.9238,
+      "step": 156
+    },
+    {
+      "epoch": 0.02512,
+      "grad_norm": 3.536035349068856,
+      "learning_rate": 5.5062166962699826e-05,
+      "loss": 5.876,
+      "step": 157
+    },
+    {
+      "epoch": 0.02528,
+      "grad_norm": 3.6687480204220875,
+      "learning_rate": 5.5417406749555954e-05,
+      "loss": 5.9221,
+      "step": 158
+    },
+    {
+      "epoch": 0.02544,
+      "grad_norm": 2.916693398767227,
+      "learning_rate": 5.5772646536412075e-05,
+      "loss": 5.833,
+      "step": 159
+    },
+    {
+      "epoch": 0.0256,
+      "grad_norm": 6.209379571704568,
+      "learning_rate": 5.612788632326821e-05,
+      "loss": 5.8987,
+      "step": 160
+    },
+    {
+      "epoch": 0.02576,
+      "grad_norm": 4.6670861927190685,
+      "learning_rate": 5.648312611012434e-05,
+      "loss": 5.8388,
+      "step": 161
+    },
+    {
+      "epoch": 0.02592,
+      "grad_norm": 2.9313538769754692,
+      "learning_rate": 5.683836589698046e-05,
+      "loss": 5.8019,
+      "step": 162
+    },
+    {
+      "epoch": 0.02608,
+      "grad_norm": 3.3041634611249986,
+      "learning_rate": 5.7193605683836594e-05,
+      "loss": 5.8025,
+      "step": 163
+    },
+    {
+      "epoch": 0.02624,
+      "grad_norm": 5.248926579136218,
+      "learning_rate": 5.754884547069272e-05,
+      "loss": 5.791,
+      "step": 164
+    },
+    {
+      "epoch": 0.0264,
+      "grad_norm": 5.8942286881494965,
+      "learning_rate": 5.7904085257548844e-05,
+      "loss": 5.7918,
+      "step": 165
+    },
+    {
+      "epoch": 0.02656,
+      "grad_norm": 3.523492733694665,
+      "learning_rate": 5.825932504440498e-05,
+      "loss": 5.8006,
+      "step": 166
+    },
+    {
+      "epoch": 0.02672,
+      "grad_norm": 7.179280168333415,
+      "learning_rate": 5.861456483126111e-05,
+      "loss": 5.7342,
+      "step": 167
+    },
+    {
+      "epoch": 0.02688,
+      "grad_norm": 4.996506500428147,
+      "learning_rate": 5.896980461811723e-05,
+      "loss": 5.7421,
+      "step": 168
+    },
+    {
+      "epoch": 0.02704,
+      "grad_norm": 3.900166772661288,
+      "learning_rate": 5.932504440497336e-05,
+      "loss": 5.7123,
+      "step": 169
+    },
+    {
+      "epoch": 0.0272,
+      "grad_norm": 10.826422726290046,
+      "learning_rate": 5.9680284191829484e-05,
+      "loss": 5.7385,
+      "step": 170
+    },
+    {
+      "epoch": 0.02736,
+      "grad_norm": 13.304874914723065,
+      "learning_rate": 6.003552397868561e-05,
+      "loss": 5.7408,
+      "step": 171
+    },
+    {
+      "epoch": 0.02752,
+      "grad_norm": 8.445243752241202,
+      "learning_rate": 6.039076376554175e-05,
+      "loss": 5.7219,
+      "step": 172
+    },
+    {
+      "epoch": 0.02768,
+      "grad_norm": 3.504353258847423,
+      "learning_rate": 6.074600355239787e-05,
+      "loss": 5.6745,
+      "step": 173
+    },
+    {
+      "epoch": 0.02784,
+      "grad_norm": 13.921697870459203,
+      "learning_rate": 6.1101243339254e-05,
+      "loss": 5.7555,
+      "step": 174
+    },
+    {
+      "epoch": 0.028,
+      "grad_norm": 19.885042048871856,
+      "learning_rate": 6.145648312611014e-05,
+      "loss": 5.8133,
+      "step": 175
+    },
+    {
+      "epoch": 0.02816,
+      "grad_norm": 11.477260257259386,
+      "learning_rate": 6.181172291296625e-05,
+      "loss": 5.7502,
+      "step": 176
+    },
+    {
+      "epoch": 0.02832,
+      "grad_norm": 12.365220450157118,
+      "learning_rate": 6.216696269982238e-05,
+      "loss": 5.6682,
+      "step": 177
+    },
+    {
+      "epoch": 0.02848,
+      "grad_norm": 14.749524845270633,
+      "learning_rate": 6.252220248667852e-05,
+      "loss": 5.7343,
+      "step": 178
+    },
+    {
+      "epoch": 0.02864,
+      "grad_norm": 2.618379997399385,
+      "learning_rate": 6.287744227353464e-05,
+      "loss": 5.6264,
+      "step": 179
+    },
+    {
+      "epoch": 0.0288,
+      "grad_norm": 5.365282184612086,
+      "learning_rate": 6.323268206039077e-05,
+      "loss": 5.6077,
+      "step": 180
+    },
+    {
+      "epoch": 0.02896,
+      "grad_norm": 4.664584753827127,
+      "learning_rate": 6.358792184724689e-05,
+      "loss": 5.6895,
+      "step": 181
+    },
+    {
+      "epoch": 0.02912,
+      "grad_norm": 3.919035092107766,
+      "learning_rate": 6.394316163410303e-05,
+      "loss": 5.5971,
+      "step": 182
+    },
+    {
+      "epoch": 0.02928,
+      "grad_norm": 3.408624745559893,
+      "learning_rate": 6.429840142095915e-05,
+      "loss": 5.5877,
+      "step": 183
+    },
+    {
+      "epoch": 0.02944,
+      "grad_norm": 7.119233560478808,
+      "learning_rate": 6.465364120781527e-05,
+      "loss": 5.5457,
+      "step": 184
+    },
+    {
+      "epoch": 0.0296,
+      "grad_norm": 11.143795931980051,
+      "learning_rate": 6.50088809946714e-05,
+      "loss": 5.6756,
+      "step": 185
+    },
+    {
+      "epoch": 0.02976,
+      "grad_norm": 3.5497527332682015,
+      "learning_rate": 6.536412078152754e-05,
+      "loss": 5.4997,
+      "step": 186
+    },
+    {
+      "epoch": 0.02992,
+      "grad_norm": 5.87663897037261,
+      "learning_rate": 6.571936056838366e-05,
+      "loss": 5.4793,
+      "step": 187
+    },
+    {
+      "epoch": 0.03008,
+      "grad_norm": 6.7367921328205815,
+      "learning_rate": 6.60746003552398e-05,
+      "loss": 5.5471,
+      "step": 188
+    },
+    {
+      "epoch": 0.03024,
+      "grad_norm": 5.336931982211002,
+      "learning_rate": 6.642984014209592e-05,
+      "loss": 5.4736,
+      "step": 189
+    },
+    {
+      "epoch": 0.0304,
+      "grad_norm": 2.8509577330465383,
+      "learning_rate": 6.678507992895204e-05,
+      "loss": 5.4982,
+      "step": 190
+    },
+    {
+      "epoch": 0.03056,
+      "grad_norm": 7.334024558674715,
+      "learning_rate": 6.714031971580817e-05,
+      "loss": 5.4593,
+      "step": 191
+    },
+    {
+      "epoch": 0.03072,
+      "grad_norm": 6.857151826111544,
+      "learning_rate": 6.74955595026643e-05,
+      "loss": 5.4987,
+      "step": 192
+    },
+    {
+      "epoch": 0.03088,
+      "grad_norm": 11.72857868199863,
+      "learning_rate": 6.785079928952043e-05,
+      "loss": 5.4978,
+      "step": 193
+    },
+    {
+      "epoch": 0.03104,
+      "grad_norm": 6.659028954379637,
+      "learning_rate": 6.820603907637657e-05,
+      "loss": 5.4939,
+      "step": 194
+    },
+    {
+      "epoch": 0.0312,
+      "grad_norm": 4.816219332648455,
+      "learning_rate": 6.856127886323269e-05,
+      "loss": 5.4574,
+      "step": 195
+    },
+    {
+      "epoch": 0.03136,
+      "grad_norm": 10.317237698084652,
+      "learning_rate": 6.891651865008881e-05,
+      "loss": 5.4666,
+      "step": 196
+    },
+    {
+      "epoch": 0.03152,
+      "grad_norm": 3.6479189586443845,
+      "learning_rate": 6.927175843694494e-05,
+      "loss": 5.398,
+      "step": 197
+    },
+    {
+      "epoch": 0.03168,
+      "grad_norm": 6.632737333904688,
+      "learning_rate": 6.962699822380106e-05,
+      "loss": 5.4476,
+      "step": 198
+    },
+    {
+      "epoch": 0.03184,
+      "grad_norm": 12.772868942445378,
+      "learning_rate": 6.99822380106572e-05,
+      "loss": 5.4683,
+      "step": 199
+    },
+    {
+      "epoch": 0.032,
+      "grad_norm": 5.363243425010833,
+      "learning_rate": 7.033747779751333e-05,
+      "loss": 5.4212,
+      "step": 200
+    },
+    {
+      "epoch": 0.03216,
+      "grad_norm": 11.269705813773186,
+      "learning_rate": 7.069271758436945e-05,
+      "loss": 5.5339,
+      "step": 201
+    },
+    {
+      "epoch": 0.03232,
+      "grad_norm": 12.424274829053584,
+      "learning_rate": 7.104795737122558e-05,
+      "loss": 5.5372,
+      "step": 202
+    },
+    {
+      "epoch": 0.03248,
+      "grad_norm": 3.42874824578121,
+      "learning_rate": 7.140319715808171e-05,
+      "loss": 5.43,
+      "step": 203
+    },
+    {
+      "epoch": 0.03264,
+      "grad_norm": 10.398064736617805,
+      "learning_rate": 7.175843694493783e-05,
+      "loss": 5.3952,
+      "step": 204
+    },
+    {
+      "epoch": 0.0328,
+      "grad_norm": 5.917209397752171,
+      "learning_rate": 7.211367673179397e-05,
+      "loss": 5.3678,
+      "step": 205
+    },
+    {
+      "epoch": 0.03296,
+      "grad_norm": 6.419677330128784,
+      "learning_rate": 7.246891651865009e-05,
+      "loss": 5.4153,
+      "step": 206
+    },
+    {
+      "epoch": 0.03312,
+      "grad_norm": 10.666759583946586,
+      "learning_rate": 7.282415630550622e-05,
+      "loss": 5.4272,
+      "step": 207
+    },
+    {
+      "epoch": 0.03328,
+      "grad_norm": 2.9949633307934285,
+      "learning_rate": 7.317939609236234e-05,
+      "loss": 5.3646,
+      "step": 208
+    },
+    {
+      "epoch": 0.03344,
+      "grad_norm": 9.822527447872957,
+      "learning_rate": 7.353463587921847e-05,
+      "loss": 5.3383,
+      "step": 209
+    },
+    {
+      "epoch": 0.0336,
+      "grad_norm": 5.599896133233406,
+      "learning_rate": 7.38898756660746e-05,
+      "loss": 5.3156,
+      "step": 210
+    },
+    {
+      "epoch": 0.03376,
+      "grad_norm": 16.20878766063811,
+      "learning_rate": 7.424511545293074e-05,
+      "loss": 5.4503,
+      "step": 211
+    },
+    {
+      "epoch": 0.03392,
+      "grad_norm": 24.184919914000034,
+      "learning_rate": 7.460035523978686e-05,
+      "loss": 5.5402,
+      "step": 212
+    },
+    {
+      "epoch": 0.03408,
+      "grad_norm": 3.640320065407538,
+      "learning_rate": 7.495559502664299e-05,
+      "loss": 5.3419,
+      "step": 213
+    },
+    {
+      "epoch": 0.03424,
+      "grad_norm": 16.659742793624538,
+      "learning_rate": 7.531083481349913e-05,
+      "loss": 5.4872,
+      "step": 214
+    },
+    {
+      "epoch": 0.0344,
+      "grad_norm": 5.114042223430297,
+      "learning_rate": 7.566607460035523e-05,
+      "loss": 5.3459,
+      "step": 215
+    },
+    {
+      "epoch": 0.03456,
+      "grad_norm": 16.629638022679206,
+      "learning_rate": 7.602131438721137e-05,
+      "loss": 5.5772,
+      "step": 216
+    },
+    {
+      "epoch": 0.03472,
+      "grad_norm": 5.871731727817278,
+      "learning_rate": 7.637655417406749e-05,
+      "loss": 5.3587,
+      "step": 217
+    },
+    {
+      "epoch": 0.03488,
+      "grad_norm": 15.355360987577962,
+      "learning_rate": 7.673179396092363e-05,
+      "loss": 5.4315,
+      "step": 218
+    },
+    {
+      "epoch": 0.03504,
+      "grad_norm": 5.774432609363165,
+      "learning_rate": 7.708703374777976e-05,
+      "loss": 5.3143,
+      "step": 219
+    },
+    {
+      "epoch": 0.0352,
+      "grad_norm": 8.75354387349498,
+      "learning_rate": 7.744227353463588e-05,
+      "loss": 5.4038,
+      "step": 220
+    },
+    {
+      "epoch": 0.03536,
+      "grad_norm": 14.732113089139183,
+      "learning_rate": 7.779751332149202e-05,
+      "loss": 5.4651,
+      "step": 221
+    },
+    {
+      "epoch": 0.03552,
+      "grad_norm": 5.757802099424904,
+      "learning_rate": 7.815275310834814e-05,
+      "loss": 5.2946,
+      "step": 222
+    },
+    {
+      "epoch": 0.03568,
+      "grad_norm": 19.271731519172715,
+      "learning_rate": 7.850799289520426e-05,
+      "loss": 5.5807,
+      "step": 223
+    },
+    {
+      "epoch": 0.03584,
+      "grad_norm": 6.777310396189338,
+      "learning_rate": 7.88632326820604e-05,
+      "loss": 5.3395,
+      "step": 224
+    },
+    {
+      "epoch": 0.036,
+      "grad_norm": 13.598995169943139,
+      "learning_rate": 7.921847246891653e-05,
+      "loss": 5.4458,
+      "step": 225
+    },
+    {
+      "epoch": 0.03616,
+      "grad_norm": 4.56587577257147,
+      "learning_rate": 7.957371225577265e-05,
+      "loss": 5.2857,
+      "step": 226
+    },
+    {
+      "epoch": 0.03632,
+      "grad_norm": 8.774202752474505,
+      "learning_rate": 7.992895204262878e-05,
+      "loss": 5.3354,
+      "step": 227
+    },
+    {
+      "epoch": 0.03648,
+      "grad_norm": 5.437882458507581,
+      "learning_rate": 8.02841918294849e-05,
+      "loss": 5.3243,
+      "step": 228
+    },
+    {
+      "epoch": 0.03664,
+      "grad_norm": 4.5882286789765665,
+      "learning_rate": 8.063943161634103e-05,
+      "loss": 5.3034,
+      "step": 229
+    },
+    {
+      "epoch": 0.0368,
+      "grad_norm": 9.266456001851875,
+      "learning_rate": 8.099467140319716e-05,
+      "loss": 5.2621,
+      "step": 230
+    },
+    {
+      "epoch": 0.03696,
+      "grad_norm": 4.979388988031083,
+      "learning_rate": 8.134991119005328e-05,
+      "loss": 5.298,
+      "step": 231
+    },
+    {
+      "epoch": 0.03712,
+      "grad_norm": 4.745679003937358,
+      "learning_rate": 8.170515097690942e-05,
+      "loss": 5.247,
+      "step": 232
+    },
+    {
+      "epoch": 0.03728,
+      "grad_norm": 3.14862533055938,
+      "learning_rate": 8.206039076376555e-05,
+      "loss": 5.2621,
+      "step": 233
+    },
+    {
+      "epoch": 0.03744,
+      "grad_norm": 6.535045594910592,
+      "learning_rate": 8.241563055062167e-05,
+      "loss": 5.2604,
+      "step": 234
+    },
+    {
+      "epoch": 0.0376,
+      "grad_norm": 4.191911817977689,
+      "learning_rate": 8.27708703374778e-05,
+      "loss": 5.2248,
+      "step": 235
+    },
+    {
+      "epoch": 0.03776,
+      "grad_norm": 4.015212401757453,
+      "learning_rate": 8.312611012433393e-05,
+      "loss": 5.1956,
+      "step": 236
+    },
+    {
+      "epoch": 0.03792,
+      "grad_norm": 4.704913374580526,
+      "learning_rate": 8.348134991119005e-05,
+      "loss": 5.1755,
+      "step": 237
+    },
+    {
+      "epoch": 0.03808,
+      "grad_norm": 3.468398347778478,
+      "learning_rate": 8.383658969804619e-05,
+      "loss": 5.1844,
+      "step": 238
+    },
+    {
+      "epoch": 0.03824,
+      "grad_norm": 2.9600650587493007,
+      "learning_rate": 8.419182948490232e-05,
+      "loss": 5.2224,
+      "step": 239
+    },
+    {
+      "epoch": 0.0384,
+      "grad_norm": 6.572656023509431,
+      "learning_rate": 8.454706927175844e-05,
+      "loss": 5.2337,
+      "step": 240
+    },
+    {
+      "epoch": 0.03856,
+      "grad_norm": 8.635755507655073,
+      "learning_rate": 8.490230905861456e-05,
+      "loss": 5.1937,
+      "step": 241
+    },
+    {
+      "epoch": 0.03872,
+      "grad_norm": 2.737276573133315,
+      "learning_rate": 8.525754884547069e-05,
+      "loss": 5.1918,
+      "step": 242
+    },
+    {
+      "epoch": 0.03888,
+      "grad_norm": 7.71700552452157,
+      "learning_rate": 8.561278863232682e-05,
+      "loss": 5.2216,
+      "step": 243
+    },
+    {
+      "epoch": 0.03904,
+      "grad_norm": 5.04726659516664,
+      "learning_rate": 8.596802841918296e-05,
+      "loss": 5.2059,
+      "step": 244
+    },
+    {
+      "epoch": 0.0392,
+      "grad_norm": 7.001512792140965,
+      "learning_rate": 8.632326820603908e-05,
+      "loss": 5.2245,
+      "step": 245
+    },
+    {
+      "epoch": 0.03936,
+      "grad_norm": 10.284897463524112,
+      "learning_rate": 8.667850799289521e-05,
+      "loss": 5.1916,
+      "step": 246
+    },
+    {
+      "epoch": 0.03952,
+      "grad_norm": 3.5162279494131026,
+      "learning_rate": 8.703374777975135e-05,
+      "loss": 5.1753,
+      "step": 247
+    },
+    {
+      "epoch": 0.03968,
+      "grad_norm": 3.424967057716724,
+      "learning_rate": 8.738898756660745e-05,
+      "loss": 5.1498,
+      "step": 248
+    },
+    {
+      "epoch": 0.03984,
+      "grad_norm": 3.0101105224457827,
+      "learning_rate": 8.774422735346359e-05,
+      "loss": 5.1999,
+      "step": 249
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 2.523743867291498,
+      "learning_rate": 8.809946714031972e-05,
+      "loss": 5.1613,
+      "step": 250
+    },
+    {
+      "epoch": 0.04016,
+      "grad_norm": 5.399497642377313,
+      "learning_rate": 8.845470692717585e-05,
+      "loss": 5.149,
+      "step": 251
+    },
+    {
+      "epoch": 0.04032,
+      "grad_norm": 5.750067528264651,
+      "learning_rate": 8.880994671403198e-05,
+      "loss": 5.0953,
+      "step": 252
+    },
+    {
+      "epoch": 0.04048,
+      "grad_norm": 4.721938312355053,
+      "learning_rate": 8.91651865008881e-05,
+      "loss": 5.0881,
+      "step": 253
+    },
+    {
+      "epoch": 0.04064,
+      "grad_norm": 13.345103486255177,
+      "learning_rate": 8.952042628774424e-05,
+      "loss": 5.192,
+      "step": 254
+    },
+    {
+      "epoch": 0.0408,
+      "grad_norm": 8.708687737615124,
+      "learning_rate": 8.987566607460036e-05,
+      "loss": 5.1956,
+      "step": 255
+    },
+    {
+      "epoch": 0.04096,
+      "grad_norm": 6.828288848747506,
+      "learning_rate": 9.023090586145648e-05,
+      "loss": 5.1435,
+      "step": 256
+    },
+    {
+      "epoch": 0.04112,
+      "grad_norm": 12.70208575204729,
+      "learning_rate": 9.058614564831261e-05,
+      "loss": 5.1697,
+      "step": 257
+    },
+    {
+      "epoch": 0.04128,
+      "grad_norm": 2.8423151093358543,
+      "learning_rate": 9.094138543516875e-05,
+      "loss": 5.1072,
+      "step": 258
+    },
+    {
+      "epoch": 0.04144,
+      "grad_norm": 7.3224010752626985,
+      "learning_rate": 9.129662522202487e-05,
+      "loss": 5.1045,
+      "step": 259
+    },
+    {
+      "epoch": 0.0416,
+      "grad_norm": 3.3694419108657625,
+      "learning_rate": 9.1651865008881e-05,
+      "loss": 5.0538,
+      "step": 260
+    },
+    {
+      "epoch": 0.04176,
+      "grad_norm": 8.590097133061764,
+      "learning_rate": 9.200710479573713e-05,
+      "loss": 5.1715,
+      "step": 261
+    },
+    {
+      "epoch": 0.04192,
+      "grad_norm": 7.437351333135087,
+      "learning_rate": 9.236234458259325e-05,
+      "loss": 5.1513,
+      "step": 262
+    },
+    {
+      "epoch": 0.04208,
+      "grad_norm": 8.118083822679122,
+      "learning_rate": 9.271758436944938e-05,
+      "loss": 5.1157,
+      "step": 263
+    },
+    {
+      "epoch": 0.04224,
+      "grad_norm": 8.186766664408657,
+      "learning_rate": 9.307282415630552e-05,
+      "loss": 5.103,
+      "step": 264
+    },
+    {
+      "epoch": 0.0424,
+      "grad_norm": 3.139516105762643,
+      "learning_rate": 9.342806394316164e-05,
+      "loss": 5.1216,
+      "step": 265
+    },
+    {
+      "epoch": 0.04256,
+      "grad_norm": 6.301245019296755,
+      "learning_rate": 9.378330373001777e-05,
+      "loss": 5.0779,
+      "step": 266
+    },
+    {
+      "epoch": 0.04272,
+      "grad_norm": 2.7435358255243187,
+      "learning_rate": 9.41385435168739e-05,
+      "loss": 5.0417,
+      "step": 267
+    },
+    {
+      "epoch": 0.04288,
+      "grad_norm": 4.6658868097254,
+      "learning_rate": 9.449378330373002e-05,
+      "loss": 4.9884,
+      "step": 268
+    },
+    {
+      "epoch": 0.04304,
+      "grad_norm": 4.988041529808235,
+      "learning_rate": 9.484902309058615e-05,
+      "loss": 5.0996,
+      "step": 269
+    },
+    {
+      "epoch": 0.0432,
+      "grad_norm": 3.833707653714462,
+      "learning_rate": 9.520426287744227e-05,
+      "loss": 4.9505,
+      "step": 270
+    },
+    {
+      "epoch": 0.04336,
+      "grad_norm": 4.371019898954924,
+      "learning_rate": 9.555950266429841e-05,
+      "loss": 5.0566,
+      "step": 271
+    },
+    {
+      "epoch": 0.04352,
+      "grad_norm": 2.8234281449604626,
+      "learning_rate": 9.591474245115454e-05,
+      "loss": 5.1073,
+      "step": 272
+    },
+    {
+      "epoch": 0.04368,
+      "grad_norm": 2.438098581847483,
+      "learning_rate": 9.626998223801066e-05,
+      "loss": 5.006,
+      "step": 273
+    },
+    {
+      "epoch": 0.04384,
+      "grad_norm": 2.719809535009068,
+      "learning_rate": 9.662522202486678e-05,
+      "loss": 5.1045,
+      "step": 274
+    },
+    {
+      "epoch": 0.044,
+      "grad_norm": 3.631097757346069,
+      "learning_rate": 9.698046181172292e-05,
+      "loss": 5.0548,
+      "step": 275
+    },
+    {
+      "epoch": 0.04416,
+      "grad_norm": 3.3686694458420785,
+      "learning_rate": 9.733570159857904e-05,
+      "loss": 5.0796,
+      "step": 276
+    },
+    {
+      "epoch": 0.04432,
+      "grad_norm": 3.3190936347932487,
+      "learning_rate": 9.769094138543518e-05,
+      "loss": 4.9818,
+      "step": 277
+    },
+    {
+      "epoch": 0.04448,
+      "grad_norm": 3.529290019156268,
+      "learning_rate": 9.80461811722913e-05,
+      "loss": 4.918,
+      "step": 278
+    },
+    {
+      "epoch": 0.04464,
+      "grad_norm": 2.483918994859683,
+      "learning_rate": 9.840142095914743e-05,
+      "loss": 5.0849,
+      "step": 279
+    },
+    {
+      "epoch": 0.0448,
+      "grad_norm": 6.670295312866141,
+      "learning_rate": 9.875666074600355e-05,
+      "loss": 5.0075,
+      "step": 280
+    },
+    {
+      "epoch": 0.04496,
+      "grad_norm": 10.150913556189485,
+      "learning_rate": 9.911190053285967e-05,
+      "loss": 5.0234,
+      "step": 281
+    },
+    {
+      "epoch": 0.04512,
+      "grad_norm": 9.062924443471173,
+      "learning_rate": 9.946714031971581e-05,
+      "loss": 5.0173,
+      "step": 282
+    },
+    {
+      "epoch": 0.04528,
+      "grad_norm": 4.119049674373488,
+      "learning_rate": 9.982238010657194e-05,
+      "loss": 5.0161,
+      "step": 283
+    },
+    {
+      "epoch": 0.04544,
+      "grad_norm": 5.366591304526708,
+      "learning_rate": 0.00010017761989342807,
+      "loss": 5.0317,
+      "step": 284
+    },
+    {
+      "epoch": 0.0456,
+      "grad_norm": 7.237592605069814,
+      "learning_rate": 0.00010053285968028419,
+      "loss": 5.0674,
+      "step": 285
+    },
+    {
+      "epoch": 0.04576,
+      "grad_norm": 5.043156516232848,
+      "learning_rate": 0.00010088809946714034,
+      "loss": 5.0023,
+      "step": 286
+    },
+    {
+      "epoch": 0.04592,
+      "grad_norm": 5.431166071138367,
+      "learning_rate": 0.00010124333925399646,
+      "loss": 5.0058,
+      "step": 287
+    },
+    {
+      "epoch": 0.04608,
+      "grad_norm": 4.5973707892275995,
+      "learning_rate": 0.00010159857904085256,
+      "loss": 5.015,
+      "step": 288
+    },
+    {
+      "epoch": 0.04624,
+      "grad_norm": 2.8050652755272476,
+      "learning_rate": 0.00010195381882770871,
+      "loss": 4.921,
+      "step": 289
+    },
+    {
+      "epoch": 0.0464,
+      "grad_norm": 3.2516809600199865,
+      "learning_rate": 0.00010230905861456483,
+      "loss": 4.9738,
+      "step": 290
+    },
+    {
+      "epoch": 0.04656,
+      "grad_norm": 3.298715427454396,
+      "learning_rate": 0.00010266429840142096,
+      "loss": 4.9471,
+      "step": 291
+    },
+    {
+      "epoch": 0.04672,
+      "grad_norm": 4.441942164232882,
+      "learning_rate": 0.0001030195381882771,
+      "loss": 4.9639,
+      "step": 292
+    },
+    {
+      "epoch": 0.04688,
+      "grad_norm": 5.195648891898555,
+      "learning_rate": 0.00010337477797513322,
+      "loss": 4.9707,
+      "step": 293
+    },
+    {
+      "epoch": 0.04704,
+      "grad_norm": 2.67195107832033,
+      "learning_rate": 0.00010373001776198935,
+      "loss": 5.0072,
+      "step": 294
+    },
+    {
+      "epoch": 0.0472,
+      "grad_norm": 6.009004295483241,
+      "learning_rate": 0.00010408525754884548,
+      "loss": 5.0399,
+      "step": 295
+    },
+    {
+      "epoch": 0.04736,
+      "grad_norm": 2.6630535423971016,
+      "learning_rate": 0.0001044404973357016,
+      "loss": 4.9041,
+      "step": 296
+    },
+    {
+      "epoch": 0.04752,
+      "grad_norm": 4.492787688007631,
+      "learning_rate": 0.00010479573712255772,
+      "loss": 4.9033,
+      "step": 297
+    },
+    {
+      "epoch": 0.04768,
+      "grad_norm": 3.844826341288209,
+      "learning_rate": 0.00010515097690941387,
+      "loss": 4.953,
+      "step": 298
+    },
+    {
+      "epoch": 0.04784,
+      "grad_norm": 3.3183676027462035,
+      "learning_rate": 0.00010550621669627,
+      "loss": 4.9762,
+      "step": 299
+    },
+    {
+      "epoch": 0.048,
+      "grad_norm": 4.893145341172718,
+      "learning_rate": 0.00010586145648312611,
+      "loss": 4.9496,
+      "step": 300
+    },
+    {
+      "epoch": 0.04816,
+      "grad_norm": 5.7474151063067165,
+      "learning_rate": 0.00010621669626998225,
+      "loss": 4.9307,
+      "step": 301
+    },
+    {
+      "epoch": 0.04832,
+      "grad_norm": 2.062019031999705,
+      "learning_rate": 0.00010657193605683837,
+      "loss": 4.8953,
+      "step": 302
+    },
+    {
+      "epoch": 0.04848,
+      "grad_norm": 4.474000763786756,
+      "learning_rate": 0.00010692717584369449,
+      "loss": 4.9406,
+      "step": 303
+    },
+    {
+      "epoch": 0.04864,
+      "grad_norm": 4.820989458152678,
+      "learning_rate": 0.00010728241563055064,
+      "loss": 4.9617,
+      "step": 304
+    },
+    {
+      "epoch": 0.0488,
+      "grad_norm": 2.8133270336753,
+      "learning_rate": 0.00010763765541740676,
+      "loss": 4.9787,
+      "step": 305
+    },
+    {
+      "epoch": 0.04896,
+      "grad_norm": 6.316591490262785,
+      "learning_rate": 0.00010799289520426288,
+      "loss": 5.0101,
+      "step": 306
+    },
+    {
+      "epoch": 0.04912,
+      "grad_norm": 5.904923861120589,
+      "learning_rate": 0.00010834813499111902,
+      "loss": 4.9195,
+      "step": 307
+    },
+    {
+      "epoch": 0.04928,
+      "grad_norm": 2.4014733456071036,
+      "learning_rate": 0.00010870337477797514,
+      "loss": 4.9972,
+      "step": 308
+    },
+    {
+      "epoch": 0.04944,
+      "grad_norm": 5.775330293577001,
+      "learning_rate": 0.00010905861456483126,
+      "loss": 4.9402,
+      "step": 309
+    },
+    {
+      "epoch": 0.0496,
+      "grad_norm": 5.087818876701451,
+      "learning_rate": 0.00010941385435168738,
+      "loss": 4.8647,
+      "step": 310
+    },
+    {
+      "epoch": 0.04976,
+      "grad_norm": 3.14142858194589,
+      "learning_rate": 0.00010976909413854353,
+      "loss": 4.9464,
+      "step": 311
+    },
+    {
+      "epoch": 0.04992,
+      "grad_norm": 6.016557242021748,
+      "learning_rate": 0.00011012433392539965,
+      "loss": 4.9883,
+      "step": 312
+    },
+    {
+      "epoch": 0.05008,
+      "grad_norm": 2.188123218546429,
+      "learning_rate": 0.00011047957371225577,
+      "loss": 4.9447,
+      "step": 313
+    },
+    {
+      "epoch": 0.05024,
+      "grad_norm": 6.160595627752611,
+      "learning_rate": 0.00011083481349911191,
+      "loss": 4.9482,
+      "step": 314
+    },
+    {
+      "epoch": 0.0504,
+      "grad_norm": 8.282119656617137,
+      "learning_rate": 0.00011119005328596803,
+      "loss": 4.937,
+      "step": 315
+    },
+    {
+      "epoch": 0.05056,
+      "grad_norm": 2.600516263459627,
+      "learning_rate": 0.00011154529307282415,
+      "loss": 4.9157,
+      "step": 316
+    },
+    {
+      "epoch": 0.05072,
+      "grad_norm": 4.612252335874222,
+      "learning_rate": 0.0001119005328596803,
+      "loss": 4.9307,
+      "step": 317
+    },
+    {
+      "epoch": 0.05088,
+      "grad_norm": 3.7788492466255184,
+      "learning_rate": 0.00011225577264653642,
+      "loss": 4.9059,
+      "step": 318
+    },
+    {
+      "epoch": 0.05104,
+      "grad_norm": 4.001957309075157,
+      "learning_rate": 0.00011261101243339254,
+      "loss": 4.8999,
+      "step": 319
+    },
+    {
+      "epoch": 0.0512,
+      "grad_norm": 4.6373334705029245,
+      "learning_rate": 0.00011296625222024868,
+      "loss": 4.9324,
+      "step": 320
+    },
+    {
+      "epoch": 0.05136,
+      "grad_norm": 2.492248511707348,
+      "learning_rate": 0.0001133214920071048,
+      "loss": 4.9702,
+      "step": 321
+    },
+    {
+      "epoch": 0.05152,
+      "grad_norm": 5.301515819189879,
+      "learning_rate": 0.00011367673179396092,
+      "loss": 4.905,
+      "step": 322
+    },
+    {
+      "epoch": 0.05168,
+      "grad_norm": 3.317466563180237,
+      "learning_rate": 0.00011403197158081707,
+      "loss": 4.8475,
+      "step": 323
+    },
+    {
+      "epoch": 0.05184,
+      "grad_norm": 3.1092105943432697,
+      "learning_rate": 0.00011438721136767319,
+      "loss": 4.9048,
+      "step": 324
+    },
+    {
+      "epoch": 0.052,
+      "grad_norm": 4.006009700121147,
+      "learning_rate": 0.00011474245115452931,
+      "loss": 4.8614,
+      "step": 325
+    },
+    {
+      "epoch": 0.05216,
+      "grad_norm": 3.12265679113318,
+      "learning_rate": 0.00011509769094138544,
+      "loss": 4.8347,
+      "step": 326
+    },
+    {
+      "epoch": 0.05232,
+      "grad_norm": 4.266892265717954,
+      "learning_rate": 0.00011545293072824157,
+      "loss": 4.9129,
+      "step": 327
+    },
+    {
+      "epoch": 0.05248,
+      "grad_norm": 2.621791347034737,
+      "learning_rate": 0.00011580817051509769,
+      "loss": 4.8292,
+      "step": 328
+    },
+    {
+      "epoch": 0.05264,
+      "grad_norm": 5.388833653965526,
+      "learning_rate": 0.00011616341030195384,
+      "loss": 4.8514,
+      "step": 329
+    },
+    {
+      "epoch": 0.0528,
+      "grad_norm": 4.158894898428984,
+      "learning_rate": 0.00011651865008880996,
+      "loss": 4.9009,
+      "step": 330
+    },
+    {
+      "epoch": 0.05296,
+      "grad_norm": 4.095697143724956,
+      "learning_rate": 0.00011687388987566608,
+      "loss": 4.8578,
+      "step": 331
+    },
+    {
+      "epoch": 0.05312,
+      "grad_norm": 8.843972935252964,
+      "learning_rate": 0.00011722912966252221,
+      "loss": 4.8892,
+      "step": 332
+    },
+    {
+      "epoch": 0.05328,
+      "grad_norm": 4.718914085773682,
+      "learning_rate": 0.00011758436944937833,
+      "loss": 4.8562,
+      "step": 333
+    },
+    {
+      "epoch": 0.05344,
+      "grad_norm": 3.7520601850810635,
+      "learning_rate": 0.00011793960923623446,
+      "loss": 4.8665,
+      "step": 334
+    },
+    {
+      "epoch": 0.0536,
+      "grad_norm": 3.6918557733489923,
+      "learning_rate": 0.00011829484902309058,
+      "loss": 4.8812,
+      "step": 335
+    },
+    {
+      "epoch": 0.05376,
+      "grad_norm": 3.1302796767965755,
+      "learning_rate": 0.00011865008880994673,
+      "loss": 4.8854,
+      "step": 336
+    },
+    {
+      "epoch": 0.05392,
+      "grad_norm": 7.31457629055535,
+      "learning_rate": 0.00011900532859680285,
+      "loss": 4.9232,
+      "step": 337
+    },
+    {
+      "epoch": 0.05408,
+      "grad_norm": 7.407687714994547,
+      "learning_rate": 0.00011936056838365897,
+      "loss": 4.8909,
+      "step": 338
+    },
+    {
+      "epoch": 0.05424,
+      "grad_norm": 2.7725102246864344,
+      "learning_rate": 0.0001197158081705151,
+      "loss": 4.7873,
+      "step": 339
+    },
+    {
+      "epoch": 0.0544,
+      "grad_norm": 4.483137706037656,
+      "learning_rate": 0.00012007104795737122,
+      "loss": 4.8493,
+      "step": 340
+    },
+    {
+      "epoch": 0.05456,
+      "grad_norm": 5.545887549833481,
+      "learning_rate": 0.00012042628774422735,
+      "loss": 4.7787,
+      "step": 341
+    },
+    {
+      "epoch": 0.05472,
+      "grad_norm": 3.0813649511961154,
+      "learning_rate": 0.0001207815275310835,
+      "loss": 4.7862,
+      "step": 342
+    },
+    {
+      "epoch": 0.05488,
+      "grad_norm": 3.958365249929929,
+      "learning_rate": 0.00012113676731793962,
+      "loss": 4.8363,
+      "step": 343
+    },
+    {
+      "epoch": 0.05504,
+      "grad_norm": 2.811686215960801,
+      "learning_rate": 0.00012149200710479574,
+      "loss": 4.8897,
+      "step": 344
+    },
+    {
+      "epoch": 0.0552,
+      "grad_norm": 2.4306244841847655,
+      "learning_rate": 0.00012184724689165187,
+      "loss": 4.8666,
+      "step": 345
+    },
+    {
+      "epoch": 0.05536,
+      "grad_norm": 2.513555680400047,
+      "learning_rate": 0.000122202486678508,
+      "loss": 4.8785,
+      "step": 346
+    },
+    {
+      "epoch": 0.05552,
+      "grad_norm": 3.5168525866044225,
+      "learning_rate": 0.00012255772646536413,
+      "loss": 4.8472,
+      "step": 347
+    },
+    {
+      "epoch": 0.05568,
+      "grad_norm": 2.660247672315408,
+      "learning_rate": 0.00012291296625222028,
+      "loss": 4.8429,
+      "step": 348
+    },
+    {
+      "epoch": 0.05584,
+      "grad_norm": 4.504632159687314,
+      "learning_rate": 0.00012326820603907637,
+      "loss": 4.8497,
+      "step": 349
+    },
+    {
+      "epoch": 0.056,
+      "grad_norm": 5.150559845525184,
+      "learning_rate": 0.0001236234458259325,
+      "loss": 4.8423,
+      "step": 350
+    },
+    {
+      "epoch": 0.05616,
+      "grad_norm": 4.995869856422735,
+      "learning_rate": 0.00012397868561278864,
+      "loss": 4.7721,
+      "step": 351
+    },
+    {
+      "epoch": 0.05632,
+      "grad_norm": 2.993313571912598,
+      "learning_rate": 0.00012433392539964476,
+      "loss": 4.808,
+      "step": 352
+    },
+    {
+      "epoch": 0.05648,
+      "grad_norm": 3.9498750909912057,
+      "learning_rate": 0.00012468916518650088,
+      "loss": 4.7778,
+      "step": 353
+    },
+    {
+      "epoch": 0.05664,
+      "grad_norm": 5.029692879798669,
+      "learning_rate": 0.00012504440497335703,
+      "loss": 4.8723,
+      "step": 354
+    },
+    {
+      "epoch": 0.0568,
+      "grad_norm": 3.5219905562569345,
+      "learning_rate": 0.00012539964476021315,
+      "loss": 4.7882,
+      "step": 355
+    },
+    {
+      "epoch": 0.05696,
+      "grad_norm": 7.835386031257312,
+      "learning_rate": 0.00012575488454706927,
+      "loss": 4.9396,
+      "step": 356
+    },
+    {
+      "epoch": 0.05712,
+      "grad_norm": 4.4613782506600455,
+      "learning_rate": 0.0001261101243339254,
+      "loss": 4.8177,
+      "step": 357
+    },
+    {
+      "epoch": 0.05728,
+      "grad_norm": 5.014475762366743,
+      "learning_rate": 0.00012646536412078154,
+      "loss": 4.8339,
+      "step": 358
+    },
+    {
+      "epoch": 0.05744,
+      "grad_norm": 5.574243927914136,
+      "learning_rate": 0.00012682060390763766,
+      "loss": 4.769,
+      "step": 359
+    },
+    {
+      "epoch": 0.0576,
+      "grad_norm": 3.641892726311011,
+      "learning_rate": 0.00012717584369449379,
+      "loss": 4.8344,
+      "step": 360
+    },
+    {
+      "epoch": 0.05776,
+      "grad_norm": 6.117693245248029,
+      "learning_rate": 0.00012753108348134993,
+      "loss": 4.8409,
+      "step": 361
+    },
+    {
+      "epoch": 0.05792,
+      "grad_norm": 6.419626809378983,
+      "learning_rate": 0.00012788632326820606,
+      "loss": 4.8264,
+      "step": 362
+    },
+    {
+      "epoch": 0.05808,
+      "grad_norm": 3.165668802757641,
+      "learning_rate": 0.00012824156305506218,
+      "loss": 4.8229,
+      "step": 363
+    },
+    {
+      "epoch": 0.05824,
+      "grad_norm": 2.777814428095496,
+      "learning_rate": 0.0001285968028419183,
+      "loss": 4.8018,
+      "step": 364
+    },
+    {
+      "epoch": 0.0584,
+      "grad_norm": 3.093350252180669,
+      "learning_rate": 0.00012895204262877442,
+      "loss": 4.8509,
+      "step": 365
+    },
+    {
+      "epoch": 0.05856,
+      "grad_norm": 4.32317775441231,
+      "learning_rate": 0.00012930728241563054,
+      "loss": 4.7364,
+      "step": 366
+    },
+    {
+      "epoch": 0.05872,
+      "grad_norm": 4.931154815143631,
+      "learning_rate": 0.0001296625222024867,
+      "loss": 4.8854,
+      "step": 367
+    },
+    {
+      "epoch": 0.05888,
+      "grad_norm": 8.271150716359521,
+      "learning_rate": 0.0001300177619893428,
+      "loss": 4.7381,
+      "step": 368
+    },
+    {
+      "epoch": 0.05904,
+      "grad_norm": 7.55850224981692,
+      "learning_rate": 0.00013037300177619893,
+      "loss": 4.8478,
+      "step": 369
+    },
+    {
+      "epoch": 0.0592,
+      "grad_norm": 2.798267532636435,
+      "learning_rate": 0.00013072824156305508,
+      "loss": 4.8142,
+      "step": 370
+    },
+    {
+      "epoch": 0.05936,
+      "grad_norm": 4.368952814161644,
+      "learning_rate": 0.0001310834813499112,
+      "loss": 4.8785,
+      "step": 371
+    },
+    {
+      "epoch": 0.05952,
+      "grad_norm": 5.412427828408207,
+      "learning_rate": 0.00013143872113676732,
+      "loss": 4.8117,
+      "step": 372
+    },
+    {
+      "epoch": 0.05968,
+      "grad_norm": 4.477292013476806,
+      "learning_rate": 0.00013179396092362347,
+      "loss": 4.9115,
+      "step": 373
+    },
+    {
+      "epoch": 0.05984,
+      "grad_norm": 5.204110412031779,
+      "learning_rate": 0.0001321492007104796,
+      "loss": 4.7803,
+      "step": 374
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 3.622262010877985,
+      "learning_rate": 0.00013250444049733571,
+      "loss": 4.8324,
+      "step": 375
+    },
+    {
+      "epoch": 0.06016,
+      "grad_norm": 6.140090265420642,
+      "learning_rate": 0.00013285968028419184,
+      "loss": 4.8213,
+      "step": 376
+    },
+    {
+      "epoch": 0.06032,
+      "grad_norm": 9.974937098945354,
+      "learning_rate": 0.00013321492007104796,
+      "loss": 4.8348,
+      "step": 377
+    },
+    {
+      "epoch": 0.06048,
+      "grad_norm": 5.652906151943218,
+      "learning_rate": 0.00013357015985790408,
+      "loss": 4.8295,
+      "step": 378
+    },
+    {
+      "epoch": 0.06064,
+      "grad_norm": 3.2882606556911553,
+      "learning_rate": 0.00013392539964476023,
+      "loss": 4.7676,
+      "step": 379
+    },
+    {
+      "epoch": 0.0608,
+      "grad_norm": 3.5642280280877445,
+      "learning_rate": 0.00013428063943161635,
+      "loss": 4.7447,
+      "step": 380
+    },
+    {
+      "epoch": 0.06096,
+      "grad_norm": 3.230916422117009,
+      "learning_rate": 0.00013463587921847247,
+      "loss": 4.7961,
+      "step": 381
+    },
+    {
+      "epoch": 0.06112,
+      "grad_norm": 3.324399848632162,
+      "learning_rate": 0.0001349911190053286,
+      "loss": 4.7807,
+      "step": 382
+    },
+    {
+      "epoch": 0.06128,
+      "grad_norm": 3.8316229823685655,
+      "learning_rate": 0.00013534635879218474,
+      "loss": 4.795,
+      "step": 383
+    },
+    {
+      "epoch": 0.06144,
+      "grad_norm": 5.459020574028291,
+      "learning_rate": 0.00013570159857904086,
+      "loss": 4.8039,
+      "step": 384
+    },
+    {
+      "epoch": 0.0616,
+      "grad_norm": 3.4844580358983945,
+      "learning_rate": 0.00013605683836589698,
+      "loss": 4.7849,
+      "step": 385
+    },
+    {
+      "epoch": 0.06176,
+      "grad_norm": 2.3402083526593067,
+      "learning_rate": 0.00013641207815275313,
+      "loss": 4.8287,
+      "step": 386
+    },
+    {
+      "epoch": 0.06192,
+      "grad_norm": 3.769321144164565,
+      "learning_rate": 0.00013676731793960925,
+      "loss": 4.7109,
+      "step": 387
+    },
+    {
+      "epoch": 0.06208,
+      "grad_norm": 3.9207160285547524,
+      "learning_rate": 0.00013712255772646537,
+      "loss": 4.7638,
+      "step": 388
+    },
+    {
+      "epoch": 0.06224,
+      "grad_norm": 2.5912074943035686,
+      "learning_rate": 0.0001374777975133215,
+      "loss": 4.6955,
+      "step": 389
+    },
+    {
+      "epoch": 0.0624,
+      "grad_norm": 2.6298625924705887,
+      "learning_rate": 0.00013783303730017762,
+      "loss": 4.7028,
+      "step": 390
+    },
+    {
+      "epoch": 0.06256,
+      "grad_norm": 3.021446908220404,
+      "learning_rate": 0.00013818827708703374,
+      "loss": 4.7826,
+      "step": 391
+    },
+    {
+      "epoch": 0.06272,
+      "grad_norm": 3.285717782584293,
+      "learning_rate": 0.00013854351687388988,
+      "loss": 4.7869,
+      "step": 392
+    },
+    {
+      "epoch": 0.06288,
+      "grad_norm": 7.260047489428337,
+      "learning_rate": 0.000138898756660746,
+      "loss": 4.8063,
+      "step": 393
+    },
+    {
+      "epoch": 0.06304,
+      "grad_norm": 8.170819447079618,
+      "learning_rate": 0.00013925399644760213,
+      "loss": 4.8276,
+      "step": 394
+    },
+    {
+      "epoch": 0.0632,
+      "grad_norm": 8.706291509799422,
+      "learning_rate": 0.00013960923623445828,
+      "loss": 4.7622,
+      "step": 395
+    },
+    {
+      "epoch": 0.06336,
+      "grad_norm": 2.664654045503026,
+      "learning_rate": 0.0001399644760213144,
+      "loss": 4.7903,
+      "step": 396
+    },
+    {
+      "epoch": 0.06352,
+      "grad_norm": 7.158947718562827,
+      "learning_rate": 0.00014031971580817052,
+      "loss": 4.823,
+      "step": 397
+    },
+    {
+      "epoch": 0.06368,
+      "grad_norm": 7.025458253890768,
+      "learning_rate": 0.00014067495559502667,
+      "loss": 4.7419,
+      "step": 398
+    },
+    {
+      "epoch": 0.06384,
+      "grad_norm": 3.0596095362522404,
+      "learning_rate": 0.0001410301953818828,
+      "loss": 4.7643,
+      "step": 399
+    },
+    {
+      "epoch": 0.064,
+      "grad_norm": 4.216896711984632,
+      "learning_rate": 0.0001413854351687389,
+      "loss": 4.7122,
+      "step": 400
+    },
+    {
+      "epoch": 0.06416,
+      "grad_norm": 5.0037890123857345,
+      "learning_rate": 0.00014174067495559503,
+      "loss": 4.7382,
+      "step": 401
+    },
+    {
+      "epoch": 0.06432,
+      "grad_norm": 4.891154850210662,
+      "learning_rate": 0.00014209591474245115,
+      "loss": 4.8159,
+      "step": 402
+    },
+    {
+      "epoch": 0.06448,
+      "grad_norm": 5.055285156092573,
+      "learning_rate": 0.00014245115452930727,
+      "loss": 4.7707,
+      "step": 403
+    },
+    {
+      "epoch": 0.06464,
+      "grad_norm": 2.9716962797962823,
+      "learning_rate": 0.00014280639431616342,
+      "loss": 4.7302,
+      "step": 404
+    },
+    {
+      "epoch": 0.0648,
+      "grad_norm": 4.306437265369655,
+      "learning_rate": 0.00014316163410301954,
+      "loss": 4.7357,
+      "step": 405
+    },
+    {
+      "epoch": 0.06496,
+      "grad_norm": 2.2292572038425056,
+      "learning_rate": 0.00014351687388987566,
+      "loss": 4.7459,
+      "step": 406
+    },
+    {
+      "epoch": 0.06512,
+      "grad_norm": 3.342753785618463,
+      "learning_rate": 0.00014387211367673179,
+      "loss": 4.772,
+      "step": 407
+    },
+    {
+      "epoch": 0.06528,
+      "grad_norm": 2.2515079966015965,
+      "learning_rate": 0.00014422735346358793,
+      "loss": 4.7436,
+      "step": 408
+    },
+    {
+      "epoch": 0.06544,
+      "grad_norm": 2.486502436700024,
+      "learning_rate": 0.00014458259325044406,
+      "loss": 4.7902,
+      "step": 409
+    },
+    {
+      "epoch": 0.0656,
+      "grad_norm": 2.830314433914487,
+      "learning_rate": 0.00014493783303730018,
+      "loss": 4.8081,
+      "step": 410
+    },
+    {
+      "epoch": 0.06576,
+      "grad_norm": 3.9123053895010726,
+      "learning_rate": 0.00014529307282415633,
+      "loss": 4.7363,
+      "step": 411
+    },
+    {
+      "epoch": 0.06592,
+      "grad_norm": 4.605205131788471,
+      "learning_rate": 0.00014564831261101245,
+      "loss": 4.7034,
+      "step": 412
+    },
+    {
+      "epoch": 0.06608,
+      "grad_norm": 4.052076851865704,
+      "learning_rate": 0.00014600355239786857,
+      "loss": 4.6997,
+      "step": 413
+    },
+    {
+      "epoch": 0.06624,
+      "grad_norm": 3.501606915732486,
+      "learning_rate": 0.0001463587921847247,
+      "loss": 4.7316,
+      "step": 414
+    },
+    {
+      "epoch": 0.0664,
+      "grad_norm": 2.7287890623691817,
+      "learning_rate": 0.0001467140319715808,
+      "loss": 4.7863,
+      "step": 415
+    },
+    {
+      "epoch": 0.06656,
+      "grad_norm": 2.837853487371347,
+      "learning_rate": 0.00014706927175843693,
+      "loss": 4.7913,
+      "step": 416
+    },
+    {
+      "epoch": 0.06672,
+      "grad_norm": 2.603892503590653,
+      "learning_rate": 0.00014742451154529308,
+      "loss": 4.7438,
+      "step": 417
+    },
+    {
+      "epoch": 0.06688,
+      "grad_norm": 2.6814581811043303,
+      "learning_rate": 0.0001477797513321492,
+      "loss": 4.7468,
+      "step": 418
+    },
+    {
+      "epoch": 0.06704,
+      "grad_norm": 2.836125413867714,
+      "learning_rate": 0.00014813499111900532,
+      "loss": 4.7933,
+      "step": 419
+    },
+    {
+      "epoch": 0.0672,
+      "grad_norm": 5.576626961671651,
+      "learning_rate": 0.00014849023090586147,
+      "loss": 4.8204,
+      "step": 420
+    },
+    {
+      "epoch": 0.06736,
+      "grad_norm": 6.6734648824209035,
+      "learning_rate": 0.0001488454706927176,
+      "loss": 4.779,
+      "step": 421
+    },
+    {
+      "epoch": 0.06752,
+      "grad_norm": 5.89711333653718,
+      "learning_rate": 0.00014920071047957371,
+      "loss": 4.6719,
+      "step": 422
+    },
+    {
+      "epoch": 0.06768,
+      "grad_norm": 2.4956989408437216,
+      "learning_rate": 0.00014955595026642986,
+      "loss": 4.6914,
+      "step": 423
+    },
+    {
+      "epoch": 0.06784,
+      "grad_norm": 8.744762418889547,
+      "learning_rate": 0.00014991119005328598,
+      "loss": 4.8054,
+      "step": 424
+    },
+    {
+      "epoch": 0.068,
+      "grad_norm": 7.105384221263611,
+      "learning_rate": 0.0001502664298401421,
+      "loss": 4.8233,
+      "step": 425
+    },
+    {
+      "epoch": 0.06816,
+      "grad_norm": 2.501842860224022,
+      "learning_rate": 0.00015062166962699825,
+      "loss": 4.7071,
+      "step": 426
+    },
+    {
+      "epoch": 0.06832,
+      "grad_norm": 5.715946324434159,
+      "learning_rate": 0.00015097690941385437,
+      "loss": 4.7176,
+      "step": 427
+    },
+    {
+      "epoch": 0.06848,
+      "grad_norm": 4.562079305922884,
+      "learning_rate": 0.00015133214920071047,
+      "loss": 4.7206,
+      "step": 428
+    },
+    {
+      "epoch": 0.06864,
+      "grad_norm": 2.5985436029077222,
+      "learning_rate": 0.00015168738898756662,
+      "loss": 4.7569,
+      "step": 429
+    },
+    {
+      "epoch": 0.0688,
+      "grad_norm": 4.816089699294097,
+      "learning_rate": 0.00015204262877442274,
+      "loss": 4.6996,
+      "step": 430
+    },
+    {
+      "epoch": 0.06896,
+      "grad_norm": 2.893778831560003,
+      "learning_rate": 0.00015239786856127886,
+      "loss": 4.7065,
+      "step": 431
+    },
+    {
+      "epoch": 0.06912,
+      "grad_norm": 3.317498604003229,
+      "learning_rate": 0.00015275310834813498,
+      "loss": 4.7534,
+      "step": 432
+    },
+    {
+      "epoch": 0.06928,
+      "grad_norm": 2.630778268654286,
+      "learning_rate": 0.00015310834813499113,
+      "loss": 4.6698,
+      "step": 433
+    },
+    {
+      "epoch": 0.06944,
+      "grad_norm": 6.365019777981565,
+      "learning_rate": 0.00015346358792184725,
+      "loss": 4.6814,
+      "step": 434
+    },
+    {
+      "epoch": 0.0696,
+      "grad_norm": 6.070603547949627,
+      "learning_rate": 0.00015381882770870337,
+      "loss": 4.7467,
+      "step": 435
+    },
+    {
+      "epoch": 0.06976,
+      "grad_norm": 2.694595164965054,
+      "learning_rate": 0.00015417406749555952,
+      "loss": 4.6758,
+      "step": 436
+    },
+    {
+      "epoch": 0.06992,
+      "grad_norm": 6.483504364440021,
+      "learning_rate": 0.00015452930728241564,
+      "loss": 4.7042,
+      "step": 437
+    },
+    {
+      "epoch": 0.07008,
+      "grad_norm": 3.0684076022529085,
+      "learning_rate": 0.00015488454706927176,
+      "loss": 4.7166,
+      "step": 438
+    },
+    {
+      "epoch": 0.07024,
+      "grad_norm": 5.040266819763534,
+      "learning_rate": 0.0001552397868561279,
+      "loss": 4.6682,
+      "step": 439
+    },
+    {
+      "epoch": 0.0704,
+      "grad_norm": 6.801880572827608,
+      "learning_rate": 0.00015559502664298403,
+      "loss": 4.6321,
+      "step": 440
+    },
+    {
+      "epoch": 0.07056,
+      "grad_norm": 2.723612006921323,
+      "learning_rate": 0.00015595026642984015,
+      "loss": 4.6595,
+      "step": 441
+    },
+    {
+      "epoch": 0.07072,
+      "grad_norm": 3.1114371275261643,
+      "learning_rate": 0.00015630550621669628,
+      "loss": 4.6853,
+      "step": 442
+    },
+    {
+      "epoch": 0.07088,
+      "grad_norm": 2.9710258706557977,
+      "learning_rate": 0.0001566607460035524,
+      "loss": 4.7197,
+      "step": 443
+    },
+    {
+      "epoch": 0.07104,
+      "grad_norm": 2.2650891057792437,
+      "learning_rate": 0.00015701598579040852,
+      "loss": 4.6675,
+      "step": 444
+    },
+    {
+      "epoch": 0.0712,
+      "grad_norm": 3.0162364594768083,
+      "learning_rate": 0.00015737122557726467,
+      "loss": 4.708,
+      "step": 445
+    },
+    {
+      "epoch": 0.07136,
+      "grad_norm": 2.212398269819425,
+      "learning_rate": 0.0001577264653641208,
+      "loss": 4.678,
+      "step": 446
+    },
+    {
+      "epoch": 0.07152,
+      "grad_norm": 3.0682828288573982,
+      "learning_rate": 0.0001580817051509769,
+      "loss": 4.6541,
+      "step": 447
+    },
+    {
+      "epoch": 0.07168,
+      "grad_norm": 3.9592508732468543,
+      "learning_rate": 0.00015843694493783306,
+      "loss": 4.6462,
+      "step": 448
+    },
+    {
+      "epoch": 0.07184,
+      "grad_norm": 3.0160406817838172,
+      "learning_rate": 0.00015879218472468918,
+      "loss": 4.6068,
+      "step": 449
+    },
+    {
+      "epoch": 0.072,
+      "grad_norm": 2.1389040488212157,
+      "learning_rate": 0.0001591474245115453,
+      "loss": 4.6136,
+      "step": 450
+    },
+    {
+      "epoch": 0.07216,
+      "grad_norm": 2.4497155285741687,
+      "learning_rate": 0.00015950266429840145,
+      "loss": 4.6386,
+      "step": 451
+    },
+    {
+      "epoch": 0.07232,
+      "grad_norm": 2.446450222856664,
+      "learning_rate": 0.00015985790408525757,
+      "loss": 4.6022,
+      "step": 452
+    },
+    {
+      "epoch": 0.07248,
+      "grad_norm": 3.8510747929254983,
+      "learning_rate": 0.0001602131438721137,
+      "loss": 4.6612,
+      "step": 453
+    },
+    {
+      "epoch": 0.07264,
+      "grad_norm": 4.147852931552536,
+      "learning_rate": 0.0001605683836589698,
+      "loss": 4.6379,
+      "step": 454
+    },
+    {
+      "epoch": 0.0728,
+      "grad_norm": 2.482801093647865,
+      "learning_rate": 0.00016092362344582593,
+      "loss": 4.6674,
+      "step": 455
+    },
+    {
+      "epoch": 0.07296,
+      "grad_norm": 2.981810890179259,
+      "learning_rate": 0.00016127886323268206,
+      "loss": 4.6887,
+      "step": 456
+    },
+    {
+      "epoch": 0.07312,
+      "grad_norm": 8.634161028676914,
+      "learning_rate": 0.00016163410301953818,
+      "loss": 4.7368,
+      "step": 457
+    },
+    {
+      "epoch": 0.07328,
+      "grad_norm": 6.6507487122230335,
+      "learning_rate": 0.00016198934280639432,
+      "loss": 4.7197,
+      "step": 458
+    },
+    {
+      "epoch": 0.07344,
+      "grad_norm": 2.659560532572739,
+      "learning_rate": 0.00016234458259325045,
+      "loss": 4.6888,
+      "step": 459
+    },
+    {
+      "epoch": 0.0736,
+      "grad_norm": 3.0736699373316068,
+      "learning_rate": 0.00016269982238010657,
+      "loss": 4.6635,
+      "step": 460
+    },
+    {
+      "epoch": 0.07376,
+      "grad_norm": 2.785446065202562,
+      "learning_rate": 0.00016305506216696272,
+      "loss": 4.6211,
+      "step": 461
+    },
+    {
+      "epoch": 0.07392,
+      "grad_norm": 3.3730120497969955,
+      "learning_rate": 0.00016341030195381884,
+      "loss": 4.64,
+      "step": 462
+    },
+    {
+      "epoch": 0.07408,
+      "grad_norm": 2.6348990910221586,
+      "learning_rate": 0.00016376554174067496,
+      "loss": 4.6218,
+      "step": 463
+    },
+    {
+      "epoch": 0.07424,
+      "grad_norm": 2.86918684718777,
+      "learning_rate": 0.0001641207815275311,
+      "loss": 4.6001,
+      "step": 464
+    },
+    {
+      "epoch": 0.0744,
+      "grad_norm": 2.020306363115699,
+      "learning_rate": 0.00016447602131438723,
+      "loss": 4.6292,
+      "step": 465
+    },
+    {
+      "epoch": 0.07456,
+      "grad_norm": 5.261542065768452,
+      "learning_rate": 0.00016483126110124335,
+      "loss": 4.7332,
+      "step": 466
+    },
+    {
+      "epoch": 0.07472,
+      "grad_norm": 6.888546768125345,
+      "learning_rate": 0.00016518650088809947,
+      "loss": 4.6872,
+      "step": 467
+    },
+    {
+      "epoch": 0.07488,
+      "grad_norm": 2.5502704188083754,
+      "learning_rate": 0.0001655417406749556,
+      "loss": 4.7435,
+      "step": 468
+    },
+    {
+      "epoch": 0.07504,
+      "grad_norm": 10.061895750322229,
+      "learning_rate": 0.0001658969804618117,
+      "loss": 4.6647,
+      "step": 469
+    },
+    {
+      "epoch": 0.0752,
+      "grad_norm": 4.8746768941008245,
+      "learning_rate": 0.00016625222024866786,
+      "loss": 4.5674,
+      "step": 470
+    },
+    {
+      "epoch": 0.07536,
+      "grad_norm": 5.955480525839466,
+      "learning_rate": 0.00016660746003552398,
+      "loss": 4.7474,
+      "step": 471
+    },
+    {
+      "epoch": 0.07552,
+      "grad_norm": 10.433538691676409,
+      "learning_rate": 0.0001669626998223801,
+      "loss": 4.7938,
+      "step": 472
+    },
+    {
+      "epoch": 0.07568,
+      "grad_norm": 4.4140424929435484,
+      "learning_rate": 0.00016731793960923625,
+      "loss": 4.6836,
+      "step": 473
+    },
+    {
+      "epoch": 0.07584,
+      "grad_norm": 3.471128747980634,
+      "learning_rate": 0.00016767317939609237,
+      "loss": 4.6714,
+      "step": 474
+    },
+    {
+      "epoch": 0.076,
+      "grad_norm": 5.358665786916061,
+      "learning_rate": 0.0001680284191829485,
+      "loss": 4.6947,
+      "step": 475
+    },
+    {
+      "epoch": 0.07616,
+      "grad_norm": 5.383323370784499,
+      "learning_rate": 0.00016838365896980464,
+      "loss": 4.6566,
+      "step": 476
+    },
+    {
+      "epoch": 0.07632,
+      "grad_norm": 4.156698241523254,
+      "learning_rate": 0.00016873889875666077,
+      "loss": 4.6288,
+      "step": 477
+    },
+    {
+      "epoch": 0.07648,
+      "grad_norm": 2.958059856991853,
+      "learning_rate": 0.0001690941385435169,
+      "loss": 4.6364,
+      "step": 478
+    },
+    {
+      "epoch": 0.07664,
+      "grad_norm": 4.435226803254213,
+      "learning_rate": 0.000169449378330373,
+      "loss": 4.5612,
+      "step": 479
+    },
+    {
+      "epoch": 0.0768,
+      "grad_norm": 4.99278875684811,
+      "learning_rate": 0.00016980461811722913,
+      "loss": 4.7081,
+      "step": 480
+    },
+    {
+      "epoch": 0.07696,
+      "grad_norm": 2.8555656586587315,
+      "learning_rate": 0.00017015985790408525,
+      "loss": 4.6234,
+      "step": 481
+    },
+    {
+      "epoch": 0.07712,
+      "grad_norm": 5.134922003894972,
+      "learning_rate": 0.00017051509769094137,
+      "loss": 4.7026,
+      "step": 482
+    },
+    {
+      "epoch": 0.07728,
+      "grad_norm": 2.475110933340091,
+      "learning_rate": 0.00017087033747779752,
+      "loss": 4.6788,
+      "step": 483
+    },
+    {
+      "epoch": 0.07744,
+      "grad_norm": 3.9329694103999913,
+      "learning_rate": 0.00017122557726465364,
+      "loss": 4.6822,
+      "step": 484
+    },
+    {
+      "epoch": 0.0776,
+      "grad_norm": 2.580007232005459,
+      "learning_rate": 0.00017158081705150976,
+      "loss": 4.6399,
+      "step": 485
+    },
+    {
+      "epoch": 0.07776,
+      "grad_norm": 5.620886391432912,
+      "learning_rate": 0.0001719360568383659,
+      "loss": 4.627,
+      "step": 486
+    },
+    {
+      "epoch": 0.07792,
+      "grad_norm": 2.6247713925289924,
+      "learning_rate": 0.00017229129662522203,
+      "loss": 4.6316,
+      "step": 487
+    },
+    {
+      "epoch": 0.07808,
+      "grad_norm": 6.644026668587743,
+      "learning_rate": 0.00017264653641207815,
+      "loss": 4.6576,
+      "step": 488
+    },
+    {
+      "epoch": 0.07824,
+      "grad_norm": 6.255799114432507,
+      "learning_rate": 0.0001730017761989343,
+      "loss": 4.651,
+      "step": 489
+    },
+    {
+      "epoch": 0.0784,
+      "grad_norm": 3.987921681877768,
+      "learning_rate": 0.00017335701598579042,
+      "loss": 4.6249,
+      "step": 490
+    },
+    {
+      "epoch": 0.07856,
+      "grad_norm": 8.24825882949861,
+      "learning_rate": 0.00017371225577264654,
+      "loss": 4.6679,
+      "step": 491
+    },
+    {
+      "epoch": 0.07872,
+      "grad_norm": 3.3238104349626227,
+      "learning_rate": 0.0001740674955595027,
+      "loss": 4.612,
+      "step": 492
+    },
+    {
+      "epoch": 0.07888,
+      "grad_norm": 6.00658127547705,
+      "learning_rate": 0.0001744227353463588,
+      "loss": 4.7096,
+      "step": 493
+    },
+    {
+      "epoch": 0.07904,
+      "grad_norm": 2.8238431869537983,
+      "learning_rate": 0.0001747779751332149,
+      "loss": 4.6271,
+      "step": 494
+    },
+    {
+      "epoch": 0.0792,
+      "grad_norm": 6.682953649766787,
+      "learning_rate": 0.00017513321492007106,
+      "loss": 4.7315,
+      "step": 495
+    },
+    {
+      "epoch": 0.07936,
+      "grad_norm": 4.157670102687912,
+      "learning_rate": 0.00017548845470692718,
+      "loss": 4.6214,
+      "step": 496
+    },
+    {
+      "epoch": 0.07952,
+      "grad_norm": 6.267874754973918,
+      "learning_rate": 0.0001758436944937833,
+      "loss": 4.6919,
+      "step": 497
+    },
+    {
+      "epoch": 0.07968,
+      "grad_norm": 3.9003215648414016,
+      "learning_rate": 0.00017619893428063945,
+      "loss": 4.576,
+      "step": 498
+    },
+    {
+      "epoch": 0.07984,
+      "grad_norm": 5.998414741180055,
+      "learning_rate": 0.00017655417406749557,
+      "loss": 4.6277,
+      "step": 499
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 4.366643153899971,
+      "learning_rate": 0.0001769094138543517,
+      "loss": 4.6364,
+      "step": 500
+    },
+    {
+      "epoch": 0.08016,
+      "grad_norm": 3.452703633339061,
+      "learning_rate": 0.00017726465364120784,
+      "loss": 4.5337,
+      "step": 501
+    },
+    {
+      "epoch": 0.08032,
+      "grad_norm": 2.818154588470989,
+      "learning_rate": 0.00017761989342806396,
+      "loss": 4.5667,
+      "step": 502
+    },
+    {
+      "epoch": 0.08048,
+      "grad_norm": 5.037229001095363,
+      "learning_rate": 0.00017797513321492008,
+      "loss": 4.5946,
+      "step": 503
+    },
+    {
+      "epoch": 0.08064,
+      "grad_norm": 3.890504540439526,
+      "learning_rate": 0.0001783303730017762,
+      "loss": 4.6169,
+      "step": 504
+    },
+    {
+      "epoch": 0.0808,
+      "grad_norm": 3.0895865354182215,
+      "learning_rate": 0.00017868561278863235,
+      "loss": 4.5782,
+      "step": 505
+    },
+    {
+      "epoch": 0.08096,
+      "grad_norm": 5.504764672013709,
+      "learning_rate": 0.00017904085257548847,
+      "loss": 4.6093,
+      "step": 506
+    },
+    {
+      "epoch": 0.08112,
+      "grad_norm": 4.935049273643551,
+      "learning_rate": 0.0001793960923623446,
+      "loss": 4.6709,
+      "step": 507
+    },
+    {
+      "epoch": 0.08128,
+      "grad_norm": 3.351114899172522,
+      "learning_rate": 0.00017975133214920072,
+      "loss": 4.5861,
+      "step": 508
+    },
+    {
+      "epoch": 0.08144,
+      "grad_norm": 3.1244842516092266,
+      "learning_rate": 0.00018010657193605684,
+      "loss": 4.6079,
+      "step": 509
+    },
+    {
+      "epoch": 0.0816,
+      "grad_norm": 3.551707686885247,
+      "learning_rate": 0.00018046181172291296,
+      "loss": 4.629,
+      "step": 510
+    },
+    {
+      "epoch": 0.08176,
+      "grad_norm": 3.313493240726521,
+      "learning_rate": 0.0001808170515097691,
+      "loss": 4.503,
+      "step": 511
+    },
+    {
+      "epoch": 0.08192,
+      "grad_norm": 3.5935956895587715,
+      "learning_rate": 0.00018117229129662523,
+      "loss": 4.6189,
+      "step": 512
+    },
+    {
+      "epoch": 0.08208,
+      "grad_norm": 3.0085708019137463,
+      "learning_rate": 0.00018152753108348135,
+      "loss": 4.5284,
+      "step": 513
+    },
+    {
+      "epoch": 0.08224,
+      "grad_norm": 4.532013009168544,
+      "learning_rate": 0.0001818827708703375,
+      "loss": 4.6097,
+      "step": 514
+    },
+    {
+      "epoch": 0.0824,
+      "grad_norm": 2.8711672675360806,
+      "learning_rate": 0.00018223801065719362,
+      "loss": 4.5557,
+      "step": 515
+    },
+    {
+      "epoch": 0.08256,
+      "grad_norm": 6.5011759643321785,
+      "learning_rate": 0.00018259325044404974,
+      "loss": 4.5706,
+      "step": 516
+    },
+    {
+      "epoch": 0.08272,
+      "grad_norm": 5.397142524775912,
+      "learning_rate": 0.0001829484902309059,
+      "loss": 4.6158,
+      "step": 517
+    },
+    {
+      "epoch": 0.08288,
+      "grad_norm": 3.4909160115891633,
+      "learning_rate": 0.000183303730017762,
+      "loss": 4.5509,
+      "step": 518
+    },
+    {
+      "epoch": 0.08304,
+      "grad_norm": 8.278770928169228,
+      "learning_rate": 0.00018365896980461813,
+      "loss": 4.631,
+      "step": 519
+    },
+    {
+      "epoch": 0.0832,
+      "grad_norm": 6.323960160693728,
+      "learning_rate": 0.00018401420959147425,
+      "loss": 4.5997,
+      "step": 520
+    },
+    {
+      "epoch": 0.08336,
+      "grad_norm": 4.797967487753939,
+      "learning_rate": 0.00018436944937833037,
+      "loss": 4.6725,
+      "step": 521
+    },
+    {
+      "epoch": 0.08352,
+      "grad_norm": 7.787835692448123,
+      "learning_rate": 0.0001847246891651865,
+      "loss": 4.6158,
+      "step": 522
+    },
+    {
+      "epoch": 0.08368,
+      "grad_norm": 2.94008585802761,
+      "learning_rate": 0.00018507992895204264,
+      "loss": 4.5018,
+      "step": 523
+    },
+    {
+      "epoch": 0.08384,
+      "grad_norm": 4.597265907678653,
+      "learning_rate": 0.00018543516873889876,
+      "loss": 4.5945,
+      "step": 524
+    },
+    {
+      "epoch": 0.084,
+      "grad_norm": 4.375406019814464,
+      "learning_rate": 0.00018579040852575489,
+      "loss": 4.615,
+      "step": 525
+    },
+    {
+      "epoch": 0.08416,
+      "grad_norm": 2.3715275182795077,
+      "learning_rate": 0.00018614564831261103,
+      "loss": 4.5076,
+      "step": 526
+    },
+    {
+      "epoch": 0.08432,
+      "grad_norm": 3.2495025231957104,
+      "learning_rate": 0.00018650088809946716,
+      "loss": 4.592,
+      "step": 527
+    },
+    {
+      "epoch": 0.08448,
+      "grad_norm": 3.251728958524214,
+      "learning_rate": 0.00018685612788632328,
+      "loss": 4.5861,
+      "step": 528
+    },
+    {
+      "epoch": 0.08464,
+      "grad_norm": 4.6671393792919025,
+      "learning_rate": 0.0001872113676731794,
+      "loss": 4.5545,
+      "step": 529
+    },
+    {
+      "epoch": 0.0848,
+      "grad_norm": 3.682120250469278,
+      "learning_rate": 0.00018756660746003555,
+      "loss": 4.6542,
+      "step": 530
+    },
+    {
+      "epoch": 0.08496,
+      "grad_norm": 2.470137855946921,
+      "learning_rate": 0.00018792184724689167,
+      "loss": 4.5269,
+      "step": 531
+    },
+    {
+      "epoch": 0.08512,
+      "grad_norm": 3.791089157594378,
+      "learning_rate": 0.0001882770870337478,
+      "loss": 4.5082,
+      "step": 532
+    },
+    {
+      "epoch": 0.08528,
+      "grad_norm": 2.8383293393205684,
+      "learning_rate": 0.0001886323268206039,
+      "loss": 4.5603,
+      "step": 533
+    },
+    {
+      "epoch": 0.08544,
+      "grad_norm": 4.544781024506043,
+      "learning_rate": 0.00018898756660746003,
+      "loss": 4.4664,
+      "step": 534
+    },
+    {
+      "epoch": 0.0856,
+      "grad_norm": 10.82415971770859,
+      "learning_rate": 0.00018934280639431615,
+      "loss": 4.5751,
+      "step": 535
+    },
+    {
+      "epoch": 0.08576,
+      "grad_norm": 11.824386790190202,
+      "learning_rate": 0.0001896980461811723,
+      "loss": 4.6677,
+      "step": 536
+    },
+    {
+      "epoch": 0.08592,
+      "grad_norm": 4.8861792867055005,
+      "learning_rate": 0.00019005328596802842,
+      "loss": 4.5572,
+      "step": 537
+    },
+    {
+      "epoch": 0.08608,
+      "grad_norm": 7.698808756623991,
+      "learning_rate": 0.00019040852575488454,
+      "loss": 4.5452,
+      "step": 538
+    },
+    {
+      "epoch": 0.08624,
+      "grad_norm": 5.23250339064728,
+      "learning_rate": 0.0001907637655417407,
+      "loss": 4.5674,
+      "step": 539
+    },
+    {
+      "epoch": 0.0864,
+      "grad_norm": 3.9240942813600075,
+      "learning_rate": 0.00019111900532859681,
+      "loss": 4.5352,
+      "step": 540
+    },
+    {
+      "epoch": 0.08656,
+      "grad_norm": 6.863072769394279,
+      "learning_rate": 0.00019147424511545294,
+      "loss": 4.5687,
+      "step": 541
+    },
+    {
+      "epoch": 0.08672,
+      "grad_norm": 2.99362259025842,
+      "learning_rate": 0.00019182948490230908,
+      "loss": 4.5756,
+      "step": 542
+    },
+    {
+      "epoch": 0.08688,
+      "grad_norm": 3.915433044691897,
+      "learning_rate": 0.0001921847246891652,
+      "loss": 4.4809,
+      "step": 543
+    },
+    {
+      "epoch": 0.08704,
+      "grad_norm": 3.6027628705528345,
+      "learning_rate": 0.00019253996447602133,
+      "loss": 4.4367,
+      "step": 544
+    },
+    {
+      "epoch": 0.0872,
+      "grad_norm": 2.5693914676091407,
+      "learning_rate": 0.00019289520426287745,
+      "loss": 4.4709,
+      "step": 545
+    },
+    {
+      "epoch": 0.08736,
+      "grad_norm": 3.839488596732451,
+      "learning_rate": 0.00019325044404973357,
+      "loss": 4.4864,
+      "step": 546
+    },
+    {
+      "epoch": 0.08752,
+      "grad_norm": 4.461169172854201,
+      "learning_rate": 0.0001936056838365897,
+      "loss": 4.4535,
+      "step": 547
+    },
+    {
+      "epoch": 0.08768,
+      "grad_norm": 3.7912602772522903,
+      "learning_rate": 0.00019396092362344584,
+      "loss": 4.4579,
+      "step": 548
+    },
+    {
+      "epoch": 0.08784,
+      "grad_norm": 2.533063269892783,
+      "learning_rate": 0.00019431616341030196,
+      "loss": 4.4365,
+      "step": 549
+    },
+    {
+      "epoch": 0.088,
+      "grad_norm": 3.43022026175714,
+      "learning_rate": 0.00019467140319715808,
+      "loss": 4.4181,
+      "step": 550
+    },
+    {
+      "epoch": 0.08816,
+      "grad_norm": 3.2029139439300836,
+      "learning_rate": 0.00019502664298401423,
+      "loss": 4.4874,
+      "step": 551
+    },
+    {
+      "epoch": 0.08832,
+      "grad_norm": 4.047446887445623,
+      "learning_rate": 0.00019538188277087035,
+      "loss": 4.395,
+      "step": 552
+    },
+    {
+      "epoch": 0.08848,
+      "grad_norm": 5.5946119984739155,
+      "learning_rate": 0.00019573712255772647,
+      "loss": 4.4259,
+      "step": 553
+    },
+    {
+      "epoch": 0.08864,
+      "grad_norm": 6.858544218846108,
+      "learning_rate": 0.0001960923623445826,
+      "loss": 4.5285,
+      "step": 554
+    },
+    {
+      "epoch": 0.0888,
+      "grad_norm": 6.642204203270011,
+      "learning_rate": 0.00019644760213143874,
+      "loss": 4.4385,
+      "step": 555
+    },
+    {
+      "epoch": 0.08896,
+      "grad_norm": 3.188349311109047,
+      "learning_rate": 0.00019680284191829486,
+      "loss": 4.4027,
+      "step": 556
+    },
+    {
+      "epoch": 0.08912,
+      "grad_norm": 8.17140086195297,
+      "learning_rate": 0.00019715808170515098,
+      "loss": 4.4647,
+      "step": 557
+    },
+    {
+      "epoch": 0.08928,
+      "grad_norm": 9.833042802406498,
+      "learning_rate": 0.0001975133214920071,
+      "loss": 4.5367,
+      "step": 558
+    },
+    {
+      "epoch": 0.08944,
+      "grad_norm": 4.760547220040306,
+      "learning_rate": 0.00019786856127886323,
+      "loss": 4.4256,
+      "step": 559
+    },
+    {
+      "epoch": 0.0896,
+      "grad_norm": 6.378038279576038,
+      "learning_rate": 0.00019822380106571935,
+      "loss": 4.5394,
+      "step": 560
+    },
+    {
+      "epoch": 0.08976,
+      "grad_norm": 5.0068888723760745,
+      "learning_rate": 0.0001985790408525755,
+      "loss": 4.5732,
+      "step": 561
+    },
+    {
+      "epoch": 0.08992,
+      "grad_norm": 7.29671167417037,
+      "learning_rate": 0.00019893428063943162,
+      "loss": 4.5458,
+      "step": 562
+    },
+    {
+      "epoch": 0.09008,
+      "grad_norm": 8.376756431865696,
+      "learning_rate": 0.00019928952042628774,
+      "loss": 4.5022,
+      "step": 563
+    },
+    {
+      "epoch": 0.09024,
+      "grad_norm": 4.311604208961875,
+      "learning_rate": 0.0001996447602131439,
+      "loss": 4.4658,
+      "step": 564
+    },
+    {
+      "epoch": 0.0904,
+      "grad_norm": 5.7134112755810715,
+      "learning_rate": 0.0002,
+      "loss": 4.5231,
+      "step": 565
+    },
+    {
+      "epoch": 0.09056,
+      "grad_norm": 5.257348086683008,
+      "learning_rate": 0.00019999999850807288,
+      "loss": 4.464,
+      "step": 566
+    },
+    {
+      "epoch": 0.09072,
+      "grad_norm": 3.6012670734860532,
+      "learning_rate": 0.00019999999403229148,
+      "loss": 4.3832,
+      "step": 567
+    },
+    {
+      "epoch": 0.09088,
+      "grad_norm": 4.551940240132202,
+      "learning_rate": 0.000199999986572656,
+      "loss": 4.3288,
+      "step": 568
+    },
+    {
+      "epoch": 0.09104,
+      "grad_norm": 3.479327809472697,
+      "learning_rate": 0.0001999999761291666,
+      "loss": 4.4966,
+      "step": 569
+    },
+    {
+      "epoch": 0.0912,
+      "grad_norm": 3.3869408394554843,
+      "learning_rate": 0.0001999999627018237,
+      "loss": 4.3554,
+      "step": 570
+    },
+    {
+      "epoch": 0.09136,
+      "grad_norm": 3.4765433863619375,
+      "learning_rate": 0.00019999994629062756,
+      "loss": 4.4191,
+      "step": 571
+    },
+    {
+      "epoch": 0.09152,
+      "grad_norm": 3.1439706842120105,
+      "learning_rate": 0.00019999992689557876,
+      "loss": 4.3668,
+      "step": 572
+    },
+    {
+      "epoch": 0.09168,
+      "grad_norm": 3.2012667914985387,
+      "learning_rate": 0.00019999990451667786,
+      "loss": 4.4392,
+      "step": 573
+    },
+    {
+      "epoch": 0.09184,
+      "grad_norm": 2.664610033539076,
+      "learning_rate": 0.00019999987915392552,
+      "loss": 4.3182,
+      "step": 574
+    },
+    {
+      "epoch": 0.092,
+      "grad_norm": 3.2592267620076405,
+      "learning_rate": 0.0001999998508073225,
+      "loss": 4.3715,
+      "step": 575
+    },
+    {
+      "epoch": 0.09216,
+      "grad_norm": 3.716304792395464,
+      "learning_rate": 0.00019999981947686966,
+      "loss": 4.3378,
+      "step": 576
+    },
+    {
+      "epoch": 0.09232,
+      "grad_norm": 3.3221976824038593,
+      "learning_rate": 0.0001999997851625679,
+      "loss": 4.3218,
+      "step": 577
+    },
+    {
+      "epoch": 0.09248,
+      "grad_norm": 2.8292929431669593,
+      "learning_rate": 0.00019999974786441828,
+      "loss": 4.3325,
+      "step": 578
+    },
+    {
+      "epoch": 0.09264,
+      "grad_norm": 4.019345820249025,
+      "learning_rate": 0.0001999997075824219,
+      "loss": 4.2778,
+      "step": 579
+    },
+    {
+      "epoch": 0.0928,
+      "grad_norm": 6.4737303699572175,
+      "learning_rate": 0.00019999966431657998,
+      "loss": 4.2801,
+      "step": 580
+    },
+    {
+      "epoch": 0.09296,
+      "grad_norm": 4.709807014065953,
+      "learning_rate": 0.00019999961806689375,
+      "loss": 4.3548,
+      "step": 581
+    },
+    {
+      "epoch": 0.09312,
+      "grad_norm": 4.829224615776024,
+      "learning_rate": 0.00019999956883336466,
+      "loss": 4.2527,
+      "step": 582
+    },
+    {
+      "epoch": 0.09328,
+      "grad_norm": 3.384020335343133,
+      "learning_rate": 0.00019999951661599412,
+      "loss": 4.3552,
+      "step": 583
+    },
+    {
+      "epoch": 0.09344,
+      "grad_norm": 3.5628128166692536,
+      "learning_rate": 0.00019999946141478376,
+      "loss": 4.3335,
+      "step": 584
+    },
+    {
+      "epoch": 0.0936,
+      "grad_norm": 2.4268893409637955,
+      "learning_rate": 0.00019999940322973516,
+      "loss": 4.1893,
+      "step": 585
+    },
+    {
+      "epoch": 0.09376,
+      "grad_norm": 3.738461504643024,
+      "learning_rate": 0.0001999993420608501,
+      "loss": 4.2518,
+      "step": 586
+    },
+    {
+      "epoch": 0.09392,
+      "grad_norm": 3.894316529856452,
+      "learning_rate": 0.00019999927790813037,
+      "loss": 4.3041,
+      "step": 587
+    },
+    {
+      "epoch": 0.09408,
+      "grad_norm": 4.479446474467455,
+      "learning_rate": 0.0001999992107715779,
+      "loss": 4.3296,
+      "step": 588
+    },
+    {
+      "epoch": 0.09424,
+      "grad_norm": 3.6519172245373217,
+      "learning_rate": 0.0001999991406511947,
+      "loss": 4.2297,
+      "step": 589
+    },
+    {
+      "epoch": 0.0944,
+      "grad_norm": 3.179258614685981,
+      "learning_rate": 0.00019999906754698287,
+      "loss": 4.3281,
+      "step": 590
+    },
+    {
+      "epoch": 0.09456,
+      "grad_norm": 2.9124290433090945,
+      "learning_rate": 0.00019999899145894457,
+      "loss": 4.2769,
+      "step": 591
+    },
+    {
+      "epoch": 0.09472,
+      "grad_norm": 3.616430089041825,
+      "learning_rate": 0.00019999891238708207,
+      "loss": 4.2173,
+      "step": 592
+    },
+    {
+      "epoch": 0.09488,
+      "grad_norm": 2.821098894765023,
+      "learning_rate": 0.00019999883033139777,
+      "loss": 4.1873,
+      "step": 593
+    },
+    {
+      "epoch": 0.09504,
+      "grad_norm": 4.127632955634765,
+      "learning_rate": 0.00019999874529189404,
+      "loss": 4.2553,
+      "step": 594
+    },
+    {
+      "epoch": 0.0952,
+      "grad_norm": 5.560835535712226,
+      "learning_rate": 0.00019999865726857348,
+      "loss": 4.1888,
+      "step": 595
+    },
+    {
+      "epoch": 0.09536,
+      "grad_norm": 2.6847718821821682,
+      "learning_rate": 0.0001999985662614387,
+      "loss": 4.2608,
+      "step": 596
+    },
+    {
+      "epoch": 0.09552,
+      "grad_norm": 3.238467473426333,
+      "learning_rate": 0.00019999847227049244,
+      "loss": 4.2924,
+      "step": 597
+    },
+    {
+      "epoch": 0.09568,
+      "grad_norm": 4.519775608940327,
+      "learning_rate": 0.00019999837529573747,
+      "loss": 4.2776,
+      "step": 598
+    },
+    {
+      "epoch": 0.09584,
+      "grad_norm": 3.698073366325793,
+      "learning_rate": 0.00019999827533717667,
+      "loss": 4.2005,
+      "step": 599
+    },
+    {
+      "epoch": 0.096,
+      "grad_norm": 3.0305907380371253,
+      "learning_rate": 0.00019999817239481308,
+      "loss": 4.1367,
+      "step": 600
+    },
+    {
+      "epoch": 0.09616,
+      "grad_norm": 4.523693051787799,
+      "learning_rate": 0.0001999980664686497,
+      "loss": 4.1254,
+      "step": 601
+    },
+    {
+      "epoch": 0.09632,
+      "grad_norm": 4.294498989719827,
+      "learning_rate": 0.00019999795755868977,
+      "loss": 4.2799,
+      "step": 602
+    },
+    {
+      "epoch": 0.09648,
+      "grad_norm": 3.221057029586693,
+      "learning_rate": 0.00019999784566493648,
+      "loss": 4.1096,
+      "step": 603
+    },
+    {
+      "epoch": 0.09664,
+      "grad_norm": 4.433245371793626,
+      "learning_rate": 0.00019999773078739318,
+      "loss": 4.2301,
+      "step": 604
+    },
+    {
+      "epoch": 0.0968,
+      "grad_norm": 5.332357395680833,
+      "learning_rate": 0.00019999761292606333,
+      "loss": 4.188,
+      "step": 605
+    },
+    {
+      "epoch": 0.09696,
+      "grad_norm": 3.399863259844689,
+      "learning_rate": 0.0001999974920809504,
+      "loss": 4.1742,
+      "step": 606
+    },
+    {
+      "epoch": 0.09712,
+      "grad_norm": 3.505633833266201,
+      "learning_rate": 0.00019999736825205804,
+      "loss": 4.2288,
+      "step": 607
+    },
+    {
+      "epoch": 0.09728,
+      "grad_norm": 3.0755654213303965,
+      "learning_rate": 0.0001999972414393899,
+      "loss": 4.1731,
+      "step": 608
+    },
+    {
+      "epoch": 0.09744,
+      "grad_norm": 3.730676624113583,
+      "learning_rate": 0.00019999711164294978,
+      "loss": 4.1918,
+      "step": 609
+    },
+    {
+      "epoch": 0.0976,
+      "grad_norm": 3.3390738546920202,
+      "learning_rate": 0.00019999697886274157,
+      "loss": 4.1116,
+      "step": 610
+    },
+    {
+      "epoch": 0.09776,
+      "grad_norm": 3.083790861260367,
+      "learning_rate": 0.00019999684309876924,
+      "loss": 4.1468,
+      "step": 611
+    },
+    {
+      "epoch": 0.09792,
+      "grad_norm": 3.3059205506524285,
+      "learning_rate": 0.0001999967043510368,
+      "loss": 4.1615,
+      "step": 612
+    },
+    {
+      "epoch": 0.09808,
+      "grad_norm": 3.950806740876017,
+      "learning_rate": 0.00019999656261954842,
+      "loss": 4.0868,
+      "step": 613
+    },
+    {
+      "epoch": 0.09824,
+      "grad_norm": 3.998775013169602,
+      "learning_rate": 0.00019999641790430835,
+      "loss": 4.1136,
+      "step": 614
+    },
+    {
+      "epoch": 0.0984,
+      "grad_norm": 2.9099602740901065,
+      "learning_rate": 0.00019999627020532085,
+      "loss": 4.1072,
+      "step": 615
+    },
+    {
+      "epoch": 0.09856,
+      "grad_norm": 3.1266881301914484,
+      "learning_rate": 0.00019999611952259034,
+      "loss": 4.1115,
+      "step": 616
+    },
+    {
+      "epoch": 0.09872,
+      "grad_norm": 2.899669670038305,
+      "learning_rate": 0.00019999596585612136,
+      "loss": 4.1159,
+      "step": 617
+    },
+    {
+      "epoch": 0.09888,
+      "grad_norm": 3.9097072950521805,
+      "learning_rate": 0.00019999580920591846,
+      "loss": 4.1821,
+      "step": 618
+    },
+    {
+      "epoch": 0.09904,
+      "grad_norm": 3.465805511068144,
+      "learning_rate": 0.0001999956495719863,
+      "loss": 4.0619,
+      "step": 619
+    },
+    {
+      "epoch": 0.0992,
+      "grad_norm": 3.7654028661663954,
+      "learning_rate": 0.0001999954869543297,
+      "loss": 4.1137,
+      "step": 620
+    },
+    {
+      "epoch": 0.09936,
+      "grad_norm": 3.935863197034617,
+      "learning_rate": 0.00019999532135295346,
+      "loss": 4.1214,
+      "step": 621
+    },
+    {
+      "epoch": 0.09952,
+      "grad_norm": 3.239810609049369,
+      "learning_rate": 0.00019999515276786255,
+      "loss": 4.0817,
+      "step": 622
+    },
+    {
+      "epoch": 0.09968,
+      "grad_norm": 4.205797395814603,
+      "learning_rate": 0.00019999498119906202,
+      "loss": 4.2188,
+      "step": 623
+    },
+    {
+      "epoch": 0.09984,
+      "grad_norm": 4.593333572331768,
+      "learning_rate": 0.0001999948066465569,
+      "loss": 4.0652,
+      "step": 624
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 3.1093637874363065,
+      "learning_rate": 0.00019999462911035249,
+      "loss": 4.1814,
+      "step": 625
+    },
+    {
+      "epoch": 0.10016,
+      "grad_norm": 4.08668096987421,
+      "learning_rate": 0.00019999444859045403,
+      "loss": 4.0876,
+      "step": 626
+    },
+    {
+      "epoch": 0.10032,
+      "grad_norm": 3.8628526677162176,
+      "learning_rate": 0.00019999426508686695,
+      "loss": 4.1439,
+      "step": 627
+    },
+    {
+      "epoch": 0.10048,
+      "grad_norm": 2.895391655231612,
+      "learning_rate": 0.00019999407859959672,
+      "loss": 4.0633,
+      "step": 628
+    },
+    {
+      "epoch": 0.10064,
+      "grad_norm": 3.6774198493649353,
+      "learning_rate": 0.0001999938891286489,
+      "loss": 4.2024,
+      "step": 629
+    },
+    {
+      "epoch": 0.1008,
+      "grad_norm": 3.2850613662566053,
+      "learning_rate": 0.0001999936966740291,
+      "loss": 4.1053,
+      "step": 630
+    },
+    {
+      "epoch": 0.10096,
+      "grad_norm": 3.831903783711723,
+      "learning_rate": 0.0001999935012357431,
+      "loss": 4.0639,
+      "step": 631
+    },
+    {
+      "epoch": 0.10112,
+      "grad_norm": 4.577241798048573,
+      "learning_rate": 0.00019999330281379675,
+      "loss": 4.1686,
+      "step": 632
+    },
+    {
+      "epoch": 0.10128,
+      "grad_norm": 3.207546355968215,
+      "learning_rate": 0.00019999310140819595,
+      "loss": 4.1252,
+      "step": 633
+    },
+    {
+      "epoch": 0.10144,
+      "grad_norm": 2.6267071309407988,
+      "learning_rate": 0.0001999928970189467,
+      "loss": 4.119,
+      "step": 634
+    },
+    {
+      "epoch": 0.1016,
+      "grad_norm": 3.1289576698199384,
+      "learning_rate": 0.0001999926896460551,
+      "loss": 4.118,
+      "step": 635
+    },
+    {
+      "epoch": 0.10176,
+      "grad_norm": 2.8150189230238234,
+      "learning_rate": 0.00019999247928952736,
+      "loss": 4.0172,
+      "step": 636
+    },
+    {
+      "epoch": 0.10192,
+      "grad_norm": 2.597115042423837,
+      "learning_rate": 0.00019999226594936972,
+      "loss": 3.9759,
+      "step": 637
+    },
+    {
+      "epoch": 0.10208,
+      "grad_norm": 3.05574660838907,
+      "learning_rate": 0.00019999204962558857,
+      "loss": 3.9694,
+      "step": 638
+    },
+    {
+      "epoch": 0.10224,
+      "grad_norm": 3.3270456990247386,
+      "learning_rate": 0.00019999183031819037,
+      "loss": 4.0133,
+      "step": 639
+    },
+    {
+      "epoch": 0.1024,
+      "grad_norm": 3.183231080093459,
+      "learning_rate": 0.00019999160802718167,
+      "loss": 4.0888,
+      "step": 640
+    },
+    {
+      "epoch": 0.10256,
+      "grad_norm": 3.1480592095221223,
+      "learning_rate": 0.0001999913827525691,
+      "loss": 4.041,
+      "step": 641
+    },
+    {
+      "epoch": 0.10272,
+      "grad_norm": 3.0625693318663845,
+      "learning_rate": 0.00019999115449435934,
+      "loss": 3.9084,
+      "step": 642
+    },
+    {
+      "epoch": 0.10288,
+      "grad_norm": 3.2675192311255774,
+      "learning_rate": 0.00019999092325255924,
+      "loss": 3.9651,
+      "step": 643
+    },
+    {
+      "epoch": 0.10304,
+      "grad_norm": 3.158425225605362,
+      "learning_rate": 0.00019999068902717572,
+      "loss": 4.0251,
+      "step": 644
+    },
+    {
+      "epoch": 0.1032,
+      "grad_norm": 3.1873876768479374,
+      "learning_rate": 0.00019999045181821572,
+      "loss": 3.9247,
+      "step": 645
+    },
+    {
+      "epoch": 0.10336,
+      "grad_norm": 3.9445131587435087,
+      "learning_rate": 0.00019999021162568634,
+      "loss": 4.005,
+      "step": 646
+    },
+    {
+      "epoch": 0.10352,
+      "grad_norm": 2.7567909831541435,
+      "learning_rate": 0.00019998996844959471,
+      "loss": 3.9863,
+      "step": 647
+    },
+    {
+      "epoch": 0.10368,
+      "grad_norm": 4.636598494538182,
+      "learning_rate": 0.0001999897222899482,
+      "loss": 4.0123,
+      "step": 648
+    },
+    {
+      "epoch": 0.10384,
+      "grad_norm": 3.5908094354868862,
+      "learning_rate": 0.00019998947314675403,
+      "loss": 3.9176,
+      "step": 649
+    },
+    {
+      "epoch": 0.104,
+      "grad_norm": 3.141963638351258,
+      "learning_rate": 0.00019998922102001967,
+      "loss": 3.9872,
+      "step": 650
+    },
+    {
+      "epoch": 0.10416,
+      "grad_norm": 2.620361525270701,
+      "learning_rate": 0.00019998896590975266,
+      "loss": 3.9173,
+      "step": 651
+    },
+    {
+      "epoch": 0.10432,
+      "grad_norm": 4.908712584310313,
+      "learning_rate": 0.0001999887078159606,
+      "loss": 3.9733,
+      "step": 652
+    },
+    {
+      "epoch": 0.10448,
+      "grad_norm": 3.753967155467085,
+      "learning_rate": 0.00019998844673865124,
+      "loss": 4.0449,
+      "step": 653
+    },
+    {
+      "epoch": 0.10464,
+      "grad_norm": 2.235509667538088,
+      "learning_rate": 0.0001999881826778323,
+      "loss": 4.0019,
+      "step": 654
+    },
+    {
+      "epoch": 0.1048,
+      "grad_norm": 3.0238502554760633,
+      "learning_rate": 0.0001999879156335117,
+      "loss": 4.0399,
+      "step": 655
+    },
+    {
+      "epoch": 0.10496,
+      "grad_norm": 5.012227747354559,
+      "learning_rate": 0.00019998764560569737,
+      "loss": 3.89,
+      "step": 656
+    },
+    {
+      "epoch": 0.10512,
+      "grad_norm": 6.217837850100138,
+      "learning_rate": 0.00019998737259439741,
+      "loss": 4.1052,
+      "step": 657
+    },
+    {
+      "epoch": 0.10528,
+      "grad_norm": 3.3925785174229586,
+      "learning_rate": 0.00019998709659961995,
+      "loss": 3.795,
+      "step": 658
+    },
+    {
+      "epoch": 0.10544,
+      "grad_norm": 5.582782466738514,
+      "learning_rate": 0.00019998681762137323,
+      "loss": 3.9532,
+      "step": 659
+    },
+    {
+      "epoch": 0.1056,
+      "grad_norm": 3.92797240271547,
+      "learning_rate": 0.00019998653565966556,
+      "loss": 3.9784,
+      "step": 660
+    },
+    {
+      "epoch": 0.10576,
+      "grad_norm": 4.289816835649352,
+      "learning_rate": 0.00019998625071450534,
+      "loss": 4.0424,
+      "step": 661
+    },
+    {
+      "epoch": 0.10592,
+      "grad_norm": 5.252578839940932,
+      "learning_rate": 0.00019998596278590112,
+      "loss": 3.9167,
+      "step": 662
+    },
+    {
+      "epoch": 0.10608,
+      "grad_norm": 2.8119588635576487,
+      "learning_rate": 0.00019998567187386147,
+      "loss": 3.8633,
+      "step": 663
+    },
+    {
+      "epoch": 0.10624,
+      "grad_norm": 3.1898534971561707,
+      "learning_rate": 0.00019998537797839507,
+      "loss": 3.9406,
+      "step": 664
+    },
+    {
+      "epoch": 0.1064,
+      "grad_norm": 2.6520945899930455,
+      "learning_rate": 0.00019998508109951067,
+      "loss": 3.9922,
+      "step": 665
+    },
+    {
+      "epoch": 0.10656,
+      "grad_norm": 3.2651607698442104,
+      "learning_rate": 0.00019998478123721714,
+      "loss": 3.803,
+      "step": 666
+    },
+    {
+      "epoch": 0.10672,
+      "grad_norm": 2.5851456342004826,
+      "learning_rate": 0.00019998447839152342,
+      "loss": 3.9387,
+      "step": 667
+    },
+    {
+      "epoch": 0.10688,
+      "grad_norm": 3.854909378650117,
+      "learning_rate": 0.00019998417256243857,
+      "loss": 3.9239,
+      "step": 668
+    },
+    {
+      "epoch": 0.10704,
+      "grad_norm": 3.0870594456474323,
+      "learning_rate": 0.0001999838637499717,
+      "loss": 3.8432,
+      "step": 669
+    },
+    {
+      "epoch": 0.1072,
+      "grad_norm": 3.333231792173839,
+      "learning_rate": 0.00019998355195413204,
+      "loss": 3.901,
+      "step": 670
+    },
+    {
+      "epoch": 0.10736,
+      "grad_norm": 3.2045635222380344,
+      "learning_rate": 0.00019998323717492888,
+      "loss": 3.8865,
+      "step": 671
+    },
+    {
+      "epoch": 0.10752,
+      "grad_norm": 2.81827297738337,
+      "learning_rate": 0.00019998291941237162,
+      "loss": 3.7388,
+      "step": 672
+    },
+    {
+      "epoch": 0.10768,
+      "grad_norm": 3.5899785255057615,
+      "learning_rate": 0.00019998259866646972,
+      "loss": 3.8493,
+      "step": 673
+    },
+    {
+      "epoch": 0.10784,
+      "grad_norm": 3.373023348857671,
+      "learning_rate": 0.00019998227493723278,
+      "loss": 3.8408,
+      "step": 674
+    },
+    {
+      "epoch": 0.108,
+      "grad_norm": 2.8633120311205342,
+      "learning_rate": 0.00019998194822467042,
+      "loss": 3.8199,
+      "step": 675
+    },
+    {
+      "epoch": 0.10816,
+      "grad_norm": 4.181818624789961,
+      "learning_rate": 0.00019998161852879243,
+      "loss": 3.8564,
+      "step": 676
+    },
+    {
+      "epoch": 0.10832,
+      "grad_norm": 3.9671817492580086,
+      "learning_rate": 0.0001999812858496086,
+      "loss": 3.7619,
+      "step": 677
+    },
+    {
+      "epoch": 0.10848,
+      "grad_norm": 3.9447150275498104,
+      "learning_rate": 0.00019998095018712895,
+      "loss": 3.8697,
+      "step": 678
+    },
+    {
+      "epoch": 0.10864,
+      "grad_norm": 5.188155977478405,
+      "learning_rate": 0.0001999806115413634,
+      "loss": 3.8241,
+      "step": 679
+    },
+    {
+      "epoch": 0.1088,
+      "grad_norm": 3.754671764209342,
+      "learning_rate": 0.0001999802699123221,
+      "loss": 3.7768,
+      "step": 680
+    },
+    {
+      "epoch": 0.10896,
+      "grad_norm": 2.9672362080843486,
+      "learning_rate": 0.00019997992530001524,
+      "loss": 3.7823,
+      "step": 681
+    },
+    {
+      "epoch": 0.10912,
+      "grad_norm": 3.773897704697294,
+      "learning_rate": 0.00019997957770445307,
+      "loss": 3.9379,
+      "step": 682
+    },
+    {
+      "epoch": 0.10928,
+      "grad_norm": 3.4929670816414644,
+      "learning_rate": 0.00019997922712564599,
+      "loss": 3.8071,
+      "step": 683
+    },
+    {
+      "epoch": 0.10944,
+      "grad_norm": 3.755229081878822,
+      "learning_rate": 0.00019997887356360447,
+      "loss": 3.7829,
+      "step": 684
+    },
+    {
+      "epoch": 0.1096,
+      "grad_norm": 2.993634833446593,
+      "learning_rate": 0.00019997851701833902,
+      "loss": 3.7655,
+      "step": 685
+    },
+    {
+      "epoch": 0.10976,
+      "grad_norm": 3.0695900269314293,
+      "learning_rate": 0.00019997815748986036,
+      "loss": 3.7138,
+      "step": 686
+    },
+    {
+      "epoch": 0.10992,
+      "grad_norm": 3.3413496827949882,
+      "learning_rate": 0.00019997779497817912,
+      "loss": 3.7652,
+      "step": 687
+    },
+    {
+      "epoch": 0.11008,
+      "grad_norm": 3.944767260556355,
+      "learning_rate": 0.00019997742948330618,
+      "loss": 3.8138,
+      "step": 688
+    },
+    {
+      "epoch": 0.11024,
+      "grad_norm": 2.4924915021896448,
+      "learning_rate": 0.0001999770610052524,
+      "loss": 3.7647,
+      "step": 689
+    },
+    {
+      "epoch": 0.1104,
+      "grad_norm": 4.425394468449353,
+      "learning_rate": 0.00019997668954402885,
+      "loss": 3.7555,
+      "step": 690
+    },
+    {
+      "epoch": 0.11056,
+      "grad_norm": 2.4911439181850255,
+      "learning_rate": 0.00019997631509964653,
+      "loss": 3.8493,
+      "step": 691
+    },
+    {
+      "epoch": 0.11072,
+      "grad_norm": 2.8837356626625232,
+      "learning_rate": 0.00019997593767211665,
+      "loss": 3.7414,
+      "step": 692
+    },
+    {
+      "epoch": 0.11088,
+      "grad_norm": 2.912719720207602,
+      "learning_rate": 0.00019997555726145047,
+      "loss": 3.8865,
+      "step": 693
+    },
+    {
+      "epoch": 0.11104,
+      "grad_norm": 3.4693862546574756,
+      "learning_rate": 0.00019997517386765936,
+      "loss": 3.8345,
+      "step": 694
+    },
+    {
+      "epoch": 0.1112,
+      "grad_norm": 4.365453578578437,
+      "learning_rate": 0.00019997478749075471,
+      "loss": 3.7322,
+      "step": 695
+    },
+    {
+      "epoch": 0.11136,
+      "grad_norm": 3.860097706038017,
+      "learning_rate": 0.00019997439813074813,
+      "loss": 3.5972,
+      "step": 696
+    },
+    {
+      "epoch": 0.11152,
+      "grad_norm": 2.742472805439115,
+      "learning_rate": 0.00019997400578765113,
+      "loss": 3.8605,
+      "step": 697
+    },
+    {
+      "epoch": 0.11168,
+      "grad_norm": 4.102638579512068,
+      "learning_rate": 0.00019997361046147552,
+      "loss": 3.7797,
+      "step": 698
+    },
+    {
+      "epoch": 0.11184,
+      "grad_norm": 2.4286993214688377,
+      "learning_rate": 0.000199973212152233,
+      "loss": 3.8629,
+      "step": 699
+    },
+    {
+      "epoch": 0.112,
+      "grad_norm": 3.1462240864931434,
+      "learning_rate": 0.00019997281085993557,
+      "loss": 3.7212,
+      "step": 700
+    },
+    {
+      "epoch": 0.11216,
+      "grad_norm": 3.053384152484509,
+      "learning_rate": 0.0001999724065845951,
+      "loss": 3.7687,
+      "step": 701
+    },
+    {
+      "epoch": 0.11232,
+      "grad_norm": 2.598654178795769,
+      "learning_rate": 0.00019997199932622367,
+      "loss": 3.6954,
+      "step": 702
+    },
+    {
+      "epoch": 0.11248,
+      "grad_norm": 2.934498248131821,
+      "learning_rate": 0.0001999715890848335,
+      "loss": 3.8117,
+      "step": 703
+    },
+    {
+      "epoch": 0.11264,
+      "grad_norm": 2.493089438936112,
+      "learning_rate": 0.00019997117586043677,
+      "loss": 3.7461,
+      "step": 704
+    },
+    {
+      "epoch": 0.1128,
+      "grad_norm": 3.3425043325750643,
+      "learning_rate": 0.00019997075965304584,
+      "loss": 3.7492,
+      "step": 705
+    },
+    {
+      "epoch": 0.11296,
+      "grad_norm": 3.0171247003912165,
+      "learning_rate": 0.00019997034046267308,
+      "loss": 3.773,
+      "step": 706
+    },
+    {
+      "epoch": 0.11312,
+      "grad_norm": 2.606771321836323,
+      "learning_rate": 0.00019996991828933106,
+      "loss": 3.7609,
+      "step": 707
+    },
+    {
+      "epoch": 0.11328,
+      "grad_norm": 3.391149833522026,
+      "learning_rate": 0.00019996949313303236,
+      "loss": 3.7922,
+      "step": 708
+    },
+    {
+      "epoch": 0.11344,
+      "grad_norm": 3.012824468688187,
+      "learning_rate": 0.00019996906499378963,
+      "loss": 3.7188,
+      "step": 709
+    },
+    {
+      "epoch": 0.1136,
+      "grad_norm": 3.7410690217968776,
+      "learning_rate": 0.00019996863387161567,
+      "loss": 3.7352,
+      "step": 710
+    },
+    {
+      "epoch": 0.11376,
+      "grad_norm": 5.073069648268586,
+      "learning_rate": 0.00019996819976652336,
+      "loss": 3.7077,
+      "step": 711
+    },
+    {
+      "epoch": 0.11392,
+      "grad_norm": 4.408015385728824,
+      "learning_rate": 0.00019996776267852565,
+      "loss": 3.6524,
+      "step": 712
+    },
+    {
+      "epoch": 0.11408,
+      "grad_norm": 4.081289629952381,
+      "learning_rate": 0.00019996732260763554,
+      "loss": 3.658,
+      "step": 713
+    },
+    {
+      "epoch": 0.11424,
+      "grad_norm": 7.393640585647136,
+      "learning_rate": 0.0001999668795538662,
+      "loss": 3.6222,
+      "step": 714
+    },
+    {
+      "epoch": 0.1144,
+      "grad_norm": 3.149606541030356,
+      "learning_rate": 0.00019996643351723084,
+      "loss": 3.6393,
+      "step": 715
+    },
+    {
+      "epoch": 0.11456,
+      "grad_norm": 3.747756633775693,
+      "learning_rate": 0.00019996598449774277,
+      "loss": 3.7813,
+      "step": 716
+    },
+    {
+      "epoch": 0.11472,
+      "grad_norm": 4.198182952016631,
+      "learning_rate": 0.0001999655324954154,
+      "loss": 3.637,
+      "step": 717
+    },
+    {
+      "epoch": 0.11488,
+      "grad_norm": 2.620232088606557,
+      "learning_rate": 0.0001999650775102622,
+      "loss": 3.6928,
+      "step": 718
+    },
+    {
+      "epoch": 0.11504,
+      "grad_norm": 3.51331091276312,
+      "learning_rate": 0.00019996461954229676,
+      "loss": 3.5561,
+      "step": 719
+    },
+    {
+      "epoch": 0.1152,
+      "grad_norm": 2.534694196609995,
+      "learning_rate": 0.0001999641585915327,
+      "loss": 3.6146,
+      "step": 720
+    },
+    {
+      "epoch": 0.11536,
+      "grad_norm": 3.9021383838926447,
+      "learning_rate": 0.00019996369465798385,
+      "loss": 3.6543,
+      "step": 721
+    },
+    {
+      "epoch": 0.11552,
+      "grad_norm": 3.6945693324313478,
+      "learning_rate": 0.000199963227741664,
+      "loss": 3.6617,
+      "step": 722
+    },
+    {
+      "epoch": 0.11568,
+      "grad_norm": 4.659293446604254,
+      "learning_rate": 0.0001999627578425871,
+      "loss": 3.6603,
+      "step": 723
+    },
+    {
+      "epoch": 0.11584,
+      "grad_norm": 5.144397747802922,
+      "learning_rate": 0.00019996228496076718,
+      "loss": 3.6397,
+      "step": 724
+    },
+    {
+      "epoch": 0.116,
+      "grad_norm": 3.3759618596900824,
+      "learning_rate": 0.00019996180909621832,
+      "loss": 3.5811,
+      "step": 725
+    },
+    {
+      "epoch": 0.11616,
+      "grad_norm": 3.194266677063017,
+      "learning_rate": 0.00019996133024895475,
+      "loss": 3.7295,
+      "step": 726
+    },
+    {
+      "epoch": 0.11632,
+      "grad_norm": 2.934593854772452,
+      "learning_rate": 0.00019996084841899072,
+      "loss": 3.6446,
+      "step": 727
+    },
+    {
+      "epoch": 0.11648,
+      "grad_norm": 2.8101812511142064,
+      "learning_rate": 0.00019996036360634065,
+      "loss": 3.5755,
+      "step": 728
+    },
+    {
+      "epoch": 0.11664,
+      "grad_norm": 3.6674803864732106,
+      "learning_rate": 0.00019995987581101896,
+      "loss": 3.6086,
+      "step": 729
+    },
+    {
+      "epoch": 0.1168,
+      "grad_norm": 3.1772028072741914,
+      "learning_rate": 0.00019995938503304027,
+      "loss": 3.5855,
+      "step": 730
+    },
+    {
+      "epoch": 0.11696,
+      "grad_norm": 3.0656097371917053,
+      "learning_rate": 0.00019995889127241915,
+      "loss": 3.6896,
+      "step": 731
+    },
+    {
+      "epoch": 0.11712,
+      "grad_norm": 3.2353743496730543,
+      "learning_rate": 0.0001999583945291704,
+      "loss": 3.6321,
+      "step": 732
+    },
+    {
+      "epoch": 0.11728,
+      "grad_norm": 3.6546933459629485,
+      "learning_rate": 0.00019995789480330882,
+      "loss": 3.5314,
+      "step": 733
+    },
+    {
+      "epoch": 0.11744,
+      "grad_norm": 3.635949606099434,
+      "learning_rate": 0.0001999573920948493,
+      "loss": 3.6251,
+      "step": 734
+    },
+    {
+      "epoch": 0.1176,
+      "grad_norm": 2.5197418007322607,
+      "learning_rate": 0.00019995688640380682,
+      "loss": 3.6278,
+      "step": 735
+    },
+    {
+      "epoch": 0.11776,
+      "grad_norm": 3.4771317537988278,
+      "learning_rate": 0.00019995637773019656,
+      "loss": 3.6591,
+      "step": 736
+    },
+    {
+      "epoch": 0.11792,
+      "grad_norm": 3.5234675893468848,
+      "learning_rate": 0.0001999558660740336,
+      "loss": 3.6834,
+      "step": 737
+    },
+    {
+      "epoch": 0.11808,
+      "grad_norm": 3.2045416332107624,
+      "learning_rate": 0.00019995535143533326,
+      "loss": 3.5606,
+      "step": 738
+    },
+    {
+      "epoch": 0.11824,
+      "grad_norm": 3.558786469105246,
+      "learning_rate": 0.0001999548338141109,
+      "loss": 3.5256,
+      "step": 739
+    },
+    {
+      "epoch": 0.1184,
+      "grad_norm": 3.3108129046672743,
+      "learning_rate": 0.0001999543132103819,
+      "loss": 3.3877,
+      "step": 740
+    },
+    {
+      "epoch": 0.11856,
+      "grad_norm": 2.7015850414819926,
+      "learning_rate": 0.0001999537896241619,
+      "loss": 3.5828,
+      "step": 741
+    },
+    {
+      "epoch": 0.11872,
+      "grad_norm": 3.258071642612808,
+      "learning_rate": 0.00019995326305546644,
+      "loss": 3.5216,
+      "step": 742
+    },
+    {
+      "epoch": 0.11888,
+      "grad_norm": 3.295939511396412,
+      "learning_rate": 0.00019995273350431126,
+      "loss": 3.5468,
+      "step": 743
+    },
+    {
+      "epoch": 0.11904,
+      "grad_norm": 2.8888932276513426,
+      "learning_rate": 0.00019995220097071217,
+      "loss": 3.5015,
+      "step": 744
+    },
+    {
+      "epoch": 0.1192,
+      "grad_norm": 3.9067238003860867,
+      "learning_rate": 0.00019995166545468503,
+      "loss": 3.6746,
+      "step": 745
+    },
+    {
+      "epoch": 0.11936,
+      "grad_norm": 3.022836062119301,
+      "learning_rate": 0.00019995112695624583,
+      "loss": 3.3817,
+      "step": 746
+    },
+    {
+      "epoch": 0.11952,
+      "grad_norm": 3.15088928263215,
+      "learning_rate": 0.00019995058547541067,
+      "loss": 3.5488,
+      "step": 747
+    },
+    {
+      "epoch": 0.11968,
+      "grad_norm": 2.962191137973988,
+      "learning_rate": 0.00019995004101219568,
+      "loss": 3.4671,
+      "step": 748
+    },
+    {
+      "epoch": 0.11984,
+      "grad_norm": 2.573916533386747,
+      "learning_rate": 0.0001999494935666171,
+      "loss": 3.4453,
+      "step": 749
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 3.1406775893345724,
+      "learning_rate": 0.00019994894313869125,
+      "loss": 3.4603,
+      "step": 750
+    },
+    {
+      "epoch": 0.12016,
+      "grad_norm": 3.6871908164827807,
+      "learning_rate": 0.00019994838972843461,
+      "loss": 3.5223,
+      "step": 751
+    },
+    {
+      "epoch": 0.12032,
+      "grad_norm": 3.5869830771990365,
+      "learning_rate": 0.00019994783333586365,
+      "loss": 3.446,
+      "step": 752
+    },
+    {
+      "epoch": 0.12048,
+      "grad_norm": 2.432721545684539,
+      "learning_rate": 0.000199947273960995,
+      "loss": 3.5323,
+      "step": 753
+    },
+    {
+      "epoch": 0.12064,
+      "grad_norm": 3.016869375892028,
+      "learning_rate": 0.00019994671160384532,
+      "loss": 3.5213,
+      "step": 754
+    },
+    {
+      "epoch": 0.1208,
+      "grad_norm": 2.5197605403447394,
+      "learning_rate": 0.0001999461462644314,
+      "loss": 3.4051,
+      "step": 755
+    },
+    {
+      "epoch": 0.12096,
+      "grad_norm": 3.632753795103309,
+      "learning_rate": 0.00019994557794277012,
+      "loss": 3.4682,
+      "step": 756
+    },
+    {
+      "epoch": 0.12112,
+      "grad_norm": 3.512650674336017,
+      "learning_rate": 0.00019994500663887842,
+      "loss": 3.4964,
+      "step": 757
+    },
+    {
+      "epoch": 0.12128,
+      "grad_norm": 3.656980437408664,
+      "learning_rate": 0.00019994443235277337,
+      "loss": 3.3921,
+      "step": 758
+    },
+    {
+      "epoch": 0.12144,
+      "grad_norm": 3.2972041705570767,
+      "learning_rate": 0.0001999438550844721,
+      "loss": 3.4129,
+      "step": 759
+    },
+    {
+      "epoch": 0.1216,
+      "grad_norm": 3.018703420670456,
+      "learning_rate": 0.0001999432748339918,
+      "loss": 3.4862,
+      "step": 760
+    },
+    {
+      "epoch": 0.12176,
+      "grad_norm": 3.1149217937938696,
+      "learning_rate": 0.00019994269160134983,
+      "loss": 3.398,
+      "step": 761
+    },
+    {
+      "epoch": 0.12192,
+      "grad_norm": 2.9952160547958044,
+      "learning_rate": 0.00019994210538656356,
+      "loss": 3.5373,
+      "step": 762
+    },
+    {
+      "epoch": 0.12208,
+      "grad_norm": 3.3032821372467924,
+      "learning_rate": 0.00019994151618965051,
+      "loss": 3.5616,
+      "step": 763
+    },
+    {
+      "epoch": 0.12224,
+      "grad_norm": 3.860139099161452,
+      "learning_rate": 0.00019994092401062825,
+      "loss": 3.5076,
+      "step": 764
+    },
+    {
+      "epoch": 0.1224,
+      "grad_norm": 2.9282431168618226,
+      "learning_rate": 0.00019994032884951445,
+      "loss": 3.4392,
+      "step": 765
+    },
+    {
+      "epoch": 0.12256,
+      "grad_norm": 3.1782548666236234,
+      "learning_rate": 0.00019993973070632682,
+      "loss": 3.4855,
+      "step": 766
+    },
+    {
+      "epoch": 0.12272,
+      "grad_norm": 4.167294113269877,
+      "learning_rate": 0.0001999391295810833,
+      "loss": 3.4311,
+      "step": 767
+    },
+    {
+      "epoch": 0.12288,
+      "grad_norm": 3.387046425051251,
+      "learning_rate": 0.0001999385254738018,
+      "loss": 3.4311,
+      "step": 768
+    },
+    {
+      "epoch": 0.12304,
+      "grad_norm": 2.8181477524739287,
+      "learning_rate": 0.00019993791838450028,
+      "loss": 3.4827,
+      "step": 769
+    },
+    {
+      "epoch": 0.1232,
+      "grad_norm": 3.761498538248364,
+      "learning_rate": 0.00019993730831319693,
+      "loss": 3.4834,
+      "step": 770
+    },
+    {
+      "epoch": 0.12336,
+      "grad_norm": 3.8170072345959687,
+      "learning_rate": 0.0001999366952599099,
+      "loss": 3.3616,
+      "step": 771
+    },
+    {
+      "epoch": 0.12352,
+      "grad_norm": 3.093908545421489,
+      "learning_rate": 0.0001999360792246575,
+      "loss": 3.4696,
+      "step": 772
+    },
+    {
+      "epoch": 0.12368,
+      "grad_norm": 3.8225532358900103,
+      "learning_rate": 0.00019993546020745812,
+      "loss": 3.3674,
+      "step": 773
+    },
+    {
+      "epoch": 0.12384,
+      "grad_norm": 3.862553234518576,
+      "learning_rate": 0.00019993483820833027,
+      "loss": 3.5156,
+      "step": 774
+    },
+    {
+      "epoch": 0.124,
+      "grad_norm": 4.26290609492955,
+      "learning_rate": 0.00019993421322729243,
+      "loss": 3.3762,
+      "step": 775
+    },
+    {
+      "epoch": 0.12416,
+      "grad_norm": 4.755019869771094,
+      "learning_rate": 0.0001999335852643633,
+      "loss": 3.4178,
+      "step": 776
+    },
+    {
+      "epoch": 0.12432,
+      "grad_norm": 3.46685419804234,
+      "learning_rate": 0.0001999329543195616,
+      "loss": 3.428,
+      "step": 777
+    },
+    {
+      "epoch": 0.12448,
+      "grad_norm": 4.286473765048274,
+      "learning_rate": 0.00019993232039290616,
+      "loss": 3.4217,
+      "step": 778
+    },
+    {
+      "epoch": 0.12464,
+      "grad_norm": 4.288237475729105,
+      "learning_rate": 0.0001999316834844159,
+      "loss": 3.5206,
+      "step": 779
+    },
+    {
+      "epoch": 0.1248,
+      "grad_norm": 6.430962819565208,
+      "learning_rate": 0.00019993104359410983,
+      "loss": 3.4318,
+      "step": 780
+    },
+    {
+      "epoch": 0.12496,
+      "grad_norm": 3.033549815484248,
+      "learning_rate": 0.00019993040072200703,
+      "loss": 3.408,
+      "step": 781
+    },
+    {
+      "epoch": 0.12512,
+      "grad_norm": 4.51541805236547,
+      "learning_rate": 0.00019992975486812667,
+      "loss": 3.3756,
+      "step": 782
+    },
+    {
+      "epoch": 0.12528,
+      "grad_norm": 3.191434390800921,
+      "learning_rate": 0.00019992910603248808,
+      "loss": 3.4266,
+      "step": 783
+    },
+    {
+      "epoch": 0.12544,
+      "grad_norm": 3.4433979338165552,
+      "learning_rate": 0.00019992845421511052,
+      "loss": 3.3899,
+      "step": 784
+    },
+    {
+      "epoch": 0.1256,
+      "grad_norm": 2.9102272290856175,
+      "learning_rate": 0.00019992779941601353,
+      "loss": 3.4305,
+      "step": 785
+    },
+    {
+      "epoch": 0.12576,
+      "grad_norm": 3.5673603159750953,
+      "learning_rate": 0.0001999271416352166,
+      "loss": 3.3562,
+      "step": 786
+    },
+    {
+      "epoch": 0.12592,
+      "grad_norm": 3.9940367140841424,
+      "learning_rate": 0.00019992648087273937,
+      "loss": 3.5011,
+      "step": 787
+    },
+    {
+      "epoch": 0.12608,
+      "grad_norm": 3.3902379867394314,
+      "learning_rate": 0.0001999258171286016,
+      "loss": 3.4305,
+      "step": 788
+    },
+    {
+      "epoch": 0.12624,
+      "grad_norm": 3.1526706237959106,
+      "learning_rate": 0.000199925150402823,
+      "loss": 3.4954,
+      "step": 789
+    },
+    {
+      "epoch": 0.1264,
+      "grad_norm": 4.2481667655710575,
+      "learning_rate": 0.00019992448069542355,
+      "loss": 3.3826,
+      "step": 790
+    },
+    {
+      "epoch": 0.12656,
+      "grad_norm": 4.202407532440794,
+      "learning_rate": 0.00019992380800642318,
+      "loss": 3.1733,
+      "step": 791
+    },
+    {
+      "epoch": 0.12672,
+      "grad_norm": 3.092484008250104,
+      "learning_rate": 0.000199923132335842,
+      "loss": 3.4393,
+      "step": 792
+    },
+    {
+      "epoch": 0.12688,
+      "grad_norm": 3.2188604776278646,
+      "learning_rate": 0.00019992245368370017,
+      "loss": 3.3171,
+      "step": 793
+    },
+    {
+      "epoch": 0.12704,
+      "grad_norm": 3.142851562840543,
+      "learning_rate": 0.0001999217720500179,
+      "loss": 3.3276,
+      "step": 794
+    },
+    {
+      "epoch": 0.1272,
+      "grad_norm": 3.932721398601908,
+      "learning_rate": 0.00019992108743481556,
+      "loss": 3.3769,
+      "step": 795
+    },
+    {
+      "epoch": 0.12736,
+      "grad_norm": 3.5562456343283104,
+      "learning_rate": 0.00019992039983811357,
+      "loss": 3.3756,
+      "step": 796
+    },
+    {
+      "epoch": 0.12752,
+      "grad_norm": 3.524935249059301,
+      "learning_rate": 0.00019991970925993245,
+      "loss": 3.318,
+      "step": 797
+    },
+    {
+      "epoch": 0.12768,
+      "grad_norm": 4.654801745406681,
+      "learning_rate": 0.0001999190157002928,
+      "loss": 3.2279,
+      "step": 798
+    },
+    {
+      "epoch": 0.12784,
+      "grad_norm": 3.7959375113100884,
+      "learning_rate": 0.00019991831915921532,
+      "loss": 3.4047,
+      "step": 799
+    },
+    {
+      "epoch": 0.128,
+      "grad_norm": 3.978215169727504,
+      "learning_rate": 0.00019991761963672082,
+      "loss": 3.3668,
+      "step": 800
+    },
+    {
+      "epoch": 0.12816,
+      "grad_norm": 4.028626938160699,
+      "learning_rate": 0.00019991691713283012,
+      "loss": 3.3873,
+      "step": 801
+    },
+    {
+      "epoch": 0.12832,
+      "grad_norm": 5.461263942409574,
+      "learning_rate": 0.00019991621164756424,
+      "loss": 3.352,
+      "step": 802
+    },
+    {
+      "epoch": 0.12848,
+      "grad_norm": 3.9024580764777204,
+      "learning_rate": 0.00019991550318094415,
+      "loss": 3.4361,
+      "step": 803
+    },
+    {
+      "epoch": 0.12864,
+      "grad_norm": 4.774661503752779,
+      "learning_rate": 0.00019991479173299103,
+      "loss": 3.29,
+      "step": 804
+    },
+    {
+      "epoch": 0.1288,
+      "grad_norm": 4.143136844924149,
+      "learning_rate": 0.00019991407730372615,
+      "loss": 3.3563,
+      "step": 805
+    },
+    {
+      "epoch": 0.12896,
+      "grad_norm": 3.369047506400054,
+      "learning_rate": 0.0001999133598931708,
+      "loss": 3.289,
+      "step": 806
+    },
+    {
+      "epoch": 0.12912,
+      "grad_norm": 4.707898886242847,
+      "learning_rate": 0.00019991263950134633,
+      "loss": 3.2078,
+      "step": 807
+    },
+    {
+      "epoch": 0.12928,
+      "grad_norm": 3.516015631593528,
+      "learning_rate": 0.00019991191612827433,
+      "loss": 3.5227,
+      "step": 808
+    },
+    {
+      "epoch": 0.12944,
+      "grad_norm": 3.2967129904889685,
+      "learning_rate": 0.00019991118977397634,
+      "loss": 3.2817,
+      "step": 809
+    },
+    {
+      "epoch": 0.1296,
+      "grad_norm": 3.4179324507829394,
+      "learning_rate": 0.00019991046043847403,
+      "loss": 3.2338,
+      "step": 810
+    },
+    {
+      "epoch": 0.12976,
+      "grad_norm": 3.6898966150239394,
+      "learning_rate": 0.00019990972812178914,
+      "loss": 3.4255,
+      "step": 811
+    },
+    {
+      "epoch": 0.12992,
+      "grad_norm": 3.1085203517941493,
+      "learning_rate": 0.00019990899282394358,
+      "loss": 3.2805,
+      "step": 812
+    },
+    {
+      "epoch": 0.13008,
+      "grad_norm": 3.5967082954471,
+      "learning_rate": 0.00019990825454495922,
+      "loss": 3.3497,
+      "step": 813
+    },
+    {
+      "epoch": 0.13024,
+      "grad_norm": 3.6172681122918444,
+      "learning_rate": 0.00019990751328485815,
+      "loss": 3.3276,
+      "step": 814
+    },
+    {
+      "epoch": 0.1304,
+      "grad_norm": 2.795799882865931,
+      "learning_rate": 0.00019990676904366246,
+      "loss": 3.3303,
+      "step": 815
+    },
+    {
+      "epoch": 0.13056,
+      "grad_norm": 3.7001816027456025,
+      "learning_rate": 0.00019990602182139435,
+      "loss": 3.3094,
+      "step": 816
+    },
+    {
+      "epoch": 0.13072,
+      "grad_norm": 2.671384500053062,
+      "learning_rate": 0.00019990527161807618,
+      "loss": 3.3585,
+      "step": 817
+    },
+    {
+      "epoch": 0.13088,
+      "grad_norm": 2.9553766524499547,
+      "learning_rate": 0.0001999045184337302,
+      "loss": 3.2521,
+      "step": 818
+    },
+    {
+      "epoch": 0.13104,
+      "grad_norm": 3.0917260372310333,
+      "learning_rate": 0.00019990376226837903,
+      "loss": 3.3179,
+      "step": 819
+    },
+    {
+      "epoch": 0.1312,
+      "grad_norm": 3.299233757380845,
+      "learning_rate": 0.00019990300312204516,
+      "loss": 3.1443,
+      "step": 820
+    },
+    {
+      "epoch": 0.13136,
+      "grad_norm": 3.28392764976188,
+      "learning_rate": 0.0001999022409947512,
+      "loss": 3.2565,
+      "step": 821
+    },
+    {
+      "epoch": 0.13152,
+      "grad_norm": 3.505028236113936,
+      "learning_rate": 0.00019990147588652,
+      "loss": 3.2925,
+      "step": 822
+    },
+    {
+      "epoch": 0.13168,
+      "grad_norm": 2.585986911688406,
+      "learning_rate": 0.00019990070779737432,
+      "loss": 3.2501,
+      "step": 823
+    },
+    {
+      "epoch": 0.13184,
+      "grad_norm": 3.8229375213905303,
+      "learning_rate": 0.0001998999367273371,
+      "loss": 3.2681,
+      "step": 824
+    },
+    {
+      "epoch": 0.132,
+      "grad_norm": 2.8002566870689507,
+      "learning_rate": 0.0001998991626764313,
+      "loss": 3.291,
+      "step": 825
+    },
+    {
+      "epoch": 0.13216,
+      "grad_norm": 4.7958763489702925,
+      "learning_rate": 0.00019989838564468007,
+      "loss": 3.3358,
+      "step": 826
+    },
+    {
+      "epoch": 0.13232,
+      "grad_norm": 2.766379815708628,
+      "learning_rate": 0.00019989760563210657,
+      "loss": 3.1682,
+      "step": 827
+    },
+    {
+      "epoch": 0.13248,
+      "grad_norm": 3.852318449617603,
+      "learning_rate": 0.0001998968226387341,
+      "loss": 3.185,
+      "step": 828
+    },
+    {
+      "epoch": 0.13264,
+      "grad_norm": 3.6846639473794003,
+      "learning_rate": 0.000199896036664586,
+      "loss": 3.2882,
+      "step": 829
+    },
+    {
+      "epoch": 0.1328,
+      "grad_norm": 3.2961518044219353,
+      "learning_rate": 0.00019989524770968573,
+      "loss": 3.194,
+      "step": 830
+    },
+    {
+      "epoch": 0.13296,
+      "grad_norm": 2.7932388163631177,
+      "learning_rate": 0.00019989445577405686,
+      "loss": 3.2869,
+      "step": 831
+    },
+    {
+      "epoch": 0.13312,
+      "grad_norm": 3.1645491355691004,
+      "learning_rate": 0.00019989366085772296,
+      "loss": 3.2803,
+      "step": 832
+    },
+    {
+      "epoch": 0.13328,
+      "grad_norm": 3.7201546488146655,
+      "learning_rate": 0.0001998928629607078,
+      "loss": 3.3361,
+      "step": 833
+    },
+    {
+      "epoch": 0.13344,
+      "grad_norm": 3.994568300317842,
+      "learning_rate": 0.00019989206208303513,
+      "loss": 3.2306,
+      "step": 834
+    },
+    {
+      "epoch": 0.1336,
+      "grad_norm": 2.59387551283259,
+      "learning_rate": 0.00019989125822472893,
+      "loss": 3.2337,
+      "step": 835
+    },
+    {
+      "epoch": 0.13376,
+      "grad_norm": 3.170423752909648,
+      "learning_rate": 0.00019989045138581313,
+      "loss": 3.2538,
+      "step": 836
+    },
+    {
+      "epoch": 0.13392,
+      "grad_norm": 3.100753400740744,
+      "learning_rate": 0.0001998896415663118,
+      "loss": 3.208,
+      "step": 837
+    },
+    {
+      "epoch": 0.13408,
+      "grad_norm": 3.1049212196305818,
+      "learning_rate": 0.00019988882876624913,
+      "loss": 3.2992,
+      "step": 838
+    },
+    {
+      "epoch": 0.13424,
+      "grad_norm": 3.3140329529536263,
+      "learning_rate": 0.00019988801298564935,
+      "loss": 3.261,
+      "step": 839
+    },
+    {
+      "epoch": 0.1344,
+      "grad_norm": 3.3377483381293827,
+      "learning_rate": 0.00019988719422453684,
+      "loss": 3.1941,
+      "step": 840
+    },
+    {
+      "epoch": 0.13456,
+      "grad_norm": 4.578827269954963,
+      "learning_rate": 0.000199886372482936,
+      "loss": 3.2682,
+      "step": 841
+    },
+    {
+      "epoch": 0.13472,
+      "grad_norm": 4.139027971807797,
+      "learning_rate": 0.00019988554776087134,
+      "loss": 3.0884,
+      "step": 842
+    },
+    {
+      "epoch": 0.13488,
+      "grad_norm": 3.6113720163816097,
+      "learning_rate": 0.00019988472005836748,
+      "loss": 3.1935,
+      "step": 843
+    },
+    {
+      "epoch": 0.13504,
+      "grad_norm": 4.922751701229775,
+      "learning_rate": 0.00019988388937544914,
+      "loss": 3.2881,
+      "step": 844
+    },
+    {
+      "epoch": 0.1352,
+      "grad_norm": 4.047246703758418,
+      "learning_rate": 0.00019988305571214107,
+      "loss": 3.2878,
+      "step": 845
+    },
+    {
+      "epoch": 0.13536,
+      "grad_norm": 4.082425920957403,
+      "learning_rate": 0.00019988221906846817,
+      "loss": 3.3051,
+      "step": 846
+    },
+    {
+      "epoch": 0.13552,
+      "grad_norm": 4.312468327249615,
+      "learning_rate": 0.0001998813794444554,
+      "loss": 3.26,
+      "step": 847
+    },
+    {
+      "epoch": 0.13568,
+      "grad_norm": 3.7362515327760994,
+      "learning_rate": 0.0001998805368401278,
+      "loss": 3.222,
+      "step": 848
+    },
+    {
+      "epoch": 0.13584,
+      "grad_norm": 4.121883968211744,
+      "learning_rate": 0.00019987969125551054,
+      "loss": 3.3123,
+      "step": 849
+    },
+    {
+      "epoch": 0.136,
+      "grad_norm": 4.667774375805058,
+      "learning_rate": 0.00019987884269062882,
+      "loss": 3.2174,
+      "step": 850
+    },
+    {
+      "epoch": 0.13616,
+      "grad_norm": 3.455674123682015,
+      "learning_rate": 0.00019987799114550798,
+      "loss": 3.1527,
+      "step": 851
+    },
+    {
+      "epoch": 0.13632,
+      "grad_norm": 3.6438953033851322,
+      "learning_rate": 0.00019987713662017342,
+      "loss": 3.1333,
+      "step": 852
+    },
+    {
+      "epoch": 0.13648,
+      "grad_norm": 3.077738540560717,
+      "learning_rate": 0.00019987627911465064,
+      "loss": 3.2328,
+      "step": 853
+    },
+    {
+      "epoch": 0.13664,
+      "grad_norm": 3.276544609947457,
+      "learning_rate": 0.00019987541862896523,
+      "loss": 3.2746,
+      "step": 854
+    },
+    {
+      "epoch": 0.1368,
+      "grad_norm": 3.394869680065702,
+      "learning_rate": 0.00019987455516314284,
+      "loss": 3.2031,
+      "step": 855
+    },
+    {
+      "epoch": 0.13696,
+      "grad_norm": 3.0347639040502314,
+      "learning_rate": 0.00019987368871720928,
+      "loss": 3.2139,
+      "step": 856
+    },
+    {
+      "epoch": 0.13712,
+      "grad_norm": 3.468165682027337,
+      "learning_rate": 0.00019987281929119036,
+      "loss": 3.2936,
+      "step": 857
+    },
+    {
+      "epoch": 0.13728,
+      "grad_norm": 3.030604888889384,
+      "learning_rate": 0.00019987194688511205,
+      "loss": 3.2146,
+      "step": 858
+    },
+    {
+      "epoch": 0.13744,
+      "grad_norm": 3.437798258422719,
+      "learning_rate": 0.0001998710714990004,
+      "loss": 3.1705,
+      "step": 859
+    },
+    {
+      "epoch": 0.1376,
+      "grad_norm": 2.9166549328456406,
+      "learning_rate": 0.00019987019313288146,
+      "loss": 3.1156,
+      "step": 860
+    },
+    {
+      "epoch": 0.13776,
+      "grad_norm": 3.984040307483887,
+      "learning_rate": 0.0001998693117867815,
+      "loss": 3.2481,
+      "step": 861
+    },
+    {
+      "epoch": 0.13792,
+      "grad_norm": 2.7473066991604913,
+      "learning_rate": 0.00019986842746072683,
+      "loss": 3.0315,
+      "step": 862
+    },
+    {
+      "epoch": 0.13808,
+      "grad_norm": 3.4956008807962275,
+      "learning_rate": 0.00019986754015474377,
+      "loss": 3.0818,
+      "step": 863
+    },
+    {
+      "epoch": 0.13824,
+      "grad_norm": 4.264692364695866,
+      "learning_rate": 0.00019986664986885888,
+      "loss": 3.2254,
+      "step": 864
+    },
+    {
+      "epoch": 0.1384,
+      "grad_norm": 4.03186175644819,
+      "learning_rate": 0.00019986575660309864,
+      "loss": 3.1418,
+      "step": 865
+    },
+    {
+      "epoch": 0.13856,
+      "grad_norm": 3.9045831251309626,
+      "learning_rate": 0.00019986486035748974,
+      "loss": 3.2019,
+      "step": 866
+    },
+    {
+      "epoch": 0.13872,
+      "grad_norm": 3.2818705918129427,
+      "learning_rate": 0.00019986396113205894,
+      "loss": 3.2285,
+      "step": 867
+    },
+    {
+      "epoch": 0.13888,
+      "grad_norm": 4.067872883935154,
+      "learning_rate": 0.00019986305892683305,
+      "loss": 3.1473,
+      "step": 868
+    },
+    {
+      "epoch": 0.13904,
+      "grad_norm": 3.2769148635341474,
+      "learning_rate": 0.000199862153741839,
+      "loss": 3.1641,
+      "step": 869
+    },
+    {
+      "epoch": 0.1392,
+      "grad_norm": 3.4749060692957623,
+      "learning_rate": 0.0001998612455771038,
+      "loss": 3.1622,
+      "step": 870
+    },
+    {
+      "epoch": 0.13936,
+      "grad_norm": 3.7627867753268305,
+      "learning_rate": 0.00019986033443265456,
+      "loss": 3.0727,
+      "step": 871
+    },
+    {
+      "epoch": 0.13952,
+      "grad_norm": 3.610775364924915,
+      "learning_rate": 0.00019985942030851842,
+      "loss": 3.2864,
+      "step": 872
+    },
+    {
+      "epoch": 0.13968,
+      "grad_norm": 7.9948367479334195,
+      "learning_rate": 0.0001998585032047227,
+      "loss": 3.2886,
+      "step": 873
+    },
+    {
+      "epoch": 0.13984,
+      "grad_norm": 3.9972203127178956,
+      "learning_rate": 0.00019985758312129477,
+      "loss": 3.068,
+      "step": 874
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 3.170562405347698,
+      "learning_rate": 0.00019985666005826205,
+      "loss": 3.1698,
+      "step": 875
+    },
+    {
+      "epoch": 0.14016,
+      "grad_norm": 2.7972101390443824,
+      "learning_rate": 0.0001998557340156521,
+      "loss": 3.2489,
+      "step": 876
+    },
+    {
+      "epoch": 0.14032,
+      "grad_norm": 3.684459290312373,
+      "learning_rate": 0.00019985480499349256,
+      "loss": 3.2909,
+      "step": 877
+    },
+    {
+      "epoch": 0.14048,
+      "grad_norm": 3.4193549725067856,
+      "learning_rate": 0.00019985387299181114,
+      "loss": 3.2049,
+      "step": 878
+    },
+    {
+      "epoch": 0.14064,
+      "grad_norm": 3.3035110335242295,
+      "learning_rate": 0.00019985293801063567,
+      "loss": 3.3059,
+      "step": 879
+    },
+    {
+      "epoch": 0.1408,
+      "grad_norm": 3.396243108765038,
+      "learning_rate": 0.000199852000049994,
+      "loss": 3.1686,
+      "step": 880
+    },
+    {
+      "epoch": 0.14096,
+      "grad_norm": 3.2958040591255306,
+      "learning_rate": 0.00019985105910991417,
+      "loss": 3.2462,
+      "step": 881
+    },
+    {
+      "epoch": 0.14112,
+      "grad_norm": 3.424174965696953,
+      "learning_rate": 0.00019985011519042424,
+      "loss": 3.1983,
+      "step": 882
+    },
+    {
+      "epoch": 0.14128,
+      "grad_norm": 3.011612462799523,
+      "learning_rate": 0.00019984916829155234,
+      "loss": 3.1787,
+      "step": 883
+    },
+    {
+      "epoch": 0.14144,
+      "grad_norm": 2.76806828598787,
+      "learning_rate": 0.00019984821841332676,
+      "loss": 3.1423,
+      "step": 884
+    },
+    {
+      "epoch": 0.1416,
+      "grad_norm": 3.0169604959656744,
+      "learning_rate": 0.00019984726555577584,
+      "loss": 3.1395,
+      "step": 885
+    },
+    {
+      "epoch": 0.14176,
+      "grad_norm": 3.4554659956336335,
+      "learning_rate": 0.00019984630971892805,
+      "loss": 3.1532,
+      "step": 886
+    },
+    {
+      "epoch": 0.14192,
+      "grad_norm": 2.842474053833977,
+      "learning_rate": 0.0001998453509028118,
+      "loss": 3.1144,
+      "step": 887
+    },
+    {
+      "epoch": 0.14208,
+      "grad_norm": 3.88291025836206,
+      "learning_rate": 0.00019984438910745583,
+      "loss": 3.119,
+      "step": 888
+    },
+    {
+      "epoch": 0.14224,
+      "grad_norm": 3.62822278181425,
+      "learning_rate": 0.00019984342433288874,
+      "loss": 3.1274,
+      "step": 889
+    },
+    {
+      "epoch": 0.1424,
+      "grad_norm": 4.010476906506429,
+      "learning_rate": 0.0001998424565791394,
+      "loss": 3.0172,
+      "step": 890
+    },
+    {
+      "epoch": 0.14256,
+      "grad_norm": 3.5419119059207445,
+      "learning_rate": 0.00019984148584623656,
+      "loss": 3.2717,
+      "step": 891
+    },
+    {
+      "epoch": 0.14272,
+      "grad_norm": 3.655361760675255,
+      "learning_rate": 0.00019984051213420935,
+      "loss": 3.2156,
+      "step": 892
+    },
+    {
+      "epoch": 0.14288,
+      "grad_norm": 3.300394116467381,
+      "learning_rate": 0.0001998395354430867,
+      "loss": 3.112,
+      "step": 893
+    },
+    {
+      "epoch": 0.14304,
+      "grad_norm": 3.9399924249840685,
+      "learning_rate": 0.0001998385557728978,
+      "loss": 3.0174,
+      "step": 894
+    },
+    {
+      "epoch": 0.1432,
+      "grad_norm": 3.4958757336697563,
+      "learning_rate": 0.00019983757312367186,
+      "loss": 3.272,
+      "step": 895
+    },
+    {
+      "epoch": 0.14336,
+      "grad_norm": 4.717464960559099,
+      "learning_rate": 0.00019983658749543826,
+      "loss": 3.2567,
+      "step": 896
+    },
+    {
+      "epoch": 0.14352,
+      "grad_norm": 3.24141726400309,
+      "learning_rate": 0.00019983559888822632,
+      "loss": 3.0601,
+      "step": 897
+    },
+    {
+      "epoch": 0.14368,
+      "grad_norm": 3.7797848214907117,
+      "learning_rate": 0.0001998346073020656,
+      "loss": 3.1979,
+      "step": 898
+    },
+    {
+      "epoch": 0.14384,
+      "grad_norm": 3.637457328394703,
+      "learning_rate": 0.00019983361273698567,
+      "loss": 3.1625,
+      "step": 899
+    },
+    {
+      "epoch": 0.144,
+      "grad_norm": 3.22449775622261,
+      "learning_rate": 0.00019983261519301623,
+      "loss": 2.9589,
+      "step": 900
+    },
+    {
+      "epoch": 0.14416,
+      "grad_norm": 3.0328430219670524,
+      "learning_rate": 0.00019983161467018701,
+      "loss": 3.0591,
+      "step": 901
+    },
+    {
+      "epoch": 0.14432,
+      "grad_norm": 3.5908504080671855,
+      "learning_rate": 0.00019983061116852789,
+      "loss": 3.1447,
+      "step": 902
+    },
+    {
+      "epoch": 0.14448,
+      "grad_norm": 2.885151750190397,
+      "learning_rate": 0.00019982960468806877,
+      "loss": 3.1384,
+      "step": 903
+    },
+    {
+      "epoch": 0.14464,
+      "grad_norm": 4.509720883141306,
+      "learning_rate": 0.00019982859522883973,
+      "loss": 3.1309,
+      "step": 904
+    },
+    {
+      "epoch": 0.1448,
+      "grad_norm": 3.147931008650893,
+      "learning_rate": 0.00019982758279087087,
+      "loss": 3.0285,
+      "step": 905
+    },
+    {
+      "epoch": 0.14496,
+      "grad_norm": 3.023028061948765,
+      "learning_rate": 0.0001998265673741924,
+      "loss": 3.1919,
+      "step": 906
+    },
+    {
+      "epoch": 0.14512,
+      "grad_norm": 4.435774610706274,
+      "learning_rate": 0.00019982554897883466,
+      "loss": 3.1077,
+      "step": 907
+    },
+    {
+      "epoch": 0.14528,
+      "grad_norm": 3.693355340321252,
+      "learning_rate": 0.00019982452760482797,
+      "loss": 3.051,
+      "step": 908
+    },
+    {
+      "epoch": 0.14544,
+      "grad_norm": 3.368116449268123,
+      "learning_rate": 0.00019982350325220284,
+      "loss": 3.0899,
+      "step": 909
+    },
+    {
+      "epoch": 0.1456,
+      "grad_norm": 2.8464574798478743,
+      "learning_rate": 0.00019982247592098982,
+      "loss": 3.0591,
+      "step": 910
+    },
+    {
+      "epoch": 0.14576,
+      "grad_norm": 3.4832186655699946,
+      "learning_rate": 0.0001998214456112196,
+      "loss": 3.0361,
+      "step": 911
+    },
+    {
+      "epoch": 0.14592,
+      "grad_norm": 4.184669679037268,
+      "learning_rate": 0.00019982041232292287,
+      "loss": 3.002,
+      "step": 912
+    },
+    {
+      "epoch": 0.14608,
+      "grad_norm": 3.887862640442508,
+      "learning_rate": 0.0001998193760561305,
+      "loss": 3.0358,
+      "step": 913
+    },
+    {
+      "epoch": 0.14624,
+      "grad_norm": 2.885376722572392,
+      "learning_rate": 0.0001998183368108734,
+      "loss": 3.0254,
+      "step": 914
+    },
+    {
+      "epoch": 0.1464,
+      "grad_norm": 3.2343819790628947,
+      "learning_rate": 0.00019981729458718258,
+      "loss": 3.0165,
+      "step": 915
+    },
+    {
+      "epoch": 0.14656,
+      "grad_norm": 2.7229005047572308,
+      "learning_rate": 0.00019981624938508917,
+      "loss": 3.0001,
+      "step": 916
+    },
+    {
+      "epoch": 0.14672,
+      "grad_norm": 3.4843496762866155,
+      "learning_rate": 0.0001998152012046243,
+      "loss": 3.1149,
+      "step": 917
+    },
+    {
+      "epoch": 0.14688,
+      "grad_norm": 3.7381319641189132,
+      "learning_rate": 0.00019981415004581926,
+      "loss": 2.9564,
+      "step": 918
+    },
+    {
+      "epoch": 0.14704,
+      "grad_norm": 3.817594616799295,
+      "learning_rate": 0.00019981309590870542,
+      "loss": 2.9902,
+      "step": 919
+    },
+    {
+      "epoch": 0.1472,
+      "grad_norm": 3.899873381173169,
+      "learning_rate": 0.00019981203879331427,
+      "loss": 2.9623,
+      "step": 920
+    },
+    {
+      "epoch": 0.14736,
+      "grad_norm": 3.2027818212485384,
+      "learning_rate": 0.00019981097869967732,
+      "loss": 3.1045,
+      "step": 921
+    },
+    {
+      "epoch": 0.14752,
+      "grad_norm": 3.429589080864773,
+      "learning_rate": 0.00019980991562782617,
+      "loss": 3.2094,
+      "step": 922
+    },
+    {
+      "epoch": 0.14768,
+      "grad_norm": 3.73334062282321,
+      "learning_rate": 0.00019980884957779263,
+      "loss": 3.0777,
+      "step": 923
+    },
+    {
+      "epoch": 0.14784,
+      "grad_norm": 2.7419167269499467,
+      "learning_rate": 0.00019980778054960843,
+      "loss": 3.0558,
+      "step": 924
+    },
+    {
+      "epoch": 0.148,
+      "grad_norm": 3.6543803666142725,
+      "learning_rate": 0.00019980670854330548,
+      "loss": 3.0366,
+      "step": 925
+    },
+    {
+      "epoch": 0.14816,
+      "grad_norm": 3.053376249581391,
+      "learning_rate": 0.00019980563355891578,
+      "loss": 3.0056,
+      "step": 926
+    },
+    {
+      "epoch": 0.14832,
+      "grad_norm": 3.9490953151644748,
+      "learning_rate": 0.00019980455559647143,
+      "loss": 3.2747,
+      "step": 927
+    },
+    {
+      "epoch": 0.14848,
+      "grad_norm": 2.884134267823431,
+      "learning_rate": 0.00019980347465600458,
+      "loss": 3.1786,
+      "step": 928
+    },
+    {
+      "epoch": 0.14864,
+      "grad_norm": 4.062896700893862,
+      "learning_rate": 0.00019980239073754744,
+      "loss": 3.1071,
+      "step": 929
+    },
+    {
+      "epoch": 0.1488,
+      "grad_norm": 5.04727267976226,
+      "learning_rate": 0.00019980130384113242,
+      "loss": 3.1085,
+      "step": 930
+    },
+    {
+      "epoch": 0.14896,
+      "grad_norm": 3.009044131555114,
+      "learning_rate": 0.0001998002139667919,
+      "loss": 3.1376,
+      "step": 931
+    },
+    {
+      "epoch": 0.14912,
+      "grad_norm": 4.5547305474365425,
+      "learning_rate": 0.00019979912111455842,
+      "loss": 3.047,
+      "step": 932
+    },
+    {
+      "epoch": 0.14928,
+      "grad_norm": 3.9561988829123793,
+      "learning_rate": 0.0001997980252844646,
+      "loss": 2.9387,
+      "step": 933
+    },
+    {
+      "epoch": 0.14944,
+      "grad_norm": 3.201341677838058,
+      "learning_rate": 0.00019979692647654313,
+      "loss": 3.0328,
+      "step": 934
+    },
+    {
+      "epoch": 0.1496,
+      "grad_norm": 3.9160226209142794,
+      "learning_rate": 0.00019979582469082678,
+      "loss": 3.0434,
+      "step": 935
+    },
+    {
+      "epoch": 0.14976,
+      "grad_norm": 3.2596521052452783,
+      "learning_rate": 0.00019979471992734844,
+      "loss": 3.0767,
+      "step": 936
+    },
+    {
+      "epoch": 0.14992,
+      "grad_norm": 3.624238117531271,
+      "learning_rate": 0.00019979361218614108,
+      "loss": 2.9499,
+      "step": 937
+    },
+    {
+      "epoch": 0.15008,
+      "grad_norm": 3.1342462489324987,
+      "learning_rate": 0.00019979250146723773,
+      "loss": 3.1372,
+      "step": 938
+    },
+    {
+      "epoch": 0.15024,
+      "grad_norm": 4.199799376644347,
+      "learning_rate": 0.00019979138777067156,
+      "loss": 3.149,
+      "step": 939
+    },
+    {
+      "epoch": 0.1504,
+      "grad_norm": 3.87406937256061,
+      "learning_rate": 0.00019979027109647582,
+      "loss": 3.0901,
+      "step": 940
+    },
+    {
+      "epoch": 0.15056,
+      "grad_norm": 3.553707191758044,
+      "learning_rate": 0.00019978915144468375,
+      "loss": 3.0733,
+      "step": 941
+    },
+    {
+      "epoch": 0.15072,
+      "grad_norm": 4.016257109226128,
+      "learning_rate": 0.00019978802881532884,
+      "loss": 3.069,
+      "step": 942
+    },
+    {
+      "epoch": 0.15088,
+      "grad_norm": 3.668838074482338,
+      "learning_rate": 0.00019978690320844456,
+      "loss": 3.0565,
+      "step": 943
+    },
+    {
+      "epoch": 0.15104,
+      "grad_norm": 3.4706999458269983,
+      "learning_rate": 0.00019978577462406448,
+      "loss": 2.9887,
+      "step": 944
+    },
+    {
+      "epoch": 0.1512,
+      "grad_norm": 3.292920551549724,
+      "learning_rate": 0.00019978464306222228,
+      "loss": 2.9999,
+      "step": 945
+    },
+    {
+      "epoch": 0.15136,
+      "grad_norm": 3.508809210354875,
+      "learning_rate": 0.00019978350852295176,
+      "loss": 3.0529,
+      "step": 946
+    },
+    {
+      "epoch": 0.15152,
+      "grad_norm": 2.965520700045978,
+      "learning_rate": 0.00019978237100628672,
+      "loss": 2.9457,
+      "step": 947
+    },
+    {
+      "epoch": 0.15168,
+      "grad_norm": 3.203836655273305,
+      "learning_rate": 0.00019978123051226114,
+      "loss": 2.9406,
+      "step": 948
+    },
+    {
+      "epoch": 0.15184,
+      "grad_norm": 3.048087201242197,
+      "learning_rate": 0.00019978008704090903,
+      "loss": 2.9238,
+      "step": 949
+    },
+    {
+      "epoch": 0.152,
+      "grad_norm": 3.562051195026566,
+      "learning_rate": 0.00019977894059226451,
+      "loss": 3.0571,
+      "step": 950
+    },
+    {
+      "epoch": 0.15216,
+      "grad_norm": 4.167737102009381,
+      "learning_rate": 0.0001997777911663618,
+      "loss": 2.979,
+      "step": 951
+    },
+    {
+      "epoch": 0.15232,
+      "grad_norm": 3.150970475610013,
+      "learning_rate": 0.0001997766387632352,
+      "loss": 3.0028,
+      "step": 952
+    },
+    {
+      "epoch": 0.15248,
+      "grad_norm": 2.8577371527940003,
+      "learning_rate": 0.0001997754833829191,
+      "loss": 2.9844,
+      "step": 953
+    },
+    {
+      "epoch": 0.15264,
+      "grad_norm": 2.8023662244975567,
+      "learning_rate": 0.00019977432502544796,
+      "loss": 2.9619,
+      "step": 954
+    },
+    {
+      "epoch": 0.1528,
+      "grad_norm": 3.049077745715692,
+      "learning_rate": 0.00019977316369085632,
+      "loss": 2.9484,
+      "step": 955
+    },
+    {
+      "epoch": 0.15296,
+      "grad_norm": 3.113580439143067,
+      "learning_rate": 0.0001997719993791789,
+      "loss": 3.1001,
+      "step": 956
+    },
+    {
+      "epoch": 0.15312,
+      "grad_norm": 3.2000142189161727,
+      "learning_rate": 0.00019977083209045038,
+      "loss": 2.972,
+      "step": 957
+    },
+    {
+      "epoch": 0.15328,
+      "grad_norm": 3.0466935912636752,
+      "learning_rate": 0.0001997696618247056,
+      "loss": 3.0507,
+      "step": 958
+    },
+    {
+      "epoch": 0.15344,
+      "grad_norm": 3.313302391428803,
+      "learning_rate": 0.00019976848858197955,
+      "loss": 2.9265,
+      "step": 959
+    },
+    {
+      "epoch": 0.1536,
+      "grad_norm": 3.2514766254876113,
+      "learning_rate": 0.00019976731236230714,
+      "loss": 2.9741,
+      "step": 960
+    },
+    {
+      "epoch": 0.15376,
+      "grad_norm": 3.9687684685383444,
+      "learning_rate": 0.00019976613316572347,
+      "loss": 2.9483,
+      "step": 961
+    },
+    {
+      "epoch": 0.15392,
+      "grad_norm": 3.4396596126626395,
+      "learning_rate": 0.0001997649509922638,
+      "loss": 3.0424,
+      "step": 962
+    },
+    {
+      "epoch": 0.15408,
+      "grad_norm": 2.984613290392909,
+      "learning_rate": 0.00019976376584196338,
+      "loss": 2.9373,
+      "step": 963
+    },
+    {
+      "epoch": 0.15424,
+      "grad_norm": 3.3420132603754826,
+      "learning_rate": 0.0001997625777148575,
+      "loss": 2.9889,
+      "step": 964
+    },
+    {
+      "epoch": 0.1544,
+      "grad_norm": 2.727699188441547,
+      "learning_rate": 0.00019976138661098171,
+      "loss": 2.8732,
+      "step": 965
+    },
+    {
+      "epoch": 0.15456,
+      "grad_norm": 2.422215583454039,
+      "learning_rate": 0.0001997601925303715,
+      "loss": 2.9027,
+      "step": 966
+    },
+    {
+      "epoch": 0.15472,
+      "grad_norm": 4.20587986805219,
+      "learning_rate": 0.00019975899547306254,
+      "loss": 2.9926,
+      "step": 967
+    },
+    {
+      "epoch": 0.15488,
+      "grad_norm": 3.057164385573266,
+      "learning_rate": 0.00019975779543909047,
+      "loss": 2.9248,
+      "step": 968
+    },
+    {
+      "epoch": 0.15504,
+      "grad_norm": 3.5869523769334717,
+      "learning_rate": 0.0001997565924284912,
+      "loss": 2.9594,
+      "step": 969
+    },
+    {
+      "epoch": 0.1552,
+      "grad_norm": 3.5755182548147055,
+      "learning_rate": 0.00019975538644130054,
+      "loss": 3.0698,
+      "step": 970
+    },
+    {
+      "epoch": 0.15536,
+      "grad_norm": 3.269951639635496,
+      "learning_rate": 0.00019975417747755453,
+      "loss": 2.9992,
+      "step": 971
+    },
+    {
+      "epoch": 0.15552,
+      "grad_norm": 2.797185444052257,
+      "learning_rate": 0.00019975296553728918,
+      "loss": 2.9942,
+      "step": 972
+    },
+    {
+      "epoch": 0.15568,
+      "grad_norm": 3.3588349801161717,
+      "learning_rate": 0.00019975175062054073,
+      "loss": 2.9494,
+      "step": 973
+    },
+    {
+      "epoch": 0.15584,
+      "grad_norm": 3.1483060761885513,
+      "learning_rate": 0.0001997505327273454,
+      "loss": 2.9945,
+      "step": 974
+    },
+    {
+      "epoch": 0.156,
+      "grad_norm": 3.0208164843464136,
+      "learning_rate": 0.00019974931185773952,
+      "loss": 2.9187,
+      "step": 975
+    },
+    {
+      "epoch": 0.15616,
+      "grad_norm": 3.602720604787198,
+      "learning_rate": 0.00019974808801175947,
+      "loss": 2.8978,
+      "step": 976
+    },
+    {
+      "epoch": 0.15632,
+      "grad_norm": 3.5261433782112204,
+      "learning_rate": 0.00019974686118944188,
+      "loss": 3.0293,
+      "step": 977
+    },
+    {
+      "epoch": 0.15648,
+      "grad_norm": 2.9992730138696047,
+      "learning_rate": 0.00019974563139082328,
+      "loss": 2.9051,
+      "step": 978
+    },
+    {
+      "epoch": 0.15664,
+      "grad_norm": 3.9762515667720497,
+      "learning_rate": 0.0001997443986159404,
+      "loss": 2.9609,
+      "step": 979
+    },
+    {
+      "epoch": 0.1568,
+      "grad_norm": 3.730539663406,
+      "learning_rate": 0.00019974316286483,
+      "loss": 3.1169,
+      "step": 980
+    },
+    {
+      "epoch": 0.15696,
+      "grad_norm": 3.966462705532249,
+      "learning_rate": 0.00019974192413752895,
+      "loss": 2.8673,
+      "step": 981
+    },
+    {
+      "epoch": 0.15712,
+      "grad_norm": 3.766499714046797,
+      "learning_rate": 0.00019974068243407424,
+      "loss": 2.8703,
+      "step": 982
+    },
+    {
+      "epoch": 0.15728,
+      "grad_norm": 3.200775013675747,
+      "learning_rate": 0.00019973943775450291,
+      "loss": 2.869,
+      "step": 983
+    },
+    {
+      "epoch": 0.15744,
+      "grad_norm": 3.521358031680468,
+      "learning_rate": 0.00019973819009885206,
+      "loss": 2.8875,
+      "step": 984
+    },
+    {
+      "epoch": 0.1576,
+      "grad_norm": 3.1473828727928277,
+      "learning_rate": 0.00019973693946715896,
+      "loss": 2.8554,
+      "step": 985
+    },
+    {
+      "epoch": 0.15776,
+      "grad_norm": 4.025021624579343,
+      "learning_rate": 0.00019973568585946095,
+      "loss": 2.8564,
+      "step": 986
+    },
+    {
+      "epoch": 0.15792,
+      "grad_norm": 2.6680470596671246,
+      "learning_rate": 0.00019973442927579538,
+      "loss": 2.9144,
+      "step": 987
+    },
+    {
+      "epoch": 0.15808,
+      "grad_norm": 3.14842464709365,
+      "learning_rate": 0.00019973316971619976,
+      "loss": 2.9307,
+      "step": 988
+    },
+    {
+      "epoch": 0.15824,
+      "grad_norm": 3.2279682622487416,
+      "learning_rate": 0.00019973190718071168,
+      "loss": 2.9811,
+      "step": 989
+    },
+    {
+      "epoch": 0.1584,
+      "grad_norm": 3.6809473511621182,
+      "learning_rate": 0.00019973064166936886,
+      "loss": 2.999,
+      "step": 990
+    },
+    {
+      "epoch": 0.15856,
+      "grad_norm": 4.007470163021897,
+      "learning_rate": 0.00019972937318220896,
+      "loss": 2.796,
+      "step": 991
+    },
+    {
+      "epoch": 0.15872,
+      "grad_norm": 3.869430103926335,
+      "learning_rate": 0.00019972810171926993,
+      "loss": 2.9062,
+      "step": 992
+    },
+    {
+      "epoch": 0.15888,
+      "grad_norm": 3.4179120324401717,
+      "learning_rate": 0.00019972682728058963,
+      "loss": 2.9389,
+      "step": 993
+    },
+    {
+      "epoch": 0.15904,
+      "grad_norm": 3.1209520597173532,
+      "learning_rate": 0.00019972554986620616,
+      "loss": 2.9678,
+      "step": 994
+    },
+    {
+      "epoch": 0.1592,
+      "grad_norm": 3.302949750383646,
+      "learning_rate": 0.00019972426947615753,
+      "loss": 2.8254,
+      "step": 995
+    },
+    {
+      "epoch": 0.15936,
+      "grad_norm": 3.0289215274851786,
+      "learning_rate": 0.00019972298611048209,
+      "loss": 2.8418,
+      "step": 996
+    },
+    {
+      "epoch": 0.15952,
+      "grad_norm": 2.8866170269532154,
+      "learning_rate": 0.000199721699769218,
+      "loss": 2.9043,
+      "step": 997
+    },
+    {
+      "epoch": 0.15968,
+      "grad_norm": 3.3539078551785244,
+      "learning_rate": 0.00019972041045240372,
+      "loss": 2.8692,
+      "step": 998
+    },
+    {
+      "epoch": 0.15984,
+      "grad_norm": 3.357647884721163,
+      "learning_rate": 0.0001997191181600777,
+      "loss": 2.8328,
+      "step": 999
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 3.8336429150575997,
+      "learning_rate": 0.00019971782289227851,
+      "loss": 2.9863,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16016,
+      "grad_norm": 3.2112188419916876,
+      "learning_rate": 0.00019971652464904476,
+      "loss": 2.7203,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16032,
+      "grad_norm": 3.1447045123749535,
+      "learning_rate": 0.00019971522343041525,
+      "loss": 2.8598,
+      "step": 1002
+    },
+    {
+      "epoch": 0.16048,
+      "grad_norm": 3.0225221854183806,
+      "learning_rate": 0.00019971391923642875,
+      "loss": 2.8737,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16064,
+      "grad_norm": 3.3451224961491777,
+      "learning_rate": 0.00019971261206712423,
+      "loss": 2.9765,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1608,
+      "grad_norm": 3.3096385974022,
+      "learning_rate": 0.0001997113019225406,
+      "loss": 2.9661,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16096,
+      "grad_norm": 3.0983712821299916,
+      "learning_rate": 0.00019970998880271707,
+      "loss": 2.996,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16112,
+      "grad_norm": 3.589317114484597,
+      "learning_rate": 0.00019970867270769277,
+      "loss": 2.9729,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16128,
+      "grad_norm": 3.842061431288389,
+      "learning_rate": 0.00019970735363750692,
+      "loss": 2.8866,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16144,
+      "grad_norm": 2.81102163597488,
+      "learning_rate": 0.00019970603159219898,
+      "loss": 2.8699,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1616,
+      "grad_norm": 2.8531862578544724,
+      "learning_rate": 0.00019970470657180832,
+      "loss": 2.8295,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16176,
+      "grad_norm": 3.6934514096973428,
+      "learning_rate": 0.00019970337857637448,
+      "loss": 2.7821,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16192,
+      "grad_norm": 3.359142807106747,
+      "learning_rate": 0.00019970204760593713,
+      "loss": 2.7527,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16208,
+      "grad_norm": 2.8625348769541357,
+      "learning_rate": 0.00019970071366053594,
+      "loss": 2.8693,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16224,
+      "grad_norm": 3.6554073130999307,
+      "learning_rate": 0.00019969937674021076,
+      "loss": 2.8461,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1624,
+      "grad_norm": 3.773491015151043,
+      "learning_rate": 0.00019969803684500147,
+      "loss": 2.9215,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16256,
+      "grad_norm": 3.2150942862255376,
+      "learning_rate": 0.000199696693974948,
+      "loss": 2.8331,
+      "step": 1016
+    },
+    {
+      "epoch": 0.16272,
+      "grad_norm": 3.1312579850060644,
+      "learning_rate": 0.0001996953481300905,
+      "loss": 2.8201,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16288,
+      "grad_norm": 3.850688252811787,
+      "learning_rate": 0.00019969399931046904,
+      "loss": 2.9072,
+      "step": 1018
+    },
+    {
+      "epoch": 0.16304,
+      "grad_norm": 3.3054343547875042,
+      "learning_rate": 0.0001996926475161239,
+      "loss": 3.011,
+      "step": 1019
+    },
+    {
+      "epoch": 0.1632,
+      "grad_norm": 3.2198791976996826,
+      "learning_rate": 0.00019969129274709547,
+      "loss": 2.8842,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16336,
+      "grad_norm": 3.446163050471033,
+      "learning_rate": 0.00019968993500342407,
+      "loss": 2.9141,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16352,
+      "grad_norm": 3.52538525829937,
+      "learning_rate": 0.00019968857428515033,
+      "loss": 2.9044,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16368,
+      "grad_norm": 2.519360884600891,
+      "learning_rate": 0.0001996872105923148,
+      "loss": 2.8688,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16384,
+      "grad_norm": 3.180499289326071,
+      "learning_rate": 0.00019968584392495814,
+      "loss": 2.8746,
+      "step": 1024
+    },
+    {
+      "epoch": 0.164,
+      "grad_norm": 3.1786082051562525,
+      "learning_rate": 0.00019968447428312113,
+      "loss": 2.9178,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16416,
+      "grad_norm": 4.257475473578598,
+      "learning_rate": 0.0001996831016668447,
+      "loss": 2.746,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16432,
+      "grad_norm": 3.445874502850991,
+      "learning_rate": 0.00019968172607616976,
+      "loss": 2.851,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16448,
+      "grad_norm": 3.3277571957567833,
+      "learning_rate": 0.00019968034751113736,
+      "loss": 2.8838,
+      "step": 1028
+    },
+    {
+      "epoch": 0.16464,
+      "grad_norm": 3.8908973276218735,
+      "learning_rate": 0.00019967896597178862,
+      "loss": 2.9009,
+      "step": 1029
+    },
+    {
+      "epoch": 0.1648,
+      "grad_norm": 3.1212800289485454,
+      "learning_rate": 0.00019967758145816484,
+      "loss": 2.9221,
+      "step": 1030
+    },
+    {
+      "epoch": 0.16496,
+      "grad_norm": 3.2421862950164786,
+      "learning_rate": 0.00019967619397030721,
+      "loss": 2.8275,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16512,
+      "grad_norm": 3.4082787245752133,
+      "learning_rate": 0.00019967480350825724,
+      "loss": 3.0162,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16528,
+      "grad_norm": 3.771392018493907,
+      "learning_rate": 0.00019967341007205635,
+      "loss": 2.7731,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16544,
+      "grad_norm": 3.8342750060096678,
+      "learning_rate": 0.00019967201366174618,
+      "loss": 2.7919,
+      "step": 1034
+    },
+    {
+      "epoch": 0.1656,
+      "grad_norm": 3.444669108376155,
+      "learning_rate": 0.00019967061427736833,
+      "loss": 2.9463,
+      "step": 1035
+    },
+    {
+      "epoch": 0.16576,
+      "grad_norm": 3.0500380029709913,
+      "learning_rate": 0.00019966921191896458,
+      "loss": 2.8135,
+      "step": 1036
+    },
+    {
+      "epoch": 0.16592,
+      "grad_norm": 3.64703866136868,
+      "learning_rate": 0.0001996678065865768,
+      "loss": 2.8432,
+      "step": 1037
+    },
+    {
+      "epoch": 0.16608,
+      "grad_norm": 2.69701323714497,
+      "learning_rate": 0.0001996663982802469,
+      "loss": 2.8739,
+      "step": 1038
+    },
+    {
+      "epoch": 0.16624,
+      "grad_norm": 2.9354152029180023,
+      "learning_rate": 0.0001996649870000169,
+      "loss": 2.838,
+      "step": 1039
+    },
+    {
+      "epoch": 0.1664,
+      "grad_norm": 3.486186927836575,
+      "learning_rate": 0.0001996635727459289,
+      "loss": 2.8566,
+      "step": 1040
+    },
+    {
+      "epoch": 0.16656,
+      "grad_norm": 2.9519999739563314,
+      "learning_rate": 0.00019966215551802513,
+      "loss": 2.8223,
+      "step": 1041
+    },
+    {
+      "epoch": 0.16672,
+      "grad_norm": 2.4143583817723444,
+      "learning_rate": 0.00019966073531634785,
+      "loss": 2.7835,
+      "step": 1042
+    },
+    {
+      "epoch": 0.16688,
+      "grad_norm": 2.662504404467416,
+      "learning_rate": 0.00019965931214093946,
+      "loss": 2.6979,
+      "step": 1043
+    },
+    {
+      "epoch": 0.16704,
+      "grad_norm": 2.8643203968355313,
+      "learning_rate": 0.0001996578859918424,
+      "loss": 2.8657,
+      "step": 1044
+    },
+    {
+      "epoch": 0.1672,
+      "grad_norm": 2.7019316984028743,
+      "learning_rate": 0.0001996564568690993,
+      "loss": 2.8117,
+      "step": 1045
+    },
+    {
+      "epoch": 0.16736,
+      "grad_norm": 3.0819116216850757,
+      "learning_rate": 0.00019965502477275267,
+      "loss": 3.0033,
+      "step": 1046
+    },
+    {
+      "epoch": 0.16752,
+      "grad_norm": 3.4633495807414563,
+      "learning_rate": 0.00019965358970284533,
+      "loss": 2.6872,
+      "step": 1047
+    },
+    {
+      "epoch": 0.16768,
+      "grad_norm": 3.7848605540165576,
+      "learning_rate": 0.00019965215165942008,
+      "loss": 2.6834,
+      "step": 1048
+    },
+    {
+      "epoch": 0.16784,
+      "grad_norm": 3.246005021033392,
+      "learning_rate": 0.00019965071064251985,
+      "loss": 2.8034,
+      "step": 1049
+    },
+    {
+      "epoch": 0.168,
+      "grad_norm": 2.7823696861299845,
+      "learning_rate": 0.00019964926665218765,
+      "loss": 2.8015,
+      "step": 1050
+    },
+    {
+      "epoch": 0.16816,
+      "grad_norm": 2.227808335841823,
+      "learning_rate": 0.0001996478196884665,
+      "loss": 2.6117,
+      "step": 1051
+    },
+    {
+      "epoch": 0.16832,
+      "grad_norm": 2.4810364699249496,
+      "learning_rate": 0.00019964636975139958,
+      "loss": 2.8073,
+      "step": 1052
+    },
+    {
+      "epoch": 0.16848,
+      "grad_norm": 2.6020684701125387,
+      "learning_rate": 0.00019964491684103025,
+      "loss": 2.7635,
+      "step": 1053
+    },
+    {
+      "epoch": 0.16864,
+      "grad_norm": 2.673417902137821,
+      "learning_rate": 0.00019964346095740175,
+      "loss": 2.7638,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1688,
+      "grad_norm": 3.2067696547278253,
+      "learning_rate": 0.0001996420021005576,
+      "loss": 2.9561,
+      "step": 1055
+    },
+    {
+      "epoch": 0.16896,
+      "grad_norm": 3.0818811161338555,
+      "learning_rate": 0.00019964054027054128,
+      "loss": 2.7345,
+      "step": 1056
+    },
+    {
+      "epoch": 0.16912,
+      "grad_norm": 3.2964475694010154,
+      "learning_rate": 0.00019963907546739646,
+      "loss": 2.7892,
+      "step": 1057
+    },
+    {
+      "epoch": 0.16928,
+      "grad_norm": 3.507243249266056,
+      "learning_rate": 0.0001996376076911668,
+      "loss": 2.7793,
+      "step": 1058
+    },
+    {
+      "epoch": 0.16944,
+      "grad_norm": 3.2103822442237706,
+      "learning_rate": 0.0001996361369418961,
+      "loss": 2.8789,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1696,
+      "grad_norm": 3.390773572000025,
+      "learning_rate": 0.0001996346632196283,
+      "loss": 2.7818,
+      "step": 1060
+    },
+    {
+      "epoch": 0.16976,
+      "grad_norm": 3.195673274235192,
+      "learning_rate": 0.0001996331865244073,
+      "loss": 2.8463,
+      "step": 1061
+    },
+    {
+      "epoch": 0.16992,
+      "grad_norm": 3.853583646199081,
+      "learning_rate": 0.00019963170685627718,
+      "loss": 2.7922,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17008,
+      "grad_norm": 3.6818658798403217,
+      "learning_rate": 0.00019963022421528215,
+      "loss": 2.7948,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17024,
+      "grad_norm": 3.658644624299246,
+      "learning_rate": 0.00019962873860146637,
+      "loss": 2.7858,
+      "step": 1064
+    },
+    {
+      "epoch": 0.1704,
+      "grad_norm": 2.976005318532947,
+      "learning_rate": 0.00019962725001487425,
+      "loss": 2.7899,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17056,
+      "grad_norm": 2.8760194129212713,
+      "learning_rate": 0.0001996257584555501,
+      "loss": 2.6415,
+      "step": 1066
+    },
+    {
+      "epoch": 0.17072,
+      "grad_norm": 3.4808374689344976,
+      "learning_rate": 0.0001996242639235385,
+      "loss": 2.7805,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17088,
+      "grad_norm": 3.223158749396025,
+      "learning_rate": 0.00019962276641888405,
+      "loss": 2.7576,
+      "step": 1068
+    },
+    {
+      "epoch": 0.17104,
+      "grad_norm": 3.0714194534305492,
+      "learning_rate": 0.00019962126594163143,
+      "loss": 2.739,
+      "step": 1069
+    },
+    {
+      "epoch": 0.1712,
+      "grad_norm": 3.403886030341253,
+      "learning_rate": 0.0001996197624918254,
+      "loss": 2.8257,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17136,
+      "grad_norm": 3.4893603222418665,
+      "learning_rate": 0.0001996182560695108,
+      "loss": 2.8415,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17152,
+      "grad_norm": 2.8537171528963734,
+      "learning_rate": 0.0001996167466747326,
+      "loss": 2.7055,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17168,
+      "grad_norm": 3.7107582940574733,
+      "learning_rate": 0.00019961523430753585,
+      "loss": 2.8412,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17184,
+      "grad_norm": 3.3323749468253276,
+      "learning_rate": 0.00019961371896796564,
+      "loss": 2.7878,
+      "step": 1074
+    },
+    {
+      "epoch": 0.172,
+      "grad_norm": 3.3414777149101034,
+      "learning_rate": 0.0001996122006560672,
+      "loss": 2.6642,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17216,
+      "grad_norm": 3.55140844570953,
+      "learning_rate": 0.00019961067937188586,
+      "loss": 2.705,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17232,
+      "grad_norm": 3.5689757469506933,
+      "learning_rate": 0.000199609155115467,
+      "loss": 2.8827,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17248,
+      "grad_norm": 4.224529520858297,
+      "learning_rate": 0.0001996076278868561,
+      "loss": 2.8601,
+      "step": 1078
+    },
+    {
+      "epoch": 0.17264,
+      "grad_norm": 3.3003249477083587,
+      "learning_rate": 0.00019960609768609873,
+      "loss": 2.7115,
+      "step": 1079
+    },
+    {
+      "epoch": 0.1728,
+      "grad_norm": 3.5688299976228546,
+      "learning_rate": 0.00019960456451324055,
+      "loss": 2.9805,
+      "step": 1080
+    },
+    {
+      "epoch": 0.17296,
+      "grad_norm": 2.938831647632015,
+      "learning_rate": 0.00019960302836832727,
+      "loss": 2.796,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17312,
+      "grad_norm": 3.0760803311214358,
+      "learning_rate": 0.00019960148925140479,
+      "loss": 2.8197,
+      "step": 1082
+    },
+    {
+      "epoch": 0.17328,
+      "grad_norm": 2.9766392471479817,
+      "learning_rate": 0.000199599947162519,
+      "loss": 2.7564,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17344,
+      "grad_norm": 3.21354694237579,
+      "learning_rate": 0.00019959840210171593,
+      "loss": 2.7052,
+      "step": 1084
+    },
+    {
+      "epoch": 0.1736,
+      "grad_norm": 2.8410164277555374,
+      "learning_rate": 0.00019959685406904166,
+      "loss": 2.7671,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17376,
+      "grad_norm": 3.4059515063240773,
+      "learning_rate": 0.00019959530306454238,
+      "loss": 2.7992,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17392,
+      "grad_norm": 2.3957246828699255,
+      "learning_rate": 0.00019959374908826437,
+      "loss": 2.7695,
+      "step": 1087
+    },
+    {
+      "epoch": 0.17408,
+      "grad_norm": 3.2019069553092248,
+      "learning_rate": 0.00019959219214025405,
+      "loss": 2.8159,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17424,
+      "grad_norm": 3.4087984359839703,
+      "learning_rate": 0.0001995906322205578,
+      "loss": 2.8968,
+      "step": 1089
+    },
+    {
+      "epoch": 0.1744,
+      "grad_norm": 3.1102593889981334,
+      "learning_rate": 0.00019958906932922223,
+      "loss": 2.7467,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17456,
+      "grad_norm": 3.9508065083875716,
+      "learning_rate": 0.00019958750346629395,
+      "loss": 2.8484,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17472,
+      "grad_norm": 2.842405100361506,
+      "learning_rate": 0.00019958593463181967,
+      "loss": 2.7133,
+      "step": 1092
+    },
+    {
+      "epoch": 0.17488,
+      "grad_norm": 3.325759818422144,
+      "learning_rate": 0.0001995843628258462,
+      "loss": 2.7126,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17504,
+      "grad_norm": 3.693719188775163,
+      "learning_rate": 0.0001995827880484205,
+      "loss": 2.8303,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1752,
+      "grad_norm": 3.6693872029911985,
+      "learning_rate": 0.00019958121029958948,
+      "loss": 2.6867,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17536,
+      "grad_norm": 2.539737987084127,
+      "learning_rate": 0.00019957962957940024,
+      "loss": 2.6484,
+      "step": 1096
+    },
+    {
+      "epoch": 0.17552,
+      "grad_norm": 3.6402299904530824,
+      "learning_rate": 0.00019957804588789996,
+      "loss": 2.8162,
+      "step": 1097
+    },
+    {
+      "epoch": 0.17568,
+      "grad_norm": 2.8792756620339093,
+      "learning_rate": 0.0001995764592251359,
+      "loss": 2.7585,
+      "step": 1098
+    },
+    {
+      "epoch": 0.17584,
+      "grad_norm": 3.2820886671926424,
+      "learning_rate": 0.0001995748695911554,
+      "loss": 2.9273,
+      "step": 1099
+    },
+    {
+      "epoch": 0.176,
+      "grad_norm": 3.3680706608821254,
+      "learning_rate": 0.00019957327698600586,
+      "loss": 2.6872,
+      "step": 1100
+    },
+    {
+      "epoch": 0.17616,
+      "grad_norm": 3.107080088658685,
+      "learning_rate": 0.00019957168140973485,
+      "loss": 2.7939,
+      "step": 1101
+    },
+    {
+      "epoch": 0.17632,
+      "grad_norm": 3.0655904113981163,
+      "learning_rate": 0.00019957008286238995,
+      "loss": 2.7864,
+      "step": 1102
+    },
+    {
+      "epoch": 0.17648,
+      "grad_norm": 3.1413572378218553,
+      "learning_rate": 0.00019956848134401888,
+      "loss": 2.7351,
+      "step": 1103
+    },
+    {
+      "epoch": 0.17664,
+      "grad_norm": 3.2720658427512883,
+      "learning_rate": 0.00019956687685466942,
+      "loss": 2.8006,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1768,
+      "grad_norm": 3.180537540349954,
+      "learning_rate": 0.0001995652693943894,
+      "loss": 2.7106,
+      "step": 1105
+    },
+    {
+      "epoch": 0.17696,
+      "grad_norm": 3.484639976572538,
+      "learning_rate": 0.00019956365896322688,
+      "loss": 2.8487,
+      "step": 1106
+    },
+    {
+      "epoch": 0.17712,
+      "grad_norm": 2.7463666193942506,
+      "learning_rate": 0.00019956204556122984,
+      "loss": 2.7414,
+      "step": 1107
+    },
+    {
+      "epoch": 0.17728,
+      "grad_norm": 2.653067725806054,
+      "learning_rate": 0.00019956042918844639,
+      "loss": 2.753,
+      "step": 1108
+    },
+    {
+      "epoch": 0.17744,
+      "grad_norm": 3.888611497391141,
+      "learning_rate": 0.00019955880984492486,
+      "loss": 2.8483,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1776,
+      "grad_norm": 2.9295686940700687,
+      "learning_rate": 0.00019955718753071348,
+      "loss": 2.8225,
+      "step": 1110
+    },
+    {
+      "epoch": 0.17776,
+      "grad_norm": 3.695176068435238,
+      "learning_rate": 0.00019955556224586073,
+      "loss": 2.6458,
+      "step": 1111
+    },
+    {
+      "epoch": 0.17792,
+      "grad_norm": 3.578331628436118,
+      "learning_rate": 0.00019955393399041506,
+      "loss": 2.651,
+      "step": 1112
+    },
+    {
+      "epoch": 0.17808,
+      "grad_norm": 2.9691679515038767,
+      "learning_rate": 0.00019955230276442505,
+      "loss": 2.8321,
+      "step": 1113
+    },
+    {
+      "epoch": 0.17824,
+      "grad_norm": 3.944209556480424,
+      "learning_rate": 0.0001995506685679394,
+      "loss": 2.8349,
+      "step": 1114
+    },
+    {
+      "epoch": 0.1784,
+      "grad_norm": 3.958531171681635,
+      "learning_rate": 0.00019954903140100686,
+      "loss": 2.7752,
+      "step": 1115
+    },
+    {
+      "epoch": 0.17856,
+      "grad_norm": 3.0914906523816232,
+      "learning_rate": 0.00019954739126367628,
+      "loss": 2.7065,
+      "step": 1116
+    },
+    {
+      "epoch": 0.17872,
+      "grad_norm": 3.6674397592579506,
+      "learning_rate": 0.00019954574815599657,
+      "loss": 2.7248,
+      "step": 1117
+    },
+    {
+      "epoch": 0.17888,
+      "grad_norm": 3.0560591488242186,
+      "learning_rate": 0.00019954410207801684,
+      "loss": 2.8623,
+      "step": 1118
+    },
+    {
+      "epoch": 0.17904,
+      "grad_norm": 3.849513792231411,
+      "learning_rate": 0.0001995424530297861,
+      "loss": 2.7629,
+      "step": 1119
+    },
+    {
+      "epoch": 0.1792,
+      "grad_norm": 2.8364833138523693,
+      "learning_rate": 0.00019954080101135366,
+      "loss": 2.9306,
+      "step": 1120
+    },
+    {
+      "epoch": 0.17936,
+      "grad_norm": 3.734033586320656,
+      "learning_rate": 0.00019953914602276876,
+      "loss": 2.7288,
+      "step": 1121
+    },
+    {
+      "epoch": 0.17952,
+      "grad_norm": 3.37027802359214,
+      "learning_rate": 0.00019953748806408078,
+      "loss": 2.7041,
+      "step": 1122
+    },
+    {
+      "epoch": 0.17968,
+      "grad_norm": 3.735299598606662,
+      "learning_rate": 0.00019953582713533918,
+      "loss": 2.8112,
+      "step": 1123
+    },
+    {
+      "epoch": 0.17984,
+      "grad_norm": 3.072663324343328,
+      "learning_rate": 0.00019953416323659356,
+      "loss": 2.6796,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 3.7969844098011123,
+      "learning_rate": 0.0001995324963678935,
+      "loss": 2.916,
+      "step": 1125
+    },
+    {
+      "epoch": 0.18016,
+      "grad_norm": 4.360494839513198,
+      "learning_rate": 0.00019953082652928884,
+      "loss": 2.6185,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18032,
+      "grad_norm": 3.1058589972942414,
+      "learning_rate": 0.00019952915372082933,
+      "loss": 2.8852,
+      "step": 1127
+    },
+    {
+      "epoch": 0.18048,
+      "grad_norm": 4.0657538232496835,
+      "learning_rate": 0.00019952747794256488,
+      "loss": 2.6707,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18064,
+      "grad_norm": 3.2232055225410954,
+      "learning_rate": 0.00019952579919454551,
+      "loss": 2.6385,
+      "step": 1129
+    },
+    {
+      "epoch": 0.1808,
+      "grad_norm": 3.4182174071564173,
+      "learning_rate": 0.00019952411747682137,
+      "loss": 2.7293,
+      "step": 1130
+    },
+    {
+      "epoch": 0.18096,
+      "grad_norm": 3.655025632129096,
+      "learning_rate": 0.00019952243278944254,
+      "loss": 2.7281,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18112,
+      "grad_norm": 3.3124644718389,
+      "learning_rate": 0.00019952074513245936,
+      "loss": 2.8107,
+      "step": 1132
+    },
+    {
+      "epoch": 0.18128,
+      "grad_norm": 3.753541069250205,
+      "learning_rate": 0.00019951905450592215,
+      "loss": 2.7445,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18144,
+      "grad_norm": 2.913282071879843,
+      "learning_rate": 0.00019951736090988135,
+      "loss": 2.7353,
+      "step": 1134
+    },
+    {
+      "epoch": 0.1816,
+      "grad_norm": 3.5015498140233796,
+      "learning_rate": 0.00019951566434438754,
+      "loss": 2.7551,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18176,
+      "grad_norm": 3.6710629821303815,
+      "learning_rate": 0.00019951396480949132,
+      "loss": 2.883,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18192,
+      "grad_norm": 2.6794105208433163,
+      "learning_rate": 0.00019951226230524339,
+      "loss": 2.7891,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18208,
+      "grad_norm": 3.371504011259695,
+      "learning_rate": 0.00019951055683169452,
+      "loss": 2.6982,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18224,
+      "grad_norm": 2.982852127709996,
+      "learning_rate": 0.0001995088483888957,
+      "loss": 2.8133,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1824,
+      "grad_norm": 3.448409063773366,
+      "learning_rate": 0.00019950713697689781,
+      "loss": 2.7854,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18256,
+      "grad_norm": 2.92905265329273,
+      "learning_rate": 0.00019950542259575195,
+      "loss": 2.7669,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18272,
+      "grad_norm": 2.7104947308253444,
+      "learning_rate": 0.00019950370524550928,
+      "loss": 2.7918,
+      "step": 1142
+    },
+    {
+      "epoch": 0.18288,
+      "grad_norm": 3.184711700717751,
+      "learning_rate": 0.00019950198492622103,
+      "loss": 2.696,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18304,
+      "grad_norm": 2.4187454228200393,
+      "learning_rate": 0.00019950026163793853,
+      "loss": 2.6822,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1832,
+      "grad_norm": 3.7581567546185637,
+      "learning_rate": 0.00019949853538071323,
+      "loss": 2.7123,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18336,
+      "grad_norm": 2.90392071563963,
+      "learning_rate": 0.00019949680615459663,
+      "loss": 2.6954,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18352,
+      "grad_norm": 3.1883808916483516,
+      "learning_rate": 0.00019949507395964028,
+      "loss": 2.7315,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18368,
+      "grad_norm": 2.655054668042209,
+      "learning_rate": 0.00019949333879589593,
+      "loss": 2.7168,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18384,
+      "grad_norm": 3.287918949493105,
+      "learning_rate": 0.00019949160066341532,
+      "loss": 2.7118,
+      "step": 1149
+    },
+    {
+      "epoch": 0.184,
+      "grad_norm": 3.0973481888578807,
+      "learning_rate": 0.0001994898595622503,
+      "loss": 2.7077,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18416,
+      "grad_norm": 3.830790112110103,
+      "learning_rate": 0.00019948811549245287,
+      "loss": 2.6884,
+      "step": 1151
+    },
+    {
+      "epoch": 0.18432,
+      "grad_norm": 3.196219738322992,
+      "learning_rate": 0.00019948636845407505,
+      "loss": 2.7047,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18448,
+      "grad_norm": 3.6299777032011855,
+      "learning_rate": 0.00019948461844716895,
+      "loss": 2.7642,
+      "step": 1153
+    },
+    {
+      "epoch": 0.18464,
+      "grad_norm": 3.4205157057257436,
+      "learning_rate": 0.0001994828654717868,
+      "loss": 2.7592,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1848,
+      "grad_norm": 2.5759129489685497,
+      "learning_rate": 0.0001994811095279809,
+      "loss": 2.7856,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18496,
+      "grad_norm": 3.292732841744169,
+      "learning_rate": 0.00019947935061580366,
+      "loss": 2.6943,
+      "step": 1156
+    },
+    {
+      "epoch": 0.18512,
+      "grad_norm": 2.8260798818071864,
+      "learning_rate": 0.00019947758873530754,
+      "loss": 2.8485,
+      "step": 1157
+    },
+    {
+      "epoch": 0.18528,
+      "grad_norm": 3.4343212898074884,
+      "learning_rate": 0.00019947582388654515,
+      "loss": 2.69,
+      "step": 1158
+    },
+    {
+      "epoch": 0.18544,
+      "grad_norm": 4.335411949622639,
+      "learning_rate": 0.0001994740560695691,
+      "loss": 2.7194,
+      "step": 1159
+    },
+    {
+      "epoch": 0.1856,
+      "grad_norm": 3.0438322552474286,
+      "learning_rate": 0.0001994722852844322,
+      "loss": 2.6078,
+      "step": 1160
+    },
+    {
+      "epoch": 0.18576,
+      "grad_norm": 3.0820264916343967,
+      "learning_rate": 0.00019947051153118724,
+      "loss": 2.6961,
+      "step": 1161
+    },
+    {
+      "epoch": 0.18592,
+      "grad_norm": 2.798022144295341,
+      "learning_rate": 0.00019946873480988713,
+      "loss": 2.7282,
+      "step": 1162
+    },
+    {
+      "epoch": 0.18608,
+      "grad_norm": 4.155220158814774,
+      "learning_rate": 0.00019946695512058495,
+      "loss": 2.7072,
+      "step": 1163
+    },
+    {
+      "epoch": 0.18624,
+      "grad_norm": 3.2614689336907676,
+      "learning_rate": 0.00019946517246333375,
+      "loss": 2.6025,
+      "step": 1164
+    },
+    {
+      "epoch": 0.1864,
+      "grad_norm": 3.540517188241641,
+      "learning_rate": 0.00019946338683818674,
+      "loss": 2.6163,
+      "step": 1165
+    },
+    {
+      "epoch": 0.18656,
+      "grad_norm": 2.860879787840734,
+      "learning_rate": 0.0001994615982451972,
+      "loss": 2.6211,
+      "step": 1166
+    },
+    {
+      "epoch": 0.18672,
+      "grad_norm": 3.981079445567166,
+      "learning_rate": 0.0001994598066844185,
+      "loss": 2.7472,
+      "step": 1167
+    },
+    {
+      "epoch": 0.18688,
+      "grad_norm": 4.251688167582394,
+      "learning_rate": 0.00019945801215590412,
+      "loss": 2.6889,
+      "step": 1168
+    },
+    {
+      "epoch": 0.18704,
+      "grad_norm": 2.9202839172619592,
+      "learning_rate": 0.00019945621465970755,
+      "loss": 2.6326,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1872,
+      "grad_norm": 3.3047734057334024,
+      "learning_rate": 0.00019945441419588246,
+      "loss": 2.6876,
+      "step": 1170
+    },
+    {
+      "epoch": 0.18736,
+      "grad_norm": 3.485586363657066,
+      "learning_rate": 0.0001994526107644826,
+      "loss": 2.5394,
+      "step": 1171
+    },
+    {
+      "epoch": 0.18752,
+      "grad_norm": 3.0471039584756388,
+      "learning_rate": 0.00019945080436556173,
+      "loss": 2.6863,
+      "step": 1172
+    },
+    {
+      "epoch": 0.18768,
+      "grad_norm": 3.677680228466432,
+      "learning_rate": 0.00019944899499917378,
+      "loss": 2.6973,
+      "step": 1173
+    },
+    {
+      "epoch": 0.18784,
+      "grad_norm": 3.2438198415052835,
+      "learning_rate": 0.0001994471826653727,
+      "loss": 2.6344,
+      "step": 1174
+    },
+    {
+      "epoch": 0.188,
+      "grad_norm": 2.911909004394838,
+      "learning_rate": 0.00019944536736421267,
+      "loss": 2.6265,
+      "step": 1175
+    },
+    {
+      "epoch": 0.18816,
+      "grad_norm": 2.9015622174941913,
+      "learning_rate": 0.0001994435490957477,
+      "loss": 2.7817,
+      "step": 1176
+    },
+    {
+      "epoch": 0.18832,
+      "grad_norm": 3.3778659489759053,
+      "learning_rate": 0.0001994417278600322,
+      "loss": 2.6697,
+      "step": 1177
+    },
+    {
+      "epoch": 0.18848,
+      "grad_norm": 2.617092196131355,
+      "learning_rate": 0.00019943990365712042,
+      "loss": 2.6746,
+      "step": 1178
+    },
+    {
+      "epoch": 0.18864,
+      "grad_norm": 3.9848626337715785,
+      "learning_rate": 0.0001994380764870668,
+      "loss": 2.6642,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1888,
+      "grad_norm": 2.4423193814905693,
+      "learning_rate": 0.00019943624634992588,
+      "loss": 2.6733,
+      "step": 1180
+    },
+    {
+      "epoch": 0.18896,
+      "grad_norm": 2.4386352434487857,
+      "learning_rate": 0.00019943441324575228,
+      "loss": 2.7376,
+      "step": 1181
+    },
+    {
+      "epoch": 0.18912,
+      "grad_norm": 3.1899487076965833,
+      "learning_rate": 0.00019943257717460066,
+      "loss": 2.6376,
+      "step": 1182
+    },
+    {
+      "epoch": 0.18928,
+      "grad_norm": 2.8994917293232585,
+      "learning_rate": 0.0001994307381365258,
+      "loss": 2.7359,
+      "step": 1183
+    },
+    {
+      "epoch": 0.18944,
+      "grad_norm": 3.193003850906536,
+      "learning_rate": 0.00019942889613158268,
+      "loss": 2.7316,
+      "step": 1184
+    },
+    {
+      "epoch": 0.1896,
+      "grad_norm": 3.514727062318123,
+      "learning_rate": 0.00019942705115982612,
+      "loss": 2.5649,
+      "step": 1185
+    },
+    {
+      "epoch": 0.18976,
+      "grad_norm": 3.8206770536846384,
+      "learning_rate": 0.00019942520322131126,
+      "loss": 2.7527,
+      "step": 1186
+    },
+    {
+      "epoch": 0.18992,
+      "grad_norm": 3.6259309774223074,
+      "learning_rate": 0.0001994233523160932,
+      "loss": 2.6196,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19008,
+      "grad_norm": 2.5777349203585063,
+      "learning_rate": 0.00019942149844422716,
+      "loss": 2.5778,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19024,
+      "grad_norm": 4.267341425308837,
+      "learning_rate": 0.00019941964160576854,
+      "loss": 2.7496,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1904,
+      "grad_norm": 3.9258341759582396,
+      "learning_rate": 0.00019941778180077263,
+      "loss": 2.6761,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19056,
+      "grad_norm": 2.987500769035036,
+      "learning_rate": 0.000199415919029295,
+      "loss": 2.6448,
+      "step": 1191
+    },
+    {
+      "epoch": 0.19072,
+      "grad_norm": 3.8403547385290766,
+      "learning_rate": 0.0001994140532913912,
+      "loss": 2.6997,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19088,
+      "grad_norm": 2.7986132265531776,
+      "learning_rate": 0.00019941218458711693,
+      "loss": 2.5513,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19104,
+      "grad_norm": 3.022501697404106,
+      "learning_rate": 0.00019941031291652788,
+      "loss": 2.6622,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1912,
+      "grad_norm": 3.0196005990150625,
+      "learning_rate": 0.00019940843827968,
+      "loss": 2.6262,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19136,
+      "grad_norm": 4.327190968722658,
+      "learning_rate": 0.00019940656067662918,
+      "loss": 2.6969,
+      "step": 1196
+    },
+    {
+      "epoch": 0.19152,
+      "grad_norm": 3.6421134481037214,
+      "learning_rate": 0.0001994046801074314,
+      "loss": 2.6123,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19168,
+      "grad_norm": 3.377142424287418,
+      "learning_rate": 0.00019940279657214288,
+      "loss": 2.6706,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19184,
+      "grad_norm": 2.7953616395429837,
+      "learning_rate": 0.0001994009100708197,
+      "loss": 2.5836,
+      "step": 1199
+    },
+    {
+      "epoch": 0.192,
+      "grad_norm": 3.9362538932586726,
+      "learning_rate": 0.0001993990206035182,
+      "loss": 2.6513,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19216,
+      "grad_norm": 3.8291605409710034,
+      "learning_rate": 0.0001993971281702948,
+      "loss": 2.7841,
+      "step": 1201
+    },
+    {
+      "epoch": 0.19232,
+      "grad_norm": 4.035104198032958,
+      "learning_rate": 0.0001993952327712059,
+      "loss": 2.6905,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19248,
+      "grad_norm": 3.6452629425513594,
+      "learning_rate": 0.00019939333440630808,
+      "loss": 2.6136,
+      "step": 1203
+    },
+    {
+      "epoch": 0.19264,
+      "grad_norm": 3.6409632585199248,
+      "learning_rate": 0.00019939143307565803,
+      "loss": 2.6989,
+      "step": 1204
+    },
+    {
+      "epoch": 0.1928,
+      "grad_norm": 3.3641320842134887,
+      "learning_rate": 0.00019938952877931242,
+      "loss": 2.6725,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19296,
+      "grad_norm": 3.7275743060585835,
+      "learning_rate": 0.00019938762151732812,
+      "loss": 2.758,
+      "step": 1206
+    },
+    {
+      "epoch": 0.19312,
+      "grad_norm": 3.302621773738114,
+      "learning_rate": 0.000199385711289762,
+      "loss": 2.5082,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19328,
+      "grad_norm": 3.103317730894929,
+      "learning_rate": 0.0001993837980966711,
+      "loss": 2.7624,
+      "step": 1208
+    },
+    {
+      "epoch": 0.19344,
+      "grad_norm": 3.146163673236022,
+      "learning_rate": 0.00019938188193811246,
+      "loss": 2.5438,
+      "step": 1209
+    },
+    {
+      "epoch": 0.1936,
+      "grad_norm": 3.8450839989516425,
+      "learning_rate": 0.0001993799628141433,
+      "loss": 2.662,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19376,
+      "grad_norm": 3.4238918041164164,
+      "learning_rate": 0.00019937804072482081,
+      "loss": 2.5826,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19392,
+      "grad_norm": 4.189092739316937,
+      "learning_rate": 0.00019937611567020244,
+      "loss": 2.5696,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19408,
+      "grad_norm": 2.897903698199734,
+      "learning_rate": 0.00019937418765034556,
+      "loss": 2.6325,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19424,
+      "grad_norm": 4.000406193353709,
+      "learning_rate": 0.00019937225666530776,
+      "loss": 2.5793,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1944,
+      "grad_norm": 3.8097508092402665,
+      "learning_rate": 0.00019937032271514658,
+      "loss": 2.5769,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19456,
+      "grad_norm": 4.322464538279433,
+      "learning_rate": 0.0001993683857999198,
+      "loss": 2.6822,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19472,
+      "grad_norm": 3.735796241876012,
+      "learning_rate": 0.00019936644591968516,
+      "loss": 2.7751,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19488,
+      "grad_norm": 3.7516864440314333,
+      "learning_rate": 0.00019936450307450057,
+      "loss": 2.6282,
+      "step": 1218
+    },
+    {
+      "epoch": 0.19504,
+      "grad_norm": 4.213123963399589,
+      "learning_rate": 0.000199362557264424,
+      "loss": 2.5785,
+      "step": 1219
+    },
+    {
+      "epoch": 0.1952,
+      "grad_norm": 3.4450209596579424,
+      "learning_rate": 0.0001993606084895135,
+      "loss": 2.5518,
+      "step": 1220
+    },
+    {
+      "epoch": 0.19536,
+      "grad_norm": 3.251858205562727,
+      "learning_rate": 0.00019935865674982722,
+      "loss": 2.5042,
+      "step": 1221
+    },
+    {
+      "epoch": 0.19552,
+      "grad_norm": 3.30287218565954,
+      "learning_rate": 0.00019935670204542346,
+      "loss": 2.5388,
+      "step": 1222
+    },
+    {
+      "epoch": 0.19568,
+      "grad_norm": 2.73004678935357,
+      "learning_rate": 0.0001993547443763604,
+      "loss": 2.5917,
+      "step": 1223
+    },
+    {
+      "epoch": 0.19584,
+      "grad_norm": 2.8732199380124803,
+      "learning_rate": 0.0001993527837426966,
+      "loss": 2.6036,
+      "step": 1224
+    },
+    {
+      "epoch": 0.196,
+      "grad_norm": 2.611616228062945,
+      "learning_rate": 0.0001993508201444905,
+      "loss": 2.6191,
+      "step": 1225
+    },
+    {
+      "epoch": 0.19616,
+      "grad_norm": 3.042133668090087,
+      "learning_rate": 0.00019934885358180072,
+      "loss": 2.601,
+      "step": 1226
+    },
+    {
+      "epoch": 0.19632,
+      "grad_norm": 3.122075143436874,
+      "learning_rate": 0.00019934688405468586,
+      "loss": 2.6174,
+      "step": 1227
+    },
+    {
+      "epoch": 0.19648,
+      "grad_norm": 2.4356028871158246,
+      "learning_rate": 0.00019934491156320478,
+      "loss": 2.6274,
+      "step": 1228
+    },
+    {
+      "epoch": 0.19664,
+      "grad_norm": 3.395252281424135,
+      "learning_rate": 0.0001993429361074163,
+      "loss": 2.7255,
+      "step": 1229
+    },
+    {
+      "epoch": 0.1968,
+      "grad_norm": 3.127489176313882,
+      "learning_rate": 0.00019934095768737935,
+      "loss": 2.6921,
+      "step": 1230
+    },
+    {
+      "epoch": 0.19696,
+      "grad_norm": 4.062849144793317,
+      "learning_rate": 0.00019933897630315298,
+      "loss": 2.611,
+      "step": 1231
+    },
+    {
+      "epoch": 0.19712,
+      "grad_norm": 3.117716052063002,
+      "learning_rate": 0.00019933699195479632,
+      "loss": 2.4658,
+      "step": 1232
+    },
+    {
+      "epoch": 0.19728,
+      "grad_norm": 3.9236256522324995,
+      "learning_rate": 0.00019933500464236855,
+      "loss": 2.5847,
+      "step": 1233
+    },
+    {
+      "epoch": 0.19744,
+      "grad_norm": 3.600839551262328,
+      "learning_rate": 0.00019933301436592902,
+      "loss": 2.6787,
+      "step": 1234
+    },
+    {
+      "epoch": 0.1976,
+      "grad_norm": 3.1179510641779267,
+      "learning_rate": 0.00019933102112553702,
+      "loss": 2.6798,
+      "step": 1235
+    },
+    {
+      "epoch": 0.19776,
+      "grad_norm": 3.4776347399790044,
+      "learning_rate": 0.00019932902492125217,
+      "loss": 2.5982,
+      "step": 1236
+    },
+    {
+      "epoch": 0.19792,
+      "grad_norm": 3.073655773820224,
+      "learning_rate": 0.0001993270257531339,
+      "loss": 2.6773,
+      "step": 1237
+    },
+    {
+      "epoch": 0.19808,
+      "grad_norm": 3.0340573947594085,
+      "learning_rate": 0.00019932502362124196,
+      "loss": 2.4832,
+      "step": 1238
+    },
+    {
+      "epoch": 0.19824,
+      "grad_norm": 3.6576742102160167,
+      "learning_rate": 0.00019932301852563602,
+      "loss": 2.5296,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1984,
+      "grad_norm": 2.8162243374041362,
+      "learning_rate": 0.00019932101046637592,
+      "loss": 2.5829,
+      "step": 1240
+    },
+    {
+      "epoch": 0.19856,
+      "grad_norm": 3.129826657651419,
+      "learning_rate": 0.00019931899944352163,
+      "loss": 2.7114,
+      "step": 1241
+    },
+    {
+      "epoch": 0.19872,
+      "grad_norm": 2.9748553379954883,
+      "learning_rate": 0.00019931698545713308,
+      "loss": 2.6666,
+      "step": 1242
+    },
+    {
+      "epoch": 0.19888,
+      "grad_norm": 3.461293445932487,
+      "learning_rate": 0.00019931496850727042,
+      "loss": 2.5114,
+      "step": 1243
+    },
+    {
+      "epoch": 0.19904,
+      "grad_norm": 3.0891471684495513,
+      "learning_rate": 0.00019931294859399382,
+      "loss": 2.7255,
+      "step": 1244
+    },
+    {
+      "epoch": 0.1992,
+      "grad_norm": 3.3477975893813574,
+      "learning_rate": 0.0001993109257173636,
+      "loss": 2.5813,
+      "step": 1245
+    },
+    {
+      "epoch": 0.19936,
+      "grad_norm": 2.9991494730265216,
+      "learning_rate": 0.00019930889987744,
+      "loss": 2.5581,
+      "step": 1246
+    },
+    {
+      "epoch": 0.19952,
+      "grad_norm": 2.5595105965455023,
+      "learning_rate": 0.00019930687107428354,
+      "loss": 2.5877,
+      "step": 1247
+    },
+    {
+      "epoch": 0.19968,
+      "grad_norm": 2.5097351446694325,
+      "learning_rate": 0.00019930483930795476,
+      "loss": 2.7272,
+      "step": 1248
+    },
+    {
+      "epoch": 0.19984,
+      "grad_norm": 2.734283357366037,
+      "learning_rate": 0.00019930280457851432,
+      "loss": 2.7014,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 2.8606479302554986,
+      "learning_rate": 0.00019930076688602283,
+      "loss": 2.6419,
+      "step": 1250
+    },
+    {
+      "epoch": 0.20016,
+      "grad_norm": 4.432618764166668,
+      "learning_rate": 0.0001992987262305412,
+      "loss": 2.5763,
+      "step": 1251
+    },
+    {
+      "epoch": 0.20032,
+      "grad_norm": 3.107613454831708,
+      "learning_rate": 0.00019929668261213024,
+      "loss": 2.6163,
+      "step": 1252
+    },
+    {
+      "epoch": 0.20048,
+      "grad_norm": 2.640961623714852,
+      "learning_rate": 0.00019929463603085096,
+      "loss": 2.6114,
+      "step": 1253
+    },
+    {
+      "epoch": 0.20064,
+      "grad_norm": 2.779617757098041,
+      "learning_rate": 0.00019929258648676443,
+      "loss": 2.6537,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2008,
+      "grad_norm": 2.845625227535846,
+      "learning_rate": 0.00019929053397993186,
+      "loss": 2.5469,
+      "step": 1255
+    },
+    {
+      "epoch": 0.20096,
+      "grad_norm": 3.2354831765398777,
+      "learning_rate": 0.00019928847851041436,
+      "loss": 2.6078,
+      "step": 1256
+    },
+    {
+      "epoch": 0.20112,
+      "grad_norm": 3.2209390160930536,
+      "learning_rate": 0.00019928642007827337,
+      "loss": 2.6451,
+      "step": 1257
+    },
+    {
+      "epoch": 0.20128,
+      "grad_norm": 3.346876053806893,
+      "learning_rate": 0.0001992843586835703,
+      "loss": 2.6224,
+      "step": 1258
+    },
+    {
+      "epoch": 0.20144,
+      "grad_norm": 3.702963002726708,
+      "learning_rate": 0.00019928229432636662,
+      "loss": 2.696,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2016,
+      "grad_norm": 3.0535904557420106,
+      "learning_rate": 0.00019928022700672394,
+      "loss": 2.8414,
+      "step": 1260
+    },
+    {
+      "epoch": 0.20176,
+      "grad_norm": 3.06179310170563,
+      "learning_rate": 0.00019927815672470396,
+      "loss": 2.4849,
+      "step": 1261
+    },
+    {
+      "epoch": 0.20192,
+      "grad_norm": 2.7559110705402206,
+      "learning_rate": 0.00019927608348036845,
+      "loss": 2.4792,
+      "step": 1262
+    },
+    {
+      "epoch": 0.20208,
+      "grad_norm": 2.6774142419670026,
+      "learning_rate": 0.00019927400727377928,
+      "loss": 2.5353,
+      "step": 1263
+    },
+    {
+      "epoch": 0.20224,
+      "grad_norm": 3.2390822839431777,
+      "learning_rate": 0.00019927192810499837,
+      "loss": 2.6149,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2024,
+      "grad_norm": 3.3473481406642818,
+      "learning_rate": 0.00019926984597408778,
+      "loss": 2.6314,
+      "step": 1265
+    },
+    {
+      "epoch": 0.20256,
+      "grad_norm": 2.2662366359017674,
+      "learning_rate": 0.00019926776088110965,
+      "loss": 2.5468,
+      "step": 1266
+    },
+    {
+      "epoch": 0.20272,
+      "grad_norm": 2.7835941898472805,
+      "learning_rate": 0.00019926567282612618,
+      "loss": 2.5296,
+      "step": 1267
+    },
+    {
+      "epoch": 0.20288,
+      "grad_norm": 3.7027481974868826,
+      "learning_rate": 0.00019926358180919967,
+      "loss": 2.674,
+      "step": 1268
+    },
+    {
+      "epoch": 0.20304,
+      "grad_norm": 3.1075636321044846,
+      "learning_rate": 0.00019926148783039254,
+      "loss": 2.5647,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2032,
+      "grad_norm": 2.748473267147318,
+      "learning_rate": 0.00019925939088976725,
+      "loss": 2.5405,
+      "step": 1270
+    },
+    {
+      "epoch": 0.20336,
+      "grad_norm": 4.378476009423442,
+      "learning_rate": 0.00019925729098738633,
+      "loss": 2.5896,
+      "step": 1271
+    },
+    {
+      "epoch": 0.20352,
+      "grad_norm": 2.9475546503603054,
+      "learning_rate": 0.00019925518812331254,
+      "loss": 2.4731,
+      "step": 1272
+    },
+    {
+      "epoch": 0.20368,
+      "grad_norm": 3.1969187386997944,
+      "learning_rate": 0.00019925308229760855,
+      "loss": 2.4555,
+      "step": 1273
+    },
+    {
+      "epoch": 0.20384,
+      "grad_norm": 2.819989564267249,
+      "learning_rate": 0.0001992509735103372,
+      "loss": 2.5131,
+      "step": 1274
+    },
+    {
+      "epoch": 0.204,
+      "grad_norm": 2.933151826856131,
+      "learning_rate": 0.0001992488617615614,
+      "loss": 2.5199,
+      "step": 1275
+    },
+    {
+      "epoch": 0.20416,
+      "grad_norm": 3.304194631170998,
+      "learning_rate": 0.00019924674705134422,
+      "loss": 2.6383,
+      "step": 1276
+    },
+    {
+      "epoch": 0.20432,
+      "grad_norm": 2.9549745379150494,
+      "learning_rate": 0.00019924462937974874,
+      "loss": 2.5501,
+      "step": 1277
+    },
+    {
+      "epoch": 0.20448,
+      "grad_norm": 3.2229975286310517,
+      "learning_rate": 0.0001992425087468381,
+      "loss": 2.609,
+      "step": 1278
+    },
+    {
+      "epoch": 0.20464,
+      "grad_norm": 4.560934573435506,
+      "learning_rate": 0.00019924038515267562,
+      "loss": 2.5616,
+      "step": 1279
+    },
+    {
+      "epoch": 0.2048,
+      "grad_norm": 2.9151897111140976,
+      "learning_rate": 0.00019923825859732464,
+      "loss": 2.5603,
+      "step": 1280
+    },
+    {
+      "epoch": 0.20496,
+      "grad_norm": 2.7653924751155654,
+      "learning_rate": 0.00019923612908084865,
+      "loss": 2.5592,
+      "step": 1281
+    },
+    {
+      "epoch": 0.20512,
+      "grad_norm": 2.9848729581502855,
+      "learning_rate": 0.00019923399660331118,
+      "loss": 2.4891,
+      "step": 1282
+    },
+    {
+      "epoch": 0.20528,
+      "grad_norm": 2.7477553420934666,
+      "learning_rate": 0.00019923186116477585,
+      "loss": 2.5871,
+      "step": 1283
+    },
+    {
+      "epoch": 0.20544,
+      "grad_norm": 2.7690244482815487,
+      "learning_rate": 0.00019922972276530636,
+      "loss": 2.4857,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2056,
+      "grad_norm": 3.215963712448856,
+      "learning_rate": 0.00019922758140496652,
+      "loss": 2.5251,
+      "step": 1285
+    },
+    {
+      "epoch": 0.20576,
+      "grad_norm": 3.5220059693292196,
+      "learning_rate": 0.00019922543708382029,
+      "loss": 2.5155,
+      "step": 1286
+    },
+    {
+      "epoch": 0.20592,
+      "grad_norm": 2.723307777632806,
+      "learning_rate": 0.00019922328980193158,
+      "loss": 2.5769,
+      "step": 1287
+    },
+    {
+      "epoch": 0.20608,
+      "grad_norm": 3.6099766010990635,
+      "learning_rate": 0.00019922113955936447,
+      "loss": 2.5276,
+      "step": 1288
+    },
+    {
+      "epoch": 0.20624,
+      "grad_norm": 4.266621011659993,
+      "learning_rate": 0.00019921898635618317,
+      "loss": 2.5477,
+      "step": 1289
+    },
+    {
+      "epoch": 0.2064,
+      "grad_norm": 2.513213968774348,
+      "learning_rate": 0.00019921683019245186,
+      "loss": 2.5089,
+      "step": 1290
+    },
+    {
+      "epoch": 0.20656,
+      "grad_norm": 3.1838566452871047,
+      "learning_rate": 0.00019921467106823494,
+      "loss": 2.4607,
+      "step": 1291
+    },
+    {
+      "epoch": 0.20672,
+      "grad_norm": 3.159054847075448,
+      "learning_rate": 0.00019921250898359677,
+      "loss": 2.6784,
+      "step": 1292
+    },
+    {
+      "epoch": 0.20688,
+      "grad_norm": 3.8398332297961035,
+      "learning_rate": 0.00019921034393860195,
+      "loss": 2.6573,
+      "step": 1293
+    },
+    {
+      "epoch": 0.20704,
+      "grad_norm": 2.85005522373777,
+      "learning_rate": 0.000199208175933315,
+      "loss": 2.6248,
+      "step": 1294
+    },
+    {
+      "epoch": 0.2072,
+      "grad_norm": 3.101847338081066,
+      "learning_rate": 0.00019920600496780062,
+      "loss": 2.5008,
+      "step": 1295
+    },
+    {
+      "epoch": 0.20736,
+      "grad_norm": 3.2010863634489106,
+      "learning_rate": 0.00019920383104212364,
+      "loss": 2.5803,
+      "step": 1296
+    },
+    {
+      "epoch": 0.20752,
+      "grad_norm": 3.2922824726258435,
+      "learning_rate": 0.0001992016541563489,
+      "loss": 2.6087,
+      "step": 1297
+    },
+    {
+      "epoch": 0.20768,
+      "grad_norm": 2.7130695587867257,
+      "learning_rate": 0.00019919947431054134,
+      "loss": 2.3653,
+      "step": 1298
+    },
+    {
+      "epoch": 0.20784,
+      "grad_norm": 4.601484627997813,
+      "learning_rate": 0.00019919729150476602,
+      "loss": 2.6085,
+      "step": 1299
+    },
+    {
+      "epoch": 0.208,
+      "grad_norm": 2.6622341654409123,
+      "learning_rate": 0.00019919510573908805,
+      "loss": 2.466,
+      "step": 1300
+    },
+    {
+      "epoch": 0.20816,
+      "grad_norm": 3.3836083127197023,
+      "learning_rate": 0.0001991929170135727,
+      "loss": 2.5531,
+      "step": 1301
+    },
+    {
+      "epoch": 0.20832,
+      "grad_norm": 2.682150927737447,
+      "learning_rate": 0.00019919072532828525,
+      "loss": 2.5089,
+      "step": 1302
+    },
+    {
+      "epoch": 0.20848,
+      "grad_norm": 2.9427470527525545,
+      "learning_rate": 0.00019918853068329106,
+      "loss": 2.5408,
+      "step": 1303
+    },
+    {
+      "epoch": 0.20864,
+      "grad_norm": 2.944053208558278,
+      "learning_rate": 0.00019918633307865565,
+      "loss": 2.4357,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2088,
+      "grad_norm": 3.0365103535656432,
+      "learning_rate": 0.0001991841325144446,
+      "loss": 2.4741,
+      "step": 1305
+    },
+    {
+      "epoch": 0.20896,
+      "grad_norm": 3.307666654093006,
+      "learning_rate": 0.00019918192899072356,
+      "loss": 2.6539,
+      "step": 1306
+    },
+    {
+      "epoch": 0.20912,
+      "grad_norm": 4.078197590503433,
+      "learning_rate": 0.00019917972250755827,
+      "loss": 2.5486,
+      "step": 1307
+    },
+    {
+      "epoch": 0.20928,
+      "grad_norm": 3.5971395350456503,
+      "learning_rate": 0.0001991775130650146,
+      "loss": 2.5357,
+      "step": 1308
+    },
+    {
+      "epoch": 0.20944,
+      "grad_norm": 3.0839262075847054,
+      "learning_rate": 0.0001991753006631585,
+      "loss": 2.5018,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2096,
+      "grad_norm": 3.1017003884333074,
+      "learning_rate": 0.00019917308530205586,
+      "loss": 2.4127,
+      "step": 1310
+    },
+    {
+      "epoch": 0.20976,
+      "grad_norm": 3.4685028849731503,
+      "learning_rate": 0.0001991708669817729,
+      "loss": 2.5751,
+      "step": 1311
+    },
+    {
+      "epoch": 0.20992,
+      "grad_norm": 3.696510497626183,
+      "learning_rate": 0.00019916864570237577,
+      "loss": 2.4805,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21008,
+      "grad_norm": 3.177594664246526,
+      "learning_rate": 0.00019916642146393076,
+      "loss": 2.5134,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21024,
+      "grad_norm": 3.499000941174155,
+      "learning_rate": 0.00019916419426650424,
+      "loss": 2.457,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2104,
+      "grad_norm": 3.263653596850981,
+      "learning_rate": 0.00019916196411016266,
+      "loss": 2.5206,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21056,
+      "grad_norm": 3.4598085855485157,
+      "learning_rate": 0.00019915973099497256,
+      "loss": 2.5778,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21072,
+      "grad_norm": 2.6861209043921743,
+      "learning_rate": 0.00019915749492100058,
+      "loss": 2.4573,
+      "step": 1317
+    },
+    {
+      "epoch": 0.21088,
+      "grad_norm": 3.089078861366087,
+      "learning_rate": 0.00019915525588831345,
+      "loss": 2.4343,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21104,
+      "grad_norm": 2.856585791976678,
+      "learning_rate": 0.00019915301389697792,
+      "loss": 2.5124,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2112,
+      "grad_norm": 3.574551437905176,
+      "learning_rate": 0.000199150768947061,
+      "loss": 2.5462,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21136,
+      "grad_norm": 2.6109839882591457,
+      "learning_rate": 0.0001991485210386296,
+      "loss": 2.5045,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21152,
+      "grad_norm": 3.905777385194769,
+      "learning_rate": 0.00019914627017175078,
+      "loss": 2.5185,
+      "step": 1322
+    },
+    {
+      "epoch": 0.21168,
+      "grad_norm": 3.1602703437555686,
+      "learning_rate": 0.00019914401634649176,
+      "loss": 2.5691,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21184,
+      "grad_norm": 2.8568305024632057,
+      "learning_rate": 0.00019914175956291973,
+      "loss": 2.4863,
+      "step": 1324
+    },
+    {
+      "epoch": 0.212,
+      "grad_norm": 3.0732291783562657,
+      "learning_rate": 0.00019913949982110207,
+      "loss": 2.5386,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21216,
+      "grad_norm": 3.023058261512091,
+      "learning_rate": 0.00019913723712110623,
+      "loss": 2.6089,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21232,
+      "grad_norm": 2.6241832528055107,
+      "learning_rate": 0.00019913497146299965,
+      "loss": 2.5335,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21248,
+      "grad_norm": 3.0589092509718863,
+      "learning_rate": 0.00019913270284684998,
+      "loss": 2.5447,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21264,
+      "grad_norm": 2.8099766263598718,
+      "learning_rate": 0.00019913043127272492,
+      "loss": 2.5409,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2128,
+      "grad_norm": 3.581617302308484,
+      "learning_rate": 0.00019912815674069222,
+      "loss": 2.5722,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21296,
+      "grad_norm": 3.2666250133331944,
+      "learning_rate": 0.0001991258792508198,
+      "loss": 2.5712,
+      "step": 1331
+    },
+    {
+      "epoch": 0.21312,
+      "grad_norm": 2.8214636233830004,
+      "learning_rate": 0.00019912359880317558,
+      "loss": 2.4194,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21328,
+      "grad_norm": 2.4695358159458576,
+      "learning_rate": 0.00019912131539782759,
+      "loss": 2.5839,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21344,
+      "grad_norm": 3.203891068931345,
+      "learning_rate": 0.000199119029034844,
+      "loss": 2.4881,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2136,
+      "grad_norm": 2.9941121583492545,
+      "learning_rate": 0.00019911673971429302,
+      "loss": 2.5653,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21376,
+      "grad_norm": 3.587713466309748,
+      "learning_rate": 0.00019911444743624291,
+      "loss": 2.3746,
+      "step": 1336
+    },
+    {
+      "epoch": 0.21392,
+      "grad_norm": 3.1542585762735555,
+      "learning_rate": 0.00019911215220076215,
+      "loss": 2.4719,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21408,
+      "grad_norm": 2.8297157873101098,
+      "learning_rate": 0.0001991098540079192,
+      "loss": 2.4438,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21424,
+      "grad_norm": 2.9917617811233383,
+      "learning_rate": 0.0001991075528577826,
+      "loss": 2.4374,
+      "step": 1339
+    },
+    {
+      "epoch": 0.2144,
+      "grad_norm": 3.088726774044467,
+      "learning_rate": 0.00019910524875042107,
+      "loss": 2.5227,
+      "step": 1340
+    },
+    {
+      "epoch": 0.21456,
+      "grad_norm": 3.0764487655848454,
+      "learning_rate": 0.0001991029416859033,
+      "loss": 2.3289,
+      "step": 1341
+    },
+    {
+      "epoch": 0.21472,
+      "grad_norm": 2.9113905993927527,
+      "learning_rate": 0.00019910063166429817,
+      "loss": 2.5389,
+      "step": 1342
+    },
+    {
+      "epoch": 0.21488,
+      "grad_norm": 2.562300888672682,
+      "learning_rate": 0.0001990983186856746,
+      "loss": 2.5391,
+      "step": 1343
+    },
+    {
+      "epoch": 0.21504,
+      "grad_norm": 3.1146450067807914,
+      "learning_rate": 0.00019909600275010158,
+      "loss": 2.4721,
+      "step": 1344
+    },
+    {
+      "epoch": 0.2152,
+      "grad_norm": 3.039879699934466,
+      "learning_rate": 0.00019909368385764826,
+      "loss": 2.5882,
+      "step": 1345
+    },
+    {
+      "epoch": 0.21536,
+      "grad_norm": 3.8437455858919045,
+      "learning_rate": 0.0001990913620083838,
+      "loss": 2.5288,
+      "step": 1346
+    },
+    {
+      "epoch": 0.21552,
+      "grad_norm": 3.6340995096020534,
+      "learning_rate": 0.0001990890372023775,
+      "loss": 2.5065,
+      "step": 1347
+    },
+    {
+      "epoch": 0.21568,
+      "grad_norm": 4.459858037944732,
+      "learning_rate": 0.0001990867094396987,
+      "loss": 2.4435,
+      "step": 1348
+    },
+    {
+      "epoch": 0.21584,
+      "grad_norm": 3.1004215598650786,
+      "learning_rate": 0.00019908437872041685,
+      "loss": 2.5075,
+      "step": 1349
+    },
+    {
+      "epoch": 0.216,
+      "grad_norm": 4.179097657078835,
+      "learning_rate": 0.00019908204504460157,
+      "loss": 2.5545,
+      "step": 1350
+    },
+    {
+      "epoch": 0.21616,
+      "grad_norm": 2.8727735533082814,
+      "learning_rate": 0.0001990797084123224,
+      "loss": 2.6242,
+      "step": 1351
+    },
+    {
+      "epoch": 0.21632,
+      "grad_norm": 4.03461762001803,
+      "learning_rate": 0.00019907736882364914,
+      "loss": 2.6477,
+      "step": 1352
+    },
+    {
+      "epoch": 0.21648,
+      "grad_norm": 2.812229266315407,
+      "learning_rate": 0.00019907502627865156,
+      "loss": 2.5612,
+      "step": 1353
+    },
+    {
+      "epoch": 0.21664,
+      "grad_norm": 4.193906547418484,
+      "learning_rate": 0.00019907268077739958,
+      "loss": 2.5959,
+      "step": 1354
+    },
+    {
+      "epoch": 0.2168,
+      "grad_norm": 3.0315032246146267,
+      "learning_rate": 0.00019907033231996314,
+      "loss": 2.553,
+      "step": 1355
+    },
+    {
+      "epoch": 0.21696,
+      "grad_norm": 4.274212694256406,
+      "learning_rate": 0.00019906798090641234,
+      "loss": 2.5186,
+      "step": 1356
+    },
+    {
+      "epoch": 0.21712,
+      "grad_norm": 2.7322762933806928,
+      "learning_rate": 0.00019906562653681736,
+      "loss": 2.5236,
+      "step": 1357
+    },
+    {
+      "epoch": 0.21728,
+      "grad_norm": 3.6488649191758316,
+      "learning_rate": 0.00019906326921124846,
+      "loss": 2.5705,
+      "step": 1358
+    },
+    {
+      "epoch": 0.21744,
+      "grad_norm": 2.5873569607646503,
+      "learning_rate": 0.0001990609089297759,
+      "loss": 2.5952,
+      "step": 1359
+    },
+    {
+      "epoch": 0.2176,
+      "grad_norm": 4.293387360372495,
+      "learning_rate": 0.00019905854569247022,
+      "loss": 2.5954,
+      "step": 1360
+    },
+    {
+      "epoch": 0.21776,
+      "grad_norm": 3.514923185144505,
+      "learning_rate": 0.0001990561794994019,
+      "loss": 2.5448,
+      "step": 1361
+    },
+    {
+      "epoch": 0.21792,
+      "grad_norm": 4.617273586760779,
+      "learning_rate": 0.00019905381035064147,
+      "loss": 2.5397,
+      "step": 1362
+    },
+    {
+      "epoch": 0.21808,
+      "grad_norm": 3.0463533331065333,
+      "learning_rate": 0.00019905143824625968,
+      "loss": 2.5195,
+      "step": 1363
+    },
+    {
+      "epoch": 0.21824,
+      "grad_norm": 3.6159462019326316,
+      "learning_rate": 0.00019904906318632732,
+      "loss": 2.485,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2184,
+      "grad_norm": 2.972758614867922,
+      "learning_rate": 0.00019904668517091523,
+      "loss": 2.4996,
+      "step": 1365
+    },
+    {
+      "epoch": 0.21856,
+      "grad_norm": 3.145555527312444,
+      "learning_rate": 0.0001990443042000944,
+      "loss": 2.4246,
+      "step": 1366
+    },
+    {
+      "epoch": 0.21872,
+      "grad_norm": 2.6855079955841274,
+      "learning_rate": 0.00019904192027393584,
+      "loss": 2.4212,
+      "step": 1367
+    },
+    {
+      "epoch": 0.21888,
+      "grad_norm": 3.5709236535773097,
+      "learning_rate": 0.00019903953339251072,
+      "loss": 2.4329,
+      "step": 1368
+    },
+    {
+      "epoch": 0.21904,
+      "grad_norm": 3.020989604293351,
+      "learning_rate": 0.0001990371435558902,
+      "loss": 2.4719,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2192,
+      "grad_norm": 2.715636405261382,
+      "learning_rate": 0.00019903475076414563,
+      "loss": 2.5252,
+      "step": 1370
+    },
+    {
+      "epoch": 0.21936,
+      "grad_norm": 2.5757365894741304,
+      "learning_rate": 0.00019903235501734845,
+      "loss": 2.4426,
+      "step": 1371
+    },
+    {
+      "epoch": 0.21952,
+      "grad_norm": 2.55601269159237,
+      "learning_rate": 0.0001990299563155701,
+      "loss": 2.4695,
+      "step": 1372
+    },
+    {
+      "epoch": 0.21968,
+      "grad_norm": 3.00418227197036,
+      "learning_rate": 0.00019902755465888211,
+      "loss": 2.543,
+      "step": 1373
+    },
+    {
+      "epoch": 0.21984,
+      "grad_norm": 2.585874805440453,
+      "learning_rate": 0.0001990251500473562,
+      "loss": 2.3696,
+      "step": 1374
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 3.395137852217611,
+      "learning_rate": 0.00019902274248106413,
+      "loss": 2.5258,
+      "step": 1375
+    },
+    {
+      "epoch": 0.22016,
+      "grad_norm": 3.182205409963505,
+      "learning_rate": 0.00019902033196007773,
+      "loss": 2.5999,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22032,
+      "grad_norm": 3.4280102582943472,
+      "learning_rate": 0.00019901791848446888,
+      "loss": 2.5028,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22048,
+      "grad_norm": 3.439558742013519,
+      "learning_rate": 0.00019901550205430964,
+      "loss": 2.5292,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22064,
+      "grad_norm": 3.3003146417354574,
+      "learning_rate": 0.00019901308266967208,
+      "loss": 2.4301,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2208,
+      "grad_norm": 3.260322788924006,
+      "learning_rate": 0.00019901066033062845,
+      "loss": 2.4624,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22096,
+      "grad_norm": 3.3889031479191845,
+      "learning_rate": 0.00019900823503725095,
+      "loss": 2.4615,
+      "step": 1381
+    },
+    {
+      "epoch": 0.22112,
+      "grad_norm": 2.858780375366984,
+      "learning_rate": 0.00019900580678961202,
+      "loss": 2.5667,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22128,
+      "grad_norm": 3.1958086600909583,
+      "learning_rate": 0.00019900337558778402,
+      "loss": 2.5691,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22144,
+      "grad_norm": 3.4431663030090958,
+      "learning_rate": 0.00019900094143183962,
+      "loss": 2.5333,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2216,
+      "grad_norm": 3.913407637334283,
+      "learning_rate": 0.00019899850432185138,
+      "loss": 2.5333,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22176,
+      "grad_norm": 2.7522487766052945,
+      "learning_rate": 0.00019899606425789198,
+      "loss": 2.4306,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22192,
+      "grad_norm": 3.29906108816268,
+      "learning_rate": 0.00019899362124003433,
+      "loss": 2.507,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22208,
+      "grad_norm": 2.7650171257539453,
+      "learning_rate": 0.00019899117526835124,
+      "loss": 2.5329,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22224,
+      "grad_norm": 2.40161070730834,
+      "learning_rate": 0.00019898872634291573,
+      "loss": 2.4323,
+      "step": 1389
+    },
+    {
+      "epoch": 0.2224,
+      "grad_norm": 3.025761611923319,
+      "learning_rate": 0.0001989862744638009,
+      "loss": 2.4213,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22256,
+      "grad_norm": 2.8440645900991046,
+      "learning_rate": 0.00019898381963107985,
+      "loss": 2.4723,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22272,
+      "grad_norm": 2.8194395341675484,
+      "learning_rate": 0.00019898136184482584,
+      "loss": 2.472,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22288,
+      "grad_norm": 2.9657988871336536,
+      "learning_rate": 0.00019897890110511226,
+      "loss": 2.5907,
+      "step": 1393
+    },
+    {
+      "epoch": 0.22304,
+      "grad_norm": 2.929012772408513,
+      "learning_rate": 0.0001989764374120125,
+      "loss": 2.4436,
+      "step": 1394
+    },
+    {
+      "epoch": 0.2232,
+      "grad_norm": 3.7353178588742897,
+      "learning_rate": 0.00019897397076560003,
+      "loss": 2.521,
+      "step": 1395
+    },
+    {
+      "epoch": 0.22336,
+      "grad_norm": 2.755347946728196,
+      "learning_rate": 0.00019897150116594852,
+      "loss": 2.4443,
+      "step": 1396
+    },
+    {
+      "epoch": 0.22352,
+      "grad_norm": 4.932514295475342,
+      "learning_rate": 0.0001989690286131316,
+      "loss": 2.5284,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22368,
+      "grad_norm": 2.894554711250316,
+      "learning_rate": 0.00019896655310722313,
+      "loss": 2.423,
+      "step": 1398
+    },
+    {
+      "epoch": 0.22384,
+      "grad_norm": 3.419504106977435,
+      "learning_rate": 0.0001989640746482969,
+      "loss": 2.4332,
+      "step": 1399
+    },
+    {
+      "epoch": 0.224,
+      "grad_norm": 3.059153087229128,
+      "learning_rate": 0.00019896159323642686,
+      "loss": 2.5196,
+      "step": 1400
+    },
+    {
+      "epoch": 0.22416,
+      "grad_norm": 3.0053672124634634,
+      "learning_rate": 0.00019895910887168713,
+      "loss": 2.445,
+      "step": 1401
+    },
+    {
+      "epoch": 0.22432,
+      "grad_norm": 3.781518656683629,
+      "learning_rate": 0.00019895662155415176,
+      "loss": 2.5128,
+      "step": 1402
+    },
+    {
+      "epoch": 0.22448,
+      "grad_norm": 3.9036917779620457,
+      "learning_rate": 0.00019895413128389502,
+      "loss": 2.5535,
+      "step": 1403
+    },
+    {
+      "epoch": 0.22464,
+      "grad_norm": 2.659706989929669,
+      "learning_rate": 0.00019895163806099117,
+      "loss": 2.4306,
+      "step": 1404
+    },
+    {
+      "epoch": 0.2248,
+      "grad_norm": 4.404153424821734,
+      "learning_rate": 0.00019894914188551464,
+      "loss": 2.5202,
+      "step": 1405
+    },
+    {
+      "epoch": 0.22496,
+      "grad_norm": 3.187834876638864,
+      "learning_rate": 0.0001989466427575399,
+      "loss": 2.3797,
+      "step": 1406
+    },
+    {
+      "epoch": 0.22512,
+      "grad_norm": 4.6479346314373755,
+      "learning_rate": 0.0001989441406771415,
+      "loss": 2.401,
+      "step": 1407
+    },
+    {
+      "epoch": 0.22528,
+      "grad_norm": 2.896632081953289,
+      "learning_rate": 0.00019894163564439417,
+      "loss": 2.449,
+      "step": 1408
+    },
+    {
+      "epoch": 0.22544,
+      "grad_norm": 3.4265004094909868,
+      "learning_rate": 0.00019893912765937258,
+      "loss": 2.5253,
+      "step": 1409
+    },
+    {
+      "epoch": 0.2256,
+      "grad_norm": 3.345735135777088,
+      "learning_rate": 0.00019893661672215157,
+      "loss": 2.4905,
+      "step": 1410
+    },
+    {
+      "epoch": 0.22576,
+      "grad_norm": 3.153601861640162,
+      "learning_rate": 0.0001989341028328061,
+      "loss": 2.5491,
+      "step": 1411
+    },
+    {
+      "epoch": 0.22592,
+      "grad_norm": 3.308973963036557,
+      "learning_rate": 0.00019893158599141114,
+      "loss": 2.4371,
+      "step": 1412
+    },
+    {
+      "epoch": 0.22608,
+      "grad_norm": 2.9679761965237406,
+      "learning_rate": 0.00019892906619804185,
+      "loss": 2.4569,
+      "step": 1413
+    },
+    {
+      "epoch": 0.22624,
+      "grad_norm": 3.0432738878913392,
+      "learning_rate": 0.00019892654345277338,
+      "loss": 2.4537,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2264,
+      "grad_norm": 2.647555225839851,
+      "learning_rate": 0.000198924017755681,
+      "loss": 2.6128,
+      "step": 1415
+    },
+    {
+      "epoch": 0.22656,
+      "grad_norm": 2.7230236329259,
+      "learning_rate": 0.00019892148910684006,
+      "loss": 2.3646,
+      "step": 1416
+    },
+    {
+      "epoch": 0.22672,
+      "grad_norm": 2.9395993486459777,
+      "learning_rate": 0.00019891895750632604,
+      "loss": 2.4206,
+      "step": 1417
+    },
+    {
+      "epoch": 0.22688,
+      "grad_norm": 3.072907996561123,
+      "learning_rate": 0.00019891642295421445,
+      "loss": 2.5059,
+      "step": 1418
+    },
+    {
+      "epoch": 0.22704,
+      "grad_norm": 2.52090788171936,
+      "learning_rate": 0.00019891388545058094,
+      "loss": 2.4353,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2272,
+      "grad_norm": 2.0991555827051176,
+      "learning_rate": 0.0001989113449955012,
+      "loss": 2.342,
+      "step": 1420
+    },
+    {
+      "epoch": 0.22736,
+      "grad_norm": 2.770315612368407,
+      "learning_rate": 0.0001989088015890511,
+      "loss": 2.4295,
+      "step": 1421
+    },
+    {
+      "epoch": 0.22752,
+      "grad_norm": 2.869124516160472,
+      "learning_rate": 0.0001989062552313065,
+      "loss": 2.3823,
+      "step": 1422
+    },
+    {
+      "epoch": 0.22768,
+      "grad_norm": 2.256621979967423,
+      "learning_rate": 0.00019890370592234332,
+      "loss": 2.4358,
+      "step": 1423
+    },
+    {
+      "epoch": 0.22784,
+      "grad_norm": 2.8818358011627034,
+      "learning_rate": 0.0001989011536622377,
+      "loss": 2.4632,
+      "step": 1424
+    },
+    {
+      "epoch": 0.228,
+      "grad_norm": 2.4807807880034054,
+      "learning_rate": 0.00019889859845106575,
+      "loss": 2.3425,
+      "step": 1425
+    },
+    {
+      "epoch": 0.22816,
+      "grad_norm": 2.550213386782815,
+      "learning_rate": 0.00019889604028890375,
+      "loss": 2.4664,
+      "step": 1426
+    },
+    {
+      "epoch": 0.22832,
+      "grad_norm": 3.2368221809089976,
+      "learning_rate": 0.00019889347917582799,
+      "loss": 2.4971,
+      "step": 1427
+    },
+    {
+      "epoch": 0.22848,
+      "grad_norm": 2.613116237516798,
+      "learning_rate": 0.00019889091511191494,
+      "loss": 2.4312,
+      "step": 1428
+    },
+    {
+      "epoch": 0.22864,
+      "grad_norm": 2.847120773726172,
+      "learning_rate": 0.00019888834809724104,
+      "loss": 2.2841,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2288,
+      "grad_norm": 2.9071689498053743,
+      "learning_rate": 0.00019888577813188295,
+      "loss": 2.3121,
+      "step": 1430
+    },
+    {
+      "epoch": 0.22896,
+      "grad_norm": 2.9509939243663204,
+      "learning_rate": 0.00019888320521591738,
+      "loss": 2.4868,
+      "step": 1431
+    },
+    {
+      "epoch": 0.22912,
+      "grad_norm": 2.7129093779553592,
+      "learning_rate": 0.000198880629349421,
+      "loss": 2.3834,
+      "step": 1432
+    },
+    {
+      "epoch": 0.22928,
+      "grad_norm": 2.668202166021356,
+      "learning_rate": 0.00019887805053247072,
+      "loss": 2.4243,
+      "step": 1433
+    },
+    {
+      "epoch": 0.22944,
+      "grad_norm": 3.502590953097065,
+      "learning_rate": 0.00019887546876514353,
+      "loss": 2.5322,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2296,
+      "grad_norm": 2.924534656351701,
+      "learning_rate": 0.00019887288404751643,
+      "loss": 2.4763,
+      "step": 1435
+    },
+    {
+      "epoch": 0.22976,
+      "grad_norm": 3.4446876847578474,
+      "learning_rate": 0.00019887029637966652,
+      "loss": 2.5084,
+      "step": 1436
+    },
+    {
+      "epoch": 0.22992,
+      "grad_norm": 2.9683182098553593,
+      "learning_rate": 0.000198867705761671,
+      "loss": 2.4358,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23008,
+      "grad_norm": 2.7677684968236407,
+      "learning_rate": 0.00019886511219360727,
+      "loss": 2.4695,
+      "step": 1438
+    },
+    {
+      "epoch": 0.23024,
+      "grad_norm": 3.1073961659979785,
+      "learning_rate": 0.00019886251567555264,
+      "loss": 2.4525,
+      "step": 1439
+    },
+    {
+      "epoch": 0.2304,
+      "grad_norm": 2.7271641712811263,
+      "learning_rate": 0.0001988599162075846,
+      "loss": 2.4626,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23056,
+      "grad_norm": 2.66904199122754,
+      "learning_rate": 0.0001988573137897807,
+      "loss": 2.433,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23072,
+      "grad_norm": 3.3435159408010664,
+      "learning_rate": 0.0001988547084222186,
+      "loss": 2.3623,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23088,
+      "grad_norm": 2.8655845790736914,
+      "learning_rate": 0.00019885210010497605,
+      "loss": 2.418,
+      "step": 1443
+    },
+    {
+      "epoch": 0.23104,
+      "grad_norm": 2.6569917226150315,
+      "learning_rate": 0.00019884948883813088,
+      "loss": 2.4478,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2312,
+      "grad_norm": 2.912605429684545,
+      "learning_rate": 0.000198846874621761,
+      "loss": 2.4628,
+      "step": 1445
+    },
+    {
+      "epoch": 0.23136,
+      "grad_norm": 3.006278033656192,
+      "learning_rate": 0.00019884425745594447,
+      "loss": 2.3472,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23152,
+      "grad_norm": 3.3092199815438548,
+      "learning_rate": 0.00019884163734075925,
+      "loss": 2.2973,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23168,
+      "grad_norm": 3.0038143342010186,
+      "learning_rate": 0.00019883901427628363,
+      "loss": 2.4797,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23184,
+      "grad_norm": 3.3145725319389143,
+      "learning_rate": 0.00019883638826259585,
+      "loss": 2.5458,
+      "step": 1449
+    },
+    {
+      "epoch": 0.232,
+      "grad_norm": 2.7342197834817394,
+      "learning_rate": 0.00019883375929977425,
+      "loss": 2.5069,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23216,
+      "grad_norm": 2.9474400056520786,
+      "learning_rate": 0.00019883112738789733,
+      "loss": 2.3742,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23232,
+      "grad_norm": 3.693230693579856,
+      "learning_rate": 0.00019882849252704355,
+      "loss": 2.3829,
+      "step": 1452
+    },
+    {
+      "epoch": 0.23248,
+      "grad_norm": 3.045894596428267,
+      "learning_rate": 0.00019882585471729157,
+      "loss": 2.4079,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23264,
+      "grad_norm": 3.287959618686943,
+      "learning_rate": 0.00019882321395872006,
+      "loss": 2.3925,
+      "step": 1454
+    },
+    {
+      "epoch": 0.2328,
+      "grad_norm": 2.9187233972082147,
+      "learning_rate": 0.00019882057025140789,
+      "loss": 2.5013,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23296,
+      "grad_norm": 3.5495484916050164,
+      "learning_rate": 0.00019881792359543386,
+      "loss": 2.4863,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23312,
+      "grad_norm": 4.385265005535355,
+      "learning_rate": 0.000198815273990877,
+      "loss": 2.3734,
+      "step": 1457
+    },
+    {
+      "epoch": 0.23328,
+      "grad_norm": 2.7457795673693846,
+      "learning_rate": 0.00019881262143781636,
+      "loss": 2.5661,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23344,
+      "grad_norm": 4.131408268790814,
+      "learning_rate": 0.00019880996593633107,
+      "loss": 2.3677,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2336,
+      "grad_norm": 3.779713113421245,
+      "learning_rate": 0.0001988073074865004,
+      "loss": 2.4691,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23376,
+      "grad_norm": 3.681078292357638,
+      "learning_rate": 0.00019880464608840363,
+      "loss": 2.3617,
+      "step": 1461
+    },
+    {
+      "epoch": 0.23392,
+      "grad_norm": 4.307718611115387,
+      "learning_rate": 0.00019880198174212018,
+      "loss": 2.4619,
+      "step": 1462
+    },
+    {
+      "epoch": 0.23408,
+      "grad_norm": 3.7197664275753475,
+      "learning_rate": 0.00019879931444772962,
+      "loss": 2.3865,
+      "step": 1463
+    },
+    {
+      "epoch": 0.23424,
+      "grad_norm": 3.4269643113515618,
+      "learning_rate": 0.00019879664420531144,
+      "loss": 2.5124,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2344,
+      "grad_norm": 3.1629421643697015,
+      "learning_rate": 0.00019879397101494535,
+      "loss": 2.4016,
+      "step": 1465
+    },
+    {
+      "epoch": 0.23456,
+      "grad_norm": 2.9351772511796153,
+      "learning_rate": 0.00019879129487671116,
+      "loss": 2.3889,
+      "step": 1466
+    },
+    {
+      "epoch": 0.23472,
+      "grad_norm": 2.3024853435024943,
+      "learning_rate": 0.00019878861579068866,
+      "loss": 2.428,
+      "step": 1467
+    },
+    {
+      "epoch": 0.23488,
+      "grad_norm": 3.2015104302446584,
+      "learning_rate": 0.0001987859337569578,
+      "loss": 2.4231,
+      "step": 1468
+    },
+    {
+      "epoch": 0.23504,
+      "grad_norm": 4.359154690191569,
+      "learning_rate": 0.00019878324877559866,
+      "loss": 2.5184,
+      "step": 1469
+    },
+    {
+      "epoch": 0.2352,
+      "grad_norm": 2.7936652495919905,
+      "learning_rate": 0.0001987805608466913,
+      "loss": 2.463,
+      "step": 1470
+    },
+    {
+      "epoch": 0.23536,
+      "grad_norm": 3.354272808650202,
+      "learning_rate": 0.00019877786997031594,
+      "loss": 2.4159,
+      "step": 1471
+    },
+    {
+      "epoch": 0.23552,
+      "grad_norm": 3.6872273519295864,
+      "learning_rate": 0.00019877517614655288,
+      "loss": 2.4377,
+      "step": 1472
+    },
+    {
+      "epoch": 0.23568,
+      "grad_norm": 3.2972428080088254,
+      "learning_rate": 0.0001987724793754825,
+      "loss": 2.3848,
+      "step": 1473
+    },
+    {
+      "epoch": 0.23584,
+      "grad_norm": 3.2586226193610925,
+      "learning_rate": 0.00019876977965718526,
+      "loss": 2.4427,
+      "step": 1474
+    },
+    {
+      "epoch": 0.236,
+      "grad_norm": 3.158268371801861,
+      "learning_rate": 0.0001987670769917417,
+      "loss": 2.3751,
+      "step": 1475
+    },
+    {
+      "epoch": 0.23616,
+      "grad_norm": 3.746931552368218,
+      "learning_rate": 0.00019876437137923253,
+      "loss": 2.2904,
+      "step": 1476
+    },
+    {
+      "epoch": 0.23632,
+      "grad_norm": 4.107627995699095,
+      "learning_rate": 0.00019876166281973835,
+      "loss": 2.4177,
+      "step": 1477
+    },
+    {
+      "epoch": 0.23648,
+      "grad_norm": 2.8450781048927576,
+      "learning_rate": 0.00019875895131334014,
+      "loss": 2.5362,
+      "step": 1478
+    },
+    {
+      "epoch": 0.23664,
+      "grad_norm": 4.2665883009711685,
+      "learning_rate": 0.0001987562368601187,
+      "loss": 2.3666,
+      "step": 1479
+    },
+    {
+      "epoch": 0.2368,
+      "grad_norm": 2.879354935946796,
+      "learning_rate": 0.00019875351946015507,
+      "loss": 2.4539,
+      "step": 1480
+    },
+    {
+      "epoch": 0.23696,
+      "grad_norm": 4.075415487081278,
+      "learning_rate": 0.00019875079911353026,
+      "loss": 2.3217,
+      "step": 1481
+    },
+    {
+      "epoch": 0.23712,
+      "grad_norm": 2.504451909540309,
+      "learning_rate": 0.00019874807582032558,
+      "loss": 2.3282,
+      "step": 1482
+    },
+    {
+      "epoch": 0.23728,
+      "grad_norm": 4.3592898201781525,
+      "learning_rate": 0.00019874534958062216,
+      "loss": 2.4625,
+      "step": 1483
+    },
+    {
+      "epoch": 0.23744,
+      "grad_norm": 2.659819073076315,
+      "learning_rate": 0.0001987426203945014,
+      "loss": 2.3359,
+      "step": 1484
+    },
+    {
+      "epoch": 0.2376,
+      "grad_norm": 3.455252246872302,
+      "learning_rate": 0.00019873988826204475,
+      "loss": 2.3179,
+      "step": 1485
+    },
+    {
+      "epoch": 0.23776,
+      "grad_norm": 2.3768306765924114,
+      "learning_rate": 0.0001987371531833337,
+      "loss": 2.3519,
+      "step": 1486
+    },
+    {
+      "epoch": 0.23792,
+      "grad_norm": 4.225645874414721,
+      "learning_rate": 0.00019873441515844988,
+      "loss": 2.4055,
+      "step": 1487
+    },
+    {
+      "epoch": 0.23808,
+      "grad_norm": 2.8298209702409145,
+      "learning_rate": 0.00019873167418747498,
+      "loss": 2.4799,
+      "step": 1488
+    },
+    {
+      "epoch": 0.23824,
+      "grad_norm": 3.645285957117695,
+      "learning_rate": 0.0001987289302704908,
+      "loss": 2.3612,
+      "step": 1489
+    },
+    {
+      "epoch": 0.2384,
+      "grad_norm": 3.0471908289112353,
+      "learning_rate": 0.00019872618340757917,
+      "loss": 2.4621,
+      "step": 1490
+    },
+    {
+      "epoch": 0.23856,
+      "grad_norm": 2.8440200420075823,
+      "learning_rate": 0.00019872343359882212,
+      "loss": 2.5696,
+      "step": 1491
+    },
+    {
+      "epoch": 0.23872,
+      "grad_norm": 2.8676390728150967,
+      "learning_rate": 0.00019872068084430167,
+      "loss": 2.457,
+      "step": 1492
+    },
+    {
+      "epoch": 0.23888,
+      "grad_norm": 2.415656284372818,
+      "learning_rate": 0.00019871792514409992,
+      "loss": 2.2336,
+      "step": 1493
+    },
+    {
+      "epoch": 0.23904,
+      "grad_norm": 2.41376702007706,
+      "learning_rate": 0.00019871516649829917,
+      "loss": 2.382,
+      "step": 1494
+    },
+    {
+      "epoch": 0.2392,
+      "grad_norm": 3.275011889922767,
+      "learning_rate": 0.00019871240490698167,
+      "loss": 2.4429,
+      "step": 1495
+    },
+    {
+      "epoch": 0.23936,
+      "grad_norm": 3.0357898936712053,
+      "learning_rate": 0.00019870964037022986,
+      "loss": 2.4371,
+      "step": 1496
+    },
+    {
+      "epoch": 0.23952,
+      "grad_norm": 2.6922658273142526,
+      "learning_rate": 0.0001987068728881262,
+      "loss": 2.5197,
+      "step": 1497
+    },
+    {
+      "epoch": 0.23968,
+      "grad_norm": 2.7957445023206526,
+      "learning_rate": 0.0001987041024607533,
+      "loss": 2.4103,
+      "step": 1498
+    },
+    {
+      "epoch": 0.23984,
+      "grad_norm": 2.7153540057783587,
+      "learning_rate": 0.00019870132908819378,
+      "loss": 2.4339,
+      "step": 1499
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 3.5252076974132924,
+      "learning_rate": 0.00019869855277053047,
+      "loss": 2.4274,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24016,
+      "grad_norm": 3.208152969261027,
+      "learning_rate": 0.0001986957735078461,
+      "loss": 2.3034,
+      "step": 1501
+    },
+    {
+      "epoch": 0.24032,
+      "grad_norm": 2.907934538911683,
+      "learning_rate": 0.00019869299130022375,
+      "loss": 2.4015,
+      "step": 1502
+    },
+    {
+      "epoch": 0.24048,
+      "grad_norm": 2.7680251427266667,
+      "learning_rate": 0.0001986902061477463,
+      "loss": 2.4249,
+      "step": 1503
+    },
+    {
+      "epoch": 0.24064,
+      "grad_norm": 3.066231271244863,
+      "learning_rate": 0.0001986874180504969,
+      "loss": 2.4981,
+      "step": 1504
+    },
+    {
+      "epoch": 0.2408,
+      "grad_norm": 3.196915241870709,
+      "learning_rate": 0.00019868462700855876,
+      "loss": 2.4706,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24096,
+      "grad_norm": 2.863314977593761,
+      "learning_rate": 0.00019868183302201514,
+      "loss": 2.3278,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24112,
+      "grad_norm": 2.6149058451234137,
+      "learning_rate": 0.00019867903609094946,
+      "loss": 2.483,
+      "step": 1507
+    },
+    {
+      "epoch": 0.24128,
+      "grad_norm": 2.986721973733325,
+      "learning_rate": 0.00019867623621544508,
+      "loss": 2.4747,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24144,
+      "grad_norm": 2.676839267183484,
+      "learning_rate": 0.00019867343339558563,
+      "loss": 2.3855,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2416,
+      "grad_norm": 2.8122669300246756,
+      "learning_rate": 0.00019867062763145467,
+      "loss": 2.3419,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24176,
+      "grad_norm": 2.606312253536779,
+      "learning_rate": 0.00019866781892313598,
+      "loss": 2.424,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24192,
+      "grad_norm": 2.9574303198186533,
+      "learning_rate": 0.00019866500727071335,
+      "loss": 2.4068,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24208,
+      "grad_norm": 2.6253198311247035,
+      "learning_rate": 0.0001986621926742707,
+      "loss": 2.3725,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24224,
+      "grad_norm": 2.926373218977815,
+      "learning_rate": 0.00019865937513389195,
+      "loss": 2.3232,
+      "step": 1514
+    },
+    {
+      "epoch": 0.2424,
+      "grad_norm": 2.1653742061138765,
+      "learning_rate": 0.00019865655464966123,
+      "loss": 2.3682,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24256,
+      "grad_norm": 2.464207322969639,
+      "learning_rate": 0.00019865373122166267,
+      "loss": 2.3718,
+      "step": 1516
+    },
+    {
+      "epoch": 0.24272,
+      "grad_norm": 3.452315208279471,
+      "learning_rate": 0.0001986509048499805,
+      "loss": 2.4508,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24288,
+      "grad_norm": 2.7000977837003313,
+      "learning_rate": 0.0001986480755346991,
+      "loss": 2.3944,
+      "step": 1518
+    },
+    {
+      "epoch": 0.24304,
+      "grad_norm": 2.8986049271117555,
+      "learning_rate": 0.00019864524327590287,
+      "loss": 2.5462,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2432,
+      "grad_norm": 3.0339347065486773,
+      "learning_rate": 0.00019864240807367634,
+      "loss": 2.4006,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24336,
+      "grad_norm": 2.6996949777387544,
+      "learning_rate": 0.00019863956992810405,
+      "loss": 2.3457,
+      "step": 1521
+    },
+    {
+      "epoch": 0.24352,
+      "grad_norm": 2.392371111728286,
+      "learning_rate": 0.00019863672883927076,
+      "loss": 2.3072,
+      "step": 1522
+    },
+    {
+      "epoch": 0.24368,
+      "grad_norm": 2.854357349269475,
+      "learning_rate": 0.0001986338848072612,
+      "loss": 2.4794,
+      "step": 1523
+    },
+    {
+      "epoch": 0.24384,
+      "grad_norm": 3.1046147703097535,
+      "learning_rate": 0.00019863103783216026,
+      "loss": 2.2896,
+      "step": 1524
+    },
+    {
+      "epoch": 0.244,
+      "grad_norm": 2.777022481559943,
+      "learning_rate": 0.00019862818791405284,
+      "loss": 2.3397,
+      "step": 1525
+    },
+    {
+      "epoch": 0.24416,
+      "grad_norm": 3.344642355043343,
+      "learning_rate": 0.00019862533505302402,
+      "loss": 2.2509,
+      "step": 1526
+    },
+    {
+      "epoch": 0.24432,
+      "grad_norm": 2.4756838554611886,
+      "learning_rate": 0.00019862247924915894,
+      "loss": 2.3932,
+      "step": 1527
+    },
+    {
+      "epoch": 0.24448,
+      "grad_norm": 3.13322901510755,
+      "learning_rate": 0.00019861962050254274,
+      "loss": 2.3608,
+      "step": 1528
+    },
+    {
+      "epoch": 0.24464,
+      "grad_norm": 2.738283511319363,
+      "learning_rate": 0.0001986167588132608,
+      "loss": 2.4304,
+      "step": 1529
+    },
+    {
+      "epoch": 0.2448,
+      "grad_norm": 2.861168436486425,
+      "learning_rate": 0.00019861389418139846,
+      "loss": 2.2993,
+      "step": 1530
+    },
+    {
+      "epoch": 0.24496,
+      "grad_norm": 2.942291729970459,
+      "learning_rate": 0.00019861102660704126,
+      "loss": 2.3165,
+      "step": 1531
+    },
+    {
+      "epoch": 0.24512,
+      "grad_norm": 3.035678833610755,
+      "learning_rate": 0.00019860815609027465,
+      "loss": 2.4532,
+      "step": 1532
+    },
+    {
+      "epoch": 0.24528,
+      "grad_norm": 2.7536343561203362,
+      "learning_rate": 0.00019860528263118438,
+      "loss": 2.391,
+      "step": 1533
+    },
+    {
+      "epoch": 0.24544,
+      "grad_norm": 2.563030502883996,
+      "learning_rate": 0.00019860240622985618,
+      "loss": 2.4125,
+      "step": 1534
+    },
+    {
+      "epoch": 0.2456,
+      "grad_norm": 2.297144088362333,
+      "learning_rate": 0.00019859952688637585,
+      "loss": 2.3225,
+      "step": 1535
+    },
+    {
+      "epoch": 0.24576,
+      "grad_norm": 2.6076931379690085,
+      "learning_rate": 0.0001985966446008293,
+      "loss": 2.3336,
+      "step": 1536
+    },
+    {
+      "epoch": 0.24592,
+      "grad_norm": 2.9220550086898927,
+      "learning_rate": 0.00019859375937330255,
+      "loss": 2.5095,
+      "step": 1537
+    },
+    {
+      "epoch": 0.24608,
+      "grad_norm": 3.267715735118693,
+      "learning_rate": 0.00019859087120388167,
+      "loss": 2.2764,
+      "step": 1538
+    },
+    {
+      "epoch": 0.24624,
+      "grad_norm": 3.32476803223203,
+      "learning_rate": 0.00019858798009265286,
+      "loss": 2.4483,
+      "step": 1539
+    },
+    {
+      "epoch": 0.2464,
+      "grad_norm": 2.32181245005628,
+      "learning_rate": 0.0001985850860397024,
+      "loss": 2.4887,
+      "step": 1540
+    },
+    {
+      "epoch": 0.24656,
+      "grad_norm": 2.92455361037686,
+      "learning_rate": 0.00019858218904511657,
+      "loss": 2.4712,
+      "step": 1541
+    },
+    {
+      "epoch": 0.24672,
+      "grad_norm": 2.2400456721282502,
+      "learning_rate": 0.00019857928910898192,
+      "loss": 2.2982,
+      "step": 1542
+    },
+    {
+      "epoch": 0.24688,
+      "grad_norm": 3.377880359445539,
+      "learning_rate": 0.00019857638623138496,
+      "loss": 2.5527,
+      "step": 1543
+    },
+    {
+      "epoch": 0.24704,
+      "grad_norm": 3.510588159352361,
+      "learning_rate": 0.00019857348041241223,
+      "loss": 2.372,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2472,
+      "grad_norm": 2.4632865000681172,
+      "learning_rate": 0.00019857057165215047,
+      "loss": 2.256,
+      "step": 1545
+    },
+    {
+      "epoch": 0.24736,
+      "grad_norm": 2.7163767969576673,
+      "learning_rate": 0.00019856765995068652,
+      "loss": 2.3154,
+      "step": 1546
+    },
+    {
+      "epoch": 0.24752,
+      "grad_norm": 2.7520315185434634,
+      "learning_rate": 0.0001985647453081072,
+      "loss": 2.3722,
+      "step": 1547
+    },
+    {
+      "epoch": 0.24768,
+      "grad_norm": 3.5630407250444525,
+      "learning_rate": 0.00019856182772449952,
+      "loss": 2.4263,
+      "step": 1548
+    },
+    {
+      "epoch": 0.24784,
+      "grad_norm": 3.0946497770534265,
+      "learning_rate": 0.00019855890719995052,
+      "loss": 2.3285,
+      "step": 1549
+    },
+    {
+      "epoch": 0.248,
+      "grad_norm": 2.8503198179259375,
+      "learning_rate": 0.00019855598373454733,
+      "loss": 2.3573,
+      "step": 1550
+    },
+    {
+      "epoch": 0.24816,
+      "grad_norm": 3.0647348470608833,
+      "learning_rate": 0.0001985530573283772,
+      "loss": 2.2983,
+      "step": 1551
+    },
+    {
+      "epoch": 0.24832,
+      "grad_norm": 2.5631605315339083,
+      "learning_rate": 0.00019855012798152742,
+      "loss": 2.367,
+      "step": 1552
+    },
+    {
+      "epoch": 0.24848,
+      "grad_norm": 3.2616253842693634,
+      "learning_rate": 0.00019854719569408548,
+      "loss": 2.2718,
+      "step": 1553
+    },
+    {
+      "epoch": 0.24864,
+      "grad_norm": 2.5598951254699696,
+      "learning_rate": 0.0001985442604661388,
+      "loss": 2.4392,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2488,
+      "grad_norm": 3.852587200406158,
+      "learning_rate": 0.00019854132229777494,
+      "loss": 2.2842,
+      "step": 1555
+    },
+    {
+      "epoch": 0.24896,
+      "grad_norm": 3.784378651956612,
+      "learning_rate": 0.00019853838118908169,
+      "loss": 2.4062,
+      "step": 1556
+    },
+    {
+      "epoch": 0.24912,
+      "grad_norm": 3.454233250141354,
+      "learning_rate": 0.00019853543714014665,
+      "loss": 2.3417,
+      "step": 1557
+    },
+    {
+      "epoch": 0.24928,
+      "grad_norm": 2.6545598056527804,
+      "learning_rate": 0.0001985324901510578,
+      "loss": 2.3716,
+      "step": 1558
+    },
+    {
+      "epoch": 0.24944,
+      "grad_norm": 2.530717776830785,
+      "learning_rate": 0.00019852954022190301,
+      "loss": 2.2887,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2496,
+      "grad_norm": 2.374544006279536,
+      "learning_rate": 0.00019852658735277032,
+      "loss": 2.3212,
+      "step": 1560
+    },
+    {
+      "epoch": 0.24976,
+      "grad_norm": 2.929032272977595,
+      "learning_rate": 0.00019852363154374784,
+      "loss": 2.346,
+      "step": 1561
+    },
+    {
+      "epoch": 0.24992,
+      "grad_norm": 3.020425234811124,
+      "learning_rate": 0.0001985206727949237,
+      "loss": 2.3636,
+      "step": 1562
+    },
+    {
+      "epoch": 0.25008,
+      "grad_norm": 3.0782143022334862,
+      "learning_rate": 0.0001985177111063863,
+      "loss": 2.2881,
+      "step": 1563
+    },
+    {
+      "epoch": 0.25024,
+      "grad_norm": 2.535771079117538,
+      "learning_rate": 0.00019851474647822394,
+      "loss": 2.2916,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2504,
+      "grad_norm": 3.3900446206565564,
+      "learning_rate": 0.0001985117789105251,
+      "loss": 2.4773,
+      "step": 1565
+    },
+    {
+      "epoch": 0.25056,
+      "grad_norm": 2.847593827901904,
+      "learning_rate": 0.00019850880840337833,
+      "loss": 2.3641,
+      "step": 1566
+    },
+    {
+      "epoch": 0.25072,
+      "grad_norm": 3.319001803312752,
+      "learning_rate": 0.00019850583495687225,
+      "loss": 2.3311,
+      "step": 1567
+    },
+    {
+      "epoch": 0.25088,
+      "grad_norm": 2.7664797333939135,
+      "learning_rate": 0.00019850285857109557,
+      "loss": 2.2776,
+      "step": 1568
+    },
+    {
+      "epoch": 0.25104,
+      "grad_norm": 3.27710488941066,
+      "learning_rate": 0.00019849987924613715,
+      "loss": 2.3149,
+      "step": 1569
+    },
+    {
+      "epoch": 0.2512,
+      "grad_norm": 2.6812691304019975,
+      "learning_rate": 0.00019849689698208588,
+      "loss": 2.3966,
+      "step": 1570
+    },
+    {
+      "epoch": 0.25136,
+      "grad_norm": 2.603904520800367,
+      "learning_rate": 0.0001984939117790307,
+      "loss": 2.4311,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25152,
+      "grad_norm": 3.102355785768007,
+      "learning_rate": 0.00019849092363706072,
+      "loss": 2.3237,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25168,
+      "grad_norm": 2.3903176812205507,
+      "learning_rate": 0.0001984879325562651,
+      "loss": 2.4357,
+      "step": 1573
+    },
+    {
+      "epoch": 0.25184,
+      "grad_norm": 2.808241618118446,
+      "learning_rate": 0.00019848493853673308,
+      "loss": 2.467,
+      "step": 1574
+    },
+    {
+      "epoch": 0.252,
+      "grad_norm": 2.680484758012559,
+      "learning_rate": 0.000198481941578554,
+      "loss": 2.3038,
+      "step": 1575
+    },
+    {
+      "epoch": 0.25216,
+      "grad_norm": 3.3548102087196394,
+      "learning_rate": 0.0001984789416818173,
+      "loss": 2.3987,
+      "step": 1576
+    },
+    {
+      "epoch": 0.25232,
+      "grad_norm": 2.814961305384181,
+      "learning_rate": 0.00019847593884661247,
+      "loss": 2.3182,
+      "step": 1577
+    },
+    {
+      "epoch": 0.25248,
+      "grad_norm": 2.725892823449959,
+      "learning_rate": 0.0001984729330730291,
+      "loss": 2.3116,
+      "step": 1578
+    },
+    {
+      "epoch": 0.25264,
+      "grad_norm": 3.4300501604936455,
+      "learning_rate": 0.00019846992436115693,
+      "loss": 2.4246,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2528,
+      "grad_norm": 2.6543908484360643,
+      "learning_rate": 0.00019846691271108567,
+      "loss": 2.3574,
+      "step": 1580
+    },
+    {
+      "epoch": 0.25296,
+      "grad_norm": 3.512699644431009,
+      "learning_rate": 0.00019846389812290524,
+      "loss": 2.4397,
+      "step": 1581
+    },
+    {
+      "epoch": 0.25312,
+      "grad_norm": 2.450817813559121,
+      "learning_rate": 0.00019846088059670555,
+      "loss": 2.1574,
+      "step": 1582
+    },
+    {
+      "epoch": 0.25328,
+      "grad_norm": 3.1783923557857676,
+      "learning_rate": 0.0001984578601325767,
+      "loss": 2.2647,
+      "step": 1583
+    },
+    {
+      "epoch": 0.25344,
+      "grad_norm": 3.1807484719699293,
+      "learning_rate": 0.00019845483673060872,
+      "loss": 2.4342,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2536,
+      "grad_norm": 2.506188595215319,
+      "learning_rate": 0.00019845181039089189,
+      "loss": 2.4008,
+      "step": 1585
+    },
+    {
+      "epoch": 0.25376,
+      "grad_norm": 2.5218225794292826,
+      "learning_rate": 0.0001984487811135165,
+      "loss": 2.3679,
+      "step": 1586
+    },
+    {
+      "epoch": 0.25392,
+      "grad_norm": 2.6441336336864465,
+      "learning_rate": 0.00019844574889857296,
+      "loss": 2.3851,
+      "step": 1587
+    },
+    {
+      "epoch": 0.25408,
+      "grad_norm": 3.4886013186494096,
+      "learning_rate": 0.0001984427137461517,
+      "loss": 2.4089,
+      "step": 1588
+    },
+    {
+      "epoch": 0.25424,
+      "grad_norm": 3.131457863140836,
+      "learning_rate": 0.0001984396756563433,
+      "loss": 2.336,
+      "step": 1589
+    },
+    {
+      "epoch": 0.2544,
+      "grad_norm": 3.323836360375625,
+      "learning_rate": 0.00019843663462923842,
+      "loss": 2.4209,
+      "step": 1590
+    },
+    {
+      "epoch": 0.25456,
+      "grad_norm": 2.6670914248773667,
+      "learning_rate": 0.0001984335906649278,
+      "loss": 2.415,
+      "step": 1591
+    },
+    {
+      "epoch": 0.25472,
+      "grad_norm": 2.7031413435528027,
+      "learning_rate": 0.0001984305437635023,
+      "loss": 2.3598,
+      "step": 1592
+    },
+    {
+      "epoch": 0.25488,
+      "grad_norm": 2.5657412872521697,
+      "learning_rate": 0.00019842749392505278,
+      "loss": 2.2983,
+      "step": 1593
+    },
+    {
+      "epoch": 0.25504,
+      "grad_norm": 2.6564216662606026,
+      "learning_rate": 0.00019842444114967028,
+      "loss": 2.362,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2552,
+      "grad_norm": 2.9329449164916426,
+      "learning_rate": 0.00019842138543744585,
+      "loss": 2.4118,
+      "step": 1595
+    },
+    {
+      "epoch": 0.25536,
+      "grad_norm": 2.6149050672829017,
+      "learning_rate": 0.00019841832678847072,
+      "loss": 2.4351,
+      "step": 1596
+    },
+    {
+      "epoch": 0.25552,
+      "grad_norm": 2.7115943536917,
+      "learning_rate": 0.0001984152652028361,
+      "loss": 2.3909,
+      "step": 1597
+    },
+    {
+      "epoch": 0.25568,
+      "grad_norm": 3.111554828619634,
+      "learning_rate": 0.0001984122006806334,
+      "loss": 2.2504,
+      "step": 1598
+    },
+    {
+      "epoch": 0.25584,
+      "grad_norm": 2.786088570689942,
+      "learning_rate": 0.00019840913322195403,
+      "loss": 2.3014,
+      "step": 1599
+    },
+    {
+      "epoch": 0.256,
+      "grad_norm": 3.994336958049368,
+      "learning_rate": 0.0001984060628268895,
+      "loss": 2.3393,
+      "step": 1600
+    },
+    {
+      "epoch": 0.25616,
+      "grad_norm": 3.5333393638136985,
+      "learning_rate": 0.00019840298949553147,
+      "loss": 2.3186,
+      "step": 1601
+    },
+    {
+      "epoch": 0.25632,
+      "grad_norm": 3.0356962460887664,
+      "learning_rate": 0.0001983999132279716,
+      "loss": 2.3772,
+      "step": 1602
+    },
+    {
+      "epoch": 0.25648,
+      "grad_norm": 3.3436237166803195,
+      "learning_rate": 0.00019839683402430173,
+      "loss": 2.2931,
+      "step": 1603
+    },
+    {
+      "epoch": 0.25664,
+      "grad_norm": 2.450812372192856,
+      "learning_rate": 0.00019839375188461368,
+      "loss": 2.3053,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2568,
+      "grad_norm": 2.5292799233592067,
+      "learning_rate": 0.00019839066680899946,
+      "loss": 2.3959,
+      "step": 1605
+    },
+    {
+      "epoch": 0.25696,
+      "grad_norm": 3.020708165355865,
+      "learning_rate": 0.00019838757879755116,
+      "loss": 2.3908,
+      "step": 1606
+    },
+    {
+      "epoch": 0.25712,
+      "grad_norm": 2.592880800652305,
+      "learning_rate": 0.00019838448785036082,
+      "loss": 2.4226,
+      "step": 1607
+    },
+    {
+      "epoch": 0.25728,
+      "grad_norm": 3.2691504229676127,
+      "learning_rate": 0.0001983813939675207,
+      "loss": 2.4066,
+      "step": 1608
+    },
+    {
+      "epoch": 0.25744,
+      "grad_norm": 2.8276000863276183,
+      "learning_rate": 0.0001983782971491232,
+      "loss": 2.3928,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2576,
+      "grad_norm": 2.956214007032562,
+      "learning_rate": 0.00019837519739526066,
+      "loss": 2.3316,
+      "step": 1610
+    },
+    {
+      "epoch": 0.25776,
+      "grad_norm": 2.444908094629176,
+      "learning_rate": 0.00019837209470602556,
+      "loss": 2.3252,
+      "step": 1611
+    },
+    {
+      "epoch": 0.25792,
+      "grad_norm": 3.3335035982169017,
+      "learning_rate": 0.0001983689890815105,
+      "loss": 2.296,
+      "step": 1612
+    },
+    {
+      "epoch": 0.25808,
+      "grad_norm": 2.2927315851452126,
+      "learning_rate": 0.00019836588052180812,
+      "loss": 2.2854,
+      "step": 1613
+    },
+    {
+      "epoch": 0.25824,
+      "grad_norm": 3.467545137598852,
+      "learning_rate": 0.00019836276902701126,
+      "loss": 2.3207,
+      "step": 1614
+    },
+    {
+      "epoch": 0.2584,
+      "grad_norm": 2.662041670827964,
+      "learning_rate": 0.00019835965459721265,
+      "loss": 2.3199,
+      "step": 1615
+    },
+    {
+      "epoch": 0.25856,
+      "grad_norm": 2.5457385858106933,
+      "learning_rate": 0.00019835653723250526,
+      "loss": 2.2768,
+      "step": 1616
+    },
+    {
+      "epoch": 0.25872,
+      "grad_norm": 2.7066409340766033,
+      "learning_rate": 0.00019835341693298216,
+      "loss": 2.2831,
+      "step": 1617
+    },
+    {
+      "epoch": 0.25888,
+      "grad_norm": 2.5922272612943025,
+      "learning_rate": 0.0001983502936987364,
+      "loss": 2.3088,
+      "step": 1618
+    },
+    {
+      "epoch": 0.25904,
+      "grad_norm": 2.9607721342280273,
+      "learning_rate": 0.0001983471675298612,
+      "loss": 2.2776,
+      "step": 1619
+    },
+    {
+      "epoch": 0.2592,
+      "grad_norm": 3.450812238333222,
+      "learning_rate": 0.0001983440384264498,
+      "loss": 2.3089,
+      "step": 1620
+    },
+    {
+      "epoch": 0.25936,
+      "grad_norm": 2.7191388373926664,
+      "learning_rate": 0.0001983409063885956,
+      "loss": 2.4102,
+      "step": 1621
+    },
+    {
+      "epoch": 0.25952,
+      "grad_norm": 3.4117382389652575,
+      "learning_rate": 0.0001983377714163921,
+      "loss": 2.408,
+      "step": 1622
+    },
+    {
+      "epoch": 0.25968,
+      "grad_norm": 2.5842249517096376,
+      "learning_rate": 0.00019833463350993276,
+      "loss": 2.2464,
+      "step": 1623
+    },
+    {
+      "epoch": 0.25984,
+      "grad_norm": 2.7842836566940248,
+      "learning_rate": 0.0001983314926693112,
+      "loss": 2.3049,
+      "step": 1624
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 2.866713536458556,
+      "learning_rate": 0.00019832834889462126,
+      "loss": 2.2884,
+      "step": 1625
+    },
+    {
+      "epoch": 0.26016,
+      "grad_norm": 2.9673972865438825,
+      "learning_rate": 0.00019832520218595663,
+      "loss": 2.2488,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26032,
+      "grad_norm": 3.0290425235589713,
+      "learning_rate": 0.00019832205254341126,
+      "loss": 2.3179,
+      "step": 1627
+    },
+    {
+      "epoch": 0.26048,
+      "grad_norm": 3.043999234434197,
+      "learning_rate": 0.00019831889996707908,
+      "loss": 2.3913,
+      "step": 1628
+    },
+    {
+      "epoch": 0.26064,
+      "grad_norm": 3.1654094829371724,
+      "learning_rate": 0.0001983157444570542,
+      "loss": 2.4659,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2608,
+      "grad_norm": 2.731964131600935,
+      "learning_rate": 0.00019831258601343078,
+      "loss": 2.2172,
+      "step": 1630
+    },
+    {
+      "epoch": 0.26096,
+      "grad_norm": 2.7841542529989374,
+      "learning_rate": 0.00019830942463630304,
+      "loss": 2.3131,
+      "step": 1631
+    },
+    {
+      "epoch": 0.26112,
+      "grad_norm": 2.7639667187742227,
+      "learning_rate": 0.00019830626032576535,
+      "loss": 2.387,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26128,
+      "grad_norm": 2.5228199385349264,
+      "learning_rate": 0.00019830309308191207,
+      "loss": 2.2279,
+      "step": 1633
+    },
+    {
+      "epoch": 0.26144,
+      "grad_norm": 2.6085114548305737,
+      "learning_rate": 0.0001982999229048377,
+      "loss": 2.4256,
+      "step": 1634
+    },
+    {
+      "epoch": 0.2616,
+      "grad_norm": 2.761488750947841,
+      "learning_rate": 0.00019829674979463696,
+      "loss": 2.283,
+      "step": 1635
+    },
+    {
+      "epoch": 0.26176,
+      "grad_norm": 2.954523460392327,
+      "learning_rate": 0.00019829357375140437,
+      "loss": 2.2864,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26192,
+      "grad_norm": 3.292975074149261,
+      "learning_rate": 0.0001982903947752348,
+      "loss": 2.3163,
+      "step": 1637
+    },
+    {
+      "epoch": 0.26208,
+      "grad_norm": 2.715635423014736,
+      "learning_rate": 0.00019828721286622306,
+      "loss": 2.3438,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26224,
+      "grad_norm": 3.2270386690565998,
+      "learning_rate": 0.00019828402802446408,
+      "loss": 2.2414,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2624,
+      "grad_norm": 3.3649157541674857,
+      "learning_rate": 0.00019828084025005297,
+      "loss": 2.3352,
+      "step": 1640
+    },
+    {
+      "epoch": 0.26256,
+      "grad_norm": 2.893129230704414,
+      "learning_rate": 0.00019827764954308478,
+      "loss": 2.2379,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26272,
+      "grad_norm": 2.593799439303848,
+      "learning_rate": 0.00019827445590365472,
+      "loss": 2.2177,
+      "step": 1642
+    },
+    {
+      "epoch": 0.26288,
+      "grad_norm": 3.0915448127978875,
+      "learning_rate": 0.0001982712593318581,
+      "loss": 2.4649,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26304,
+      "grad_norm": 2.779650274723147,
+      "learning_rate": 0.0001982680598277903,
+      "loss": 2.2397,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2632,
+      "grad_norm": 3.3232338857869346,
+      "learning_rate": 0.00019826485739154678,
+      "loss": 2.3085,
+      "step": 1645
+    },
+    {
+      "epoch": 0.26336,
+      "grad_norm": 3.406190487625826,
+      "learning_rate": 0.00019826165202322312,
+      "loss": 2.2403,
+      "step": 1646
+    },
+    {
+      "epoch": 0.26352,
+      "grad_norm": 2.9257970687434756,
+      "learning_rate": 0.00019825844372291495,
+      "loss": 2.4308,
+      "step": 1647
+    },
+    {
+      "epoch": 0.26368,
+      "grad_norm": 2.569438596373319,
+      "learning_rate": 0.00019825523249071797,
+      "loss": 2.326,
+      "step": 1648
+    },
+    {
+      "epoch": 0.26384,
+      "grad_norm": 2.5453208112870036,
+      "learning_rate": 0.00019825201832672807,
+      "loss": 2.2895,
+      "step": 1649
+    },
+    {
+      "epoch": 0.264,
+      "grad_norm": 3.0031939053211287,
+      "learning_rate": 0.00019824880123104106,
+      "loss": 2.3778,
+      "step": 1650
+    },
+    {
+      "epoch": 0.26416,
+      "grad_norm": 5.236479765424396,
+      "learning_rate": 0.000198245581203753,
+      "loss": 2.3492,
+      "step": 1651
+    },
+    {
+      "epoch": 0.26432,
+      "grad_norm": 2.7022127793556168,
+      "learning_rate": 0.00019824235824495996,
+      "loss": 2.4104,
+      "step": 1652
+    },
+    {
+      "epoch": 0.26448,
+      "grad_norm": 3.5269005518868815,
+      "learning_rate": 0.0001982391323547581,
+      "loss": 2.2584,
+      "step": 1653
+    },
+    {
+      "epoch": 0.26464,
+      "grad_norm": 2.994709546108483,
+      "learning_rate": 0.0001982359035332437,
+      "loss": 2.309,
+      "step": 1654
+    },
+    {
+      "epoch": 0.2648,
+      "grad_norm": 3.4378287308889584,
+      "learning_rate": 0.0001982326717805131,
+      "loss": 2.2792,
+      "step": 1655
+    },
+    {
+      "epoch": 0.26496,
+      "grad_norm": 3.054688744094265,
+      "learning_rate": 0.00019822943709666267,
+      "loss": 2.3183,
+      "step": 1656
+    },
+    {
+      "epoch": 0.26512,
+      "grad_norm": 2.5510960307970523,
+      "learning_rate": 0.00019822619948178897,
+      "loss": 2.3046,
+      "step": 1657
+    },
+    {
+      "epoch": 0.26528,
+      "grad_norm": 2.7368842265946167,
+      "learning_rate": 0.00019822295893598863,
+      "loss": 2.3488,
+      "step": 1658
+    },
+    {
+      "epoch": 0.26544,
+      "grad_norm": 2.466774161620559,
+      "learning_rate": 0.0001982197154593583,
+      "loss": 2.1188,
+      "step": 1659
+    },
+    {
+      "epoch": 0.2656,
+      "grad_norm": 2.3983831531262974,
+      "learning_rate": 0.0001982164690519948,
+      "loss": 2.2863,
+      "step": 1660
+    },
+    {
+      "epoch": 0.26576,
+      "grad_norm": 2.5247668435413164,
+      "learning_rate": 0.00019821321971399497,
+      "loss": 2.285,
+      "step": 1661
+    },
+    {
+      "epoch": 0.26592,
+      "grad_norm": 3.2486216257693457,
+      "learning_rate": 0.0001982099674454558,
+      "loss": 2.4982,
+      "step": 1662
+    },
+    {
+      "epoch": 0.26608,
+      "grad_norm": 2.4397189723674133,
+      "learning_rate": 0.00019820671224647432,
+      "loss": 2.2069,
+      "step": 1663
+    },
+    {
+      "epoch": 0.26624,
+      "grad_norm": 2.449630596325863,
+      "learning_rate": 0.0001982034541171476,
+      "loss": 2.2134,
+      "step": 1664
+    },
+    {
+      "epoch": 0.2664,
+      "grad_norm": 2.4757032172296514,
+      "learning_rate": 0.00019820019305757293,
+      "loss": 2.2179,
+      "step": 1665
+    },
+    {
+      "epoch": 0.26656,
+      "grad_norm": 2.399122630924837,
+      "learning_rate": 0.0001981969290678476,
+      "loss": 2.3602,
+      "step": 1666
+    },
+    {
+      "epoch": 0.26672,
+      "grad_norm": 3.1156907370285944,
+      "learning_rate": 0.00019819366214806898,
+      "loss": 2.3332,
+      "step": 1667
+    },
+    {
+      "epoch": 0.26688,
+      "grad_norm": 2.6046485441384184,
+      "learning_rate": 0.00019819039229833455,
+      "loss": 2.3686,
+      "step": 1668
+    },
+    {
+      "epoch": 0.26704,
+      "grad_norm": 2.653072927344337,
+      "learning_rate": 0.00019818711951874191,
+      "loss": 2.4103,
+      "step": 1669
+    },
+    {
+      "epoch": 0.2672,
+      "grad_norm": 3.6389345279767853,
+      "learning_rate": 0.00019818384380938873,
+      "loss": 2.4015,
+      "step": 1670
+    },
+    {
+      "epoch": 0.26736,
+      "grad_norm": 2.670353113679647,
+      "learning_rate": 0.00019818056517037266,
+      "loss": 2.2457,
+      "step": 1671
+    },
+    {
+      "epoch": 0.26752,
+      "grad_norm": 2.809220295009736,
+      "learning_rate": 0.0001981772836017916,
+      "loss": 2.2614,
+      "step": 1672
+    },
+    {
+      "epoch": 0.26768,
+      "grad_norm": 2.855757472941043,
+      "learning_rate": 0.0001981739991037435,
+      "loss": 2.346,
+      "step": 1673
+    },
+    {
+      "epoch": 0.26784,
+      "grad_norm": 3.0173036469581795,
+      "learning_rate": 0.0001981707116763263,
+      "loss": 2.2137,
+      "step": 1674
+    },
+    {
+      "epoch": 0.268,
+      "grad_norm": 3.152305055047542,
+      "learning_rate": 0.0001981674213196381,
+      "loss": 2.3205,
+      "step": 1675
+    },
+    {
+      "epoch": 0.26816,
+      "grad_norm": 2.882000934187908,
+      "learning_rate": 0.0001981641280337771,
+      "loss": 2.268,
+      "step": 1676
+    },
+    {
+      "epoch": 0.26832,
+      "grad_norm": 2.414201275842011,
+      "learning_rate": 0.00019816083181884158,
+      "loss": 2.161,
+      "step": 1677
+    },
+    {
+      "epoch": 0.26848,
+      "grad_norm": 2.97973541406651,
+      "learning_rate": 0.00019815753267492987,
+      "loss": 2.2376,
+      "step": 1678
+    },
+    {
+      "epoch": 0.26864,
+      "grad_norm": 2.8504767864617873,
+      "learning_rate": 0.0001981542306021404,
+      "loss": 2.1756,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2688,
+      "grad_norm": 2.719703848594572,
+      "learning_rate": 0.0001981509256005717,
+      "loss": 2.3918,
+      "step": 1680
+    },
+    {
+      "epoch": 0.26896,
+      "grad_norm": 3.1423025656839467,
+      "learning_rate": 0.0001981476176703224,
+      "loss": 2.208,
+      "step": 1681
+    },
+    {
+      "epoch": 0.26912,
+      "grad_norm": 3.63849228785236,
+      "learning_rate": 0.00019814430681149126,
+      "loss": 2.3049,
+      "step": 1682
+    },
+    {
+      "epoch": 0.26928,
+      "grad_norm": 3.063027026960093,
+      "learning_rate": 0.00019814099302417696,
+      "loss": 2.3192,
+      "step": 1683
+    },
+    {
+      "epoch": 0.26944,
+      "grad_norm": 4.081152264856389,
+      "learning_rate": 0.00019813767630847846,
+      "loss": 2.3726,
+      "step": 1684
+    },
+    {
+      "epoch": 0.2696,
+      "grad_norm": 2.560621374880146,
+      "learning_rate": 0.0001981343566644947,
+      "loss": 2.3265,
+      "step": 1685
+    },
+    {
+      "epoch": 0.26976,
+      "grad_norm": 3.320839729050535,
+      "learning_rate": 0.0001981310340923247,
+      "loss": 2.4747,
+      "step": 1686
+    },
+    {
+      "epoch": 0.26992,
+      "grad_norm": 2.678771521426955,
+      "learning_rate": 0.00019812770859206768,
+      "loss": 2.3425,
+      "step": 1687
+    },
+    {
+      "epoch": 0.27008,
+      "grad_norm": 2.489959586844413,
+      "learning_rate": 0.0001981243801638228,
+      "loss": 2.2775,
+      "step": 1688
+    },
+    {
+      "epoch": 0.27024,
+      "grad_norm": 2.7491681604589937,
+      "learning_rate": 0.0001981210488076894,
+      "loss": 2.3457,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2704,
+      "grad_norm": 2.2980367848658916,
+      "learning_rate": 0.00019811771452376688,
+      "loss": 2.2896,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27056,
+      "grad_norm": 2.9248533531034555,
+      "learning_rate": 0.00019811437731215477,
+      "loss": 2.2203,
+      "step": 1691
+    },
+    {
+      "epoch": 0.27072,
+      "grad_norm": 2.538209498143175,
+      "learning_rate": 0.00019811103717295257,
+      "loss": 2.3028,
+      "step": 1692
+    },
+    {
+      "epoch": 0.27088,
+      "grad_norm": 2.942701241976463,
+      "learning_rate": 0.00019810769410625998,
+      "loss": 2.3686,
+      "step": 1693
+    },
+    {
+      "epoch": 0.27104,
+      "grad_norm": 3.601585927183573,
+      "learning_rate": 0.00019810434811217675,
+      "loss": 2.2368,
+      "step": 1694
+    },
+    {
+      "epoch": 0.2712,
+      "grad_norm": 2.5304449881379525,
+      "learning_rate": 0.00019810099919080276,
+      "loss": 2.2559,
+      "step": 1695
+    },
+    {
+      "epoch": 0.27136,
+      "grad_norm": 4.178933783852623,
+      "learning_rate": 0.0001980976473422379,
+      "loss": 2.2896,
+      "step": 1696
+    },
+    {
+      "epoch": 0.27152,
+      "grad_norm": 2.78828112145337,
+      "learning_rate": 0.00019809429256658213,
+      "loss": 2.2305,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27168,
+      "grad_norm": 2.4494821139594296,
+      "learning_rate": 0.00019809093486393566,
+      "loss": 2.3029,
+      "step": 1698
+    },
+    {
+      "epoch": 0.27184,
+      "grad_norm": 2.041288047558559,
+      "learning_rate": 0.00019808757423439861,
+      "loss": 2.3419,
+      "step": 1699
+    },
+    {
+      "epoch": 0.272,
+      "grad_norm": 2.6193939822235075,
+      "learning_rate": 0.00019808421067807127,
+      "loss": 2.2541,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27216,
+      "grad_norm": 2.846747437340764,
+      "learning_rate": 0.00019808084419505403,
+      "loss": 2.2565,
+      "step": 1701
+    },
+    {
+      "epoch": 0.27232,
+      "grad_norm": 2.954907240791223,
+      "learning_rate": 0.0001980774747854473,
+      "loss": 2.2233,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27248,
+      "grad_norm": 2.6690999675256015,
+      "learning_rate": 0.00019807410244935162,
+      "loss": 2.2946,
+      "step": 1703
+    },
+    {
+      "epoch": 0.27264,
+      "grad_norm": 2.3194977774497327,
+      "learning_rate": 0.00019807072718686767,
+      "loss": 2.284,
+      "step": 1704
+    },
+    {
+      "epoch": 0.2728,
+      "grad_norm": 2.930056485990181,
+      "learning_rate": 0.00019806734899809611,
+      "loss": 2.288,
+      "step": 1705
+    },
+    {
+      "epoch": 0.27296,
+      "grad_norm": 2.3737852069780576,
+      "learning_rate": 0.00019806396788313774,
+      "loss": 2.2984,
+      "step": 1706
+    },
+    {
+      "epoch": 0.27312,
+      "grad_norm": 2.4669218049180746,
+      "learning_rate": 0.00019806058384209347,
+      "loss": 2.2759,
+      "step": 1707
+    },
+    {
+      "epoch": 0.27328,
+      "grad_norm": 2.4018937662889357,
+      "learning_rate": 0.00019805719687506428,
+      "loss": 2.2441,
+      "step": 1708
+    },
+    {
+      "epoch": 0.27344,
+      "grad_norm": 3.2930150614757827,
+      "learning_rate": 0.00019805380698215122,
+      "loss": 2.3165,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2736,
+      "grad_norm": 3.1761112149123014,
+      "learning_rate": 0.00019805041416345542,
+      "loss": 2.2424,
+      "step": 1710
+    },
+    {
+      "epoch": 0.27376,
+      "grad_norm": 3.212896125652509,
+      "learning_rate": 0.00019804701841907816,
+      "loss": 2.4816,
+      "step": 1711
+    },
+    {
+      "epoch": 0.27392,
+      "grad_norm": 2.7963459862814317,
+      "learning_rate": 0.00019804361974912072,
+      "loss": 2.2388,
+      "step": 1712
+    },
+    {
+      "epoch": 0.27408,
+      "grad_norm": 3.1448116119197995,
+      "learning_rate": 0.00019804021815368454,
+      "loss": 2.2952,
+      "step": 1713
+    },
+    {
+      "epoch": 0.27424,
+      "grad_norm": 3.074275310169415,
+      "learning_rate": 0.0001980368136328711,
+      "loss": 2.3223,
+      "step": 1714
+    },
+    {
+      "epoch": 0.2744,
+      "grad_norm": 2.150417573208974,
+      "learning_rate": 0.000198033406186782,
+      "loss": 2.1356,
+      "step": 1715
+    },
+    {
+      "epoch": 0.27456,
+      "grad_norm": 2.3327935724423163,
+      "learning_rate": 0.0001980299958155189,
+      "loss": 2.209,
+      "step": 1716
+    },
+    {
+      "epoch": 0.27472,
+      "grad_norm": 3.168427492174659,
+      "learning_rate": 0.0001980265825191836,
+      "loss": 2.2551,
+      "step": 1717
+    },
+    {
+      "epoch": 0.27488,
+      "grad_norm": 2.549758214081435,
+      "learning_rate": 0.0001980231662978779,
+      "loss": 2.2933,
+      "step": 1718
+    },
+    {
+      "epoch": 0.27504,
+      "grad_norm": 2.819486098453655,
+      "learning_rate": 0.00019801974715170375,
+      "loss": 2.2937,
+      "step": 1719
+    },
+    {
+      "epoch": 0.2752,
+      "grad_norm": 2.673740640973969,
+      "learning_rate": 0.00019801632508076317,
+      "loss": 2.3179,
+      "step": 1720
+    },
+    {
+      "epoch": 0.27536,
+      "grad_norm": 3.159512358422118,
+      "learning_rate": 0.0001980129000851583,
+      "loss": 2.2855,
+      "step": 1721
+    },
+    {
+      "epoch": 0.27552,
+      "grad_norm": 2.304778856021597,
+      "learning_rate": 0.00019800947216499132,
+      "loss": 2.2032,
+      "step": 1722
+    },
+    {
+      "epoch": 0.27568,
+      "grad_norm": 2.5755667851388573,
+      "learning_rate": 0.00019800604132036448,
+      "loss": 2.2435,
+      "step": 1723
+    },
+    {
+      "epoch": 0.27584,
+      "grad_norm": 2.8300305643307375,
+      "learning_rate": 0.00019800260755138016,
+      "loss": 2.1724,
+      "step": 1724
+    },
+    {
+      "epoch": 0.276,
+      "grad_norm": 3.2610220823264933,
+      "learning_rate": 0.00019799917085814084,
+      "loss": 2.2784,
+      "step": 1725
+    },
+    {
+      "epoch": 0.27616,
+      "grad_norm": 2.95316018973005,
+      "learning_rate": 0.0001979957312407491,
+      "loss": 2.2675,
+      "step": 1726
+    },
+    {
+      "epoch": 0.27632,
+      "grad_norm": 2.9486530472713746,
+      "learning_rate": 0.00019799228869930749,
+      "loss": 2.2549,
+      "step": 1727
+    },
+    {
+      "epoch": 0.27648,
+      "grad_norm": 2.545217480642293,
+      "learning_rate": 0.00019798884323391877,
+      "loss": 2.3924,
+      "step": 1728
+    },
+    {
+      "epoch": 0.27664,
+      "grad_norm": 2.500670734525634,
+      "learning_rate": 0.00019798539484468578,
+      "loss": 2.1172,
+      "step": 1729
+    },
+    {
+      "epoch": 0.2768,
+      "grad_norm": 2.160750182634486,
+      "learning_rate": 0.0001979819435317114,
+      "loss": 2.2977,
+      "step": 1730
+    },
+    {
+      "epoch": 0.27696,
+      "grad_norm": 3.345855381200494,
+      "learning_rate": 0.00019797848929509857,
+      "loss": 2.2741,
+      "step": 1731
+    },
+    {
+      "epoch": 0.27712,
+      "grad_norm": 2.7995183154571275,
+      "learning_rate": 0.0001979750321349504,
+      "loss": 2.3475,
+      "step": 1732
+    },
+    {
+      "epoch": 0.27728,
+      "grad_norm": 2.9503733977340185,
+      "learning_rate": 0.00019797157205137005,
+      "loss": 2.3248,
+      "step": 1733
+    },
+    {
+      "epoch": 0.27744,
+      "grad_norm": 2.499312960318029,
+      "learning_rate": 0.00019796810904446072,
+      "loss": 2.3851,
+      "step": 1734
+    },
+    {
+      "epoch": 0.2776,
+      "grad_norm": 3.118698367024418,
+      "learning_rate": 0.00019796464311432578,
+      "loss": 2.2778,
+      "step": 1735
+    },
+    {
+      "epoch": 0.27776,
+      "grad_norm": 2.598162798513969,
+      "learning_rate": 0.00019796117426106865,
+      "loss": 2.2875,
+      "step": 1736
+    },
+    {
+      "epoch": 0.27792,
+      "grad_norm": 3.054408320647199,
+      "learning_rate": 0.00019795770248479285,
+      "loss": 2.2718,
+      "step": 1737
+    },
+    {
+      "epoch": 0.27808,
+      "grad_norm": 2.5956523408076713,
+      "learning_rate": 0.0001979542277856019,
+      "loss": 2.2267,
+      "step": 1738
+    },
+    {
+      "epoch": 0.27824,
+      "grad_norm": 3.0213955038804436,
+      "learning_rate": 0.00019795075016359955,
+      "loss": 2.2721,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2784,
+      "grad_norm": 3.061777050092099,
+      "learning_rate": 0.00019794726961888955,
+      "loss": 2.3267,
+      "step": 1740
+    },
+    {
+      "epoch": 0.27856,
+      "grad_norm": 2.831549297422062,
+      "learning_rate": 0.00019794378615157573,
+      "loss": 2.3016,
+      "step": 1741
+    },
+    {
+      "epoch": 0.27872,
+      "grad_norm": 2.616451763366058,
+      "learning_rate": 0.00019794029976176206,
+      "loss": 2.2558,
+      "step": 1742
+    },
+    {
+      "epoch": 0.27888,
+      "grad_norm": 3.084173385058616,
+      "learning_rate": 0.00019793681044955256,
+      "loss": 2.3699,
+      "step": 1743
+    },
+    {
+      "epoch": 0.27904,
+      "grad_norm": 2.7264297378822286,
+      "learning_rate": 0.00019793331821505133,
+      "loss": 2.2199,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2792,
+      "grad_norm": 2.5212816406723726,
+      "learning_rate": 0.0001979298230583626,
+      "loss": 2.2821,
+      "step": 1745
+    },
+    {
+      "epoch": 0.27936,
+      "grad_norm": 3.1204231670164084,
+      "learning_rate": 0.00019792632497959067,
+      "loss": 2.3535,
+      "step": 1746
+    },
+    {
+      "epoch": 0.27952,
+      "grad_norm": 2.907049930392064,
+      "learning_rate": 0.00019792282397883986,
+      "loss": 2.2433,
+      "step": 1747
+    },
+    {
+      "epoch": 0.27968,
+      "grad_norm": 2.8274269036144677,
+      "learning_rate": 0.0001979193200562147,
+      "loss": 2.1702,
+      "step": 1748
+    },
+    {
+      "epoch": 0.27984,
+      "grad_norm": 3.0262087416085097,
+      "learning_rate": 0.00019791581321181972,
+      "loss": 2.2717,
+      "step": 1749
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 2.404874157940834,
+      "learning_rate": 0.0001979123034457595,
+      "loss": 2.2874,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28016,
+      "grad_norm": 2.9160293179273253,
+      "learning_rate": 0.00019790879075813885,
+      "loss": 2.1964,
+      "step": 1751
+    },
+    {
+      "epoch": 0.28032,
+      "grad_norm": 2.797010280957688,
+      "learning_rate": 0.00019790527514906256,
+      "loss": 2.1977,
+      "step": 1752
+    },
+    {
+      "epoch": 0.28048,
+      "grad_norm": 2.8197556928466074,
+      "learning_rate": 0.00019790175661863556,
+      "loss": 2.3659,
+      "step": 1753
+    },
+    {
+      "epoch": 0.28064,
+      "grad_norm": 2.3880121873806672,
+      "learning_rate": 0.00019789823516696273,
+      "loss": 2.2895,
+      "step": 1754
+    },
+    {
+      "epoch": 0.2808,
+      "grad_norm": 3.0869406059878925,
+      "learning_rate": 0.00019789471079414926,
+      "loss": 2.2104,
+      "step": 1755
+    },
+    {
+      "epoch": 0.28096,
+      "grad_norm": 2.724248990157237,
+      "learning_rate": 0.00019789118350030025,
+      "loss": 2.2343,
+      "step": 1756
+    },
+    {
+      "epoch": 0.28112,
+      "grad_norm": 3.558775343119277,
+      "learning_rate": 0.00019788765328552095,
+      "loss": 2.3012,
+      "step": 1757
+    },
+    {
+      "epoch": 0.28128,
+      "grad_norm": 2.7627410401139656,
+      "learning_rate": 0.00019788412014991674,
+      "loss": 2.3097,
+      "step": 1758
+    },
+    {
+      "epoch": 0.28144,
+      "grad_norm": 2.6593864772601985,
+      "learning_rate": 0.00019788058409359301,
+      "loss": 2.2491,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2816,
+      "grad_norm": 2.372567837518144,
+      "learning_rate": 0.00019787704511665526,
+      "loss": 2.2238,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28176,
+      "grad_norm": 2.87665676207782,
+      "learning_rate": 0.00019787350321920914,
+      "loss": 2.274,
+      "step": 1761
+    },
+    {
+      "epoch": 0.28192,
+      "grad_norm": 2.8622908058069036,
+      "learning_rate": 0.00019786995840136027,
+      "loss": 2.234,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28208,
+      "grad_norm": 3.291680070919589,
+      "learning_rate": 0.00019786641066321447,
+      "loss": 2.2239,
+      "step": 1763
+    },
+    {
+      "epoch": 0.28224,
+      "grad_norm": 2.9122431818948695,
+      "learning_rate": 0.00019786286000487757,
+      "loss": 2.1587,
+      "step": 1764
+    },
+    {
+      "epoch": 0.2824,
+      "grad_norm": 2.295060195812565,
+      "learning_rate": 0.00019785930642645554,
+      "loss": 2.238,
+      "step": 1765
+    },
+    {
+      "epoch": 0.28256,
+      "grad_norm": 3.0090733660093223,
+      "learning_rate": 0.0001978557499280544,
+      "loss": 2.3735,
+      "step": 1766
+    },
+    {
+      "epoch": 0.28272,
+      "grad_norm": 2.7330210683713183,
+      "learning_rate": 0.00019785219050978028,
+      "loss": 2.2844,
+      "step": 1767
+    },
+    {
+      "epoch": 0.28288,
+      "grad_norm": 3.823449920531867,
+      "learning_rate": 0.00019784862817173938,
+      "loss": 2.3911,
+      "step": 1768
+    },
+    {
+      "epoch": 0.28304,
+      "grad_norm": 3.3357984564959366,
+      "learning_rate": 0.00019784506291403798,
+      "loss": 2.2677,
+      "step": 1769
+    },
+    {
+      "epoch": 0.2832,
+      "grad_norm": 3.574935919891878,
+      "learning_rate": 0.00019784149473678247,
+      "loss": 2.3761,
+      "step": 1770
+    },
+    {
+      "epoch": 0.28336,
+      "grad_norm": 2.7237758988702407,
+      "learning_rate": 0.00019783792364007935,
+      "loss": 2.2516,
+      "step": 1771
+    },
+    {
+      "epoch": 0.28352,
+      "grad_norm": 3.5217896820124928,
+      "learning_rate": 0.00019783434962403513,
+      "loss": 2.3514,
+      "step": 1772
+    },
+    {
+      "epoch": 0.28368,
+      "grad_norm": 2.582394232327547,
+      "learning_rate": 0.0001978307726887565,
+      "loss": 2.3824,
+      "step": 1773
+    },
+    {
+      "epoch": 0.28384,
+      "grad_norm": 2.7064356476582003,
+      "learning_rate": 0.00019782719283435016,
+      "loss": 2.261,
+      "step": 1774
+    },
+    {
+      "epoch": 0.284,
+      "grad_norm": 2.689163794832635,
+      "learning_rate": 0.00019782361006092294,
+      "loss": 2.2387,
+      "step": 1775
+    },
+    {
+      "epoch": 0.28416,
+      "grad_norm": 3.1547070935874517,
+      "learning_rate": 0.00019782002436858175,
+      "loss": 2.2195,
+      "step": 1776
+    },
+    {
+      "epoch": 0.28432,
+      "grad_norm": 3.2951862396834364,
+      "learning_rate": 0.00019781643575743357,
+      "loss": 2.3235,
+      "step": 1777
+    },
+    {
+      "epoch": 0.28448,
+      "grad_norm": 2.8047154665591316,
+      "learning_rate": 0.00019781284422758545,
+      "loss": 2.2373,
+      "step": 1778
+    },
+    {
+      "epoch": 0.28464,
+      "grad_norm": 2.527235701696705,
+      "learning_rate": 0.00019780924977914462,
+      "loss": 2.3787,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2848,
+      "grad_norm": 3.7887291603562554,
+      "learning_rate": 0.00019780565241221828,
+      "loss": 2.311,
+      "step": 1780
+    },
+    {
+      "epoch": 0.28496,
+      "grad_norm": 2.387274352948454,
+      "learning_rate": 0.00019780205212691378,
+      "loss": 2.2021,
+      "step": 1781
+    },
+    {
+      "epoch": 0.28512,
+      "grad_norm": 2.732435283274794,
+      "learning_rate": 0.00019779844892333856,
+      "loss": 2.2161,
+      "step": 1782
+    },
+    {
+      "epoch": 0.28528,
+      "grad_norm": 2.6659510738619745,
+      "learning_rate": 0.00019779484280160015,
+      "loss": 2.2244,
+      "step": 1783
+    },
+    {
+      "epoch": 0.28544,
+      "grad_norm": 3.0822842923404075,
+      "learning_rate": 0.00019779123376180614,
+      "loss": 2.2538,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2856,
+      "grad_norm": 2.7160474232771965,
+      "learning_rate": 0.00019778762180406418,
+      "loss": 2.1932,
+      "step": 1785
+    },
+    {
+      "epoch": 0.28576,
+      "grad_norm": 3.154568310753041,
+      "learning_rate": 0.00019778400692848212,
+      "loss": 2.3423,
+      "step": 1786
+    },
+    {
+      "epoch": 0.28592,
+      "grad_norm": 2.6493252804542964,
+      "learning_rate": 0.00019778038913516774,
+      "loss": 2.2255,
+      "step": 1787
+    },
+    {
+      "epoch": 0.28608,
+      "grad_norm": 3.0884836878548407,
+      "learning_rate": 0.00019777676842422904,
+      "loss": 2.257,
+      "step": 1788
+    },
+    {
+      "epoch": 0.28624,
+      "grad_norm": 4.052502540756648,
+      "learning_rate": 0.00019777314479577408,
+      "loss": 2.1403,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2864,
+      "grad_norm": 2.6763358757694364,
+      "learning_rate": 0.0001977695182499109,
+      "loss": 2.2626,
+      "step": 1790
+    },
+    {
+      "epoch": 0.28656,
+      "grad_norm": 3.284488749874395,
+      "learning_rate": 0.0001977658887867478,
+      "loss": 2.3028,
+      "step": 1791
+    },
+    {
+      "epoch": 0.28672,
+      "grad_norm": 2.2423761956146184,
+      "learning_rate": 0.00019776225640639298,
+      "loss": 2.3598,
+      "step": 1792
+    },
+    {
+      "epoch": 0.28688,
+      "grad_norm": 3.1201187219974167,
+      "learning_rate": 0.00019775862110895493,
+      "loss": 2.1971,
+      "step": 1793
+    },
+    {
+      "epoch": 0.28704,
+      "grad_norm": 2.4679404131311653,
+      "learning_rate": 0.00019775498289454206,
+      "loss": 2.3328,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2872,
+      "grad_norm": 2.9305635016116716,
+      "learning_rate": 0.00019775134176326295,
+      "loss": 2.2229,
+      "step": 1795
+    },
+    {
+      "epoch": 0.28736,
+      "grad_norm": 2.5174398139055545,
+      "learning_rate": 0.00019774769771522624,
+      "loss": 2.2508,
+      "step": 1796
+    },
+    {
+      "epoch": 0.28752,
+      "grad_norm": 3.7776150746983173,
+      "learning_rate": 0.00019774405075054067,
+      "loss": 2.3008,
+      "step": 1797
+    },
+    {
+      "epoch": 0.28768,
+      "grad_norm": 2.3327547962992012,
+      "learning_rate": 0.00019774040086931508,
+      "loss": 2.2477,
+      "step": 1798
+    },
+    {
+      "epoch": 0.28784,
+      "grad_norm": 3.7646678776863745,
+      "learning_rate": 0.0001977367480716583,
+      "loss": 2.2808,
+      "step": 1799
+    },
+    {
+      "epoch": 0.288,
+      "grad_norm": 3.4099830177747434,
+      "learning_rate": 0.0001977330923576794,
+      "loss": 2.2639,
+      "step": 1800
+    },
+    {
+      "epoch": 0.28816,
+      "grad_norm": 2.5980640463029654,
+      "learning_rate": 0.00019772943372748744,
+      "loss": 2.2807,
+      "step": 1801
+    },
+    {
+      "epoch": 0.28832,
+      "grad_norm": 3.890852477020094,
+      "learning_rate": 0.00019772577218119157,
+      "loss": 2.2505,
+      "step": 1802
+    },
+    {
+      "epoch": 0.28848,
+      "grad_norm": 3.7594181712414967,
+      "learning_rate": 0.00019772210771890105,
+      "loss": 2.2594,
+      "step": 1803
+    },
+    {
+      "epoch": 0.28864,
+      "grad_norm": 3.2119518641993006,
+      "learning_rate": 0.00019771844034072527,
+      "loss": 2.2724,
+      "step": 1804
+    },
+    {
+      "epoch": 0.2888,
+      "grad_norm": 2.8704268132982995,
+      "learning_rate": 0.0001977147700467736,
+      "loss": 2.3257,
+      "step": 1805
+    },
+    {
+      "epoch": 0.28896,
+      "grad_norm": 3.3462909965190044,
+      "learning_rate": 0.00019771109683715555,
+      "loss": 2.3625,
+      "step": 1806
+    },
+    {
+      "epoch": 0.28912,
+      "grad_norm": 2.5565533706666974,
+      "learning_rate": 0.0001977074207119808,
+      "loss": 2.3355,
+      "step": 1807
+    },
+    {
+      "epoch": 0.28928,
+      "grad_norm": 3.4138592999116377,
+      "learning_rate": 0.00019770374167135896,
+      "loss": 2.1507,
+      "step": 1808
+    },
+    {
+      "epoch": 0.28944,
+      "grad_norm": 2.3086388028556644,
+      "learning_rate": 0.00019770005971539984,
+      "loss": 2.264,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2896,
+      "grad_norm": 2.8891622867450653,
+      "learning_rate": 0.00019769637484421334,
+      "loss": 2.1947,
+      "step": 1810
+    },
+    {
+      "epoch": 0.28976,
+      "grad_norm": 2.8213440230898867,
+      "learning_rate": 0.0001976926870579093,
+      "loss": 2.1932,
+      "step": 1811
+    },
+    {
+      "epoch": 0.28992,
+      "grad_norm": 2.4875923597849394,
+      "learning_rate": 0.0001976889963565979,
+      "loss": 2.2941,
+      "step": 1812
+    },
+    {
+      "epoch": 0.29008,
+      "grad_norm": 2.5597170912900693,
+      "learning_rate": 0.00019768530274038914,
+      "loss": 2.2974,
+      "step": 1813
+    },
+    {
+      "epoch": 0.29024,
+      "grad_norm": 2.3161696853787217,
+      "learning_rate": 0.00019768160620939335,
+      "loss": 2.2802,
+      "step": 1814
+    },
+    {
+      "epoch": 0.2904,
+      "grad_norm": 2.370030367532865,
+      "learning_rate": 0.00019767790676372072,
+      "loss": 2.2227,
+      "step": 1815
+    },
+    {
+      "epoch": 0.29056,
+      "grad_norm": 3.2904477109633095,
+      "learning_rate": 0.0001976742044034817,
+      "loss": 2.1492,
+      "step": 1816
+    },
+    {
+      "epoch": 0.29072,
+      "grad_norm": 2.3325332626748523,
+      "learning_rate": 0.00019767049912878675,
+      "loss": 2.2563,
+      "step": 1817
+    },
+    {
+      "epoch": 0.29088,
+      "grad_norm": 3.708275532897084,
+      "learning_rate": 0.0001976667909397464,
+      "loss": 2.3689,
+      "step": 1818
+    },
+    {
+      "epoch": 0.29104,
+      "grad_norm": 2.8378338535308396,
+      "learning_rate": 0.00019766307983647136,
+      "loss": 2.1431,
+      "step": 1819
+    },
+    {
+      "epoch": 0.2912,
+      "grad_norm": 3.6268525422597837,
+      "learning_rate": 0.0001976593658190723,
+      "loss": 2.2954,
+      "step": 1820
+    },
+    {
+      "epoch": 0.29136,
+      "grad_norm": 3.1400672580041618,
+      "learning_rate": 0.0001976556488876601,
+      "loss": 2.1469,
+      "step": 1821
+    },
+    {
+      "epoch": 0.29152,
+      "grad_norm": 2.9079745395293366,
+      "learning_rate": 0.00019765192904234563,
+      "loss": 2.3353,
+      "step": 1822
+    },
+    {
+      "epoch": 0.29168,
+      "grad_norm": 3.0470919966797214,
+      "learning_rate": 0.00019764820628323987,
+      "loss": 2.268,
+      "step": 1823
+    },
+    {
+      "epoch": 0.29184,
+      "grad_norm": 2.887183848609249,
+      "learning_rate": 0.00019764448061045396,
+      "loss": 2.2506,
+      "step": 1824
+    },
+    {
+      "epoch": 0.292,
+      "grad_norm": 3.0838972957067847,
+      "learning_rate": 0.00019764075202409898,
+      "loss": 2.2748,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29216,
+      "grad_norm": 3.195832663783932,
+      "learning_rate": 0.00019763702052428628,
+      "loss": 2.2462,
+      "step": 1826
+    },
+    {
+      "epoch": 0.29232,
+      "grad_norm": 2.8927696125083338,
+      "learning_rate": 0.00019763328611112715,
+      "loss": 2.2673,
+      "step": 1827
+    },
+    {
+      "epoch": 0.29248,
+      "grad_norm": 3.104462324754358,
+      "learning_rate": 0.00019762954878473302,
+      "loss": 2.1306,
+      "step": 1828
+    },
+    {
+      "epoch": 0.29264,
+      "grad_norm": 2.4256209746780657,
+      "learning_rate": 0.0001976258085452154,
+      "loss": 2.3163,
+      "step": 1829
+    },
+    {
+      "epoch": 0.2928,
+      "grad_norm": 2.3852411917827845,
+      "learning_rate": 0.00019762206539268592,
+      "loss": 2.2373,
+      "step": 1830
+    },
+    {
+      "epoch": 0.29296,
+      "grad_norm": 3.1090101333389693,
+      "learning_rate": 0.00019761831932725628,
+      "loss": 2.2139,
+      "step": 1831
+    },
+    {
+      "epoch": 0.29312,
+      "grad_norm": 2.5375946758439754,
+      "learning_rate": 0.00019761457034903824,
+      "loss": 2.2356,
+      "step": 1832
+    },
+    {
+      "epoch": 0.29328,
+      "grad_norm": 2.6699419594333165,
+      "learning_rate": 0.00019761081845814363,
+      "loss": 2.2082,
+      "step": 1833
+    },
+    {
+      "epoch": 0.29344,
+      "grad_norm": 3.1337787568325433,
+      "learning_rate": 0.00019760706365468442,
+      "loss": 2.1818,
+      "step": 1834
+    },
+    {
+      "epoch": 0.2936,
+      "grad_norm": 2.62567432309139,
+      "learning_rate": 0.0001976033059387727,
+      "loss": 2.3495,
+      "step": 1835
+    },
+    {
+      "epoch": 0.29376,
+      "grad_norm": 2.8845270235304694,
+      "learning_rate": 0.00019759954531052052,
+      "loss": 2.2848,
+      "step": 1836
+    },
+    {
+      "epoch": 0.29392,
+      "grad_norm": 2.754670018879713,
+      "learning_rate": 0.00019759578177004015,
+      "loss": 2.3463,
+      "step": 1837
+    },
+    {
+      "epoch": 0.29408,
+      "grad_norm": 2.681985269489923,
+      "learning_rate": 0.00019759201531744383,
+      "loss": 2.249,
+      "step": 1838
+    },
+    {
+      "epoch": 0.29424,
+      "grad_norm": 3.0224917806760248,
+      "learning_rate": 0.00019758824595284404,
+      "loss": 2.2714,
+      "step": 1839
+    },
+    {
+      "epoch": 0.2944,
+      "grad_norm": 3.1390196048690235,
+      "learning_rate": 0.00019758447367635316,
+      "loss": 2.1705,
+      "step": 1840
+    },
+    {
+      "epoch": 0.29456,
+      "grad_norm": 2.847321142862892,
+      "learning_rate": 0.0001975806984880838,
+      "loss": 2.1949,
+      "step": 1841
+    },
+    {
+      "epoch": 0.29472,
+      "grad_norm": 3.2364715161562065,
+      "learning_rate": 0.00019757692038814851,
+      "loss": 2.1724,
+      "step": 1842
+    },
+    {
+      "epoch": 0.29488,
+      "grad_norm": 2.730945100170825,
+      "learning_rate": 0.00019757313937666018,
+      "loss": 2.294,
+      "step": 1843
+    },
+    {
+      "epoch": 0.29504,
+      "grad_norm": 2.8948957581166215,
+      "learning_rate": 0.00019756935545373153,
+      "loss": 2.2349,
+      "step": 1844
+    },
+    {
+      "epoch": 0.2952,
+      "grad_norm": 2.220898874395032,
+      "learning_rate": 0.00019756556861947548,
+      "loss": 2.2973,
+      "step": 1845
+    },
+    {
+      "epoch": 0.29536,
+      "grad_norm": 2.9667724029251046,
+      "learning_rate": 0.00019756177887400505,
+      "loss": 2.1908,
+      "step": 1846
+    },
+    {
+      "epoch": 0.29552,
+      "grad_norm": 2.2453759808781357,
+      "learning_rate": 0.00019755798621743327,
+      "loss": 2.2231,
+      "step": 1847
+    },
+    {
+      "epoch": 0.29568,
+      "grad_norm": 3.1999123946144175,
+      "learning_rate": 0.00019755419064987334,
+      "loss": 2.264,
+      "step": 1848
+    },
+    {
+      "epoch": 0.29584,
+      "grad_norm": 3.181764947907635,
+      "learning_rate": 0.00019755039217143853,
+      "loss": 2.2609,
+      "step": 1849
+    },
+    {
+      "epoch": 0.296,
+      "grad_norm": 2.9689402947484766,
+      "learning_rate": 0.00019754659078224216,
+      "loss": 2.1946,
+      "step": 1850
+    },
+    {
+      "epoch": 0.29616,
+      "grad_norm": 2.815091605513438,
+      "learning_rate": 0.00019754278648239767,
+      "loss": 2.1441,
+      "step": 1851
+    },
+    {
+      "epoch": 0.29632,
+      "grad_norm": 2.4152561601661255,
+      "learning_rate": 0.00019753897927201854,
+      "loss": 2.3263,
+      "step": 1852
+    },
+    {
+      "epoch": 0.29648,
+      "grad_norm": 2.604469445647426,
+      "learning_rate": 0.00019753516915121843,
+      "loss": 2.1703,
+      "step": 1853
+    },
+    {
+      "epoch": 0.29664,
+      "grad_norm": 2.274984881111874,
+      "learning_rate": 0.00019753135612011097,
+      "loss": 2.2412,
+      "step": 1854
+    },
+    {
+      "epoch": 0.2968,
+      "grad_norm": 2.6507521705709354,
+      "learning_rate": 0.00019752754017880996,
+      "loss": 2.2369,
+      "step": 1855
+    },
+    {
+      "epoch": 0.29696,
+      "grad_norm": 2.304162794795012,
+      "learning_rate": 0.00019752372132742928,
+      "loss": 2.1789,
+      "step": 1856
+    },
+    {
+      "epoch": 0.29712,
+      "grad_norm": 2.6904244268386694,
+      "learning_rate": 0.00019751989956608287,
+      "loss": 2.2041,
+      "step": 1857
+    },
+    {
+      "epoch": 0.29728,
+      "grad_norm": 3.122137609352522,
+      "learning_rate": 0.00019751607489488474,
+      "loss": 2.2345,
+      "step": 1858
+    },
+    {
+      "epoch": 0.29744,
+      "grad_norm": 2.995058271227329,
+      "learning_rate": 0.00019751224731394905,
+      "loss": 2.2417,
+      "step": 1859
+    },
+    {
+      "epoch": 0.2976,
+      "grad_norm": 2.3901151535095835,
+      "learning_rate": 0.00019750841682338995,
+      "loss": 2.1833,
+      "step": 1860
+    },
+    {
+      "epoch": 0.29776,
+      "grad_norm": 3.110288620387538,
+      "learning_rate": 0.00019750458342332182,
+      "loss": 2.219,
+      "step": 1861
+    },
+    {
+      "epoch": 0.29792,
+      "grad_norm": 2.9009192149026966,
+      "learning_rate": 0.000197500747113859,
+      "loss": 2.0376,
+      "step": 1862
+    },
+    {
+      "epoch": 0.29808,
+      "grad_norm": 2.8244549079603005,
+      "learning_rate": 0.00019749690789511593,
+      "loss": 2.1449,
+      "step": 1863
+    },
+    {
+      "epoch": 0.29824,
+      "grad_norm": 2.6633819443699847,
+      "learning_rate": 0.00019749306576720722,
+      "loss": 2.2447,
+      "step": 1864
+    },
+    {
+      "epoch": 0.2984,
+      "grad_norm": 2.9182618952944726,
+      "learning_rate": 0.0001974892207302475,
+      "loss": 2.14,
+      "step": 1865
+    },
+    {
+      "epoch": 0.29856,
+      "grad_norm": 2.598237254965205,
+      "learning_rate": 0.00019748537278435144,
+      "loss": 2.3306,
+      "step": 1866
+    },
+    {
+      "epoch": 0.29872,
+      "grad_norm": 3.1066146796433167,
+      "learning_rate": 0.00019748152192963393,
+      "loss": 2.284,
+      "step": 1867
+    },
+    {
+      "epoch": 0.29888,
+      "grad_norm": 2.9520238469276383,
+      "learning_rate": 0.00019747766816620988,
+      "loss": 2.0826,
+      "step": 1868
+    },
+    {
+      "epoch": 0.29904,
+      "grad_norm": 2.1439655664389208,
+      "learning_rate": 0.00019747381149419423,
+      "loss": 2.2035,
+      "step": 1869
+    },
+    {
+      "epoch": 0.2992,
+      "grad_norm": 3.405657435498947,
+      "learning_rate": 0.00019746995191370208,
+      "loss": 2.2375,
+      "step": 1870
+    },
+    {
+      "epoch": 0.29936,
+      "grad_norm": 2.6843395785551616,
+      "learning_rate": 0.0001974660894248486,
+      "loss": 2.2021,
+      "step": 1871
+    },
+    {
+      "epoch": 0.29952,
+      "grad_norm": 3.6010211297340677,
+      "learning_rate": 0.00019746222402774904,
+      "loss": 2.2603,
+      "step": 1872
+    },
+    {
+      "epoch": 0.29968,
+      "grad_norm": 2.3393991627889337,
+      "learning_rate": 0.0001974583557225187,
+      "loss": 2.2504,
+      "step": 1873
+    },
+    {
+      "epoch": 0.29984,
+      "grad_norm": 3.300749043584047,
+      "learning_rate": 0.0001974544845092731,
+      "loss": 2.215,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 2.9450650789683266,
+      "learning_rate": 0.00019745061038812763,
+      "loss": 2.2628,
+      "step": 1875
+    },
+    {
+      "epoch": 0.30016,
+      "grad_norm": 2.890176292599796,
+      "learning_rate": 0.00019744673335919796,
+      "loss": 2.2296,
+      "step": 1876
+    },
+    {
+      "epoch": 0.30032,
+      "grad_norm": 3.286957779167102,
+      "learning_rate": 0.00019744285342259977,
+      "loss": 2.2664,
+      "step": 1877
+    },
+    {
+      "epoch": 0.30048,
+      "grad_norm": 3.261354075981635,
+      "learning_rate": 0.0001974389705784488,
+      "loss": 2.2656,
+      "step": 1878
+    },
+    {
+      "epoch": 0.30064,
+      "grad_norm": 2.7686429678459237,
+      "learning_rate": 0.00019743508482686096,
+      "loss": 2.2237,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3008,
+      "grad_norm": 2.481825566182892,
+      "learning_rate": 0.00019743119616795216,
+      "loss": 2.1607,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30096,
+      "grad_norm": 2.601867056157973,
+      "learning_rate": 0.00019742730460183843,
+      "loss": 2.2527,
+      "step": 1881
+    },
+    {
+      "epoch": 0.30112,
+      "grad_norm": 2.534603607478141,
+      "learning_rate": 0.0001974234101286359,
+      "loss": 2.2591,
+      "step": 1882
+    },
+    {
+      "epoch": 0.30128,
+      "grad_norm": 3.5303063820783027,
+      "learning_rate": 0.00019741951274846077,
+      "loss": 2.2024,
+      "step": 1883
+    },
+    {
+      "epoch": 0.30144,
+      "grad_norm": 3.2084432224433455,
+      "learning_rate": 0.00019741561246142936,
+      "loss": 2.2569,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3016,
+      "grad_norm": 2.356010318027483,
+      "learning_rate": 0.000197411709267658,
+      "loss": 2.2956,
+      "step": 1885
+    },
+    {
+      "epoch": 0.30176,
+      "grad_norm": 3.162924491705986,
+      "learning_rate": 0.00019740780316726317,
+      "loss": 2.2384,
+      "step": 1886
+    },
+    {
+      "epoch": 0.30192,
+      "grad_norm": 2.6419387137839987,
+      "learning_rate": 0.00019740389416036144,
+      "loss": 2.3173,
+      "step": 1887
+    },
+    {
+      "epoch": 0.30208,
+      "grad_norm": 3.169972761441896,
+      "learning_rate": 0.00019739998224706945,
+      "loss": 2.1363,
+      "step": 1888
+    },
+    {
+      "epoch": 0.30224,
+      "grad_norm": 2.693922582924529,
+      "learning_rate": 0.00019739606742750392,
+      "loss": 2.189,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3024,
+      "grad_norm": 3.1009375214424146,
+      "learning_rate": 0.00019739214970178166,
+      "loss": 2.0996,
+      "step": 1890
+    },
+    {
+      "epoch": 0.30256,
+      "grad_norm": 2.6404671102136184,
+      "learning_rate": 0.00019738822907001955,
+      "loss": 2.2192,
+      "step": 1891
+    },
+    {
+      "epoch": 0.30272,
+      "grad_norm": 2.5320865948347118,
+      "learning_rate": 0.00019738430553233461,
+      "loss": 2.2172,
+      "step": 1892
+    },
+    {
+      "epoch": 0.30288,
+      "grad_norm": 2.0827270320864004,
+      "learning_rate": 0.00019738037908884388,
+      "loss": 2.2319,
+      "step": 1893
+    },
+    {
+      "epoch": 0.30304,
+      "grad_norm": 3.0974785932993476,
+      "learning_rate": 0.00019737644973966454,
+      "loss": 2.2693,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3032,
+      "grad_norm": 2.6476783129045827,
+      "learning_rate": 0.00019737251748491383,
+      "loss": 2.1914,
+      "step": 1895
+    },
+    {
+      "epoch": 0.30336,
+      "grad_norm": 3.1637976326025448,
+      "learning_rate": 0.0001973685823247091,
+      "loss": 2.2135,
+      "step": 1896
+    },
+    {
+      "epoch": 0.30352,
+      "grad_norm": 2.310722089133447,
+      "learning_rate": 0.00019736464425916777,
+      "loss": 2.2295,
+      "step": 1897
+    },
+    {
+      "epoch": 0.30368,
+      "grad_norm": 3.325118702896636,
+      "learning_rate": 0.00019736070328840727,
+      "loss": 2.1784,
+      "step": 1898
+    },
+    {
+      "epoch": 0.30384,
+      "grad_norm": 2.9710384740550575,
+      "learning_rate": 0.0001973567594125453,
+      "loss": 2.2542,
+      "step": 1899
+    },
+    {
+      "epoch": 0.304,
+      "grad_norm": 2.6986673835436874,
+      "learning_rate": 0.0001973528126316995,
+      "loss": 2.2916,
+      "step": 1900
+    },
+    {
+      "epoch": 0.30416,
+      "grad_norm": 2.5354787371344205,
+      "learning_rate": 0.0001973488629459876,
+      "loss": 2.2414,
+      "step": 1901
+    },
+    {
+      "epoch": 0.30432,
+      "grad_norm": 2.782439355789952,
+      "learning_rate": 0.00019734491035552752,
+      "loss": 2.2027,
+      "step": 1902
+    },
+    {
+      "epoch": 0.30448,
+      "grad_norm": 2.9869249285219546,
+      "learning_rate": 0.00019734095486043714,
+      "loss": 2.2207,
+      "step": 1903
+    },
+    {
+      "epoch": 0.30464,
+      "grad_norm": 2.6628382684438217,
+      "learning_rate": 0.00019733699646083447,
+      "loss": 2.2636,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3048,
+      "grad_norm": 2.602295648225074,
+      "learning_rate": 0.00019733303515683773,
+      "loss": 2.3585,
+      "step": 1905
+    },
+    {
+      "epoch": 0.30496,
+      "grad_norm": 2.37631524883476,
+      "learning_rate": 0.000197329070948565,
+      "loss": 2.1228,
+      "step": 1906
+    },
+    {
+      "epoch": 0.30512,
+      "grad_norm": 2.740968923539444,
+      "learning_rate": 0.00019732510383613466,
+      "loss": 2.1235,
+      "step": 1907
+    },
+    {
+      "epoch": 0.30528,
+      "grad_norm": 2.755249750428689,
+      "learning_rate": 0.000197321133819665,
+      "loss": 2.314,
+      "step": 1908
+    },
+    {
+      "epoch": 0.30544,
+      "grad_norm": 2.4268826546472577,
+      "learning_rate": 0.00019731716089927453,
+      "loss": 2.1752,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3056,
+      "grad_norm": 3.0140490778192714,
+      "learning_rate": 0.0001973131850750818,
+      "loss": 2.1201,
+      "step": 1910
+    },
+    {
+      "epoch": 0.30576,
+      "grad_norm": 2.713012438067155,
+      "learning_rate": 0.00019730920634720539,
+      "loss": 2.1717,
+      "step": 1911
+    },
+    {
+      "epoch": 0.30592,
+      "grad_norm": 3.8135693615747677,
+      "learning_rate": 0.00019730522471576406,
+      "loss": 2.2493,
+      "step": 1912
+    },
+    {
+      "epoch": 0.30608,
+      "grad_norm": 2.4944626164419668,
+      "learning_rate": 0.00019730124018087667,
+      "loss": 2.0644,
+      "step": 1913
+    },
+    {
+      "epoch": 0.30624,
+      "grad_norm": 2.7328036258175756,
+      "learning_rate": 0.000197297252742662,
+      "loss": 2.1083,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3064,
+      "grad_norm": 2.7178487483066474,
+      "learning_rate": 0.0001972932624012391,
+      "loss": 2.1982,
+      "step": 1915
+    },
+    {
+      "epoch": 0.30656,
+      "grad_norm": 2.4884901936842385,
+      "learning_rate": 0.00019728926915672703,
+      "loss": 2.35,
+      "step": 1916
+    },
+    {
+      "epoch": 0.30672,
+      "grad_norm": 2.7230700874297744,
+      "learning_rate": 0.00019728527300924494,
+      "loss": 2.1502,
+      "step": 1917
+    },
+    {
+      "epoch": 0.30688,
+      "grad_norm": 3.0752712317440904,
+      "learning_rate": 0.00019728127395891206,
+      "loss": 2.4546,
+      "step": 1918
+    },
+    {
+      "epoch": 0.30704,
+      "grad_norm": 3.4391071823492885,
+      "learning_rate": 0.00019727727200584768,
+      "loss": 2.3016,
+      "step": 1919
+    },
+    {
+      "epoch": 0.3072,
+      "grad_norm": 2.957704264030024,
+      "learning_rate": 0.0001972732671501713,
+      "loss": 2.132,
+      "step": 1920
+    },
+    {
+      "epoch": 0.30736,
+      "grad_norm": 3.224418420387995,
+      "learning_rate": 0.00019726925939200234,
+      "loss": 2.2555,
+      "step": 1921
+    },
+    {
+      "epoch": 0.30752,
+      "grad_norm": 2.1758445626012346,
+      "learning_rate": 0.00019726524873146044,
+      "loss": 2.2447,
+      "step": 1922
+    },
+    {
+      "epoch": 0.30768,
+      "grad_norm": 2.8796912268345998,
+      "learning_rate": 0.00019726123516866523,
+      "loss": 2.1208,
+      "step": 1923
+    },
+    {
+      "epoch": 0.30784,
+      "grad_norm": 3.0229710797791958,
+      "learning_rate": 0.0001972572187037365,
+      "loss": 2.1897,
+      "step": 1924
+    },
+    {
+      "epoch": 0.308,
+      "grad_norm": 3.038276154164901,
+      "learning_rate": 0.0001972531993367941,
+      "loss": 2.1856,
+      "step": 1925
+    },
+    {
+      "epoch": 0.30816,
+      "grad_norm": 3.0132309896655762,
+      "learning_rate": 0.0001972491770679579,
+      "loss": 2.1404,
+      "step": 1926
+    },
+    {
+      "epoch": 0.30832,
+      "grad_norm": 2.993780138066659,
+      "learning_rate": 0.00019724515189734797,
+      "loss": 2.1804,
+      "step": 1927
+    },
+    {
+      "epoch": 0.30848,
+      "grad_norm": 2.58300088899997,
+      "learning_rate": 0.00019724112382508443,
+      "loss": 2.1516,
+      "step": 1928
+    },
+    {
+      "epoch": 0.30864,
+      "grad_norm": 4.3441181897033925,
+      "learning_rate": 0.00019723709285128743,
+      "loss": 2.2711,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3088,
+      "grad_norm": 2.9838711778444846,
+      "learning_rate": 0.0001972330589760773,
+      "loss": 2.2424,
+      "step": 1930
+    },
+    {
+      "epoch": 0.30896,
+      "grad_norm": 3.1859547813552176,
+      "learning_rate": 0.00019722902219957436,
+      "loss": 2.176,
+      "step": 1931
+    },
+    {
+      "epoch": 0.30912,
+      "grad_norm": 2.750279819919116,
+      "learning_rate": 0.00019722498252189907,
+      "loss": 2.2208,
+      "step": 1932
+    },
+    {
+      "epoch": 0.30928,
+      "grad_norm": 2.896221523892521,
+      "learning_rate": 0.00019722093994317197,
+      "loss": 2.2738,
+      "step": 1933
+    },
+    {
+      "epoch": 0.30944,
+      "grad_norm": 2.5855422213057637,
+      "learning_rate": 0.0001972168944635137,
+      "loss": 2.2132,
+      "step": 1934
+    },
+    {
+      "epoch": 0.3096,
+      "grad_norm": 3.9760740042230474,
+      "learning_rate": 0.00019721284608304496,
+      "loss": 2.2943,
+      "step": 1935
+    },
+    {
+      "epoch": 0.30976,
+      "grad_norm": 2.268772412306113,
+      "learning_rate": 0.00019720879480188654,
+      "loss": 2.2652,
+      "step": 1936
+    },
+    {
+      "epoch": 0.30992,
+      "grad_norm": 3.354344144671542,
+      "learning_rate": 0.00019720474062015934,
+      "loss": 2.2473,
+      "step": 1937
+    },
+    {
+      "epoch": 0.31008,
+      "grad_norm": 3.1488429553430737,
+      "learning_rate": 0.00019720068353798432,
+      "loss": 2.2771,
+      "step": 1938
+    },
+    {
+      "epoch": 0.31024,
+      "grad_norm": 4.109283528005654,
+      "learning_rate": 0.00019719662355548252,
+      "loss": 2.2189,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3104,
+      "grad_norm": 2.736507694309741,
+      "learning_rate": 0.00019719256067277509,
+      "loss": 2.1964,
+      "step": 1940
+    },
+    {
+      "epoch": 0.31056,
+      "grad_norm": 3.8350072821914845,
+      "learning_rate": 0.00019718849488998332,
+      "loss": 2.1897,
+      "step": 1941
+    },
+    {
+      "epoch": 0.31072,
+      "grad_norm": 2.733528425431521,
+      "learning_rate": 0.00019718442620722846,
+      "loss": 2.2849,
+      "step": 1942
+    },
+    {
+      "epoch": 0.31088,
+      "grad_norm": 3.3735631603611367,
+      "learning_rate": 0.00019718035462463194,
+      "loss": 2.1584,
+      "step": 1943
+    },
+    {
+      "epoch": 0.31104,
+      "grad_norm": 2.710496227745055,
+      "learning_rate": 0.00019717628014231524,
+      "loss": 2.2276,
+      "step": 1944
+    },
+    {
+      "epoch": 0.3112,
+      "grad_norm": 3.3588802401980242,
+      "learning_rate": 0.00019717220276039993,
+      "loss": 2.2898,
+      "step": 1945
+    },
+    {
+      "epoch": 0.31136,
+      "grad_norm": 3.589090577247786,
+      "learning_rate": 0.0001971681224790077,
+      "loss": 2.3346,
+      "step": 1946
+    },
+    {
+      "epoch": 0.31152,
+      "grad_norm": 3.4143072801356826,
+      "learning_rate": 0.00019716403929826028,
+      "loss": 2.2571,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31168,
+      "grad_norm": 2.388662057356473,
+      "learning_rate": 0.0001971599532182795,
+      "loss": 2.1604,
+      "step": 1948
+    },
+    {
+      "epoch": 0.31184,
+      "grad_norm": 2.748998562431776,
+      "learning_rate": 0.00019715586423918731,
+      "loss": 2.1342,
+      "step": 1949
+    },
+    {
+      "epoch": 0.312,
+      "grad_norm": 2.910067705003068,
+      "learning_rate": 0.0001971517723611057,
+      "loss": 2.2353,
+      "step": 1950
+    },
+    {
+      "epoch": 0.31216,
+      "grad_norm": 3.3607713969342403,
+      "learning_rate": 0.00019714767758415675,
+      "loss": 2.1666,
+      "step": 1951
+    },
+    {
+      "epoch": 0.31232,
+      "grad_norm": 3.2182319709299776,
+      "learning_rate": 0.0001971435799084627,
+      "loss": 2.2242,
+      "step": 1952
+    },
+    {
+      "epoch": 0.31248,
+      "grad_norm": 2.360177583150763,
+      "learning_rate": 0.00019713947933414574,
+      "loss": 2.2059,
+      "step": 1953
+    },
+    {
+      "epoch": 0.31264,
+      "grad_norm": 2.2922734794263224,
+      "learning_rate": 0.00019713537586132826,
+      "loss": 2.1553,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3128,
+      "grad_norm": 2.703045222936906,
+      "learning_rate": 0.00019713126949013273,
+      "loss": 2.1899,
+      "step": 1955
+    },
+    {
+      "epoch": 0.31296,
+      "grad_norm": 2.465676128024452,
+      "learning_rate": 0.00019712716022068165,
+      "loss": 2.0987,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31312,
+      "grad_norm": 2.8655524936373897,
+      "learning_rate": 0.00019712304805309764,
+      "loss": 2.1843,
+      "step": 1957
+    },
+    {
+      "epoch": 0.31328,
+      "grad_norm": 2.774698338614816,
+      "learning_rate": 0.0001971189329875034,
+      "loss": 2.2006,
+      "step": 1958
+    },
+    {
+      "epoch": 0.31344,
+      "grad_norm": 3.071475172146556,
+      "learning_rate": 0.00019711481502402168,
+      "loss": 2.2534,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3136,
+      "grad_norm": 2.5213615864420498,
+      "learning_rate": 0.00019711069416277543,
+      "loss": 2.141,
+      "step": 1960
+    },
+    {
+      "epoch": 0.31376,
+      "grad_norm": 3.3390749628965604,
+      "learning_rate": 0.00019710657040388757,
+      "loss": 2.1245,
+      "step": 1961
+    },
+    {
+      "epoch": 0.31392,
+      "grad_norm": 2.989415061909898,
+      "learning_rate": 0.00019710244374748114,
+      "loss": 2.2097,
+      "step": 1962
+    },
+    {
+      "epoch": 0.31408,
+      "grad_norm": 2.8829426348598823,
+      "learning_rate": 0.00019709831419367928,
+      "loss": 2.1539,
+      "step": 1963
+    },
+    {
+      "epoch": 0.31424,
+      "grad_norm": 3.1468899686225265,
+      "learning_rate": 0.0001970941817426052,
+      "loss": 2.2848,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3144,
+      "grad_norm": 2.898914544492722,
+      "learning_rate": 0.00019709004639438223,
+      "loss": 2.2539,
+      "step": 1965
+    },
+    {
+      "epoch": 0.31456,
+      "grad_norm": 2.8753703227567415,
+      "learning_rate": 0.00019708590814913378,
+      "loss": 2.2326,
+      "step": 1966
+    },
+    {
+      "epoch": 0.31472,
+      "grad_norm": 2.5692357288559102,
+      "learning_rate": 0.00019708176700698326,
+      "loss": 2.1496,
+      "step": 1967
+    },
+    {
+      "epoch": 0.31488,
+      "grad_norm": 2.2047166671821032,
+      "learning_rate": 0.0001970776229680543,
+      "loss": 2.126,
+      "step": 1968
+    },
+    {
+      "epoch": 0.31504,
+      "grad_norm": 2.297429274615516,
+      "learning_rate": 0.0001970734760324705,
+      "loss": 2.2293,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3152,
+      "grad_norm": 2.927269906936613,
+      "learning_rate": 0.00019706932620035565,
+      "loss": 2.1781,
+      "step": 1970
+    },
+    {
+      "epoch": 0.31536,
+      "grad_norm": 2.1705848091358226,
+      "learning_rate": 0.00019706517347183353,
+      "loss": 2.1924,
+      "step": 1971
+    },
+    {
+      "epoch": 0.31552,
+      "grad_norm": 2.8229030576077085,
+      "learning_rate": 0.0001970610178470281,
+      "loss": 2.1561,
+      "step": 1972
+    },
+    {
+      "epoch": 0.31568,
+      "grad_norm": 3.4985094201924247,
+      "learning_rate": 0.00019705685932606333,
+      "loss": 2.3233,
+      "step": 1973
+    },
+    {
+      "epoch": 0.31584,
+      "grad_norm": 3.2649445229747283,
+      "learning_rate": 0.00019705269790906327,
+      "loss": 2.105,
+      "step": 1974
+    },
+    {
+      "epoch": 0.316,
+      "grad_norm": 2.669387925418763,
+      "learning_rate": 0.00019704853359615214,
+      "loss": 2.1148,
+      "step": 1975
+    },
+    {
+      "epoch": 0.31616,
+      "grad_norm": 2.565056311060403,
+      "learning_rate": 0.0001970443663874542,
+      "loss": 2.1545,
+      "step": 1976
+    },
+    {
+      "epoch": 0.31632,
+      "grad_norm": 2.4550379185639084,
+      "learning_rate": 0.00019704019628309375,
+      "loss": 2.0627,
+      "step": 1977
+    },
+    {
+      "epoch": 0.31648,
+      "grad_norm": 3.2426414098039533,
+      "learning_rate": 0.00019703602328319527,
+      "loss": 2.1707,
+      "step": 1978
+    },
+    {
+      "epoch": 0.31664,
+      "grad_norm": 2.2809028059942693,
+      "learning_rate": 0.00019703184738788324,
+      "loss": 2.1033,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3168,
+      "grad_norm": 2.7130427942226225,
+      "learning_rate": 0.00019702766859728228,
+      "loss": 2.1745,
+      "step": 1980
+    },
+    {
+      "epoch": 0.31696,
+      "grad_norm": 2.466350532134857,
+      "learning_rate": 0.00019702348691151707,
+      "loss": 2.272,
+      "step": 1981
+    },
+    {
+      "epoch": 0.31712,
+      "grad_norm": 2.677254308590208,
+      "learning_rate": 0.00019701930233071238,
+      "loss": 2.0721,
+      "step": 1982
+    },
+    {
+      "epoch": 0.31728,
+      "grad_norm": 3.710588809480669,
+      "learning_rate": 0.00019701511485499307,
+      "loss": 2.1465,
+      "step": 1983
+    },
+    {
+      "epoch": 0.31744,
+      "grad_norm": 2.8092979309622295,
+      "learning_rate": 0.00019701092448448412,
+      "loss": 2.1617,
+      "step": 1984
+    },
+    {
+      "epoch": 0.3176,
+      "grad_norm": 3.150186169981232,
+      "learning_rate": 0.00019700673121931056,
+      "loss": 2.1331,
+      "step": 1985
+    },
+    {
+      "epoch": 0.31776,
+      "grad_norm": 2.674358549263957,
+      "learning_rate": 0.00019700253505959748,
+      "loss": 2.124,
+      "step": 1986
+    },
+    {
+      "epoch": 0.31792,
+      "grad_norm": 2.9537621741368363,
+      "learning_rate": 0.0001969983360054701,
+      "loss": 2.1507,
+      "step": 1987
+    },
+    {
+      "epoch": 0.31808,
+      "grad_norm": 2.37104508205031,
+      "learning_rate": 0.00019699413405705366,
+      "loss": 2.1947,
+      "step": 1988
+    },
+    {
+      "epoch": 0.31824,
+      "grad_norm": 2.5161290257566686,
+      "learning_rate": 0.00019698992921447368,
+      "loss": 2.1232,
+      "step": 1989
+    },
+    {
+      "epoch": 0.3184,
+      "grad_norm": 2.3180056279129744,
+      "learning_rate": 0.0001969857214778555,
+      "loss": 2.2562,
+      "step": 1990
+    },
+    {
+      "epoch": 0.31856,
+      "grad_norm": 2.9030720062737654,
+      "learning_rate": 0.0001969815108473247,
+      "loss": 2.1398,
+      "step": 1991
+    },
+    {
+      "epoch": 0.31872,
+      "grad_norm": 2.9949061110949122,
+      "learning_rate": 0.000196977297323007,
+      "loss": 2.2364,
+      "step": 1992
+    },
+    {
+      "epoch": 0.31888,
+      "grad_norm": 2.3696132726849255,
+      "learning_rate": 0.00019697308090502798,
+      "loss": 2.1826,
+      "step": 1993
+    },
+    {
+      "epoch": 0.31904,
+      "grad_norm": 2.9582020625983723,
+      "learning_rate": 0.00019696886159351355,
+      "loss": 2.2336,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3192,
+      "grad_norm": 2.4313774079170942,
+      "learning_rate": 0.00019696463938858962,
+      "loss": 2.1917,
+      "step": 1995
+    },
+    {
+      "epoch": 0.31936,
+      "grad_norm": 2.5931690628135424,
+      "learning_rate": 0.00019696041429038212,
+      "loss": 2.2286,
+      "step": 1996
+    },
+    {
+      "epoch": 0.31952,
+      "grad_norm": 2.775797104489944,
+      "learning_rate": 0.00019695618629901715,
+      "loss": 2.1135,
+      "step": 1997
+    },
+    {
+      "epoch": 0.31968,
+      "grad_norm": 2.4175321640123735,
+      "learning_rate": 0.00019695195541462085,
+      "loss": 2.1241,
+      "step": 1998
+    },
+    {
+      "epoch": 0.31984,
+      "grad_norm": 3.3752147131770056,
+      "learning_rate": 0.0001969477216373195,
+      "loss": 2.1761,
+      "step": 1999
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 2.5433311398520715,
+      "learning_rate": 0.0001969434849672394,
+      "loss": 2.1614,
+      "step": 2000
+    },
+    {
+      "epoch": 0.32016,
+      "grad_norm": 3.0187324888774,
+      "learning_rate": 0.00019693924540450694,
+      "loss": 2.2196,
+      "step": 2001
+    },
+    {
+      "epoch": 0.32032,
+      "grad_norm": 3.2608081548847174,
+      "learning_rate": 0.00019693500294924867,
+      "loss": 2.0942,
+      "step": 2002
+    },
+    {
+      "epoch": 0.32048,
+      "grad_norm": 2.858930059600435,
+      "learning_rate": 0.00019693075760159116,
+      "loss": 2.1441,
+      "step": 2003
+    },
+    {
+      "epoch": 0.32064,
+      "grad_norm": 2.941633458009233,
+      "learning_rate": 0.0001969265093616611,
+      "loss": 2.1368,
+      "step": 2004
+    },
+    {
+      "epoch": 0.3208,
+      "grad_norm": 2.549504027164352,
+      "learning_rate": 0.00019692225822958524,
+      "loss": 2.0097,
+      "step": 2005
+    },
+    {
+      "epoch": 0.32096,
+      "grad_norm": 2.5930188424772056,
+      "learning_rate": 0.00019691800420549037,
+      "loss": 2.1027,
+      "step": 2006
+    },
+    {
+      "epoch": 0.32112,
+      "grad_norm": 3.1962401641544465,
+      "learning_rate": 0.00019691374728950353,
+      "loss": 2.2894,
+      "step": 2007
+    },
+    {
+      "epoch": 0.32128,
+      "grad_norm": 2.4296669680296707,
+      "learning_rate": 0.00019690948748175168,
+      "loss": 2.2519,
+      "step": 2008
+    },
+    {
+      "epoch": 0.32144,
+      "grad_norm": 2.4661054878759763,
+      "learning_rate": 0.00019690522478236195,
+      "loss": 2.1227,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3216,
+      "grad_norm": 2.6113658878804693,
+      "learning_rate": 0.0001969009591914615,
+      "loss": 2.2721,
+      "step": 2010
+    },
+    {
+      "epoch": 0.32176,
+      "grad_norm": 2.6264699623091188,
+      "learning_rate": 0.00019689669070917763,
+      "loss": 2.3011,
+      "step": 2011
+    },
+    {
+      "epoch": 0.32192,
+      "grad_norm": 3.8440007499245508,
+      "learning_rate": 0.00019689241933563772,
+      "loss": 2.3459,
+      "step": 2012
+    },
+    {
+      "epoch": 0.32208,
+      "grad_norm": 2.282972227199808,
+      "learning_rate": 0.00019688814507096914,
+      "loss": 2.1024,
+      "step": 2013
+    },
+    {
+      "epoch": 0.32224,
+      "grad_norm": 3.3139630416640675,
+      "learning_rate": 0.00019688386791529957,
+      "loss": 2.2407,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3224,
+      "grad_norm": 2.6324951489064774,
+      "learning_rate": 0.0001968795878687565,
+      "loss": 2.2687,
+      "step": 2015
+    },
+    {
+      "epoch": 0.32256,
+      "grad_norm": 3.246058701652056,
+      "learning_rate": 0.00019687530493146772,
+      "loss": 2.2141,
+      "step": 2016
+    },
+    {
+      "epoch": 0.32272,
+      "grad_norm": 2.8462975983745937,
+      "learning_rate": 0.00019687101910356104,
+      "loss": 2.2003,
+      "step": 2017
+    },
+    {
+      "epoch": 0.32288,
+      "grad_norm": 2.5666584031141357,
+      "learning_rate": 0.00019686673038516426,
+      "loss": 2.1262,
+      "step": 2018
+    },
+    {
+      "epoch": 0.32304,
+      "grad_norm": 2.7044078858906957,
+      "learning_rate": 0.0001968624387764054,
+      "loss": 2.2358,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3232,
+      "grad_norm": 2.4780890427974662,
+      "learning_rate": 0.0001968581442774125,
+      "loss": 2.1423,
+      "step": 2020
+    },
+    {
+      "epoch": 0.32336,
+      "grad_norm": 2.643494183006831,
+      "learning_rate": 0.00019685384688831374,
+      "loss": 2.2197,
+      "step": 2021
+    },
+    {
+      "epoch": 0.32352,
+      "grad_norm": 3.8874875235683164,
+      "learning_rate": 0.0001968495466092373,
+      "loss": 2.2362,
+      "step": 2022
+    },
+    {
+      "epoch": 0.32368,
+      "grad_norm": 3.3863899927545664,
+      "learning_rate": 0.00019684524344031155,
+      "loss": 2.2176,
+      "step": 2023
+    },
+    {
+      "epoch": 0.32384,
+      "grad_norm": 2.6658528055306165,
+      "learning_rate": 0.00019684093738166482,
+      "loss": 2.1147,
+      "step": 2024
+    },
+    {
+      "epoch": 0.324,
+      "grad_norm": 3.83681077633359,
+      "learning_rate": 0.00019683662843342564,
+      "loss": 2.2209,
+      "step": 2025
+    },
+    {
+      "epoch": 0.32416,
+      "grad_norm": 3.564439794437094,
+      "learning_rate": 0.00019683231659572257,
+      "loss": 2.1671,
+      "step": 2026
+    },
+    {
+      "epoch": 0.32432,
+      "grad_norm": 2.5767374536514485,
+      "learning_rate": 0.00019682800186868426,
+      "loss": 2.1109,
+      "step": 2027
+    },
+    {
+      "epoch": 0.32448,
+      "grad_norm": 2.6313083390558094,
+      "learning_rate": 0.00019682368425243948,
+      "loss": 2.1513,
+      "step": 2028
+    },
+    {
+      "epoch": 0.32464,
+      "grad_norm": 2.4919030703737928,
+      "learning_rate": 0.00019681936374711705,
+      "loss": 2.1588,
+      "step": 2029
+    },
+    {
+      "epoch": 0.3248,
+      "grad_norm": 2.9699261894841347,
+      "learning_rate": 0.00019681504035284588,
+      "loss": 2.1652,
+      "step": 2030
+    },
+    {
+      "epoch": 0.32496,
+      "grad_norm": 2.9794755595550306,
+      "learning_rate": 0.000196810714069755,
+      "loss": 2.2165,
+      "step": 2031
+    },
+    {
+      "epoch": 0.32512,
+      "grad_norm": 2.596735841327562,
+      "learning_rate": 0.00019680638489797344,
+      "loss": 2.1744,
+      "step": 2032
+    },
+    {
+      "epoch": 0.32528,
+      "grad_norm": 3.148353393025252,
+      "learning_rate": 0.00019680205283763047,
+      "loss": 2.1724,
+      "step": 2033
+    },
+    {
+      "epoch": 0.32544,
+      "grad_norm": 2.483085443560206,
+      "learning_rate": 0.00019679771788885526,
+      "loss": 2.1926,
+      "step": 2034
+    },
+    {
+      "epoch": 0.3256,
+      "grad_norm": 2.8633696518375036,
+      "learning_rate": 0.0001967933800517772,
+      "loss": 2.1619,
+      "step": 2035
+    },
+    {
+      "epoch": 0.32576,
+      "grad_norm": 2.7387138313810575,
+      "learning_rate": 0.00019678903932652574,
+      "loss": 2.1431,
+      "step": 2036
+    },
+    {
+      "epoch": 0.32592,
+      "grad_norm": 2.555551925083661,
+      "learning_rate": 0.00019678469571323038,
+      "loss": 2.3085,
+      "step": 2037
+    },
+    {
+      "epoch": 0.32608,
+      "grad_norm": 2.595004705465071,
+      "learning_rate": 0.00019678034921202073,
+      "loss": 2.0013,
+      "step": 2038
+    },
+    {
+      "epoch": 0.32624,
+      "grad_norm": 3.0726316798586217,
+      "learning_rate": 0.00019677599982302648,
+      "loss": 2.1178,
+      "step": 2039
+    },
+    {
+      "epoch": 0.3264,
+      "grad_norm": 2.8097882277668296,
+      "learning_rate": 0.0001967716475463774,
+      "loss": 2.2274,
+      "step": 2040
+    },
+    {
+      "epoch": 0.32656,
+      "grad_norm": 2.8405827566258233,
+      "learning_rate": 0.00019676729238220336,
+      "loss": 2.111,
+      "step": 2041
+    },
+    {
+      "epoch": 0.32672,
+      "grad_norm": 3.0798288094306616,
+      "learning_rate": 0.00019676293433063436,
+      "loss": 2.1405,
+      "step": 2042
+    },
+    {
+      "epoch": 0.32688,
+      "grad_norm": 2.2879370404815753,
+      "learning_rate": 0.00019675857339180038,
+      "loss": 2.1597,
+      "step": 2043
+    },
+    {
+      "epoch": 0.32704,
+      "grad_norm": 3.88384988997996,
+      "learning_rate": 0.00019675420956583156,
+      "loss": 2.2344,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3272,
+      "grad_norm": 2.505185820561978,
+      "learning_rate": 0.00019674984285285811,
+      "loss": 2.0732,
+      "step": 2045
+    },
+    {
+      "epoch": 0.32736,
+      "grad_norm": 3.070852502538797,
+      "learning_rate": 0.00019674547325301034,
+      "loss": 2.118,
+      "step": 2046
+    },
+    {
+      "epoch": 0.32752,
+      "grad_norm": 3.091693974811029,
+      "learning_rate": 0.0001967411007664186,
+      "loss": 2.2795,
+      "step": 2047
+    },
+    {
+      "epoch": 0.32768,
+      "grad_norm": 2.5728593297404365,
+      "learning_rate": 0.0001967367253932134,
+      "loss": 2.0707,
+      "step": 2048
+    },
+    {
+      "epoch": 0.32784,
+      "grad_norm": 2.17131459448443,
+      "learning_rate": 0.00019673234713352527,
+      "loss": 2.2524,
+      "step": 2049
+    },
+    {
+      "epoch": 0.328,
+      "grad_norm": 2.832175946197687,
+      "learning_rate": 0.00019672796598748483,
+      "loss": 2.2097,
+      "step": 2050
+    },
+    {
+      "epoch": 0.32816,
+      "grad_norm": 2.5553947594114392,
+      "learning_rate": 0.00019672358195522286,
+      "loss": 2.1019,
+      "step": 2051
+    },
+    {
+      "epoch": 0.32832,
+      "grad_norm": 2.7134670661583717,
+      "learning_rate": 0.00019671919503687013,
+      "loss": 2.223,
+      "step": 2052
+    },
+    {
+      "epoch": 0.32848,
+      "grad_norm": 2.700545033880592,
+      "learning_rate": 0.0001967148052325576,
+      "loss": 2.2776,
+      "step": 2053
+    },
+    {
+      "epoch": 0.32864,
+      "grad_norm": 2.597934809785572,
+      "learning_rate": 0.00019671041254241614,
+      "loss": 2.0552,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3288,
+      "grad_norm": 2.6278035902617196,
+      "learning_rate": 0.00019670601696657693,
+      "loss": 2.1799,
+      "step": 2055
+    },
+    {
+      "epoch": 0.32896,
+      "grad_norm": 2.261990072585754,
+      "learning_rate": 0.00019670161850517107,
+      "loss": 2.1998,
+      "step": 2056
+    },
+    {
+      "epoch": 0.32912,
+      "grad_norm": 2.3617243124687626,
+      "learning_rate": 0.00019669721715832986,
+      "loss": 2.1068,
+      "step": 2057
+    },
+    {
+      "epoch": 0.32928,
+      "grad_norm": 2.408946243695195,
+      "learning_rate": 0.00019669281292618456,
+      "loss": 1.9994,
+      "step": 2058
+    },
+    {
+      "epoch": 0.32944,
+      "grad_norm": 3.015954568296941,
+      "learning_rate": 0.00019668840580886664,
+      "loss": 2.1514,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3296,
+      "grad_norm": 2.545173698226165,
+      "learning_rate": 0.00019668399580650755,
+      "loss": 2.1721,
+      "step": 2060
+    },
+    {
+      "epoch": 0.32976,
+      "grad_norm": 2.2100172920775885,
+      "learning_rate": 0.00019667958291923893,
+      "loss": 2.1319,
+      "step": 2061
+    },
+    {
+      "epoch": 0.32992,
+      "grad_norm": 3.060800456079053,
+      "learning_rate": 0.00019667516714719243,
+      "loss": 2.1939,
+      "step": 2062
+    },
+    {
+      "epoch": 0.33008,
+      "grad_norm": 2.7166642692334206,
+      "learning_rate": 0.00019667074849049983,
+      "loss": 2.0311,
+      "step": 2063
+    },
+    {
+      "epoch": 0.33024,
+      "grad_norm": 2.19044746887785,
+      "learning_rate": 0.00019666632694929293,
+      "loss": 2.128,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3304,
+      "grad_norm": 2.719986992708382,
+      "learning_rate": 0.00019666190252370371,
+      "loss": 2.1657,
+      "step": 2065
+    },
+    {
+      "epoch": 0.33056,
+      "grad_norm": 2.4031770916767283,
+      "learning_rate": 0.00019665747521386417,
+      "loss": 2.2337,
+      "step": 2066
+    },
+    {
+      "epoch": 0.33072,
+      "grad_norm": 2.8991436723825803,
+      "learning_rate": 0.0001966530450199064,
+      "loss": 2.1128,
+      "step": 2067
+    },
+    {
+      "epoch": 0.33088,
+      "grad_norm": 2.4143986915868005,
+      "learning_rate": 0.00019664861194196263,
+      "loss": 2.1971,
+      "step": 2068
+    },
+    {
+      "epoch": 0.33104,
+      "grad_norm": 2.7335088585279257,
+      "learning_rate": 0.0001966441759801651,
+      "loss": 2.1143,
+      "step": 2069
+    },
+    {
+      "epoch": 0.3312,
+      "grad_norm": 2.9195353492276146,
+      "learning_rate": 0.00019663973713464618,
+      "loss": 2.1375,
+      "step": 2070
+    },
+    {
+      "epoch": 0.33136,
+      "grad_norm": 2.7489415671521304,
+      "learning_rate": 0.00019663529540553832,
+      "loss": 2.088,
+      "step": 2071
+    },
+    {
+      "epoch": 0.33152,
+      "grad_norm": 2.898540470837924,
+      "learning_rate": 0.00019663085079297404,
+      "loss": 2.2645,
+      "step": 2072
+    },
+    {
+      "epoch": 0.33168,
+      "grad_norm": 2.4527446700451874,
+      "learning_rate": 0.000196626403297086,
+      "loss": 2.0694,
+      "step": 2073
+    },
+    {
+      "epoch": 0.33184,
+      "grad_norm": 2.8348847433330415,
+      "learning_rate": 0.00019662195291800689,
+      "loss": 2.1006,
+      "step": 2074
+    },
+    {
+      "epoch": 0.332,
+      "grad_norm": 2.2872882544088484,
+      "learning_rate": 0.0001966174996558695,
+      "loss": 2.2087,
+      "step": 2075
+    },
+    {
+      "epoch": 0.33216,
+      "grad_norm": 2.491983204741037,
+      "learning_rate": 0.00019661304351080672,
+      "loss": 2.1536,
+      "step": 2076
+    },
+    {
+      "epoch": 0.33232,
+      "grad_norm": 2.609663490018398,
+      "learning_rate": 0.00019660858448295146,
+      "loss": 2.2285,
+      "step": 2077
+    },
+    {
+      "epoch": 0.33248,
+      "grad_norm": 2.764029111979444,
+      "learning_rate": 0.00019660412257243684,
+      "loss": 2.2005,
+      "step": 2078
+    },
+    {
+      "epoch": 0.33264,
+      "grad_norm": 2.2650778305741697,
+      "learning_rate": 0.00019659965777939594,
+      "loss": 2.1246,
+      "step": 2079
+    },
+    {
+      "epoch": 0.3328,
+      "grad_norm": 2.7320939549407695,
+      "learning_rate": 0.00019659519010396207,
+      "loss": 2.0333,
+      "step": 2080
+    },
+    {
+      "epoch": 0.33296,
+      "grad_norm": 2.259991646371364,
+      "learning_rate": 0.00019659071954626845,
+      "loss": 2.1124,
+      "step": 2081
+    },
+    {
+      "epoch": 0.33312,
+      "grad_norm": 2.8213894747435604,
+      "learning_rate": 0.00019658624610644848,
+      "loss": 2.193,
+      "step": 2082
+    },
+    {
+      "epoch": 0.33328,
+      "grad_norm": 2.720378141024468,
+      "learning_rate": 0.0001965817697846357,
+      "loss": 2.269,
+      "step": 2083
+    },
+    {
+      "epoch": 0.33344,
+      "grad_norm": 3.3761590605756524,
+      "learning_rate": 0.00019657729058096361,
+      "loss": 2.2388,
+      "step": 2084
+    },
+    {
+      "epoch": 0.3336,
+      "grad_norm": 2.504016740939359,
+      "learning_rate": 0.00019657280849556593,
+      "loss": 2.1079,
+      "step": 2085
+    },
+    {
+      "epoch": 0.33376,
+      "grad_norm": 3.032138704978993,
+      "learning_rate": 0.00019656832352857637,
+      "loss": 2.2057,
+      "step": 2086
+    },
+    {
+      "epoch": 0.33392,
+      "grad_norm": 2.8006715719601214,
+      "learning_rate": 0.00019656383568012874,
+      "loss": 2.1948,
+      "step": 2087
+    },
+    {
+      "epoch": 0.33408,
+      "grad_norm": 2.859902208018383,
+      "learning_rate": 0.00019655934495035697,
+      "loss": 2.0388,
+      "step": 2088
+    },
+    {
+      "epoch": 0.33424,
+      "grad_norm": 2.8716211307186224,
+      "learning_rate": 0.00019655485133939502,
+      "loss": 2.164,
+      "step": 2089
+    },
+    {
+      "epoch": 0.3344,
+      "grad_norm": 2.784965379407728,
+      "learning_rate": 0.00019655035484737702,
+      "loss": 2.2228,
+      "step": 2090
+    },
+    {
+      "epoch": 0.33456,
+      "grad_norm": 2.753362302734115,
+      "learning_rate": 0.00019654585547443715,
+      "loss": 2.1226,
+      "step": 2091
+    },
+    {
+      "epoch": 0.33472,
+      "grad_norm": 2.1546468951736624,
+      "learning_rate": 0.00019654135322070957,
+      "loss": 2.1869,
+      "step": 2092
+    },
+    {
+      "epoch": 0.33488,
+      "grad_norm": 3.423199042254143,
+      "learning_rate": 0.00019653684808632873,
+      "loss": 2.1747,
+      "step": 2093
+    },
+    {
+      "epoch": 0.33504,
+      "grad_norm": 3.3441956341810872,
+      "learning_rate": 0.000196532340071429,
+      "loss": 2.1338,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3352,
+      "grad_norm": 3.2235413388374665,
+      "learning_rate": 0.0001965278291761449,
+      "loss": 2.2454,
+      "step": 2095
+    },
+    {
+      "epoch": 0.33536,
+      "grad_norm": 2.7873027167394477,
+      "learning_rate": 0.00019652331540061105,
+      "loss": 2.0883,
+      "step": 2096
+    },
+    {
+      "epoch": 0.33552,
+      "grad_norm": 3.1255871341344337,
+      "learning_rate": 0.00019651879874496212,
+      "loss": 2.1826,
+      "step": 2097
+    },
+    {
+      "epoch": 0.33568,
+      "grad_norm": 3.737950933511494,
+      "learning_rate": 0.00019651427920933286,
+      "loss": 2.1301,
+      "step": 2098
+    },
+    {
+      "epoch": 0.33584,
+      "grad_norm": 2.3706397075960006,
+      "learning_rate": 0.00019650975679385814,
+      "loss": 2.2003,
+      "step": 2099
+    },
+    {
+      "epoch": 0.336,
+      "grad_norm": 2.8307178979963634,
+      "learning_rate": 0.00019650523149867294,
+      "loss": 2.0937,
+      "step": 2100
+    },
+    {
+      "epoch": 0.33616,
+      "grad_norm": 2.0813417982324975,
+      "learning_rate": 0.00019650070332391222,
+      "loss": 2.1571,
+      "step": 2101
+    },
+    {
+      "epoch": 0.33632,
+      "grad_norm": 3.080720032904708,
+      "learning_rate": 0.00019649617226971113,
+      "loss": 2.1474,
+      "step": 2102
+    },
+    {
+      "epoch": 0.33648,
+      "grad_norm": 2.321871318040983,
+      "learning_rate": 0.00019649163833620488,
+      "loss": 2.143,
+      "step": 2103
+    },
+    {
+      "epoch": 0.33664,
+      "grad_norm": 2.797904381453168,
+      "learning_rate": 0.00019648710152352875,
+      "loss": 2.0925,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3368,
+      "grad_norm": 2.5655861241849096,
+      "learning_rate": 0.0001964825618318181,
+      "loss": 2.3028,
+      "step": 2105
+    },
+    {
+      "epoch": 0.33696,
+      "grad_norm": 2.557259032432377,
+      "learning_rate": 0.0001964780192612084,
+      "loss": 2.175,
+      "step": 2106
+    },
+    {
+      "epoch": 0.33712,
+      "grad_norm": 2.655342119335462,
+      "learning_rate": 0.00019647347381183519,
+      "loss": 2.1771,
+      "step": 2107
+    },
+    {
+      "epoch": 0.33728,
+      "grad_norm": 2.6733392091319232,
+      "learning_rate": 0.00019646892548383412,
+      "loss": 2.1223,
+      "step": 2108
+    },
+    {
+      "epoch": 0.33744,
+      "grad_norm": 3.090053374911528,
+      "learning_rate": 0.00019646437427734084,
+      "loss": 2.2466,
+      "step": 2109
+    },
+    {
+      "epoch": 0.3376,
+      "grad_norm": 2.3683574307410096,
+      "learning_rate": 0.00019645982019249122,
+      "loss": 2.122,
+      "step": 2110
+    },
+    {
+      "epoch": 0.33776,
+      "grad_norm": 2.6051587488005383,
+      "learning_rate": 0.00019645526322942112,
+      "loss": 2.1567,
+      "step": 2111
+    },
+    {
+      "epoch": 0.33792,
+      "grad_norm": 2.2630765201466367,
+      "learning_rate": 0.0001964507033882665,
+      "loss": 2.0142,
+      "step": 2112
+    },
+    {
+      "epoch": 0.33808,
+      "grad_norm": 3.308657023736578,
+      "learning_rate": 0.00019644614066916346,
+      "loss": 2.195,
+      "step": 2113
+    },
+    {
+      "epoch": 0.33824,
+      "grad_norm": 2.11387327787874,
+      "learning_rate": 0.0001964415750722481,
+      "loss": 2.2857,
+      "step": 2114
+    },
+    {
+      "epoch": 0.3384,
+      "grad_norm": 2.828747144188151,
+      "learning_rate": 0.00019643700659765668,
+      "loss": 2.2783,
+      "step": 2115
+    },
+    {
+      "epoch": 0.33856,
+      "grad_norm": 2.7130084968465753,
+      "learning_rate": 0.00019643243524552549,
+      "loss": 2.1431,
+      "step": 2116
+    },
+    {
+      "epoch": 0.33872,
+      "grad_norm": 2.648383589797381,
+      "learning_rate": 0.00019642786101599095,
+      "loss": 2.1138,
+      "step": 2117
+    },
+    {
+      "epoch": 0.33888,
+      "grad_norm": 2.9527571321174704,
+      "learning_rate": 0.00019642328390918955,
+      "loss": 2.1233,
+      "step": 2118
+    },
+    {
+      "epoch": 0.33904,
+      "grad_norm": 2.6343563068772653,
+      "learning_rate": 0.0001964187039252579,
+      "loss": 2.1451,
+      "step": 2119
+    },
+    {
+      "epoch": 0.3392,
+      "grad_norm": 2.7620656608507934,
+      "learning_rate": 0.00019641412106433256,
+      "loss": 2.1208,
+      "step": 2120
+    },
+    {
+      "epoch": 0.33936,
+      "grad_norm": 2.393324249835985,
+      "learning_rate": 0.0001964095353265504,
+      "loss": 2.1139,
+      "step": 2121
+    },
+    {
+      "epoch": 0.33952,
+      "grad_norm": 2.4408739027897814,
+      "learning_rate": 0.00019640494671204818,
+      "loss": 2.0369,
+      "step": 2122
+    },
+    {
+      "epoch": 0.33968,
+      "grad_norm": 2.5442074225835953,
+      "learning_rate": 0.00019640035522096278,
+      "loss": 2.0574,
+      "step": 2123
+    },
+    {
+      "epoch": 0.33984,
+      "grad_norm": 2.9705897519219646,
+      "learning_rate": 0.0001963957608534313,
+      "loss": 2.0266,
+      "step": 2124
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 3.4028271552119502,
+      "learning_rate": 0.00019639116360959074,
+      "loss": 2.2251,
+      "step": 2125
+    },
+    {
+      "epoch": 0.34016,
+      "grad_norm": 2.5168790820019713,
+      "learning_rate": 0.00019638656348957832,
+      "loss": 2.1542,
+      "step": 2126
+    },
+    {
+      "epoch": 0.34032,
+      "grad_norm": 3.157279508697921,
+      "learning_rate": 0.0001963819604935313,
+      "loss": 2.0054,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34048,
+      "grad_norm": 2.7873835208603794,
+      "learning_rate": 0.00019637735462158705,
+      "loss": 2.1432,
+      "step": 2128
+    },
+    {
+      "epoch": 0.34064,
+      "grad_norm": 2.9834602042760134,
+      "learning_rate": 0.00019637274587388292,
+      "loss": 2.27,
+      "step": 2129
+    },
+    {
+      "epoch": 0.3408,
+      "grad_norm": 2.3786982500583784,
+      "learning_rate": 0.00019636813425055654,
+      "loss": 2.1735,
+      "step": 2130
+    },
+    {
+      "epoch": 0.34096,
+      "grad_norm": 2.869829449470534,
+      "learning_rate": 0.00019636351975174543,
+      "loss": 2.164,
+      "step": 2131
+    },
+    {
+      "epoch": 0.34112,
+      "grad_norm": 3.1023519669779756,
+      "learning_rate": 0.0001963589023775873,
+      "loss": 2.1031,
+      "step": 2132
+    },
+    {
+      "epoch": 0.34128,
+      "grad_norm": 2.5330708597938685,
+      "learning_rate": 0.00019635428212821997,
+      "loss": 2.117,
+      "step": 2133
+    },
+    {
+      "epoch": 0.34144,
+      "grad_norm": 2.4617684336426118,
+      "learning_rate": 0.00019634965900378122,
+      "loss": 2.1673,
+      "step": 2134
+    },
+    {
+      "epoch": 0.3416,
+      "grad_norm": 2.588775593191545,
+      "learning_rate": 0.00019634503300440908,
+      "loss": 2.1496,
+      "step": 2135
+    },
+    {
+      "epoch": 0.34176,
+      "grad_norm": 2.698929983135224,
+      "learning_rate": 0.0001963404041302415,
+      "loss": 2.0556,
+      "step": 2136
+    },
+    {
+      "epoch": 0.34192,
+      "grad_norm": 2.9415038709332433,
+      "learning_rate": 0.00019633577238141667,
+      "loss": 2.0902,
+      "step": 2137
+    },
+    {
+      "epoch": 0.34208,
+      "grad_norm": 2.7956040894918917,
+      "learning_rate": 0.00019633113775807277,
+      "loss": 2.1567,
+      "step": 2138
+    },
+    {
+      "epoch": 0.34224,
+      "grad_norm": 2.790438303127627,
+      "learning_rate": 0.0001963265002603481,
+      "loss": 2.1182,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3424,
+      "grad_norm": 3.5514921865870823,
+      "learning_rate": 0.00019632185988838102,
+      "loss": 2.2221,
+      "step": 2140
+    },
+    {
+      "epoch": 0.34256,
+      "grad_norm": 2.5066808066429918,
+      "learning_rate": 0.00019631721664230996,
+      "loss": 2.1804,
+      "step": 2141
+    },
+    {
+      "epoch": 0.34272,
+      "grad_norm": 3.505478830574228,
+      "learning_rate": 0.00019631257052227352,
+      "loss": 2.2105,
+      "step": 2142
+    },
+    {
+      "epoch": 0.34288,
+      "grad_norm": 3.222583618284806,
+      "learning_rate": 0.00019630792152841036,
+      "loss": 2.2033,
+      "step": 2143
+    },
+    {
+      "epoch": 0.34304,
+      "grad_norm": 2.6142864005303723,
+      "learning_rate": 0.0001963032696608591,
+      "loss": 2.1008,
+      "step": 2144
+    },
+    {
+      "epoch": 0.3432,
+      "grad_norm": 2.5888827769793274,
+      "learning_rate": 0.00019629861491975864,
+      "loss": 2.1784,
+      "step": 2145
+    },
+    {
+      "epoch": 0.34336,
+      "grad_norm": 2.60689406216281,
+      "learning_rate": 0.0001962939573052478,
+      "loss": 2.1781,
+      "step": 2146
+    },
+    {
+      "epoch": 0.34352,
+      "grad_norm": 2.9760354433631733,
+      "learning_rate": 0.00019628929681746562,
+      "loss": 2.1061,
+      "step": 2147
+    },
+    {
+      "epoch": 0.34368,
+      "grad_norm": 2.946167419394839,
+      "learning_rate": 0.00019628463345655112,
+      "loss": 2.1483,
+      "step": 2148
+    },
+    {
+      "epoch": 0.34384,
+      "grad_norm": 2.695718035162574,
+      "learning_rate": 0.00019627996722264344,
+      "loss": 2.1256,
+      "step": 2149
+    },
+    {
+      "epoch": 0.344,
+      "grad_norm": 2.9037145479443724,
+      "learning_rate": 0.00019627529811588188,
+      "loss": 2.1529,
+      "step": 2150
+    },
+    {
+      "epoch": 0.34416,
+      "grad_norm": 3.2671575519650613,
+      "learning_rate": 0.00019627062613640568,
+      "loss": 2.0848,
+      "step": 2151
+    },
+    {
+      "epoch": 0.34432,
+      "grad_norm": 3.0604636554192797,
+      "learning_rate": 0.0001962659512843543,
+      "loss": 1.9179,
+      "step": 2152
+    },
+    {
+      "epoch": 0.34448,
+      "grad_norm": 2.4340390089187296,
+      "learning_rate": 0.00019626127355986718,
+      "loss": 2.0395,
+      "step": 2153
+    },
+    {
+      "epoch": 0.34464,
+      "grad_norm": 2.6921167752221353,
+      "learning_rate": 0.00019625659296308392,
+      "loss": 2.1953,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3448,
+      "grad_norm": 3.136729914048863,
+      "learning_rate": 0.0001962519094941442,
+      "loss": 2.1701,
+      "step": 2155
+    },
+    {
+      "epoch": 0.34496,
+      "grad_norm": 2.308564474453527,
+      "learning_rate": 0.00019624722315318778,
+      "loss": 2.1977,
+      "step": 2156
+    },
+    {
+      "epoch": 0.34512,
+      "grad_norm": 2.5628149200071055,
+      "learning_rate": 0.00019624253394035442,
+      "loss": 2.1495,
+      "step": 2157
+    },
+    {
+      "epoch": 0.34528,
+      "grad_norm": 2.4573079334085395,
+      "learning_rate": 0.0001962378418557841,
+      "loss": 2.1788,
+      "step": 2158
+    },
+    {
+      "epoch": 0.34544,
+      "grad_norm": 2.635917718250497,
+      "learning_rate": 0.00019623314689961685,
+      "loss": 2.1181,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3456,
+      "grad_norm": 3.3866230664533075,
+      "learning_rate": 0.0001962284490719927,
+      "loss": 2.2112,
+      "step": 2160
+    },
+    {
+      "epoch": 0.34576,
+      "grad_norm": 2.61758879028882,
+      "learning_rate": 0.00019622374837305185,
+      "loss": 2.0782,
+      "step": 2161
+    },
+    {
+      "epoch": 0.34592,
+      "grad_norm": 3.0274867999788566,
+      "learning_rate": 0.00019621904480293453,
+      "loss": 2.1221,
+      "step": 2162
+    },
+    {
+      "epoch": 0.34608,
+      "grad_norm": 2.531220531379471,
+      "learning_rate": 0.00019621433836178117,
+      "loss": 2.1347,
+      "step": 2163
+    },
+    {
+      "epoch": 0.34624,
+      "grad_norm": 3.4035218963824674,
+      "learning_rate": 0.00019620962904973212,
+      "loss": 2.0829,
+      "step": 2164
+    },
+    {
+      "epoch": 0.3464,
+      "grad_norm": 2.726119676451726,
+      "learning_rate": 0.00019620491686692796,
+      "loss": 2.0743,
+      "step": 2165
+    },
+    {
+      "epoch": 0.34656,
+      "grad_norm": 3.719235425860108,
+      "learning_rate": 0.0001962002018135092,
+      "loss": 2.1504,
+      "step": 2166
+    },
+    {
+      "epoch": 0.34672,
+      "grad_norm": 2.827829568194481,
+      "learning_rate": 0.0001961954838896167,
+      "loss": 2.0306,
+      "step": 2167
+    },
+    {
+      "epoch": 0.34688,
+      "grad_norm": 2.3501934968431466,
+      "learning_rate": 0.00019619076309539105,
+      "loss": 2.0671,
+      "step": 2168
+    },
+    {
+      "epoch": 0.34704,
+      "grad_norm": 2.7954163949284627,
+      "learning_rate": 0.0001961860394309732,
+      "loss": 2.1204,
+      "step": 2169
+    },
+    {
+      "epoch": 0.3472,
+      "grad_norm": 3.2405554522057254,
+      "learning_rate": 0.00019618131289650411,
+      "loss": 2.0892,
+      "step": 2170
+    },
+    {
+      "epoch": 0.34736,
+      "grad_norm": 2.5707407726197125,
+      "learning_rate": 0.0001961765834921248,
+      "loss": 2.2621,
+      "step": 2171
+    },
+    {
+      "epoch": 0.34752,
+      "grad_norm": 3.5109268222493357,
+      "learning_rate": 0.0001961718512179764,
+      "loss": 2.0572,
+      "step": 2172
+    },
+    {
+      "epoch": 0.34768,
+      "grad_norm": 2.4826470915793104,
+      "learning_rate": 0.00019616711607420005,
+      "loss": 2.1303,
+      "step": 2173
+    },
+    {
+      "epoch": 0.34784,
+      "grad_norm": 3.1691193049412067,
+      "learning_rate": 0.00019616237806093713,
+      "loss": 2.1871,
+      "step": 2174
+    },
+    {
+      "epoch": 0.348,
+      "grad_norm": 3.3442022450048747,
+      "learning_rate": 0.00019615763717832892,
+      "loss": 2.1969,
+      "step": 2175
+    },
+    {
+      "epoch": 0.34816,
+      "grad_norm": 3.26083050275074,
+      "learning_rate": 0.00019615289342651696,
+      "loss": 2.1498,
+      "step": 2176
+    },
+    {
+      "epoch": 0.34832,
+      "grad_norm": 4.3694243492976055,
+      "learning_rate": 0.0001961481468056428,
+      "loss": 2.1457,
+      "step": 2177
+    },
+    {
+      "epoch": 0.34848,
+      "grad_norm": 2.9421280917372394,
+      "learning_rate": 0.000196143397315848,
+      "loss": 2.1178,
+      "step": 2178
+    },
+    {
+      "epoch": 0.34864,
+      "grad_norm": 4.348150014873862,
+      "learning_rate": 0.00019613864495727435,
+      "loss": 2.1249,
+      "step": 2179
+    },
+    {
+      "epoch": 0.3488,
+      "grad_norm": 3.1993253800773775,
+      "learning_rate": 0.0001961338897300636,
+      "loss": 2.1773,
+      "step": 2180
+    },
+    {
+      "epoch": 0.34896,
+      "grad_norm": 4.260939059320171,
+      "learning_rate": 0.00019612913163435768,
+      "loss": 2.0749,
+      "step": 2181
+    },
+    {
+      "epoch": 0.34912,
+      "grad_norm": 2.5347529936310185,
+      "learning_rate": 0.00019612437067029857,
+      "loss": 2.0269,
+      "step": 2182
+    },
+    {
+      "epoch": 0.34928,
+      "grad_norm": 2.8567325753668227,
+      "learning_rate": 0.00019611960683802826,
+      "loss": 2.0896,
+      "step": 2183
+    },
+    {
+      "epoch": 0.34944,
+      "grad_norm": 3.4368834078350528,
+      "learning_rate": 0.00019611484013768896,
+      "loss": 2.2207,
+      "step": 2184
+    },
+    {
+      "epoch": 0.3496,
+      "grad_norm": 2.6136817522214906,
+      "learning_rate": 0.00019611007056942288,
+      "loss": 2.1135,
+      "step": 2185
+    },
+    {
+      "epoch": 0.34976,
+      "grad_norm": 3.98106818858096,
+      "learning_rate": 0.00019610529813337234,
+      "loss": 2.089,
+      "step": 2186
+    },
+    {
+      "epoch": 0.34992,
+      "grad_norm": 2.499109310176253,
+      "learning_rate": 0.00019610052282967977,
+      "loss": 2.133,
+      "step": 2187
+    },
+    {
+      "epoch": 0.35008,
+      "grad_norm": 3.581549227532897,
+      "learning_rate": 0.0001960957446584876,
+      "loss": 2.1479,
+      "step": 2188
+    },
+    {
+      "epoch": 0.35024,
+      "grad_norm": 2.427290776774226,
+      "learning_rate": 0.00019609096361993843,
+      "loss": 2.1487,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3504,
+      "grad_norm": 3.422074628741724,
+      "learning_rate": 0.00019608617971417496,
+      "loss": 2.0775,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35056,
+      "grad_norm": 2.5533437303950666,
+      "learning_rate": 0.00019608139294133986,
+      "loss": 2.1163,
+      "step": 2191
+    },
+    {
+      "epoch": 0.35072,
+      "grad_norm": 2.9918008948609276,
+      "learning_rate": 0.00019607660330157604,
+      "loss": 2.1632,
+      "step": 2192
+    },
+    {
+      "epoch": 0.35088,
+      "grad_norm": 2.229109870605333,
+      "learning_rate": 0.00019607181079502633,
+      "loss": 2.1169,
+      "step": 2193
+    },
+    {
+      "epoch": 0.35104,
+      "grad_norm": 3.239762244421452,
+      "learning_rate": 0.00019606701542183383,
+      "loss": 2.0756,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3512,
+      "grad_norm": 2.4802745065796423,
+      "learning_rate": 0.00019606221718214157,
+      "loss": 2.1274,
+      "step": 2195
+    },
+    {
+      "epoch": 0.35136,
+      "grad_norm": 3.333725895950341,
+      "learning_rate": 0.00019605741607609268,
+      "loss": 2.1579,
+      "step": 2196
+    },
+    {
+      "epoch": 0.35152,
+      "grad_norm": 2.677924006744866,
+      "learning_rate": 0.0001960526121038305,
+      "loss": 2.1254,
+      "step": 2197
+    },
+    {
+      "epoch": 0.35168,
+      "grad_norm": 3.103999897959286,
+      "learning_rate": 0.00019604780526549833,
+      "loss": 2.123,
+      "step": 2198
+    },
+    {
+      "epoch": 0.35184,
+      "grad_norm": 2.711800313911488,
+      "learning_rate": 0.0001960429955612396,
+      "loss": 2.2259,
+      "step": 2199
+    },
+    {
+      "epoch": 0.352,
+      "grad_norm": 2.8320908478619593,
+      "learning_rate": 0.00019603818299119785,
+      "loss": 2.2373,
+      "step": 2200
+    },
+    {
+      "epoch": 0.35216,
+      "grad_norm": 2.8997941387369335,
+      "learning_rate": 0.00019603336755551665,
+      "loss": 2.0886,
+      "step": 2201
+    },
+    {
+      "epoch": 0.35232,
+      "grad_norm": 2.419764139384246,
+      "learning_rate": 0.0001960285492543397,
+      "loss": 2.1483,
+      "step": 2202
+    },
+    {
+      "epoch": 0.35248,
+      "grad_norm": 2.6820170324983446,
+      "learning_rate": 0.00019602372808781078,
+      "loss": 2.1033,
+      "step": 2203
+    },
+    {
+      "epoch": 0.35264,
+      "grad_norm": 3.3337354901823733,
+      "learning_rate": 0.00019601890405607372,
+      "loss": 2.102,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3528,
+      "grad_norm": 3.234871792908033,
+      "learning_rate": 0.00019601407715927246,
+      "loss": 2.0671,
+      "step": 2205
+    },
+    {
+      "epoch": 0.35296,
+      "grad_norm": 4.169507544958776,
+      "learning_rate": 0.00019600924739755107,
+      "loss": 2.1919,
+      "step": 2206
+    },
+    {
+      "epoch": 0.35312,
+      "grad_norm": 3.31624807642208,
+      "learning_rate": 0.00019600441477105363,
+      "loss": 2.1904,
+      "step": 2207
+    },
+    {
+      "epoch": 0.35328,
+      "grad_norm": 3.746790156654608,
+      "learning_rate": 0.00019599957927992433,
+      "loss": 2.207,
+      "step": 2208
+    },
+    {
+      "epoch": 0.35344,
+      "grad_norm": 3.3369017272288066,
+      "learning_rate": 0.0001959947409243075,
+      "loss": 2.1534,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3536,
+      "grad_norm": 4.023588522980532,
+      "learning_rate": 0.0001959898997043475,
+      "loss": 2.1525,
+      "step": 2210
+    },
+    {
+      "epoch": 0.35376,
+      "grad_norm": 2.8745906703710653,
+      "learning_rate": 0.0001959850556201887,
+      "loss": 2.0746,
+      "step": 2211
+    },
+    {
+      "epoch": 0.35392,
+      "grad_norm": 3.077939720079665,
+      "learning_rate": 0.00019598020867197575,
+      "loss": 2.1003,
+      "step": 2212
+    },
+    {
+      "epoch": 0.35408,
+      "grad_norm": 2.6882304614980344,
+      "learning_rate": 0.00019597535885985324,
+      "loss": 2.0435,
+      "step": 2213
+    },
+    {
+      "epoch": 0.35424,
+      "grad_norm": 2.6942848514819704,
+      "learning_rate": 0.00019597050618396585,
+      "loss": 2.0766,
+      "step": 2214
+    },
+    {
+      "epoch": 0.3544,
+      "grad_norm": 2.9187011349804477,
+      "learning_rate": 0.0001959656506444584,
+      "loss": 2.0979,
+      "step": 2215
+    },
+    {
+      "epoch": 0.35456,
+      "grad_norm": 3.231490609118702,
+      "learning_rate": 0.00019596079224147578,
+      "loss": 2.0754,
+      "step": 2216
+    },
+    {
+      "epoch": 0.35472,
+      "grad_norm": 3.094058740266997,
+      "learning_rate": 0.00019595593097516292,
+      "loss": 1.9782,
+      "step": 2217
+    },
+    {
+      "epoch": 0.35488,
+      "grad_norm": 2.3448181519738287,
+      "learning_rate": 0.00019595106684566492,
+      "loss": 2.2544,
+      "step": 2218
+    },
+    {
+      "epoch": 0.35504,
+      "grad_norm": 2.5966865324758315,
+      "learning_rate": 0.0001959461998531269,
+      "loss": 2.0467,
+      "step": 2219
+    },
+    {
+      "epoch": 0.3552,
+      "grad_norm": 2.5197168729801085,
+      "learning_rate": 0.0001959413299976941,
+      "loss": 2.2562,
+      "step": 2220
+    },
+    {
+      "epoch": 0.35536,
+      "grad_norm": 4.244270105406763,
+      "learning_rate": 0.0001959364572795118,
+      "loss": 2.0465,
+      "step": 2221
+    },
+    {
+      "epoch": 0.35552,
+      "grad_norm": 2.172868376516801,
+      "learning_rate": 0.00019593158169872542,
+      "loss": 2.2042,
+      "step": 2222
+    },
+    {
+      "epoch": 0.35568,
+      "grad_norm": 3.3801743206839543,
+      "learning_rate": 0.00019592670325548043,
+      "loss": 2.0489,
+      "step": 2223
+    },
+    {
+      "epoch": 0.35584,
+      "grad_norm": 2.5190342874381084,
+      "learning_rate": 0.00019592182194992237,
+      "loss": 2.2715,
+      "step": 2224
+    },
+    {
+      "epoch": 0.356,
+      "grad_norm": 3.159745262437379,
+      "learning_rate": 0.00019591693778219695,
+      "loss": 2.1363,
+      "step": 2225
+    },
+    {
+      "epoch": 0.35616,
+      "grad_norm": 2.380244604332379,
+      "learning_rate": 0.00019591205075244984,
+      "loss": 2.0297,
+      "step": 2226
+    },
+    {
+      "epoch": 0.35632,
+      "grad_norm": 2.875030309753666,
+      "learning_rate": 0.00019590716086082693,
+      "loss": 2.0962,
+      "step": 2227
+    },
+    {
+      "epoch": 0.35648,
+      "grad_norm": 3.10354050659477,
+      "learning_rate": 0.00019590226810747407,
+      "loss": 2.2233,
+      "step": 2228
+    },
+    {
+      "epoch": 0.35664,
+      "grad_norm": 2.403142677651436,
+      "learning_rate": 0.00019589737249253725,
+      "loss": 2.125,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3568,
+      "grad_norm": 3.4656148143144367,
+      "learning_rate": 0.00019589247401616259,
+      "loss": 2.0898,
+      "step": 2230
+    },
+    {
+      "epoch": 0.35696,
+      "grad_norm": 2.972511517861812,
+      "learning_rate": 0.00019588757267849626,
+      "loss": 2.134,
+      "step": 2231
+    },
+    {
+      "epoch": 0.35712,
+      "grad_norm": 3.763019631079269,
+      "learning_rate": 0.00019588266847968448,
+      "loss": 2.2033,
+      "step": 2232
+    },
+    {
+      "epoch": 0.35728,
+      "grad_norm": 4.536438969067655,
+      "learning_rate": 0.00019587776141987358,
+      "loss": 2.0491,
+      "step": 2233
+    },
+    {
+      "epoch": 0.35744,
+      "grad_norm": 2.6500992213743366,
+      "learning_rate": 0.00019587285149920997,
+      "loss": 2.0583,
+      "step": 2234
+    },
+    {
+      "epoch": 0.3576,
+      "grad_norm": 3.209591469684893,
+      "learning_rate": 0.00019586793871784018,
+      "loss": 2.0605,
+      "step": 2235
+    },
+    {
+      "epoch": 0.35776,
+      "grad_norm": 2.653056800113723,
+      "learning_rate": 0.00019586302307591082,
+      "loss": 2.125,
+      "step": 2236
+    },
+    {
+      "epoch": 0.35792,
+      "grad_norm": 2.366202946355986,
+      "learning_rate": 0.0001958581045735685,
+      "loss": 2.0844,
+      "step": 2237
+    },
+    {
+      "epoch": 0.35808,
+      "grad_norm": 2.4507910292482458,
+      "learning_rate": 0.00019585318321096006,
+      "loss": 2.019,
+      "step": 2238
+    },
+    {
+      "epoch": 0.35824,
+      "grad_norm": 2.4676693255667157,
+      "learning_rate": 0.00019584825898823227,
+      "loss": 2.0296,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3584,
+      "grad_norm": 3.541119738466235,
+      "learning_rate": 0.00019584333190553213,
+      "loss": 2.0517,
+      "step": 2240
+    },
+    {
+      "epoch": 0.35856,
+      "grad_norm": 3.2143724956051343,
+      "learning_rate": 0.00019583840196300657,
+      "loss": 2.1668,
+      "step": 2241
+    },
+    {
+      "epoch": 0.35872,
+      "grad_norm": 2.6648870042888384,
+      "learning_rate": 0.0001958334691608028,
+      "loss": 2.0638,
+      "step": 2242
+    },
+    {
+      "epoch": 0.35888,
+      "grad_norm": 3.6954029459644935,
+      "learning_rate": 0.00019582853349906794,
+      "loss": 2.0463,
+      "step": 2243
+    },
+    {
+      "epoch": 0.35904,
+      "grad_norm": 3.2784091011427807,
+      "learning_rate": 0.00019582359497794926,
+      "loss": 2.1654,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3592,
+      "grad_norm": 3.1473197572464793,
+      "learning_rate": 0.00019581865359759415,
+      "loss": 2.1718,
+      "step": 2245
+    },
+    {
+      "epoch": 0.35936,
+      "grad_norm": 2.884891411887831,
+      "learning_rate": 0.00019581370935815005,
+      "loss": 2.0811,
+      "step": 2246
+    },
+    {
+      "epoch": 0.35952,
+      "grad_norm": 2.7474854614155046,
+      "learning_rate": 0.00019580876225976446,
+      "loss": 2.103,
+      "step": 2247
+    },
+    {
+      "epoch": 0.35968,
+      "grad_norm": 3.2468359865342444,
+      "learning_rate": 0.00019580381230258501,
+      "loss": 2.0827,
+      "step": 2248
+    },
+    {
+      "epoch": 0.35984,
+      "grad_norm": 2.4639182858690902,
+      "learning_rate": 0.00019579885948675942,
+      "loss": 2.1117,
+      "step": 2249
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 2.3586345149804546,
+      "learning_rate": 0.00019579390381243544,
+      "loss": 2.2423,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36016,
+      "grad_norm": 3.0127316461034526,
+      "learning_rate": 0.00019578894527976098,
+      "loss": 2.1232,
+      "step": 2251
+    },
+    {
+      "epoch": 0.36032,
+      "grad_norm": 2.5292615746162097,
+      "learning_rate": 0.00019578398388888396,
+      "loss": 2.1876,
+      "step": 2252
+    },
+    {
+      "epoch": 0.36048,
+      "grad_norm": 2.492317982175676,
+      "learning_rate": 0.0001957790196399524,
+      "loss": 2.0109,
+      "step": 2253
+    },
+    {
+      "epoch": 0.36064,
+      "grad_norm": 2.6809825758691814,
+      "learning_rate": 0.0001957740525331145,
+      "loss": 2.2422,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3608,
+      "grad_norm": 3.3303746539597636,
+      "learning_rate": 0.00019576908256851842,
+      "loss": 2.1603,
+      "step": 2255
+    },
+    {
+      "epoch": 0.36096,
+      "grad_norm": 2.581588804360019,
+      "learning_rate": 0.00019576410974631245,
+      "loss": 2.0641,
+      "step": 2256
+    },
+    {
+      "epoch": 0.36112,
+      "grad_norm": 2.279888617889013,
+      "learning_rate": 0.000195759134066645,
+      "loss": 2.1801,
+      "step": 2257
+    },
+    {
+      "epoch": 0.36128,
+      "grad_norm": 2.9252289779122513,
+      "learning_rate": 0.00019575415552966453,
+      "loss": 2.1074,
+      "step": 2258
+    },
+    {
+      "epoch": 0.36144,
+      "grad_norm": 2.6620878850117777,
+      "learning_rate": 0.00019574917413551962,
+      "loss": 2.244,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3616,
+      "grad_norm": 3.1739217897129026,
+      "learning_rate": 0.00019574418988435883,
+      "loss": 2.071,
+      "step": 2260
+    },
+    {
+      "epoch": 0.36176,
+      "grad_norm": 2.500912352228855,
+      "learning_rate": 0.00019573920277633091,
+      "loss": 2.1506,
+      "step": 2261
+    },
+    {
+      "epoch": 0.36192,
+      "grad_norm": 2.219689532602336,
+      "learning_rate": 0.00019573421281158474,
+      "loss": 2.0921,
+      "step": 2262
+    },
+    {
+      "epoch": 0.36208,
+      "grad_norm": 2.851007686777188,
+      "learning_rate": 0.00019572921999026912,
+      "loss": 2.0915,
+      "step": 2263
+    },
+    {
+      "epoch": 0.36224,
+      "grad_norm": 2.2873528759994453,
+      "learning_rate": 0.00019572422431253308,
+      "loss": 2.1622,
+      "step": 2264
+    },
+    {
+      "epoch": 0.3624,
+      "grad_norm": 2.8183076096461277,
+      "learning_rate": 0.0001957192257785257,
+      "loss": 2.1964,
+      "step": 2265
+    },
+    {
+      "epoch": 0.36256,
+      "grad_norm": 2.536484217668223,
+      "learning_rate": 0.00019571422438839604,
+      "loss": 1.9727,
+      "step": 2266
+    },
+    {
+      "epoch": 0.36272,
+      "grad_norm": 2.6501258080992427,
+      "learning_rate": 0.00019570922014229344,
+      "loss": 1.9874,
+      "step": 2267
+    },
+    {
+      "epoch": 0.36288,
+      "grad_norm": 2.651712033981178,
+      "learning_rate": 0.00019570421304036714,
+      "loss": 1.9709,
+      "step": 2268
+    },
+    {
+      "epoch": 0.36304,
+      "grad_norm": 2.5209530619231555,
+      "learning_rate": 0.0001956992030827666,
+      "loss": 2.0918,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3632,
+      "grad_norm": 2.411888427499294,
+      "learning_rate": 0.0001956941902696413,
+      "loss": 2.0883,
+      "step": 2270
+    },
+    {
+      "epoch": 0.36336,
+      "grad_norm": 3.0896723260163395,
+      "learning_rate": 0.0001956891746011408,
+      "loss": 2.156,
+      "step": 2271
+    },
+    {
+      "epoch": 0.36352,
+      "grad_norm": 2.5864303167465232,
+      "learning_rate": 0.00019568415607741475,
+      "loss": 2.0852,
+      "step": 2272
+    },
+    {
+      "epoch": 0.36368,
+      "grad_norm": 2.491871189117126,
+      "learning_rate": 0.00019567913469861292,
+      "loss": 2.0436,
+      "step": 2273
+    },
+    {
+      "epoch": 0.36384,
+      "grad_norm": 2.2814671763066103,
+      "learning_rate": 0.0001956741104648851,
+      "loss": 1.9469,
+      "step": 2274
+    },
+    {
+      "epoch": 0.364,
+      "grad_norm": 2.8867132664481283,
+      "learning_rate": 0.00019566908337638128,
+      "loss": 2.0488,
+      "step": 2275
+    },
+    {
+      "epoch": 0.36416,
+      "grad_norm": 3.237351193982214,
+      "learning_rate": 0.0001956640534332514,
+      "loss": 2.1457,
+      "step": 2276
+    },
+    {
+      "epoch": 0.36432,
+      "grad_norm": 2.8254041458795838,
+      "learning_rate": 0.00019565902063564555,
+      "loss": 2.0656,
+      "step": 2277
+    },
+    {
+      "epoch": 0.36448,
+      "grad_norm": 2.741772068967418,
+      "learning_rate": 0.00019565398498371393,
+      "loss": 2.1662,
+      "step": 2278
+    },
+    {
+      "epoch": 0.36464,
+      "grad_norm": 3.0084284773843843,
+      "learning_rate": 0.0001956489464776068,
+      "loss": 1.9543,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3648,
+      "grad_norm": 2.280317006904864,
+      "learning_rate": 0.00019564390511747444,
+      "loss": 2.1101,
+      "step": 2280
+    },
+    {
+      "epoch": 0.36496,
+      "grad_norm": 2.9420332323119247,
+      "learning_rate": 0.00019563886090346734,
+      "loss": 2.0176,
+      "step": 2281
+    },
+    {
+      "epoch": 0.36512,
+      "grad_norm": 2.8485676850683506,
+      "learning_rate": 0.000195633813835736,
+      "loss": 2.1848,
+      "step": 2282
+    },
+    {
+      "epoch": 0.36528,
+      "grad_norm": 3.2756075145541486,
+      "learning_rate": 0.00019562876391443097,
+      "loss": 2.2179,
+      "step": 2283
+    },
+    {
+      "epoch": 0.36544,
+      "grad_norm": 2.5577457440003246,
+      "learning_rate": 0.00019562371113970303,
+      "loss": 2.0801,
+      "step": 2284
+    },
+    {
+      "epoch": 0.3656,
+      "grad_norm": 2.8030405615616303,
+      "learning_rate": 0.00019561865551170283,
+      "loss": 2.0319,
+      "step": 2285
+    },
+    {
+      "epoch": 0.36576,
+      "grad_norm": 3.5322060393928187,
+      "learning_rate": 0.00019561359703058128,
+      "loss": 2.0938,
+      "step": 2286
+    },
+    {
+      "epoch": 0.36592,
+      "grad_norm": 2.95198873011555,
+      "learning_rate": 0.00019560853569648937,
+      "loss": 2.0709,
+      "step": 2287
+    },
+    {
+      "epoch": 0.36608,
+      "grad_norm": 3.994594909790021,
+      "learning_rate": 0.00019560347150957802,
+      "loss": 2.0302,
+      "step": 2288
+    },
+    {
+      "epoch": 0.36624,
+      "grad_norm": 3.048220989026546,
+      "learning_rate": 0.0001955984044699984,
+      "loss": 2.1264,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3664,
+      "grad_norm": 4.121099261702788,
+      "learning_rate": 0.00019559333457790174,
+      "loss": 2.0727,
+      "step": 2290
+    },
+    {
+      "epoch": 0.36656,
+      "grad_norm": 2.9617447035488653,
+      "learning_rate": 0.00019558826183343925,
+      "loss": 2.059,
+      "step": 2291
+    },
+    {
+      "epoch": 0.36672,
+      "grad_norm": 3.329943092728877,
+      "learning_rate": 0.00019558318623676227,
+      "loss": 2.0947,
+      "step": 2292
+    },
+    {
+      "epoch": 0.36688,
+      "grad_norm": 2.7870641652203396,
+      "learning_rate": 0.0001955781077880223,
+      "loss": 2.0613,
+      "step": 2293
+    },
+    {
+      "epoch": 0.36704,
+      "grad_norm": 3.0722102161855003,
+      "learning_rate": 0.00019557302648737094,
+      "loss": 1.9649,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3672,
+      "grad_norm": 2.478649411482598,
+      "learning_rate": 0.00019556794233495966,
+      "loss": 2.1094,
+      "step": 2295
+    },
+    {
+      "epoch": 0.36736,
+      "grad_norm": 2.9157291400754572,
+      "learning_rate": 0.00019556285533094028,
+      "loss": 2.0807,
+      "step": 2296
+    },
+    {
+      "epoch": 0.36752,
+      "grad_norm": 2.742189871947558,
+      "learning_rate": 0.00019555776547546452,
+      "loss": 2.0443,
+      "step": 2297
+    },
+    {
+      "epoch": 0.36768,
+      "grad_norm": 3.5378895726182082,
+      "learning_rate": 0.00019555267276868428,
+      "loss": 2.0896,
+      "step": 2298
+    },
+    {
+      "epoch": 0.36784,
+      "grad_norm": 2.766432397347542,
+      "learning_rate": 0.00019554757721075152,
+      "loss": 2.1675,
+      "step": 2299
+    },
+    {
+      "epoch": 0.368,
+      "grad_norm": 2.690260474587619,
+      "learning_rate": 0.0001955424788018183,
+      "loss": 2.1039,
+      "step": 2300
+    },
+    {
+      "epoch": 0.36816,
+      "grad_norm": 2.8039210440347566,
+      "learning_rate": 0.00019553737754203672,
+      "loss": 1.999,
+      "step": 2301
+    },
+    {
+      "epoch": 0.36832,
+      "grad_norm": 2.7288934547417574,
+      "learning_rate": 0.000195532273431559,
+      "loss": 2.1167,
+      "step": 2302
+    },
+    {
+      "epoch": 0.36848,
+      "grad_norm": 2.8825882934780203,
+      "learning_rate": 0.00019552716647053747,
+      "loss": 2.2057,
+      "step": 2303
+    },
+    {
+      "epoch": 0.36864,
+      "grad_norm": 2.852636285558402,
+      "learning_rate": 0.00019552205665912447,
+      "loss": 2.0607,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3688,
+      "grad_norm": 2.4815510271783983,
+      "learning_rate": 0.0001955169439974725,
+      "loss": 2.0046,
+      "step": 2305
+    },
+    {
+      "epoch": 0.36896,
+      "grad_norm": 2.4027079204427984,
+      "learning_rate": 0.0001955118284857341,
+      "loss": 2.1261,
+      "step": 2306
+    },
+    {
+      "epoch": 0.36912,
+      "grad_norm": 2.512375031605561,
+      "learning_rate": 0.00019550671012406192,
+      "loss": 2.0176,
+      "step": 2307
+    },
+    {
+      "epoch": 0.36928,
+      "grad_norm": 2.627404528295075,
+      "learning_rate": 0.00019550158891260864,
+      "loss": 2.0708,
+      "step": 2308
+    },
+    {
+      "epoch": 0.36944,
+      "grad_norm": 2.5543280132209754,
+      "learning_rate": 0.00019549646485152717,
+      "loss": 1.9765,
+      "step": 2309
+    },
+    {
+      "epoch": 0.3696,
+      "grad_norm": 3.387121292960407,
+      "learning_rate": 0.0001954913379409703,
+      "loss": 2.2123,
+      "step": 2310
+    },
+    {
+      "epoch": 0.36976,
+      "grad_norm": 2.2872432703828296,
+      "learning_rate": 0.00019548620818109105,
+      "loss": 2.023,
+      "step": 2311
+    },
+    {
+      "epoch": 0.36992,
+      "grad_norm": 2.689485993644162,
+      "learning_rate": 0.00019548107557204253,
+      "loss": 1.9962,
+      "step": 2312
+    },
+    {
+      "epoch": 0.37008,
+      "grad_norm": 2.30111538420132,
+      "learning_rate": 0.0001954759401139778,
+      "loss": 2.1587,
+      "step": 2313
+    },
+    {
+      "epoch": 0.37024,
+      "grad_norm": 3.0077314683678362,
+      "learning_rate": 0.00019547080180705014,
+      "loss": 2.0728,
+      "step": 2314
+    },
+    {
+      "epoch": 0.3704,
+      "grad_norm": 2.4306165513145976,
+      "learning_rate": 0.00019546566065141286,
+      "loss": 2.0781,
+      "step": 2315
+    },
+    {
+      "epoch": 0.37056,
+      "grad_norm": 2.662128138228974,
+      "learning_rate": 0.00019546051664721942,
+      "loss": 2.0713,
+      "step": 2316
+    },
+    {
+      "epoch": 0.37072,
+      "grad_norm": 3.0121601636053725,
+      "learning_rate": 0.00019545536979462323,
+      "loss": 2.2144,
+      "step": 2317
+    },
+    {
+      "epoch": 0.37088,
+      "grad_norm": 2.94810911115486,
+      "learning_rate": 0.00019545022009377793,
+      "loss": 2.0793,
+      "step": 2318
+    },
+    {
+      "epoch": 0.37104,
+      "grad_norm": 2.5594144910956858,
+      "learning_rate": 0.0001954450675448371,
+      "loss": 2.0289,
+      "step": 2319
+    },
+    {
+      "epoch": 0.3712,
+      "grad_norm": 2.4022206474386594,
+      "learning_rate": 0.00019543991214795458,
+      "loss": 2.1436,
+      "step": 2320
+    },
+    {
+      "epoch": 0.37136,
+      "grad_norm": 2.4526038396914784,
+      "learning_rate": 0.00019543475390328415,
+      "loss": 2.1278,
+      "step": 2321
+    },
+    {
+      "epoch": 0.37152,
+      "grad_norm": 2.4047515965434028,
+      "learning_rate": 0.0001954295928109797,
+      "loss": 2.1512,
+      "step": 2322
+    },
+    {
+      "epoch": 0.37168,
+      "grad_norm": 3.5027409260029856,
+      "learning_rate": 0.00019542442887119526,
+      "loss": 2.0004,
+      "step": 2323
+    },
+    {
+      "epoch": 0.37184,
+      "grad_norm": 2.326929378319313,
+      "learning_rate": 0.0001954192620840849,
+      "loss": 2.003,
+      "step": 2324
+    },
+    {
+      "epoch": 0.372,
+      "grad_norm": 3.821706866482892,
+      "learning_rate": 0.00019541409244980281,
+      "loss": 2.0693,
+      "step": 2325
+    },
+    {
+      "epoch": 0.37216,
+      "grad_norm": 2.9301291708813957,
+      "learning_rate": 0.00019540891996850328,
+      "loss": 2.0787,
+      "step": 2326
+    },
+    {
+      "epoch": 0.37232,
+      "grad_norm": 2.856883931637023,
+      "learning_rate": 0.00019540374464034054,
+      "loss": 2.0467,
+      "step": 2327
+    },
+    {
+      "epoch": 0.37248,
+      "grad_norm": 1.9840756119428833,
+      "learning_rate": 0.00019539856646546912,
+      "loss": 2.0446,
+      "step": 2328
+    },
+    {
+      "epoch": 0.37264,
+      "grad_norm": 3.142220948929022,
+      "learning_rate": 0.00019539338544404347,
+      "loss": 2.1016,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3728,
+      "grad_norm": 3.1505966898647055,
+      "learning_rate": 0.0001953882015762182,
+      "loss": 2.1964,
+      "step": 2330
+    },
+    {
+      "epoch": 0.37296,
+      "grad_norm": 2.7786497036896294,
+      "learning_rate": 0.00019538301486214798,
+      "loss": 2.1487,
+      "step": 2331
+    },
+    {
+      "epoch": 0.37312,
+      "grad_norm": 3.4619941323857435,
+      "learning_rate": 0.0001953778253019876,
+      "loss": 2.0897,
+      "step": 2332
+    },
+    {
+      "epoch": 0.37328,
+      "grad_norm": 3.4812552589642647,
+      "learning_rate": 0.0001953726328958919,
+      "loss": 2.0641,
+      "step": 2333
+    },
+    {
+      "epoch": 0.37344,
+      "grad_norm": 3.4040733798942044,
+      "learning_rate": 0.00019536743764401578,
+      "loss": 2.0927,
+      "step": 2334
+    },
+    {
+      "epoch": 0.3736,
+      "grad_norm": 2.97912399178212,
+      "learning_rate": 0.0001953622395465143,
+      "loss": 2.0544,
+      "step": 2335
+    },
+    {
+      "epoch": 0.37376,
+      "grad_norm": 3.309441163159355,
+      "learning_rate": 0.00019535703860354254,
+      "loss": 2.0961,
+      "step": 2336
+    },
+    {
+      "epoch": 0.37392,
+      "grad_norm": 3.416904141890157,
+      "learning_rate": 0.0001953518348152557,
+      "loss": 1.973,
+      "step": 2337
+    },
+    {
+      "epoch": 0.37408,
+      "grad_norm": 2.8901552866801525,
+      "learning_rate": 0.00019534662818180908,
+      "loss": 2.0309,
+      "step": 2338
+    },
+    {
+      "epoch": 0.37424,
+      "grad_norm": 2.975586360190211,
+      "learning_rate": 0.00019534141870335798,
+      "loss": 2.0425,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3744,
+      "grad_norm": 2.447715500392001,
+      "learning_rate": 0.0001953362063800579,
+      "loss": 2.0109,
+      "step": 2340
+    },
+    {
+      "epoch": 0.37456,
+      "grad_norm": 2.3385267145509028,
+      "learning_rate": 0.00019533099121206432,
+      "loss": 2.0085,
+      "step": 2341
+    },
+    {
+      "epoch": 0.37472,
+      "grad_norm": 2.924956689802994,
+      "learning_rate": 0.00019532577319953287,
+      "loss": 1.9975,
+      "step": 2342
+    },
+    {
+      "epoch": 0.37488,
+      "grad_norm": 2.5359284218419558,
+      "learning_rate": 0.00019532055234261925,
+      "loss": 2.038,
+      "step": 2343
+    },
+    {
+      "epoch": 0.37504,
+      "grad_norm": 3.275453914071824,
+      "learning_rate": 0.0001953153286414793,
+      "loss": 2.1149,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3752,
+      "grad_norm": 3.0311630837755246,
+      "learning_rate": 0.00019531010209626875,
+      "loss": 2.0601,
+      "step": 2345
+    },
+    {
+      "epoch": 0.37536,
+      "grad_norm": 2.184232506460452,
+      "learning_rate": 0.00019530487270714368,
+      "loss": 2.06,
+      "step": 2346
+    },
+    {
+      "epoch": 0.37552,
+      "grad_norm": 2.967521286804141,
+      "learning_rate": 0.00019529964047426007,
+      "loss": 2.1485,
+      "step": 2347
+    },
+    {
+      "epoch": 0.37568,
+      "grad_norm": 2.713434533385111,
+      "learning_rate": 0.00019529440539777407,
+      "loss": 2.1554,
+      "step": 2348
+    },
+    {
+      "epoch": 0.37584,
+      "grad_norm": 2.934883631619277,
+      "learning_rate": 0.00019528916747784186,
+      "loss": 2.0102,
+      "step": 2349
+    },
+    {
+      "epoch": 0.376,
+      "grad_norm": 2.365581292162257,
+      "learning_rate": 0.00019528392671461974,
+      "loss": 1.9834,
+      "step": 2350
+    },
+    {
+      "epoch": 0.37616,
+      "grad_norm": 3.0894188157135187,
+      "learning_rate": 0.00019527868310826408,
+      "loss": 2.0829,
+      "step": 2351
+    },
+    {
+      "epoch": 0.37632,
+      "grad_norm": 3.032609098420297,
+      "learning_rate": 0.00019527343665893136,
+      "loss": 1.997,
+      "step": 2352
+    },
+    {
+      "epoch": 0.37648,
+      "grad_norm": 2.395607355206257,
+      "learning_rate": 0.00019526818736677814,
+      "loss": 2.1493,
+      "step": 2353
+    },
+    {
+      "epoch": 0.37664,
+      "grad_norm": 2.7410920757125425,
+      "learning_rate": 0.000195262935231961,
+      "loss": 2.0314,
+      "step": 2354
+    },
+    {
+      "epoch": 0.3768,
+      "grad_norm": 2.134857207862604,
+      "learning_rate": 0.00019525768025463672,
+      "loss": 2.0248,
+      "step": 2355
+    },
+    {
+      "epoch": 0.37696,
+      "grad_norm": 2.3664978655347553,
+      "learning_rate": 0.00019525242243496202,
+      "loss": 2.0557,
+      "step": 2356
+    },
+    {
+      "epoch": 0.37712,
+      "grad_norm": 2.2364403019801773,
+      "learning_rate": 0.00019524716177309385,
+      "loss": 1.9629,
+      "step": 2357
+    },
+    {
+      "epoch": 0.37728,
+      "grad_norm": 2.618819727192434,
+      "learning_rate": 0.00019524189826918917,
+      "loss": 2.1668,
+      "step": 2358
+    },
+    {
+      "epoch": 0.37744,
+      "grad_norm": 2.5916649957169113,
+      "learning_rate": 0.00019523663192340504,
+      "loss": 1.8663,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3776,
+      "grad_norm": 3.0490786263965903,
+      "learning_rate": 0.00019523136273589858,
+      "loss": 2.1564,
+      "step": 2360
+    },
+    {
+      "epoch": 0.37776,
+      "grad_norm": 2.4325397505498567,
+      "learning_rate": 0.00019522609070682704,
+      "loss": 2.136,
+      "step": 2361
+    },
+    {
+      "epoch": 0.37792,
+      "grad_norm": 2.7498756683984817,
+      "learning_rate": 0.0001952208158363477,
+      "loss": 1.9544,
+      "step": 2362
+    },
+    {
+      "epoch": 0.37808,
+      "grad_norm": 2.633029585161244,
+      "learning_rate": 0.0001952155381246179,
+      "loss": 2.0448,
+      "step": 2363
+    },
+    {
+      "epoch": 0.37824,
+      "grad_norm": 2.038712277505896,
+      "learning_rate": 0.00019521025757179524,
+      "loss": 2.1109,
+      "step": 2364
+    },
+    {
+      "epoch": 0.3784,
+      "grad_norm": 2.5831323123125647,
+      "learning_rate": 0.00019520497417803724,
+      "loss": 2.0353,
+      "step": 2365
+    },
+    {
+      "epoch": 0.37856,
+      "grad_norm": 2.6358815182320483,
+      "learning_rate": 0.00019519968794350155,
+      "loss": 2.0643,
+      "step": 2366
+    },
+    {
+      "epoch": 0.37872,
+      "grad_norm": 2.9170426355111987,
+      "learning_rate": 0.00019519439886834587,
+      "loss": 2.1344,
+      "step": 2367
+    },
+    {
+      "epoch": 0.37888,
+      "grad_norm": 2.666436921641179,
+      "learning_rate": 0.00019518910695272805,
+      "loss": 2.0159,
+      "step": 2368
+    },
+    {
+      "epoch": 0.37904,
+      "grad_norm": 2.839380351104755,
+      "learning_rate": 0.00019518381219680596,
+      "loss": 2.0286,
+      "step": 2369
+    },
+    {
+      "epoch": 0.3792,
+      "grad_norm": 2.3440998457039544,
+      "learning_rate": 0.00019517851460073765,
+      "loss": 2.0545,
+      "step": 2370
+    },
+    {
+      "epoch": 0.37936,
+      "grad_norm": 2.332431387141718,
+      "learning_rate": 0.00019517321416468112,
+      "loss": 2.107,
+      "step": 2371
+    },
+    {
+      "epoch": 0.37952,
+      "grad_norm": 2.552224699390129,
+      "learning_rate": 0.0001951679108887946,
+      "loss": 2.0343,
+      "step": 2372
+    },
+    {
+      "epoch": 0.37968,
+      "grad_norm": 2.7708674190490767,
+      "learning_rate": 0.00019516260477323626,
+      "loss": 2.0157,
+      "step": 2373
+    },
+    {
+      "epoch": 0.37984,
+      "grad_norm": 2.20044919224978,
+      "learning_rate": 0.00019515729581816445,
+      "loss": 2.1258,
+      "step": 2374
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 3.079894959686821,
+      "learning_rate": 0.00019515198402373765,
+      "loss": 2.0515,
+      "step": 2375
+    },
+    {
+      "epoch": 0.38016,
+      "grad_norm": 2.4312178797119124,
+      "learning_rate": 0.00019514666939011424,
+      "loss": 2.0675,
+      "step": 2376
+    },
+    {
+      "epoch": 0.38032,
+      "grad_norm": 2.9888255938150303,
+      "learning_rate": 0.00019514135191745288,
+      "loss": 2.0271,
+      "step": 2377
+    },
+    {
+      "epoch": 0.38048,
+      "grad_norm": 2.2327576885218403,
+      "learning_rate": 0.0001951360316059122,
+      "loss": 1.9786,
+      "step": 2378
+    },
+    {
+      "epoch": 0.38064,
+      "grad_norm": 2.999646989115491,
+      "learning_rate": 0.00019513070845565098,
+      "loss": 2.0438,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3808,
+      "grad_norm": 2.933145662386062,
+      "learning_rate": 0.00019512538246682802,
+      "loss": 2.0208,
+      "step": 2380
+    },
+    {
+      "epoch": 0.38096,
+      "grad_norm": 2.3182740185479735,
+      "learning_rate": 0.0001951200536396023,
+      "loss": 2.0444,
+      "step": 2381
+    },
+    {
+      "epoch": 0.38112,
+      "grad_norm": 3.1949114838231325,
+      "learning_rate": 0.00019511472197413277,
+      "loss": 2.0248,
+      "step": 2382
+    },
+    {
+      "epoch": 0.38128,
+      "grad_norm": 2.7151602635592265,
+      "learning_rate": 0.0001951093874705785,
+      "loss": 1.9934,
+      "step": 2383
+    },
+    {
+      "epoch": 0.38144,
+      "grad_norm": 2.6280848106656345,
+      "learning_rate": 0.00019510405012909874,
+      "loss": 2.1201,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3816,
+      "grad_norm": 2.674913653016646,
+      "learning_rate": 0.0001950987099498527,
+      "loss": 1.9445,
+      "step": 2385
+    },
+    {
+      "epoch": 0.38176,
+      "grad_norm": 2.0674066971285265,
+      "learning_rate": 0.0001950933669329997,
+      "loss": 2.0061,
+      "step": 2386
+    },
+    {
+      "epoch": 0.38192,
+      "grad_norm": 2.5272502830066688,
+      "learning_rate": 0.0001950880210786992,
+      "loss": 2.035,
+      "step": 2387
+    },
+    {
+      "epoch": 0.38208,
+      "grad_norm": 2.6702728896662324,
+      "learning_rate": 0.0001950826723871107,
+      "loss": 2.0141,
+      "step": 2388
+    },
+    {
+      "epoch": 0.38224,
+      "grad_norm": 2.5804395407070873,
+      "learning_rate": 0.00019507732085839385,
+      "loss": 2.1289,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3824,
+      "grad_norm": 2.9630967036036835,
+      "learning_rate": 0.00019507196649270824,
+      "loss": 2.1301,
+      "step": 2390
+    },
+    {
+      "epoch": 0.38256,
+      "grad_norm": 2.380692449910074,
+      "learning_rate": 0.00019506660929021372,
+      "loss": 2.0025,
+      "step": 2391
+    },
+    {
+      "epoch": 0.38272,
+      "grad_norm": 2.292336047453249,
+      "learning_rate": 0.00019506124925107007,
+      "loss": 2.0948,
+      "step": 2392
+    },
+    {
+      "epoch": 0.38288,
+      "grad_norm": 2.696200261018932,
+      "learning_rate": 0.0001950558863754373,
+      "loss": 1.9313,
+      "step": 2393
+    },
+    {
+      "epoch": 0.38304,
+      "grad_norm": 2.5262174687988908,
+      "learning_rate": 0.00019505052066347539,
+      "loss": 1.9976,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3832,
+      "grad_norm": 2.930885171243685,
+      "learning_rate": 0.0001950451521153444,
+      "loss": 2.0293,
+      "step": 2395
+    },
+    {
+      "epoch": 0.38336,
+      "grad_norm": 2.2584886726494866,
+      "learning_rate": 0.00019503978073120464,
+      "loss": 2.031,
+      "step": 2396
+    },
+    {
+      "epoch": 0.38352,
+      "grad_norm": 2.551105226948089,
+      "learning_rate": 0.00019503440651121625,
+      "loss": 2.1288,
+      "step": 2397
+    },
+    {
+      "epoch": 0.38368,
+      "grad_norm": 2.999110095695281,
+      "learning_rate": 0.00019502902945553967,
+      "loss": 2.0449,
+      "step": 2398
+    },
+    {
+      "epoch": 0.38384,
+      "grad_norm": 2.52867213414762,
+      "learning_rate": 0.00019502364956433532,
+      "loss": 2.035,
+      "step": 2399
+    },
+    {
+      "epoch": 0.384,
+      "grad_norm": 2.6482266476812497,
+      "learning_rate": 0.00019501826683776373,
+      "loss": 2.0484,
+      "step": 2400
+    },
+    {
+      "epoch": 0.38416,
+      "grad_norm": 3.06171645494503,
+      "learning_rate": 0.0001950128812759855,
+      "loss": 2.0811,
+      "step": 2401
+    },
+    {
+      "epoch": 0.38432,
+      "grad_norm": 4.069421884144295,
+      "learning_rate": 0.0001950074928791614,
+      "loss": 2.1908,
+      "step": 2402
+    },
+    {
+      "epoch": 0.38448,
+      "grad_norm": 2.5028908901569515,
+      "learning_rate": 0.0001950021016474521,
+      "loss": 2.0373,
+      "step": 2403
+    },
+    {
+      "epoch": 0.38464,
+      "grad_norm": 4.145071096382648,
+      "learning_rate": 0.00019499670758101853,
+      "loss": 2.037,
+      "step": 2404
+    },
+    {
+      "epoch": 0.3848,
+      "grad_norm": 2.695789890095866,
+      "learning_rate": 0.00019499131068002163,
+      "loss": 2.0944,
+      "step": 2405
+    },
+    {
+      "epoch": 0.38496,
+      "grad_norm": 3.997262406358371,
+      "learning_rate": 0.00019498591094462242,
+      "loss": 1.9737,
+      "step": 2406
+    },
+    {
+      "epoch": 0.38512,
+      "grad_norm": 3.543968089350587,
+      "learning_rate": 0.00019498050837498206,
+      "loss": 1.9489,
+      "step": 2407
+    },
+    {
+      "epoch": 0.38528,
+      "grad_norm": 3.1178482927630378,
+      "learning_rate": 0.00019497510297126174,
+      "loss": 2.0644,
+      "step": 2408
+    },
+    {
+      "epoch": 0.38544,
+      "grad_norm": 2.678284178034168,
+      "learning_rate": 0.00019496969473362273,
+      "loss": 2.0532,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3856,
+      "grad_norm": 2.6658536557455057,
+      "learning_rate": 0.00019496428366222638,
+      "loss": 2.0104,
+      "step": 2410
+    },
+    {
+      "epoch": 0.38576,
+      "grad_norm": 2.750863383075438,
+      "learning_rate": 0.0001949588697572342,
+      "loss": 1.9993,
+      "step": 2411
+    },
+    {
+      "epoch": 0.38592,
+      "grad_norm": 2.4423698578572575,
+      "learning_rate": 0.00019495345301880773,
+      "loss": 2.0679,
+      "step": 2412
+    },
+    {
+      "epoch": 0.38608,
+      "grad_norm": 3.870393356903461,
+      "learning_rate": 0.0001949480334471086,
+      "loss": 2.0145,
+      "step": 2413
+    },
+    {
+      "epoch": 0.38624,
+      "grad_norm": 2.552682021313978,
+      "learning_rate": 0.0001949426110422985,
+      "loss": 1.89,
+      "step": 2414
+    },
+    {
+      "epoch": 0.3864,
+      "grad_norm": 3.2536748370415918,
+      "learning_rate": 0.00019493718580453922,
+      "loss": 1.9874,
+      "step": 2415
+    },
+    {
+      "epoch": 0.38656,
+      "grad_norm": 2.4426565265962283,
+      "learning_rate": 0.00019493175773399262,
+      "loss": 2.0905,
+      "step": 2416
+    },
+    {
+      "epoch": 0.38672,
+      "grad_norm": 2.724519697171787,
+      "learning_rate": 0.00019492632683082076,
+      "loss": 2.1623,
+      "step": 2417
+    },
+    {
+      "epoch": 0.38688,
+      "grad_norm": 3.2357240083467236,
+      "learning_rate": 0.00019492089309518563,
+      "loss": 2.1244,
+      "step": 2418
+    },
+    {
+      "epoch": 0.38704,
+      "grad_norm": 3.270190579415226,
+      "learning_rate": 0.0001949154565272493,
+      "loss": 2.0925,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3872,
+      "grad_norm": 3.021689407154576,
+      "learning_rate": 0.00019491001712717413,
+      "loss": 1.973,
+      "step": 2420
+    },
+    {
+      "epoch": 0.38736,
+      "grad_norm": 2.7283860975551617,
+      "learning_rate": 0.0001949045748951223,
+      "loss": 1.9992,
+      "step": 2421
+    },
+    {
+      "epoch": 0.38752,
+      "grad_norm": 2.5708288719108823,
+      "learning_rate": 0.00019489912983125624,
+      "loss": 2.0718,
+      "step": 2422
+    },
+    {
+      "epoch": 0.38768,
+      "grad_norm": 3.1569056973424154,
+      "learning_rate": 0.00019489368193573845,
+      "loss": 2.0032,
+      "step": 2423
+    },
+    {
+      "epoch": 0.38784,
+      "grad_norm": 2.34927769028785,
+      "learning_rate": 0.00019488823120873146,
+      "loss": 1.921,
+      "step": 2424
+    },
+    {
+      "epoch": 0.388,
+      "grad_norm": 3.0827529631292965,
+      "learning_rate": 0.00019488277765039792,
+      "loss": 1.9788,
+      "step": 2425
+    },
+    {
+      "epoch": 0.38816,
+      "grad_norm": 2.4264116671671405,
+      "learning_rate": 0.00019487732126090054,
+      "loss": 2.1017,
+      "step": 2426
+    },
+    {
+      "epoch": 0.38832,
+      "grad_norm": 3.5691985400176187,
+      "learning_rate": 0.00019487186204040217,
+      "loss": 2.1553,
+      "step": 2427
+    },
+    {
+      "epoch": 0.38848,
+      "grad_norm": 2.869549483792387,
+      "learning_rate": 0.00019486639998906566,
+      "loss": 2.142,
+      "step": 2428
+    },
+    {
+      "epoch": 0.38864,
+      "grad_norm": 2.9893612016474,
+      "learning_rate": 0.000194860935107054,
+      "loss": 2.0419,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3888,
+      "grad_norm": 2.593417159424906,
+      "learning_rate": 0.0001948554673945303,
+      "loss": 2.0293,
+      "step": 2430
+    },
+    {
+      "epoch": 0.38896,
+      "grad_norm": 2.3732457548193664,
+      "learning_rate": 0.00019484999685165765,
+      "loss": 2.0997,
+      "step": 2431
+    },
+    {
+      "epoch": 0.38912,
+      "grad_norm": 2.278320252052254,
+      "learning_rate": 0.0001948445234785993,
+      "loss": 1.9529,
+      "step": 2432
+    },
+    {
+      "epoch": 0.38928,
+      "grad_norm": 2.4558711076115918,
+      "learning_rate": 0.00019483904727551855,
+      "loss": 1.9988,
+      "step": 2433
+    },
+    {
+      "epoch": 0.38944,
+      "grad_norm": 2.3979827916352225,
+      "learning_rate": 0.00019483356824257886,
+      "loss": 2.1605,
+      "step": 2434
+    },
+    {
+      "epoch": 0.3896,
+      "grad_norm": 3.4461043460828873,
+      "learning_rate": 0.0001948280863799437,
+      "loss": 2.0859,
+      "step": 2435
+    },
+    {
+      "epoch": 0.38976,
+      "grad_norm": 2.4855317659499048,
+      "learning_rate": 0.00019482260168777657,
+      "loss": 1.9693,
+      "step": 2436
+    },
+    {
+      "epoch": 0.38992,
+      "grad_norm": 3.487333134539431,
+      "learning_rate": 0.0001948171141662412,
+      "loss": 2.0308,
+      "step": 2437
+    },
+    {
+      "epoch": 0.39008,
+      "grad_norm": 2.5685996013967225,
+      "learning_rate": 0.00019481162381550132,
+      "loss": 1.969,
+      "step": 2438
+    },
+    {
+      "epoch": 0.39024,
+      "grad_norm": 2.649423842740734,
+      "learning_rate": 0.00019480613063572072,
+      "loss": 1.966,
+      "step": 2439
+    },
+    {
+      "epoch": 0.3904,
+      "grad_norm": 2.319677898482677,
+      "learning_rate": 0.00019480063462706335,
+      "loss": 1.9951,
+      "step": 2440
+    },
+    {
+      "epoch": 0.39056,
+      "grad_norm": 2.8777703411038424,
+      "learning_rate": 0.00019479513578969315,
+      "loss": 2.0426,
+      "step": 2441
+    },
+    {
+      "epoch": 0.39072,
+      "grad_norm": 2.5947970481178952,
+      "learning_rate": 0.00019478963412377428,
+      "loss": 2.1778,
+      "step": 2442
+    },
+    {
+      "epoch": 0.39088,
+      "grad_norm": 2.891352824336549,
+      "learning_rate": 0.0001947841296294708,
+      "loss": 1.9809,
+      "step": 2443
+    },
+    {
+      "epoch": 0.39104,
+      "grad_norm": 2.6090345147387497,
+      "learning_rate": 0.00019477862230694705,
+      "loss": 2.0572,
+      "step": 2444
+    },
+    {
+      "epoch": 0.3912,
+      "grad_norm": 2.54828688993774,
+      "learning_rate": 0.00019477311215636726,
+      "loss": 2.0238,
+      "step": 2445
+    },
+    {
+      "epoch": 0.39136,
+      "grad_norm": 2.7651531219037992,
+      "learning_rate": 0.00019476759917789593,
+      "loss": 1.9603,
+      "step": 2446
+    },
+    {
+      "epoch": 0.39152,
+      "grad_norm": 2.1130345112995634,
+      "learning_rate": 0.00019476208337169756,
+      "loss": 2.0904,
+      "step": 2447
+    },
+    {
+      "epoch": 0.39168,
+      "grad_norm": 2.948129398626233,
+      "learning_rate": 0.00019475656473793668,
+      "loss": 2.0243,
+      "step": 2448
+    },
+    {
+      "epoch": 0.39184,
+      "grad_norm": 2.265338033582148,
+      "learning_rate": 0.00019475104327677797,
+      "loss": 1.9678,
+      "step": 2449
+    },
+    {
+      "epoch": 0.392,
+      "grad_norm": 3.355910478754272,
+      "learning_rate": 0.00019474551898838622,
+      "loss": 1.9353,
+      "step": 2450
+    },
+    {
+      "epoch": 0.39216,
+      "grad_norm": 2.3659878496634326,
+      "learning_rate": 0.00019473999187292622,
+      "loss": 2.0579,
+      "step": 2451
+    },
+    {
+      "epoch": 0.39232,
+      "grad_norm": 3.121907943323793,
+      "learning_rate": 0.00019473446193056293,
+      "loss": 2.0737,
+      "step": 2452
+    },
+    {
+      "epoch": 0.39248,
+      "grad_norm": 2.555741111103568,
+      "learning_rate": 0.00019472892916146134,
+      "loss": 2.0699,
+      "step": 2453
+    },
+    {
+      "epoch": 0.39264,
+      "grad_norm": 2.7543161865345764,
+      "learning_rate": 0.00019472339356578654,
+      "loss": 1.9716,
+      "step": 2454
+    },
+    {
+      "epoch": 0.3928,
+      "grad_norm": 2.5307238662859812,
+      "learning_rate": 0.00019471785514370366,
+      "loss": 1.9675,
+      "step": 2455
+    },
+    {
+      "epoch": 0.39296,
+      "grad_norm": 2.4848987474588435,
+      "learning_rate": 0.00019471231389537805,
+      "loss": 1.955,
+      "step": 2456
+    },
+    {
+      "epoch": 0.39312,
+      "grad_norm": 2.4889016024937596,
+      "learning_rate": 0.000194706769820975,
+      "loss": 2.0187,
+      "step": 2457
+    },
+    {
+      "epoch": 0.39328,
+      "grad_norm": 3.0281586697884864,
+      "learning_rate": 0.00019470122292065994,
+      "loss": 1.978,
+      "step": 2458
+    },
+    {
+      "epoch": 0.39344,
+      "grad_norm": 2.5327445748940076,
+      "learning_rate": 0.00019469567319459837,
+      "loss": 2.0653,
+      "step": 2459
+    },
+    {
+      "epoch": 0.3936,
+      "grad_norm": 2.6141069435583626,
+      "learning_rate": 0.00019469012064295588,
+      "loss": 2.089,
+      "step": 2460
+    },
+    {
+      "epoch": 0.39376,
+      "grad_norm": 2.7088324718492314,
+      "learning_rate": 0.0001946845652658982,
+      "loss": 2.1633,
+      "step": 2461
+    },
+    {
+      "epoch": 0.39392,
+      "grad_norm": 3.042216768487742,
+      "learning_rate": 0.00019467900706359107,
+      "loss": 2.0379,
+      "step": 2462
+    },
+    {
+      "epoch": 0.39408,
+      "grad_norm": 2.9889990701375617,
+      "learning_rate": 0.0001946734460362003,
+      "loss": 2.0069,
+      "step": 2463
+    },
+    {
+      "epoch": 0.39424,
+      "grad_norm": 3.19980346925694,
+      "learning_rate": 0.00019466788218389187,
+      "loss": 2.0104,
+      "step": 2464
+    },
+    {
+      "epoch": 0.3944,
+      "grad_norm": 2.35391529386432,
+      "learning_rate": 0.00019466231550683175,
+      "loss": 2.0729,
+      "step": 2465
+    },
+    {
+      "epoch": 0.39456,
+      "grad_norm": 2.902980823288791,
+      "learning_rate": 0.00019465674600518612,
+      "loss": 2.1155,
+      "step": 2466
+    },
+    {
+      "epoch": 0.39472,
+      "grad_norm": 2.8002891943547787,
+      "learning_rate": 0.00019465117367912108,
+      "loss": 2.0121,
+      "step": 2467
+    },
+    {
+      "epoch": 0.39488,
+      "grad_norm": 2.73476250026196,
+      "learning_rate": 0.00019464559852880298,
+      "loss": 2.0531,
+      "step": 2468
+    },
+    {
+      "epoch": 0.39504,
+      "grad_norm": 2.276782084918487,
+      "learning_rate": 0.00019464002055439813,
+      "loss": 2.0977,
+      "step": 2469
+    },
+    {
+      "epoch": 0.3952,
+      "grad_norm": 2.3717248319483333,
+      "learning_rate": 0.00019463443975607293,
+      "loss": 2.145,
+      "step": 2470
+    },
+    {
+      "epoch": 0.39536,
+      "grad_norm": 3.280650082859922,
+      "learning_rate": 0.000194628856133994,
+      "loss": 2.1252,
+      "step": 2471
+    },
+    {
+      "epoch": 0.39552,
+      "grad_norm": 2.4862543182892525,
+      "learning_rate": 0.00019462326968832785,
+      "loss": 2.0008,
+      "step": 2472
+    },
+    {
+      "epoch": 0.39568,
+      "grad_norm": 2.4584993363944325,
+      "learning_rate": 0.00019461768041924122,
+      "loss": 2.0923,
+      "step": 2473
+    },
+    {
+      "epoch": 0.39584,
+      "grad_norm": 2.273862327591804,
+      "learning_rate": 0.00019461208832690087,
+      "loss": 2.1009,
+      "step": 2474
+    },
+    {
+      "epoch": 0.396,
+      "grad_norm": 2.171394614735974,
+      "learning_rate": 0.00019460649341147368,
+      "loss": 2.068,
+      "step": 2475
+    },
+    {
+      "epoch": 0.39616,
+      "grad_norm": 2.7129045028911727,
+      "learning_rate": 0.00019460089567312658,
+      "loss": 2.1149,
+      "step": 2476
+    },
+    {
+      "epoch": 0.39632,
+      "grad_norm": 2.452335531787434,
+      "learning_rate": 0.0001945952951120266,
+      "loss": 2.0091,
+      "step": 2477
+    },
+    {
+      "epoch": 0.39648,
+      "grad_norm": 2.8391428089926456,
+      "learning_rate": 0.00019458969172834087,
+      "loss": 1.9526,
+      "step": 2478
+    },
+    {
+      "epoch": 0.39664,
+      "grad_norm": 2.971059758673504,
+      "learning_rate": 0.00019458408552223654,
+      "loss": 2.1607,
+      "step": 2479
+    },
+    {
+      "epoch": 0.3968,
+      "grad_norm": 2.53803694496759,
+      "learning_rate": 0.00019457847649388094,
+      "loss": 2.0062,
+      "step": 2480
+    },
+    {
+      "epoch": 0.39696,
+      "grad_norm": 2.5285732954761446,
+      "learning_rate": 0.00019457286464344143,
+      "loss": 2.0646,
+      "step": 2481
+    },
+    {
+      "epoch": 0.39712,
+      "grad_norm": 3.1524370101875197,
+      "learning_rate": 0.00019456724997108545,
+      "loss": 1.962,
+      "step": 2482
+    },
+    {
+      "epoch": 0.39728,
+      "grad_norm": 3.197694227509552,
+      "learning_rate": 0.0001945616324769805,
+      "loss": 1.9546,
+      "step": 2483
+    },
+    {
+      "epoch": 0.39744,
+      "grad_norm": 2.810771325821981,
+      "learning_rate": 0.00019455601216129424,
+      "loss": 2.053,
+      "step": 2484
+    },
+    {
+      "epoch": 0.3976,
+      "grad_norm": 2.5745475966281783,
+      "learning_rate": 0.00019455038902419438,
+      "loss": 1.9215,
+      "step": 2485
+    },
+    {
+      "epoch": 0.39776,
+      "grad_norm": 2.3602350302987127,
+      "learning_rate": 0.00019454476306584865,
+      "loss": 2.0488,
+      "step": 2486
+    },
+    {
+      "epoch": 0.39792,
+      "grad_norm": 2.08046699601508,
+      "learning_rate": 0.00019453913428642497,
+      "loss": 2.0305,
+      "step": 2487
+    },
+    {
+      "epoch": 0.39808,
+      "grad_norm": 3.1258804715950608,
+      "learning_rate": 0.00019453350268609128,
+      "loss": 1.936,
+      "step": 2488
+    },
+    {
+      "epoch": 0.39824,
+      "grad_norm": 2.269439830612028,
+      "learning_rate": 0.00019452786826501563,
+      "loss": 2.0518,
+      "step": 2489
+    },
+    {
+      "epoch": 0.3984,
+      "grad_norm": 2.479586323936769,
+      "learning_rate": 0.00019452223102336613,
+      "loss": 1.9507,
+      "step": 2490
+    },
+    {
+      "epoch": 0.39856,
+      "grad_norm": 2.9250318522092433,
+      "learning_rate": 0.000194516590961311,
+      "loss": 2.0405,
+      "step": 2491
+    },
+    {
+      "epoch": 0.39872,
+      "grad_norm": 2.2697741646936196,
+      "learning_rate": 0.0001945109480790185,
+      "loss": 1.9584,
+      "step": 2492
+    },
+    {
+      "epoch": 0.39888,
+      "grad_norm": 2.40446713443292,
+      "learning_rate": 0.00019450530237665703,
+      "loss": 1.97,
+      "step": 2493
+    },
+    {
+      "epoch": 0.39904,
+      "grad_norm": 2.938518945811268,
+      "learning_rate": 0.00019449965385439505,
+      "loss": 2.0902,
+      "step": 2494
+    },
+    {
+      "epoch": 0.3992,
+      "grad_norm": 2.5624391801661055,
+      "learning_rate": 0.0001944940025124011,
+      "loss": 1.9609,
+      "step": 2495
+    },
+    {
+      "epoch": 0.39936,
+      "grad_norm": 2.7421196222904505,
+      "learning_rate": 0.00019448834835084382,
+      "loss": 1.9339,
+      "step": 2496
+    },
+    {
+      "epoch": 0.39952,
+      "grad_norm": 2.978566766733678,
+      "learning_rate": 0.0001944826913698919,
+      "loss": 2.1905,
+      "step": 2497
+    },
+    {
+      "epoch": 0.39968,
+      "grad_norm": 2.6947716686595324,
+      "learning_rate": 0.0001944770315697141,
+      "loss": 2.1272,
+      "step": 2498
+    },
+    {
+      "epoch": 0.39984,
+      "grad_norm": 2.702494364625341,
+      "learning_rate": 0.0001944713689504794,
+      "loss": 2.1169,
+      "step": 2499
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.592658226379581,
+      "learning_rate": 0.00019446570351235672,
+      "loss": 2.0736,
+      "step": 2500
+    },
+    {
+      "epoch": 0.40016,
+      "grad_norm": 2.909377489043446,
+      "learning_rate": 0.00019446003525551507,
+      "loss": 1.8842,
+      "step": 2501
+    },
+    {
+      "epoch": 0.40032,
+      "grad_norm": 2.657468592216221,
+      "learning_rate": 0.0001944543641801236,
+      "loss": 2.044,
+      "step": 2502
+    },
+    {
+      "epoch": 0.40048,
+      "grad_norm": 2.435439079299335,
+      "learning_rate": 0.00019444869028635155,
+      "loss": 1.9825,
+      "step": 2503
+    },
+    {
+      "epoch": 0.40064,
+      "grad_norm": 2.6878374540316177,
+      "learning_rate": 0.00019444301357436823,
+      "loss": 1.834,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4008,
+      "grad_norm": 2.8423083115106174,
+      "learning_rate": 0.00019443733404434296,
+      "loss": 1.9849,
+      "step": 2505
+    },
+    {
+      "epoch": 0.40096,
+      "grad_norm": 4.046513287776709,
+      "learning_rate": 0.00019443165169644527,
+      "loss": 2.0981,
+      "step": 2506
+    },
+    {
+      "epoch": 0.40112,
+      "grad_norm": 2.81737740799602,
+      "learning_rate": 0.00019442596653084472,
+      "loss": 1.9897,
+      "step": 2507
+    },
+    {
+      "epoch": 0.40128,
+      "grad_norm": 4.905895472431032,
+      "learning_rate": 0.0001944202785477109,
+      "loss": 2.0526,
+      "step": 2508
+    },
+    {
+      "epoch": 0.40144,
+      "grad_norm": 3.155011012175926,
+      "learning_rate": 0.00019441458774721357,
+      "loss": 2.0625,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4016,
+      "grad_norm": 3.9314146483553145,
+      "learning_rate": 0.00019440889412952248,
+      "loss": 2.0431,
+      "step": 2510
+    },
+    {
+      "epoch": 0.40176,
+      "grad_norm": 2.2803038310459622,
+      "learning_rate": 0.0001944031976948076,
+      "loss": 1.8989,
+      "step": 2511
+    },
+    {
+      "epoch": 0.40192,
+      "grad_norm": 3.4455907546020783,
+      "learning_rate": 0.00019439749844323885,
+      "loss": 2.048,
+      "step": 2512
+    },
+    {
+      "epoch": 0.40208,
+      "grad_norm": 3.1256080759655105,
+      "learning_rate": 0.00019439179637498628,
+      "loss": 2.1185,
+      "step": 2513
+    },
+    {
+      "epoch": 0.40224,
+      "grad_norm": 2.9962842158143843,
+      "learning_rate": 0.00019438609149022005,
+      "loss": 2.0108,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4024,
+      "grad_norm": 2.5497187082678803,
+      "learning_rate": 0.0001943803837891104,
+      "loss": 2.0289,
+      "step": 2515
+    },
+    {
+      "epoch": 0.40256,
+      "grad_norm": 2.67734438062472,
+      "learning_rate": 0.00019437467327182765,
+      "loss": 2.0858,
+      "step": 2516
+    },
+    {
+      "epoch": 0.40272,
+      "grad_norm": 2.323127544044031,
+      "learning_rate": 0.00019436895993854211,
+      "loss": 1.9814,
+      "step": 2517
+    },
+    {
+      "epoch": 0.40288,
+      "grad_norm": 2.1914072530524447,
+      "learning_rate": 0.00019436324378942435,
+      "loss": 1.9989,
+      "step": 2518
+    },
+    {
+      "epoch": 0.40304,
+      "grad_norm": 3.020967642364364,
+      "learning_rate": 0.00019435752482464488,
+      "loss": 2.1609,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4032,
+      "grad_norm": 2.113425299077887,
+      "learning_rate": 0.00019435180304437438,
+      "loss": 1.9655,
+      "step": 2520
+    },
+    {
+      "epoch": 0.40336,
+      "grad_norm": 3.6501427664402004,
+      "learning_rate": 0.00019434607844878357,
+      "loss": 1.9983,
+      "step": 2521
+    },
+    {
+      "epoch": 0.40352,
+      "grad_norm": 2.576445253556566,
+      "learning_rate": 0.0001943403510380432,
+      "loss": 2.0289,
+      "step": 2522
+    },
+    {
+      "epoch": 0.40368,
+      "grad_norm": 2.3503097517705376,
+      "learning_rate": 0.00019433462081232426,
+      "loss": 1.9747,
+      "step": 2523
+    },
+    {
+      "epoch": 0.40384,
+      "grad_norm": 2.6146486928262145,
+      "learning_rate": 0.00019432888777179774,
+      "loss": 1.981,
+      "step": 2524
+    },
+    {
+      "epoch": 0.404,
+      "grad_norm": 2.471805577722846,
+      "learning_rate": 0.0001943231519166346,
+      "loss": 2.0098,
+      "step": 2525
+    },
+    {
+      "epoch": 0.40416,
+      "grad_norm": 2.351528071629898,
+      "learning_rate": 0.00019431741324700606,
+      "loss": 2.0426,
+      "step": 2526
+    },
+    {
+      "epoch": 0.40432,
+      "grad_norm": 3.0274581793531823,
+      "learning_rate": 0.00019431167176308338,
+      "loss": 2.0121,
+      "step": 2527
+    },
+    {
+      "epoch": 0.40448,
+      "grad_norm": 2.624912816337265,
+      "learning_rate": 0.0001943059274650378,
+      "loss": 2.03,
+      "step": 2528
+    },
+    {
+      "epoch": 0.40464,
+      "grad_norm": 2.90102147026496,
+      "learning_rate": 0.00019430018035304078,
+      "loss": 2.0314,
+      "step": 2529
+    },
+    {
+      "epoch": 0.4048,
+      "grad_norm": 2.2961603848057393,
+      "learning_rate": 0.00019429443042726378,
+      "loss": 2.0235,
+      "step": 2530
+    },
+    {
+      "epoch": 0.40496,
+      "grad_norm": 2.4554085320686405,
+      "learning_rate": 0.0001942886776878784,
+      "loss": 1.9889,
+      "step": 2531
+    },
+    {
+      "epoch": 0.40512,
+      "grad_norm": 2.814035583221214,
+      "learning_rate": 0.00019428292213505627,
+      "loss": 2.0707,
+      "step": 2532
+    },
+    {
+      "epoch": 0.40528,
+      "grad_norm": 3.1162980146721027,
+      "learning_rate": 0.00019427716376896909,
+      "loss": 2.0017,
+      "step": 2533
+    },
+    {
+      "epoch": 0.40544,
+      "grad_norm": 2.4463769368941986,
+      "learning_rate": 0.00019427140258978875,
+      "loss": 1.9723,
+      "step": 2534
+    },
+    {
+      "epoch": 0.4056,
+      "grad_norm": 2.737522639739841,
+      "learning_rate": 0.0001942656385976871,
+      "loss": 2.0856,
+      "step": 2535
+    },
+    {
+      "epoch": 0.40576,
+      "grad_norm": 3.254498546122873,
+      "learning_rate": 0.00019425987179283617,
+      "loss": 2.01,
+      "step": 2536
+    },
+    {
+      "epoch": 0.40592,
+      "grad_norm": 2.5319632977637108,
+      "learning_rate": 0.00019425410217540801,
+      "loss": 2.0335,
+      "step": 2537
+    },
+    {
+      "epoch": 0.40608,
+      "grad_norm": 4.354292311388595,
+      "learning_rate": 0.0001942483297455748,
+      "loss": 1.9946,
+      "step": 2538
+    },
+    {
+      "epoch": 0.40624,
+      "grad_norm": 2.239930356457109,
+      "learning_rate": 0.00019424255450350877,
+      "loss": 1.971,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4064,
+      "grad_norm": 2.798166071730291,
+      "learning_rate": 0.0001942367764493822,
+      "loss": 2.1202,
+      "step": 2540
+    },
+    {
+      "epoch": 0.40656,
+      "grad_norm": 2.757016296902186,
+      "learning_rate": 0.00019423099558336758,
+      "loss": 1.994,
+      "step": 2541
+    },
+    {
+      "epoch": 0.40672,
+      "grad_norm": 3.226657015562643,
+      "learning_rate": 0.00019422521190563737,
+      "loss": 2.0521,
+      "step": 2542
+    },
+    {
+      "epoch": 0.40688,
+      "grad_norm": 3.1226593534547624,
+      "learning_rate": 0.0001942194254163641,
+      "loss": 2.1749,
+      "step": 2543
+    },
+    {
+      "epoch": 0.40704,
+      "grad_norm": 3.2176528049889668,
+      "learning_rate": 0.00019421363611572047,
+      "loss": 2.051,
+      "step": 2544
+    },
+    {
+      "epoch": 0.4072,
+      "grad_norm": 2.9698336309354896,
+      "learning_rate": 0.00019420784400387924,
+      "loss": 1.9534,
+      "step": 2545
+    },
+    {
+      "epoch": 0.40736,
+      "grad_norm": 3.4299519110827417,
+      "learning_rate": 0.00019420204908101322,
+      "loss": 2.0016,
+      "step": 2546
+    },
+    {
+      "epoch": 0.40752,
+      "grad_norm": 2.502565397284444,
+      "learning_rate": 0.0001941962513472953,
+      "loss": 2.0996,
+      "step": 2547
+    },
+    {
+      "epoch": 0.40768,
+      "grad_norm": 2.3830186966936386,
+      "learning_rate": 0.00019419045080289852,
+      "loss": 2.0821,
+      "step": 2548
+    },
+    {
+      "epoch": 0.40784,
+      "grad_norm": 2.478796395340954,
+      "learning_rate": 0.00019418464744799593,
+      "loss": 1.933,
+      "step": 2549
+    },
+    {
+      "epoch": 0.408,
+      "grad_norm": 2.6016485682762585,
+      "learning_rate": 0.0001941788412827607,
+      "loss": 2.0224,
+      "step": 2550
+    },
+    {
+      "epoch": 0.40816,
+      "grad_norm": 2.3847854332839336,
+      "learning_rate": 0.00019417303230736608,
+      "loss": 2.0287,
+      "step": 2551
+    },
+    {
+      "epoch": 0.40832,
+      "grad_norm": 2.52621535635978,
+      "learning_rate": 0.0001941672205219854,
+      "loss": 2.0142,
+      "step": 2552
+    },
+    {
+      "epoch": 0.40848,
+      "grad_norm": 2.2433963421650125,
+      "learning_rate": 0.00019416140592679208,
+      "loss": 2.009,
+      "step": 2553
+    },
+    {
+      "epoch": 0.40864,
+      "grad_norm": 2.837832346688312,
+      "learning_rate": 0.0001941555885219596,
+      "loss": 2.1465,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4088,
+      "grad_norm": 2.3045343840302372,
+      "learning_rate": 0.0001941497683076616,
+      "loss": 2.0526,
+      "step": 2555
+    },
+    {
+      "epoch": 0.40896,
+      "grad_norm": 2.159889604788752,
+      "learning_rate": 0.00019414394528407164,
+      "loss": 1.9864,
+      "step": 2556
+    },
+    {
+      "epoch": 0.40912,
+      "grad_norm": 2.164351915005906,
+      "learning_rate": 0.00019413811945136357,
+      "loss": 2.0356,
+      "step": 2557
+    },
+    {
+      "epoch": 0.40928,
+      "grad_norm": 2.945403415304578,
+      "learning_rate": 0.0001941322908097112,
+      "loss": 2.1372,
+      "step": 2558
+    },
+    {
+      "epoch": 0.40944,
+      "grad_norm": 3.106987062468803,
+      "learning_rate": 0.0001941264593592884,
+      "loss": 2.1422,
+      "step": 2559
+    },
+    {
+      "epoch": 0.4096,
+      "grad_norm": 2.0303050956911517,
+      "learning_rate": 0.00019412062510026928,
+      "loss": 1.9918,
+      "step": 2560
+    },
+    {
+      "epoch": 0.40976,
+      "grad_norm": 2.596702097763419,
+      "learning_rate": 0.0001941147880328278,
+      "loss": 2.0465,
+      "step": 2561
+    },
+    {
+      "epoch": 0.40992,
+      "grad_norm": 2.169094356052711,
+      "learning_rate": 0.0001941089481571382,
+      "loss": 1.9706,
+      "step": 2562
+    },
+    {
+      "epoch": 0.41008,
+      "grad_norm": 2.544416993178893,
+      "learning_rate": 0.0001941031054733747,
+      "loss": 2.0114,
+      "step": 2563
+    },
+    {
+      "epoch": 0.41024,
+      "grad_norm": 3.0635049030189627,
+      "learning_rate": 0.00019409725998171168,
+      "loss": 2.0218,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4104,
+      "grad_norm": 3.5112169159985873,
+      "learning_rate": 0.0001940914116823235,
+      "loss": 2.1314,
+      "step": 2565
+    },
+    {
+      "epoch": 0.41056,
+      "grad_norm": 3.9297105659522225,
+      "learning_rate": 0.00019408556057538476,
+      "loss": 1.9678,
+      "step": 2566
+    },
+    {
+      "epoch": 0.41072,
+      "grad_norm": 4.212598205296388,
+      "learning_rate": 0.00019407970666106996,
+      "loss": 2.0387,
+      "step": 2567
+    },
+    {
+      "epoch": 0.41088,
+      "grad_norm": 2.8476708074365713,
+      "learning_rate": 0.0001940738499395538,
+      "loss": 1.9572,
+      "step": 2568
+    },
+    {
+      "epoch": 0.41104,
+      "grad_norm": 2.833907580234802,
+      "learning_rate": 0.000194067990411011,
+      "loss": 1.9064,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4112,
+      "grad_norm": 3.309754048135546,
+      "learning_rate": 0.0001940621280756165,
+      "loss": 1.9621,
+      "step": 2570
+    },
+    {
+      "epoch": 0.41136,
+      "grad_norm": 3.441593892145916,
+      "learning_rate": 0.00019405626293354512,
+      "loss": 2.0332,
+      "step": 2571
+    },
+    {
+      "epoch": 0.41152,
+      "grad_norm": 2.799212547497694,
+      "learning_rate": 0.00019405039498497195,
+      "loss": 1.9265,
+      "step": 2572
+    },
+    {
+      "epoch": 0.41168,
+      "grad_norm": 2.9940286596443966,
+      "learning_rate": 0.000194044524230072,
+      "loss": 2.0426,
+      "step": 2573
+    },
+    {
+      "epoch": 0.41184,
+      "grad_norm": 2.9232614861968806,
+      "learning_rate": 0.0001940386506690205,
+      "loss": 2.0246,
+      "step": 2574
+    },
+    {
+      "epoch": 0.412,
+      "grad_norm": 2.2576297586724303,
+      "learning_rate": 0.00019403277430199268,
+      "loss": 2.012,
+      "step": 2575
+    },
+    {
+      "epoch": 0.41216,
+      "grad_norm": 2.5698520394835582,
+      "learning_rate": 0.0001940268951291639,
+      "loss": 1.953,
+      "step": 2576
+    },
+    {
+      "epoch": 0.41232,
+      "grad_norm": 2.5067069540376634,
+      "learning_rate": 0.00019402101315070958,
+      "loss": 1.9927,
+      "step": 2577
+    },
+    {
+      "epoch": 0.41248,
+      "grad_norm": 2.3373532396403744,
+      "learning_rate": 0.00019401512836680524,
+      "loss": 1.9972,
+      "step": 2578
+    },
+    {
+      "epoch": 0.41264,
+      "grad_norm": 2.5026851053900847,
+      "learning_rate": 0.00019400924077762647,
+      "loss": 2.1028,
+      "step": 2579
+    },
+    {
+      "epoch": 0.4128,
+      "grad_norm": 2.837858858346277,
+      "learning_rate": 0.00019400335038334893,
+      "loss": 2.028,
+      "step": 2580
+    },
+    {
+      "epoch": 0.41296,
+      "grad_norm": 2.6179023551716183,
+      "learning_rate": 0.0001939974571841484,
+      "loss": 1.968,
+      "step": 2581
+    },
+    {
+      "epoch": 0.41312,
+      "grad_norm": 2.420071641167019,
+      "learning_rate": 0.00019399156118020072,
+      "loss": 1.8962,
+      "step": 2582
+    },
+    {
+      "epoch": 0.41328,
+      "grad_norm": 2.1663541746918034,
+      "learning_rate": 0.00019398566237168182,
+      "loss": 2.0126,
+      "step": 2583
+    },
+    {
+      "epoch": 0.41344,
+      "grad_norm": 2.5875867842724687,
+      "learning_rate": 0.00019397976075876766,
+      "loss": 1.9884,
+      "step": 2584
+    },
+    {
+      "epoch": 0.4136,
+      "grad_norm": 2.8176044502133157,
+      "learning_rate": 0.00019397385634163444,
+      "loss": 2.0026,
+      "step": 2585
+    },
+    {
+      "epoch": 0.41376,
+      "grad_norm": 2.948518646240711,
+      "learning_rate": 0.00019396794912045825,
+      "loss": 1.9275,
+      "step": 2586
+    },
+    {
+      "epoch": 0.41392,
+      "grad_norm": 2.5394874445256836,
+      "learning_rate": 0.00019396203909541538,
+      "loss": 2.0334,
+      "step": 2587
+    },
+    {
+      "epoch": 0.41408,
+      "grad_norm": 2.8632324955672757,
+      "learning_rate": 0.0001939561262666822,
+      "loss": 1.9429,
+      "step": 2588
+    },
+    {
+      "epoch": 0.41424,
+      "grad_norm": 2.512011973287286,
+      "learning_rate": 0.0001939502106344351,
+      "loss": 1.9486,
+      "step": 2589
+    },
+    {
+      "epoch": 0.4144,
+      "grad_norm": 2.597510148879353,
+      "learning_rate": 0.00019394429219885063,
+      "loss": 1.9817,
+      "step": 2590
+    },
+    {
+      "epoch": 0.41456,
+      "grad_norm": 2.4872358670765418,
+      "learning_rate": 0.00019393837096010535,
+      "loss": 1.9867,
+      "step": 2591
+    },
+    {
+      "epoch": 0.41472,
+      "grad_norm": 3.7200455478897014,
+      "learning_rate": 0.000193932446918376,
+      "loss": 1.9606,
+      "step": 2592
+    },
+    {
+      "epoch": 0.41488,
+      "grad_norm": 2.1765447426084945,
+      "learning_rate": 0.0001939265200738393,
+      "loss": 1.9678,
+      "step": 2593
+    },
+    {
+      "epoch": 0.41504,
+      "grad_norm": 3.5287781315295383,
+      "learning_rate": 0.00019392059042667209,
+      "loss": 1.9892,
+      "step": 2594
+    },
+    {
+      "epoch": 0.4152,
+      "grad_norm": 2.866071193714827,
+      "learning_rate": 0.00019391465797705132,
+      "loss": 1.9444,
+      "step": 2595
+    },
+    {
+      "epoch": 0.41536,
+      "grad_norm": 3.018056114696415,
+      "learning_rate": 0.00019390872272515402,
+      "loss": 1.9148,
+      "step": 2596
+    },
+    {
+      "epoch": 0.41552,
+      "grad_norm": 3.6972366845803846,
+      "learning_rate": 0.00019390278467115723,
+      "loss": 1.9138,
+      "step": 2597
+    },
+    {
+      "epoch": 0.41568,
+      "grad_norm": 3.669283886195012,
+      "learning_rate": 0.0001938968438152382,
+      "loss": 2.0656,
+      "step": 2598
+    },
+    {
+      "epoch": 0.41584,
+      "grad_norm": 2.2816798719046063,
+      "learning_rate": 0.00019389090015757421,
+      "loss": 1.9668,
+      "step": 2599
+    },
+    {
+      "epoch": 0.416,
+      "grad_norm": 2.3018852716198657,
+      "learning_rate": 0.0001938849536983425,
+      "loss": 1.8805,
+      "step": 2600
+    },
+    {
+      "epoch": 0.41616,
+      "grad_norm": 2.414673085147852,
+      "learning_rate": 0.00019387900443772064,
+      "loss": 1.9903,
+      "step": 2601
+    },
+    {
+      "epoch": 0.41632,
+      "grad_norm": 2.8885756809723313,
+      "learning_rate": 0.00019387305237588608,
+      "loss": 2.034,
+      "step": 2602
+    },
+    {
+      "epoch": 0.41648,
+      "grad_norm": 2.3992988970367746,
+      "learning_rate": 0.0001938670975130164,
+      "loss": 2.0166,
+      "step": 2603
+    },
+    {
+      "epoch": 0.41664,
+      "grad_norm": 2.492111765262342,
+      "learning_rate": 0.0001938611398492893,
+      "loss": 2.0762,
+      "step": 2604
+    },
+    {
+      "epoch": 0.4168,
+      "grad_norm": 2.9100761962407087,
+      "learning_rate": 0.0001938551793848826,
+      "loss": 1.9872,
+      "step": 2605
+    },
+    {
+      "epoch": 0.41696,
+      "grad_norm": 2.092826771699031,
+      "learning_rate": 0.00019384921611997406,
+      "loss": 2.0456,
+      "step": 2606
+    },
+    {
+      "epoch": 0.41712,
+      "grad_norm": 2.4518095146248426,
+      "learning_rate": 0.00019384325005474172,
+      "loss": 2.0278,
+      "step": 2607
+    },
+    {
+      "epoch": 0.41728,
+      "grad_norm": 2.4083251074327325,
+      "learning_rate": 0.0001938372811893635,
+      "loss": 2.0575,
+      "step": 2608
+    },
+    {
+      "epoch": 0.41744,
+      "grad_norm": 2.6005586093989796,
+      "learning_rate": 0.00019383130952401752,
+      "loss": 1.9927,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4176,
+      "grad_norm": 2.3245845152000073,
+      "learning_rate": 0.00019382533505888204,
+      "loss": 1.9842,
+      "step": 2610
+    },
+    {
+      "epoch": 0.41776,
+      "grad_norm": 2.774436251000822,
+      "learning_rate": 0.00019381935779413527,
+      "loss": 1.9782,
+      "step": 2611
+    },
+    {
+      "epoch": 0.41792,
+      "grad_norm": 2.4686667926828307,
+      "learning_rate": 0.00019381337772995556,
+      "loss": 2.098,
+      "step": 2612
+    },
+    {
+      "epoch": 0.41808,
+      "grad_norm": 2.066919483366623,
+      "learning_rate": 0.00019380739486652137,
+      "loss": 1.9702,
+      "step": 2613
+    },
+    {
+      "epoch": 0.41824,
+      "grad_norm": 2.847202111999368,
+      "learning_rate": 0.0001938014092040112,
+      "loss": 1.9201,
+      "step": 2614
+    },
+    {
+      "epoch": 0.4184,
+      "grad_norm": 3.173729820663912,
+      "learning_rate": 0.00019379542074260365,
+      "loss": 2.0519,
+      "step": 2615
+    },
+    {
+      "epoch": 0.41856,
+      "grad_norm": 2.960267855853576,
+      "learning_rate": 0.00019378942948247745,
+      "loss": 2.022,
+      "step": 2616
+    },
+    {
+      "epoch": 0.41872,
+      "grad_norm": 2.747361936349983,
+      "learning_rate": 0.00019378343542381132,
+      "loss": 2.0532,
+      "step": 2617
+    },
+    {
+      "epoch": 0.41888,
+      "grad_norm": 2.3607117735392324,
+      "learning_rate": 0.00019377743856678414,
+      "loss": 1.9165,
+      "step": 2618
+    },
+    {
+      "epoch": 0.41904,
+      "grad_norm": 2.426986736504945,
+      "learning_rate": 0.00019377143891157484,
+      "loss": 1.9666,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4192,
+      "grad_norm": 2.44774976780791,
+      "learning_rate": 0.00019376543645836244,
+      "loss": 1.8967,
+      "step": 2620
+    },
+    {
+      "epoch": 0.41936,
+      "grad_norm": 2.1256083169321776,
+      "learning_rate": 0.00019375943120732606,
+      "loss": 2.0155,
+      "step": 2621
+    },
+    {
+      "epoch": 0.41952,
+      "grad_norm": 2.1682579776035125,
+      "learning_rate": 0.00019375342315864486,
+      "loss": 2.0261,
+      "step": 2622
+    },
+    {
+      "epoch": 0.41968,
+      "grad_norm": 2.876133807620392,
+      "learning_rate": 0.00019374741231249814,
+      "loss": 1.9522,
+      "step": 2623
+    },
+    {
+      "epoch": 0.41984,
+      "grad_norm": 2.8487687566518014,
+      "learning_rate": 0.00019374139866906522,
+      "loss": 2.0349,
+      "step": 2624
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 2.72362881409183,
+      "learning_rate": 0.00019373538222852558,
+      "loss": 2.0524,
+      "step": 2625
+    },
+    {
+      "epoch": 0.42016,
+      "grad_norm": 2.30540164572955,
+      "learning_rate": 0.0001937293629910587,
+      "loss": 1.9787,
+      "step": 2626
+    },
+    {
+      "epoch": 0.42032,
+      "grad_norm": 2.5721203504615255,
+      "learning_rate": 0.00019372334095684422,
+      "loss": 1.9404,
+      "step": 2627
+    },
+    {
+      "epoch": 0.42048,
+      "grad_norm": 2.276810977941027,
+      "learning_rate": 0.0001937173161260618,
+      "loss": 1.9599,
+      "step": 2628
+    },
+    {
+      "epoch": 0.42064,
+      "grad_norm": 2.291258011564743,
+      "learning_rate": 0.00019371128849889123,
+      "loss": 1.9623,
+      "step": 2629
+    },
+    {
+      "epoch": 0.4208,
+      "grad_norm": 2.0228547785032585,
+      "learning_rate": 0.00019370525807551242,
+      "loss": 2.0331,
+      "step": 2630
+    },
+    {
+      "epoch": 0.42096,
+      "grad_norm": 2.872397413020353,
+      "learning_rate": 0.0001936992248561052,
+      "loss": 2.036,
+      "step": 2631
+    },
+    {
+      "epoch": 0.42112,
+      "grad_norm": 2.293727989836843,
+      "learning_rate": 0.00019369318884084963,
+      "loss": 1.9988,
+      "step": 2632
+    },
+    {
+      "epoch": 0.42128,
+      "grad_norm": 2.7216949559985557,
+      "learning_rate": 0.00019368715002992585,
+      "loss": 1.9702,
+      "step": 2633
+    },
+    {
+      "epoch": 0.42144,
+      "grad_norm": 2.225312158894147,
+      "learning_rate": 0.00019368110842351403,
+      "loss": 1.9985,
+      "step": 2634
+    },
+    {
+      "epoch": 0.4216,
+      "grad_norm": 2.703080529142754,
+      "learning_rate": 0.00019367506402179446,
+      "loss": 2.1484,
+      "step": 2635
+    },
+    {
+      "epoch": 0.42176,
+      "grad_norm": 2.972884122629039,
+      "learning_rate": 0.00019366901682494746,
+      "loss": 1.9659,
+      "step": 2636
+    },
+    {
+      "epoch": 0.42192,
+      "grad_norm": 2.8791174086837805,
+      "learning_rate": 0.00019366296683315348,
+      "loss": 2.0811,
+      "step": 2637
+    },
+    {
+      "epoch": 0.42208,
+      "grad_norm": 2.768966116685456,
+      "learning_rate": 0.00019365691404659306,
+      "loss": 1.9992,
+      "step": 2638
+    },
+    {
+      "epoch": 0.42224,
+      "grad_norm": 2.8818992390381246,
+      "learning_rate": 0.0001936508584654468,
+      "loss": 1.9971,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4224,
+      "grad_norm": 3.054420741789267,
+      "learning_rate": 0.00019364480008989538,
+      "loss": 1.9935,
+      "step": 2640
+    },
+    {
+      "epoch": 0.42256,
+      "grad_norm": 2.452308130246276,
+      "learning_rate": 0.00019363873892011958,
+      "loss": 1.967,
+      "step": 2641
+    },
+    {
+      "epoch": 0.42272,
+      "grad_norm": 2.980424575111919,
+      "learning_rate": 0.00019363267495630027,
+      "loss": 1.9917,
+      "step": 2642
+    },
+    {
+      "epoch": 0.42288,
+      "grad_norm": 2.119150254026084,
+      "learning_rate": 0.0001936266081986184,
+      "loss": 2.0683,
+      "step": 2643
+    },
+    {
+      "epoch": 0.42304,
+      "grad_norm": 2.4648236164535184,
+      "learning_rate": 0.00019362053864725493,
+      "loss": 1.9769,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4232,
+      "grad_norm": 2.4703344882874574,
+      "learning_rate": 0.000193614466302391,
+      "loss": 1.9562,
+      "step": 2645
+    },
+    {
+      "epoch": 0.42336,
+      "grad_norm": 2.7047616692283825,
+      "learning_rate": 0.0001936083911642078,
+      "loss": 2.0371,
+      "step": 2646
+    },
+    {
+      "epoch": 0.42352,
+      "grad_norm": 2.5724402255746015,
+      "learning_rate": 0.00019360231323288664,
+      "loss": 1.9996,
+      "step": 2647
+    },
+    {
+      "epoch": 0.42368,
+      "grad_norm": 3.2037651083158964,
+      "learning_rate": 0.00019359623250860884,
+      "loss": 2.019,
+      "step": 2648
+    },
+    {
+      "epoch": 0.42384,
+      "grad_norm": 2.4083767683400676,
+      "learning_rate": 0.00019359014899155582,
+      "loss": 1.9857,
+      "step": 2649
+    },
+    {
+      "epoch": 0.424,
+      "grad_norm": 3.6378771151645664,
+      "learning_rate": 0.00019358406268190918,
+      "loss": 1.9825,
+      "step": 2650
+    },
+    {
+      "epoch": 0.42416,
+      "grad_norm": 2.7113979747731656,
+      "learning_rate": 0.00019357797357985045,
+      "loss": 2.031,
+      "step": 2651
+    },
+    {
+      "epoch": 0.42432,
+      "grad_norm": 3.2440867234726656,
+      "learning_rate": 0.00019357188168556136,
+      "loss": 1.9995,
+      "step": 2652
+    },
+    {
+      "epoch": 0.42448,
+      "grad_norm": 2.8274013713018586,
+      "learning_rate": 0.00019356578699922364,
+      "loss": 2.1137,
+      "step": 2653
+    },
+    {
+      "epoch": 0.42464,
+      "grad_norm": 2.6847028537397404,
+      "learning_rate": 0.00019355968952101916,
+      "loss": 2.0252,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4248,
+      "grad_norm": 3.6841321421021864,
+      "learning_rate": 0.00019355358925112994,
+      "loss": 2.0491,
+      "step": 2655
+    },
+    {
+      "epoch": 0.42496,
+      "grad_norm": 2.802391111037597,
+      "learning_rate": 0.0001935474861897379,
+      "loss": 2.0349,
+      "step": 2656
+    },
+    {
+      "epoch": 0.42512,
+      "grad_norm": 3.1722069994897155,
+      "learning_rate": 0.00019354138033702518,
+      "loss": 1.8701,
+      "step": 2657
+    },
+    {
+      "epoch": 0.42528,
+      "grad_norm": 2.3358559025197856,
+      "learning_rate": 0.00019353527169317398,
+      "loss": 2.0089,
+      "step": 2658
+    },
+    {
+      "epoch": 0.42544,
+      "grad_norm": 2.6518831991968166,
+      "learning_rate": 0.00019352916025836656,
+      "loss": 2.0034,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4256,
+      "grad_norm": 2.059723610209733,
+      "learning_rate": 0.0001935230460327853,
+      "loss": 1.979,
+      "step": 2660
+    },
+    {
+      "epoch": 0.42576,
+      "grad_norm": 3.7855531465628074,
+      "learning_rate": 0.00019351692901661263,
+      "loss": 1.9757,
+      "step": 2661
+    },
+    {
+      "epoch": 0.42592,
+      "grad_norm": 2.167046951368944,
+      "learning_rate": 0.00019351080921003105,
+      "loss": 1.9667,
+      "step": 2662
+    },
+    {
+      "epoch": 0.42608,
+      "grad_norm": 3.374702660169513,
+      "learning_rate": 0.0001935046866132232,
+      "loss": 2.0087,
+      "step": 2663
+    },
+    {
+      "epoch": 0.42624,
+      "grad_norm": 2.4739773941769734,
+      "learning_rate": 0.00019349856122637173,
+      "loss": 1.9013,
+      "step": 2664
+    },
+    {
+      "epoch": 0.4264,
+      "grad_norm": 2.6785868703075626,
+      "learning_rate": 0.00019349243304965944,
+      "loss": 1.96,
+      "step": 2665
+    },
+    {
+      "epoch": 0.42656,
+      "grad_norm": 2.5940165887871096,
+      "learning_rate": 0.00019348630208326922,
+      "loss": 2.0254,
+      "step": 2666
+    },
+    {
+      "epoch": 0.42672,
+      "grad_norm": 2.4367395621924017,
+      "learning_rate": 0.0001934801683273839,
+      "loss": 2.0152,
+      "step": 2667
+    },
+    {
+      "epoch": 0.42688,
+      "grad_norm": 2.5072537336243603,
+      "learning_rate": 0.00019347403178218665,
+      "loss": 1.8641,
+      "step": 2668
+    },
+    {
+      "epoch": 0.42704,
+      "grad_norm": 2.739190821287464,
+      "learning_rate": 0.00019346789244786044,
+      "loss": 1.9947,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4272,
+      "grad_norm": 3.873773932836421,
+      "learning_rate": 0.00019346175032458854,
+      "loss": 2.0555,
+      "step": 2670
+    },
+    {
+      "epoch": 0.42736,
+      "grad_norm": 2.80743405905864,
+      "learning_rate": 0.00019345560541255416,
+      "loss": 1.9087,
+      "step": 2671
+    },
+    {
+      "epoch": 0.42752,
+      "grad_norm": 2.6711784501504696,
+      "learning_rate": 0.00019344945771194075,
+      "loss": 1.9807,
+      "step": 2672
+    },
+    {
+      "epoch": 0.42768,
+      "grad_norm": 2.36128511020638,
+      "learning_rate": 0.00019344330722293168,
+      "loss": 1.9165,
+      "step": 2673
+    },
+    {
+      "epoch": 0.42784,
+      "grad_norm": 2.838588861295859,
+      "learning_rate": 0.00019343715394571045,
+      "loss": 1.9922,
+      "step": 2674
+    },
+    {
+      "epoch": 0.428,
+      "grad_norm": 2.2068785073375823,
+      "learning_rate": 0.0001934309978804607,
+      "loss": 2.0328,
+      "step": 2675
+    },
+    {
+      "epoch": 0.42816,
+      "grad_norm": 2.499267795093583,
+      "learning_rate": 0.00019342483902736615,
+      "loss": 1.9752,
+      "step": 2676
+    },
+    {
+      "epoch": 0.42832,
+      "grad_norm": 2.489284972673199,
+      "learning_rate": 0.00019341867738661053,
+      "loss": 2.0176,
+      "step": 2677
+    },
+    {
+      "epoch": 0.42848,
+      "grad_norm": 3.5564923115094786,
+      "learning_rate": 0.0001934125129583777,
+      "loss": 2.0493,
+      "step": 2678
+    },
+    {
+      "epoch": 0.42864,
+      "grad_norm": 3.2787819185613376,
+      "learning_rate": 0.00019340634574285156,
+      "loss": 2.0907,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4288,
+      "grad_norm": 2.6854988936126047,
+      "learning_rate": 0.00019340017574021623,
+      "loss": 1.8896,
+      "step": 2680
+    },
+    {
+      "epoch": 0.42896,
+      "grad_norm": 2.617433201453412,
+      "learning_rate": 0.00019339400295065572,
+      "loss": 1.9874,
+      "step": 2681
+    },
+    {
+      "epoch": 0.42912,
+      "grad_norm": 2.6746110552815687,
+      "learning_rate": 0.00019338782737435425,
+      "loss": 2.079,
+      "step": 2682
+    },
+    {
+      "epoch": 0.42928,
+      "grad_norm": 2.4096886523184238,
+      "learning_rate": 0.00019338164901149614,
+      "loss": 2.0179,
+      "step": 2683
+    },
+    {
+      "epoch": 0.42944,
+      "grad_norm": 2.651022742990473,
+      "learning_rate": 0.00019337546786226564,
+      "loss": 2.1289,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4296,
+      "grad_norm": 2.5771985223287937,
+      "learning_rate": 0.00019336928392684727,
+      "loss": 1.8988,
+      "step": 2685
+    },
+    {
+      "epoch": 0.42976,
+      "grad_norm": 2.8771137333353747,
+      "learning_rate": 0.00019336309720542548,
+      "loss": 2.1394,
+      "step": 2686
+    },
+    {
+      "epoch": 0.42992,
+      "grad_norm": 2.5688844342164927,
+      "learning_rate": 0.00019335690769818498,
+      "loss": 2.1213,
+      "step": 2687
+    },
+    {
+      "epoch": 0.43008,
+      "grad_norm": 2.16132623257917,
+      "learning_rate": 0.00019335071540531033,
+      "loss": 1.964,
+      "step": 2688
+    },
+    {
+      "epoch": 0.43024,
+      "grad_norm": 2.9664525259678602,
+      "learning_rate": 0.0001933445203269864,
+      "loss": 1.893,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4304,
+      "grad_norm": 2.1225277049229443,
+      "learning_rate": 0.00019333832246339798,
+      "loss": 1.9598,
+      "step": 2690
+    },
+    {
+      "epoch": 0.43056,
+      "grad_norm": 2.7836325604223178,
+      "learning_rate": 0.00019333212181473,
+      "loss": 2.0058,
+      "step": 2691
+    },
+    {
+      "epoch": 0.43072,
+      "grad_norm": 2.329583816464554,
+      "learning_rate": 0.00019332591838116755,
+      "loss": 2.0524,
+      "step": 2692
+    },
+    {
+      "epoch": 0.43088,
+      "grad_norm": 3.5682610642665042,
+      "learning_rate": 0.0001933197121628957,
+      "loss": 2.0372,
+      "step": 2693
+    },
+    {
+      "epoch": 0.43104,
+      "grad_norm": 2.3068136302294393,
+      "learning_rate": 0.00019331350316009957,
+      "loss": 1.9264,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4312,
+      "grad_norm": 2.5998532308167634,
+      "learning_rate": 0.0001933072913729645,
+      "loss": 1.9921,
+      "step": 2695
+    },
+    {
+      "epoch": 0.43136,
+      "grad_norm": 2.801227353856751,
+      "learning_rate": 0.00019330107680167583,
+      "loss": 1.8193,
+      "step": 2696
+    },
+    {
+      "epoch": 0.43152,
+      "grad_norm": 3.5438227623937912,
+      "learning_rate": 0.00019329485944641898,
+      "loss": 1.9324,
+      "step": 2697
+    },
+    {
+      "epoch": 0.43168,
+      "grad_norm": 2.761062211157971,
+      "learning_rate": 0.00019328863930737943,
+      "loss": 1.9928,
+      "step": 2698
+    },
+    {
+      "epoch": 0.43184,
+      "grad_norm": 3.355578291175513,
+      "learning_rate": 0.00019328241638474284,
+      "loss": 2.0328,
+      "step": 2699
+    },
+    {
+      "epoch": 0.432,
+      "grad_norm": 2.6881704164795184,
+      "learning_rate": 0.0001932761906786949,
+      "loss": 2.1249,
+      "step": 2700
+    },
+    {
+      "epoch": 0.43216,
+      "grad_norm": 2.270800398694423,
+      "learning_rate": 0.0001932699621894213,
+      "loss": 1.9408,
+      "step": 2701
+    },
+    {
+      "epoch": 0.43232,
+      "grad_norm": 2.9097443396722067,
+      "learning_rate": 0.00019326373091710798,
+      "loss": 1.8518,
+      "step": 2702
+    },
+    {
+      "epoch": 0.43248,
+      "grad_norm": 2.3898800198944707,
+      "learning_rate": 0.00019325749686194082,
+      "loss": 2.0556,
+      "step": 2703
+    },
+    {
+      "epoch": 0.43264,
+      "grad_norm": 3.315119444063561,
+      "learning_rate": 0.00019325126002410584,
+      "loss": 1.9363,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4328,
+      "grad_norm": 2.703654583906907,
+      "learning_rate": 0.00019324502040378913,
+      "loss": 1.908,
+      "step": 2705
+    },
+    {
+      "epoch": 0.43296,
+      "grad_norm": 3.0299293888720493,
+      "learning_rate": 0.0001932387780011769,
+      "loss": 2.0721,
+      "step": 2706
+    },
+    {
+      "epoch": 0.43312,
+      "grad_norm": 2.726381597669828,
+      "learning_rate": 0.0001932325328164554,
+      "loss": 1.9546,
+      "step": 2707
+    },
+    {
+      "epoch": 0.43328,
+      "grad_norm": 2.6127961545667526,
+      "learning_rate": 0.00019322628484981093,
+      "loss": 2.022,
+      "step": 2708
+    },
+    {
+      "epoch": 0.43344,
+      "grad_norm": 2.438409656693187,
+      "learning_rate": 0.00019322003410143,
+      "loss": 1.8677,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4336,
+      "grad_norm": 2.601008466767552,
+      "learning_rate": 0.00019321378057149911,
+      "loss": 1.9131,
+      "step": 2710
+    },
+    {
+      "epoch": 0.43376,
+      "grad_norm": 3.529022516888536,
+      "learning_rate": 0.0001932075242602048,
+      "loss": 2.1562,
+      "step": 2711
+    },
+    {
+      "epoch": 0.43392,
+      "grad_norm": 3.517320259831619,
+      "learning_rate": 0.0001932012651677338,
+      "loss": 1.8761,
+      "step": 2712
+    },
+    {
+      "epoch": 0.43408,
+      "grad_norm": 3.0611865060151926,
+      "learning_rate": 0.00019319500329427288,
+      "loss": 1.9519,
+      "step": 2713
+    },
+    {
+      "epoch": 0.43424,
+      "grad_norm": 2.2353920857656924,
+      "learning_rate": 0.0001931887386400088,
+      "loss": 1.9935,
+      "step": 2714
+    },
+    {
+      "epoch": 0.4344,
+      "grad_norm": 3.1964459382664865,
+      "learning_rate": 0.0001931824712051286,
+      "loss": 2.0358,
+      "step": 2715
+    },
+    {
+      "epoch": 0.43456,
+      "grad_norm": 2.648124393105986,
+      "learning_rate": 0.00019317620098981922,
+      "loss": 1.8898,
+      "step": 2716
+    },
+    {
+      "epoch": 0.43472,
+      "grad_norm": 3.150044875308562,
+      "learning_rate": 0.00019316992799426778,
+      "loss": 2.0214,
+      "step": 2717
+    },
+    {
+      "epoch": 0.43488,
+      "grad_norm": 2.7126356935471008,
+      "learning_rate": 0.00019316365221866145,
+      "loss": 2.0004,
+      "step": 2718
+    },
+    {
+      "epoch": 0.43504,
+      "grad_norm": 2.7018669858116655,
+      "learning_rate": 0.0001931573736631875,
+      "loss": 2.1121,
+      "step": 2719
+    },
+    {
+      "epoch": 0.4352,
+      "grad_norm": 3.1238687411986565,
+      "learning_rate": 0.00019315109232803326,
+      "loss": 2.0992,
+      "step": 2720
+    },
+    {
+      "epoch": 0.43536,
+      "grad_norm": 2.576705233040193,
+      "learning_rate": 0.00019314480821338613,
+      "loss": 2.1208,
+      "step": 2721
+    },
+    {
+      "epoch": 0.43552,
+      "grad_norm": 2.7501016855489735,
+      "learning_rate": 0.00019313852131943367,
+      "loss": 1.9491,
+      "step": 2722
+    },
+    {
+      "epoch": 0.43568,
+      "grad_norm": 2.5041964921996698,
+      "learning_rate": 0.00019313223164636346,
+      "loss": 2.0254,
+      "step": 2723
+    },
+    {
+      "epoch": 0.43584,
+      "grad_norm": 3.313951444472303,
+      "learning_rate": 0.00019312593919436317,
+      "loss": 1.9793,
+      "step": 2724
+    },
+    {
+      "epoch": 0.436,
+      "grad_norm": 2.3581805975138774,
+      "learning_rate": 0.00019311964396362058,
+      "loss": 1.8779,
+      "step": 2725
+    },
+    {
+      "epoch": 0.43616,
+      "grad_norm": 2.441554670996049,
+      "learning_rate": 0.00019311334595432345,
+      "loss": 2.0508,
+      "step": 2726
+    },
+    {
+      "epoch": 0.43632,
+      "grad_norm": 2.58351987510524,
+      "learning_rate": 0.00019310704516665975,
+      "loss": 1.9257,
+      "step": 2727
+    },
+    {
+      "epoch": 0.43648,
+      "grad_norm": 2.692326774384115,
+      "learning_rate": 0.00019310074160081754,
+      "loss": 2.1231,
+      "step": 2728
+    },
+    {
+      "epoch": 0.43664,
+      "grad_norm": 2.4669411974791497,
+      "learning_rate": 0.00019309443525698485,
+      "loss": 1.8553,
+      "step": 2729
+    },
+    {
+      "epoch": 0.4368,
+      "grad_norm": 2.633948489800821,
+      "learning_rate": 0.00019308812613534986,
+      "loss": 2.0537,
+      "step": 2730
+    },
+    {
+      "epoch": 0.43696,
+      "grad_norm": 2.745484859086153,
+      "learning_rate": 0.00019308181423610083,
+      "loss": 1.9199,
+      "step": 2731
+    },
+    {
+      "epoch": 0.43712,
+      "grad_norm": 2.903814726697649,
+      "learning_rate": 0.00019307549955942607,
+      "loss": 1.9278,
+      "step": 2732
+    },
+    {
+      "epoch": 0.43728,
+      "grad_norm": 2.58136536255751,
+      "learning_rate": 0.00019306918210551408,
+      "loss": 1.9352,
+      "step": 2733
+    },
+    {
+      "epoch": 0.43744,
+      "grad_norm": 2.1790877378642493,
+      "learning_rate": 0.00019306286187455328,
+      "loss": 1.9612,
+      "step": 2734
+    },
+    {
+      "epoch": 0.4376,
+      "grad_norm": 3.0158629293157015,
+      "learning_rate": 0.00019305653886673229,
+      "loss": 2.0306,
+      "step": 2735
+    },
+    {
+      "epoch": 0.43776,
+      "grad_norm": 2.5443796479549183,
+      "learning_rate": 0.0001930502130822398,
+      "loss": 1.9029,
+      "step": 2736
+    },
+    {
+      "epoch": 0.43792,
+      "grad_norm": 2.692328655763128,
+      "learning_rate": 0.0001930438845212645,
+      "loss": 2.0799,
+      "step": 2737
+    },
+    {
+      "epoch": 0.43808,
+      "grad_norm": 3.299659644020777,
+      "learning_rate": 0.0001930375531839953,
+      "loss": 2.0035,
+      "step": 2738
+    },
+    {
+      "epoch": 0.43824,
+      "grad_norm": 3.055267361799677,
+      "learning_rate": 0.00019303121907062108,
+      "loss": 2.0739,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4384,
+      "grad_norm": 2.5809418672434585,
+      "learning_rate": 0.00019302488218133083,
+      "loss": 2.0011,
+      "step": 2740
+    },
+    {
+      "epoch": 0.43856,
+      "grad_norm": 2.9074359120625655,
+      "learning_rate": 0.00019301854251631363,
+      "loss": 2.0165,
+      "step": 2741
+    },
+    {
+      "epoch": 0.43872,
+      "grad_norm": 2.7569106014939853,
+      "learning_rate": 0.0001930122000757587,
+      "loss": 2.0557,
+      "step": 2742
+    },
+    {
+      "epoch": 0.43888,
+      "grad_norm": 3.114448708925145,
+      "learning_rate": 0.00019300585485985524,
+      "loss": 1.919,
+      "step": 2743
+    },
+    {
+      "epoch": 0.43904,
+      "grad_norm": 2.625089765466122,
+      "learning_rate": 0.0001929995068687926,
+      "loss": 1.9656,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4392,
+      "grad_norm": 3.18154149890084,
+      "learning_rate": 0.00019299315610276017,
+      "loss": 2.002,
+      "step": 2745
+    },
+    {
+      "epoch": 0.43936,
+      "grad_norm": 2.5958559126934646,
+      "learning_rate": 0.0001929868025619475,
+      "loss": 1.9294,
+      "step": 2746
+    },
+    {
+      "epoch": 0.43952,
+      "grad_norm": 2.082898674849693,
+      "learning_rate": 0.0001929804462465441,
+      "loss": 1.9843,
+      "step": 2747
+    },
+    {
+      "epoch": 0.43968,
+      "grad_norm": 2.3829290808286197,
+      "learning_rate": 0.00019297408715673967,
+      "loss": 1.976,
+      "step": 2748
+    },
+    {
+      "epoch": 0.43984,
+      "grad_norm": 2.6729253431685343,
+      "learning_rate": 0.000192967725292724,
+      "loss": 1.9823,
+      "step": 2749
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 3.8539740273293583,
+      "learning_rate": 0.00019296136065468687,
+      "loss": 2.0024,
+      "step": 2750
+    },
+    {
+      "epoch": 0.44016,
+      "grad_norm": 2.8582700683684004,
+      "learning_rate": 0.00019295499324281816,
+      "loss": 1.8935,
+      "step": 2751
+    },
+    {
+      "epoch": 0.44032,
+      "grad_norm": 3.5405835492755253,
+      "learning_rate": 0.00019294862305730792,
+      "loss": 2.1543,
+      "step": 2752
+    },
+    {
+      "epoch": 0.44048,
+      "grad_norm": 2.9019869754104137,
+      "learning_rate": 0.0001929422500983462,
+      "loss": 1.9827,
+      "step": 2753
+    },
+    {
+      "epoch": 0.44064,
+      "grad_norm": 2.5909493183631764,
+      "learning_rate": 0.0001929358743661232,
+      "loss": 1.9876,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4408,
+      "grad_norm": 4.135420264006849,
+      "learning_rate": 0.0001929294958608291,
+      "loss": 2.0136,
+      "step": 2755
+    },
+    {
+      "epoch": 0.44096,
+      "grad_norm": 2.232814407292551,
+      "learning_rate": 0.00019292311458265427,
+      "loss": 1.9333,
+      "step": 2756
+    },
+    {
+      "epoch": 0.44112,
+      "grad_norm": 3.931646565088408,
+      "learning_rate": 0.00019291673053178907,
+      "loss": 2.0051,
+      "step": 2757
+    },
+    {
+      "epoch": 0.44128,
+      "grad_norm": 2.8740724792081167,
+      "learning_rate": 0.00019291034370842405,
+      "loss": 2.0584,
+      "step": 2758
+    },
+    {
+      "epoch": 0.44144,
+      "grad_norm": 4.438159375114478,
+      "learning_rate": 0.00019290395411274976,
+      "loss": 1.9909,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4416,
+      "grad_norm": 2.589369750199308,
+      "learning_rate": 0.00019289756174495687,
+      "loss": 1.9013,
+      "step": 2760
+    },
+    {
+      "epoch": 0.44176,
+      "grad_norm": 3.4099990520473957,
+      "learning_rate": 0.00019289116660523608,
+      "loss": 2.0186,
+      "step": 2761
+    },
+    {
+      "epoch": 0.44192,
+      "grad_norm": 5.121242326720547,
+      "learning_rate": 0.00019288476869377825,
+      "loss": 2.0303,
+      "step": 2762
+    },
+    {
+      "epoch": 0.44208,
+      "grad_norm": 2.63867358551157,
+      "learning_rate": 0.00019287836801077426,
+      "loss": 1.9815,
+      "step": 2763
+    },
+    {
+      "epoch": 0.44224,
+      "grad_norm": 4.129141857506268,
+      "learning_rate": 0.00019287196455641513,
+      "loss": 1.9145,
+      "step": 2764
+    },
+    {
+      "epoch": 0.4424,
+      "grad_norm": 2.4959031014371256,
+      "learning_rate": 0.00019286555833089187,
+      "loss": 2.0867,
+      "step": 2765
+    },
+    {
+      "epoch": 0.44256,
+      "grad_norm": 3.1852249819605434,
+      "learning_rate": 0.00019285914933439568,
+      "loss": 1.9672,
+      "step": 2766
+    },
+    {
+      "epoch": 0.44272,
+      "grad_norm": 2.738683474626768,
+      "learning_rate": 0.0001928527375671178,
+      "loss": 2.116,
+      "step": 2767
+    },
+    {
+      "epoch": 0.44288,
+      "grad_norm": 2.6336899922362274,
+      "learning_rate": 0.00019284632302924952,
+      "loss": 2.0165,
+      "step": 2768
+    },
+    {
+      "epoch": 0.44304,
+      "grad_norm": 3.6613833528069626,
+      "learning_rate": 0.00019283990572098225,
+      "loss": 1.9445,
+      "step": 2769
+    },
+    {
+      "epoch": 0.4432,
+      "grad_norm": 2.6982199485119827,
+      "learning_rate": 0.0001928334856425075,
+      "loss": 1.9354,
+      "step": 2770
+    },
+    {
+      "epoch": 0.44336,
+      "grad_norm": 3.5542500528043064,
+      "learning_rate": 0.00019282706279401677,
+      "loss": 2.0061,
+      "step": 2771
+    },
+    {
+      "epoch": 0.44352,
+      "grad_norm": 2.582486879170646,
+      "learning_rate": 0.00019282063717570175,
+      "loss": 1.9205,
+      "step": 2772
+    },
+    {
+      "epoch": 0.44368,
+      "grad_norm": 3.0872034114251634,
+      "learning_rate": 0.0001928142087877542,
+      "loss": 1.9802,
+      "step": 2773
+    },
+    {
+      "epoch": 0.44384,
+      "grad_norm": 2.6603342752397,
+      "learning_rate": 0.00019280777763036592,
+      "loss": 1.9316,
+      "step": 2774
+    },
+    {
+      "epoch": 0.444,
+      "grad_norm": 3.218080078131902,
+      "learning_rate": 0.00019280134370372874,
+      "loss": 2.0577,
+      "step": 2775
+    },
+    {
+      "epoch": 0.44416,
+      "grad_norm": 3.1019951432067816,
+      "learning_rate": 0.00019279490700803474,
+      "loss": 1.8919,
+      "step": 2776
+    },
+    {
+      "epoch": 0.44432,
+      "grad_norm": 2.433784459194874,
+      "learning_rate": 0.0001927884675434759,
+      "loss": 1.975,
+      "step": 2777
+    },
+    {
+      "epoch": 0.44448,
+      "grad_norm": 3.1262769624584945,
+      "learning_rate": 0.0001927820253102444,
+      "loss": 2.08,
+      "step": 2778
+    },
+    {
+      "epoch": 0.44464,
+      "grad_norm": 2.7298777042931532,
+      "learning_rate": 0.00019277558030853244,
+      "loss": 2.0272,
+      "step": 2779
+    },
+    {
+      "epoch": 0.4448,
+      "grad_norm": 2.99947179204662,
+      "learning_rate": 0.0001927691325385324,
+      "loss": 1.9458,
+      "step": 2780
+    },
+    {
+      "epoch": 0.44496,
+      "grad_norm": 2.9062107519791014,
+      "learning_rate": 0.0001927626820004366,
+      "loss": 2.0258,
+      "step": 2781
+    },
+    {
+      "epoch": 0.44512,
+      "grad_norm": 2.7371568538988846,
+      "learning_rate": 0.00019275622869443753,
+      "loss": 1.9442,
+      "step": 2782
+    },
+    {
+      "epoch": 0.44528,
+      "grad_norm": 3.0593750234347015,
+      "learning_rate": 0.00019274977262072777,
+      "loss": 1.8984,
+      "step": 2783
+    },
+    {
+      "epoch": 0.44544,
+      "grad_norm": 2.1217671631008153,
+      "learning_rate": 0.00019274331377949994,
+      "loss": 1.9187,
+      "step": 2784
+    },
+    {
+      "epoch": 0.4456,
+      "grad_norm": 2.8701491690974694,
+      "learning_rate": 0.00019273685217094676,
+      "loss": 2.0664,
+      "step": 2785
+    },
+    {
+      "epoch": 0.44576,
+      "grad_norm": 2.138604077792513,
+      "learning_rate": 0.00019273038779526105,
+      "loss": 2.0216,
+      "step": 2786
+    },
+    {
+      "epoch": 0.44592,
+      "grad_norm": 2.4721361004026834,
+      "learning_rate": 0.0001927239206526357,
+      "loss": 1.9367,
+      "step": 2787
+    },
+    {
+      "epoch": 0.44608,
+      "grad_norm": 2.519443656841295,
+      "learning_rate": 0.00019271745074326367,
+      "loss": 1.9023,
+      "step": 2788
+    },
+    {
+      "epoch": 0.44624,
+      "grad_norm": 2.8809921841396804,
+      "learning_rate": 0.000192710978067338,
+      "loss": 2.0102,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4464,
+      "grad_norm": 2.746178497647614,
+      "learning_rate": 0.00019270450262505185,
+      "loss": 2.0078,
+      "step": 2790
+    },
+    {
+      "epoch": 0.44656,
+      "grad_norm": 2.2986655666502305,
+      "learning_rate": 0.0001926980244165984,
+      "loss": 1.8878,
+      "step": 2791
+    },
+    {
+      "epoch": 0.44672,
+      "grad_norm": 2.487769834482932,
+      "learning_rate": 0.000192691543442171,
+      "loss": 1.9203,
+      "step": 2792
+    },
+    {
+      "epoch": 0.44688,
+      "grad_norm": 3.4128358836831953,
+      "learning_rate": 0.00019268505970196302,
+      "loss": 1.9716,
+      "step": 2793
+    },
+    {
+      "epoch": 0.44704,
+      "grad_norm": 2.2728700680759752,
+      "learning_rate": 0.0001926785731961679,
+      "loss": 2.0322,
+      "step": 2794
+    },
+    {
+      "epoch": 0.4472,
+      "grad_norm": 2.412861226933938,
+      "learning_rate": 0.0001926720839249792,
+      "loss": 2.0088,
+      "step": 2795
+    },
+    {
+      "epoch": 0.44736,
+      "grad_norm": 2.131745812812329,
+      "learning_rate": 0.00019266559188859053,
+      "loss": 1.9406,
+      "step": 2796
+    },
+    {
+      "epoch": 0.44752,
+      "grad_norm": 2.775007910031862,
+      "learning_rate": 0.00019265909708719568,
+      "loss": 2.0752,
+      "step": 2797
+    },
+    {
+      "epoch": 0.44768,
+      "grad_norm": 2.4413390402841095,
+      "learning_rate": 0.00019265259952098838,
+      "loss": 1.9504,
+      "step": 2798
+    },
+    {
+      "epoch": 0.44784,
+      "grad_norm": 2.5869662022457387,
+      "learning_rate": 0.0001926460991901625,
+      "loss": 1.9799,
+      "step": 2799
+    },
+    {
+      "epoch": 0.448,
+      "grad_norm": 2.9960091921384384,
+      "learning_rate": 0.00019263959609491202,
+      "loss": 2.0179,
+      "step": 2800
+    },
+    {
+      "epoch": 0.44816,
+      "grad_norm": 3.044794891614827,
+      "learning_rate": 0.000192633090235431,
+      "loss": 1.9306,
+      "step": 2801
+    },
+    {
+      "epoch": 0.44832,
+      "grad_norm": 2.2239879325162177,
+      "learning_rate": 0.00019262658161191355,
+      "loss": 2.0041,
+      "step": 2802
+    },
+    {
+      "epoch": 0.44848,
+      "grad_norm": 2.4707472240864736,
+      "learning_rate": 0.00019262007022455383,
+      "loss": 1.9943,
+      "step": 2803
+    },
+    {
+      "epoch": 0.44864,
+      "grad_norm": 2.2203540183046133,
+      "learning_rate": 0.0001926135560735462,
+      "loss": 1.9651,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4488,
+      "grad_norm": 2.470290670792892,
+      "learning_rate": 0.000192607039159085,
+      "loss": 1.93,
+      "step": 2805
+    },
+    {
+      "epoch": 0.44896,
+      "grad_norm": 2.4944130613184425,
+      "learning_rate": 0.00019260051948136474,
+      "loss": 1.9567,
+      "step": 2806
+    },
+    {
+      "epoch": 0.44912,
+      "grad_norm": 2.063900961395211,
+      "learning_rate": 0.0001925939970405799,
+      "loss": 1.9126,
+      "step": 2807
+    },
+    {
+      "epoch": 0.44928,
+      "grad_norm": 2.3718650171463027,
+      "learning_rate": 0.00019258747183692506,
+      "loss": 1.9674,
+      "step": 2808
+    },
+    {
+      "epoch": 0.44944,
+      "grad_norm": 2.3852355372113014,
+      "learning_rate": 0.00019258094387059505,
+      "loss": 2.029,
+      "step": 2809
+    },
+    {
+      "epoch": 0.4496,
+      "grad_norm": 2.9073392016941813,
+      "learning_rate": 0.00019257441314178453,
+      "loss": 1.9567,
+      "step": 2810
+    },
+    {
+      "epoch": 0.44976,
+      "grad_norm": 2.752707160370044,
+      "learning_rate": 0.0001925678796506884,
+      "loss": 1.9664,
+      "step": 2811
+    },
+    {
+      "epoch": 0.44992,
+      "grad_norm": 3.706218343540193,
+      "learning_rate": 0.00019256134339750168,
+      "loss": 1.949,
+      "step": 2812
+    },
+    {
+      "epoch": 0.45008,
+      "grad_norm": 2.442381114041105,
+      "learning_rate": 0.00019255480438241934,
+      "loss": 1.9801,
+      "step": 2813
+    },
+    {
+      "epoch": 0.45024,
+      "grad_norm": 2.477732984002641,
+      "learning_rate": 0.00019254826260563647,
+      "loss": 1.9895,
+      "step": 2814
+    },
+    {
+      "epoch": 0.4504,
+      "grad_norm": 2.613756382797513,
+      "learning_rate": 0.0001925417180673483,
+      "loss": 2.0232,
+      "step": 2815
+    },
+    {
+      "epoch": 0.45056,
+      "grad_norm": 2.6180170119887567,
+      "learning_rate": 0.00019253517076775015,
+      "loss": 2.0186,
+      "step": 2816
+    },
+    {
+      "epoch": 0.45072,
+      "grad_norm": 2.5900308728538075,
+      "learning_rate": 0.00019252862070703733,
+      "loss": 2.0042,
+      "step": 2817
+    },
+    {
+      "epoch": 0.45088,
+      "grad_norm": 2.6332815918996015,
+      "learning_rate": 0.00019252206788540526,
+      "loss": 1.9316,
+      "step": 2818
+    },
+    {
+      "epoch": 0.45104,
+      "grad_norm": 2.2983003928230006,
+      "learning_rate": 0.00019251551230304955,
+      "loss": 2.0724,
+      "step": 2819
+    },
+    {
+      "epoch": 0.4512,
+      "grad_norm": 2.70376560040202,
+      "learning_rate": 0.00019250895396016572,
+      "loss": 2.0175,
+      "step": 2820
+    },
+    {
+      "epoch": 0.45136,
+      "grad_norm": 2.637033490220021,
+      "learning_rate": 0.00019250239285694952,
+      "loss": 2.026,
+      "step": 2821
+    },
+    {
+      "epoch": 0.45152,
+      "grad_norm": 2.9190531450320076,
+      "learning_rate": 0.00019249582899359672,
+      "loss": 1.9063,
+      "step": 2822
+    },
+    {
+      "epoch": 0.45168,
+      "grad_norm": 2.2723677612094786,
+      "learning_rate": 0.00019248926237030317,
+      "loss": 1.9652,
+      "step": 2823
+    },
+    {
+      "epoch": 0.45184,
+      "grad_norm": 3.541581661804517,
+      "learning_rate": 0.00019248269298726477,
+      "loss": 2.0448,
+      "step": 2824
+    },
+    {
+      "epoch": 0.452,
+      "grad_norm": 2.358607573200597,
+      "learning_rate": 0.0001924761208446776,
+      "loss": 1.927,
+      "step": 2825
+    },
+    {
+      "epoch": 0.45216,
+      "grad_norm": 3.1804656103212245,
+      "learning_rate": 0.00019246954594273773,
+      "loss": 1.96,
+      "step": 2826
+    },
+    {
+      "epoch": 0.45232,
+      "grad_norm": 2.8220393262849948,
+      "learning_rate": 0.00019246296828164135,
+      "loss": 1.8941,
+      "step": 2827
+    },
+    {
+      "epoch": 0.45248,
+      "grad_norm": 3.0475127011116103,
+      "learning_rate": 0.00019245638786158473,
+      "loss": 1.9741,
+      "step": 2828
+    },
+    {
+      "epoch": 0.45264,
+      "grad_norm": 2.8151286528266812,
+      "learning_rate": 0.00019244980468276422,
+      "loss": 1.9075,
+      "step": 2829
+    },
+    {
+      "epoch": 0.4528,
+      "grad_norm": 2.597762278561114,
+      "learning_rate": 0.00019244321874537624,
+      "loss": 1.9641,
+      "step": 2830
+    },
+    {
+      "epoch": 0.45296,
+      "grad_norm": 2.6727659885034165,
+      "learning_rate": 0.00019243663004961734,
+      "loss": 2.0447,
+      "step": 2831
+    },
+    {
+      "epoch": 0.45312,
+      "grad_norm": 2.871654476513311,
+      "learning_rate": 0.00019243003859568405,
+      "loss": 1.9507,
+      "step": 2832
+    },
+    {
+      "epoch": 0.45328,
+      "grad_norm": 2.5560669638111273,
+      "learning_rate": 0.00019242344438377313,
+      "loss": 1.9661,
+      "step": 2833
+    },
+    {
+      "epoch": 0.45344,
+      "grad_norm": 2.5000223951575293,
+      "learning_rate": 0.0001924168474140813,
+      "loss": 1.9732,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4536,
+      "grad_norm": 2.6230672150172634,
+      "learning_rate": 0.0001924102476868054,
+      "loss": 1.9794,
+      "step": 2835
+    },
+    {
+      "epoch": 0.45376,
+      "grad_norm": 3.355390737784659,
+      "learning_rate": 0.00019240364520214237,
+      "loss": 1.9057,
+      "step": 2836
+    },
+    {
+      "epoch": 0.45392,
+      "grad_norm": 3.147054468623704,
+      "learning_rate": 0.00019239703996028925,
+      "loss": 2.0805,
+      "step": 2837
+    },
+    {
+      "epoch": 0.45408,
+      "grad_norm": 3.5091630699743095,
+      "learning_rate": 0.00019239043196144303,
+      "loss": 2.0141,
+      "step": 2838
+    },
+    {
+      "epoch": 0.45424,
+      "grad_norm": 2.867656735183716,
+      "learning_rate": 0.000192383821205801,
+      "loss": 1.8877,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4544,
+      "grad_norm": 3.527080140786561,
+      "learning_rate": 0.00019237720769356033,
+      "loss": 2.0527,
+      "step": 2840
+    },
+    {
+      "epoch": 0.45456,
+      "grad_norm": 2.6178910093659553,
+      "learning_rate": 0.0001923705914249184,
+      "loss": 2.0909,
+      "step": 2841
+    },
+    {
+      "epoch": 0.45472,
+      "grad_norm": 4.055548491968067,
+      "learning_rate": 0.00019236397240007263,
+      "loss": 2.123,
+      "step": 2842
+    },
+    {
+      "epoch": 0.45488,
+      "grad_norm": 2.875668103336539,
+      "learning_rate": 0.0001923573506192205,
+      "loss": 1.9987,
+      "step": 2843
+    },
+    {
+      "epoch": 0.45504,
+      "grad_norm": 2.1935652952282267,
+      "learning_rate": 0.00019235072608255957,
+      "loss": 1.9368,
+      "step": 2844
+    },
+    {
+      "epoch": 0.4552,
+      "grad_norm": 2.7172117613123987,
+      "learning_rate": 0.00019234409879028758,
+      "loss": 1.9271,
+      "step": 2845
+    },
+    {
+      "epoch": 0.45536,
+      "grad_norm": 2.7472618559956308,
+      "learning_rate": 0.00019233746874260227,
+      "loss": 1.8622,
+      "step": 2846
+    },
+    {
+      "epoch": 0.45552,
+      "grad_norm": 2.3836762447566358,
+      "learning_rate": 0.0001923308359397014,
+      "loss": 2.0002,
+      "step": 2847
+    },
+    {
+      "epoch": 0.45568,
+      "grad_norm": 2.715492951348223,
+      "learning_rate": 0.00019232420038178293,
+      "loss": 1.9435,
+      "step": 2848
+    },
+    {
+      "epoch": 0.45584,
+      "grad_norm": 2.3842261321172042,
+      "learning_rate": 0.00019231756206904488,
+      "loss": 1.8751,
+      "step": 2849
+    },
+    {
+      "epoch": 0.456,
+      "grad_norm": 2.4832617040792964,
+      "learning_rate": 0.00019231092100168527,
+      "loss": 1.958,
+      "step": 2850
+    },
+    {
+      "epoch": 0.45616,
+      "grad_norm": 3.021921531057869,
+      "learning_rate": 0.00019230427717990232,
+      "loss": 1.9788,
+      "step": 2851
+    },
+    {
+      "epoch": 0.45632,
+      "grad_norm": 2.1757465002413516,
+      "learning_rate": 0.0001922976306038942,
+      "loss": 2.0245,
+      "step": 2852
+    },
+    {
+      "epoch": 0.45648,
+      "grad_norm": 1.9737053089084522,
+      "learning_rate": 0.0001922909812738593,
+      "loss": 1.9876,
+      "step": 2853
+    },
+    {
+      "epoch": 0.45664,
+      "grad_norm": 2.7732585003420285,
+      "learning_rate": 0.00019228432918999599,
+      "loss": 2.0051,
+      "step": 2854
+    },
+    {
+      "epoch": 0.4568,
+      "grad_norm": 2.2072785128355226,
+      "learning_rate": 0.00019227767435250277,
+      "loss": 1.8834,
+      "step": 2855
+    },
+    {
+      "epoch": 0.45696,
+      "grad_norm": 3.935576608930947,
+      "learning_rate": 0.00019227101676157823,
+      "loss": 2.0107,
+      "step": 2856
+    },
+    {
+      "epoch": 0.45712,
+      "grad_norm": 2.703162037723679,
+      "learning_rate": 0.000192264356417421,
+      "loss": 1.9011,
+      "step": 2857
+    },
+    {
+      "epoch": 0.45728,
+      "grad_norm": 2.2492515912394095,
+      "learning_rate": 0.0001922576933202298,
+      "loss": 1.9378,
+      "step": 2858
+    },
+    {
+      "epoch": 0.45744,
+      "grad_norm": 2.5614579756749385,
+      "learning_rate": 0.00019225102747020344,
+      "loss": 1.8871,
+      "step": 2859
+    },
+    {
+      "epoch": 0.4576,
+      "grad_norm": 2.585832071415096,
+      "learning_rate": 0.00019224435886754088,
+      "loss": 1.9832,
+      "step": 2860
+    },
+    {
+      "epoch": 0.45776,
+      "grad_norm": 2.5572873229883997,
+      "learning_rate": 0.00019223768751244107,
+      "loss": 1.9008,
+      "step": 2861
+    },
+    {
+      "epoch": 0.45792,
+      "grad_norm": 2.8183807650761934,
+      "learning_rate": 0.00019223101340510307,
+      "loss": 1.9953,
+      "step": 2862
+    },
+    {
+      "epoch": 0.45808,
+      "grad_norm": 2.3070760573914013,
+      "learning_rate": 0.000192224336545726,
+      "loss": 1.9172,
+      "step": 2863
+    },
+    {
+      "epoch": 0.45824,
+      "grad_norm": 2.4053631032122436,
+      "learning_rate": 0.00019221765693450916,
+      "loss": 2.1262,
+      "step": 2864
+    },
+    {
+      "epoch": 0.4584,
+      "grad_norm": 2.8729007293550657,
+      "learning_rate": 0.00019221097457165176,
+      "loss": 1.997,
+      "step": 2865
+    },
+    {
+      "epoch": 0.45856,
+      "grad_norm": 2.369361067620973,
+      "learning_rate": 0.00019220428945735326,
+      "loss": 1.9143,
+      "step": 2866
+    },
+    {
+      "epoch": 0.45872,
+      "grad_norm": 2.716675876312113,
+      "learning_rate": 0.00019219760159181314,
+      "loss": 1.9058,
+      "step": 2867
+    },
+    {
+      "epoch": 0.45888,
+      "grad_norm": 2.5887286289281084,
+      "learning_rate": 0.0001921909109752309,
+      "loss": 1.9419,
+      "step": 2868
+    },
+    {
+      "epoch": 0.45904,
+      "grad_norm": 2.615935648609441,
+      "learning_rate": 0.00019218421760780623,
+      "loss": 1.9309,
+      "step": 2869
+    },
+    {
+      "epoch": 0.4592,
+      "grad_norm": 2.5538225715851186,
+      "learning_rate": 0.00019217752148973882,
+      "loss": 2.0421,
+      "step": 2870
+    },
+    {
+      "epoch": 0.45936,
+      "grad_norm": 2.5414872412978036,
+      "learning_rate": 0.0001921708226212285,
+      "loss": 2.0046,
+      "step": 2871
+    },
+    {
+      "epoch": 0.45952,
+      "grad_norm": 2.352954355054497,
+      "learning_rate": 0.00019216412100247517,
+      "loss": 1.9241,
+      "step": 2872
+    },
+    {
+      "epoch": 0.45968,
+      "grad_norm": 2.6540803165922005,
+      "learning_rate": 0.0001921574166336787,
+      "loss": 1.9534,
+      "step": 2873
+    },
+    {
+      "epoch": 0.45984,
+      "grad_norm": 3.380924979622367,
+      "learning_rate": 0.00019215070951503926,
+      "loss": 2.0247,
+      "step": 2874
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 2.4294209456490417,
+      "learning_rate": 0.0001921439996467569,
+      "loss": 1.9432,
+      "step": 2875
+    },
+    {
+      "epoch": 0.46016,
+      "grad_norm": 2.956234071913719,
+      "learning_rate": 0.00019213728702903186,
+      "loss": 2.0427,
+      "step": 2876
+    },
+    {
+      "epoch": 0.46032,
+      "grad_norm": 2.89757566012767,
+      "learning_rate": 0.00019213057166206444,
+      "loss": 1.9206,
+      "step": 2877
+    },
+    {
+      "epoch": 0.46048,
+      "grad_norm": 3.160214477712574,
+      "learning_rate": 0.000192123853546055,
+      "loss": 1.9418,
+      "step": 2878
+    },
+    {
+      "epoch": 0.46064,
+      "grad_norm": 2.4008641237192117,
+      "learning_rate": 0.00019211713268120403,
+      "loss": 1.8951,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4608,
+      "grad_norm": 2.7370047931825625,
+      "learning_rate": 0.00019211040906771202,
+      "loss": 1.9775,
+      "step": 2880
+    },
+    {
+      "epoch": 0.46096,
+      "grad_norm": 2.254201113781316,
+      "learning_rate": 0.00019210368270577966,
+      "loss": 1.9812,
+      "step": 2881
+    },
+    {
+      "epoch": 0.46112,
+      "grad_norm": 2.965175132009835,
+      "learning_rate": 0.0001920969535956076,
+      "loss": 2.0517,
+      "step": 2882
+    },
+    {
+      "epoch": 0.46128,
+      "grad_norm": 2.8830089619459285,
+      "learning_rate": 0.00019209022173739663,
+      "loss": 1.9481,
+      "step": 2883
+    },
+    {
+      "epoch": 0.46144,
+      "grad_norm": 2.3129056009829276,
+      "learning_rate": 0.00019208348713134764,
+      "loss": 1.9111,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4616,
+      "grad_norm": 2.783894868307797,
+      "learning_rate": 0.00019207674977766157,
+      "loss": 1.9689,
+      "step": 2885
+    },
+    {
+      "epoch": 0.46176,
+      "grad_norm": 3.1236759010385673,
+      "learning_rate": 0.00019207000967653949,
+      "loss": 2.0074,
+      "step": 2886
+    },
+    {
+      "epoch": 0.46192,
+      "grad_norm": 2.540755883282408,
+      "learning_rate": 0.00019206326682818245,
+      "loss": 2.0545,
+      "step": 2887
+    },
+    {
+      "epoch": 0.46208,
+      "grad_norm": 2.9451784046982357,
+      "learning_rate": 0.00019205652123279168,
+      "loss": 2.0525,
+      "step": 2888
+    },
+    {
+      "epoch": 0.46224,
+      "grad_norm": 2.071852600530951,
+      "learning_rate": 0.00019204977289056846,
+      "loss": 1.9111,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4624,
+      "grad_norm": 2.331541570043893,
+      "learning_rate": 0.00019204302180171414,
+      "loss": 1.8689,
+      "step": 2890
+    },
+    {
+      "epoch": 0.46256,
+      "grad_norm": 2.3472596142913975,
+      "learning_rate": 0.00019203626796643016,
+      "loss": 1.9768,
+      "step": 2891
+    },
+    {
+      "epoch": 0.46272,
+      "grad_norm": 2.5870299814640974,
+      "learning_rate": 0.00019202951138491808,
+      "loss": 1.9474,
+      "step": 2892
+    },
+    {
+      "epoch": 0.46288,
+      "grad_norm": 2.9966491251078002,
+      "learning_rate": 0.0001920227520573795,
+      "loss": 2.1137,
+      "step": 2893
+    },
+    {
+      "epoch": 0.46304,
+      "grad_norm": 2.1555915896882776,
+      "learning_rate": 0.00019201598998401604,
+      "loss": 1.8854,
+      "step": 2894
+    },
+    {
+      "epoch": 0.4632,
+      "grad_norm": 2.432034173071389,
+      "learning_rate": 0.00019200922516502954,
+      "loss": 1.9681,
+      "step": 2895
+    },
+    {
+      "epoch": 0.46336,
+      "grad_norm": 2.2459375493725307,
+      "learning_rate": 0.00019200245760062185,
+      "loss": 1.9363,
+      "step": 2896
+    },
+    {
+      "epoch": 0.46352,
+      "grad_norm": 2.0921406372729345,
+      "learning_rate": 0.00019199568729099488,
+      "loss": 1.8644,
+      "step": 2897
+    },
+    {
+      "epoch": 0.46368,
+      "grad_norm": 2.535627983107855,
+      "learning_rate": 0.00019198891423635064,
+      "loss": 1.8926,
+      "step": 2898
+    },
+    {
+      "epoch": 0.46384,
+      "grad_norm": 2.3613059641885865,
+      "learning_rate": 0.00019198213843689126,
+      "loss": 2.0103,
+      "step": 2899
+    },
+    {
+      "epoch": 0.464,
+      "grad_norm": 2.477696722834747,
+      "learning_rate": 0.0001919753598928189,
+      "loss": 2.0696,
+      "step": 2900
+    },
+    {
+      "epoch": 0.46416,
+      "grad_norm": 2.607309338912845,
+      "learning_rate": 0.0001919685786043358,
+      "loss": 1.833,
+      "step": 2901
+    },
+    {
+      "epoch": 0.46432,
+      "grad_norm": 3.4964505641547157,
+      "learning_rate": 0.00019196179457164433,
+      "loss": 2.082,
+      "step": 2902
+    },
+    {
+      "epoch": 0.46448,
+      "grad_norm": 2.2305226191088385,
+      "learning_rate": 0.00019195500779494693,
+      "loss": 1.9011,
+      "step": 2903
+    },
+    {
+      "epoch": 0.46464,
+      "grad_norm": 2.523265616264017,
+      "learning_rate": 0.0001919482182744461,
+      "loss": 1.9607,
+      "step": 2904
+    },
+    {
+      "epoch": 0.4648,
+      "grad_norm": 3.0198839082771913,
+      "learning_rate": 0.0001919414260103444,
+      "loss": 1.9861,
+      "step": 2905
+    },
+    {
+      "epoch": 0.46496,
+      "grad_norm": 3.6403463675214645,
+      "learning_rate": 0.0001919346310028445,
+      "loss": 1.9289,
+      "step": 2906
+    },
+    {
+      "epoch": 0.46512,
+      "grad_norm": 2.6363171371266696,
+      "learning_rate": 0.0001919278332521492,
+      "loss": 1.9836,
+      "step": 2907
+    },
+    {
+      "epoch": 0.46528,
+      "grad_norm": 3.985861698551258,
+      "learning_rate": 0.0001919210327584613,
+      "loss": 1.8434,
+      "step": 2908
+    },
+    {
+      "epoch": 0.46544,
+      "grad_norm": 2.6152977712831067,
+      "learning_rate": 0.00019191422952198375,
+      "loss": 1.8894,
+      "step": 2909
+    },
+    {
+      "epoch": 0.4656,
+      "grad_norm": 2.628059361780515,
+      "learning_rate": 0.00019190742354291952,
+      "loss": 2.0151,
+      "step": 2910
+    },
+    {
+      "epoch": 0.46576,
+      "grad_norm": 2.792275839861921,
+      "learning_rate": 0.00019190061482147165,
+      "loss": 2.013,
+      "step": 2911
+    },
+    {
+      "epoch": 0.46592,
+      "grad_norm": 2.242183288107548,
+      "learning_rate": 0.00019189380335784338,
+      "loss": 1.8831,
+      "step": 2912
+    },
+    {
+      "epoch": 0.46608,
+      "grad_norm": 3.41355436368681,
+      "learning_rate": 0.0001918869891522379,
+      "loss": 1.9673,
+      "step": 2913
+    },
+    {
+      "epoch": 0.46624,
+      "grad_norm": 2.9662599139477344,
+      "learning_rate": 0.0001918801722048586,
+      "loss": 2.1184,
+      "step": 2914
+    },
+    {
+      "epoch": 0.4664,
+      "grad_norm": 2.9526551514806094,
+      "learning_rate": 0.0001918733525159088,
+      "loss": 1.8643,
+      "step": 2915
+    },
+    {
+      "epoch": 0.46656,
+      "grad_norm": 2.413104574179478,
+      "learning_rate": 0.00019186653008559205,
+      "loss": 1.9208,
+      "step": 2916
+    },
+    {
+      "epoch": 0.46672,
+      "grad_norm": 2.131690495103008,
+      "learning_rate": 0.00019185970491411193,
+      "loss": 1.7851,
+      "step": 2917
+    },
+    {
+      "epoch": 0.46688,
+      "grad_norm": 2.3788191608570646,
+      "learning_rate": 0.00019185287700167205,
+      "loss": 1.9815,
+      "step": 2918
+    },
+    {
+      "epoch": 0.46704,
+      "grad_norm": 2.916887004238271,
+      "learning_rate": 0.00019184604634847616,
+      "loss": 2.0106,
+      "step": 2919
+    },
+    {
+      "epoch": 0.4672,
+      "grad_norm": 2.661068418253033,
+      "learning_rate": 0.0001918392129547281,
+      "loss": 2.0585,
+      "step": 2920
+    },
+    {
+      "epoch": 0.46736,
+      "grad_norm": 3.44308576020634,
+      "learning_rate": 0.00019183237682063173,
+      "loss": 1.981,
+      "step": 2921
+    },
+    {
+      "epoch": 0.46752,
+      "grad_norm": 2.2580204264303476,
+      "learning_rate": 0.00019182553794639103,
+      "loss": 1.9801,
+      "step": 2922
+    },
+    {
+      "epoch": 0.46768,
+      "grad_norm": 2.2802594001162326,
+      "learning_rate": 0.0001918186963322101,
+      "loss": 1.9338,
+      "step": 2923
+    },
+    {
+      "epoch": 0.46784,
+      "grad_norm": 2.616750557854618,
+      "learning_rate": 0.00019181185197829307,
+      "loss": 1.9257,
+      "step": 2924
+    },
+    {
+      "epoch": 0.468,
+      "grad_norm": 2.272847908135625,
+      "learning_rate": 0.00019180500488484418,
+      "loss": 1.8775,
+      "step": 2925
+    },
+    {
+      "epoch": 0.46816,
+      "grad_norm": 2.272897597826903,
+      "learning_rate": 0.0001917981550520677,
+      "loss": 1.9623,
+      "step": 2926
+    },
+    {
+      "epoch": 0.46832,
+      "grad_norm": 2.2169127948765572,
+      "learning_rate": 0.00019179130248016802,
+      "loss": 1.9846,
+      "step": 2927
+    },
+    {
+      "epoch": 0.46848,
+      "grad_norm": 2.375604920032883,
+      "learning_rate": 0.00019178444716934964,
+      "loss": 2.0231,
+      "step": 2928
+    },
+    {
+      "epoch": 0.46864,
+      "grad_norm": 2.112693614025373,
+      "learning_rate": 0.0001917775891198171,
+      "loss": 2.0203,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4688,
+      "grad_norm": 2.7629548853917405,
+      "learning_rate": 0.00019177072833177506,
+      "loss": 1.9254,
+      "step": 2930
+    },
+    {
+      "epoch": 0.46896,
+      "grad_norm": 2.137859797428623,
+      "learning_rate": 0.0001917638648054282,
+      "loss": 1.851,
+      "step": 2931
+    },
+    {
+      "epoch": 0.46912,
+      "grad_norm": 2.7830803967592277,
+      "learning_rate": 0.00019175699854098133,
+      "loss": 1.9024,
+      "step": 2932
+    },
+    {
+      "epoch": 0.46928,
+      "grad_norm": 2.817582106979147,
+      "learning_rate": 0.00019175012953863928,
+      "loss": 2.0399,
+      "step": 2933
+    },
+    {
+      "epoch": 0.46944,
+      "grad_norm": 2.3614954266735753,
+      "learning_rate": 0.00019174325779860714,
+      "loss": 1.9091,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4696,
+      "grad_norm": 2.489118152140196,
+      "learning_rate": 0.00019173638332108982,
+      "loss": 1.9197,
+      "step": 2935
+    },
+    {
+      "epoch": 0.46976,
+      "grad_norm": 2.217228453420316,
+      "learning_rate": 0.00019172950610629247,
+      "loss": 1.9969,
+      "step": 2936
+    },
+    {
+      "epoch": 0.46992,
+      "grad_norm": 2.8335546165048435,
+      "learning_rate": 0.00019172262615442035,
+      "loss": 1.8573,
+      "step": 2937
+    },
+    {
+      "epoch": 0.47008,
+      "grad_norm": 2.650909989094047,
+      "learning_rate": 0.00019171574346567877,
+      "loss": 1.9918,
+      "step": 2938
+    },
+    {
+      "epoch": 0.47024,
+      "grad_norm": 2.681577722944498,
+      "learning_rate": 0.000191708858040273,
+      "loss": 1.9332,
+      "step": 2939
+    },
+    {
+      "epoch": 0.4704,
+      "grad_norm": 2.431330219350318,
+      "learning_rate": 0.00019170196987840853,
+      "loss": 1.9259,
+      "step": 2940
+    },
+    {
+      "epoch": 0.47056,
+      "grad_norm": 2.7631541782430586,
+      "learning_rate": 0.00019169507898029092,
+      "loss": 1.9567,
+      "step": 2941
+    },
+    {
+      "epoch": 0.47072,
+      "grad_norm": 2.2461508524933804,
+      "learning_rate": 0.00019168818534612576,
+      "loss": 1.9084,
+      "step": 2942
+    },
+    {
+      "epoch": 0.47088,
+      "grad_norm": 2.519757660776704,
+      "learning_rate": 0.00019168128897611875,
+      "loss": 1.9556,
+      "step": 2943
+    },
+    {
+      "epoch": 0.47104,
+      "grad_norm": 2.8821228189179195,
+      "learning_rate": 0.0001916743898704757,
+      "loss": 1.9848,
+      "step": 2944
+    },
+    {
+      "epoch": 0.4712,
+      "grad_norm": 2.504044268654983,
+      "learning_rate": 0.0001916674880294024,
+      "loss": 1.9345,
+      "step": 2945
+    },
+    {
+      "epoch": 0.47136,
+      "grad_norm": 2.9092753831940925,
+      "learning_rate": 0.00019166058345310486,
+      "loss": 1.9154,
+      "step": 2946
+    },
+    {
+      "epoch": 0.47152,
+      "grad_norm": 2.393160652800562,
+      "learning_rate": 0.00019165367614178903,
+      "loss": 1.9155,
+      "step": 2947
+    },
+    {
+      "epoch": 0.47168,
+      "grad_norm": 2.789193661632494,
+      "learning_rate": 0.0001916467660956611,
+      "loss": 1.96,
+      "step": 2948
+    },
+    {
+      "epoch": 0.47184,
+      "grad_norm": 2.317107189217505,
+      "learning_rate": 0.0001916398533149272,
+      "loss": 1.7828,
+      "step": 2949
+    },
+    {
+      "epoch": 0.472,
+      "grad_norm": 2.3884122310487372,
+      "learning_rate": 0.00019163293779979362,
+      "loss": 1.9444,
+      "step": 2950
+    },
+    {
+      "epoch": 0.47216,
+      "grad_norm": 2.2850270688190695,
+      "learning_rate": 0.00019162601955046669,
+      "loss": 1.8841,
+      "step": 2951
+    },
+    {
+      "epoch": 0.47232,
+      "grad_norm": 2.817831332947141,
+      "learning_rate": 0.00019161909856715284,
+      "loss": 1.9065,
+      "step": 2952
+    },
+    {
+      "epoch": 0.47248,
+      "grad_norm": 2.4082488057985376,
+      "learning_rate": 0.0001916121748500586,
+      "loss": 1.9002,
+      "step": 2953
+    },
+    {
+      "epoch": 0.47264,
+      "grad_norm": 3.150184594235518,
+      "learning_rate": 0.00019160524839939057,
+      "loss": 1.904,
+      "step": 2954
+    },
+    {
+      "epoch": 0.4728,
+      "grad_norm": 2.744540907695351,
+      "learning_rate": 0.0001915983192153554,
+      "loss": 1.9294,
+      "step": 2955
+    },
+    {
+      "epoch": 0.47296,
+      "grad_norm": 3.0346184986206937,
+      "learning_rate": 0.00019159138729815984,
+      "loss": 1.9487,
+      "step": 2956
+    },
+    {
+      "epoch": 0.47312,
+      "grad_norm": 1.9883806912029534,
+      "learning_rate": 0.00019158445264801078,
+      "loss": 1.9946,
+      "step": 2957
+    },
+    {
+      "epoch": 0.47328,
+      "grad_norm": 3.2492528236556337,
+      "learning_rate": 0.0001915775152651151,
+      "loss": 2.0075,
+      "step": 2958
+    },
+    {
+      "epoch": 0.47344,
+      "grad_norm": 2.409354852925433,
+      "learning_rate": 0.00019157057514967976,
+      "loss": 2.0456,
+      "step": 2959
+    },
+    {
+      "epoch": 0.4736,
+      "grad_norm": 2.075829884024126,
+      "learning_rate": 0.00019156363230191196,
+      "loss": 1.8775,
+      "step": 2960
+    },
+    {
+      "epoch": 0.47376,
+      "grad_norm": 2.54145387960791,
+      "learning_rate": 0.00019155668672201875,
+      "loss": 2.0392,
+      "step": 2961
+    },
+    {
+      "epoch": 0.47392,
+      "grad_norm": 2.336366525746161,
+      "learning_rate": 0.00019154973841020741,
+      "loss": 1.9074,
+      "step": 2962
+    },
+    {
+      "epoch": 0.47408,
+      "grad_norm": 2.751822503016587,
+      "learning_rate": 0.00019154278736668528,
+      "loss": 1.8712,
+      "step": 2963
+    },
+    {
+      "epoch": 0.47424,
+      "grad_norm": 2.3306292247153504,
+      "learning_rate": 0.0001915358335916598,
+      "loss": 2.022,
+      "step": 2964
+    },
+    {
+      "epoch": 0.4744,
+      "grad_norm": 2.6254122868330887,
+      "learning_rate": 0.0001915288770853384,
+      "loss": 2.0478,
+      "step": 2965
+    },
+    {
+      "epoch": 0.47456,
+      "grad_norm": 3.5072730038214983,
+      "learning_rate": 0.00019152191784792874,
+      "loss": 1.9188,
+      "step": 2966
+    },
+    {
+      "epoch": 0.47472,
+      "grad_norm": 2.4962213840555063,
+      "learning_rate": 0.00019151495587963835,
+      "loss": 1.9555,
+      "step": 2967
+    },
+    {
+      "epoch": 0.47488,
+      "grad_norm": 3.1954094314212966,
+      "learning_rate": 0.00019150799118067505,
+      "loss": 1.7777,
+      "step": 2968
+    },
+    {
+      "epoch": 0.47504,
+      "grad_norm": 2.179769192282817,
+      "learning_rate": 0.0001915010237512466,
+      "loss": 2.0522,
+      "step": 2969
+    },
+    {
+      "epoch": 0.4752,
+      "grad_norm": 3.012275616893334,
+      "learning_rate": 0.000191494053591561,
+      "loss": 1.9086,
+      "step": 2970
+    },
+    {
+      "epoch": 0.47536,
+      "grad_norm": 2.65338555601258,
+      "learning_rate": 0.00019148708070182609,
+      "loss": 1.8743,
+      "step": 2971
+    },
+    {
+      "epoch": 0.47552,
+      "grad_norm": 2.915378607850128,
+      "learning_rate": 0.00019148010508225003,
+      "loss": 1.9839,
+      "step": 2972
+    },
+    {
+      "epoch": 0.47568,
+      "grad_norm": 2.6944407305381515,
+      "learning_rate": 0.00019147312673304095,
+      "loss": 2.038,
+      "step": 2973
+    },
+    {
+      "epoch": 0.47584,
+      "grad_norm": 3.3439132210381577,
+      "learning_rate": 0.00019146614565440705,
+      "loss": 1.9077,
+      "step": 2974
+    },
+    {
+      "epoch": 0.476,
+      "grad_norm": 2.369040341907794,
+      "learning_rate": 0.00019145916184655664,
+      "loss": 2.0194,
+      "step": 2975
+    },
+    {
+      "epoch": 0.47616,
+      "grad_norm": 3.042735852579854,
+      "learning_rate": 0.0001914521753096981,
+      "loss": 1.9986,
+      "step": 2976
+    },
+    {
+      "epoch": 0.47632,
+      "grad_norm": 2.4823497797669973,
+      "learning_rate": 0.00019144518604403992,
+      "loss": 1.9415,
+      "step": 2977
+    },
+    {
+      "epoch": 0.47648,
+      "grad_norm": 2.2875955605700855,
+      "learning_rate": 0.00019143819404979063,
+      "loss": 1.9851,
+      "step": 2978
+    },
+    {
+      "epoch": 0.47664,
+      "grad_norm": 2.622129530966095,
+      "learning_rate": 0.00019143119932715888,
+      "loss": 1.9699,
+      "step": 2979
+    },
+    {
+      "epoch": 0.4768,
+      "grad_norm": 2.7420809524730023,
+      "learning_rate": 0.00019142420187635338,
+      "loss": 1.9341,
+      "step": 2980
+    },
+    {
+      "epoch": 0.47696,
+      "grad_norm": 2.0525810205569357,
+      "learning_rate": 0.0001914172016975829,
+      "loss": 2.0844,
+      "step": 2981
+    },
+    {
+      "epoch": 0.47712,
+      "grad_norm": 2.718823015983253,
+      "learning_rate": 0.00019141019879105633,
+      "loss": 1.9018,
+      "step": 2982
+    },
+    {
+      "epoch": 0.47728,
+      "grad_norm": 2.5260813620829436,
+      "learning_rate": 0.00019140319315698263,
+      "loss": 2.0218,
+      "step": 2983
+    },
+    {
+      "epoch": 0.47744,
+      "grad_norm": 2.725810564332601,
+      "learning_rate": 0.00019139618479557084,
+      "loss": 2.0032,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4776,
+      "grad_norm": 2.377162517687332,
+      "learning_rate": 0.00019138917370703008,
+      "loss": 1.8424,
+      "step": 2985
+    },
+    {
+      "epoch": 0.47776,
+      "grad_norm": 2.4005612203924747,
+      "learning_rate": 0.00019138215989156953,
+      "loss": 1.9309,
+      "step": 2986
+    },
+    {
+      "epoch": 0.47792,
+      "grad_norm": 2.664666205935313,
+      "learning_rate": 0.0001913751433493985,
+      "loss": 1.9393,
+      "step": 2987
+    },
+    {
+      "epoch": 0.47808,
+      "grad_norm": 2.2556357450535374,
+      "learning_rate": 0.0001913681240807263,
+      "loss": 1.943,
+      "step": 2988
+    },
+    {
+      "epoch": 0.47824,
+      "grad_norm": 2.2538636182914398,
+      "learning_rate": 0.00019136110208576245,
+      "loss": 1.8887,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4784,
+      "grad_norm": 2.446875190115401,
+      "learning_rate": 0.00019135407736471644,
+      "loss": 1.9408,
+      "step": 2990
+    },
+    {
+      "epoch": 0.47856,
+      "grad_norm": 2.50634489679612,
+      "learning_rate": 0.0001913470499177979,
+      "loss": 1.8928,
+      "step": 2991
+    },
+    {
+      "epoch": 0.47872,
+      "grad_norm": 2.6317604472563723,
+      "learning_rate": 0.00019134001974521647,
+      "loss": 1.9358,
+      "step": 2992
+    },
+    {
+      "epoch": 0.47888,
+      "grad_norm": 2.7966289707058407,
+      "learning_rate": 0.00019133298684718194,
+      "loss": 1.9347,
+      "step": 2993
+    },
+    {
+      "epoch": 0.47904,
+      "grad_norm": 2.9409123256305563,
+      "learning_rate": 0.0001913259512239042,
+      "loss": 1.8724,
+      "step": 2994
+    },
+    {
+      "epoch": 0.4792,
+      "grad_norm": 2.407386786416185,
+      "learning_rate": 0.00019131891287559312,
+      "loss": 1.9035,
+      "step": 2995
+    },
+    {
+      "epoch": 0.47936,
+      "grad_norm": 2.972568195237764,
+      "learning_rate": 0.00019131187180245875,
+      "loss": 1.9422,
+      "step": 2996
+    },
+    {
+      "epoch": 0.47952,
+      "grad_norm": 2.1447539659829347,
+      "learning_rate": 0.00019130482800471118,
+      "loss": 1.8817,
+      "step": 2997
+    },
+    {
+      "epoch": 0.47968,
+      "grad_norm": 2.568842865702065,
+      "learning_rate": 0.0001912977814825606,
+      "loss": 1.9324,
+      "step": 2998
+    },
+    {
+      "epoch": 0.47984,
+      "grad_norm": 3.1350884110015467,
+      "learning_rate": 0.00019129073223621726,
+      "loss": 2.012,
+      "step": 2999
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2495702725973348,
+      "learning_rate": 0.0001912836802658915,
+      "loss": 1.8829,
+      "step": 3000
+    },
+    {
+      "epoch": 0.48016,
+      "grad_norm": 2.7799609883381793,
+      "learning_rate": 0.0001912766255717937,
+      "loss": 1.8952,
+      "step": 3001
+    },
+    {
+      "epoch": 0.48032,
+      "grad_norm": 2.7797610951757123,
+      "learning_rate": 0.0001912695681541344,
+      "loss": 1.9989,
+      "step": 3002
+    },
+    {
+      "epoch": 0.48048,
+      "grad_norm": 2.9166759889933886,
+      "learning_rate": 0.0001912625080131242,
+      "loss": 1.8951,
+      "step": 3003
+    },
+    {
+      "epoch": 0.48064,
+      "grad_norm": 2.1806665275630692,
+      "learning_rate": 0.00019125544514897376,
+      "loss": 1.8645,
+      "step": 3004
+    },
+    {
+      "epoch": 0.4808,
+      "grad_norm": 2.445299059472722,
+      "learning_rate": 0.00019124837956189378,
+      "loss": 1.9677,
+      "step": 3005
+    },
+    {
+      "epoch": 0.48096,
+      "grad_norm": 2.763016427704139,
+      "learning_rate": 0.00019124131125209512,
+      "loss": 2.0082,
+      "step": 3006
+    },
+    {
+      "epoch": 0.48112,
+      "grad_norm": 2.738168340394027,
+      "learning_rate": 0.00019123424021978867,
+      "loss": 2.0428,
+      "step": 3007
+    },
+    {
+      "epoch": 0.48128,
+      "grad_norm": 2.5792714658792115,
+      "learning_rate": 0.0001912271664651855,
+      "loss": 1.9664,
+      "step": 3008
+    },
+    {
+      "epoch": 0.48144,
+      "grad_norm": 2.9061143933652853,
+      "learning_rate": 0.00019122008998849656,
+      "loss": 1.8697,
+      "step": 3009
+    },
+    {
+      "epoch": 0.4816,
+      "grad_norm": 2.3427878007529497,
+      "learning_rate": 0.00019121301078993307,
+      "loss": 1.8924,
+      "step": 3010
+    },
+    {
+      "epoch": 0.48176,
+      "grad_norm": 2.381107517120459,
+      "learning_rate": 0.00019120592886970626,
+      "loss": 1.9186,
+      "step": 3011
+    },
+    {
+      "epoch": 0.48192,
+      "grad_norm": 2.6624886071389255,
+      "learning_rate": 0.00019119884422802742,
+      "loss": 1.9559,
+      "step": 3012
+    },
+    {
+      "epoch": 0.48208,
+      "grad_norm": 2.901774126576396,
+      "learning_rate": 0.00019119175686510795,
+      "loss": 1.9082,
+      "step": 3013
+    },
+    {
+      "epoch": 0.48224,
+      "grad_norm": 2.1737415898380603,
+      "learning_rate": 0.00019118466678115934,
+      "loss": 2.0578,
+      "step": 3014
+    },
+    {
+      "epoch": 0.4824,
+      "grad_norm": 2.6482749047028347,
+      "learning_rate": 0.00019117757397639316,
+      "loss": 1.9595,
+      "step": 3015
+    },
+    {
+      "epoch": 0.48256,
+      "grad_norm": 2.8770232856297104,
+      "learning_rate": 0.00019117047845102102,
+      "loss": 1.8799,
+      "step": 3016
+    },
+    {
+      "epoch": 0.48272,
+      "grad_norm": 2.1459871299928586,
+      "learning_rate": 0.00019116338020525465,
+      "loss": 1.9477,
+      "step": 3017
+    },
+    {
+      "epoch": 0.48288,
+      "grad_norm": 2.2489546033309034,
+      "learning_rate": 0.00019115627923930586,
+      "loss": 1.9543,
+      "step": 3018
+    },
+    {
+      "epoch": 0.48304,
+      "grad_norm": 2.529880579645887,
+      "learning_rate": 0.00019114917555338652,
+      "loss": 1.9005,
+      "step": 3019
+    },
+    {
+      "epoch": 0.4832,
+      "grad_norm": 2.3262111060985546,
+      "learning_rate": 0.0001911420691477086,
+      "loss": 1.9283,
+      "step": 3020
+    },
+    {
+      "epoch": 0.48336,
+      "grad_norm": 2.4450898083947292,
+      "learning_rate": 0.00019113496002248415,
+      "loss": 1.9301,
+      "step": 3021
+    },
+    {
+      "epoch": 0.48352,
+      "grad_norm": 2.356618587447162,
+      "learning_rate": 0.00019112784817792528,
+      "loss": 1.8541,
+      "step": 3022
+    },
+    {
+      "epoch": 0.48368,
+      "grad_norm": 2.5068862894750206,
+      "learning_rate": 0.00019112073361424422,
+      "loss": 1.8428,
+      "step": 3023
+    },
+    {
+      "epoch": 0.48384,
+      "grad_norm": 2.496496734671923,
+      "learning_rate": 0.00019111361633165325,
+      "loss": 2.0123,
+      "step": 3024
+    },
+    {
+      "epoch": 0.484,
+      "grad_norm": 2.814239438062519,
+      "learning_rate": 0.00019110649633036472,
+      "loss": 1.938,
+      "step": 3025
+    },
+    {
+      "epoch": 0.48416,
+      "grad_norm": 2.3466181587136172,
+      "learning_rate": 0.0001910993736105911,
+      "loss": 1.8518,
+      "step": 3026
+    },
+    {
+      "epoch": 0.48432,
+      "grad_norm": 3.3374360891741595,
+      "learning_rate": 0.00019109224817254495,
+      "loss": 1.9323,
+      "step": 3027
+    },
+    {
+      "epoch": 0.48448,
+      "grad_norm": 2.4690738956914,
+      "learning_rate": 0.0001910851200164388,
+      "loss": 2.0258,
+      "step": 3028
+    },
+    {
+      "epoch": 0.48464,
+      "grad_norm": 2.7542739822173172,
+      "learning_rate": 0.00019107798914248542,
+      "loss": 1.9621,
+      "step": 3029
+    },
+    {
+      "epoch": 0.4848,
+      "grad_norm": 2.80518798933369,
+      "learning_rate": 0.00019107085555089754,
+      "loss": 1.9609,
+      "step": 3030
+    },
+    {
+      "epoch": 0.48496,
+      "grad_norm": 2.60368691307603,
+      "learning_rate": 0.00019106371924188805,
+      "loss": 1.8167,
+      "step": 3031
+    },
+    {
+      "epoch": 0.48512,
+      "grad_norm": 2.1897238130232957,
+      "learning_rate": 0.0001910565802156699,
+      "loss": 1.8919,
+      "step": 3032
+    },
+    {
+      "epoch": 0.48528,
+      "grad_norm": 2.9071509027790987,
+      "learning_rate": 0.000191049438472456,
+      "loss": 1.9215,
+      "step": 3033
+    },
+    {
+      "epoch": 0.48544,
+      "grad_norm": 2.3706505264622235,
+      "learning_rate": 0.0001910422940124596,
+      "loss": 1.9494,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4856,
+      "grad_norm": 3.209621736690457,
+      "learning_rate": 0.00019103514683589378,
+      "loss": 1.9952,
+      "step": 3035
+    },
+    {
+      "epoch": 0.48576,
+      "grad_norm": 2.675477880303465,
+      "learning_rate": 0.00019102799694297183,
+      "loss": 1.8721,
+      "step": 3036
+    },
+    {
+      "epoch": 0.48592,
+      "grad_norm": 2.8169339844958845,
+      "learning_rate": 0.00019102084433390712,
+      "loss": 1.9959,
+      "step": 3037
+    },
+    {
+      "epoch": 0.48608,
+      "grad_norm": 2.9407352862897462,
+      "learning_rate": 0.00019101368900891304,
+      "loss": 1.9052,
+      "step": 3038
+    },
+    {
+      "epoch": 0.48624,
+      "grad_norm": 2.5818318891530954,
+      "learning_rate": 0.00019100653096820307,
+      "loss": 1.9611,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4864,
+      "grad_norm": 2.8323996305323806,
+      "learning_rate": 0.00019099937021199086,
+      "loss": 1.9064,
+      "step": 3040
+    },
+    {
+      "epoch": 0.48656,
+      "grad_norm": 2.0808319620849622,
+      "learning_rate": 0.00019099220674049003,
+      "loss": 1.9417,
+      "step": 3041
+    },
+    {
+      "epoch": 0.48672,
+      "grad_norm": 2.397734170319552,
+      "learning_rate": 0.00019098504055391433,
+      "loss": 1.9306,
+      "step": 3042
+    },
+    {
+      "epoch": 0.48688,
+      "grad_norm": 2.5093911035775602,
+      "learning_rate": 0.00019097787165247762,
+      "loss": 1.9907,
+      "step": 3043
+    },
+    {
+      "epoch": 0.48704,
+      "grad_norm": 2.642028952260672,
+      "learning_rate": 0.00019097070003639376,
+      "loss": 2.0083,
+      "step": 3044
+    },
+    {
+      "epoch": 0.4872,
+      "grad_norm": 2.530653235637901,
+      "learning_rate": 0.0001909635257058768,
+      "loss": 1.9176,
+      "step": 3045
+    },
+    {
+      "epoch": 0.48736,
+      "grad_norm": 2.2408866514222017,
+      "learning_rate": 0.00019095634866114075,
+      "loss": 1.9365,
+      "step": 3046
+    },
+    {
+      "epoch": 0.48752,
+      "grad_norm": 2.3344465896132154,
+      "learning_rate": 0.00019094916890239982,
+      "loss": 1.9068,
+      "step": 3047
+    },
+    {
+      "epoch": 0.48768,
+      "grad_norm": 2.6679277806662953,
+      "learning_rate": 0.00019094198642986824,
+      "loss": 1.8307,
+      "step": 3048
+    },
+    {
+      "epoch": 0.48784,
+      "grad_norm": 2.5150817400559706,
+      "learning_rate": 0.00019093480124376025,
+      "loss": 1.9972,
+      "step": 3049
+    },
+    {
+      "epoch": 0.488,
+      "grad_norm": 2.5645331292667772,
+      "learning_rate": 0.0001909276133442903,
+      "loss": 1.8243,
+      "step": 3050
+    },
+    {
+      "epoch": 0.48816,
+      "grad_norm": 2.590267107742736,
+      "learning_rate": 0.0001909204227316729,
+      "loss": 1.8684,
+      "step": 3051
+    },
+    {
+      "epoch": 0.48832,
+      "grad_norm": 2.378699003379048,
+      "learning_rate": 0.00019091322940612254,
+      "loss": 2.0089,
+      "step": 3052
+    },
+    {
+      "epoch": 0.48848,
+      "grad_norm": 2.462751453746808,
+      "learning_rate": 0.00019090603336785388,
+      "loss": 1.9074,
+      "step": 3053
+    },
+    {
+      "epoch": 0.48864,
+      "grad_norm": 2.4596373946576304,
+      "learning_rate": 0.0001908988346170817,
+      "loss": 2.0363,
+      "step": 3054
+    },
+    {
+      "epoch": 0.4888,
+      "grad_norm": 3.0298885029089426,
+      "learning_rate": 0.0001908916331540207,
+      "loss": 2.0223,
+      "step": 3055
+    },
+    {
+      "epoch": 0.48896,
+      "grad_norm": 2.2657414427484373,
+      "learning_rate": 0.0001908844289788858,
+      "loss": 1.8766,
+      "step": 3056
+    },
+    {
+      "epoch": 0.48912,
+      "grad_norm": 2.4390486433861187,
+      "learning_rate": 0.000190877222091892,
+      "loss": 1.9849,
+      "step": 3057
+    },
+    {
+      "epoch": 0.48928,
+      "grad_norm": 2.974764470979545,
+      "learning_rate": 0.0001908700124932543,
+      "loss": 1.8961,
+      "step": 3058
+    },
+    {
+      "epoch": 0.48944,
+      "grad_norm": 2.588316168863133,
+      "learning_rate": 0.0001908628001831879,
+      "loss": 1.8322,
+      "step": 3059
+    },
+    {
+      "epoch": 0.4896,
+      "grad_norm": 2.093284582025604,
+      "learning_rate": 0.00019085558516190786,
+      "loss": 1.8421,
+      "step": 3060
+    },
+    {
+      "epoch": 0.48976,
+      "grad_norm": 2.8531224445125467,
+      "learning_rate": 0.00019084836742962958,
+      "loss": 1.8555,
+      "step": 3061
+    },
+    {
+      "epoch": 0.48992,
+      "grad_norm": 2.8262828545481895,
+      "learning_rate": 0.00019084114698656842,
+      "loss": 1.9514,
+      "step": 3062
+    },
+    {
+      "epoch": 0.49008,
+      "grad_norm": 2.860414729071258,
+      "learning_rate": 0.0001908339238329398,
+      "loss": 1.8251,
+      "step": 3063
+    },
+    {
+      "epoch": 0.49024,
+      "grad_norm": 2.0234082774324462,
+      "learning_rate": 0.0001908266979689592,
+      "loss": 1.9025,
+      "step": 3064
+    },
+    {
+      "epoch": 0.4904,
+      "grad_norm": 2.3956478534194545,
+      "learning_rate": 0.0001908194693948423,
+      "loss": 1.9197,
+      "step": 3065
+    },
+    {
+      "epoch": 0.49056,
+      "grad_norm": 2.360949669749447,
+      "learning_rate": 0.00019081223811080478,
+      "loss": 1.8793,
+      "step": 3066
+    },
+    {
+      "epoch": 0.49072,
+      "grad_norm": 2.705240554443098,
+      "learning_rate": 0.00019080500411706246,
+      "loss": 1.9616,
+      "step": 3067
+    },
+    {
+      "epoch": 0.49088,
+      "grad_norm": 2.5683520294535866,
+      "learning_rate": 0.00019079776741383106,
+      "loss": 1.8578,
+      "step": 3068
+    },
+    {
+      "epoch": 0.49104,
+      "grad_norm": 2.6274720473116497,
+      "learning_rate": 0.00019079052800132662,
+      "loss": 1.9252,
+      "step": 3069
+    },
+    {
+      "epoch": 0.4912,
+      "grad_norm": 2.215828775949358,
+      "learning_rate": 0.0001907832858797651,
+      "loss": 1.8796,
+      "step": 3070
+    },
+    {
+      "epoch": 0.49136,
+      "grad_norm": 2.356356144549243,
+      "learning_rate": 0.00019077604104936264,
+      "loss": 1.9206,
+      "step": 3071
+    },
+    {
+      "epoch": 0.49152,
+      "grad_norm": 2.9102568515507983,
+      "learning_rate": 0.00019076879351033536,
+      "loss": 2.0021,
+      "step": 3072
+    },
+    {
+      "epoch": 0.49168,
+      "grad_norm": 2.2134973719330895,
+      "learning_rate": 0.00019076154326289958,
+      "loss": 1.9542,
+      "step": 3073
+    },
+    {
+      "epoch": 0.49184,
+      "grad_norm": 2.3041118243810206,
+      "learning_rate": 0.00019075429030727156,
+      "loss": 1.8363,
+      "step": 3074
+    },
+    {
+      "epoch": 0.492,
+      "grad_norm": 2.5525843358150557,
+      "learning_rate": 0.0001907470346436678,
+      "loss": 1.9254,
+      "step": 3075
+    },
+    {
+      "epoch": 0.49216,
+      "grad_norm": 2.6169194431397167,
+      "learning_rate": 0.00019073977627230476,
+      "loss": 1.9139,
+      "step": 3076
+    },
+    {
+      "epoch": 0.49232,
+      "grad_norm": 3.3554588003116526,
+      "learning_rate": 0.00019073251519339897,
+      "loss": 1.8205,
+      "step": 3077
+    },
+    {
+      "epoch": 0.49248,
+      "grad_norm": 2.3759804435358047,
+      "learning_rate": 0.00019072525140716718,
+      "loss": 1.9068,
+      "step": 3078
+    },
+    {
+      "epoch": 0.49264,
+      "grad_norm": 2.431491830976655,
+      "learning_rate": 0.00019071798491382608,
+      "loss": 1.9289,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4928,
+      "grad_norm": 2.5113669176213422,
+      "learning_rate": 0.0001907107157135925,
+      "loss": 1.816,
+      "step": 3080
+    },
+    {
+      "epoch": 0.49296,
+      "grad_norm": 2.504288399764401,
+      "learning_rate": 0.00019070344380668332,
+      "loss": 1.874,
+      "step": 3081
+    },
+    {
+      "epoch": 0.49312,
+      "grad_norm": 2.5518590361732003,
+      "learning_rate": 0.00019069616919331557,
+      "loss": 1.9322,
+      "step": 3082
+    },
+    {
+      "epoch": 0.49328,
+      "grad_norm": 2.6286000863650316,
+      "learning_rate": 0.0001906888918737063,
+      "loss": 1.9803,
+      "step": 3083
+    },
+    {
+      "epoch": 0.49344,
+      "grad_norm": 2.263817327439831,
+      "learning_rate": 0.00019068161184807264,
+      "loss": 1.8143,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4936,
+      "grad_norm": 2.2492170800993794,
+      "learning_rate": 0.00019067432911663179,
+      "loss": 1.8966,
+      "step": 3085
+    },
+    {
+      "epoch": 0.49376,
+      "grad_norm": 2.2818655451481473,
+      "learning_rate": 0.0001906670436796011,
+      "loss": 1.8323,
+      "step": 3086
+    },
+    {
+      "epoch": 0.49392,
+      "grad_norm": 2.500145389137928,
+      "learning_rate": 0.00019065975553719794,
+      "loss": 1.8684,
+      "step": 3087
+    },
+    {
+      "epoch": 0.49408,
+      "grad_norm": 2.2683897410697087,
+      "learning_rate": 0.0001906524646896398,
+      "loss": 1.8851,
+      "step": 3088
+    },
+    {
+      "epoch": 0.49424,
+      "grad_norm": 2.342705375336455,
+      "learning_rate": 0.0001906451711371442,
+      "loss": 2.0626,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4944,
+      "grad_norm": 2.0875420795204263,
+      "learning_rate": 0.00019063787487992877,
+      "loss": 1.849,
+      "step": 3090
+    },
+    {
+      "epoch": 0.49456,
+      "grad_norm": 2.2013815641351524,
+      "learning_rate": 0.00019063057591821122,
+      "loss": 1.8191,
+      "step": 3091
+    },
+    {
+      "epoch": 0.49472,
+      "grad_norm": 2.455410248054929,
+      "learning_rate": 0.00019062327425220933,
+      "loss": 1.9461,
+      "step": 3092
+    },
+    {
+      "epoch": 0.49488,
+      "grad_norm": 2.1238896135671164,
+      "learning_rate": 0.00019061596988214103,
+      "loss": 1.8047,
+      "step": 3093
+    },
+    {
+      "epoch": 0.49504,
+      "grad_norm": 2.9671117059308347,
+      "learning_rate": 0.00019060866280822423,
+      "loss": 1.9371,
+      "step": 3094
+    },
+    {
+      "epoch": 0.4952,
+      "grad_norm": 2.471249907204298,
+      "learning_rate": 0.00019060135303067694,
+      "loss": 1.9047,
+      "step": 3095
+    },
+    {
+      "epoch": 0.49536,
+      "grad_norm": 2.688808756789046,
+      "learning_rate": 0.0001905940405497173,
+      "loss": 1.9,
+      "step": 3096
+    },
+    {
+      "epoch": 0.49552,
+      "grad_norm": 2.4940122783814,
+      "learning_rate": 0.0001905867253655635,
+      "loss": 1.8891,
+      "step": 3097
+    },
+    {
+      "epoch": 0.49568,
+      "grad_norm": 2.3165457923194284,
+      "learning_rate": 0.00019057940747843383,
+      "loss": 1.9858,
+      "step": 3098
+    },
+    {
+      "epoch": 0.49584,
+      "grad_norm": 2.174557566453346,
+      "learning_rate": 0.0001905720868885466,
+      "loss": 1.8353,
+      "step": 3099
+    },
+    {
+      "epoch": 0.496,
+      "grad_norm": 2.1323428105546607,
+      "learning_rate": 0.0001905647635961203,
+      "loss": 1.9668,
+      "step": 3100
+    },
+    {
+      "epoch": 0.49616,
+      "grad_norm": 2.2498380028428944,
+      "learning_rate": 0.0001905574376013734,
+      "loss": 1.8115,
+      "step": 3101
+    },
+    {
+      "epoch": 0.49632,
+      "grad_norm": 2.678837418525421,
+      "learning_rate": 0.00019055010890452453,
+      "loss": 1.9285,
+      "step": 3102
+    },
+    {
+      "epoch": 0.49648,
+      "grad_norm": 2.5280373454969043,
+      "learning_rate": 0.00019054277750579237,
+      "loss": 1.913,
+      "step": 3103
+    },
+    {
+      "epoch": 0.49664,
+      "grad_norm": 2.2389019926418743,
+      "learning_rate": 0.00019053544340539563,
+      "loss": 1.8724,
+      "step": 3104
+    },
+    {
+      "epoch": 0.4968,
+      "grad_norm": 3.0435535376247085,
+      "learning_rate": 0.0001905281066035532,
+      "loss": 1.8366,
+      "step": 3105
+    },
+    {
+      "epoch": 0.49696,
+      "grad_norm": 2.479160852392145,
+      "learning_rate": 0.00019052076710048398,
+      "loss": 1.8877,
+      "step": 3106
+    },
+    {
+      "epoch": 0.49712,
+      "grad_norm": 2.8665274336766084,
+      "learning_rate": 0.00019051342489640697,
+      "loss": 1.7933,
+      "step": 3107
+    },
+    {
+      "epoch": 0.49728,
+      "grad_norm": 2.688059447799018,
+      "learning_rate": 0.00019050607999154129,
+      "loss": 1.87,
+      "step": 3108
+    },
+    {
+      "epoch": 0.49744,
+      "grad_norm": 2.476761322866557,
+      "learning_rate": 0.00019049873238610602,
+      "loss": 1.7676,
+      "step": 3109
+    },
+    {
+      "epoch": 0.4976,
+      "grad_norm": 3.340825408806492,
+      "learning_rate": 0.00019049138208032047,
+      "loss": 1.9374,
+      "step": 3110
+    },
+    {
+      "epoch": 0.49776,
+      "grad_norm": 2.954302903399583,
+      "learning_rate": 0.00019048402907440392,
+      "loss": 2.0015,
+      "step": 3111
+    },
+    {
+      "epoch": 0.49792,
+      "grad_norm": 2.727435445784985,
+      "learning_rate": 0.0001904766733685758,
+      "loss": 1.8559,
+      "step": 3112
+    },
+    {
+      "epoch": 0.49808,
+      "grad_norm": 2.521063977100494,
+      "learning_rate": 0.0001904693149630556,
+      "loss": 1.8648,
+      "step": 3113
+    },
+    {
+      "epoch": 0.49824,
+      "grad_norm": 3.8926495172815097,
+      "learning_rate": 0.00019046195385806285,
+      "loss": 1.9176,
+      "step": 3114
+    },
+    {
+      "epoch": 0.4984,
+      "grad_norm": 2.512426310354057,
+      "learning_rate": 0.0001904545900538172,
+      "loss": 2.0751,
+      "step": 3115
+    },
+    {
+      "epoch": 0.49856,
+      "grad_norm": 2.522448193334787,
+      "learning_rate": 0.00019044722355053843,
+      "loss": 1.9293,
+      "step": 3116
+    },
+    {
+      "epoch": 0.49872,
+      "grad_norm": 2.9326245827241024,
+      "learning_rate": 0.00019043985434844627,
+      "loss": 1.9334,
+      "step": 3117
+    },
+    {
+      "epoch": 0.49888,
+      "grad_norm": 2.675529632454356,
+      "learning_rate": 0.00019043248244776065,
+      "loss": 2.0072,
+      "step": 3118
+    },
+    {
+      "epoch": 0.49904,
+      "grad_norm": 3.2216113233005026,
+      "learning_rate": 0.00019042510784870152,
+      "loss": 1.9524,
+      "step": 3119
+    },
+    {
+      "epoch": 0.4992,
+      "grad_norm": 2.7176732066292657,
+      "learning_rate": 0.00019041773055148898,
+      "loss": 1.798,
+      "step": 3120
+    },
+    {
+      "epoch": 0.49936,
+      "grad_norm": 2.567974370847852,
+      "learning_rate": 0.00019041035055634303,
+      "loss": 1.8967,
+      "step": 3121
+    },
+    {
+      "epoch": 0.49952,
+      "grad_norm": 2.2302733013556586,
+      "learning_rate": 0.00019040296786348404,
+      "loss": 1.9115,
+      "step": 3122
+    },
+    {
+      "epoch": 0.49968,
+      "grad_norm": 2.2302733013556586,
+      "learning_rate": 0.00019040296786348404,
+      "loss": 1.9127,
+      "step": 3123
+    },
+    {
+      "epoch": 0.49984,
+      "grad_norm": 4.181647660318779,
+      "learning_rate": 0.0001903955824731322,
+      "loss": 1.9968,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 2.339239894628699,
+      "learning_rate": 0.00019038819438550787,
+      "loss": 2.0377,
+      "step": 3125
+    },
+    {
+      "epoch": 0.50016,
+      "grad_norm": 3.9189587851168484,
+      "learning_rate": 0.00019038080360083158,
+      "loss": 1.9666,
+      "step": 3126
+    },
+    {
+      "epoch": 0.50032,
+      "grad_norm": 2.446279962781154,
+      "learning_rate": 0.00019037341011932376,
+      "loss": 2.0246,
+      "step": 3127
+    },
+    {
+      "epoch": 0.50048,
+      "grad_norm": 2.5338424005211553,
+      "learning_rate": 0.00019036601394120507,
+      "loss": 1.8258,
+      "step": 3128
+    },
+    {
+      "epoch": 0.50064,
+      "grad_norm": 2.588320425545139,
+      "learning_rate": 0.00019035861506669623,
+      "loss": 1.9086,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5008,
+      "grad_norm": 3.049636003827965,
+      "learning_rate": 0.00019035121349601796,
+      "loss": 1.9039,
+      "step": 3130
+    },
+    {
+      "epoch": 0.50096,
+      "grad_norm": 2.3981578604838667,
+      "learning_rate": 0.00019034380922939117,
+      "loss": 1.7121,
+      "step": 3131
+    },
+    {
+      "epoch": 0.50112,
+      "grad_norm": 2.8291520796498357,
+      "learning_rate": 0.00019033640226703673,
+      "loss": 1.891,
+      "step": 3132
+    },
+    {
+      "epoch": 0.50128,
+      "grad_norm": 2.3945961916615057,
+      "learning_rate": 0.00019032899260917568,
+      "loss": 1.9612,
+      "step": 3133
+    },
+    {
+      "epoch": 0.50144,
+      "grad_norm": 3.284331005567369,
+      "learning_rate": 0.00019032158025602912,
+      "loss": 1.8762,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5016,
+      "grad_norm": 2.8666699273895713,
+      "learning_rate": 0.0001903141652078182,
+      "loss": 1.9337,
+      "step": 3135
+    },
+    {
+      "epoch": 0.50176,
+      "grad_norm": 2.474232760149717,
+      "learning_rate": 0.00019030674746476421,
+      "loss": 1.9605,
+      "step": 3136
+    },
+    {
+      "epoch": 0.50192,
+      "grad_norm": 2.2366560755634683,
+      "learning_rate": 0.0001902993270270885,
+      "loss": 1.9082,
+      "step": 3137
+    },
+    {
+      "epoch": 0.50208,
+      "grad_norm": 3.0487459651713626,
+      "learning_rate": 0.0001902919038950124,
+      "loss": 1.9283,
+      "step": 3138
+    },
+    {
+      "epoch": 0.50224,
+      "grad_norm": 2.2720472867263037,
+      "learning_rate": 0.00019028447806875747,
+      "loss": 1.9051,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5024,
+      "grad_norm": 2.4567930727108904,
+      "learning_rate": 0.00019027704954854526,
+      "loss": 1.9973,
+      "step": 3140
+    },
+    {
+      "epoch": 0.50256,
+      "grad_norm": 3.145379696276254,
+      "learning_rate": 0.00019026961833459746,
+      "loss": 1.9548,
+      "step": 3141
+    },
+    {
+      "epoch": 0.50272,
+      "grad_norm": 2.3367144516442546,
+      "learning_rate": 0.00019026218442713577,
+      "loss": 1.8416,
+      "step": 3142
+    },
+    {
+      "epoch": 0.50288,
+      "grad_norm": 2.160706728912866,
+      "learning_rate": 0.00019025474782638203,
+      "loss": 1.826,
+      "step": 3143
+    },
+    {
+      "epoch": 0.50304,
+      "grad_norm": 3.3458761632302982,
+      "learning_rate": 0.00019024730853255814,
+      "loss": 1.9581,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5032,
+      "grad_norm": 3.7289142124421244,
+      "learning_rate": 0.00019023986654588608,
+      "loss": 1.8849,
+      "step": 3145
+    },
+    {
+      "epoch": 0.50336,
+      "grad_norm": 2.488187534686263,
+      "learning_rate": 0.00019023242186658786,
+      "loss": 1.924,
+      "step": 3146
+    },
+    {
+      "epoch": 0.50352,
+      "grad_norm": 2.806295863376601,
+      "learning_rate": 0.00019022497449488568,
+      "loss": 1.9653,
+      "step": 3147
+    },
+    {
+      "epoch": 0.50368,
+      "grad_norm": 2.690917380479888,
+      "learning_rate": 0.00019021752443100173,
+      "loss": 1.8727,
+      "step": 3148
+    },
+    {
+      "epoch": 0.50384,
+      "grad_norm": 2.916619091822501,
+      "learning_rate": 0.0001902100716751583,
+      "loss": 1.8542,
+      "step": 3149
+    },
+    {
+      "epoch": 0.504,
+      "grad_norm": 2.616313691086378,
+      "learning_rate": 0.00019020261622757777,
+      "loss": 1.929,
+      "step": 3150
+    },
+    {
+      "epoch": 0.50416,
+      "grad_norm": 2.2483098123620895,
+      "learning_rate": 0.00019019515808848264,
+      "loss": 1.9257,
+      "step": 3151
+    },
+    {
+      "epoch": 0.50432,
+      "grad_norm": 2.4283966312174,
+      "learning_rate": 0.00019018769725809542,
+      "loss": 1.8543,
+      "step": 3152
+    },
+    {
+      "epoch": 0.50448,
+      "grad_norm": 2.643071707413299,
+      "learning_rate": 0.0001901802337366387,
+      "loss": 1.8955,
+      "step": 3153
+    },
+    {
+      "epoch": 0.50464,
+      "grad_norm": 2.3218542444643675,
+      "learning_rate": 0.00019017276752433525,
+      "loss": 1.8612,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5048,
+      "grad_norm": 3.368680581908801,
+      "learning_rate": 0.0001901652986214078,
+      "loss": 1.9687,
+      "step": 3155
+    },
+    {
+      "epoch": 0.50496,
+      "grad_norm": 3.179680913524329,
+      "learning_rate": 0.00019015782702807917,
+      "loss": 1.8171,
+      "step": 3156
+    },
+    {
+      "epoch": 0.50512,
+      "grad_norm": 2.5273226816969996,
+      "learning_rate": 0.00019015035274457242,
+      "loss": 1.9211,
+      "step": 3157
+    },
+    {
+      "epoch": 0.50528,
+      "grad_norm": 2.5659030749068394,
+      "learning_rate": 0.0001901428757711105,
+      "loss": 1.8694,
+      "step": 3158
+    },
+    {
+      "epoch": 0.50544,
+      "grad_norm": 2.5659030749068394,
+      "learning_rate": 0.0001901428757711105,
+      "loss": 1.9413,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5056,
+      "grad_norm": 3.2272051394289205,
+      "learning_rate": 0.0001901353961079165,
+      "loss": 1.922,
+      "step": 3160
+    },
+    {
+      "epoch": 0.50576,
+      "grad_norm": 2.1036737938556573,
+      "learning_rate": 0.00019012791375521364,
+      "loss": 1.8103,
+      "step": 3161
+    },
+    {
+      "epoch": 0.50592,
+      "grad_norm": 2.925008734565273,
+      "learning_rate": 0.00019012042871322516,
+      "loss": 1.9581,
+      "step": 3162
+    },
+    {
+      "epoch": 0.50608,
+      "grad_norm": 2.5284549777349365,
+      "learning_rate": 0.00019011294098217438,
+      "loss": 1.8295,
+      "step": 3163
+    },
+    {
+      "epoch": 0.50624,
+      "grad_norm": 2.3284911286690275,
+      "learning_rate": 0.00019010545056228478,
+      "loss": 1.9034,
+      "step": 3164
+    },
+    {
+      "epoch": 0.5064,
+      "grad_norm": 2.3230464929624595,
+      "learning_rate": 0.00019009795745377983,
+      "loss": 1.8747,
+      "step": 3165
+    },
+    {
+      "epoch": 0.50656,
+      "grad_norm": 2.4879189474817167,
+      "learning_rate": 0.0001900904616568831,
+      "loss": 1.8141,
+      "step": 3166
+    },
+    {
+      "epoch": 0.50672,
+      "grad_norm": 2.468673589753658,
+      "learning_rate": 0.0001900829631718183,
+      "loss": 1.8616,
+      "step": 3167
+    },
+    {
+      "epoch": 0.50688,
+      "grad_norm": 2.558972065857889,
+      "learning_rate": 0.0001900754619988091,
+      "loss": 1.8247,
+      "step": 3168
+    },
+    {
+      "epoch": 0.50704,
+      "grad_norm": 3.04360747992521,
+      "learning_rate": 0.0001900679581380794,
+      "loss": 1.945,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5072,
+      "grad_norm": 2.5110721848964985,
+      "learning_rate": 0.00019006045158985306,
+      "loss": 1.9462,
+      "step": 3170
+    },
+    {
+      "epoch": 0.50736,
+      "grad_norm": 2.3526552005455117,
+      "learning_rate": 0.0001900529423543541,
+      "loss": 1.9092,
+      "step": 3171
+    },
+    {
+      "epoch": 0.50752,
+      "grad_norm": 2.6313894981346326,
+      "learning_rate": 0.0001900454304318065,
+      "loss": 1.969,
+      "step": 3172
+    },
+    {
+      "epoch": 0.50768,
+      "grad_norm": 2.239637200303824,
+      "learning_rate": 0.00019003791582243454,
+      "loss": 1.844,
+      "step": 3173
+    },
+    {
+      "epoch": 0.50784,
+      "grad_norm": 2.321908475066396,
+      "learning_rate": 0.00019003039852646231,
+      "loss": 1.876,
+      "step": 3174
+    },
+    {
+      "epoch": 0.508,
+      "grad_norm": 2.8005046111087637,
+      "learning_rate": 0.00019002287854411423,
+      "loss": 1.8878,
+      "step": 3175
+    },
+    {
+      "epoch": 0.50816,
+      "grad_norm": 2.4501499264804165,
+      "learning_rate": 0.0001900153558756146,
+      "loss": 1.8521,
+      "step": 3176
+    },
+    {
+      "epoch": 0.50832,
+      "grad_norm": 3.052086385185825,
+      "learning_rate": 0.0001900078305211879,
+      "loss": 1.9436,
+      "step": 3177
+    },
+    {
+      "epoch": 0.50848,
+      "grad_norm": 2.197642211345802,
+      "learning_rate": 0.00019000030248105874,
+      "loss": 1.8498,
+      "step": 3178
+    },
+    {
+      "epoch": 0.50864,
+      "grad_norm": 2.430330163028931,
+      "learning_rate": 0.00018999277175545167,
+      "loss": 1.9435,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5088,
+      "grad_norm": 2.729696654673811,
+      "learning_rate": 0.00018998523834459142,
+      "loss": 1.9346,
+      "step": 3180
+    },
+    {
+      "epoch": 0.50896,
+      "grad_norm": 2.5928987701412196,
+      "learning_rate": 0.00018997770224870277,
+      "loss": 1.9109,
+      "step": 3181
+    },
+    {
+      "epoch": 0.50912,
+      "grad_norm": 2.3524324091005826,
+      "learning_rate": 0.0001899701634680106,
+      "loss": 1.8177,
+      "step": 3182
+    },
+    {
+      "epoch": 0.50928,
+      "grad_norm": 2.305147995760629,
+      "learning_rate": 0.00018996262200273988,
+      "loss": 1.8421,
+      "step": 3183
+    },
+    {
+      "epoch": 0.50944,
+      "grad_norm": 2.416380688324242,
+      "learning_rate": 0.0001899550778531156,
+      "loss": 1.8255,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5096,
+      "grad_norm": 2.3610846252527313,
+      "learning_rate": 0.00018994753101936286,
+      "loss": 1.8365,
+      "step": 3185
+    },
+    {
+      "epoch": 0.50976,
+      "grad_norm": 2.853920330544436,
+      "learning_rate": 0.0001899399815017069,
+      "loss": 1.8882,
+      "step": 3186
+    },
+    {
+      "epoch": 0.50992,
+      "grad_norm": 2.278377045030013,
+      "learning_rate": 0.0001899324293003729,
+      "loss": 1.918,
+      "step": 3187
+    },
+    {
+      "epoch": 0.51008,
+      "grad_norm": 2.165713876065681,
+      "learning_rate": 0.00018992487441558628,
+      "loss": 2.026,
+      "step": 3188
+    },
+    {
+      "epoch": 0.51024,
+      "grad_norm": 2.6378646112908304,
+      "learning_rate": 0.00018991731684757244,
+      "loss": 1.8632,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5104,
+      "grad_norm": 2.2424630879643335,
+      "learning_rate": 0.00018990975659655688,
+      "loss": 1.8641,
+      "step": 3190
+    },
+    {
+      "epoch": 0.51056,
+      "grad_norm": 2.2063043968024454,
+      "learning_rate": 0.00018990219366276517,
+      "loss": 1.9045,
+      "step": 3191
+    },
+    {
+      "epoch": 0.51072,
+      "grad_norm": 2.574299472213771,
+      "learning_rate": 0.00018989462804642305,
+      "loss": 1.8576,
+      "step": 3192
+    },
+    {
+      "epoch": 0.51088,
+      "grad_norm": 2.819404138138879,
+      "learning_rate": 0.00018988705974775621,
+      "loss": 1.9454,
+      "step": 3193
+    },
+    {
+      "epoch": 0.51104,
+      "grad_norm": 2.660066633377504,
+      "learning_rate": 0.00018987948876699048,
+      "loss": 1.9319,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5112,
+      "grad_norm": 2.8200662082197936,
+      "learning_rate": 0.00018987191510435176,
+      "loss": 1.8917,
+      "step": 3195
+    },
+    {
+      "epoch": 0.51136,
+      "grad_norm": 3.117021113441628,
+      "learning_rate": 0.0001898643387600661,
+      "loss": 1.85,
+      "step": 3196
+    },
+    {
+      "epoch": 0.51152,
+      "grad_norm": 2.285626215920963,
+      "learning_rate": 0.00018985675973435945,
+      "loss": 1.9982,
+      "step": 3197
+    },
+    {
+      "epoch": 0.51168,
+      "grad_norm": 2.290989208264769,
+      "learning_rate": 0.00018984917802745806,
+      "loss": 1.9373,
+      "step": 3198
+    },
+    {
+      "epoch": 0.51184,
+      "grad_norm": 2.4150318962720947,
+      "learning_rate": 0.00018984159363958812,
+      "loss": 1.9543,
+      "step": 3199
+    },
+    {
+      "epoch": 0.512,
+      "grad_norm": 2.4453184127307663,
+      "learning_rate": 0.0001898340065709759,
+      "loss": 1.83,
+      "step": 3200
+    },
+    {
+      "epoch": 0.51216,
+      "grad_norm": 2.260440931404344,
+      "learning_rate": 0.0001898264168218479,
+      "loss": 1.9028,
+      "step": 3201
+    },
+    {
+      "epoch": 0.51232,
+      "grad_norm": 2.2991927737921096,
+      "learning_rate": 0.00018981882439243045,
+      "loss": 1.8479,
+      "step": 3202
+    },
+    {
+      "epoch": 0.51248,
+      "grad_norm": 2.7859248789724487,
+      "learning_rate": 0.0001898112292829502,
+      "loss": 1.9044,
+      "step": 3203
+    },
+    {
+      "epoch": 0.51264,
+      "grad_norm": 2.3484178611644837,
+      "learning_rate": 0.00018980363149363375,
+      "loss": 1.8238,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5128,
+      "grad_norm": 3.1419372730515414,
+      "learning_rate": 0.00018979603102470773,
+      "loss": 1.9163,
+      "step": 3205
+    },
+    {
+      "epoch": 0.51296,
+      "grad_norm": 2.7124894035901885,
+      "learning_rate": 0.00018978842787639903,
+      "loss": 1.8666,
+      "step": 3206
+    },
+    {
+      "epoch": 0.51312,
+      "grad_norm": 2.88513506041206,
+      "learning_rate": 0.00018978082204893447,
+      "loss": 1.9247,
+      "step": 3207
+    },
+    {
+      "epoch": 0.51328,
+      "grad_norm": 2.4025326220503245,
+      "learning_rate": 0.000189773213542541,
+      "loss": 1.7417,
+      "step": 3208
+    },
+    {
+      "epoch": 0.51344,
+      "grad_norm": 2.8135288376877337,
+      "learning_rate": 0.00018976560235744566,
+      "loss": 1.8575,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5136,
+      "grad_norm": 2.6961900995303787,
+      "learning_rate": 0.00018975798849387557,
+      "loss": 1.9017,
+      "step": 3210
+    },
+    {
+      "epoch": 0.51376,
+      "grad_norm": 3.042424891134434,
+      "learning_rate": 0.00018975037195205786,
+      "loss": 1.8855,
+      "step": 3211
+    },
+    {
+      "epoch": 0.51392,
+      "grad_norm": 2.3153479935755272,
+      "learning_rate": 0.00018974275273221983,
+      "loss": 1.8543,
+      "step": 3212
+    },
+    {
+      "epoch": 0.51408,
+      "grad_norm": 2.4182186939741785,
+      "learning_rate": 0.00018973513083458883,
+      "loss": 1.9397,
+      "step": 3213
+    },
+    {
+      "epoch": 0.51424,
+      "grad_norm": 2.610701231035211,
+      "learning_rate": 0.0001897275062593923,
+      "loss": 1.8211,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5144,
+      "grad_norm": 2.5985680898191097,
+      "learning_rate": 0.00018971987900685776,
+      "loss": 1.936,
+      "step": 3215
+    },
+    {
+      "epoch": 0.51456,
+      "grad_norm": 2.558159632574935,
+      "learning_rate": 0.0001897122490772127,
+      "loss": 1.8298,
+      "step": 3216
+    },
+    {
+      "epoch": 0.51472,
+      "grad_norm": 2.2732283023438358,
+      "learning_rate": 0.00018970461647068492,
+      "loss": 1.9669,
+      "step": 3217
+    },
+    {
+      "epoch": 0.51488,
+      "grad_norm": 2.3872371572882485,
+      "learning_rate": 0.00018969698118750206,
+      "loss": 1.8753,
+      "step": 3218
+    },
+    {
+      "epoch": 0.51504,
+      "grad_norm": 2.34024803441212,
+      "learning_rate": 0.00018968934322789198,
+      "loss": 1.8237,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5152,
+      "grad_norm": 2.4659954496426337,
+      "learning_rate": 0.00018968170259208258,
+      "loss": 1.8071,
+      "step": 3220
+    },
+    {
+      "epoch": 0.51536,
+      "grad_norm": 1.983485720734264,
+      "learning_rate": 0.00018967405928030186,
+      "loss": 1.8105,
+      "step": 3221
+    },
+    {
+      "epoch": 0.51552,
+      "grad_norm": 1.9626371618868732,
+      "learning_rate": 0.0001896664132927779,
+      "loss": 1.9778,
+      "step": 3222
+    },
+    {
+      "epoch": 0.51568,
+      "grad_norm": 2.5231137711605918,
+      "learning_rate": 0.0001896587646297388,
+      "loss": 1.8469,
+      "step": 3223
+    },
+    {
+      "epoch": 0.51584,
+      "grad_norm": 2.9903924127145673,
+      "learning_rate": 0.00018965111329141284,
+      "loss": 1.9287,
+      "step": 3224
+    },
+    {
+      "epoch": 0.516,
+      "grad_norm": 2.2888192366916247,
+      "learning_rate": 0.00018964345927802826,
+      "loss": 1.8558,
+      "step": 3225
+    },
+    {
+      "epoch": 0.51616,
+      "grad_norm": 2.297902425458101,
+      "learning_rate": 0.0001896358025898135,
+      "loss": 1.8534,
+      "step": 3226
+    },
+    {
+      "epoch": 0.51632,
+      "grad_norm": 2.2617399975405768,
+      "learning_rate": 0.00018962814322699703,
+      "loss": 1.8408,
+      "step": 3227
+    },
+    {
+      "epoch": 0.51648,
+      "grad_norm": 2.502232214080828,
+      "learning_rate": 0.00018962048118980734,
+      "loss": 1.8383,
+      "step": 3228
+    },
+    {
+      "epoch": 0.51664,
+      "grad_norm": 2.2920665283209023,
+      "learning_rate": 0.00018961281647847307,
+      "loss": 1.8735,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5168,
+      "grad_norm": 2.873526253520362,
+      "learning_rate": 0.00018960514909322295,
+      "loss": 1.7919,
+      "step": 3230
+    },
+    {
+      "epoch": 0.51696,
+      "grad_norm": 2.5597548581100225,
+      "learning_rate": 0.00018959747903428572,
+      "loss": 1.8735,
+      "step": 3231
+    },
+    {
+      "epoch": 0.51712,
+      "grad_norm": 2.3162288557765915,
+      "learning_rate": 0.00018958980630189032,
+      "loss": 1.7341,
+      "step": 3232
+    },
+    {
+      "epoch": 0.51728,
+      "grad_norm": 2.390001484356675,
+      "learning_rate": 0.00018958213089626563,
+      "loss": 1.9583,
+      "step": 3233
+    },
+    {
+      "epoch": 0.51744,
+      "grad_norm": 2.4286212787037664,
+      "learning_rate": 0.00018957445281764067,
+      "loss": 1.913,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5176,
+      "grad_norm": 2.2633967276926814,
+      "learning_rate": 0.00018956677206624458,
+      "loss": 1.8718,
+      "step": 3235
+    },
+    {
+      "epoch": 0.51776,
+      "grad_norm": 3.3266287379360855,
+      "learning_rate": 0.00018955908864230653,
+      "loss": 1.8815,
+      "step": 3236
+    },
+    {
+      "epoch": 0.51792,
+      "grad_norm": 2.3552521291241892,
+      "learning_rate": 0.00018955140254605578,
+      "loss": 1.7935,
+      "step": 3237
+    },
+    {
+      "epoch": 0.51808,
+      "grad_norm": 2.0513774287242983,
+      "learning_rate": 0.00018954371377772163,
+      "loss": 1.763,
+      "step": 3238
+    },
+    {
+      "epoch": 0.51824,
+      "grad_norm": 2.5947326521269005,
+      "learning_rate": 0.0001895360223375336,
+      "loss": 1.9374,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5184,
+      "grad_norm": 2.2471125084103645,
+      "learning_rate": 0.00018952832822572108,
+      "loss": 1.7684,
+      "step": 3240
+    },
+    {
+      "epoch": 0.51856,
+      "grad_norm": 2.919932922012266,
+      "learning_rate": 0.00018952063144251372,
+      "loss": 2.0456,
+      "step": 3241
+    },
+    {
+      "epoch": 0.51872,
+      "grad_norm": 2.5432322878719034,
+      "learning_rate": 0.00018951293198814115,
+      "loss": 1.7607,
+      "step": 3242
+    },
+    {
+      "epoch": 0.51888,
+      "grad_norm": 2.578028293781798,
+      "learning_rate": 0.00018950522986283314,
+      "loss": 1.7981,
+      "step": 3243
+    },
+    {
+      "epoch": 0.51904,
+      "grad_norm": 3.350441904583565,
+      "learning_rate": 0.00018949752506681948,
+      "loss": 1.9298,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5192,
+      "grad_norm": 2.8136799252983056,
+      "learning_rate": 0.00018948981760033012,
+      "loss": 1.7881,
+      "step": 3245
+    },
+    {
+      "epoch": 0.51936,
+      "grad_norm": 2.4683272011595316,
+      "learning_rate": 0.00018948210746359495,
+      "loss": 1.9223,
+      "step": 3246
+    },
+    {
+      "epoch": 0.51952,
+      "grad_norm": 2.0799921205755214,
+      "learning_rate": 0.00018947439465684412,
+      "loss": 1.8984,
+      "step": 3247
+    },
+    {
+      "epoch": 0.51968,
+      "grad_norm": 2.6966439373770483,
+      "learning_rate": 0.00018946667918030773,
+      "loss": 1.9397,
+      "step": 3248
+    },
+    {
+      "epoch": 0.51984,
+      "grad_norm": 2.797817787599836,
+      "learning_rate": 0.00018945896103421598,
+      "loss": 1.8119,
+      "step": 3249
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 3.5105588219292008,
+      "learning_rate": 0.0001894512402187992,
+      "loss": 1.9238,
+      "step": 3250
+    },
+    {
+      "epoch": 0.52016,
+      "grad_norm": 3.053030388870718,
+      "learning_rate": 0.00018944351673428776,
+      "loss": 1.8946,
+      "step": 3251
+    },
+    {
+      "epoch": 0.52032,
+      "grad_norm": 2.865539971585009,
+      "learning_rate": 0.0001894357905809121,
+      "loss": 1.8501,
+      "step": 3252
+    },
+    {
+      "epoch": 0.52048,
+      "grad_norm": 3.5098557184036574,
+      "learning_rate": 0.00018942806175890276,
+      "loss": 1.8837,
+      "step": 3253
+    },
+    {
+      "epoch": 0.52064,
+      "grad_norm": 2.4295064380839846,
+      "learning_rate": 0.00018942033026849038,
+      "loss": 1.9092,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5208,
+      "grad_norm": 2.6277184344775475,
+      "learning_rate": 0.00018941259610990568,
+      "loss": 1.8142,
+      "step": 3255
+    },
+    {
+      "epoch": 0.52096,
+      "grad_norm": 3.109377303585755,
+      "learning_rate": 0.0001894048592833794,
+      "loss": 2.0151,
+      "step": 3256
+    },
+    {
+      "epoch": 0.52112,
+      "grad_norm": 2.6284071897765453,
+      "learning_rate": 0.00018939711978914234,
+      "loss": 1.8105,
+      "step": 3257
+    },
+    {
+      "epoch": 0.52128,
+      "grad_norm": 2.482257703560573,
+      "learning_rate": 0.00018938937762742552,
+      "loss": 1.8881,
+      "step": 3258
+    },
+    {
+      "epoch": 0.52144,
+      "grad_norm": 2.7241284695023062,
+      "learning_rate": 0.00018938163279845992,
+      "loss": 1.9406,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5216,
+      "grad_norm": 2.9457511643964462,
+      "learning_rate": 0.00018937388530247665,
+      "loss": 1.8849,
+      "step": 3260
+    },
+    {
+      "epoch": 0.52176,
+      "grad_norm": 2.872944063110346,
+      "learning_rate": 0.00018936613513970685,
+      "loss": 1.913,
+      "step": 3261
+    },
+    {
+      "epoch": 0.52192,
+      "grad_norm": 2.6330713032052038,
+      "learning_rate": 0.00018935838231038182,
+      "loss": 1.8613,
+      "step": 3262
+    },
+    {
+      "epoch": 0.52208,
+      "grad_norm": 2.3817851388203612,
+      "learning_rate": 0.00018935062681473287,
+      "loss": 1.7661,
+      "step": 3263
+    },
+    {
+      "epoch": 0.52224,
+      "grad_norm": 2.5320184587442554,
+      "learning_rate": 0.0001893428686529914,
+      "loss": 1.8221,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5224,
+      "grad_norm": 2.4454492207350476,
+      "learning_rate": 0.00018933510782538892,
+      "loss": 2.0059,
+      "step": 3265
+    },
+    {
+      "epoch": 0.52256,
+      "grad_norm": 2.5958955710832012,
+      "learning_rate": 0.000189327344332157,
+      "loss": 1.7837,
+      "step": 3266
+    },
+    {
+      "epoch": 0.52272,
+      "grad_norm": 2.093470674572211,
+      "learning_rate": 0.0001893195781735273,
+      "loss": 1.7619,
+      "step": 3267
+    },
+    {
+      "epoch": 0.52288,
+      "grad_norm": 2.3857514752055162,
+      "learning_rate": 0.00018931180934973155,
+      "loss": 1.8821,
+      "step": 3268
+    },
+    {
+      "epoch": 0.52304,
+      "grad_norm": 2.499062112754082,
+      "learning_rate": 0.00018930403786100153,
+      "loss": 1.8525,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5232,
+      "grad_norm": 2.330481598355447,
+      "learning_rate": 0.00018929626370756914,
+      "loss": 1.8861,
+      "step": 3270
+    },
+    {
+      "epoch": 0.52336,
+      "grad_norm": 2.679613296576877,
+      "learning_rate": 0.00018928848688966638,
+      "loss": 1.9289,
+      "step": 3271
+    },
+    {
+      "epoch": 0.52352,
+      "grad_norm": 2.359971948066754,
+      "learning_rate": 0.00018928070740752525,
+      "loss": 1.7833,
+      "step": 3272
+    },
+    {
+      "epoch": 0.52368,
+      "grad_norm": 2.9739895925257622,
+      "learning_rate": 0.00018927292526137795,
+      "loss": 1.8922,
+      "step": 3273
+    },
+    {
+      "epoch": 0.52384,
+      "grad_norm": 2.244747310457343,
+      "learning_rate": 0.00018926514045145663,
+      "loss": 2.0385,
+      "step": 3274
+    },
+    {
+      "epoch": 0.524,
+      "grad_norm": 2.613255539835434,
+      "learning_rate": 0.00018925735297799356,
+      "loss": 1.8827,
+      "step": 3275
+    },
+    {
+      "epoch": 0.52416,
+      "grad_norm": 2.5617700453346184,
+      "learning_rate": 0.00018924956284122114,
+      "loss": 1.8943,
+      "step": 3276
+    },
+    {
+      "epoch": 0.52432,
+      "grad_norm": 2.157581418896464,
+      "learning_rate": 0.00018924177004137188,
+      "loss": 1.8515,
+      "step": 3277
+    },
+    {
+      "epoch": 0.52448,
+      "grad_norm": 2.3088093563810252,
+      "learning_rate": 0.00018923397457867817,
+      "loss": 1.8162,
+      "step": 3278
+    },
+    {
+      "epoch": 0.52464,
+      "grad_norm": 2.383400659176987,
+      "learning_rate": 0.0001892261764533727,
+      "loss": 1.7597,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5248,
+      "grad_norm": 2.719816765539249,
+      "learning_rate": 0.00018921837566568818,
+      "loss": 2.0129,
+      "step": 3280
+    },
+    {
+      "epoch": 0.52496,
+      "grad_norm": 2.7692541821609944,
+      "learning_rate": 0.00018921057221585728,
+      "loss": 1.8849,
+      "step": 3281
+    },
+    {
+      "epoch": 0.52512,
+      "grad_norm": 2.289899110215865,
+      "learning_rate": 0.00018920276610411295,
+      "loss": 1.8333,
+      "step": 3282
+    },
+    {
+      "epoch": 0.52528,
+      "grad_norm": 2.506861077914194,
+      "learning_rate": 0.00018919495733068804,
+      "loss": 1.901,
+      "step": 3283
+    },
+    {
+      "epoch": 0.52544,
+      "grad_norm": 2.6865401919537373,
+      "learning_rate": 0.00018918714589581556,
+      "loss": 1.8723,
+      "step": 3284
+    },
+    {
+      "epoch": 0.5256,
+      "grad_norm": 2.6639023205708767,
+      "learning_rate": 0.00018917933179972863,
+      "loss": 1.821,
+      "step": 3285
+    },
+    {
+      "epoch": 0.52576,
+      "grad_norm": 2.817139770451563,
+      "learning_rate": 0.00018917151504266037,
+      "loss": 1.8642,
+      "step": 3286
+    },
+    {
+      "epoch": 0.52592,
+      "grad_norm": 2.2036084287424336,
+      "learning_rate": 0.00018916369562484405,
+      "loss": 1.8852,
+      "step": 3287
+    },
+    {
+      "epoch": 0.52608,
+      "grad_norm": 2.1724530740260533,
+      "learning_rate": 0.00018915587354651297,
+      "loss": 1.7188,
+      "step": 3288
+    },
+    {
+      "epoch": 0.52624,
+      "grad_norm": 2.0903118754962065,
+      "learning_rate": 0.00018914804880790052,
+      "loss": 1.8258,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5264,
+      "grad_norm": 2.452459688725725,
+      "learning_rate": 0.00018914022140924025,
+      "loss": 1.934,
+      "step": 3290
+    },
+    {
+      "epoch": 0.52656,
+      "grad_norm": 2.4480285730545854,
+      "learning_rate": 0.00018913239135076563,
+      "loss": 1.8403,
+      "step": 3291
+    },
+    {
+      "epoch": 0.52672,
+      "grad_norm": 2.694699649496421,
+      "learning_rate": 0.00018912455863271034,
+      "loss": 1.9698,
+      "step": 3292
+    },
+    {
+      "epoch": 0.52688,
+      "grad_norm": 2.2486370671400264,
+      "learning_rate": 0.00018911672325530807,
+      "loss": 1.8816,
+      "step": 3293
+    },
+    {
+      "epoch": 0.52704,
+      "grad_norm": 2.397368678417437,
+      "learning_rate": 0.00018910888521879264,
+      "loss": 1.81,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5272,
+      "grad_norm": 2.666379366241654,
+      "learning_rate": 0.0001891010445233979,
+      "loss": 1.8538,
+      "step": 3295
+    },
+    {
+      "epoch": 0.52736,
+      "grad_norm": 2.281667800711112,
+      "learning_rate": 0.00018909320116935788,
+      "loss": 1.8113,
+      "step": 3296
+    },
+    {
+      "epoch": 0.52752,
+      "grad_norm": 2.328094708054904,
+      "learning_rate": 0.0001890853551569065,
+      "loss": 1.8924,
+      "step": 3297
+    },
+    {
+      "epoch": 0.52768,
+      "grad_norm": 2.724552829964935,
+      "learning_rate": 0.00018907750648627797,
+      "loss": 1.9277,
+      "step": 3298
+    },
+    {
+      "epoch": 0.52784,
+      "grad_norm": 2.8005049025726603,
+      "learning_rate": 0.00018906965515770645,
+      "loss": 1.9678,
+      "step": 3299
+    },
+    {
+      "epoch": 0.528,
+      "grad_norm": 2.7923175234903694,
+      "learning_rate": 0.0001890618011714262,
+      "loss": 1.8763,
+      "step": 3300
+    },
+    {
+      "epoch": 0.52816,
+      "grad_norm": 2.455643465419677,
+      "learning_rate": 0.00018905394452767158,
+      "loss": 1.8993,
+      "step": 3301
+    },
+    {
+      "epoch": 0.52832,
+      "grad_norm": 2.745434717768068,
+      "learning_rate": 0.000189046085226677,
+      "loss": 1.9468,
+      "step": 3302
+    },
+    {
+      "epoch": 0.52848,
+      "grad_norm": 2.265232185005679,
+      "learning_rate": 0.00018903822326867703,
+      "loss": 1.8096,
+      "step": 3303
+    },
+    {
+      "epoch": 0.52864,
+      "grad_norm": 2.426097459565744,
+      "learning_rate": 0.0001890303586539062,
+      "loss": 1.8962,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5288,
+      "grad_norm": 2.5388690428554086,
+      "learning_rate": 0.00018902249138259917,
+      "loss": 1.8413,
+      "step": 3305
+    },
+    {
+      "epoch": 0.52896,
+      "grad_norm": 2.2798899605343315,
+      "learning_rate": 0.00018901462145499076,
+      "loss": 1.8129,
+      "step": 3306
+    },
+    {
+      "epoch": 0.52912,
+      "grad_norm": 3.054908077469524,
+      "learning_rate": 0.00018900674887131573,
+      "loss": 1.8526,
+      "step": 3307
+    },
+    {
+      "epoch": 0.52928,
+      "grad_norm": 2.541800890895723,
+      "learning_rate": 0.00018899887363180905,
+      "loss": 1.865,
+      "step": 3308
+    },
+    {
+      "epoch": 0.52944,
+      "grad_norm": 2.1971628001710863,
+      "learning_rate": 0.00018899099573670562,
+      "loss": 1.8717,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5296,
+      "grad_norm": 1.963804512067701,
+      "learning_rate": 0.00018898311518624055,
+      "loss": 1.8256,
+      "step": 3310
+    },
+    {
+      "epoch": 0.52976,
+      "grad_norm": 2.5191643022603714,
+      "learning_rate": 0.000188975231980649,
+      "loss": 1.8733,
+      "step": 3311
+    },
+    {
+      "epoch": 0.52992,
+      "grad_norm": 2.189209995746007,
+      "learning_rate": 0.00018896734612016617,
+      "loss": 1.9224,
+      "step": 3312
+    },
+    {
+      "epoch": 0.53008,
+      "grad_norm": 2.6186588367360852,
+      "learning_rate": 0.0001889594576050274,
+      "loss": 1.9728,
+      "step": 3313
+    },
+    {
+      "epoch": 0.53024,
+      "grad_norm": 2.5400535843361465,
+      "learning_rate": 0.000188951566435468,
+      "loss": 1.8982,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5304,
+      "grad_norm": 2.2034004462810497,
+      "learning_rate": 0.00018894367261172352,
+      "loss": 1.8146,
+      "step": 3315
+    },
+    {
+      "epoch": 0.53056,
+      "grad_norm": 2.3446912504540554,
+      "learning_rate": 0.0001889357761340294,
+      "loss": 1.8859,
+      "step": 3316
+    },
+    {
+      "epoch": 0.53072,
+      "grad_norm": 1.958088615987502,
+      "learning_rate": 0.00018892787700262135,
+      "loss": 1.8006,
+      "step": 3317
+    },
+    {
+      "epoch": 0.53088,
+      "grad_norm": 2.382558070133807,
+      "learning_rate": 0.00018891997521773503,
+      "loss": 1.8892,
+      "step": 3318
+    },
+    {
+      "epoch": 0.53104,
+      "grad_norm": 2.269713680495375,
+      "learning_rate": 0.00018891207077960623,
+      "loss": 1.8511,
+      "step": 3319
+    },
+    {
+      "epoch": 0.5312,
+      "grad_norm": 2.6991493276917726,
+      "learning_rate": 0.00018890416368847076,
+      "loss": 1.8799,
+      "step": 3320
+    },
+    {
+      "epoch": 0.53136,
+      "grad_norm": 2.2211596691377253,
+      "learning_rate": 0.00018889625394456462,
+      "loss": 1.8538,
+      "step": 3321
+    },
+    {
+      "epoch": 0.53152,
+      "grad_norm": 2.4969015297987935,
+      "learning_rate": 0.00018888834154812378,
+      "loss": 1.9664,
+      "step": 3322
+    },
+    {
+      "epoch": 0.53168,
+      "grad_norm": 2.5810057238833854,
+      "learning_rate": 0.0001888804264993844,
+      "loss": 1.9102,
+      "step": 3323
+    },
+    {
+      "epoch": 0.53184,
+      "grad_norm": 2.279591167329188,
+      "learning_rate": 0.00018887250879858253,
+      "loss": 1.8298,
+      "step": 3324
+    },
+    {
+      "epoch": 0.532,
+      "grad_norm": 2.6482184893367133,
+      "learning_rate": 0.00018886458844595455,
+      "loss": 1.9329,
+      "step": 3325
+    },
+    {
+      "epoch": 0.53216,
+      "grad_norm": 2.009818610064851,
+      "learning_rate": 0.00018885666544173676,
+      "loss": 1.9282,
+      "step": 3326
+    },
+    {
+      "epoch": 0.53232,
+      "grad_norm": 2.2390466351391685,
+      "learning_rate": 0.0001888487397861655,
+      "loss": 1.9065,
+      "step": 3327
+    },
+    {
+      "epoch": 0.53248,
+      "grad_norm": 1.8648219692859052,
+      "learning_rate": 0.00018884081147947732,
+      "loss": 1.8296,
+      "step": 3328
+    },
+    {
+      "epoch": 0.53264,
+      "grad_norm": 2.256046303199721,
+      "learning_rate": 0.0001888328805219088,
+      "loss": 1.9619,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5328,
+      "grad_norm": 2.4888365782985464,
+      "learning_rate": 0.00018882494691369655,
+      "loss": 1.8855,
+      "step": 3330
+    },
+    {
+      "epoch": 0.53296,
+      "grad_norm": 2.7517039870675455,
+      "learning_rate": 0.00018881701065507733,
+      "loss": 1.7856,
+      "step": 3331
+    },
+    {
+      "epoch": 0.53312,
+      "grad_norm": 2.6453236877078754,
+      "learning_rate": 0.0001888090717462879,
+      "loss": 1.8966,
+      "step": 3332
+    },
+    {
+      "epoch": 0.53328,
+      "grad_norm": 2.2946125107341517,
+      "learning_rate": 0.00018880113018756523,
+      "loss": 1.9171,
+      "step": 3333
+    },
+    {
+      "epoch": 0.53344,
+      "grad_norm": 2.123103871135371,
+      "learning_rate": 0.00018879318597914618,
+      "loss": 1.841,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5336,
+      "grad_norm": 2.58500481081868,
+      "learning_rate": 0.0001887852391212679,
+      "loss": 1.864,
+      "step": 3335
+    },
+    {
+      "epoch": 0.53376,
+      "grad_norm": 2.3375520240632834,
+      "learning_rate": 0.00018877728961416742,
+      "loss": 1.8576,
+      "step": 3336
+    },
+    {
+      "epoch": 0.53392,
+      "grad_norm": 3.0571609370147357,
+      "learning_rate": 0.000188769337458082,
+      "loss": 1.8939,
+      "step": 3337
+    },
+    {
+      "epoch": 0.53408,
+      "grad_norm": 2.6193248054771523,
+      "learning_rate": 0.00018876138265324887,
+      "loss": 1.9489,
+      "step": 3338
+    },
+    {
+      "epoch": 0.53424,
+      "grad_norm": 2.296158636795969,
+      "learning_rate": 0.00018875342519990545,
+      "loss": 1.8672,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5344,
+      "grad_norm": 2.409386305389288,
+      "learning_rate": 0.00018874546509828913,
+      "loss": 1.9497,
+      "step": 3340
+    },
+    {
+      "epoch": 0.53456,
+      "grad_norm": 2.0951096631371566,
+      "learning_rate": 0.00018873750234863748,
+      "loss": 1.8692,
+      "step": 3341
+    },
+    {
+      "epoch": 0.53472,
+      "grad_norm": 2.08614663327132,
+      "learning_rate": 0.00018872953695118802,
+      "loss": 1.8134,
+      "step": 3342
+    },
+    {
+      "epoch": 0.53488,
+      "grad_norm": 2.0613622343037012,
+      "learning_rate": 0.0001887215689061785,
+      "loss": 1.8024,
+      "step": 3343
+    },
+    {
+      "epoch": 0.53504,
+      "grad_norm": 2.634875744434,
+      "learning_rate": 0.00018871359821384664,
+      "loss": 1.9627,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5352,
+      "grad_norm": 2.7672399180544844,
+      "learning_rate": 0.00018870562487443028,
+      "loss": 1.9171,
+      "step": 3345
+    },
+    {
+      "epoch": 0.53536,
+      "grad_norm": 2.136978250413692,
+      "learning_rate": 0.00018869764888816733,
+      "loss": 1.7224,
+      "step": 3346
+    },
+    {
+      "epoch": 0.53552,
+      "grad_norm": 2.455621633545787,
+      "learning_rate": 0.00018868967025529579,
+      "loss": 1.8542,
+      "step": 3347
+    },
+    {
+      "epoch": 0.53568,
+      "grad_norm": 2.4544497821659017,
+      "learning_rate": 0.0001886816889760537,
+      "loss": 1.8471,
+      "step": 3348
+    },
+    {
+      "epoch": 0.53584,
+      "grad_norm": 2.3099299591875115,
+      "learning_rate": 0.00018867370505067923,
+      "loss": 1.8725,
+      "step": 3349
+    },
+    {
+      "epoch": 0.536,
+      "grad_norm": 2.661435395512796,
+      "learning_rate": 0.00018866571847941065,
+      "loss": 1.8437,
+      "step": 3350
+    },
+    {
+      "epoch": 0.53616,
+      "grad_norm": 2.3752408454458798,
+      "learning_rate": 0.0001886577292624862,
+      "loss": 1.8645,
+      "step": 3351
+    },
+    {
+      "epoch": 0.53632,
+      "grad_norm": 2.456884432775735,
+      "learning_rate": 0.0001886497374001443,
+      "loss": 1.838,
+      "step": 3352
+    },
+    {
+      "epoch": 0.53648,
+      "grad_norm": 2.3456988200307287,
+      "learning_rate": 0.00018864174289262345,
+      "loss": 1.9127,
+      "step": 3353
+    },
+    {
+      "epoch": 0.53664,
+      "grad_norm": 2.8968775588690283,
+      "learning_rate": 0.0001886337457401621,
+      "loss": 1.9351,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5368,
+      "grad_norm": 2.445254155878241,
+      "learning_rate": 0.00018862574594299898,
+      "loss": 1.9385,
+      "step": 3355
+    },
+    {
+      "epoch": 0.53696,
+      "grad_norm": 2.3674400910389153,
+      "learning_rate": 0.0001886177435013727,
+      "loss": 1.8518,
+      "step": 3356
+    },
+    {
+      "epoch": 0.53712,
+      "grad_norm": 2.62327486080274,
+      "learning_rate": 0.0001886097384155221,
+      "loss": 1.8168,
+      "step": 3357
+    },
+    {
+      "epoch": 0.53728,
+      "grad_norm": 2.536880772055434,
+      "learning_rate": 0.00018860173068568602,
+      "loss": 1.8891,
+      "step": 3358
+    },
+    {
+      "epoch": 0.53744,
+      "grad_norm": 2.15527021657818,
+      "learning_rate": 0.00018859372031210342,
+      "loss": 1.82,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5376,
+      "grad_norm": 2.520303045369166,
+      "learning_rate": 0.0001885857072950133,
+      "loss": 1.8211,
+      "step": 3360
+    },
+    {
+      "epoch": 0.53776,
+      "grad_norm": 2.4127853058709032,
+      "learning_rate": 0.00018857769163465476,
+      "loss": 1.8134,
+      "step": 3361
+    },
+    {
+      "epoch": 0.53792,
+      "grad_norm": 2.348993247129188,
+      "learning_rate": 0.00018856967333126694,
+      "loss": 1.8539,
+      "step": 3362
+    },
+    {
+      "epoch": 0.53808,
+      "grad_norm": 2.726110231674797,
+      "learning_rate": 0.00018856165238508913,
+      "loss": 1.9393,
+      "step": 3363
+    },
+    {
+      "epoch": 0.53824,
+      "grad_norm": 2.768533864207901,
+      "learning_rate": 0.0001885536287963607,
+      "loss": 1.9842,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5384,
+      "grad_norm": 2.5057327899575332,
+      "learning_rate": 0.000188545602565321,
+      "loss": 1.8376,
+      "step": 3365
+    },
+    {
+      "epoch": 0.53856,
+      "grad_norm": 2.4190749582721707,
+      "learning_rate": 0.00018853757369220956,
+      "loss": 1.8445,
+      "step": 3366
+    },
+    {
+      "epoch": 0.53872,
+      "grad_norm": 3.039547262687638,
+      "learning_rate": 0.00018852954217726592,
+      "loss": 1.9311,
+      "step": 3367
+    },
+    {
+      "epoch": 0.53888,
+      "grad_norm": 2.6801993419427705,
+      "learning_rate": 0.00018852150802072974,
+      "loss": 1.8518,
+      "step": 3368
+    },
+    {
+      "epoch": 0.53904,
+      "grad_norm": 2.3798571957924337,
+      "learning_rate": 0.00018851347122284074,
+      "loss": 1.8614,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5392,
+      "grad_norm": 2.326027799130371,
+      "learning_rate": 0.00018850543178383874,
+      "loss": 1.8748,
+      "step": 3370
+    },
+    {
+      "epoch": 0.53936,
+      "grad_norm": 2.1432439737195437,
+      "learning_rate": 0.00018849738970396363,
+      "loss": 1.7881,
+      "step": 3371
+    },
+    {
+      "epoch": 0.53952,
+      "grad_norm": 3.208783642262571,
+      "learning_rate": 0.00018848934498345537,
+      "loss": 1.913,
+      "step": 3372
+    },
+    {
+      "epoch": 0.53968,
+      "grad_norm": 3.033857250341029,
+      "learning_rate": 0.000188481297622554,
+      "loss": 1.8248,
+      "step": 3373
+    },
+    {
+      "epoch": 0.53984,
+      "grad_norm": 2.4002573926513153,
+      "learning_rate": 0.00018847324762149965,
+      "loss": 1.8609,
+      "step": 3374
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 3.03326191347397,
+      "learning_rate": 0.00018846519498053247,
+      "loss": 1.8684,
+      "step": 3375
+    },
+    {
+      "epoch": 0.54016,
+      "grad_norm": 2.389067014264998,
+      "learning_rate": 0.00018845713969989282,
+      "loss": 1.8549,
+      "step": 3376
+    },
+    {
+      "epoch": 0.54032,
+      "grad_norm": 2.689989246628773,
+      "learning_rate": 0.000188449081779821,
+      "loss": 1.7851,
+      "step": 3377
+    },
+    {
+      "epoch": 0.54048,
+      "grad_norm": 2.4190896334666463,
+      "learning_rate": 0.00018844102122055745,
+      "loss": 1.9268,
+      "step": 3378
+    },
+    {
+      "epoch": 0.54064,
+      "grad_norm": 3.61180450138509,
+      "learning_rate": 0.00018843295802234273,
+      "loss": 1.8637,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5408,
+      "grad_norm": 2.0603465568908788,
+      "learning_rate": 0.0001884248921854174,
+      "loss": 1.7786,
+      "step": 3380
+    },
+    {
+      "epoch": 0.54096,
+      "grad_norm": 2.690950400360028,
+      "learning_rate": 0.0001884168237100221,
+      "loss": 1.9245,
+      "step": 3381
+    },
+    {
+      "epoch": 0.54112,
+      "grad_norm": 2.0834699146653977,
+      "learning_rate": 0.00018840875259639763,
+      "loss": 1.7722,
+      "step": 3382
+    },
+    {
+      "epoch": 0.54128,
+      "grad_norm": 2.5723597418635222,
+      "learning_rate": 0.00018840067884478483,
+      "loss": 1.8465,
+      "step": 3383
+    },
+    {
+      "epoch": 0.54144,
+      "grad_norm": 2.4715129077944846,
+      "learning_rate": 0.00018839260245542456,
+      "loss": 1.9575,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5416,
+      "grad_norm": 2.5885077246146473,
+      "learning_rate": 0.00018838452342855787,
+      "loss": 1.8977,
+      "step": 3385
+    },
+    {
+      "epoch": 0.54176,
+      "grad_norm": 2.2594931674637886,
+      "learning_rate": 0.00018837644176442577,
+      "loss": 1.7785,
+      "step": 3386
+    },
+    {
+      "epoch": 0.54192,
+      "grad_norm": 2.8055198302408413,
+      "learning_rate": 0.0001883683574632694,
+      "loss": 1.8844,
+      "step": 3387
+    },
+    {
+      "epoch": 0.54208,
+      "grad_norm": 2.2504248355630025,
+      "learning_rate": 0.00018836027052533005,
+      "loss": 1.8244,
+      "step": 3388
+    },
+    {
+      "epoch": 0.54224,
+      "grad_norm": 2.878178590435609,
+      "learning_rate": 0.00018835218095084898,
+      "loss": 1.8132,
+      "step": 3389
+    },
+    {
+      "epoch": 0.5424,
+      "grad_norm": 2.0853425402797043,
+      "learning_rate": 0.00018834408874006755,
+      "loss": 1.8689,
+      "step": 3390
+    },
+    {
+      "epoch": 0.54256,
+      "grad_norm": 2.8635640190528147,
+      "learning_rate": 0.00018833599389322725,
+      "loss": 1.8288,
+      "step": 3391
+    },
+    {
+      "epoch": 0.54272,
+      "grad_norm": 2.400902229857274,
+      "learning_rate": 0.0001883278964105696,
+      "loss": 1.8218,
+      "step": 3392
+    },
+    {
+      "epoch": 0.54288,
+      "grad_norm": 2.0447799713281265,
+      "learning_rate": 0.00018831979629233624,
+      "loss": 1.7931,
+      "step": 3393
+    },
+    {
+      "epoch": 0.54304,
+      "grad_norm": 2.57679898338439,
+      "learning_rate": 0.00018831169353876885,
+      "loss": 1.7914,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5432,
+      "grad_norm": 2.5417925957552603,
+      "learning_rate": 0.00018830358815010923,
+      "loss": 1.8776,
+      "step": 3395
+    },
+    {
+      "epoch": 0.54336,
+      "grad_norm": 2.4383593493808307,
+      "learning_rate": 0.00018829548012659918,
+      "loss": 1.7891,
+      "step": 3396
+    },
+    {
+      "epoch": 0.54352,
+      "grad_norm": 2.1148893725857927,
+      "learning_rate": 0.0001882873694684807,
+      "loss": 1.7291,
+      "step": 3397
+    },
+    {
+      "epoch": 0.54368,
+      "grad_norm": 2.989355850095863,
+      "learning_rate": 0.00018827925617599574,
+      "loss": 1.8332,
+      "step": 3398
+    },
+    {
+      "epoch": 0.54384,
+      "grad_norm": 2.387486490160427,
+      "learning_rate": 0.0001882711402493864,
+      "loss": 1.9071,
+      "step": 3399
+    },
+    {
+      "epoch": 0.544,
+      "grad_norm": 2.5493519883306615,
+      "learning_rate": 0.00018826302168889488,
+      "loss": 1.7881,
+      "step": 3400
+    },
+    {
+      "epoch": 0.54416,
+      "grad_norm": 2.394995919722691,
+      "learning_rate": 0.0001882549004947634,
+      "loss": 1.8369,
+      "step": 3401
+    },
+    {
+      "epoch": 0.54432,
+      "grad_norm": 2.217204631708528,
+      "learning_rate": 0.00018824677666723428,
+      "loss": 1.9459,
+      "step": 3402
+    },
+    {
+      "epoch": 0.54448,
+      "grad_norm": 2.2707507425114697,
+      "learning_rate": 0.00018823865020654995,
+      "loss": 1.8219,
+      "step": 3403
+    },
+    {
+      "epoch": 0.54464,
+      "grad_norm": 2.569205719901865,
+      "learning_rate": 0.00018823052111295287,
+      "loss": 1.9288,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5448,
+      "grad_norm": 2.526699492056423,
+      "learning_rate": 0.00018822238938668562,
+      "loss": 1.9156,
+      "step": 3405
+    },
+    {
+      "epoch": 0.54496,
+      "grad_norm": 3.88345444632863,
+      "learning_rate": 0.0001882142550279908,
+      "loss": 1.8591,
+      "step": 3406
+    },
+    {
+      "epoch": 0.54512,
+      "grad_norm": 2.2813594525014977,
+      "learning_rate": 0.00018820611803711118,
+      "loss": 1.8587,
+      "step": 3407
+    },
+    {
+      "epoch": 0.54528,
+      "grad_norm": 2.7304723343771085,
+      "learning_rate": 0.0001881979784142895,
+      "loss": 1.791,
+      "step": 3408
+    },
+    {
+      "epoch": 0.54544,
+      "grad_norm": 2.07748926601973,
+      "learning_rate": 0.00018818983615976866,
+      "loss": 1.824,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5456,
+      "grad_norm": 2.758297467951474,
+      "learning_rate": 0.00018818169127379162,
+      "loss": 1.9488,
+      "step": 3410
+    },
+    {
+      "epoch": 0.54576,
+      "grad_norm": 2.644781413036678,
+      "learning_rate": 0.0001881735437566014,
+      "loss": 1.946,
+      "step": 3411
+    },
+    {
+      "epoch": 0.54592,
+      "grad_norm": 2.8348163743796717,
+      "learning_rate": 0.00018816539360844115,
+      "loss": 1.8633,
+      "step": 3412
+    },
+    {
+      "epoch": 0.54608,
+      "grad_norm": 2.2948651189836977,
+      "learning_rate": 0.000188157240829554,
+      "loss": 1.8898,
+      "step": 3413
+    },
+    {
+      "epoch": 0.54624,
+      "grad_norm": 2.6561884400478037,
+      "learning_rate": 0.00018814908542018325,
+      "loss": 1.8953,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5464,
+      "grad_norm": 2.3176690519182794,
+      "learning_rate": 0.00018814092738057224,
+      "loss": 1.8394,
+      "step": 3415
+    },
+    {
+      "epoch": 0.54656,
+      "grad_norm": 2.9849774582815085,
+      "learning_rate": 0.0001881327667109644,
+      "loss": 1.8658,
+      "step": 3416
+    },
+    {
+      "epoch": 0.54672,
+      "grad_norm": 2.678326370540075,
+      "learning_rate": 0.00018812460341160317,
+      "loss": 1.8425,
+      "step": 3417
+    },
+    {
+      "epoch": 0.54688,
+      "grad_norm": 2.263664447819656,
+      "learning_rate": 0.00018811643748273225,
+      "loss": 1.8335,
+      "step": 3418
+    },
+    {
+      "epoch": 0.54704,
+      "grad_norm": 2.6588339082151697,
+      "learning_rate": 0.00018810826892459522,
+      "loss": 1.872,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5472,
+      "grad_norm": 2.165861942072059,
+      "learning_rate": 0.0001881000977374358,
+      "loss": 1.7818,
+      "step": 3420
+    },
+    {
+      "epoch": 0.54736,
+      "grad_norm": 2.4264605226725617,
+      "learning_rate": 0.00018809192392149786,
+      "loss": 1.8799,
+      "step": 3421
+    },
+    {
+      "epoch": 0.54752,
+      "grad_norm": 2.6756519104290475,
+      "learning_rate": 0.00018808374747702528,
+      "loss": 1.8215,
+      "step": 3422
+    },
+    {
+      "epoch": 0.54768,
+      "grad_norm": 2.5506326786429954,
+      "learning_rate": 0.000188075568404262,
+      "loss": 1.8257,
+      "step": 3423
+    },
+    {
+      "epoch": 0.54784,
+      "grad_norm": 2.774261397607938,
+      "learning_rate": 0.00018806738670345215,
+      "loss": 1.8072,
+      "step": 3424
+    },
+    {
+      "epoch": 0.548,
+      "grad_norm": 2.3386744464710914,
+      "learning_rate": 0.00018805920237483978,
+      "loss": 1.7539,
+      "step": 3425
+    },
+    {
+      "epoch": 0.54816,
+      "grad_norm": 2.6920825876065075,
+      "learning_rate": 0.00018805101541866911,
+      "loss": 1.9156,
+      "step": 3426
+    },
+    {
+      "epoch": 0.54832,
+      "grad_norm": 2.906885014820122,
+      "learning_rate": 0.00018804282583518447,
+      "loss": 1.839,
+      "step": 3427
+    },
+    {
+      "epoch": 0.54848,
+      "grad_norm": 2.3348008359919072,
+      "learning_rate": 0.00018803463362463018,
+      "loss": 1.8214,
+      "step": 3428
+    },
+    {
+      "epoch": 0.54864,
+      "grad_norm": 2.838967816420769,
+      "learning_rate": 0.00018802643878725072,
+      "loss": 1.8871,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5488,
+      "grad_norm": 3.1822990836520693,
+      "learning_rate": 0.0001880182413232906,
+      "loss": 1.8277,
+      "step": 3430
+    },
+    {
+      "epoch": 0.54896,
+      "grad_norm": 2.5536906920126508,
+      "learning_rate": 0.0001880100412329944,
+      "loss": 1.8072,
+      "step": 3431
+    },
+    {
+      "epoch": 0.54912,
+      "grad_norm": 2.1758520554472964,
+      "learning_rate": 0.00018800183851660682,
+      "loss": 1.9404,
+      "step": 3432
+    },
+    {
+      "epoch": 0.54928,
+      "grad_norm": 3.0064774372334644,
+      "learning_rate": 0.00018799363317437262,
+      "loss": 1.8804,
+      "step": 3433
+    },
+    {
+      "epoch": 0.54944,
+      "grad_norm": 2.5169050528092844,
+      "learning_rate": 0.0001879854252065366,
+      "loss": 1.8393,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5496,
+      "grad_norm": 2.6205346565868273,
+      "learning_rate": 0.0001879772146133437,
+      "loss": 1.8764,
+      "step": 3435
+    },
+    {
+      "epoch": 0.54976,
+      "grad_norm": 2.9165008852918444,
+      "learning_rate": 0.00018796900139503894,
+      "loss": 2.0305,
+      "step": 3436
+    },
+    {
+      "epoch": 0.54992,
+      "grad_norm": 2.7170204948824703,
+      "learning_rate": 0.00018796078555186733,
+      "loss": 1.7858,
+      "step": 3437
+    },
+    {
+      "epoch": 0.55008,
+      "grad_norm": 2.762388941902107,
+      "learning_rate": 0.00018795256708407408,
+      "loss": 1.7989,
+      "step": 3438
+    },
+    {
+      "epoch": 0.55024,
+      "grad_norm": 2.9233872918479986,
+      "learning_rate": 0.00018794434599190437,
+      "loss": 1.8329,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5504,
+      "grad_norm": 2.0921736041920336,
+      "learning_rate": 0.00018793612227560353,
+      "loss": 1.7741,
+      "step": 3440
+    },
+    {
+      "epoch": 0.55056,
+      "grad_norm": 2.4863865420862896,
+      "learning_rate": 0.00018792789593541695,
+      "loss": 1.8133,
+      "step": 3441
+    },
+    {
+      "epoch": 0.55072,
+      "grad_norm": 2.124782824364368,
+      "learning_rate": 0.00018791966697159006,
+      "loss": 1.8106,
+      "step": 3442
+    },
+    {
+      "epoch": 0.55088,
+      "grad_norm": 2.373226022666757,
+      "learning_rate": 0.00018791143538436845,
+      "loss": 1.7782,
+      "step": 3443
+    },
+    {
+      "epoch": 0.55104,
+      "grad_norm": 2.4154310769567915,
+      "learning_rate": 0.00018790320117399766,
+      "loss": 1.8785,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5512,
+      "grad_norm": 3.0167687755030816,
+      "learning_rate": 0.00018789496434072349,
+      "loss": 1.8831,
+      "step": 3445
+    },
+    {
+      "epoch": 0.55136,
+      "grad_norm": 2.4013262015033363,
+      "learning_rate": 0.00018788672488479163,
+      "loss": 1.8586,
+      "step": 3446
+    },
+    {
+      "epoch": 0.55152,
+      "grad_norm": 2.8490718256671155,
+      "learning_rate": 0.00018787848280644797,
+      "loss": 1.7367,
+      "step": 3447
+    },
+    {
+      "epoch": 0.55168,
+      "grad_norm": 2.571302058407648,
+      "learning_rate": 0.00018787023810593842,
+      "loss": 1.9198,
+      "step": 3448
+    },
+    {
+      "epoch": 0.55184,
+      "grad_norm": 2.377849347759235,
+      "learning_rate": 0.00018786199078350902,
+      "loss": 2.0384,
+      "step": 3449
+    },
+    {
+      "epoch": 0.552,
+      "grad_norm": 2.0778406221874905,
+      "learning_rate": 0.00018785374083940586,
+      "loss": 1.7993,
+      "step": 3450
+    },
+    {
+      "epoch": 0.55216,
+      "grad_norm": 2.735045512511833,
+      "learning_rate": 0.00018784548827387507,
+      "loss": 1.9511,
+      "step": 3451
+    },
+    {
+      "epoch": 0.55232,
+      "grad_norm": 2.320953494536637,
+      "learning_rate": 0.0001878372330871629,
+      "loss": 1.9169,
+      "step": 3452
+    },
+    {
+      "epoch": 0.55248,
+      "grad_norm": 2.6766613392957064,
+      "learning_rate": 0.0001878289752795157,
+      "loss": 1.913,
+      "step": 3453
+    },
+    {
+      "epoch": 0.55264,
+      "grad_norm": 3.2212691158914994,
+      "learning_rate": 0.00018782071485117988,
+      "loss": 1.7917,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5528,
+      "grad_norm": 2.390503710832981,
+      "learning_rate": 0.00018781245180240187,
+      "loss": 1.9539,
+      "step": 3455
+    },
+    {
+      "epoch": 0.55296,
+      "grad_norm": 3.972280335258419,
+      "learning_rate": 0.00018780418613342827,
+      "loss": 1.8178,
+      "step": 3456
+    },
+    {
+      "epoch": 0.55312,
+      "grad_norm": 2.9795053094990975,
+      "learning_rate": 0.0001877959178445057,
+      "loss": 1.8774,
+      "step": 3457
+    },
+    {
+      "epoch": 0.55328,
+      "grad_norm": 3.2775679309066734,
+      "learning_rate": 0.00018778764693588087,
+      "loss": 1.8679,
+      "step": 3458
+    },
+    {
+      "epoch": 0.55344,
+      "grad_norm": 2.8982084273659283,
+      "learning_rate": 0.00018777937340780056,
+      "loss": 1.8892,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5536,
+      "grad_norm": 2.181165454450738,
+      "learning_rate": 0.00018777109726051167,
+      "loss": 1.7555,
+      "step": 3460
+    },
+    {
+      "epoch": 0.55376,
+      "grad_norm": 2.5034793756501776,
+      "learning_rate": 0.00018776281849426115,
+      "loss": 1.8406,
+      "step": 3461
+    },
+    {
+      "epoch": 0.55392,
+      "grad_norm": 2.85565780695327,
+      "learning_rate": 0.00018775453710929597,
+      "loss": 1.8268,
+      "step": 3462
+    },
+    {
+      "epoch": 0.55408,
+      "grad_norm": 2.328717865243124,
+      "learning_rate": 0.0001877462531058633,
+      "loss": 1.9225,
+      "step": 3463
+    },
+    {
+      "epoch": 0.55424,
+      "grad_norm": 2.420348546027227,
+      "learning_rate": 0.0001877379664842103,
+      "loss": 1.801,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5544,
+      "grad_norm": 2.365390232344128,
+      "learning_rate": 0.00018772967724458423,
+      "loss": 1.8028,
+      "step": 3465
+    },
+    {
+      "epoch": 0.55456,
+      "grad_norm": 2.5037428332624763,
+      "learning_rate": 0.0001877213853872324,
+      "loss": 1.7889,
+      "step": 3466
+    },
+    {
+      "epoch": 0.55472,
+      "grad_norm": 3.0971433575250886,
+      "learning_rate": 0.0001877130909124023,
+      "loss": 1.8472,
+      "step": 3467
+    },
+    {
+      "epoch": 0.55488,
+      "grad_norm": 2.4368360517086134,
+      "learning_rate": 0.00018770479382034135,
+      "loss": 1.8136,
+      "step": 3468
+    },
+    {
+      "epoch": 0.55504,
+      "grad_norm": 2.4357568378895347,
+      "learning_rate": 0.00018769649411129713,
+      "loss": 1.9289,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5552,
+      "grad_norm": 2.155339515338668,
+      "learning_rate": 0.00018768819178551733,
+      "loss": 1.8691,
+      "step": 3470
+    },
+    {
+      "epoch": 0.55536,
+      "grad_norm": 3.269038880536527,
+      "learning_rate": 0.0001876798868432497,
+      "loss": 1.8383,
+      "step": 3471
+    },
+    {
+      "epoch": 0.55552,
+      "grad_norm": 3.0878492770640293,
+      "learning_rate": 0.00018767157928474196,
+      "loss": 1.7836,
+      "step": 3472
+    },
+    {
+      "epoch": 0.55568,
+      "grad_norm": 2.411200502340821,
+      "learning_rate": 0.00018766326911024206,
+      "loss": 1.8703,
+      "step": 3473
+    },
+    {
+      "epoch": 0.55584,
+      "grad_norm": 2.6546026322324243,
+      "learning_rate": 0.00018765495631999795,
+      "loss": 1.9271,
+      "step": 3474
+    },
+    {
+      "epoch": 0.556,
+      "grad_norm": 2.4327614877042034,
+      "learning_rate": 0.00018764664091425767,
+      "loss": 1.8413,
+      "step": 3475
+    },
+    {
+      "epoch": 0.55616,
+      "grad_norm": 2.4316370947627406,
+      "learning_rate": 0.00018763832289326934,
+      "loss": 1.9053,
+      "step": 3476
+    },
+    {
+      "epoch": 0.55632,
+      "grad_norm": 2.5368139986433995,
+      "learning_rate": 0.00018763000225728119,
+      "loss": 1.806,
+      "step": 3477
+    },
+    {
+      "epoch": 0.55648,
+      "grad_norm": 2.0620562114153653,
+      "learning_rate": 0.0001876216790065414,
+      "loss": 1.7768,
+      "step": 3478
+    },
+    {
+      "epoch": 0.55664,
+      "grad_norm": 2.1703398907838833,
+      "learning_rate": 0.00018761335314129842,
+      "loss": 1.8479,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5568,
+      "grad_norm": 2.2931014983431424,
+      "learning_rate": 0.00018760502466180065,
+      "loss": 1.8191,
+      "step": 3480
+    },
+    {
+      "epoch": 0.55696,
+      "grad_norm": 2.41190790458806,
+      "learning_rate": 0.0001875966935682966,
+      "loss": 1.9021,
+      "step": 3481
+    },
+    {
+      "epoch": 0.55712,
+      "grad_norm": 2.206138065942389,
+      "learning_rate": 0.00018758835986103486,
+      "loss": 1.9103,
+      "step": 3482
+    },
+    {
+      "epoch": 0.55728,
+      "grad_norm": 2.445824292307528,
+      "learning_rate": 0.00018758002354026407,
+      "loss": 1.8097,
+      "step": 3483
+    },
+    {
+      "epoch": 0.55744,
+      "grad_norm": 2.474886815070142,
+      "learning_rate": 0.00018757168460623301,
+      "loss": 1.7885,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5576,
+      "grad_norm": 2.4372587697015464,
+      "learning_rate": 0.0001875633430591905,
+      "loss": 1.944,
+      "step": 3485
+    },
+    {
+      "epoch": 0.55776,
+      "grad_norm": 2.3724226454114703,
+      "learning_rate": 0.0001875549988993854,
+      "loss": 1.7932,
+      "step": 3486
+    },
+    {
+      "epoch": 0.55792,
+      "grad_norm": 2.3118326982879585,
+      "learning_rate": 0.0001875466521270667,
+      "loss": 1.8814,
+      "step": 3487
+    },
+    {
+      "epoch": 0.55808,
+      "grad_norm": 2.7994716655044822,
+      "learning_rate": 0.0001875383027424835,
+      "loss": 1.9775,
+      "step": 3488
+    },
+    {
+      "epoch": 0.55824,
+      "grad_norm": 2.492930909524291,
+      "learning_rate": 0.0001875299507458849,
+      "loss": 1.9385,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5584,
+      "grad_norm": 2.2086363309485098,
+      "learning_rate": 0.0001875215961375201,
+      "loss": 1.8058,
+      "step": 3490
+    },
+    {
+      "epoch": 0.55856,
+      "grad_norm": 2.2903098872465946,
+      "learning_rate": 0.00018751323891763843,
+      "loss": 1.8388,
+      "step": 3491
+    },
+    {
+      "epoch": 0.55872,
+      "grad_norm": 2.7138029125559684,
+      "learning_rate": 0.00018750487908648924,
+      "loss": 1.8372,
+      "step": 3492
+    },
+    {
+      "epoch": 0.55888,
+      "grad_norm": 2.740122629816534,
+      "learning_rate": 0.00018749651664432194,
+      "loss": 1.7334,
+      "step": 3493
+    },
+    {
+      "epoch": 0.55904,
+      "grad_norm": 2.7299882263192092,
+      "learning_rate": 0.00018748815159138607,
+      "loss": 1.8071,
+      "step": 3494
+    },
+    {
+      "epoch": 0.5592,
+      "grad_norm": 2.273846767540308,
+      "learning_rate": 0.00018747978392793125,
+      "loss": 1.7204,
+      "step": 3495
+    },
+    {
+      "epoch": 0.55936,
+      "grad_norm": 3.3539459891600774,
+      "learning_rate": 0.00018747141365420714,
+      "loss": 1.8445,
+      "step": 3496
+    },
+    {
+      "epoch": 0.55952,
+      "grad_norm": 2.2594013254872314,
+      "learning_rate": 0.00018746304077046352,
+      "loss": 1.7623,
+      "step": 3497
+    },
+    {
+      "epoch": 0.55968,
+      "grad_norm": 3.0289850529122315,
+      "learning_rate": 0.0001874546652769502,
+      "loss": 1.9361,
+      "step": 3498
+    },
+    {
+      "epoch": 0.55984,
+      "grad_norm": 2.7496814298365977,
+      "learning_rate": 0.00018744628717391713,
+      "loss": 1.8917,
+      "step": 3499
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 2.852718889513737,
+      "learning_rate": 0.00018743790646161425,
+      "loss": 1.8338,
+      "step": 3500
+    },
+    {
+      "epoch": 0.56016,
+      "grad_norm": 2.0828589558283257,
+      "learning_rate": 0.00018742952314029166,
+      "loss": 1.845,
+      "step": 3501
+    },
+    {
+      "epoch": 0.56032,
+      "grad_norm": 2.4747542216204454,
+      "learning_rate": 0.00018742113721019948,
+      "loss": 1.8138,
+      "step": 3502
+    },
+    {
+      "epoch": 0.56048,
+      "grad_norm": 2.2014613763745685,
+      "learning_rate": 0.00018741274867158797,
+      "loss": 1.7826,
+      "step": 3503
+    },
+    {
+      "epoch": 0.56064,
+      "grad_norm": 2.755788211373433,
+      "learning_rate": 0.0001874043575247074,
+      "loss": 1.7838,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5608,
+      "grad_norm": 2.103244488106163,
+      "learning_rate": 0.00018739596376980816,
+      "loss": 1.8427,
+      "step": 3505
+    },
+    {
+      "epoch": 0.56096,
+      "grad_norm": 2.1210410459276385,
+      "learning_rate": 0.00018738756740714074,
+      "loss": 1.7779,
+      "step": 3506
+    },
+    {
+      "epoch": 0.56112,
+      "grad_norm": 2.4999373177357387,
+      "learning_rate": 0.00018737916843695562,
+      "loss": 1.7851,
+      "step": 3507
+    },
+    {
+      "epoch": 0.56128,
+      "grad_norm": 3.374350704194289,
+      "learning_rate": 0.00018737076685950346,
+      "loss": 1.8906,
+      "step": 3508
+    },
+    {
+      "epoch": 0.56144,
+      "grad_norm": 2.5016709650362876,
+      "learning_rate": 0.00018736236267503491,
+      "loss": 1.8516,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5616,
+      "grad_norm": 2.88989660314022,
+      "learning_rate": 0.00018735395588380076,
+      "loss": 1.8714,
+      "step": 3510
+    },
+    {
+      "epoch": 0.56176,
+      "grad_norm": 2.141260389335806,
+      "learning_rate": 0.00018734554648605184,
+      "loss": 1.7984,
+      "step": 3511
+    },
+    {
+      "epoch": 0.56192,
+      "grad_norm": 3.1470072027090312,
+      "learning_rate": 0.00018733713448203912,
+      "loss": 1.8378,
+      "step": 3512
+    },
+    {
+      "epoch": 0.56208,
+      "grad_norm": 2.54524981690084,
+      "learning_rate": 0.00018732871987201357,
+      "loss": 1.8443,
+      "step": 3513
+    },
+    {
+      "epoch": 0.56224,
+      "grad_norm": 2.3248591466410904,
+      "learning_rate": 0.00018732030265622627,
+      "loss": 1.7294,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5624,
+      "grad_norm": 2.2928184110005394,
+      "learning_rate": 0.00018731188283492835,
+      "loss": 1.7407,
+      "step": 3515
+    },
+    {
+      "epoch": 0.56256,
+      "grad_norm": 2.514404434951277,
+      "learning_rate": 0.0001873034604083711,
+      "loss": 1.9119,
+      "step": 3516
+    },
+    {
+      "epoch": 0.56272,
+      "grad_norm": 2.843165443673217,
+      "learning_rate": 0.0001872950353768058,
+      "loss": 1.945,
+      "step": 3517
+    },
+    {
+      "epoch": 0.56288,
+      "grad_norm": 2.7985112505903693,
+      "learning_rate": 0.00018728660774048383,
+      "loss": 1.7887,
+      "step": 3518
+    },
+    {
+      "epoch": 0.56304,
+      "grad_norm": 2.884514528375924,
+      "learning_rate": 0.0001872781774996567,
+      "loss": 1.8777,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5632,
+      "grad_norm": 2.673101134464127,
+      "learning_rate": 0.00018726974465457593,
+      "loss": 1.8577,
+      "step": 3520
+    },
+    {
+      "epoch": 0.56336,
+      "grad_norm": 2.8830628371190286,
+      "learning_rate": 0.00018726130920549311,
+      "loss": 1.7924,
+      "step": 3521
+    },
+    {
+      "epoch": 0.56352,
+      "grad_norm": 3.29017896015454,
+      "learning_rate": 0.00018725287115266,
+      "loss": 1.8162,
+      "step": 3522
+    },
+    {
+      "epoch": 0.56368,
+      "grad_norm": 2.8277147559678064,
+      "learning_rate": 0.00018724443049632835,
+      "loss": 1.7901,
+      "step": 3523
+    },
+    {
+      "epoch": 0.56384,
+      "grad_norm": 2.4774821915685155,
+      "learning_rate": 0.00018723598723675004,
+      "loss": 1.7947,
+      "step": 3524
+    },
+    {
+      "epoch": 0.564,
+      "grad_norm": 2.434934990503685,
+      "learning_rate": 0.00018722754137417698,
+      "loss": 1.8229,
+      "step": 3525
+    },
+    {
+      "epoch": 0.56416,
+      "grad_norm": 2.4983701762822217,
+      "learning_rate": 0.00018721909290886116,
+      "loss": 1.8463,
+      "step": 3526
+    },
+    {
+      "epoch": 0.56432,
+      "grad_norm": 2.57041599694329,
+      "learning_rate": 0.00018721064184105472,
+      "loss": 1.8189,
+      "step": 3527
+    },
+    {
+      "epoch": 0.56448,
+      "grad_norm": 2.245432389917245,
+      "learning_rate": 0.00018720218817100984,
+      "loss": 1.8347,
+      "step": 3528
+    },
+    {
+      "epoch": 0.56464,
+      "grad_norm": 2.2777146235179644,
+      "learning_rate": 0.0001871937318989787,
+      "loss": 1.854,
+      "step": 3529
+    },
+    {
+      "epoch": 0.5648,
+      "grad_norm": 2.584326082954648,
+      "learning_rate": 0.00018718527302521366,
+      "loss": 1.825,
+      "step": 3530
+    },
+    {
+      "epoch": 0.56496,
+      "grad_norm": 2.3721134631710736,
+      "learning_rate": 0.00018717681154996712,
+      "loss": 1.7631,
+      "step": 3531
+    },
+    {
+      "epoch": 0.56512,
+      "grad_norm": 2.411546702319119,
+      "learning_rate": 0.00018716834747349156,
+      "loss": 1.7844,
+      "step": 3532
+    },
+    {
+      "epoch": 0.56528,
+      "grad_norm": 2.383564903798149,
+      "learning_rate": 0.00018715988079603949,
+      "loss": 1.8095,
+      "step": 3533
+    },
+    {
+      "epoch": 0.56544,
+      "grad_norm": 2.5645000295905462,
+      "learning_rate": 0.0001871514115178636,
+      "loss": 1.8297,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5656,
+      "grad_norm": 2.6639146169247354,
+      "learning_rate": 0.0001871429396392166,
+      "loss": 1.9107,
+      "step": 3535
+    },
+    {
+      "epoch": 0.56576,
+      "grad_norm": 2.1624620523939067,
+      "learning_rate": 0.00018713446516035125,
+      "loss": 1.8458,
+      "step": 3536
+    },
+    {
+      "epoch": 0.56592,
+      "grad_norm": 2.134818432576096,
+      "learning_rate": 0.00018712598808152042,
+      "loss": 1.8155,
+      "step": 3537
+    },
+    {
+      "epoch": 0.56608,
+      "grad_norm": 2.6108846299749113,
+      "learning_rate": 0.00018711750840297707,
+      "loss": 1.8287,
+      "step": 3538
+    },
+    {
+      "epoch": 0.56624,
+      "grad_norm": 2.1118741444938776,
+      "learning_rate": 0.00018710902612497423,
+      "loss": 1.7836,
+      "step": 3539
+    },
+    {
+      "epoch": 0.5664,
+      "grad_norm": 2.172852586587922,
+      "learning_rate": 0.00018710054124776497,
+      "loss": 1.8279,
+      "step": 3540
+    },
+    {
+      "epoch": 0.56656,
+      "grad_norm": 2.8405883921891912,
+      "learning_rate": 0.00018709205377160246,
+      "loss": 1.9,
+      "step": 3541
+    },
+    {
+      "epoch": 0.56672,
+      "grad_norm": 2.2075906690251776,
+      "learning_rate": 0.00018708356369673997,
+      "loss": 1.7846,
+      "step": 3542
+    },
+    {
+      "epoch": 0.56688,
+      "grad_norm": 2.2275031582383766,
+      "learning_rate": 0.0001870750710234309,
+      "loss": 1.903,
+      "step": 3543
+    },
+    {
+      "epoch": 0.56704,
+      "grad_norm": 3.6628515762666902,
+      "learning_rate": 0.00018706657575192853,
+      "loss": 1.869,
+      "step": 3544
+    },
+    {
+      "epoch": 0.5672,
+      "grad_norm": 2.1609506592777405,
+      "learning_rate": 0.0001870580778824864,
+      "loss": 1.7934,
+      "step": 3545
+    },
+    {
+      "epoch": 0.56736,
+      "grad_norm": 4.486195760683823,
+      "learning_rate": 0.00018704957741535813,
+      "loss": 1.8048,
+      "step": 3546
+    },
+    {
+      "epoch": 0.56752,
+      "grad_norm": 2.614261968890891,
+      "learning_rate": 0.00018704107435079726,
+      "loss": 1.7812,
+      "step": 3547
+    },
+    {
+      "epoch": 0.56768,
+      "grad_norm": 3.150658572401506,
+      "learning_rate": 0.0001870325686890576,
+      "loss": 1.888,
+      "step": 3548
+    },
+    {
+      "epoch": 0.56784,
+      "grad_norm": 2.158582334814209,
+      "learning_rate": 0.0001870240604303929,
+      "loss": 1.8745,
+      "step": 3549
+    },
+    {
+      "epoch": 0.568,
+      "grad_norm": 4.779394859403512,
+      "learning_rate": 0.00018701554957505705,
+      "loss": 1.9204,
+      "step": 3550
+    },
+    {
+      "epoch": 0.56816,
+      "grad_norm": 2.720689347509794,
+      "learning_rate": 0.00018700703612330402,
+      "loss": 1.7736,
+      "step": 3551
+    },
+    {
+      "epoch": 0.56832,
+      "grad_norm": 4.64531600560486,
+      "learning_rate": 0.00018699852007538777,
+      "loss": 1.9043,
+      "step": 3552
+    },
+    {
+      "epoch": 0.56848,
+      "grad_norm": 2.6134435071483124,
+      "learning_rate": 0.00018699000143156247,
+      "loss": 1.8479,
+      "step": 3553
+    },
+    {
+      "epoch": 0.56864,
+      "grad_norm": 3.554767993523918,
+      "learning_rate": 0.00018698148019208227,
+      "loss": 1.8995,
+      "step": 3554
+    },
+    {
+      "epoch": 0.5688,
+      "grad_norm": 2.765958409738011,
+      "learning_rate": 0.00018697295635720148,
+      "loss": 1.8564,
+      "step": 3555
+    },
+    {
+      "epoch": 0.56896,
+      "grad_norm": 3.042347542093229,
+      "learning_rate": 0.00018696442992717437,
+      "loss": 1.9041,
+      "step": 3556
+    },
+    {
+      "epoch": 0.56912,
+      "grad_norm": 2.3273721671492424,
+      "learning_rate": 0.0001869559009022554,
+      "loss": 1.839,
+      "step": 3557
+    },
+    {
+      "epoch": 0.56928,
+      "grad_norm": 3.454171265830914,
+      "learning_rate": 0.00018694736928269906,
+      "loss": 1.8669,
+      "step": 3558
+    },
+    {
+      "epoch": 0.56944,
+      "grad_norm": 2.0265933545701933,
+      "learning_rate": 0.00018693883506875993,
+      "loss": 1.8145,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5696,
+      "grad_norm": 2.797774141943942,
+      "learning_rate": 0.00018693029826069265,
+      "loss": 1.9067,
+      "step": 3560
+    },
+    {
+      "epoch": 0.56976,
+      "grad_norm": 2.386851286167712,
+      "learning_rate": 0.00018692175885875193,
+      "loss": 1.8288,
+      "step": 3561
+    },
+    {
+      "epoch": 0.56992,
+      "grad_norm": 2.3684201014380797,
+      "learning_rate": 0.00018691321686319256,
+      "loss": 1.9634,
+      "step": 3562
+    },
+    {
+      "epoch": 0.57008,
+      "grad_norm": 2.269984806259415,
+      "learning_rate": 0.00018690467227426945,
+      "loss": 1.8454,
+      "step": 3563
+    },
+    {
+      "epoch": 0.57024,
+      "grad_norm": 3.1417724257495134,
+      "learning_rate": 0.0001868961250922376,
+      "loss": 1.9614,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5704,
+      "grad_norm": 2.8270905669330992,
+      "learning_rate": 0.00018688757531735195,
+      "loss": 1.8691,
+      "step": 3565
+    },
+    {
+      "epoch": 0.57056,
+      "grad_norm": 3.4171733742750328,
+      "learning_rate": 0.00018687902294986768,
+      "loss": 1.7843,
+      "step": 3566
+    },
+    {
+      "epoch": 0.57072,
+      "grad_norm": 3.0301609590755314,
+      "learning_rate": 0.00018687046799003995,
+      "loss": 1.9608,
+      "step": 3567
+    },
+    {
+      "epoch": 0.57088,
+      "grad_norm": 2.7970130935392787,
+      "learning_rate": 0.00018686191043812403,
+      "loss": 1.84,
+      "step": 3568
+    },
+    {
+      "epoch": 0.57104,
+      "grad_norm": 2.997927077880955,
+      "learning_rate": 0.00018685335029437533,
+      "loss": 1.8275,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5712,
+      "grad_norm": 2.554759776319462,
+      "learning_rate": 0.00018684478755904915,
+      "loss": 1.8427,
+      "step": 3570
+    },
+    {
+      "epoch": 0.57136,
+      "grad_norm": 2.611463591453398,
+      "learning_rate": 0.00018683622223240106,
+      "loss": 1.7508,
+      "step": 3571
+    },
+    {
+      "epoch": 0.57152,
+      "grad_norm": 2.0658736999299743,
+      "learning_rate": 0.00018682765431468667,
+      "loss": 1.7804,
+      "step": 3572
+    },
+    {
+      "epoch": 0.57168,
+      "grad_norm": 2.978832954188104,
+      "learning_rate": 0.0001868190838061616,
+      "loss": 2.0102,
+      "step": 3573
+    },
+    {
+      "epoch": 0.57184,
+      "grad_norm": 2.2990784088574543,
+      "learning_rate": 0.00018681051070708153,
+      "loss": 1.828,
+      "step": 3574
+    },
+    {
+      "epoch": 0.572,
+      "grad_norm": 2.704181714900325,
+      "learning_rate": 0.00018680193501770235,
+      "loss": 1.8789,
+      "step": 3575
+    },
+    {
+      "epoch": 0.57216,
+      "grad_norm": 2.936912268871741,
+      "learning_rate": 0.0001867933567382799,
+      "loss": 1.82,
+      "step": 3576
+    },
+    {
+      "epoch": 0.57232,
+      "grad_norm": 2.931428481454991,
+      "learning_rate": 0.00018678477586907017,
+      "loss": 1.7625,
+      "step": 3577
+    },
+    {
+      "epoch": 0.57248,
+      "grad_norm": 3.3301988016633697,
+      "learning_rate": 0.0001867761924103292,
+      "loss": 1.9032,
+      "step": 3578
+    },
+    {
+      "epoch": 0.57264,
+      "grad_norm": 3.434426590949717,
+      "learning_rate": 0.00018676760636231307,
+      "loss": 1.9522,
+      "step": 3579
+    },
+    {
+      "epoch": 0.5728,
+      "grad_norm": 2.348494051352949,
+      "learning_rate": 0.000186759017725278,
+      "loss": 1.8419,
+      "step": 3580
+    },
+    {
+      "epoch": 0.57296,
+      "grad_norm": 2.73847593559306,
+      "learning_rate": 0.00018675042649948028,
+      "loss": 1.9287,
+      "step": 3581
+    },
+    {
+      "epoch": 0.57312,
+      "grad_norm": 2.5603403889672665,
+      "learning_rate": 0.00018674183268517622,
+      "loss": 1.8947,
+      "step": 3582
+    },
+    {
+      "epoch": 0.57328,
+      "grad_norm": 2.656882793047803,
+      "learning_rate": 0.00018673323628262228,
+      "loss": 1.8054,
+      "step": 3583
+    },
+    {
+      "epoch": 0.57344,
+      "grad_norm": 2.8650121844848226,
+      "learning_rate": 0.00018672463729207494,
+      "loss": 1.8135,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5736,
+      "grad_norm": 2.8263206175044524,
+      "learning_rate": 0.0001867160357137908,
+      "loss": 1.7963,
+      "step": 3585
+    },
+    {
+      "epoch": 0.57376,
+      "grad_norm": 3.0286604973879165,
+      "learning_rate": 0.00018670743154802652,
+      "loss": 1.8435,
+      "step": 3586
+    },
+    {
+      "epoch": 0.57392,
+      "grad_norm": 3.068457289470642,
+      "learning_rate": 0.00018669882479503883,
+      "loss": 1.7808,
+      "step": 3587
+    },
+    {
+      "epoch": 0.57408,
+      "grad_norm": 3.398911598834206,
+      "learning_rate": 0.00018669021545508456,
+      "loss": 1.8796,
+      "step": 3588
+    },
+    {
+      "epoch": 0.57424,
+      "grad_norm": 2.978003847435973,
+      "learning_rate": 0.00018668160352842056,
+      "loss": 1.8256,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5744,
+      "grad_norm": 2.9024183809948942,
+      "learning_rate": 0.00018667298901530382,
+      "loss": 1.7497,
+      "step": 3590
+    },
+    {
+      "epoch": 0.57456,
+      "grad_norm": 2.8933935196057647,
+      "learning_rate": 0.00018666437191599137,
+      "loss": 1.842,
+      "step": 3591
+    },
+    {
+      "epoch": 0.57472,
+      "grad_norm": 2.7354518482109067,
+      "learning_rate": 0.00018665575223074038,
+      "loss": 1.7892,
+      "step": 3592
+    },
+    {
+      "epoch": 0.57488,
+      "grad_norm": 2.112034189427756,
+      "learning_rate": 0.000186647129959808,
+      "loss": 1.8075,
+      "step": 3593
+    },
+    {
+      "epoch": 0.57504,
+      "grad_norm": 2.467331433095774,
+      "learning_rate": 0.00018663850510345148,
+      "loss": 1.7652,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5752,
+      "grad_norm": 2.676610223155541,
+      "learning_rate": 0.00018662987766192825,
+      "loss": 1.7713,
+      "step": 3595
+    },
+    {
+      "epoch": 0.57536,
+      "grad_norm": 2.6703119933011803,
+      "learning_rate": 0.0001866212476354957,
+      "loss": 1.7598,
+      "step": 3596
+    },
+    {
+      "epoch": 0.57552,
+      "grad_norm": 2.4704658201506473,
+      "learning_rate": 0.00018661261502441134,
+      "loss": 1.8162,
+      "step": 3597
+    },
+    {
+      "epoch": 0.57568,
+      "grad_norm": 2.4670541054662927,
+      "learning_rate": 0.00018660397982893275,
+      "loss": 1.9093,
+      "step": 3598
+    },
+    {
+      "epoch": 0.57584,
+      "grad_norm": 2.3137983831702376,
+      "learning_rate": 0.00018659534204931762,
+      "loss": 1.79,
+      "step": 3599
+    },
+    {
+      "epoch": 0.576,
+      "grad_norm": 2.6124087298922563,
+      "learning_rate": 0.00018658670168582363,
+      "loss": 1.7256,
+      "step": 3600
+    },
+    {
+      "epoch": 0.57616,
+      "grad_norm": 2.160675420526837,
+      "learning_rate": 0.00018657805873870865,
+      "loss": 1.8309,
+      "step": 3601
+    },
+    {
+      "epoch": 0.57632,
+      "grad_norm": 2.6905295281006567,
+      "learning_rate": 0.00018656941320823058,
+      "loss": 1.8509,
+      "step": 3602
+    },
+    {
+      "epoch": 0.57648,
+      "grad_norm": 2.997376156449377,
+      "learning_rate": 0.00018656076509464736,
+      "loss": 1.752,
+      "step": 3603
+    },
+    {
+      "epoch": 0.57664,
+      "grad_norm": 2.6427212123404273,
+      "learning_rate": 0.000186552114398217,
+      "loss": 1.847,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5768,
+      "grad_norm": 3.3392613053833484,
+      "learning_rate": 0.00018654346111919768,
+      "loss": 1.7874,
+      "step": 3605
+    },
+    {
+      "epoch": 0.57696,
+      "grad_norm": 2.269905140028578,
+      "learning_rate": 0.0001865348052578476,
+      "loss": 1.8385,
+      "step": 3606
+    },
+    {
+      "epoch": 0.57712,
+      "grad_norm": 3.782212120419357,
+      "learning_rate": 0.00018652614681442505,
+      "loss": 1.8713,
+      "step": 3607
+    },
+    {
+      "epoch": 0.57728,
+      "grad_norm": 2.2290278035237643,
+      "learning_rate": 0.0001865174857891883,
+      "loss": 1.8562,
+      "step": 3608
+    },
+    {
+      "epoch": 0.57744,
+      "grad_norm": 3.7425055523142943,
+      "learning_rate": 0.0001865088221823959,
+      "loss": 1.7492,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5776,
+      "grad_norm": 2.53518112914595,
+      "learning_rate": 0.00018650015599430626,
+      "loss": 1.8765,
+      "step": 3610
+    },
+    {
+      "epoch": 0.57776,
+      "grad_norm": 3.2768467310224945,
+      "learning_rate": 0.00018649148722517804,
+      "loss": 1.8069,
+      "step": 3611
+    },
+    {
+      "epoch": 0.57792,
+      "grad_norm": 2.918664025595221,
+      "learning_rate": 0.00018648281587526987,
+      "loss": 1.8416,
+      "step": 3612
+    },
+    {
+      "epoch": 0.57808,
+      "grad_norm": 2.3911649146127534,
+      "learning_rate": 0.00018647414194484046,
+      "loss": 1.7768,
+      "step": 3613
+    },
+    {
+      "epoch": 0.57824,
+      "grad_norm": 3.535825487934179,
+      "learning_rate": 0.0001864654654341487,
+      "loss": 1.9955,
+      "step": 3614
+    },
+    {
+      "epoch": 0.5784,
+      "grad_norm": 2.984391273447457,
+      "learning_rate": 0.00018645678634345342,
+      "loss": 1.8219,
+      "step": 3615
+    },
+    {
+      "epoch": 0.57856,
+      "grad_norm": 3.4674813689705486,
+      "learning_rate": 0.00018644810467301364,
+      "loss": 1.8957,
+      "step": 3616
+    },
+    {
+      "epoch": 0.57872,
+      "grad_norm": 3.215239747380621,
+      "learning_rate": 0.00018643942042308837,
+      "loss": 1.9195,
+      "step": 3617
+    },
+    {
+      "epoch": 0.57888,
+      "grad_norm": 2.8251500640534544,
+      "learning_rate": 0.00018643073359393678,
+      "loss": 1.7185,
+      "step": 3618
+    },
+    {
+      "epoch": 0.57904,
+      "grad_norm": 2.3840404713403176,
+      "learning_rate": 0.00018642204418581798,
+      "loss": 1.9346,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5792,
+      "grad_norm": 3.0234648894423506,
+      "learning_rate": 0.00018641335219899134,
+      "loss": 1.7688,
+      "step": 3620
+    },
+    {
+      "epoch": 0.57936,
+      "grad_norm": 2.2018004574385497,
+      "learning_rate": 0.00018640465763371621,
+      "loss": 1.7465,
+      "step": 3621
+    },
+    {
+      "epoch": 0.57952,
+      "grad_norm": 2.598289371629964,
+      "learning_rate": 0.00018639596049025198,
+      "loss": 1.7867,
+      "step": 3622
+    },
+    {
+      "epoch": 0.57968,
+      "grad_norm": 2.246567648697747,
+      "learning_rate": 0.00018638726076885818,
+      "loss": 1.8832,
+      "step": 3623
+    },
+    {
+      "epoch": 0.57984,
+      "grad_norm": 2.1747893162990515,
+      "learning_rate": 0.0001863785584697944,
+      "loss": 1.9092,
+      "step": 3624
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 2.2715428144178906,
+      "learning_rate": 0.0001863698535933203,
+      "loss": 1.7951,
+      "step": 3625
+    },
+    {
+      "epoch": 0.58016,
+      "grad_norm": 2.4418303655479936,
+      "learning_rate": 0.00018636114613969565,
+      "loss": 1.8356,
+      "step": 3626
+    },
+    {
+      "epoch": 0.58032,
+      "grad_norm": 2.6955478890445113,
+      "learning_rate": 0.00018635243610918022,
+      "loss": 1.9217,
+      "step": 3627
+    },
+    {
+      "epoch": 0.58048,
+      "grad_norm": 2.168953987856845,
+      "learning_rate": 0.00018634372350203392,
+      "loss": 1.7139,
+      "step": 3628
+    },
+    {
+      "epoch": 0.58064,
+      "grad_norm": 2.296302340634005,
+      "learning_rate": 0.00018633500831851673,
+      "loss": 1.8851,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5808,
+      "grad_norm": 2.1996295992594375,
+      "learning_rate": 0.0001863262905588887,
+      "loss": 1.8277,
+      "step": 3630
+    },
+    {
+      "epoch": 0.58096,
+      "grad_norm": 2.9190635491991537,
+      "learning_rate": 0.00018631757022340994,
+      "loss": 1.7669,
+      "step": 3631
+    },
+    {
+      "epoch": 0.58112,
+      "grad_norm": 2.2912474601193638,
+      "learning_rate": 0.00018630884731234067,
+      "loss": 1.8037,
+      "step": 3632
+    },
+    {
+      "epoch": 0.58128,
+      "grad_norm": 2.280736923899674,
+      "learning_rate": 0.00018630012182594116,
+      "loss": 1.7586,
+      "step": 3633
+    },
+    {
+      "epoch": 0.58144,
+      "grad_norm": 2.3924720243239905,
+      "learning_rate": 0.0001862913937644718,
+      "loss": 1.8472,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5816,
+      "grad_norm": 2.6273430986840323,
+      "learning_rate": 0.00018628266312819294,
+      "loss": 1.8968,
+      "step": 3635
+    },
+    {
+      "epoch": 0.58176,
+      "grad_norm": 2.43203526826688,
+      "learning_rate": 0.00018627392991736515,
+      "loss": 1.8846,
+      "step": 3636
+    },
+    {
+      "epoch": 0.58192,
+      "grad_norm": 2.4957305038357105,
+      "learning_rate": 0.00018626519413224903,
+      "loss": 1.7476,
+      "step": 3637
+    },
+    {
+      "epoch": 0.58208,
+      "grad_norm": 3.0143128612390413,
+      "learning_rate": 0.0001862564557731052,
+      "loss": 1.7653,
+      "step": 3638
+    },
+    {
+      "epoch": 0.58224,
+      "grad_norm": 2.4818362057180727,
+      "learning_rate": 0.0001862477148401944,
+      "loss": 1.7079,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5824,
+      "grad_norm": 2.9540632052827833,
+      "learning_rate": 0.0001862389713337775,
+      "loss": 1.8926,
+      "step": 3640
+    },
+    {
+      "epoch": 0.58256,
+      "grad_norm": 2.2371926124156105,
+      "learning_rate": 0.00018623022525411537,
+      "loss": 1.7982,
+      "step": 3641
+    },
+    {
+      "epoch": 0.58272,
+      "grad_norm": 3.0167034374283945,
+      "learning_rate": 0.0001862214766014689,
+      "loss": 1.7736,
+      "step": 3642
+    },
+    {
+      "epoch": 0.58288,
+      "grad_norm": 2.7957132137824807,
+      "learning_rate": 0.00018621272537609927,
+      "loss": 1.8746,
+      "step": 3643
+    },
+    {
+      "epoch": 0.58304,
+      "grad_norm": 2.656867312017569,
+      "learning_rate": 0.0001862039715782675,
+      "loss": 1.7728,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5832,
+      "grad_norm": 2.3041309348643093,
+      "learning_rate": 0.00018619521520823486,
+      "loss": 1.8431,
+      "step": 3645
+    },
+    {
+      "epoch": 0.58336,
+      "grad_norm": 2.2647390858942855,
+      "learning_rate": 0.00018618645626626261,
+      "loss": 1.8012,
+      "step": 3646
+    },
+    {
+      "epoch": 0.58352,
+      "grad_norm": 2.6986092684974703,
+      "learning_rate": 0.00018617769475261207,
+      "loss": 1.8405,
+      "step": 3647
+    },
+    {
+      "epoch": 0.58368,
+      "grad_norm": 2.554439494634937,
+      "learning_rate": 0.0001861689306675447,
+      "loss": 1.7804,
+      "step": 3648
+    },
+    {
+      "epoch": 0.58384,
+      "grad_norm": 2.4230441461021863,
+      "learning_rate": 0.00018616016401132196,
+      "loss": 1.8467,
+      "step": 3649
+    },
+    {
+      "epoch": 0.584,
+      "grad_norm": 2.2811229571868377,
+      "learning_rate": 0.0001861513947842055,
+      "loss": 1.8266,
+      "step": 3650
+    },
+    {
+      "epoch": 0.58416,
+      "grad_norm": 2.0096702605769727,
+      "learning_rate": 0.000186142622986457,
+      "loss": 1.789,
+      "step": 3651
+    },
+    {
+      "epoch": 0.58432,
+      "grad_norm": 2.1492476354474075,
+      "learning_rate": 0.00018613384861833812,
+      "loss": 1.8572,
+      "step": 3652
+    },
+    {
+      "epoch": 0.58448,
+      "grad_norm": 2.228593050402,
+      "learning_rate": 0.0001861250716801107,
+      "loss": 1.8587,
+      "step": 3653
+    },
+    {
+      "epoch": 0.58464,
+      "grad_norm": 2.180468987591554,
+      "learning_rate": 0.00018611629217203662,
+      "loss": 1.8795,
+      "step": 3654
+    },
+    {
+      "epoch": 0.5848,
+      "grad_norm": 2.6751025011856524,
+      "learning_rate": 0.0001861075100943779,
+      "loss": 1.8515,
+      "step": 3655
+    },
+    {
+      "epoch": 0.58496,
+      "grad_norm": 2.813948956166206,
+      "learning_rate": 0.00018609872544739654,
+      "loss": 1.8003,
+      "step": 3656
+    },
+    {
+      "epoch": 0.58512,
+      "grad_norm": 2.3241263846819202,
+      "learning_rate": 0.00018608993823135468,
+      "loss": 1.7482,
+      "step": 3657
+    },
+    {
+      "epoch": 0.58528,
+      "grad_norm": 2.840681396319432,
+      "learning_rate": 0.0001860811484465145,
+      "loss": 1.7588,
+      "step": 3658
+    },
+    {
+      "epoch": 0.58544,
+      "grad_norm": 2.1661937094916643,
+      "learning_rate": 0.00018607235609313828,
+      "loss": 1.7319,
+      "step": 3659
+    },
+    {
+      "epoch": 0.5856,
+      "grad_norm": 2.7156466885217863,
+      "learning_rate": 0.0001860635611714884,
+      "loss": 1.8542,
+      "step": 3660
+    },
+    {
+      "epoch": 0.58576,
+      "grad_norm": 2.668084136959145,
+      "learning_rate": 0.00018605476368182723,
+      "loss": 1.8618,
+      "step": 3661
+    },
+    {
+      "epoch": 0.58592,
+      "grad_norm": 2.8752059666358014,
+      "learning_rate": 0.00018604596362441732,
+      "loss": 1.9416,
+      "step": 3662
+    },
+    {
+      "epoch": 0.58608,
+      "grad_norm": 2.412339482921963,
+      "learning_rate": 0.00018603716099952123,
+      "loss": 1.7658,
+      "step": 3663
+    },
+    {
+      "epoch": 0.58624,
+      "grad_norm": 2.526048502145221,
+      "learning_rate": 0.00018602835580740164,
+      "loss": 1.8134,
+      "step": 3664
+    },
+    {
+      "epoch": 0.5864,
+      "grad_norm": 2.233068613792209,
+      "learning_rate": 0.00018601954804832127,
+      "loss": 1.8528,
+      "step": 3665
+    },
+    {
+      "epoch": 0.58656,
+      "grad_norm": 2.5155319342794766,
+      "learning_rate": 0.00018601073772254293,
+      "loss": 1.7794,
+      "step": 3666
+    },
+    {
+      "epoch": 0.58672,
+      "grad_norm": 2.1076673556416576,
+      "learning_rate": 0.0001860019248303295,
+      "loss": 1.8759,
+      "step": 3667
+    },
+    {
+      "epoch": 0.58688,
+      "grad_norm": 2.2217525762611334,
+      "learning_rate": 0.00018599310937194395,
+      "loss": 1.7333,
+      "step": 3668
+    },
+    {
+      "epoch": 0.58704,
+      "grad_norm": 3.315416187222373,
+      "learning_rate": 0.0001859842913476493,
+      "loss": 1.9044,
+      "step": 3669
+    },
+    {
+      "epoch": 0.5872,
+      "grad_norm": 2.331390237574158,
+      "learning_rate": 0.0001859754707577087,
+      "loss": 1.7671,
+      "step": 3670
+    },
+    {
+      "epoch": 0.58736,
+      "grad_norm": 2.4364973698484858,
+      "learning_rate": 0.00018596664760238535,
+      "loss": 1.7754,
+      "step": 3671
+    },
+    {
+      "epoch": 0.58752,
+      "grad_norm": 2.534412919429506,
+      "learning_rate": 0.00018595782188194252,
+      "loss": 1.7599,
+      "step": 3672
+    },
+    {
+      "epoch": 0.58768,
+      "grad_norm": 2.2833417349504144,
+      "learning_rate": 0.0001859489935966435,
+      "loss": 1.9669,
+      "step": 3673
+    },
+    {
+      "epoch": 0.58784,
+      "grad_norm": 2.18756481455411,
+      "learning_rate": 0.00018594016274675176,
+      "loss": 1.842,
+      "step": 3674
+    },
+    {
+      "epoch": 0.588,
+      "grad_norm": 2.456531305321912,
+      "learning_rate": 0.0001859313293325308,
+      "loss": 1.7516,
+      "step": 3675
+    },
+    {
+      "epoch": 0.58816,
+      "grad_norm": 2.4045201269536687,
+      "learning_rate": 0.0001859224933542442,
+      "loss": 1.7473,
+      "step": 3676
+    },
+    {
+      "epoch": 0.58832,
+      "grad_norm": 2.1502169443840176,
+      "learning_rate": 0.00018591365481215557,
+      "loss": 1.8581,
+      "step": 3677
+    },
+    {
+      "epoch": 0.58848,
+      "grad_norm": 2.3897110834989013,
+      "learning_rate": 0.00018590481370652867,
+      "loss": 1.6795,
+      "step": 3678
+    },
+    {
+      "epoch": 0.58864,
+      "grad_norm": 2.7640269336472083,
+      "learning_rate": 0.00018589597003762734,
+      "loss": 1.7745,
+      "step": 3679
+    },
+    {
+      "epoch": 0.5888,
+      "grad_norm": 2.2589246987373794,
+      "learning_rate": 0.0001858871238057154,
+      "loss": 1.7936,
+      "step": 3680
+    },
+    {
+      "epoch": 0.58896,
+      "grad_norm": 2.4157783279784635,
+      "learning_rate": 0.00018587827501105687,
+      "loss": 1.8157,
+      "step": 3681
+    },
+    {
+      "epoch": 0.58912,
+      "grad_norm": 2.295676314007636,
+      "learning_rate": 0.0001858694236539157,
+      "loss": 1.8829,
+      "step": 3682
+    },
+    {
+      "epoch": 0.58928,
+      "grad_norm": 3.6346095220316115,
+      "learning_rate": 0.00018586056973455607,
+      "loss": 1.8995,
+      "step": 3683
+    },
+    {
+      "epoch": 0.58944,
+      "grad_norm": 2.319709812397154,
+      "learning_rate": 0.00018585171325324215,
+      "loss": 1.7584,
+      "step": 3684
+    },
+    {
+      "epoch": 0.5896,
+      "grad_norm": 3.1572437296234575,
+      "learning_rate": 0.00018584285421023824,
+      "loss": 1.7497,
+      "step": 3685
+    },
+    {
+      "epoch": 0.58976,
+      "grad_norm": 2.820784227432587,
+      "learning_rate": 0.00018583399260580863,
+      "loss": 1.7898,
+      "step": 3686
+    },
+    {
+      "epoch": 0.58992,
+      "grad_norm": 3.0628333722058803,
+      "learning_rate": 0.00018582512844021772,
+      "loss": 1.9394,
+      "step": 3687
+    },
+    {
+      "epoch": 0.59008,
+      "grad_norm": 2.4879119791235076,
+      "learning_rate": 0.00018581626171373009,
+      "loss": 1.8607,
+      "step": 3688
+    },
+    {
+      "epoch": 0.59024,
+      "grad_norm": 2.758411881563917,
+      "learning_rate": 0.00018580739242661021,
+      "loss": 1.793,
+      "step": 3689
+    },
+    {
+      "epoch": 0.5904,
+      "grad_norm": 2.254410719320245,
+      "learning_rate": 0.00018579852057912282,
+      "loss": 1.7669,
+      "step": 3690
+    },
+    {
+      "epoch": 0.59056,
+      "grad_norm": 3.2272156055104,
+      "learning_rate": 0.00018578964617153257,
+      "loss": 1.8149,
+      "step": 3691
+    },
+    {
+      "epoch": 0.59072,
+      "grad_norm": 2.635331344721307,
+      "learning_rate": 0.0001857807692041043,
+      "loss": 1.7925,
+      "step": 3692
+    },
+    {
+      "epoch": 0.59088,
+      "grad_norm": 2.7179618125223337,
+      "learning_rate": 0.00018577188967710287,
+      "loss": 1.8378,
+      "step": 3693
+    },
+    {
+      "epoch": 0.59104,
+      "grad_norm": 2.4552256979665943,
+      "learning_rate": 0.0001857630075907932,
+      "loss": 1.8113,
+      "step": 3694
+    },
+    {
+      "epoch": 0.5912,
+      "grad_norm": 2.6465435750347077,
+      "learning_rate": 0.0001857541229454404,
+      "loss": 1.7614,
+      "step": 3695
+    },
+    {
+      "epoch": 0.59136,
+      "grad_norm": 2.657997656673406,
+      "learning_rate": 0.00018574523574130952,
+      "loss": 1.8567,
+      "step": 3696
+    },
+    {
+      "epoch": 0.59152,
+      "grad_norm": 2.9544020704254796,
+      "learning_rate": 0.00018573634597866573,
+      "loss": 1.726,
+      "step": 3697
+    },
+    {
+      "epoch": 0.59168,
+      "grad_norm": 2.4937940105478864,
+      "learning_rate": 0.0001857274536577743,
+      "loss": 1.819,
+      "step": 3698
+    },
+    {
+      "epoch": 0.59184,
+      "grad_norm": 2.8796973323648505,
+      "learning_rate": 0.00018571855877890062,
+      "loss": 1.8433,
+      "step": 3699
+    },
+    {
+      "epoch": 0.592,
+      "grad_norm": 2.9099242368421083,
+      "learning_rate": 0.00018570966134231,
+      "loss": 1.8037,
+      "step": 3700
+    },
+    {
+      "epoch": 0.59216,
+      "grad_norm": 2.6928943827889555,
+      "learning_rate": 0.00018570076134826803,
+      "loss": 1.7412,
+      "step": 3701
+    },
+    {
+      "epoch": 0.59232,
+      "grad_norm": 2.2379465677044936,
+      "learning_rate": 0.00018569185879704016,
+      "loss": 1.7742,
+      "step": 3702
+    },
+    {
+      "epoch": 0.59248,
+      "grad_norm": 3.028452260066137,
+      "learning_rate": 0.00018568295368889213,
+      "loss": 1.7197,
+      "step": 3703
+    },
+    {
+      "epoch": 0.59264,
+      "grad_norm": 3.318498545190936,
+      "learning_rate": 0.0001856740460240896,
+      "loss": 1.9081,
+      "step": 3704
+    },
+    {
+      "epoch": 0.5928,
+      "grad_norm": 2.623583359957062,
+      "learning_rate": 0.00018566513580289836,
+      "loss": 1.8526,
+      "step": 3705
+    },
+    {
+      "epoch": 0.59296,
+      "grad_norm": 2.6420343009342973,
+      "learning_rate": 0.00018565622302558433,
+      "loss": 1.7906,
+      "step": 3706
+    },
+    {
+      "epoch": 0.59312,
+      "grad_norm": 2.854446296203913,
+      "learning_rate": 0.0001856473076924134,
+      "loss": 1.8287,
+      "step": 3707
+    },
+    {
+      "epoch": 0.59328,
+      "grad_norm": 3.0199712044581424,
+      "learning_rate": 0.00018563838980365164,
+      "loss": 1.7406,
+      "step": 3708
+    },
+    {
+      "epoch": 0.59344,
+      "grad_norm": 2.444669093853658,
+      "learning_rate": 0.00018562946935956506,
+      "loss": 1.7621,
+      "step": 3709
+    },
+    {
+      "epoch": 0.5936,
+      "grad_norm": 2.887017606898198,
+      "learning_rate": 0.00018562054636041993,
+      "loss": 1.8352,
+      "step": 3710
+    },
+    {
+      "epoch": 0.59376,
+      "grad_norm": 2.3095138775279573,
+      "learning_rate": 0.00018561162080648244,
+      "loss": 1.9851,
+      "step": 3711
+    },
+    {
+      "epoch": 0.59392,
+      "grad_norm": 2.2786046498989134,
+      "learning_rate": 0.00018560269269801894,
+      "loss": 1.8142,
+      "step": 3712
+    },
+    {
+      "epoch": 0.59408,
+      "grad_norm": 2.929029955218469,
+      "learning_rate": 0.00018559376203529585,
+      "loss": 1.9086,
+      "step": 3713
+    },
+    {
+      "epoch": 0.59424,
+      "grad_norm": 2.585602567587634,
+      "learning_rate": 0.00018558482881857958,
+      "loss": 1.7971,
+      "step": 3714
+    },
+    {
+      "epoch": 0.5944,
+      "grad_norm": 2.71646945498548,
+      "learning_rate": 0.0001855758930481368,
+      "loss": 1.8222,
+      "step": 3715
+    },
+    {
+      "epoch": 0.59456,
+      "grad_norm": 2.5386757057414973,
+      "learning_rate": 0.000185566954724234,
+      "loss": 1.8509,
+      "step": 3716
+    },
+    {
+      "epoch": 0.59472,
+      "grad_norm": 2.2020804607830486,
+      "learning_rate": 0.00018555801384713796,
+      "loss": 1.7753,
+      "step": 3717
+    },
+    {
+      "epoch": 0.59488,
+      "grad_norm": 2.904764233581712,
+      "learning_rate": 0.0001855490704171155,
+      "loss": 1.7854,
+      "step": 3718
+    },
+    {
+      "epoch": 0.59504,
+      "grad_norm": 1.9901016324331615,
+      "learning_rate": 0.0001855401244344334,
+      "loss": 1.7537,
+      "step": 3719
+    },
+    {
+      "epoch": 0.5952,
+      "grad_norm": 2.159821233333846,
+      "learning_rate": 0.00018553117589935862,
+      "loss": 1.8106,
+      "step": 3720
+    },
+    {
+      "epoch": 0.59536,
+      "grad_norm": 2.551874539390924,
+      "learning_rate": 0.0001855222248121582,
+      "loss": 1.705,
+      "step": 3721
+    },
+    {
+      "epoch": 0.59552,
+      "grad_norm": 2.1938141924773187,
+      "learning_rate": 0.00018551327117309922,
+      "loss": 1.7548,
+      "step": 3722
+    },
+    {
+      "epoch": 0.59568,
+      "grad_norm": 2.1192434813191663,
+      "learning_rate": 0.00018550431498244883,
+      "loss": 1.722,
+      "step": 3723
+    },
+    {
+      "epoch": 0.59584,
+      "grad_norm": 2.5800467908149174,
+      "learning_rate": 0.0001854953562404743,
+      "loss": 1.8303,
+      "step": 3724
+    },
+    {
+      "epoch": 0.596,
+      "grad_norm": 2.604943589938889,
+      "learning_rate": 0.0001854863949474429,
+      "loss": 1.808,
+      "step": 3725
+    },
+    {
+      "epoch": 0.59616,
+      "grad_norm": 2.394294553933113,
+      "learning_rate": 0.00018547743110362204,
+      "loss": 1.8067,
+      "step": 3726
+    },
+    {
+      "epoch": 0.59632,
+      "grad_norm": 3.2687994594382475,
+      "learning_rate": 0.00018546846470927916,
+      "loss": 1.9293,
+      "step": 3727
+    },
+    {
+      "epoch": 0.59648,
+      "grad_norm": 3.05616552814709,
+      "learning_rate": 0.00018545949576468187,
+      "loss": 1.8101,
+      "step": 3728
+    },
+    {
+      "epoch": 0.59664,
+      "grad_norm": 2.9168046182836442,
+      "learning_rate": 0.00018545052427009775,
+      "loss": 1.7474,
+      "step": 3729
+    },
+    {
+      "epoch": 0.5968,
+      "grad_norm": 2.5962993296225654,
+      "learning_rate": 0.0001854415502257945,
+      "loss": 1.9239,
+      "step": 3730
+    },
+    {
+      "epoch": 0.59696,
+      "grad_norm": 2.4293974752370358,
+      "learning_rate": 0.00018543257363203987,
+      "loss": 1.8333,
+      "step": 3731
+    },
+    {
+      "epoch": 0.59712,
+      "grad_norm": 1.9293417374240185,
+      "learning_rate": 0.00018542359448910176,
+      "loss": 1.8164,
+      "step": 3732
+    },
+    {
+      "epoch": 0.59728,
+      "grad_norm": 3.3254765220976426,
+      "learning_rate": 0.00018541461279724804,
+      "loss": 1.8174,
+      "step": 3733
+    },
+    {
+      "epoch": 0.59744,
+      "grad_norm": 2.2339300668528566,
+      "learning_rate": 0.00018540562855674675,
+      "loss": 1.6889,
+      "step": 3734
+    },
+    {
+      "epoch": 0.5976,
+      "grad_norm": 3.0092772045790026,
+      "learning_rate": 0.00018539664176786595,
+      "loss": 1.838,
+      "step": 3735
+    },
+    {
+      "epoch": 0.59776,
+      "grad_norm": 2.5518414180158646,
+      "learning_rate": 0.0001853876524308738,
+      "loss": 1.7772,
+      "step": 3736
+    },
+    {
+      "epoch": 0.59792,
+      "grad_norm": 2.341490497710116,
+      "learning_rate": 0.00018537866054603853,
+      "loss": 1.7685,
+      "step": 3737
+    },
+    {
+      "epoch": 0.59808,
+      "grad_norm": 2.7958052102215776,
+      "learning_rate": 0.00018536966611362843,
+      "loss": 1.9733,
+      "step": 3738
+    },
+    {
+      "epoch": 0.59824,
+      "grad_norm": 2.4434117851748773,
+      "learning_rate": 0.0001853606691339119,
+      "loss": 1.8012,
+      "step": 3739
+    },
+    {
+      "epoch": 0.5984,
+      "grad_norm": 2.508602854623596,
+      "learning_rate": 0.00018535166960715737,
+      "loss": 1.8337,
+      "step": 3740
+    },
+    {
+      "epoch": 0.59856,
+      "grad_norm": 2.54188929329449,
+      "learning_rate": 0.00018534266753363342,
+      "loss": 1.8347,
+      "step": 3741
+    },
+    {
+      "epoch": 0.59872,
+      "grad_norm": 2.8836209126525802,
+      "learning_rate": 0.00018533366291360862,
+      "loss": 1.8349,
+      "step": 3742
+    },
+    {
+      "epoch": 0.59888,
+      "grad_norm": 3.6037482752719208,
+      "learning_rate": 0.00018532465574735163,
+      "loss": 1.9343,
+      "step": 3743
+    },
+    {
+      "epoch": 0.59904,
+      "grad_norm": 2.7236184313508303,
+      "learning_rate": 0.00018531564603513126,
+      "loss": 1.7864,
+      "step": 3744
+    },
+    {
+      "epoch": 0.5992,
+      "grad_norm": 2.595390615808817,
+      "learning_rate": 0.00018530663377721633,
+      "loss": 1.7815,
+      "step": 3745
+    },
+    {
+      "epoch": 0.59936,
+      "grad_norm": 2.912677395606222,
+      "learning_rate": 0.00018529761897387577,
+      "loss": 1.712,
+      "step": 3746
+    },
+    {
+      "epoch": 0.59952,
+      "grad_norm": 2.7220311333815133,
+      "learning_rate": 0.00018528860162537854,
+      "loss": 1.8046,
+      "step": 3747
+    },
+    {
+      "epoch": 0.59968,
+      "grad_norm": 2.2507516948273256,
+      "learning_rate": 0.0001852795817319937,
+      "loss": 1.7514,
+      "step": 3748
+    },
+    {
+      "epoch": 0.59984,
+      "grad_norm": 2.2514718321672333,
+      "learning_rate": 0.00018527055929399043,
+      "loss": 1.7876,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 2.4035823322481673,
+      "learning_rate": 0.00018526153431163792,
+      "loss": 1.9392,
+      "step": 3750
+    },
+    {
+      "epoch": 0.60016,
+      "grad_norm": 2.3795036439917587,
+      "learning_rate": 0.00018525250678520544,
+      "loss": 1.7404,
+      "step": 3751
+    },
+    {
+      "epoch": 0.60032,
+      "grad_norm": 2.781460281151083,
+      "learning_rate": 0.0001852434767149624,
+      "loss": 1.7965,
+      "step": 3752
+    },
+    {
+      "epoch": 0.60048,
+      "grad_norm": 2.5679087217992866,
+      "learning_rate": 0.00018523444410117822,
+      "loss": 1.8113,
+      "step": 3753
+    },
+    {
+      "epoch": 0.60064,
+      "grad_norm": 2.89574034548788,
+      "learning_rate": 0.00018522540894412245,
+      "loss": 1.8107,
+      "step": 3754
+    },
+    {
+      "epoch": 0.6008,
+      "grad_norm": 2.3456542252938393,
+      "learning_rate": 0.00018521637124406466,
+      "loss": 1.7821,
+      "step": 3755
+    },
+    {
+      "epoch": 0.60096,
+      "grad_norm": 2.889647503525885,
+      "learning_rate": 0.00018520733100127452,
+      "loss": 1.7597,
+      "step": 3756
+    },
+    {
+      "epoch": 0.60112,
+      "grad_norm": 2.493425340758384,
+      "learning_rate": 0.0001851982882160218,
+      "loss": 1.7466,
+      "step": 3757
+    },
+    {
+      "epoch": 0.60128,
+      "grad_norm": 2.286770479855649,
+      "learning_rate": 0.0001851892428885763,
+      "loss": 1.7793,
+      "step": 3758
+    },
+    {
+      "epoch": 0.60144,
+      "grad_norm": 2.3035093153659143,
+      "learning_rate": 0.00018518019501920788,
+      "loss": 1.8127,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6016,
+      "grad_norm": 2.6892854296138706,
+      "learning_rate": 0.0001851711446081866,
+      "loss": 1.729,
+      "step": 3760
+    },
+    {
+      "epoch": 0.60176,
+      "grad_norm": 2.593514616008025,
+      "learning_rate": 0.00018516209165578246,
+      "loss": 1.9334,
+      "step": 3761
+    },
+    {
+      "epoch": 0.60192,
+      "grad_norm": 3.8935581484846926,
+      "learning_rate": 0.0001851530361622656,
+      "loss": 1.7322,
+      "step": 3762
+    },
+    {
+      "epoch": 0.60208,
+      "grad_norm": 2.544311059114329,
+      "learning_rate": 0.00018514397812790625,
+      "loss": 1.7885,
+      "step": 3763
+    },
+    {
+      "epoch": 0.60224,
+      "grad_norm": 3.006940560609085,
+      "learning_rate": 0.00018513491755297462,
+      "loss": 1.7745,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6024,
+      "grad_norm": 2.5798382998898166,
+      "learning_rate": 0.00018512585443774114,
+      "loss": 1.7473,
+      "step": 3765
+    },
+    {
+      "epoch": 0.60256,
+      "grad_norm": 3.1456718757519484,
+      "learning_rate": 0.0001851167887824762,
+      "loss": 1.7977,
+      "step": 3766
+    },
+    {
+      "epoch": 0.60272,
+      "grad_norm": 3.5810710293593573,
+      "learning_rate": 0.0001851077205874503,
+      "loss": 1.9136,
+      "step": 3767
+    },
+    {
+      "epoch": 0.60288,
+      "grad_norm": 3.367713829850611,
+      "learning_rate": 0.00018509864985293402,
+      "loss": 1.8631,
+      "step": 3768
+    },
+    {
+      "epoch": 0.60304,
+      "grad_norm": 2.4385374499134684,
+      "learning_rate": 0.00018508957657919803,
+      "loss": 1.8606,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6032,
+      "grad_norm": 2.3478795518016784,
+      "learning_rate": 0.0001850805007665131,
+      "loss": 1.8436,
+      "step": 3770
+    },
+    {
+      "epoch": 0.60336,
+      "grad_norm": 2.733204227679743,
+      "learning_rate": 0.00018507142241514995,
+      "loss": 1.7601,
+      "step": 3771
+    },
+    {
+      "epoch": 0.60352,
+      "grad_norm": 2.7549766626720835,
+      "learning_rate": 0.00018506234152537956,
+      "loss": 1.7672,
+      "step": 3772
+    },
+    {
+      "epoch": 0.60368,
+      "grad_norm": 2.710744154339187,
+      "learning_rate": 0.00018505325809747283,
+      "loss": 1.7799,
+      "step": 3773
+    },
+    {
+      "epoch": 0.60384,
+      "grad_norm": 2.4207131032394122,
+      "learning_rate": 0.00018504417213170083,
+      "loss": 1.7929,
+      "step": 3774
+    },
+    {
+      "epoch": 0.604,
+      "grad_norm": 3.666167801089502,
+      "learning_rate": 0.00018503508362833465,
+      "loss": 1.8206,
+      "step": 3775
+    },
+    {
+      "epoch": 0.60416,
+      "grad_norm": 2.480478341870648,
+      "learning_rate": 0.00018502599258764547,
+      "loss": 1.8161,
+      "step": 3776
+    },
+    {
+      "epoch": 0.60432,
+      "grad_norm": 3.1027523424919954,
+      "learning_rate": 0.00018501689900990456,
+      "loss": 1.8025,
+      "step": 3777
+    },
+    {
+      "epoch": 0.60448,
+      "grad_norm": 2.427467798490756,
+      "learning_rate": 0.0001850078028953833,
+      "loss": 1.7076,
+      "step": 3778
+    },
+    {
+      "epoch": 0.60464,
+      "grad_norm": 3.2459920838280554,
+      "learning_rate": 0.00018499870424435307,
+      "loss": 1.7805,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6048,
+      "grad_norm": 2.459330019512458,
+      "learning_rate": 0.00018498960305708532,
+      "loss": 1.8629,
+      "step": 3780
+    },
+    {
+      "epoch": 0.60496,
+      "grad_norm": 2.337554912877655,
+      "learning_rate": 0.00018498049933385167,
+      "loss": 1.7672,
+      "step": 3781
+    },
+    {
+      "epoch": 0.60512,
+      "grad_norm": 2.203005820459682,
+      "learning_rate": 0.00018497139307492377,
+      "loss": 1.7376,
+      "step": 3782
+    },
+    {
+      "epoch": 0.60528,
+      "grad_norm": 2.5097563381558996,
+      "learning_rate": 0.0001849622842805733,
+      "loss": 1.7957,
+      "step": 3783
+    },
+    {
+      "epoch": 0.60544,
+      "grad_norm": 2.3560042393604443,
+      "learning_rate": 0.00018495317295107207,
+      "loss": 1.7529,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6056,
+      "grad_norm": 1.9130278178516766,
+      "learning_rate": 0.00018494405908669198,
+      "loss": 1.8291,
+      "step": 3785
+    },
+    {
+      "epoch": 0.60576,
+      "grad_norm": 2.6900684494304605,
+      "learning_rate": 0.0001849349426877049,
+      "loss": 1.8178,
+      "step": 3786
+    },
+    {
+      "epoch": 0.60592,
+      "grad_norm": 2.6269244376485408,
+      "learning_rate": 0.00018492582375438294,
+      "loss": 1.82,
+      "step": 3787
+    },
+    {
+      "epoch": 0.60608,
+      "grad_norm": 2.479621683167451,
+      "learning_rate": 0.00018491670228699812,
+      "loss": 1.8525,
+      "step": 3788
+    },
+    {
+      "epoch": 0.60624,
+      "grad_norm": 2.2363132168395774,
+      "learning_rate": 0.00018490757828582263,
+      "loss": 1.7285,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6064,
+      "grad_norm": 2.5868856665248745,
+      "learning_rate": 0.00018489845175112874,
+      "loss": 1.6687,
+      "step": 3790
+    },
+    {
+      "epoch": 0.60656,
+      "grad_norm": 2.267034164683301,
+      "learning_rate": 0.00018488932268318877,
+      "loss": 1.8881,
+      "step": 3791
+    },
+    {
+      "epoch": 0.60672,
+      "grad_norm": 2.6685428697630384,
+      "learning_rate": 0.0001848801910822751,
+      "loss": 1.8649,
+      "step": 3792
+    },
+    {
+      "epoch": 0.60688,
+      "grad_norm": 2.1959140340665773,
+      "learning_rate": 0.0001848710569486602,
+      "loss": 1.7824,
+      "step": 3793
+    },
+    {
+      "epoch": 0.60704,
+      "grad_norm": 3.341334370590738,
+      "learning_rate": 0.00018486192028261664,
+      "loss": 1.8897,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6072,
+      "grad_norm": 2.4581825472394994,
+      "learning_rate": 0.00018485278108441703,
+      "loss": 1.7259,
+      "step": 3795
+    },
+    {
+      "epoch": 0.60736,
+      "grad_norm": 3.010295437049093,
+      "learning_rate": 0.00018484363935433412,
+      "loss": 1.7253,
+      "step": 3796
+    },
+    {
+      "epoch": 0.60752,
+      "grad_norm": 2.6402239983938607,
+      "learning_rate": 0.0001848344950926406,
+      "loss": 1.7574,
+      "step": 3797
+    },
+    {
+      "epoch": 0.60768,
+      "grad_norm": 2.9390195423016956,
+      "learning_rate": 0.00018482534829960935,
+      "loss": 1.792,
+      "step": 3798
+    },
+    {
+      "epoch": 0.60784,
+      "grad_norm": 2.536510367071869,
+      "learning_rate": 0.0001848161989755133,
+      "loss": 1.7296,
+      "step": 3799
+    },
+    {
+      "epoch": 0.608,
+      "grad_norm": 2.0739237457749398,
+      "learning_rate": 0.00018480704712062554,
+      "loss": 1.7223,
+      "step": 3800
+    },
+    {
+      "epoch": 0.60816,
+      "grad_norm": 2.915181640641759,
+      "learning_rate": 0.000184797892735219,
+      "loss": 1.7307,
+      "step": 3801
+    },
+    {
+      "epoch": 0.60832,
+      "grad_norm": 2.2449365070854506,
+      "learning_rate": 0.00018478873581956692,
+      "loss": 1.8522,
+      "step": 3802
+    },
+    {
+      "epoch": 0.60848,
+      "grad_norm": 2.7993565175700788,
+      "learning_rate": 0.0001847795763739425,
+      "loss": 1.7792,
+      "step": 3803
+    },
+    {
+      "epoch": 0.60864,
+      "grad_norm": 2.2576515263626633,
+      "learning_rate": 0.00018477041439861913,
+      "loss": 1.7655,
+      "step": 3804
+    },
+    {
+      "epoch": 0.6088,
+      "grad_norm": 2.4530253394603645,
+      "learning_rate": 0.00018476124989387006,
+      "loss": 1.7499,
+      "step": 3805
+    },
+    {
+      "epoch": 0.60896,
+      "grad_norm": 2.519865350902593,
+      "learning_rate": 0.0001847520828599688,
+      "loss": 1.8091,
+      "step": 3806
+    },
+    {
+      "epoch": 0.60912,
+      "grad_norm": 2.1889397766490784,
+      "learning_rate": 0.00018474291329718888,
+      "loss": 1.7387,
+      "step": 3807
+    },
+    {
+      "epoch": 0.60928,
+      "grad_norm": 2.579725670776706,
+      "learning_rate": 0.00018473374120580392,
+      "loss": 1.7451,
+      "step": 3808
+    },
+    {
+      "epoch": 0.60944,
+      "grad_norm": 2.481682440841195,
+      "learning_rate": 0.0001847245665860876,
+      "loss": 1.6961,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6096,
+      "grad_norm": 2.3256250581862306,
+      "learning_rate": 0.0001847153894383137,
+      "loss": 1.7355,
+      "step": 3810
+    },
+    {
+      "epoch": 0.60976,
+      "grad_norm": 2.1713199037165323,
+      "learning_rate": 0.00018470620976275597,
+      "loss": 1.7509,
+      "step": 3811
+    },
+    {
+      "epoch": 0.60992,
+      "grad_norm": 2.9540828432325488,
+      "learning_rate": 0.0001846970275596884,
+      "loss": 1.8532,
+      "step": 3812
+    },
+    {
+      "epoch": 0.61008,
+      "grad_norm": 2.072055885394398,
+      "learning_rate": 0.00018468784282938493,
+      "loss": 1.7308,
+      "step": 3813
+    },
+    {
+      "epoch": 0.61024,
+      "grad_norm": 2.5531462809029866,
+      "learning_rate": 0.00018467865557211962,
+      "loss": 1.8826,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6104,
+      "grad_norm": 2.2332819706882217,
+      "learning_rate": 0.00018466946578816665,
+      "loss": 1.7626,
+      "step": 3815
+    },
+    {
+      "epoch": 0.61056,
+      "grad_norm": 2.3820408939523783,
+      "learning_rate": 0.0001846602734778002,
+      "loss": 1.7753,
+      "step": 3816
+    },
+    {
+      "epoch": 0.61072,
+      "grad_norm": 2.7744961863831,
+      "learning_rate": 0.00018465107864129452,
+      "loss": 1.8688,
+      "step": 3817
+    },
+    {
+      "epoch": 0.61088,
+      "grad_norm": 2.0834860080246242,
+      "learning_rate": 0.000184641881278924,
+      "loss": 1.8176,
+      "step": 3818
+    },
+    {
+      "epoch": 0.61104,
+      "grad_norm": 2.2475227378710234,
+      "learning_rate": 0.00018463268139096313,
+      "loss": 1.8034,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6112,
+      "grad_norm": 2.454106520757057,
+      "learning_rate": 0.00018462347897768631,
+      "loss": 1.9204,
+      "step": 3820
+    },
+    {
+      "epoch": 0.61136,
+      "grad_norm": 2.305380325695404,
+      "learning_rate": 0.00018461427403936823,
+      "loss": 1.6837,
+      "step": 3821
+    },
+    {
+      "epoch": 0.61152,
+      "grad_norm": 2.384642731261471,
+      "learning_rate": 0.00018460506657628348,
+      "loss": 1.8551,
+      "step": 3822
+    },
+    {
+      "epoch": 0.61168,
+      "grad_norm": 2.8100131938976354,
+      "learning_rate": 0.00018459585658870683,
+      "loss": 1.8726,
+      "step": 3823
+    },
+    {
+      "epoch": 0.61184,
+      "grad_norm": 2.7414628928035167,
+      "learning_rate": 0.0001845866440769131,
+      "loss": 1.7628,
+      "step": 3824
+    },
+    {
+      "epoch": 0.612,
+      "grad_norm": 2.4548049693975065,
+      "learning_rate": 0.00018457742904117716,
+      "loss": 1.7749,
+      "step": 3825
+    },
+    {
+      "epoch": 0.61216,
+      "grad_norm": 2.795441405608909,
+      "learning_rate": 0.00018456821148177399,
+      "loss": 1.8503,
+      "step": 3826
+    },
+    {
+      "epoch": 0.61232,
+      "grad_norm": 2.600089251041259,
+      "learning_rate": 0.00018455899139897862,
+      "loss": 1.8392,
+      "step": 3827
+    },
+    {
+      "epoch": 0.61248,
+      "grad_norm": 2.0220469436043125,
+      "learning_rate": 0.00018454976879306613,
+      "loss": 1.7359,
+      "step": 3828
+    },
+    {
+      "epoch": 0.61264,
+      "grad_norm": 2.3336155472309894,
+      "learning_rate": 0.00018454054366431178,
+      "loss": 1.6902,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6128,
+      "grad_norm": 2.5974676700579797,
+      "learning_rate": 0.00018453131601299077,
+      "loss": 1.7953,
+      "step": 3830
+    },
+    {
+      "epoch": 0.61296,
+      "grad_norm": 2.679806767782253,
+      "learning_rate": 0.0001845220858393785,
+      "loss": 1.8288,
+      "step": 3831
+    },
+    {
+      "epoch": 0.61312,
+      "grad_norm": 3.0143267065498827,
+      "learning_rate": 0.00018451285314375032,
+      "loss": 1.7492,
+      "step": 3832
+    },
+    {
+      "epoch": 0.61328,
+      "grad_norm": 2.836672260177742,
+      "learning_rate": 0.00018450361792638175,
+      "loss": 1.7527,
+      "step": 3833
+    },
+    {
+      "epoch": 0.61344,
+      "grad_norm": 2.6580150539884655,
+      "learning_rate": 0.00018449438018754835,
+      "loss": 1.7516,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6136,
+      "grad_norm": 2.431794781171838,
+      "learning_rate": 0.0001844851399275258,
+      "loss": 1.7933,
+      "step": 3835
+    },
+    {
+      "epoch": 0.61376,
+      "grad_norm": 2.4752893094054076,
+      "learning_rate": 0.00018447589714658976,
+      "loss": 1.8414,
+      "step": 3836
+    },
+    {
+      "epoch": 0.61392,
+      "grad_norm": 2.2049374076834285,
+      "learning_rate": 0.00018446665184501603,
+      "loss": 1.7561,
+      "step": 3837
+    },
+    {
+      "epoch": 0.61408,
+      "grad_norm": 2.448401657208174,
+      "learning_rate": 0.00018445740402308052,
+      "loss": 1.8272,
+      "step": 3838
+    },
+    {
+      "epoch": 0.61424,
+      "grad_norm": 2.2913724489884983,
+      "learning_rate": 0.00018444815368105912,
+      "loss": 1.7584,
+      "step": 3839
+    },
+    {
+      "epoch": 0.6144,
+      "grad_norm": 2.3707339980414717,
+      "learning_rate": 0.00018443890081922787,
+      "loss": 1.7085,
+      "step": 3840
+    },
+    {
+      "epoch": 0.61456,
+      "grad_norm": 2.6615144847439502,
+      "learning_rate": 0.00018442964543786288,
+      "loss": 1.7799,
+      "step": 3841
+    },
+    {
+      "epoch": 0.61472,
+      "grad_norm": 2.315768229601935,
+      "learning_rate": 0.00018442038753724028,
+      "loss": 1.6433,
+      "step": 3842
+    },
+    {
+      "epoch": 0.61488,
+      "grad_norm": 2.429830574796721,
+      "learning_rate": 0.00018441112711763632,
+      "loss": 1.819,
+      "step": 3843
+    },
+    {
+      "epoch": 0.61504,
+      "grad_norm": 2.2426957158329768,
+      "learning_rate": 0.00018440186417932735,
+      "loss": 1.8066,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6152,
+      "grad_norm": 2.571037517327211,
+      "learning_rate": 0.00018439259872258973,
+      "loss": 1.8574,
+      "step": 3845
+    },
+    {
+      "epoch": 0.61536,
+      "grad_norm": 2.8421861534020603,
+      "learning_rate": 0.0001843833307476999,
+      "loss": 1.8465,
+      "step": 3846
+    },
+    {
+      "epoch": 0.61552,
+      "grad_norm": 1.9478065712249006,
+      "learning_rate": 0.0001843740602549345,
+      "loss": 1.8212,
+      "step": 3847
+    },
+    {
+      "epoch": 0.61568,
+      "grad_norm": 2.6823519508165536,
+      "learning_rate": 0.00018436478724457006,
+      "loss": 1.7755,
+      "step": 3848
+    },
+    {
+      "epoch": 0.61584,
+      "grad_norm": 2.4834593506724683,
+      "learning_rate": 0.0001843555117168833,
+      "loss": 1.8055,
+      "step": 3849
+    },
+    {
+      "epoch": 0.616,
+      "grad_norm": 2.272835532979468,
+      "learning_rate": 0.000184346233672151,
+      "loss": 1.7516,
+      "step": 3850
+    },
+    {
+      "epoch": 0.61616,
+      "grad_norm": 2.25539196770585,
+      "learning_rate": 0.00018433695311064998,
+      "loss": 1.8927,
+      "step": 3851
+    },
+    {
+      "epoch": 0.61632,
+      "grad_norm": 2.3947389081724695,
+      "learning_rate": 0.00018432767003265718,
+      "loss": 1.7425,
+      "step": 3852
+    },
+    {
+      "epoch": 0.61648,
+      "grad_norm": 2.1686582929982716,
+      "learning_rate": 0.00018431838443844957,
+      "loss": 1.7315,
+      "step": 3853
+    },
+    {
+      "epoch": 0.61664,
+      "grad_norm": 2.5005302159609406,
+      "learning_rate": 0.00018430909632830423,
+      "loss": 1.7108,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6168,
+      "grad_norm": 2.2834706266516007,
+      "learning_rate": 0.00018429980570249834,
+      "loss": 1.7633,
+      "step": 3855
+    },
+    {
+      "epoch": 0.61696,
+      "grad_norm": 2.8536330112815858,
+      "learning_rate": 0.00018429051256130908,
+      "loss": 1.8776,
+      "step": 3856
+    },
+    {
+      "epoch": 0.61712,
+      "grad_norm": 2.333732148344624,
+      "learning_rate": 0.00018428121690501373,
+      "loss": 1.7246,
+      "step": 3857
+    },
+    {
+      "epoch": 0.61728,
+      "grad_norm": 3.0300046383652974,
+      "learning_rate": 0.00018427191873388968,
+      "loss": 1.8465,
+      "step": 3858
+    },
+    {
+      "epoch": 0.61744,
+      "grad_norm": 2.2426619852434047,
+      "learning_rate": 0.00018426261804821436,
+      "loss": 1.7866,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6176,
+      "grad_norm": 2.5053702321485067,
+      "learning_rate": 0.00018425331484826532,
+      "loss": 1.8071,
+      "step": 3860
+    },
+    {
+      "epoch": 0.61776,
+      "grad_norm": 2.4175995443403093,
+      "learning_rate": 0.00018424400913432014,
+      "loss": 1.7643,
+      "step": 3861
+    },
+    {
+      "epoch": 0.61792,
+      "grad_norm": 2.3255904987796385,
+      "learning_rate": 0.0001842347009066565,
+      "loss": 1.834,
+      "step": 3862
+    },
+    {
+      "epoch": 0.61808,
+      "grad_norm": 2.41305524485664,
+      "learning_rate": 0.0001842253901655521,
+      "loss": 1.7967,
+      "step": 3863
+    },
+    {
+      "epoch": 0.61824,
+      "grad_norm": 2.404687276641847,
+      "learning_rate": 0.00018421607691128478,
+      "loss": 1.804,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6184,
+      "grad_norm": 2.4262758049036446,
+      "learning_rate": 0.00018420676114413244,
+      "loss": 1.7074,
+      "step": 3865
+    },
+    {
+      "epoch": 0.61856,
+      "grad_norm": 2.4835572546815894,
+      "learning_rate": 0.0001841974428643731,
+      "loss": 1.723,
+      "step": 3866
+    },
+    {
+      "epoch": 0.61872,
+      "grad_norm": 2.3341316659740956,
+      "learning_rate": 0.0001841881220722847,
+      "loss": 1.7897,
+      "step": 3867
+    },
+    {
+      "epoch": 0.61888,
+      "grad_norm": 2.33498162862105,
+      "learning_rate": 0.00018417879876814542,
+      "loss": 1.7296,
+      "step": 3868
+    },
+    {
+      "epoch": 0.61904,
+      "grad_norm": 2.7111050965746615,
+      "learning_rate": 0.00018416947295223345,
+      "loss": 1.7449,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6192,
+      "grad_norm": 2.2534750842706632,
+      "learning_rate": 0.00018416014462482706,
+      "loss": 1.787,
+      "step": 3870
+    },
+    {
+      "epoch": 0.61936,
+      "grad_norm": 2.767571735507206,
+      "learning_rate": 0.00018415081378620458,
+      "loss": 1.7772,
+      "step": 3871
+    },
+    {
+      "epoch": 0.61952,
+      "grad_norm": 2.026375488140731,
+      "learning_rate": 0.00018414148043664446,
+      "loss": 1.7838,
+      "step": 3872
+    },
+    {
+      "epoch": 0.61968,
+      "grad_norm": 2.403567803200363,
+      "learning_rate": 0.00018413214457642515,
+      "loss": 1.7351,
+      "step": 3873
+    },
+    {
+      "epoch": 0.61984,
+      "grad_norm": 2.5478540593828094,
+      "learning_rate": 0.00018412280620582524,
+      "loss": 1.8551,
+      "step": 3874
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 2.555128626900423,
+      "learning_rate": 0.0001841134653251234,
+      "loss": 1.869,
+      "step": 3875
+    },
+    {
+      "epoch": 0.62016,
+      "grad_norm": 2.7966104163276446,
+      "learning_rate": 0.0001841041219345983,
+      "loss": 1.8055,
+      "step": 3876
+    },
+    {
+      "epoch": 0.62032,
+      "grad_norm": 2.117468749878133,
+      "learning_rate": 0.00018409477603452873,
+      "loss": 1.759,
+      "step": 3877
+    },
+    {
+      "epoch": 0.62048,
+      "grad_norm": 2.234179243440923,
+      "learning_rate": 0.00018408542762519362,
+      "loss": 1.7122,
+      "step": 3878
+    },
+    {
+      "epoch": 0.62064,
+      "grad_norm": 2.4780784116730437,
+      "learning_rate": 0.00018407607670687184,
+      "loss": 1.7954,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6208,
+      "grad_norm": 2.3134686621726406,
+      "learning_rate": 0.00018406672327984247,
+      "loss": 1.7225,
+      "step": 3880
+    },
+    {
+      "epoch": 0.62096,
+      "grad_norm": 2.502074382162755,
+      "learning_rate": 0.00018405736734438457,
+      "loss": 1.8197,
+      "step": 3881
+    },
+    {
+      "epoch": 0.62112,
+      "grad_norm": 2.402901455396003,
+      "learning_rate": 0.0001840480089007773,
+      "loss": 1.7765,
+      "step": 3882
+    },
+    {
+      "epoch": 0.62128,
+      "grad_norm": 2.6851523069508527,
+      "learning_rate": 0.00018403864794929993,
+      "loss": 1.7485,
+      "step": 3883
+    },
+    {
+      "epoch": 0.62144,
+      "grad_norm": 2.0689908655904605,
+      "learning_rate": 0.00018402928449023177,
+      "loss": 1.7951,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6216,
+      "grad_norm": 2.5882654072190925,
+      "learning_rate": 0.00018401991852385222,
+      "loss": 1.6686,
+      "step": 3885
+    },
+    {
+      "epoch": 0.62176,
+      "grad_norm": 2.4561023447312236,
+      "learning_rate": 0.00018401055005044068,
+      "loss": 1.8314,
+      "step": 3886
+    },
+    {
+      "epoch": 0.62192,
+      "grad_norm": 2.1014171699047695,
+      "learning_rate": 0.00018400117907027674,
+      "loss": 1.7762,
+      "step": 3887
+    },
+    {
+      "epoch": 0.62208,
+      "grad_norm": 2.110062981472109,
+      "learning_rate": 0.00018399180558364007,
+      "loss": 1.7424,
+      "step": 3888
+    },
+    {
+      "epoch": 0.62224,
+      "grad_norm": 2.221718741541341,
+      "learning_rate": 0.00018398242959081026,
+      "loss": 1.6734,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6224,
+      "grad_norm": 2.106345040482968,
+      "learning_rate": 0.0001839730510920672,
+      "loss": 1.9065,
+      "step": 3890
+    },
+    {
+      "epoch": 0.62256,
+      "grad_norm": 2.7522575339587236,
+      "learning_rate": 0.0001839636700876906,
+      "loss": 1.7552,
+      "step": 3891
+    },
+    {
+      "epoch": 0.62272,
+      "grad_norm": 2.1678705783433476,
+      "learning_rate": 0.00018395428657796043,
+      "loss": 1.7681,
+      "step": 3892
+    },
+    {
+      "epoch": 0.62288,
+      "grad_norm": 2.0522229206578286,
+      "learning_rate": 0.0001839449005631567,
+      "loss": 1.7679,
+      "step": 3893
+    },
+    {
+      "epoch": 0.62304,
+      "grad_norm": 2.209844524363344,
+      "learning_rate": 0.00018393551204355944,
+      "loss": 1.7079,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6232,
+      "grad_norm": 2.3595435810584435,
+      "learning_rate": 0.0001839261210194488,
+      "loss": 1.7356,
+      "step": 3895
+    },
+    {
+      "epoch": 0.62336,
+      "grad_norm": 2.7754865066009855,
+      "learning_rate": 0.00018391672749110507,
+      "loss": 1.8383,
+      "step": 3896
+    },
+    {
+      "epoch": 0.62352,
+      "grad_norm": 2.1231882386654326,
+      "learning_rate": 0.0001839073314588084,
+      "loss": 1.817,
+      "step": 3897
+    },
+    {
+      "epoch": 0.62368,
+      "grad_norm": 2.5273351697263893,
+      "learning_rate": 0.00018389793292283927,
+      "loss": 1.7395,
+      "step": 3898
+    },
+    {
+      "epoch": 0.62384,
+      "grad_norm": 2.2610252731465437,
+      "learning_rate": 0.00018388853188347804,
+      "loss": 1.7043,
+      "step": 3899
+    },
+    {
+      "epoch": 0.624,
+      "grad_norm": 2.6648250727645,
+      "learning_rate": 0.0001838791283410053,
+      "loss": 1.75,
+      "step": 3900
+    },
+    {
+      "epoch": 0.62416,
+      "grad_norm": 2.8458319287655502,
+      "learning_rate": 0.00018386972229570155,
+      "loss": 1.8287,
+      "step": 3901
+    },
+    {
+      "epoch": 0.62432,
+      "grad_norm": 3.022292612858847,
+      "learning_rate": 0.00018386031374784752,
+      "loss": 1.7648,
+      "step": 3902
+    },
+    {
+      "epoch": 0.62448,
+      "grad_norm": 2.5429704288979416,
+      "learning_rate": 0.00018385090269772389,
+      "loss": 1.797,
+      "step": 3903
+    },
+    {
+      "epoch": 0.62464,
+      "grad_norm": 2.8189543723146335,
+      "learning_rate": 0.00018384148914561154,
+      "loss": 1.7746,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6248,
+      "grad_norm": 2.0177242908213744,
+      "learning_rate": 0.00018383207309179134,
+      "loss": 1.7386,
+      "step": 3905
+    },
+    {
+      "epoch": 0.62496,
+      "grad_norm": 3.5747344166338877,
+      "learning_rate": 0.0001838226545365442,
+      "loss": 1.7918,
+      "step": 3906
+    },
+    {
+      "epoch": 0.62512,
+      "grad_norm": 2.2722830711628546,
+      "learning_rate": 0.00018381323348015124,
+      "loss": 1.754,
+      "step": 3907
+    },
+    {
+      "epoch": 0.62528,
+      "grad_norm": 3.7065052692446927,
+      "learning_rate": 0.00018380380992289344,
+      "loss": 1.8102,
+      "step": 3908
+    },
+    {
+      "epoch": 0.62544,
+      "grad_norm": 2.161802090345177,
+      "learning_rate": 0.0001837943838650521,
+      "loss": 1.7026,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6256,
+      "grad_norm": 3.109642499493198,
+      "learning_rate": 0.00018378495530690852,
+      "loss": 1.86,
+      "step": 3910
+    },
+    {
+      "epoch": 0.62576,
+      "grad_norm": 2.4413575638088916,
+      "learning_rate": 0.00018377552424874387,
+      "loss": 1.8332,
+      "step": 3911
+    },
+    {
+      "epoch": 0.62592,
+      "grad_norm": 2.2866338937957296,
+      "learning_rate": 0.0001837660906908397,
+      "loss": 1.7631,
+      "step": 3912
+    },
+    {
+      "epoch": 0.62608,
+      "grad_norm": 2.1396537719322297,
+      "learning_rate": 0.00018375665463347742,
+      "loss": 1.8407,
+      "step": 3913
+    },
+    {
+      "epoch": 0.62624,
+      "grad_norm": 3.242802431861534,
+      "learning_rate": 0.00018374721607693862,
+      "loss": 1.7783,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6264,
+      "grad_norm": 2.8109078614678844,
+      "learning_rate": 0.00018373777502150493,
+      "loss": 1.7495,
+      "step": 3915
+    },
+    {
+      "epoch": 0.62656,
+      "grad_norm": 2.9020482142621242,
+      "learning_rate": 0.00018372833146745805,
+      "loss": 1.708,
+      "step": 3916
+    },
+    {
+      "epoch": 0.62672,
+      "grad_norm": 2.4268450656552156,
+      "learning_rate": 0.00018371888541507975,
+      "loss": 1.7969,
+      "step": 3917
+    },
+    {
+      "epoch": 0.62688,
+      "grad_norm": 2.88716239525153,
+      "learning_rate": 0.00018370943686465193,
+      "loss": 1.6131,
+      "step": 3918
+    },
+    {
+      "epoch": 0.62704,
+      "grad_norm": 2.541214305238535,
+      "learning_rate": 0.00018369998581645644,
+      "loss": 1.8016,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6272,
+      "grad_norm": 2.5803589857049296,
+      "learning_rate": 0.00018369053227077539,
+      "loss": 1.7875,
+      "step": 3920
+    },
+    {
+      "epoch": 0.62736,
+      "grad_norm": 3.1080572692801898,
+      "learning_rate": 0.0001836810762278908,
+      "loss": 1.7932,
+      "step": 3921
+    },
+    {
+      "epoch": 0.62752,
+      "grad_norm": 3.0715824148004547,
+      "learning_rate": 0.0001836716176880848,
+      "loss": 1.8657,
+      "step": 3922
+    },
+    {
+      "epoch": 0.62768,
+      "grad_norm": 2.320986662409173,
+      "learning_rate": 0.00018366215665163965,
+      "loss": 1.7543,
+      "step": 3923
+    },
+    {
+      "epoch": 0.62784,
+      "grad_norm": 3.2332775717110263,
+      "learning_rate": 0.00018365269311883768,
+      "loss": 1.7405,
+      "step": 3924
+    },
+    {
+      "epoch": 0.628,
+      "grad_norm": 2.3866708631626223,
+      "learning_rate": 0.00018364322708996125,
+      "loss": 1.7504,
+      "step": 3925
+    },
+    {
+      "epoch": 0.62816,
+      "grad_norm": 2.908901657482571,
+      "learning_rate": 0.0001836337585652928,
+      "loss": 1.7267,
+      "step": 3926
+    },
+    {
+      "epoch": 0.62832,
+      "grad_norm": 2.663069644836349,
+      "learning_rate": 0.00018362428754511485,
+      "loss": 1.8076,
+      "step": 3927
+    },
+    {
+      "epoch": 0.62848,
+      "grad_norm": 2.1908569747017754,
+      "learning_rate": 0.00018361481402970998,
+      "loss": 1.7363,
+      "step": 3928
+    },
+    {
+      "epoch": 0.62864,
+      "grad_norm": 2.192680764368909,
+      "learning_rate": 0.00018360533801936094,
+      "loss": 1.7315,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6288,
+      "grad_norm": 2.2446933398521973,
+      "learning_rate": 0.00018359585951435045,
+      "loss": 1.7388,
+      "step": 3930
+    },
+    {
+      "epoch": 0.62896,
+      "grad_norm": 2.9957726450776545,
+      "learning_rate": 0.00018358637851496133,
+      "loss": 1.79,
+      "step": 3931
+    },
+    {
+      "epoch": 0.62912,
+      "grad_norm": 2.445899089157214,
+      "learning_rate": 0.0001835768950214764,
+      "loss": 1.8143,
+      "step": 3932
+    },
+    {
+      "epoch": 0.62928,
+      "grad_norm": 3.5437663165289544,
+      "learning_rate": 0.0001835674090341788,
+      "loss": 1.7526,
+      "step": 3933
+    },
+    {
+      "epoch": 0.62944,
+      "grad_norm": 2.2358600819280534,
+      "learning_rate": 0.00018355792055335145,
+      "loss": 1.6769,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6296,
+      "grad_norm": 3.277677565116881,
+      "learning_rate": 0.0001835484295792775,
+      "loss": 1.8569,
+      "step": 3935
+    },
+    {
+      "epoch": 0.62976,
+      "grad_norm": 2.3580699757823016,
+      "learning_rate": 0.00018353893611224012,
+      "loss": 1.7935,
+      "step": 3936
+    },
+    {
+      "epoch": 0.62992,
+      "grad_norm": 2.4498906660638915,
+      "learning_rate": 0.00018352944015252268,
+      "loss": 1.7341,
+      "step": 3937
+    },
+    {
+      "epoch": 0.63008,
+      "grad_norm": 2.073991695651375,
+      "learning_rate": 0.0001835199417004084,
+      "loss": 1.8753,
+      "step": 3938
+    },
+    {
+      "epoch": 0.63024,
+      "grad_norm": 2.312282477922332,
+      "learning_rate": 0.0001835104407561808,
+      "loss": 1.7528,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6304,
+      "grad_norm": 2.3987993288652874,
+      "learning_rate": 0.00018350093732012333,
+      "loss": 1.6999,
+      "step": 3940
+    },
+    {
+      "epoch": 0.63056,
+      "grad_norm": 2.612735657837787,
+      "learning_rate": 0.00018349143139251956,
+      "loss": 1.8046,
+      "step": 3941
+    },
+    {
+      "epoch": 0.63072,
+      "grad_norm": 2.7348737449578633,
+      "learning_rate": 0.00018348192297365314,
+      "loss": 1.7969,
+      "step": 3942
+    },
+    {
+      "epoch": 0.63088,
+      "grad_norm": 2.04933140819564,
+      "learning_rate": 0.00018347241206380778,
+      "loss": 1.731,
+      "step": 3943
+    },
+    {
+      "epoch": 0.63104,
+      "grad_norm": 3.0563197062292122,
+      "learning_rate": 0.0001834628986632673,
+      "loss": 1.7976,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6312,
+      "grad_norm": 2.087301333448516,
+      "learning_rate": 0.0001834533827723155,
+      "loss": 1.7577,
+      "step": 3945
+    },
+    {
+      "epoch": 0.63136,
+      "grad_norm": 2.6904376641690066,
+      "learning_rate": 0.0001834438643912364,
+      "loss": 1.7452,
+      "step": 3946
+    },
+    {
+      "epoch": 0.63152,
+      "grad_norm": 2.4851958536969434,
+      "learning_rate": 0.00018343434352031395,
+      "loss": 1.7997,
+      "step": 3947
+    },
+    {
+      "epoch": 0.63168,
+      "grad_norm": 2.4123624145862737,
+      "learning_rate": 0.00018342482015983226,
+      "loss": 1.7455,
+      "step": 3948
+    },
+    {
+      "epoch": 0.63184,
+      "grad_norm": 2.5904578842821078,
+      "learning_rate": 0.00018341529431007553,
+      "loss": 1.7055,
+      "step": 3949
+    },
+    {
+      "epoch": 0.632,
+      "grad_norm": 2.6801947192260367,
+      "learning_rate": 0.00018340576597132796,
+      "loss": 1.7066,
+      "step": 3950
+    },
+    {
+      "epoch": 0.63216,
+      "grad_norm": 2.5668537255881123,
+      "learning_rate": 0.00018339623514387386,
+      "loss": 1.7461,
+      "step": 3951
+    },
+    {
+      "epoch": 0.63232,
+      "grad_norm": 2.717491323658394,
+      "learning_rate": 0.0001833867018279976,
+      "loss": 1.7951,
+      "step": 3952
+    },
+    {
+      "epoch": 0.63248,
+      "grad_norm": 2.5450519646675556,
+      "learning_rate": 0.0001833771660239837,
+      "loss": 1.7598,
+      "step": 3953
+    },
+    {
+      "epoch": 0.63264,
+      "grad_norm": 2.8660971249131153,
+      "learning_rate": 0.00018336762773211667,
+      "loss": 1.742,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6328,
+      "grad_norm": 2.394457468632069,
+      "learning_rate": 0.00018335808695268106,
+      "loss": 1.7595,
+      "step": 3955
+    },
+    {
+      "epoch": 0.63296,
+      "grad_norm": 2.838014029077957,
+      "learning_rate": 0.00018334854368596166,
+      "loss": 1.7975,
+      "step": 3956
+    },
+    {
+      "epoch": 0.63312,
+      "grad_norm": 2.5522542053055077,
+      "learning_rate": 0.00018333899793224312,
+      "loss": 1.7018,
+      "step": 3957
+    },
+    {
+      "epoch": 0.63328,
+      "grad_norm": 3.032326500671245,
+      "learning_rate": 0.00018332944969181035,
+      "loss": 1.7336,
+      "step": 3958
+    },
+    {
+      "epoch": 0.63344,
+      "grad_norm": 2.6051405410712714,
+      "learning_rate": 0.0001833198989649482,
+      "loss": 1.7198,
+      "step": 3959
+    },
+    {
+      "epoch": 0.6336,
+      "grad_norm": 2.022936119230939,
+      "learning_rate": 0.00018331034575194168,
+      "loss": 1.7123,
+      "step": 3960
+    },
+    {
+      "epoch": 0.63376,
+      "grad_norm": 2.7509046864527136,
+      "learning_rate": 0.00018330079005307584,
+      "loss": 1.801,
+      "step": 3961
+    },
+    {
+      "epoch": 0.63392,
+      "grad_norm": 2.625945490574445,
+      "learning_rate": 0.00018329123186863581,
+      "loss": 1.8389,
+      "step": 3962
+    },
+    {
+      "epoch": 0.63408,
+      "grad_norm": 2.0223628408559873,
+      "learning_rate": 0.0001832816711989068,
+      "loss": 1.8571,
+      "step": 3963
+    },
+    {
+      "epoch": 0.63424,
+      "grad_norm": 2.0772241958217657,
+      "learning_rate": 0.00018327210804417406,
+      "loss": 1.8115,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6344,
+      "grad_norm": 3.401383041545589,
+      "learning_rate": 0.00018326254240472295,
+      "loss": 1.81,
+      "step": 3965
+    },
+    {
+      "epoch": 0.63456,
+      "grad_norm": 2.6073644847728756,
+      "learning_rate": 0.0001832529742808389,
+      "loss": 1.6748,
+      "step": 3966
+    },
+    {
+      "epoch": 0.63472,
+      "grad_norm": 2.188171323647655,
+      "learning_rate": 0.00018324340367280745,
+      "loss": 1.6853,
+      "step": 3967
+    },
+    {
+      "epoch": 0.63488,
+      "grad_norm": 2.6567824717836452,
+      "learning_rate": 0.00018323383058091412,
+      "loss": 1.7435,
+      "step": 3968
+    },
+    {
+      "epoch": 0.63504,
+      "grad_norm": 2.90848460910365,
+      "learning_rate": 0.00018322425500544456,
+      "loss": 1.8438,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6352,
+      "grad_norm": 2.984859014459228,
+      "learning_rate": 0.00018321467694668448,
+      "loss": 1.7521,
+      "step": 3970
+    },
+    {
+      "epoch": 0.63536,
+      "grad_norm": 2.8621591395875683,
+      "learning_rate": 0.00018320509640491972,
+      "loss": 1.688,
+      "step": 3971
+    },
+    {
+      "epoch": 0.63552,
+      "grad_norm": 1.966074613318486,
+      "learning_rate": 0.00018319551338043613,
+      "loss": 1.7278,
+      "step": 3972
+    },
+    {
+      "epoch": 0.63568,
+      "grad_norm": 2.728073965993802,
+      "learning_rate": 0.00018318592787351962,
+      "loss": 1.8554,
+      "step": 3973
+    },
+    {
+      "epoch": 0.63584,
+      "grad_norm": 2.4801350162786897,
+      "learning_rate": 0.00018317633988445626,
+      "loss": 1.7609,
+      "step": 3974
+    },
+    {
+      "epoch": 0.636,
+      "grad_norm": 2.337141360521373,
+      "learning_rate": 0.0001831667494135321,
+      "loss": 1.7307,
+      "step": 3975
+    },
+    {
+      "epoch": 0.63616,
+      "grad_norm": 2.5293724907729835,
+      "learning_rate": 0.00018315715646103335,
+      "loss": 1.6697,
+      "step": 3976
+    },
+    {
+      "epoch": 0.63632,
+      "grad_norm": 2.6484547208401787,
+      "learning_rate": 0.0001831475610272462,
+      "loss": 1.6491,
+      "step": 3977
+    },
+    {
+      "epoch": 0.63648,
+      "grad_norm": 2.9063936965165342,
+      "learning_rate": 0.00018313796311245702,
+      "loss": 1.8251,
+      "step": 3978
+    },
+    {
+      "epoch": 0.63664,
+      "grad_norm": 2.3672266299291684,
+      "learning_rate": 0.00018312836271695216,
+      "loss": 1.7211,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6368,
+      "grad_norm": 2.352910673565735,
+      "learning_rate": 0.00018311875984101806,
+      "loss": 1.8463,
+      "step": 3980
+    },
+    {
+      "epoch": 0.63696,
+      "grad_norm": 2.94390399478437,
+      "learning_rate": 0.0001831091544849413,
+      "loss": 1.8387,
+      "step": 3981
+    },
+    {
+      "epoch": 0.63712,
+      "grad_norm": 2.3865695769348725,
+      "learning_rate": 0.00018309954664900847,
+      "loss": 1.8496,
+      "step": 3982
+    },
+    {
+      "epoch": 0.63728,
+      "grad_norm": 2.5652337017569624,
+      "learning_rate": 0.00018308993633350628,
+      "loss": 1.7819,
+      "step": 3983
+    },
+    {
+      "epoch": 0.63744,
+      "grad_norm": 2.852489038909165,
+      "learning_rate": 0.0001830803235387214,
+      "loss": 1.8052,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6376,
+      "grad_norm": 2.287629342470392,
+      "learning_rate": 0.00018307070826494078,
+      "loss": 1.758,
+      "step": 3985
+    },
+    {
+      "epoch": 0.63776,
+      "grad_norm": 2.346133472797147,
+      "learning_rate": 0.00018306109051245125,
+      "loss": 1.752,
+      "step": 3986
+    },
+    {
+      "epoch": 0.63792,
+      "grad_norm": 2.003112809328744,
+      "learning_rate": 0.00018305147028153983,
+      "loss": 1.7335,
+      "step": 3987
+    },
+    {
+      "epoch": 0.63808,
+      "grad_norm": 2.089566148790766,
+      "learning_rate": 0.00018304184757249355,
+      "loss": 1.719,
+      "step": 3988
+    },
+    {
+      "epoch": 0.63824,
+      "grad_norm": 2.4557562358067764,
+      "learning_rate": 0.00018303222238559954,
+      "loss": 1.8703,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6384,
+      "grad_norm": 2.324234626431796,
+      "learning_rate": 0.00018302259472114502,
+      "loss": 1.658,
+      "step": 3990
+    },
+    {
+      "epoch": 0.63856,
+      "grad_norm": 2.2480914989260077,
+      "learning_rate": 0.00018301296457941724,
+      "loss": 1.7426,
+      "step": 3991
+    },
+    {
+      "epoch": 0.63872,
+      "grad_norm": 2.835350841255376,
+      "learning_rate": 0.00018300333196070355,
+      "loss": 1.8926,
+      "step": 3992
+    },
+    {
+      "epoch": 0.63888,
+      "grad_norm": 2.2429519777466442,
+      "learning_rate": 0.00018299369686529138,
+      "loss": 1.835,
+      "step": 3993
+    },
+    {
+      "epoch": 0.63904,
+      "grad_norm": 2.6224151402679374,
+      "learning_rate": 0.00018298405929346822,
+      "loss": 1.7671,
+      "step": 3994
+    },
+    {
+      "epoch": 0.6392,
+      "grad_norm": 2.8004893201236185,
+      "learning_rate": 0.0001829744192455217,
+      "loss": 1.7311,
+      "step": 3995
+    },
+    {
+      "epoch": 0.63936,
+      "grad_norm": 2.2604194431593,
+      "learning_rate": 0.00018296477672173937,
+      "loss": 1.7435,
+      "step": 3996
+    },
+    {
+      "epoch": 0.63952,
+      "grad_norm": 2.9743796310805948,
+      "learning_rate": 0.00018295513172240902,
+      "loss": 1.8242,
+      "step": 3997
+    },
+    {
+      "epoch": 0.63968,
+      "grad_norm": 1.984022374892304,
+      "learning_rate": 0.00018294548424781842,
+      "loss": 1.6913,
+      "step": 3998
+    },
+    {
+      "epoch": 0.63984,
+      "grad_norm": 2.307870651463064,
+      "learning_rate": 0.0001829358342982554,
+      "loss": 1.8999,
+      "step": 3999
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.663189548443845,
+      "learning_rate": 0.00018292618187400796,
+      "loss": 1.8315,
+      "step": 4000
+    },
+    {
+      "epoch": 0.64016,
+      "grad_norm": 2.006908398109579,
+      "learning_rate": 0.0001829165269753641,
+      "loss": 1.7138,
+      "step": 4001
+    },
+    {
+      "epoch": 0.64032,
+      "grad_norm": 2.1292129394396837,
+      "learning_rate": 0.00018290686960261188,
+      "loss": 1.7596,
+      "step": 4002
+    },
+    {
+      "epoch": 0.64048,
+      "grad_norm": 2.442421261959698,
+      "learning_rate": 0.00018289720975603948,
+      "loss": 1.771,
+      "step": 4003
+    },
+    {
+      "epoch": 0.64064,
+      "grad_norm": 2.451623471352566,
+      "learning_rate": 0.00018288754743593514,
+      "loss": 1.7817,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6408,
+      "grad_norm": 2.3561486393462077,
+      "learning_rate": 0.00018287788264258716,
+      "loss": 1.7831,
+      "step": 4005
+    },
+    {
+      "epoch": 0.64096,
+      "grad_norm": 2.7283221008631973,
+      "learning_rate": 0.0001828682153762839,
+      "loss": 1.7811,
+      "step": 4006
+    },
+    {
+      "epoch": 0.64112,
+      "grad_norm": 2.2140650307173373,
+      "learning_rate": 0.0001828585456373139,
+      "loss": 1.8019,
+      "step": 4007
+    },
+    {
+      "epoch": 0.64128,
+      "grad_norm": 2.3258849001446174,
+      "learning_rate": 0.0001828488734259656,
+      "loss": 1.7751,
+      "step": 4008
+    },
+    {
+      "epoch": 0.64144,
+      "grad_norm": 2.299833831303794,
+      "learning_rate": 0.00018283919874252766,
+      "loss": 1.814,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6416,
+      "grad_norm": 1.9595148056898317,
+      "learning_rate": 0.00018282952158728873,
+      "loss": 1.6515,
+      "step": 4010
+    },
+    {
+      "epoch": 0.64176,
+      "grad_norm": 2.442426007801907,
+      "learning_rate": 0.00018281984196053758,
+      "loss": 1.797,
+      "step": 4011
+    },
+    {
+      "epoch": 0.64192,
+      "grad_norm": 2.165694062907112,
+      "learning_rate": 0.000182810159862563,
+      "loss": 1.753,
+      "step": 4012
+    },
+    {
+      "epoch": 0.64208,
+      "grad_norm": 2.4821265849138143,
+      "learning_rate": 0.00018280047529365396,
+      "loss": 1.7592,
+      "step": 4013
+    },
+    {
+      "epoch": 0.64224,
+      "grad_norm": 2.11877049587336,
+      "learning_rate": 0.00018279078825409937,
+      "loss": 1.6859,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6424,
+      "grad_norm": 2.2835844306627373,
+      "learning_rate": 0.0001827810987441883,
+      "loss": 1.6745,
+      "step": 4015
+    },
+    {
+      "epoch": 0.64256,
+      "grad_norm": 2.7220107726182072,
+      "learning_rate": 0.00018277140676420986,
+      "loss": 1.7754,
+      "step": 4016
+    },
+    {
+      "epoch": 0.64272,
+      "grad_norm": 2.277060576714843,
+      "learning_rate": 0.00018276171231445325,
+      "loss": 1.6417,
+      "step": 4017
+    },
+    {
+      "epoch": 0.64288,
+      "grad_norm": 2.549476459241694,
+      "learning_rate": 0.00018275201539520773,
+      "loss": 1.7351,
+      "step": 4018
+    },
+    {
+      "epoch": 0.64304,
+      "grad_norm": 2.158932276145997,
+      "learning_rate": 0.00018274231600676267,
+      "loss": 1.7716,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6432,
+      "grad_norm": 2.801892344037783,
+      "learning_rate": 0.00018273261414940748,
+      "loss": 1.8178,
+      "step": 4020
+    },
+    {
+      "epoch": 0.64336,
+      "grad_norm": 2.2334781440504,
+      "learning_rate": 0.00018272290982343164,
+      "loss": 1.7164,
+      "step": 4021
+    },
+    {
+      "epoch": 0.64352,
+      "grad_norm": 2.4331148619119287,
+      "learning_rate": 0.00018271320302912466,
+      "loss": 1.7418,
+      "step": 4022
+    },
+    {
+      "epoch": 0.64368,
+      "grad_norm": 2.796111221961187,
+      "learning_rate": 0.0001827034937667763,
+      "loss": 1.7693,
+      "step": 4023
+    },
+    {
+      "epoch": 0.64384,
+      "grad_norm": 2.5020312943747887,
+      "learning_rate": 0.00018269378203667616,
+      "loss": 1.6894,
+      "step": 4024
+    },
+    {
+      "epoch": 0.644,
+      "grad_norm": 2.632225093309004,
+      "learning_rate": 0.00018268406783911406,
+      "loss": 1.831,
+      "step": 4025
+    },
+    {
+      "epoch": 0.64416,
+      "grad_norm": 2.52860034649152,
+      "learning_rate": 0.00018267435117437985,
+      "loss": 1.8182,
+      "step": 4026
+    },
+    {
+      "epoch": 0.64432,
+      "grad_norm": 2.5561036153555503,
+      "learning_rate": 0.0001826646320427635,
+      "loss": 1.8127,
+      "step": 4027
+    },
+    {
+      "epoch": 0.64448,
+      "grad_norm": 2.6201095942171464,
+      "learning_rate": 0.00018265491044455496,
+      "loss": 1.7379,
+      "step": 4028
+    },
+    {
+      "epoch": 0.64464,
+      "grad_norm": 2.381597051970297,
+      "learning_rate": 0.00018264518638004434,
+      "loss": 1.7596,
+      "step": 4029
+    },
+    {
+      "epoch": 0.6448,
+      "grad_norm": 2.430197648858659,
+      "learning_rate": 0.00018263545984952177,
+      "loss": 1.7255,
+      "step": 4030
+    },
+    {
+      "epoch": 0.64496,
+      "grad_norm": 2.3112946380787274,
+      "learning_rate": 0.0001826257308532775,
+      "loss": 1.8286,
+      "step": 4031
+    },
+    {
+      "epoch": 0.64512,
+      "grad_norm": 3.3310286224843213,
+      "learning_rate": 0.00018261599939160185,
+      "loss": 1.8763,
+      "step": 4032
+    },
+    {
+      "epoch": 0.64528,
+      "grad_norm": 2.529200939205646,
+      "learning_rate": 0.00018260626546478514,
+      "loss": 1.8038,
+      "step": 4033
+    },
+    {
+      "epoch": 0.64544,
+      "grad_norm": 2.715055056815836,
+      "learning_rate": 0.00018259652907311785,
+      "loss": 1.7873,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6456,
+      "grad_norm": 2.209673449391739,
+      "learning_rate": 0.00018258679021689045,
+      "loss": 1.7416,
+      "step": 4035
+    },
+    {
+      "epoch": 0.64576,
+      "grad_norm": 2.216066145674316,
+      "learning_rate": 0.0001825770488963936,
+      "loss": 1.7832,
+      "step": 4036
+    },
+    {
+      "epoch": 0.64592,
+      "grad_norm": 2.2997550973119187,
+      "learning_rate": 0.00018256730511191792,
+      "loss": 1.8201,
+      "step": 4037
+    },
+    {
+      "epoch": 0.64608,
+      "grad_norm": 2.7466689878791724,
+      "learning_rate": 0.0001825575588637542,
+      "loss": 1.8079,
+      "step": 4038
+    },
+    {
+      "epoch": 0.64624,
+      "grad_norm": 2.3133365012334814,
+      "learning_rate": 0.00018254781015219321,
+      "loss": 1.7112,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6464,
+      "grad_norm": 2.5818954852198526,
+      "learning_rate": 0.00018253805897752585,
+      "loss": 1.7817,
+      "step": 4040
+    },
+    {
+      "epoch": 0.64656,
+      "grad_norm": 3.0066491634194503,
+      "learning_rate": 0.00018252830534004307,
+      "loss": 1.7938,
+      "step": 4041
+    },
+    {
+      "epoch": 0.64672,
+      "grad_norm": 2.1644643629750724,
+      "learning_rate": 0.00018251854924003594,
+      "loss": 1.6856,
+      "step": 4042
+    },
+    {
+      "epoch": 0.64688,
+      "grad_norm": 2.1535001936968055,
+      "learning_rate": 0.0001825087906777955,
+      "loss": 1.6789,
+      "step": 4043
+    },
+    {
+      "epoch": 0.64704,
+      "grad_norm": 2.503999481784546,
+      "learning_rate": 0.000182499029653613,
+      "loss": 1.7413,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6472,
+      "grad_norm": 2.667462519934095,
+      "learning_rate": 0.0001824892661677797,
+      "loss": 1.6594,
+      "step": 4045
+    },
+    {
+      "epoch": 0.64736,
+      "grad_norm": 2.389583685481015,
+      "learning_rate": 0.00018247950022058683,
+      "loss": 1.8488,
+      "step": 4046
+    },
+    {
+      "epoch": 0.64752,
+      "grad_norm": 2.0129853919283134,
+      "learning_rate": 0.0001824697318123259,
+      "loss": 1.7881,
+      "step": 4047
+    },
+    {
+      "epoch": 0.64768,
+      "grad_norm": 2.5294495221464937,
+      "learning_rate": 0.00018245996094328834,
+      "loss": 1.7428,
+      "step": 4048
+    },
+    {
+      "epoch": 0.64784,
+      "grad_norm": 2.227940600596213,
+      "learning_rate": 0.0001824501876137657,
+      "loss": 1.7841,
+      "step": 4049
+    },
+    {
+      "epoch": 0.648,
+      "grad_norm": 2.476476296245851,
+      "learning_rate": 0.0001824404118240496,
+      "loss": 1.8047,
+      "step": 4050
+    },
+    {
+      "epoch": 0.64816,
+      "grad_norm": 2.693026567399526,
+      "learning_rate": 0.00018243063357443177,
+      "loss": 1.7541,
+      "step": 4051
+    },
+    {
+      "epoch": 0.64832,
+      "grad_norm": 2.234275964062659,
+      "learning_rate": 0.00018242085286520392,
+      "loss": 1.7064,
+      "step": 4052
+    },
+    {
+      "epoch": 0.64848,
+      "grad_norm": 2.511441807624459,
+      "learning_rate": 0.00018241106969665792,
+      "loss": 1.772,
+      "step": 4053
+    },
+    {
+      "epoch": 0.64864,
+      "grad_norm": 2.4807390757381893,
+      "learning_rate": 0.00018240128406908567,
+      "loss": 1.7688,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6488,
+      "grad_norm": 2.6189082329814504,
+      "learning_rate": 0.00018239149598277922,
+      "loss": 1.7394,
+      "step": 4055
+    },
+    {
+      "epoch": 0.64896,
+      "grad_norm": 2.8495863779318564,
+      "learning_rate": 0.00018238170543803054,
+      "loss": 1.7868,
+      "step": 4056
+    },
+    {
+      "epoch": 0.64912,
+      "grad_norm": 2.1679182980279443,
+      "learning_rate": 0.00018237191243513184,
+      "loss": 1.6813,
+      "step": 4057
+    },
+    {
+      "epoch": 0.64928,
+      "grad_norm": 2.2705409333289808,
+      "learning_rate": 0.0001823621169743753,
+      "loss": 1.7925,
+      "step": 4058
+    },
+    {
+      "epoch": 0.64944,
+      "grad_norm": 2.417898027344546,
+      "learning_rate": 0.00018235231905605322,
+      "loss": 1.6865,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6496,
+      "grad_norm": 2.4681534713582742,
+      "learning_rate": 0.00018234251868045787,
+      "loss": 1.8669,
+      "step": 4060
+    },
+    {
+      "epoch": 0.64976,
+      "grad_norm": 2.915164016456882,
+      "learning_rate": 0.00018233271584788184,
+      "loss": 1.7756,
+      "step": 4061
+    },
+    {
+      "epoch": 0.64992,
+      "grad_norm": 2.128569881279019,
+      "learning_rate": 0.00018232291055861747,
+      "loss": 1.7918,
+      "step": 4062
+    },
+    {
+      "epoch": 0.65008,
+      "grad_norm": 2.130565990486457,
+      "learning_rate": 0.00018231310281295745,
+      "loss": 1.7282,
+      "step": 4063
+    },
+    {
+      "epoch": 0.65024,
+      "grad_norm": 2.126011001891583,
+      "learning_rate": 0.00018230329261119437,
+      "loss": 1.7349,
+      "step": 4064
+    },
+    {
+      "epoch": 0.6504,
+      "grad_norm": 2.66173759989954,
+      "learning_rate": 0.00018229347995362094,
+      "loss": 1.7581,
+      "step": 4065
+    },
+    {
+      "epoch": 0.65056,
+      "grad_norm": 2.3640291239977937,
+      "learning_rate": 0.00018228366484053003,
+      "loss": 1.8338,
+      "step": 4066
+    },
+    {
+      "epoch": 0.65072,
+      "grad_norm": 2.222220838973321,
+      "learning_rate": 0.00018227384727221443,
+      "loss": 1.8568,
+      "step": 4067
+    },
+    {
+      "epoch": 0.65088,
+      "grad_norm": 2.236657714163993,
+      "learning_rate": 0.00018226402724896712,
+      "loss": 1.7486,
+      "step": 4068
+    },
+    {
+      "epoch": 0.65104,
+      "grad_norm": 2.1191922005461783,
+      "learning_rate": 0.00018225420477108113,
+      "loss": 1.7614,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6512,
+      "grad_norm": 2.375733663704465,
+      "learning_rate": 0.00018224437983884947,
+      "loss": 1.7501,
+      "step": 4070
+    },
+    {
+      "epoch": 0.65136,
+      "grad_norm": 2.6129668008773828,
+      "learning_rate": 0.0001822345524525654,
+      "loss": 1.7245,
+      "step": 4071
+    },
+    {
+      "epoch": 0.65152,
+      "grad_norm": 2.255248450987644,
+      "learning_rate": 0.00018222472261252213,
+      "loss": 1.781,
+      "step": 4072
+    },
+    {
+      "epoch": 0.65168,
+      "grad_norm": 2.2206462974758536,
+      "learning_rate": 0.00018221489031901295,
+      "loss": 1.7424,
+      "step": 4073
+    },
+    {
+      "epoch": 0.65184,
+      "grad_norm": 2.595111912488735,
+      "learning_rate": 0.0001822050555723312,
+      "loss": 1.7039,
+      "step": 4074
+    },
+    {
+      "epoch": 0.652,
+      "grad_norm": 2.4096832436026236,
+      "learning_rate": 0.00018219521837277037,
+      "loss": 1.7429,
+      "step": 4075
+    },
+    {
+      "epoch": 0.65216,
+      "grad_norm": 2.2881074013931504,
+      "learning_rate": 0.00018218537872062405,
+      "loss": 1.771,
+      "step": 4076
+    },
+    {
+      "epoch": 0.65232,
+      "grad_norm": 2.5393325140551553,
+      "learning_rate": 0.00018217553661618576,
+      "loss": 1.7805,
+      "step": 4077
+    },
+    {
+      "epoch": 0.65248,
+      "grad_norm": 2.4140057281265905,
+      "learning_rate": 0.00018216569205974922,
+      "loss": 1.7505,
+      "step": 4078
+    },
+    {
+      "epoch": 0.65264,
+      "grad_norm": 2.644182436495167,
+      "learning_rate": 0.0001821558450516081,
+      "loss": 1.7845,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6528,
+      "grad_norm": 2.458582621988876,
+      "learning_rate": 0.00018214599559205632,
+      "loss": 1.7646,
+      "step": 4080
+    },
+    {
+      "epoch": 0.65296,
+      "grad_norm": 3.117627996610232,
+      "learning_rate": 0.00018213614368138773,
+      "loss": 1.8002,
+      "step": 4081
+    },
+    {
+      "epoch": 0.65312,
+      "grad_norm": 2.089827123817035,
+      "learning_rate": 0.00018212628931989629,
+      "loss": 1.7217,
+      "step": 4082
+    },
+    {
+      "epoch": 0.65328,
+      "grad_norm": 2.5223123770515716,
+      "learning_rate": 0.00018211643250787605,
+      "loss": 1.7909,
+      "step": 4083
+    },
+    {
+      "epoch": 0.65344,
+      "grad_norm": 2.9431890999342736,
+      "learning_rate": 0.00018210657324562113,
+      "loss": 1.7634,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6536,
+      "grad_norm": 2.261940356006971,
+      "learning_rate": 0.00018209671153342568,
+      "loss": 1.7712,
+      "step": 4085
+    },
+    {
+      "epoch": 0.65376,
+      "grad_norm": 2.6283998966215827,
+      "learning_rate": 0.000182086847371584,
+      "loss": 1.6945,
+      "step": 4086
+    },
+    {
+      "epoch": 0.65392,
+      "grad_norm": 2.7957309335727802,
+      "learning_rate": 0.00018207698076039042,
+      "loss": 1.8117,
+      "step": 4087
+    },
+    {
+      "epoch": 0.65408,
+      "grad_norm": 2.331678368546447,
+      "learning_rate": 0.00018206711170013932,
+      "loss": 1.8674,
+      "step": 4088
+    },
+    {
+      "epoch": 0.65424,
+      "grad_norm": 2.4727800886840194,
+      "learning_rate": 0.0001820572401911252,
+      "loss": 1.7096,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6544,
+      "grad_norm": 2.6527485811638036,
+      "learning_rate": 0.0001820473662336426,
+      "loss": 1.7126,
+      "step": 4090
+    },
+    {
+      "epoch": 0.65456,
+      "grad_norm": 3.2305537260063892,
+      "learning_rate": 0.00018203748982798613,
+      "loss": 1.6886,
+      "step": 4091
+    },
+    {
+      "epoch": 0.65472,
+      "grad_norm": 2.4054644598245845,
+      "learning_rate": 0.00018202761097445053,
+      "loss": 1.6636,
+      "step": 4092
+    },
+    {
+      "epoch": 0.65488,
+      "grad_norm": 3.269055028345457,
+      "learning_rate": 0.00018201772967333052,
+      "loss": 1.7549,
+      "step": 4093
+    },
+    {
+      "epoch": 0.65504,
+      "grad_norm": 2.2052178896416614,
+      "learning_rate": 0.000182007845924921,
+      "loss": 1.6748,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6552,
+      "grad_norm": 2.753175857229788,
+      "learning_rate": 0.00018199795972951682,
+      "loss": 1.7521,
+      "step": 4095
+    },
+    {
+      "epoch": 0.65536,
+      "grad_norm": 2.385574154795833,
+      "learning_rate": 0.00018198807108741304,
+      "loss": 1.8928,
+      "step": 4096
+    },
+    {
+      "epoch": 0.65552,
+      "grad_norm": 2.2670400959326806,
+      "learning_rate": 0.00018197817999890465,
+      "loss": 1.8152,
+      "step": 4097
+    },
+    {
+      "epoch": 0.65568,
+      "grad_norm": 3.08279795177092,
+      "learning_rate": 0.00018196828646428688,
+      "loss": 1.7388,
+      "step": 4098
+    },
+    {
+      "epoch": 0.65584,
+      "grad_norm": 2.2512434584874437,
+      "learning_rate": 0.00018195839048385483,
+      "loss": 1.7568,
+      "step": 4099
+    },
+    {
+      "epoch": 0.656,
+      "grad_norm": 3.2660013481892465,
+      "learning_rate": 0.00018194849205790386,
+      "loss": 1.8183,
+      "step": 4100
+    },
+    {
+      "epoch": 0.65616,
+      "grad_norm": 2.6769560399990953,
+      "learning_rate": 0.0001819385911867293,
+      "loss": 1.7869,
+      "step": 4101
+    },
+    {
+      "epoch": 0.65632,
+      "grad_norm": 3.072163726991119,
+      "learning_rate": 0.00018192868787062654,
+      "loss": 1.7199,
+      "step": 4102
+    },
+    {
+      "epoch": 0.65648,
+      "grad_norm": 2.2683758991323035,
+      "learning_rate": 0.00018191878210989116,
+      "loss": 1.773,
+      "step": 4103
+    },
+    {
+      "epoch": 0.65664,
+      "grad_norm": 2.982753713615425,
+      "learning_rate": 0.00018190887390481865,
+      "loss": 1.7424,
+      "step": 4104
+    },
+    {
+      "epoch": 0.6568,
+      "grad_norm": 2.7400363316767233,
+      "learning_rate": 0.00018189896325570475,
+      "loss": 1.7575,
+      "step": 4105
+    },
+    {
+      "epoch": 0.65696,
+      "grad_norm": 2.9018159464592372,
+      "learning_rate": 0.0001818890501628451,
+      "loss": 1.7206,
+      "step": 4106
+    },
+    {
+      "epoch": 0.65712,
+      "grad_norm": 2.4598279235125604,
+      "learning_rate": 0.0001818791346265355,
+      "loss": 1.7703,
+      "step": 4107
+    },
+    {
+      "epoch": 0.65728,
+      "grad_norm": 2.632524682024661,
+      "learning_rate": 0.00018186921664707184,
+      "loss": 1.7739,
+      "step": 4108
+    },
+    {
+      "epoch": 0.65744,
+      "grad_norm": 2.323807138446457,
+      "learning_rate": 0.00018185929622475006,
+      "loss": 1.6996,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6576,
+      "grad_norm": 2.5891569948283224,
+      "learning_rate": 0.00018184937335986615,
+      "loss": 1.7681,
+      "step": 4110
+    },
+    {
+      "epoch": 0.65776,
+      "grad_norm": 2.370482055354385,
+      "learning_rate": 0.00018183944805271623,
+      "loss": 1.8148,
+      "step": 4111
+    },
+    {
+      "epoch": 0.65792,
+      "grad_norm": 2.6371790243232973,
+      "learning_rate": 0.0001818295203035964,
+      "loss": 1.6885,
+      "step": 4112
+    },
+    {
+      "epoch": 0.65808,
+      "grad_norm": 2.37116779629792,
+      "learning_rate": 0.00018181959011280294,
+      "loss": 1.7172,
+      "step": 4113
+    },
+    {
+      "epoch": 0.65824,
+      "grad_norm": 2.067435256235846,
+      "learning_rate": 0.00018180965748063215,
+      "loss": 1.6719,
+      "step": 4114
+    },
+    {
+      "epoch": 0.6584,
+      "grad_norm": 2.5940854409026324,
+      "learning_rate": 0.00018179972240738037,
+      "loss": 1.707,
+      "step": 4115
+    },
+    {
+      "epoch": 0.65856,
+      "grad_norm": 2.239674111324054,
+      "learning_rate": 0.00018178978489334408,
+      "loss": 1.7352,
+      "step": 4116
+    },
+    {
+      "epoch": 0.65872,
+      "grad_norm": 2.694130822872592,
+      "learning_rate": 0.00018177984493881978,
+      "loss": 1.8361,
+      "step": 4117
+    },
+    {
+      "epoch": 0.65888,
+      "grad_norm": 2.549229339992599,
+      "learning_rate": 0.0001817699025441041,
+      "loss": 1.8019,
+      "step": 4118
+    },
+    {
+      "epoch": 0.65904,
+      "grad_norm": 2.7146452133263654,
+      "learning_rate": 0.00018175995770949364,
+      "loss": 1.7967,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6592,
+      "grad_norm": 2.433205346728389,
+      "learning_rate": 0.00018175001043528518,
+      "loss": 1.7741,
+      "step": 4120
+    },
+    {
+      "epoch": 0.65936,
+      "grad_norm": 2.2799769398576832,
+      "learning_rate": 0.00018174006072177554,
+      "loss": 1.7437,
+      "step": 4121
+    },
+    {
+      "epoch": 0.65952,
+      "grad_norm": 2.146864172893088,
+      "learning_rate": 0.00018173010856926164,
+      "loss": 1.7387,
+      "step": 4122
+    },
+    {
+      "epoch": 0.65968,
+      "grad_norm": 2.4829565701933642,
+      "learning_rate": 0.00018172015397804035,
+      "loss": 1.7707,
+      "step": 4123
+    },
+    {
+      "epoch": 0.65984,
+      "grad_norm": 2.3201926370747104,
+      "learning_rate": 0.00018171019694840875,
+      "loss": 1.7209,
+      "step": 4124
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 2.8828310173288467,
+      "learning_rate": 0.0001817002374806639,
+      "loss": 1.7722,
+      "step": 4125
+    },
+    {
+      "epoch": 0.66016,
+      "grad_norm": 2.188418937095823,
+      "learning_rate": 0.00018169027557510305,
+      "loss": 1.6684,
+      "step": 4126
+    },
+    {
+      "epoch": 0.66032,
+      "grad_norm": 2.92482423256392,
+      "learning_rate": 0.00018168031123202342,
+      "loss": 1.8091,
+      "step": 4127
+    },
+    {
+      "epoch": 0.66048,
+      "grad_norm": 2.606172795967608,
+      "learning_rate": 0.00018167034445172234,
+      "loss": 1.7718,
+      "step": 4128
+    },
+    {
+      "epoch": 0.66064,
+      "grad_norm": 2.59199952046704,
+      "learning_rate": 0.00018166037523449714,
+      "loss": 1.738,
+      "step": 4129
+    },
+    {
+      "epoch": 0.6608,
+      "grad_norm": 2.2476780218397763,
+      "learning_rate": 0.00018165040358064535,
+      "loss": 1.6949,
+      "step": 4130
+    },
+    {
+      "epoch": 0.66096,
+      "grad_norm": 2.2407497882395315,
+      "learning_rate": 0.00018164042949046452,
+      "loss": 1.7834,
+      "step": 4131
+    },
+    {
+      "epoch": 0.66112,
+      "grad_norm": 2.2916757539806936,
+      "learning_rate": 0.00018163045296425224,
+      "loss": 1.7516,
+      "step": 4132
+    },
+    {
+      "epoch": 0.66128,
+      "grad_norm": 2.2437005384066406,
+      "learning_rate": 0.00018162047400230614,
+      "loss": 1.8202,
+      "step": 4133
+    },
+    {
+      "epoch": 0.66144,
+      "grad_norm": 2.365274591329446,
+      "learning_rate": 0.00018161049260492407,
+      "loss": 1.7583,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6616,
+      "grad_norm": 2.2770993370673263,
+      "learning_rate": 0.00018160050877240381,
+      "loss": 1.8185,
+      "step": 4135
+    },
+    {
+      "epoch": 0.66176,
+      "grad_norm": 2.769230262838687,
+      "learning_rate": 0.0001815905225050433,
+      "loss": 1.6698,
+      "step": 4136
+    },
+    {
+      "epoch": 0.66192,
+      "grad_norm": 2.3648971614567182,
+      "learning_rate": 0.00018158053380314045,
+      "loss": 1.7381,
+      "step": 4137
+    },
+    {
+      "epoch": 0.66208,
+      "grad_norm": 2.6478314618196914,
+      "learning_rate": 0.00018157054266699336,
+      "loss": 1.7298,
+      "step": 4138
+    },
+    {
+      "epoch": 0.66224,
+      "grad_norm": 2.4028584409164324,
+      "learning_rate": 0.00018156054909690013,
+      "loss": 1.7104,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6624,
+      "grad_norm": 2.4292821648827925,
+      "learning_rate": 0.00018155055309315897,
+      "loss": 1.7836,
+      "step": 4140
+    },
+    {
+      "epoch": 0.66256,
+      "grad_norm": 2.355397975995207,
+      "learning_rate": 0.00018154055465606814,
+      "loss": 1.8558,
+      "step": 4141
+    },
+    {
+      "epoch": 0.66272,
+      "grad_norm": 2.517457523141124,
+      "learning_rate": 0.00018153055378592595,
+      "loss": 1.8035,
+      "step": 4142
+    },
+    {
+      "epoch": 0.66288,
+      "grad_norm": 2.54550451103212,
+      "learning_rate": 0.00018152055048303086,
+      "loss": 1.6975,
+      "step": 4143
+    },
+    {
+      "epoch": 0.66304,
+      "grad_norm": 2.804848823798977,
+      "learning_rate": 0.00018151054474768133,
+      "loss": 1.8582,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6632,
+      "grad_norm": 2.757277926430398,
+      "learning_rate": 0.00018150053658017593,
+      "loss": 1.7408,
+      "step": 4145
+    },
+    {
+      "epoch": 0.66336,
+      "grad_norm": 3.1003558644498197,
+      "learning_rate": 0.00018149052598081326,
+      "loss": 1.809,
+      "step": 4146
+    },
+    {
+      "epoch": 0.66352,
+      "grad_norm": 2.9660079456957718,
+      "learning_rate": 0.00018148051294989204,
+      "loss": 1.729,
+      "step": 4147
+    },
+    {
+      "epoch": 0.66368,
+      "grad_norm": 2.860693597152219,
+      "learning_rate": 0.00018147049748771106,
+      "loss": 1.681,
+      "step": 4148
+    },
+    {
+      "epoch": 0.66384,
+      "grad_norm": 2.5268792820582386,
+      "learning_rate": 0.00018146047959456912,
+      "loss": 1.6887,
+      "step": 4149
+    },
+    {
+      "epoch": 0.664,
+      "grad_norm": 2.282893302400664,
+      "learning_rate": 0.00018145045927076518,
+      "loss": 1.7209,
+      "step": 4150
+    },
+    {
+      "epoch": 0.66416,
+      "grad_norm": 2.30232856697325,
+      "learning_rate": 0.00018144043651659826,
+      "loss": 1.7461,
+      "step": 4151
+    },
+    {
+      "epoch": 0.66432,
+      "grad_norm": 2.359380890194696,
+      "learning_rate": 0.00018143041133236734,
+      "loss": 1.716,
+      "step": 4152
+    },
+    {
+      "epoch": 0.66448,
+      "grad_norm": 2.6204708694540195,
+      "learning_rate": 0.00018142038371837162,
+      "loss": 1.722,
+      "step": 4153
+    },
+    {
+      "epoch": 0.66464,
+      "grad_norm": 2.094980966177822,
+      "learning_rate": 0.00018141035367491028,
+      "loss": 1.7721,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6648,
+      "grad_norm": 2.407773387660773,
+      "learning_rate": 0.00018140032120228262,
+      "loss": 1.8011,
+      "step": 4155
+    },
+    {
+      "epoch": 0.66496,
+      "grad_norm": 2.497045021633959,
+      "learning_rate": 0.000181390286300788,
+      "loss": 1.7934,
+      "step": 4156
+    },
+    {
+      "epoch": 0.66512,
+      "grad_norm": 2.2407237795745183,
+      "learning_rate": 0.00018138024897072582,
+      "loss": 1.7019,
+      "step": 4157
+    },
+    {
+      "epoch": 0.66528,
+      "grad_norm": 2.947830388847961,
+      "learning_rate": 0.00018137020921239557,
+      "loss": 1.7846,
+      "step": 4158
+    },
+    {
+      "epoch": 0.66544,
+      "grad_norm": 2.549767519863699,
+      "learning_rate": 0.0001813601670260969,
+      "loss": 1.7999,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6656,
+      "grad_norm": 2.521270173632819,
+      "learning_rate": 0.0001813501224121294,
+      "loss": 1.8414,
+      "step": 4160
+    },
+    {
+      "epoch": 0.66576,
+      "grad_norm": 2.5678090238010314,
+      "learning_rate": 0.00018134007537079275,
+      "loss": 1.7379,
+      "step": 4161
+    },
+    {
+      "epoch": 0.66592,
+      "grad_norm": 2.572069153006768,
+      "learning_rate": 0.00018133002590238676,
+      "loss": 1.7792,
+      "step": 4162
+    },
+    {
+      "epoch": 0.66608,
+      "grad_norm": 2.604442178678226,
+      "learning_rate": 0.00018131997400721135,
+      "loss": 1.7875,
+      "step": 4163
+    },
+    {
+      "epoch": 0.66624,
+      "grad_norm": 2.269993062799941,
+      "learning_rate": 0.0001813099196855664,
+      "loss": 1.6049,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6664,
+      "grad_norm": 2.69568392863177,
+      "learning_rate": 0.00018129986293775194,
+      "loss": 1.7336,
+      "step": 4165
+    },
+    {
+      "epoch": 0.66656,
+      "grad_norm": 2.183856778052499,
+      "learning_rate": 0.000181289803764068,
+      "loss": 1.7203,
+      "step": 4166
+    },
+    {
+      "epoch": 0.66672,
+      "grad_norm": 2.569531423692532,
+      "learning_rate": 0.0001812797421648148,
+      "loss": 1.749,
+      "step": 4167
+    },
+    {
+      "epoch": 0.66688,
+      "grad_norm": 2.3299369519685187,
+      "learning_rate": 0.00018126967814029253,
+      "loss": 1.7185,
+      "step": 4168
+    },
+    {
+      "epoch": 0.66704,
+      "grad_norm": 2.3334456642457457,
+      "learning_rate": 0.00018125961169080147,
+      "loss": 1.6712,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6672,
+      "grad_norm": 2.9758414069859924,
+      "learning_rate": 0.000181249542816642,
+      "loss": 1.843,
+      "step": 4170
+    },
+    {
+      "epoch": 0.66736,
+      "grad_norm": 2.5814638638881466,
+      "learning_rate": 0.00018123947151811457,
+      "loss": 1.779,
+      "step": 4171
+    },
+    {
+      "epoch": 0.66752,
+      "grad_norm": 2.5828955417483033,
+      "learning_rate": 0.0001812293977955197,
+      "loss": 1.735,
+      "step": 4172
+    },
+    {
+      "epoch": 0.66768,
+      "grad_norm": 3.2480040008415685,
+      "learning_rate": 0.00018121932164915795,
+      "loss": 1.7691,
+      "step": 4173
+    },
+    {
+      "epoch": 0.66784,
+      "grad_norm": 2.444311315978393,
+      "learning_rate": 0.00018120924307933,
+      "loss": 1.664,
+      "step": 4174
+    },
+    {
+      "epoch": 0.668,
+      "grad_norm": 2.181996327128147,
+      "learning_rate": 0.0001811991620863366,
+      "loss": 1.7508,
+      "step": 4175
+    },
+    {
+      "epoch": 0.66816,
+      "grad_norm": 2.5069984288798937,
+      "learning_rate": 0.0001811890786704785,
+      "loss": 1.6802,
+      "step": 4176
+    },
+    {
+      "epoch": 0.66832,
+      "grad_norm": 2.5923337464351244,
+      "learning_rate": 0.00018117899283205658,
+      "loss": 1.6675,
+      "step": 4177
+    },
+    {
+      "epoch": 0.66848,
+      "grad_norm": 2.20378459918076,
+      "learning_rate": 0.00018116890457137182,
+      "loss": 1.8173,
+      "step": 4178
+    },
+    {
+      "epoch": 0.66864,
+      "grad_norm": 2.510266296495825,
+      "learning_rate": 0.0001811588138887252,
+      "loss": 1.7833,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6688,
+      "grad_norm": 2.0481736863994144,
+      "learning_rate": 0.00018114872078441784,
+      "loss": 1.7386,
+      "step": 4180
+    },
+    {
+      "epoch": 0.66896,
+      "grad_norm": 2.3827439716766246,
+      "learning_rate": 0.00018113862525875093,
+      "loss": 1.6765,
+      "step": 4181
+    },
+    {
+      "epoch": 0.66912,
+      "grad_norm": 2.225979022971787,
+      "learning_rate": 0.00018112852731202566,
+      "loss": 1.7569,
+      "step": 4182
+    },
+    {
+      "epoch": 0.66928,
+      "grad_norm": 2.130228121617639,
+      "learning_rate": 0.00018111842694454335,
+      "loss": 1.6868,
+      "step": 4183
+    },
+    {
+      "epoch": 0.66944,
+      "grad_norm": 2.1042602157030683,
+      "learning_rate": 0.00018110832415660537,
+      "loss": 1.7761,
+      "step": 4184
+    },
+    {
+      "epoch": 0.6696,
+      "grad_norm": 2.60553449758409,
+      "learning_rate": 0.0001810982189485132,
+      "loss": 1.7297,
+      "step": 4185
+    },
+    {
+      "epoch": 0.66976,
+      "grad_norm": 2.432734856282604,
+      "learning_rate": 0.00018108811132056834,
+      "loss": 1.6949,
+      "step": 4186
+    },
+    {
+      "epoch": 0.66992,
+      "grad_norm": 2.4325566260342515,
+      "learning_rate": 0.0001810780012730724,
+      "loss": 1.7706,
+      "step": 4187
+    },
+    {
+      "epoch": 0.67008,
+      "grad_norm": 2.6001053385441972,
+      "learning_rate": 0.00018106788880632705,
+      "loss": 1.7081,
+      "step": 4188
+    },
+    {
+      "epoch": 0.67024,
+      "grad_norm": 2.780115052686373,
+      "learning_rate": 0.000181057773920634,
+      "loss": 1.7817,
+      "step": 4189
+    },
+    {
+      "epoch": 0.6704,
+      "grad_norm": 2.4057815770920117,
+      "learning_rate": 0.0001810476566162951,
+      "loss": 1.6597,
+      "step": 4190
+    },
+    {
+      "epoch": 0.67056,
+      "grad_norm": 2.4599894816059877,
+      "learning_rate": 0.00018103753689361225,
+      "loss": 1.8085,
+      "step": 4191
+    },
+    {
+      "epoch": 0.67072,
+      "grad_norm": 2.5395674888319424,
+      "learning_rate": 0.0001810274147528874,
+      "loss": 1.7932,
+      "step": 4192
+    },
+    {
+      "epoch": 0.67088,
+      "grad_norm": 2.623759179855902,
+      "learning_rate": 0.00018101729019442248,
+      "loss": 1.7155,
+      "step": 4193
+    },
+    {
+      "epoch": 0.67104,
+      "grad_norm": 2.5343496551942097,
+      "learning_rate": 0.00018100716321851974,
+      "loss": 1.7009,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6712,
+      "grad_norm": 2.5343496551942097,
+      "learning_rate": 0.00018100716321851974,
+      "loss": 1.8576,
+      "step": 4195
+    },
+    {
+      "epoch": 0.67136,
+      "grad_norm": 2.3003025583152046,
+      "learning_rate": 0.00018099703382548128,
+      "loss": 1.615,
+      "step": 4196
+    },
+    {
+      "epoch": 0.67152,
+      "grad_norm": 2.0452856653791702,
+      "learning_rate": 0.00018098690201560935,
+      "loss": 1.7455,
+      "step": 4197
+    },
+    {
+      "epoch": 0.67168,
+      "grad_norm": 2.7339757982767203,
+      "learning_rate": 0.0001809767677892063,
+      "loss": 1.7422,
+      "step": 4198
+    },
+    {
+      "epoch": 0.67184,
+      "grad_norm": 2.5464462722601566,
+      "learning_rate": 0.00018096663114657444,
+      "loss": 1.7447,
+      "step": 4199
+    },
+    {
+      "epoch": 0.672,
+      "grad_norm": 2.640243914915602,
+      "learning_rate": 0.00018095649208801632,
+      "loss": 1.6977,
+      "step": 4200
+    },
+    {
+      "epoch": 0.67216,
+      "grad_norm": 2.9216105111042223,
+      "learning_rate": 0.00018094635061383446,
+      "loss": 1.7811,
+      "step": 4201
+    },
+    {
+      "epoch": 0.67232,
+      "grad_norm": 1.987340775836971,
+      "learning_rate": 0.00018093620672433142,
+      "loss": 1.8105,
+      "step": 4202
+    },
+    {
+      "epoch": 0.67248,
+      "grad_norm": 2.399283766047064,
+      "learning_rate": 0.00018092606041980992,
+      "loss": 1.688,
+      "step": 4203
+    },
+    {
+      "epoch": 0.67264,
+      "grad_norm": 2.4592010121259413,
+      "learning_rate": 0.0001809159117005727,
+      "loss": 1.807,
+      "step": 4204
+    },
+    {
+      "epoch": 0.6728,
+      "grad_norm": 2.49253728745244,
+      "learning_rate": 0.00018090576056692257,
+      "loss": 1.7118,
+      "step": 4205
+    },
+    {
+      "epoch": 0.67296,
+      "grad_norm": 2.018088555678824,
+      "learning_rate": 0.00018089560701916243,
+      "loss": 1.7945,
+      "step": 4206
+    },
+    {
+      "epoch": 0.67312,
+      "grad_norm": 2.166928275618172,
+      "learning_rate": 0.0001808854510575953,
+      "loss": 1.7807,
+      "step": 4207
+    },
+    {
+      "epoch": 0.67328,
+      "grad_norm": 2.313000587003232,
+      "learning_rate": 0.00018087529268252412,
+      "loss": 1.8188,
+      "step": 4208
+    },
+    {
+      "epoch": 0.67344,
+      "grad_norm": 2.5578837394529015,
+      "learning_rate": 0.00018086513189425208,
+      "loss": 1.7353,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6736,
+      "grad_norm": 2.004510492383818,
+      "learning_rate": 0.0001808549686930823,
+      "loss": 1.6864,
+      "step": 4210
+    },
+    {
+      "epoch": 0.67376,
+      "grad_norm": 2.3679933045582016,
+      "learning_rate": 0.00018084480307931813,
+      "loss": 1.8743,
+      "step": 4211
+    },
+    {
+      "epoch": 0.67392,
+      "grad_norm": 2.3164237373387113,
+      "learning_rate": 0.00018083463505326282,
+      "loss": 1.631,
+      "step": 4212
+    },
+    {
+      "epoch": 0.67408,
+      "grad_norm": 2.6485714700507748,
+      "learning_rate": 0.00018082446461521978,
+      "loss": 1.7597,
+      "step": 4213
+    },
+    {
+      "epoch": 0.67424,
+      "grad_norm": 2.1801546203036475,
+      "learning_rate": 0.0001808142917654925,
+      "loss": 1.7075,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6744,
+      "grad_norm": 2.3695991339543507,
+      "learning_rate": 0.00018080411650438452,
+      "loss": 1.7763,
+      "step": 4215
+    },
+    {
+      "epoch": 0.67456,
+      "grad_norm": 2.641032982842576,
+      "learning_rate": 0.00018079393883219943,
+      "loss": 1.8082,
+      "step": 4216
+    },
+    {
+      "epoch": 0.67472,
+      "grad_norm": 2.953055756537416,
+      "learning_rate": 0.0001807837587492409,
+      "loss": 1.6941,
+      "step": 4217
+    },
+    {
+      "epoch": 0.67488,
+      "grad_norm": 2.374934138650253,
+      "learning_rate": 0.0001807735762558128,
+      "loss": 1.8059,
+      "step": 4218
+    },
+    {
+      "epoch": 0.67504,
+      "grad_norm": 2.8283025317401718,
+      "learning_rate": 0.00018076339135221882,
+      "loss": 1.7874,
+      "step": 4219
+    },
+    {
+      "epoch": 0.6752,
+      "grad_norm": 2.235133612464719,
+      "learning_rate": 0.00018075320403876296,
+      "loss": 1.6509,
+      "step": 4220
+    },
+    {
+      "epoch": 0.67536,
+      "grad_norm": 2.6031199320435094,
+      "learning_rate": 0.00018074301431574913,
+      "loss": 1.7668,
+      "step": 4221
+    },
+    {
+      "epoch": 0.67552,
+      "grad_norm": 2.411784345093072,
+      "learning_rate": 0.00018073282218348147,
+      "loss": 1.7665,
+      "step": 4222
+    },
+    {
+      "epoch": 0.67568,
+      "grad_norm": 3.117562541813238,
+      "learning_rate": 0.00018072262764226398,
+      "loss": 1.8241,
+      "step": 4223
+    },
+    {
+      "epoch": 0.67584,
+      "grad_norm": 2.206421544983744,
+      "learning_rate": 0.00018071243069240093,
+      "loss": 1.6233,
+      "step": 4224
+    },
+    {
+      "epoch": 0.676,
+      "grad_norm": 2.256332434028847,
+      "learning_rate": 0.00018070223133419654,
+      "loss": 1.7053,
+      "step": 4225
+    },
+    {
+      "epoch": 0.67616,
+      "grad_norm": 2.1889222588551456,
+      "learning_rate": 0.0001806920295679552,
+      "loss": 1.6661,
+      "step": 4226
+    },
+    {
+      "epoch": 0.67632,
+      "grad_norm": 3.198089259619162,
+      "learning_rate": 0.00018068182539398122,
+      "loss": 1.7466,
+      "step": 4227
+    },
+    {
+      "epoch": 0.67648,
+      "grad_norm": 2.584381425423621,
+      "learning_rate": 0.00018067161881257917,
+      "loss": 1.7696,
+      "step": 4228
+    },
+    {
+      "epoch": 0.67664,
+      "grad_norm": 2.367777023714503,
+      "learning_rate": 0.00018066140982405358,
+      "loss": 1.7529,
+      "step": 4229
+    },
+    {
+      "epoch": 0.6768,
+      "grad_norm": 2.088589217854261,
+      "learning_rate": 0.00018065119842870905,
+      "loss": 1.7091,
+      "step": 4230
+    },
+    {
+      "epoch": 0.67696,
+      "grad_norm": 2.4591787925169943,
+      "learning_rate": 0.00018064098462685027,
+      "loss": 1.7042,
+      "step": 4231
+    },
+    {
+      "epoch": 0.67712,
+      "grad_norm": 2.266887819459206,
+      "learning_rate": 0.000180630768418782,
+      "loss": 1.7219,
+      "step": 4232
+    },
+    {
+      "epoch": 0.67728,
+      "grad_norm": 2.2737921520677027,
+      "learning_rate": 0.00018062054980480912,
+      "loss": 1.7866,
+      "step": 4233
+    },
+    {
+      "epoch": 0.67744,
+      "grad_norm": 2.8964351361510485,
+      "learning_rate": 0.0001806103287852365,
+      "loss": 1.7619,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6776,
+      "grad_norm": 2.6076180067751116,
+      "learning_rate": 0.0001806001053603691,
+      "loss": 1.7733,
+      "step": 4235
+    },
+    {
+      "epoch": 0.67776,
+      "grad_norm": 2.366136973793912,
+      "learning_rate": 0.00018058987953051205,
+      "loss": 1.6155,
+      "step": 4236
+    },
+    {
+      "epoch": 0.67792,
+      "grad_norm": 3.4905829249584075,
+      "learning_rate": 0.0001805796512959704,
+      "loss": 1.7227,
+      "step": 4237
+    },
+    {
+      "epoch": 0.67808,
+      "grad_norm": 3.4096710193490094,
+      "learning_rate": 0.0001805694206570494,
+      "loss": 1.7301,
+      "step": 4238
+    },
+    {
+      "epoch": 0.67824,
+      "grad_norm": 2.814791769924778,
+      "learning_rate": 0.00018055918761405426,
+      "loss": 1.7578,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6784,
+      "grad_norm": 2.425331252425981,
+      "learning_rate": 0.00018054895216729037,
+      "loss": 1.7358,
+      "step": 4240
+    },
+    {
+      "epoch": 0.67856,
+      "grad_norm": 2.9841074339993403,
+      "learning_rate": 0.0001805387143170631,
+      "loss": 1.7312,
+      "step": 4241
+    },
+    {
+      "epoch": 0.67872,
+      "grad_norm": 2.1065125630266333,
+      "learning_rate": 0.00018052847406367797,
+      "loss": 1.6797,
+      "step": 4242
+    },
+    {
+      "epoch": 0.67888,
+      "grad_norm": 2.445773276687304,
+      "learning_rate": 0.0001805182314074405,
+      "loss": 1.7146,
+      "step": 4243
+    },
+    {
+      "epoch": 0.67904,
+      "grad_norm": 2.6434456211026545,
+      "learning_rate": 0.00018050798634865634,
+      "loss": 1.8087,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6792,
+      "grad_norm": 2.3456037119615605,
+      "learning_rate": 0.00018049773888763118,
+      "loss": 1.8104,
+      "step": 4245
+    },
+    {
+      "epoch": 0.67936,
+      "grad_norm": 2.273267141071625,
+      "learning_rate": 0.00018048748902467078,
+      "loss": 1.663,
+      "step": 4246
+    },
+    {
+      "epoch": 0.67952,
+      "grad_norm": 2.159567560147982,
+      "learning_rate": 0.000180477236760081,
+      "loss": 1.7409,
+      "step": 4247
+    },
+    {
+      "epoch": 0.67968,
+      "grad_norm": 2.5536113095953343,
+      "learning_rate": 0.00018046698209416774,
+      "loss": 1.776,
+      "step": 4248
+    },
+    {
+      "epoch": 0.67984,
+      "grad_norm": 2.207793763717156,
+      "learning_rate": 0.000180456725027237,
+      "loss": 1.749,
+      "step": 4249
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.4781517824629824,
+      "learning_rate": 0.00018044646555959482,
+      "loss": 1.713,
+      "step": 4250
+    },
+    {
+      "epoch": 0.68016,
+      "grad_norm": 2.516335133675631,
+      "learning_rate": 0.00018043620369154732,
+      "loss": 1.7013,
+      "step": 4251
+    },
+    {
+      "epoch": 0.68032,
+      "grad_norm": 2.687975107584278,
+      "learning_rate": 0.0001804259394234007,
+      "loss": 1.7791,
+      "step": 4252
+    },
+    {
+      "epoch": 0.68048,
+      "grad_norm": 2.630968472612907,
+      "learning_rate": 0.00018041567275546127,
+      "loss": 1.7049,
+      "step": 4253
+    },
+    {
+      "epoch": 0.68064,
+      "grad_norm": 2.4198837848907817,
+      "learning_rate": 0.00018040540368803533,
+      "loss": 1.7362,
+      "step": 4254
+    },
+    {
+      "epoch": 0.6808,
+      "grad_norm": 2.22180566287773,
+      "learning_rate": 0.00018039513222142933,
+      "loss": 1.6489,
+      "step": 4255
+    },
+    {
+      "epoch": 0.68096,
+      "grad_norm": 1.9989801874563748,
+      "learning_rate": 0.00018038485835594967,
+      "loss": 1.7496,
+      "step": 4256
+    },
+    {
+      "epoch": 0.68112,
+      "grad_norm": 2.9024260771061887,
+      "learning_rate": 0.00018037458209190302,
+      "loss": 1.6682,
+      "step": 4257
+    },
+    {
+      "epoch": 0.68128,
+      "grad_norm": 2.3769031445840487,
+      "learning_rate": 0.00018036430342959593,
+      "loss": 1.7484,
+      "step": 4258
+    },
+    {
+      "epoch": 0.68144,
+      "grad_norm": 2.9016110255931715,
+      "learning_rate": 0.00018035402236933515,
+      "loss": 1.7497,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6816,
+      "grad_norm": 2.435653150738402,
+      "learning_rate": 0.00018034373891142742,
+      "loss": 1.674,
+      "step": 4260
+    },
+    {
+      "epoch": 0.68176,
+      "grad_norm": 3.015853745815256,
+      "learning_rate": 0.0001803334530561796,
+      "loss": 1.8126,
+      "step": 4261
+    },
+    {
+      "epoch": 0.68192,
+      "grad_norm": 2.2455316042125983,
+      "learning_rate": 0.00018032316480389858,
+      "loss": 1.6994,
+      "step": 4262
+    },
+    {
+      "epoch": 0.68208,
+      "grad_norm": 2.3584370768076455,
+      "learning_rate": 0.0001803128741548914,
+      "loss": 1.8233,
+      "step": 4263
+    },
+    {
+      "epoch": 0.68224,
+      "grad_norm": 2.534894687789647,
+      "learning_rate": 0.00018030258110946504,
+      "loss": 1.7714,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6824,
+      "grad_norm": 2.3469159213227253,
+      "learning_rate": 0.00018029228566792673,
+      "loss": 1.8259,
+      "step": 4265
+    },
+    {
+      "epoch": 0.68256,
+      "grad_norm": 2.451083499998481,
+      "learning_rate": 0.00018028198783058357,
+      "loss": 1.7076,
+      "step": 4266
+    },
+    {
+      "epoch": 0.68272,
+      "grad_norm": 2.4408880418124514,
+      "learning_rate": 0.0001802716875977429,
+      "loss": 1.7105,
+      "step": 4267
+    },
+    {
+      "epoch": 0.68288,
+      "grad_norm": 2.2894422567402093,
+      "learning_rate": 0.00018026138496971203,
+      "loss": 1.6606,
+      "step": 4268
+    },
+    {
+      "epoch": 0.68304,
+      "grad_norm": 2.635790509783681,
+      "learning_rate": 0.00018025107994679837,
+      "loss": 1.6897,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6832,
+      "grad_norm": 2.3205385086017403,
+      "learning_rate": 0.00018024077252930945,
+      "loss": 1.7904,
+      "step": 4270
+    },
+    {
+      "epoch": 0.68336,
+      "grad_norm": 2.030447180572716,
+      "learning_rate": 0.0001802304627175528,
+      "loss": 1.7755,
+      "step": 4271
+    },
+    {
+      "epoch": 0.68352,
+      "grad_norm": 2.011678484503883,
+      "learning_rate": 0.00018022015051183602,
+      "loss": 1.6861,
+      "step": 4272
+    },
+    {
+      "epoch": 0.68368,
+      "grad_norm": 2.8455153877501425,
+      "learning_rate": 0.00018020983591246687,
+      "loss": 1.7721,
+      "step": 4273
+    },
+    {
+      "epoch": 0.68384,
+      "grad_norm": 2.2995884572465735,
+      "learning_rate": 0.0001801995189197531,
+      "loss": 1.7634,
+      "step": 4274
+    },
+    {
+      "epoch": 0.684,
+      "grad_norm": 2.228689696975175,
+      "learning_rate": 0.0001801891995340025,
+      "loss": 1.7947,
+      "step": 4275
+    },
+    {
+      "epoch": 0.68416,
+      "grad_norm": 2.2510050479807187,
+      "learning_rate": 0.00018017887775552307,
+      "loss": 1.7934,
+      "step": 4276
+    },
+    {
+      "epoch": 0.68432,
+      "grad_norm": 2.5079008022914366,
+      "learning_rate": 0.00018016855358462278,
+      "loss": 1.768,
+      "step": 4277
+    },
+    {
+      "epoch": 0.68448,
+      "grad_norm": 2.891433585833976,
+      "learning_rate": 0.00018015822702160966,
+      "loss": 1.7997,
+      "step": 4278
+    },
+    {
+      "epoch": 0.68464,
+      "grad_norm": 2.372163494679463,
+      "learning_rate": 0.0001801478980667918,
+      "loss": 1.7582,
+      "step": 4279
+    },
+    {
+      "epoch": 0.6848,
+      "grad_norm": 2.148414036517006,
+      "learning_rate": 0.0001801375667204775,
+      "loss": 1.7856,
+      "step": 4280
+    },
+    {
+      "epoch": 0.68496,
+      "grad_norm": 2.2739035022762106,
+      "learning_rate": 0.00018012723298297499,
+      "loss": 1.6853,
+      "step": 4281
+    },
+    {
+      "epoch": 0.68512,
+      "grad_norm": 2.626390467236614,
+      "learning_rate": 0.00018011689685459254,
+      "loss": 1.8089,
+      "step": 4282
+    },
+    {
+      "epoch": 0.68528,
+      "grad_norm": 2.4485521009522353,
+      "learning_rate": 0.0001801065583356387,
+      "loss": 1.7326,
+      "step": 4283
+    },
+    {
+      "epoch": 0.68544,
+      "grad_norm": 2.1125832592427494,
+      "learning_rate": 0.00018009621742642184,
+      "loss": 1.7102,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6856,
+      "grad_norm": 2.031518730740922,
+      "learning_rate": 0.00018008587412725059,
+      "loss": 1.7046,
+      "step": 4285
+    },
+    {
+      "epoch": 0.68576,
+      "grad_norm": 2.8871174509237187,
+      "learning_rate": 0.00018007552843843355,
+      "loss": 1.8399,
+      "step": 4286
+    },
+    {
+      "epoch": 0.68592,
+      "grad_norm": 2.670859935585853,
+      "learning_rate": 0.00018006518036027945,
+      "loss": 1.6789,
+      "step": 4287
+    },
+    {
+      "epoch": 0.68608,
+      "grad_norm": 2.2301971956634254,
+      "learning_rate": 0.000180054829893097,
+      "loss": 1.7337,
+      "step": 4288
+    },
+    {
+      "epoch": 0.68624,
+      "grad_norm": 2.5125735797936524,
+      "learning_rate": 0.00018004447703719506,
+      "loss": 1.7354,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6864,
+      "grad_norm": 2.454162849078172,
+      "learning_rate": 0.00018003412179288261,
+      "loss": 1.826,
+      "step": 4290
+    },
+    {
+      "epoch": 0.68656,
+      "grad_norm": 2.642354258137277,
+      "learning_rate": 0.00018002376416046857,
+      "loss": 1.6625,
+      "step": 4291
+    },
+    {
+      "epoch": 0.68672,
+      "grad_norm": 2.1834513354448,
+      "learning_rate": 0.000180013404140262,
+      "loss": 1.7246,
+      "step": 4292
+    },
+    {
+      "epoch": 0.68688,
+      "grad_norm": 2.165504318010599,
+      "learning_rate": 0.00018000304173257202,
+      "loss": 1.8518,
+      "step": 4293
+    },
+    {
+      "epoch": 0.68704,
+      "grad_norm": 2.1369621906771847,
+      "learning_rate": 0.00017999267693770792,
+      "loss": 1.8321,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6872,
+      "grad_norm": 2.4709223548096877,
+      "learning_rate": 0.00017998230975597887,
+      "loss": 1.7257,
+      "step": 4295
+    },
+    {
+      "epoch": 0.68736,
+      "grad_norm": 2.482958705194428,
+      "learning_rate": 0.00017997194018769423,
+      "loss": 1.6167,
+      "step": 4296
+    },
+    {
+      "epoch": 0.68752,
+      "grad_norm": 2.6652120597596656,
+      "learning_rate": 0.00017996156823316342,
+      "loss": 1.7621,
+      "step": 4297
+    },
+    {
+      "epoch": 0.68768,
+      "grad_norm": 2.8990110987777875,
+      "learning_rate": 0.00017995119389269595,
+      "loss": 1.6577,
+      "step": 4298
+    },
+    {
+      "epoch": 0.68784,
+      "grad_norm": 2.4358549878641025,
+      "learning_rate": 0.00017994081716660136,
+      "loss": 1.803,
+      "step": 4299
+    },
+    {
+      "epoch": 0.688,
+      "grad_norm": 2.1714949693722017,
+      "learning_rate": 0.00017993043805518926,
+      "loss": 1.6794,
+      "step": 4300
+    },
+    {
+      "epoch": 0.68816,
+      "grad_norm": 2.625269290427474,
+      "learning_rate": 0.00017992005655876936,
+      "loss": 1.785,
+      "step": 4301
+    },
+    {
+      "epoch": 0.68832,
+      "grad_norm": 2.0661488398157726,
+      "learning_rate": 0.00017990967267765142,
+      "loss": 1.6779,
+      "step": 4302
+    },
+    {
+      "epoch": 0.68848,
+      "grad_norm": 2.572351597651381,
+      "learning_rate": 0.0001798992864121453,
+      "loss": 1.7783,
+      "step": 4303
+    },
+    {
+      "epoch": 0.68864,
+      "grad_norm": 2.461362662326193,
+      "learning_rate": 0.00017988889776256092,
+      "loss": 1.7436,
+      "step": 4304
+    },
+    {
+      "epoch": 0.6888,
+      "grad_norm": 3.084667739885667,
+      "learning_rate": 0.0001798785067292082,
+      "loss": 1.8278,
+      "step": 4305
+    },
+    {
+      "epoch": 0.68896,
+      "grad_norm": 2.8333039503018504,
+      "learning_rate": 0.00017986811331239725,
+      "loss": 1.6862,
+      "step": 4306
+    },
+    {
+      "epoch": 0.68912,
+      "grad_norm": 3.6377557898620987,
+      "learning_rate": 0.00017985771751243817,
+      "loss": 1.6572,
+      "step": 4307
+    },
+    {
+      "epoch": 0.68928,
+      "grad_norm": 2.3466814784353405,
+      "learning_rate": 0.0001798473193296412,
+      "loss": 1.7486,
+      "step": 4308
+    },
+    {
+      "epoch": 0.68944,
+      "grad_norm": 3.45373262840042,
+      "learning_rate": 0.00017983691876431653,
+      "loss": 1.7441,
+      "step": 4309
+    },
+    {
+      "epoch": 0.6896,
+      "grad_norm": 2.196963479341605,
+      "learning_rate": 0.00017982651581677456,
+      "loss": 1.6725,
+      "step": 4310
+    },
+    {
+      "epoch": 0.68976,
+      "grad_norm": 2.196963479341605,
+      "learning_rate": 0.00017982651581677456,
+      "loss": 1.6649,
+      "step": 4311
+    },
+    {
+      "epoch": 0.68992,
+      "grad_norm": 4.540704113724911,
+      "learning_rate": 0.00017981611048732568,
+      "loss": 1.7674,
+      "step": 4312
+    },
+    {
+      "epoch": 0.69008,
+      "grad_norm": 2.3562721826522264,
+      "learning_rate": 0.00017980570277628037,
+      "loss": 1.7265,
+      "step": 4313
+    },
+    {
+      "epoch": 0.69024,
+      "grad_norm": 2.4261779845494886,
+      "learning_rate": 0.00017979529268394913,
+      "loss": 1.6337,
+      "step": 4314
+    },
+    {
+      "epoch": 0.6904,
+      "grad_norm": 2.721159102586794,
+      "learning_rate": 0.0001797848802106427,
+      "loss": 1.6716,
+      "step": 4315
+    },
+    {
+      "epoch": 0.69056,
+      "grad_norm": 2.4476978371605016,
+      "learning_rate": 0.00017977446535667166,
+      "loss": 1.725,
+      "step": 4316
+    },
+    {
+      "epoch": 0.69072,
+      "grad_norm": 2.9817130048754508,
+      "learning_rate": 0.00017976404812234686,
+      "loss": 1.8548,
+      "step": 4317
+    },
+    {
+      "epoch": 0.69088,
+      "grad_norm": 2.8548460767619694,
+      "learning_rate": 0.00017975362850797906,
+      "loss": 1.6609,
+      "step": 4318
+    },
+    {
+      "epoch": 0.69104,
+      "grad_norm": 2.8307267402804417,
+      "learning_rate": 0.0001797432065138792,
+      "loss": 1.7968,
+      "step": 4319
+    },
+    {
+      "epoch": 0.6912,
+      "grad_norm": 2.190620579917876,
+      "learning_rate": 0.00017973278214035825,
+      "loss": 1.7116,
+      "step": 4320
+    },
+    {
+      "epoch": 0.69136,
+      "grad_norm": 2.270068179375658,
+      "learning_rate": 0.00017972235538772727,
+      "loss": 1.6487,
+      "step": 4321
+    },
+    {
+      "epoch": 0.69152,
+      "grad_norm": 2.120783299841827,
+      "learning_rate": 0.00017971192625629737,
+      "loss": 1.6475,
+      "step": 4322
+    },
+    {
+      "epoch": 0.69168,
+      "grad_norm": 2.2138564374856853,
+      "learning_rate": 0.00017970149474637976,
+      "loss": 1.8173,
+      "step": 4323
+    },
+    {
+      "epoch": 0.69184,
+      "grad_norm": 2.6867321463016003,
+      "learning_rate": 0.00017969106085828567,
+      "loss": 1.7347,
+      "step": 4324
+    },
+    {
+      "epoch": 0.692,
+      "grad_norm": 2.752129301288876,
+      "learning_rate": 0.00017968062459232647,
+      "loss": 1.7567,
+      "step": 4325
+    },
+    {
+      "epoch": 0.69216,
+      "grad_norm": 2.3541125853990654,
+      "learning_rate": 0.0001796701859488135,
+      "loss": 1.8238,
+      "step": 4326
+    },
+    {
+      "epoch": 0.69232,
+      "grad_norm": 2.222699261949288,
+      "learning_rate": 0.0001796597449280583,
+      "loss": 1.7237,
+      "step": 4327
+    },
+    {
+      "epoch": 0.69248,
+      "grad_norm": 2.4183395069163924,
+      "learning_rate": 0.00017964930153037235,
+      "loss": 1.6978,
+      "step": 4328
+    },
+    {
+      "epoch": 0.69264,
+      "grad_norm": 2.2381158813494264,
+      "learning_rate": 0.00017963885575606733,
+      "loss": 1.6881,
+      "step": 4329
+    },
+    {
+      "epoch": 0.6928,
+      "grad_norm": 2.330898900985596,
+      "learning_rate": 0.0001796284076054549,
+      "loss": 1.6458,
+      "step": 4330
+    },
+    {
+      "epoch": 0.69296,
+      "grad_norm": 2.214734799832376,
+      "learning_rate": 0.0001796179570788468,
+      "loss": 1.7994,
+      "step": 4331
+    },
+    {
+      "epoch": 0.69312,
+      "grad_norm": 2.1164728939326563,
+      "learning_rate": 0.00017960750417655492,
+      "loss": 1.6802,
+      "step": 4332
+    },
+    {
+      "epoch": 0.69328,
+      "grad_norm": 2.5606309581555218,
+      "learning_rate": 0.00017959704889889108,
+      "loss": 1.7718,
+      "step": 4333
+    },
+    {
+      "epoch": 0.69344,
+      "grad_norm": 2.1904438786409948,
+      "learning_rate": 0.0001795865912461673,
+      "loss": 1.77,
+      "step": 4334
+    },
+    {
+      "epoch": 0.6936,
+      "grad_norm": 2.2360257818131997,
+      "learning_rate": 0.0001795761312186956,
+      "loss": 1.7647,
+      "step": 4335
+    },
+    {
+      "epoch": 0.69376,
+      "grad_norm": 2.7096849746710414,
+      "learning_rate": 0.0001795656688167881,
+      "loss": 1.707,
+      "step": 4336
+    },
+    {
+      "epoch": 0.69392,
+      "grad_norm": 2.0871507982474515,
+      "learning_rate": 0.000179555204040757,
+      "loss": 1.7779,
+      "step": 4337
+    },
+    {
+      "epoch": 0.69408,
+      "grad_norm": 2.4678192625972666,
+      "learning_rate": 0.00017954473689091454,
+      "loss": 1.7295,
+      "step": 4338
+    },
+    {
+      "epoch": 0.69424,
+      "grad_norm": 2.3482793653061664,
+      "learning_rate": 0.000179534267367573,
+      "loss": 1.7525,
+      "step": 4339
+    },
+    {
+      "epoch": 0.6944,
+      "grad_norm": 2.5551311679969637,
+      "learning_rate": 0.00017952379547104484,
+      "loss": 1.7362,
+      "step": 4340
+    },
+    {
+      "epoch": 0.69456,
+      "grad_norm": 2.4943042167648706,
+      "learning_rate": 0.0001795133212016425,
+      "loss": 1.7556,
+      "step": 4341
+    },
+    {
+      "epoch": 0.69472,
+      "grad_norm": 2.87100066429232,
+      "learning_rate": 0.00017950284455967852,
+      "loss": 1.7378,
+      "step": 4342
+    },
+    {
+      "epoch": 0.69488,
+      "grad_norm": 2.6197468461109485,
+      "learning_rate": 0.0001794923655454655,
+      "loss": 1.7267,
+      "step": 4343
+    },
+    {
+      "epoch": 0.69504,
+      "grad_norm": 2.5026984658329736,
+      "learning_rate": 0.00017948188415931613,
+      "loss": 1.7472,
+      "step": 4344
+    },
+    {
+      "epoch": 0.6952,
+      "grad_norm": 2.1974675326696147,
+      "learning_rate": 0.00017947140040154314,
+      "loss": 1.7148,
+      "step": 4345
+    },
+    {
+      "epoch": 0.69536,
+      "grad_norm": 2.473207161170142,
+      "learning_rate": 0.00017946091427245938,
+      "loss": 1.8607,
+      "step": 4346
+    },
+    {
+      "epoch": 0.69552,
+      "grad_norm": 2.095252576428565,
+      "learning_rate": 0.00017945042577237775,
+      "loss": 1.6864,
+      "step": 4347
+    },
+    {
+      "epoch": 0.69568,
+      "grad_norm": 2.4244564123413626,
+      "learning_rate": 0.00017943993490161114,
+      "loss": 1.8075,
+      "step": 4348
+    },
+    {
+      "epoch": 0.69584,
+      "grad_norm": 1.8750288685734444,
+      "learning_rate": 0.00017942944166047267,
+      "loss": 1.6788,
+      "step": 4349
+    },
+    {
+      "epoch": 0.696,
+      "grad_norm": 3.485095562822839,
+      "learning_rate": 0.00017941894604927536,
+      "loss": 1.6877,
+      "step": 4350
+    },
+    {
+      "epoch": 0.69616,
+      "grad_norm": 2.3404841106315706,
+      "learning_rate": 0.00017940844806833242,
+      "loss": 1.6598,
+      "step": 4351
+    },
+    {
+      "epoch": 0.69632,
+      "grad_norm": 2.8499069435165643,
+      "learning_rate": 0.00017939794771795715,
+      "loss": 1.6792,
+      "step": 4352
+    },
+    {
+      "epoch": 0.69648,
+      "grad_norm": 2.817957890543654,
+      "learning_rate": 0.00017938744499846278,
+      "loss": 1.7736,
+      "step": 4353
+    },
+    {
+      "epoch": 0.69664,
+      "grad_norm": 3.4718230908824426,
+      "learning_rate": 0.00017937693991016275,
+      "loss": 1.7463,
+      "step": 4354
+    },
+    {
+      "epoch": 0.6968,
+      "grad_norm": 2.939047152215051,
+      "learning_rate": 0.00017936643245337048,
+      "loss": 1.87,
+      "step": 4355
+    },
+    {
+      "epoch": 0.69696,
+      "grad_norm": 2.6133433864686606,
+      "learning_rate": 0.00017935592262839954,
+      "loss": 1.7738,
+      "step": 4356
+    },
+    {
+      "epoch": 0.69712,
+      "grad_norm": 2.4442204864346406,
+      "learning_rate": 0.00017934541043556346,
+      "loss": 1.7744,
+      "step": 4357
+    },
+    {
+      "epoch": 0.69728,
+      "grad_norm": 2.400127589464533,
+      "learning_rate": 0.00017933489587517598,
+      "loss": 1.7016,
+      "step": 4358
+    },
+    {
+      "epoch": 0.69744,
+      "grad_norm": 3.016428507114298,
+      "learning_rate": 0.00017932437894755082,
+      "loss": 1.6323,
+      "step": 4359
+    },
+    {
+      "epoch": 0.6976,
+      "grad_norm": 2.098363178449492,
+      "learning_rate": 0.00017931385965300173,
+      "loss": 1.6956,
+      "step": 4360
+    },
+    {
+      "epoch": 0.69776,
+      "grad_norm": 3.2485814061686216,
+      "learning_rate": 0.0001793033379918427,
+      "loss": 1.7353,
+      "step": 4361
+    },
+    {
+      "epoch": 0.69792,
+      "grad_norm": 1.9842646600009264,
+      "learning_rate": 0.0001792928139643876,
+      "loss": 1.7521,
+      "step": 4362
+    },
+    {
+      "epoch": 0.69808,
+      "grad_norm": 2.2358632342918052,
+      "learning_rate": 0.00017928228757095046,
+      "loss": 1.6562,
+      "step": 4363
+    },
+    {
+      "epoch": 0.69824,
+      "grad_norm": 2.119540952011801,
+      "learning_rate": 0.00017927175881184542,
+      "loss": 1.6466,
+      "step": 4364
+    },
+    {
+      "epoch": 0.6984,
+      "grad_norm": 2.430437329889527,
+      "learning_rate": 0.00017926122768738658,
+      "loss": 1.69,
+      "step": 4365
+    },
+    {
+      "epoch": 0.69856,
+      "grad_norm": 2.1316743197890653,
+      "learning_rate": 0.0001792506941978882,
+      "loss": 1.7383,
+      "step": 4366
+    },
+    {
+      "epoch": 0.69872,
+      "grad_norm": 1.7955228810210115,
+      "learning_rate": 0.0001792401583436646,
+      "loss": 1.6954,
+      "step": 4367
+    },
+    {
+      "epoch": 0.69888,
+      "grad_norm": 2.3924174827295586,
+      "learning_rate": 0.00017922962012503012,
+      "loss": 1.8424,
+      "step": 4368
+    },
+    {
+      "epoch": 0.69904,
+      "grad_norm": 2.2026604838961257,
+      "learning_rate": 0.00017921907954229923,
+      "loss": 1.7982,
+      "step": 4369
+    },
+    {
+      "epoch": 0.6992,
+      "grad_norm": 2.404868552467124,
+      "learning_rate": 0.00017920853659578648,
+      "loss": 1.7099,
+      "step": 4370
+    },
+    {
+      "epoch": 0.69936,
+      "grad_norm": 2.196512646992055,
+      "learning_rate": 0.00017919799128580637,
+      "loss": 1.6848,
+      "step": 4371
+    },
+    {
+      "epoch": 0.69952,
+      "grad_norm": 2.2529813992904693,
+      "learning_rate": 0.00017918744361267362,
+      "loss": 1.682,
+      "step": 4372
+    },
+    {
+      "epoch": 0.69968,
+      "grad_norm": 2.730234849623514,
+      "learning_rate": 0.00017917689357670294,
+      "loss": 1.7132,
+      "step": 4373
+    },
+    {
+      "epoch": 0.69984,
+      "grad_norm": 2.5813673682269807,
+      "learning_rate": 0.0001791663411782091,
+      "loss": 1.6525,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 2.1676330975915974,
+      "learning_rate": 0.00017915578641750703,
+      "loss": 1.7141,
+      "step": 4375
+    },
+    {
+      "epoch": 0.70016,
+      "grad_norm": 1.9323319987873127,
+      "learning_rate": 0.00017914522929491164,
+      "loss": 1.7304,
+      "step": 4376
+    },
+    {
+      "epoch": 0.70032,
+      "grad_norm": 2.4527765295927475,
+      "learning_rate": 0.00017913466981073794,
+      "loss": 1.6806,
+      "step": 4377
+    },
+    {
+      "epoch": 0.70048,
+      "grad_norm": 2.9852391532573153,
+      "learning_rate": 0.000179124107965301,
+      "loss": 1.7203,
+      "step": 4378
+    },
+    {
+      "epoch": 0.70064,
+      "grad_norm": 2.751575524228845,
+      "learning_rate": 0.00017911354375891594,
+      "loss": 1.6565,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7008,
+      "grad_norm": 2.098061205103975,
+      "learning_rate": 0.00017910297719189803,
+      "loss": 1.6538,
+      "step": 4380
+    },
+    {
+      "epoch": 0.70096,
+      "grad_norm": 2.323520180122517,
+      "learning_rate": 0.00017909240826456254,
+      "loss": 1.7085,
+      "step": 4381
+    },
+    {
+      "epoch": 0.70112,
+      "grad_norm": 3.06027896763079,
+      "learning_rate": 0.00017908183697722487,
+      "loss": 1.7715,
+      "step": 4382
+    },
+    {
+      "epoch": 0.70128,
+      "grad_norm": 2.5008061122990712,
+      "learning_rate": 0.0001790712633302004,
+      "loss": 1.6789,
+      "step": 4383
+    },
+    {
+      "epoch": 0.70144,
+      "grad_norm": 2.357854532944269,
+      "learning_rate": 0.00017906068732380465,
+      "loss": 1.6032,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7016,
+      "grad_norm": 2.4371181036967218,
+      "learning_rate": 0.0001790501089583532,
+      "loss": 1.6241,
+      "step": 4385
+    },
+    {
+      "epoch": 0.70176,
+      "grad_norm": 2.07466349234962,
+      "learning_rate": 0.00017903952823416168,
+      "loss": 1.7403,
+      "step": 4386
+    },
+    {
+      "epoch": 0.70192,
+      "grad_norm": 2.8333748943415773,
+      "learning_rate": 0.0001790289451515458,
+      "loss": 1.6433,
+      "step": 4387
+    },
+    {
+      "epoch": 0.70208,
+      "grad_norm": 2.6323184082802626,
+      "learning_rate": 0.0001790183597108214,
+      "loss": 1.6714,
+      "step": 4388
+    },
+    {
+      "epoch": 0.70224,
+      "grad_norm": 2.9048032195572744,
+      "learning_rate": 0.00017900777191230422,
+      "loss": 1.7007,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7024,
+      "grad_norm": 4.348608559833807,
+      "learning_rate": 0.0001789971817563103,
+      "loss": 1.825,
+      "step": 4390
+    },
+    {
+      "epoch": 0.70256,
+      "grad_norm": 2.3172159744499536,
+      "learning_rate": 0.00017898658924315555,
+      "loss": 1.6874,
+      "step": 4391
+    },
+    {
+      "epoch": 0.70272,
+      "grad_norm": 2.8998060790952933,
+      "learning_rate": 0.0001789759943731561,
+      "loss": 1.6517,
+      "step": 4392
+    },
+    {
+      "epoch": 0.70288,
+      "grad_norm": 2.0394000090244866,
+      "learning_rate": 0.0001789653971466281,
+      "loss": 1.6871,
+      "step": 4393
+    },
+    {
+      "epoch": 0.70304,
+      "grad_norm": 2.6431763477203076,
+      "learning_rate": 0.00017895479756388768,
+      "loss": 1.6385,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7032,
+      "grad_norm": 2.2109100997935265,
+      "learning_rate": 0.0001789441956252511,
+      "loss": 1.775,
+      "step": 4395
+    },
+    {
+      "epoch": 0.70336,
+      "grad_norm": 2.759152606603681,
+      "learning_rate": 0.0001789335913310348,
+      "loss": 1.6671,
+      "step": 4396
+    },
+    {
+      "epoch": 0.70352,
+      "grad_norm": 2.1098974156890455,
+      "learning_rate": 0.00017892298468155519,
+      "loss": 1.7561,
+      "step": 4397
+    },
+    {
+      "epoch": 0.70368,
+      "grad_norm": 2.6471703849813926,
+      "learning_rate": 0.00017891237567712866,
+      "loss": 1.7859,
+      "step": 4398
+    },
+    {
+      "epoch": 0.70384,
+      "grad_norm": 2.4907690478388482,
+      "learning_rate": 0.00017890176431807183,
+      "loss": 1.7246,
+      "step": 4399
+    },
+    {
+      "epoch": 0.704,
+      "grad_norm": 2.2006562003586314,
+      "learning_rate": 0.00017889115060470135,
+      "loss": 1.7203,
+      "step": 4400
+    },
+    {
+      "epoch": 0.70416,
+      "grad_norm": 2.556667035060147,
+      "learning_rate": 0.00017888053453733388,
+      "loss": 1.7693,
+      "step": 4401
+    },
+    {
+      "epoch": 0.70432,
+      "grad_norm": 2.669829806157551,
+      "learning_rate": 0.0001788699161162862,
+      "loss": 1.6404,
+      "step": 4402
+    },
+    {
+      "epoch": 0.70448,
+      "grad_norm": 2.6331282494507477,
+      "learning_rate": 0.00017885929534187516,
+      "loss": 1.7778,
+      "step": 4403
+    },
+    {
+      "epoch": 0.70464,
+      "grad_norm": 2.6444545526551586,
+      "learning_rate": 0.00017884867221441763,
+      "loss": 1.7129,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7048,
+      "grad_norm": 2.1387235943319287,
+      "learning_rate": 0.00017883804673423065,
+      "loss": 1.7306,
+      "step": 4405
+    },
+    {
+      "epoch": 0.70496,
+      "grad_norm": 2.431910812828967,
+      "learning_rate": 0.00017882741890163124,
+      "loss": 1.6458,
+      "step": 4406
+    },
+    {
+      "epoch": 0.70512,
+      "grad_norm": 2.324950964926763,
+      "learning_rate": 0.00017881678871693648,
+      "loss": 1.7667,
+      "step": 4407
+    },
+    {
+      "epoch": 0.70528,
+      "grad_norm": 3.6975743659324634,
+      "learning_rate": 0.0001788061561804636,
+      "loss": 1.6446,
+      "step": 4408
+    },
+    {
+      "epoch": 0.70544,
+      "grad_norm": 2.695820804050279,
+      "learning_rate": 0.00017879552129252985,
+      "loss": 1.7414,
+      "step": 4409
+    },
+    {
+      "epoch": 0.7056,
+      "grad_norm": 3.002914925176497,
+      "learning_rate": 0.0001787848840534526,
+      "loss": 1.7481,
+      "step": 4410
+    },
+    {
+      "epoch": 0.70576,
+      "grad_norm": 2.5991484063068078,
+      "learning_rate": 0.00017877424446354918,
+      "loss": 1.5918,
+      "step": 4411
+    },
+    {
+      "epoch": 0.70592,
+      "grad_norm": 2.3874906301529095,
+      "learning_rate": 0.0001787636025231371,
+      "loss": 1.5713,
+      "step": 4412
+    },
+    {
+      "epoch": 0.70608,
+      "grad_norm": 3.065228941179938,
+      "learning_rate": 0.00017875295823253386,
+      "loss": 1.659,
+      "step": 4413
+    },
+    {
+      "epoch": 0.70624,
+      "grad_norm": 3.5578892017375456,
+      "learning_rate": 0.00017874231159205713,
+      "loss": 1.8349,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7064,
+      "grad_norm": 2.672549365155214,
+      "learning_rate": 0.00017873166260202455,
+      "loss": 1.7081,
+      "step": 4415
+    },
+    {
+      "epoch": 0.70656,
+      "grad_norm": 2.6059450129513633,
+      "learning_rate": 0.0001787210112627539,
+      "loss": 1.7246,
+      "step": 4416
+    },
+    {
+      "epoch": 0.70672,
+      "grad_norm": 2.391201529209389,
+      "learning_rate": 0.00017871035757456295,
+      "loss": 1.8027,
+      "step": 4417
+    },
+    {
+      "epoch": 0.70688,
+      "grad_norm": 3.0505601674269744,
+      "learning_rate": 0.00017869970153776968,
+      "loss": 1.6586,
+      "step": 4418
+    },
+    {
+      "epoch": 0.70704,
+      "grad_norm": 2.5902944605801004,
+      "learning_rate": 0.00017868904315269195,
+      "loss": 1.7092,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7072,
+      "grad_norm": 3.063169502571489,
+      "learning_rate": 0.00017867838241964785,
+      "loss": 1.8406,
+      "step": 4420
+    },
+    {
+      "epoch": 0.70736,
+      "grad_norm": 2.381465358019842,
+      "learning_rate": 0.00017866771933895546,
+      "loss": 1.7048,
+      "step": 4421
+    },
+    {
+      "epoch": 0.70752,
+      "grad_norm": 2.51940935305762,
+      "learning_rate": 0.00017865705391093294,
+      "loss": 1.6642,
+      "step": 4422
+    },
+    {
+      "epoch": 0.70768,
+      "grad_norm": 2.7717646556285866,
+      "learning_rate": 0.00017864638613589856,
+      "loss": 1.6954,
+      "step": 4423
+    },
+    {
+      "epoch": 0.70784,
+      "grad_norm": 2.2438227577111,
+      "learning_rate": 0.00017863571601417063,
+      "loss": 1.6617,
+      "step": 4424
+    },
+    {
+      "epoch": 0.708,
+      "grad_norm": 2.299493986811642,
+      "learning_rate": 0.00017862504354606751,
+      "loss": 1.7421,
+      "step": 4425
+    },
+    {
+      "epoch": 0.70816,
+      "grad_norm": 2.7076158283495153,
+      "learning_rate": 0.00017861436873190765,
+      "loss": 1.6846,
+      "step": 4426
+    },
+    {
+      "epoch": 0.70832,
+      "grad_norm": 2.491090697765309,
+      "learning_rate": 0.0001786036915720096,
+      "loss": 1.7407,
+      "step": 4427
+    },
+    {
+      "epoch": 0.70848,
+      "grad_norm": 2.3203072109723055,
+      "learning_rate": 0.00017859301206669191,
+      "loss": 1.6627,
+      "step": 4428
+    },
+    {
+      "epoch": 0.70864,
+      "grad_norm": 2.119039791195135,
+      "learning_rate": 0.0001785823302162733,
+      "loss": 1.6831,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7088,
+      "grad_norm": 2.1847914732655833,
+      "learning_rate": 0.00017857164602107242,
+      "loss": 1.6413,
+      "step": 4430
+    },
+    {
+      "epoch": 0.70896,
+      "grad_norm": 2.5953061520746914,
+      "learning_rate": 0.00017856095948140811,
+      "loss": 1.8239,
+      "step": 4431
+    },
+    {
+      "epoch": 0.70912,
+      "grad_norm": 2.5860500234792747,
+      "learning_rate": 0.0001785502705975993,
+      "loss": 1.6934,
+      "step": 4432
+    },
+    {
+      "epoch": 0.70928,
+      "grad_norm": 2.3485028103911056,
+      "learning_rate": 0.0001785395793699648,
+      "loss": 1.6888,
+      "step": 4433
+    },
+    {
+      "epoch": 0.70944,
+      "grad_norm": 2.2308372135124888,
+      "learning_rate": 0.00017852888579882373,
+      "loss": 1.6834,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7096,
+      "grad_norm": 1.8836034277036213,
+      "learning_rate": 0.00017851818988449516,
+      "loss": 1.6618,
+      "step": 4435
+    },
+    {
+      "epoch": 0.70976,
+      "grad_norm": 2.1239017887329923,
+      "learning_rate": 0.00017850749162729817,
+      "loss": 1.676,
+      "step": 4436
+    },
+    {
+      "epoch": 0.70992,
+      "grad_norm": 2.4684881783964143,
+      "learning_rate": 0.00017849679102755207,
+      "loss": 1.6652,
+      "step": 4437
+    },
+    {
+      "epoch": 0.71008,
+      "grad_norm": 2.374171601589023,
+      "learning_rate": 0.0001784860880855761,
+      "loss": 1.7271,
+      "step": 4438
+    },
+    {
+      "epoch": 0.71024,
+      "grad_norm": 2.5708417727298847,
+      "learning_rate": 0.00017847538280168964,
+      "loss": 1.6828,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7104,
+      "grad_norm": 2.2084993415616134,
+      "learning_rate": 0.00017846467517621206,
+      "loss": 1.7564,
+      "step": 4440
+    },
+    {
+      "epoch": 0.71056,
+      "grad_norm": 2.108872014264611,
+      "learning_rate": 0.00017845396520946294,
+      "loss": 1.7495,
+      "step": 4441
+    },
+    {
+      "epoch": 0.71072,
+      "grad_norm": 2.368245329113838,
+      "learning_rate": 0.00017844325290176181,
+      "loss": 1.7918,
+      "step": 4442
+    },
+    {
+      "epoch": 0.71088,
+      "grad_norm": 2.6620874687816727,
+      "learning_rate": 0.00017843253825342833,
+      "loss": 1.694,
+      "step": 4443
+    },
+    {
+      "epoch": 0.71104,
+      "grad_norm": 2.1414026575510023,
+      "learning_rate": 0.00017842182126478218,
+      "loss": 1.665,
+      "step": 4444
+    },
+    {
+      "epoch": 0.7112,
+      "grad_norm": 2.154882593351219,
+      "learning_rate": 0.00017841110193614316,
+      "loss": 1.6759,
+      "step": 4445
+    },
+    {
+      "epoch": 0.71136,
+      "grad_norm": 2.4555944943722694,
+      "learning_rate": 0.0001784003802678311,
+      "loss": 1.7028,
+      "step": 4446
+    },
+    {
+      "epoch": 0.71152,
+      "grad_norm": 2.284725706322637,
+      "learning_rate": 0.00017838965626016596,
+      "loss": 1.6675,
+      "step": 4447
+    },
+    {
+      "epoch": 0.71168,
+      "grad_norm": 2.359636200701578,
+      "learning_rate": 0.0001783789299134677,
+      "loss": 1.6357,
+      "step": 4448
+    },
+    {
+      "epoch": 0.71184,
+      "grad_norm": 2.987411807158091,
+      "learning_rate": 0.0001783682012280564,
+      "loss": 1.6773,
+      "step": 4449
+    },
+    {
+      "epoch": 0.712,
+      "grad_norm": 2.0726114812884173,
+      "learning_rate": 0.00017835747020425213,
+      "loss": 1.6712,
+      "step": 4450
+    },
+    {
+      "epoch": 0.71216,
+      "grad_norm": 2.9231804586158634,
+      "learning_rate": 0.00017834673684237516,
+      "loss": 1.7037,
+      "step": 4451
+    },
+    {
+      "epoch": 0.71232,
+      "grad_norm": 2.4629750491364573,
+      "learning_rate": 0.0001783360011427457,
+      "loss": 1.7422,
+      "step": 4452
+    },
+    {
+      "epoch": 0.71248,
+      "grad_norm": 3.1730652737829588,
+      "learning_rate": 0.00017832526310568412,
+      "loss": 1.7378,
+      "step": 4453
+    },
+    {
+      "epoch": 0.71264,
+      "grad_norm": 2.302534326360817,
+      "learning_rate": 0.00017831452273151083,
+      "loss": 1.7067,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7128,
+      "grad_norm": 2.5493842050553566,
+      "learning_rate": 0.0001783037800205463,
+      "loss": 1.6895,
+      "step": 4455
+    },
+    {
+      "epoch": 0.71296,
+      "grad_norm": 2.159033763618607,
+      "learning_rate": 0.00017829303497311107,
+      "loss": 1.6371,
+      "step": 4456
+    },
+    {
+      "epoch": 0.71312,
+      "grad_norm": 2.4765228197583773,
+      "learning_rate": 0.00017828228758952577,
+      "loss": 1.6876,
+      "step": 4457
+    },
+    {
+      "epoch": 0.71328,
+      "grad_norm": 2.4247684053555703,
+      "learning_rate": 0.00017827153787011105,
+      "loss": 1.6881,
+      "step": 4458
+    },
+    {
+      "epoch": 0.71344,
+      "grad_norm": 2.3166699782151583,
+      "learning_rate": 0.00017826078581518773,
+      "loss": 1.7054,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7136,
+      "grad_norm": 2.6071459715533987,
+      "learning_rate": 0.00017825003142507657,
+      "loss": 1.7774,
+      "step": 4460
+    },
+    {
+      "epoch": 0.71376,
+      "grad_norm": 2.236430432744168,
+      "learning_rate": 0.00017823927470009848,
+      "loss": 1.7407,
+      "step": 4461
+    },
+    {
+      "epoch": 0.71392,
+      "grad_norm": 2.3463282945262374,
+      "learning_rate": 0.00017822851564057448,
+      "loss": 1.6741,
+      "step": 4462
+    },
+    {
+      "epoch": 0.71408,
+      "grad_norm": 2.4111066507110404,
+      "learning_rate": 0.00017821775424682555,
+      "loss": 1.7057,
+      "step": 4463
+    },
+    {
+      "epoch": 0.71424,
+      "grad_norm": 2.4864840639573735,
+      "learning_rate": 0.00017820699051917282,
+      "loss": 1.7267,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7144,
+      "grad_norm": 2.0564760321025384,
+      "learning_rate": 0.0001781962244579374,
+      "loss": 1.6677,
+      "step": 4465
+    },
+    {
+      "epoch": 0.71456,
+      "grad_norm": 2.613296382018561,
+      "learning_rate": 0.00017818545606344065,
+      "loss": 1.632,
+      "step": 4466
+    },
+    {
+      "epoch": 0.71472,
+      "grad_norm": 2.3490227804532333,
+      "learning_rate": 0.00017817468533600381,
+      "loss": 1.6618,
+      "step": 4467
+    },
+    {
+      "epoch": 0.71488,
+      "grad_norm": 2.64580247476497,
+      "learning_rate": 0.00017816391227594823,
+      "loss": 1.5664,
+      "step": 4468
+    },
+    {
+      "epoch": 0.71504,
+      "grad_norm": 2.4480448612718932,
+      "learning_rate": 0.00017815313688359546,
+      "loss": 1.637,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7152,
+      "grad_norm": 2.0116145606252926,
+      "learning_rate": 0.00017814235915926695,
+      "loss": 1.6728,
+      "step": 4470
+    },
+    {
+      "epoch": 0.71536,
+      "grad_norm": 3.50720726619715,
+      "learning_rate": 0.00017813157910328428,
+      "loss": 1.6482,
+      "step": 4471
+    },
+    {
+      "epoch": 0.71552,
+      "grad_norm": 2.216882617859541,
+      "learning_rate": 0.00017812079671596917,
+      "loss": 1.7549,
+      "step": 4472
+    },
+    {
+      "epoch": 0.71568,
+      "grad_norm": 2.9443967050043116,
+      "learning_rate": 0.0001781100119976433,
+      "loss": 1.7072,
+      "step": 4473
+    },
+    {
+      "epoch": 0.71584,
+      "grad_norm": 2.920768430041164,
+      "learning_rate": 0.0001780992249486285,
+      "loss": 1.6892,
+      "step": 4474
+    },
+    {
+      "epoch": 0.716,
+      "grad_norm": 2.8244146983258096,
+      "learning_rate": 0.00017808843556924663,
+      "loss": 1.7547,
+      "step": 4475
+    },
+    {
+      "epoch": 0.71616,
+      "grad_norm": 3.074067503583119,
+      "learning_rate": 0.0001780776438598196,
+      "loss": 1.6507,
+      "step": 4476
+    },
+    {
+      "epoch": 0.71632,
+      "grad_norm": 2.096410456605335,
+      "learning_rate": 0.0001780668498206695,
+      "loss": 1.7202,
+      "step": 4477
+    },
+    {
+      "epoch": 0.71648,
+      "grad_norm": 2.3562178365484487,
+      "learning_rate": 0.0001780560534521183,
+      "loss": 1.6238,
+      "step": 4478
+    },
+    {
+      "epoch": 0.71664,
+      "grad_norm": 2.7907887407814203,
+      "learning_rate": 0.00017804525475448825,
+      "loss": 1.7827,
+      "step": 4479
+    },
+    {
+      "epoch": 0.7168,
+      "grad_norm": 2.3869734768058706,
+      "learning_rate": 0.00017803445372810151,
+      "loss": 1.7329,
+      "step": 4480
+    },
+    {
+      "epoch": 0.71696,
+      "grad_norm": 2.1823933643161744,
+      "learning_rate": 0.0001780236503732804,
+      "loss": 1.5302,
+      "step": 4481
+    },
+    {
+      "epoch": 0.71712,
+      "grad_norm": 2.904050668981875,
+      "learning_rate": 0.0001780128446903472,
+      "loss": 1.6454,
+      "step": 4482
+    },
+    {
+      "epoch": 0.71728,
+      "grad_norm": 2.365406956665868,
+      "learning_rate": 0.00017800203667962442,
+      "loss": 1.6355,
+      "step": 4483
+    },
+    {
+      "epoch": 0.71744,
+      "grad_norm": 3.0362284489277216,
+      "learning_rate": 0.00017799122634143454,
+      "loss": 1.6805,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7176,
+      "grad_norm": 2.455808731465816,
+      "learning_rate": 0.00017798041367610013,
+      "loss": 1.7162,
+      "step": 4485
+    },
+    {
+      "epoch": 0.71776,
+      "grad_norm": 2.8344644597113264,
+      "learning_rate": 0.00017796959868394377,
+      "loss": 1.6356,
+      "step": 4486
+    },
+    {
+      "epoch": 0.71792,
+      "grad_norm": 2.4825383637417073,
+      "learning_rate": 0.00017795878136528818,
+      "loss": 1.6723,
+      "step": 4487
+    },
+    {
+      "epoch": 0.71808,
+      "grad_norm": 2.9794762053829977,
+      "learning_rate": 0.0001779479617204562,
+      "loss": 1.6478,
+      "step": 4488
+    },
+    {
+      "epoch": 0.71824,
+      "grad_norm": 2.4930686363223837,
+      "learning_rate": 0.00017793713974977057,
+      "loss": 1.644,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7184,
+      "grad_norm": 2.1835895904928333,
+      "learning_rate": 0.0001779263154535543,
+      "loss": 1.5871,
+      "step": 4490
+    },
+    {
+      "epoch": 0.71856,
+      "grad_norm": 2.548779070910138,
+      "learning_rate": 0.00017791548883213033,
+      "loss": 1.71,
+      "step": 4491
+    },
+    {
+      "epoch": 0.71872,
+      "grad_norm": 2.3438118332849847,
+      "learning_rate": 0.0001779046598858217,
+      "loss": 1.7358,
+      "step": 4492
+    },
+    {
+      "epoch": 0.71888,
+      "grad_norm": 2.744271357409253,
+      "learning_rate": 0.00017789382861495151,
+      "loss": 1.8066,
+      "step": 4493
+    },
+    {
+      "epoch": 0.71904,
+      "grad_norm": 2.0358345541838663,
+      "learning_rate": 0.000177882995019843,
+      "loss": 1.7128,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7192,
+      "grad_norm": 2.6675485410554933,
+      "learning_rate": 0.0001778721591008194,
+      "loss": 1.6259,
+      "step": 4495
+    },
+    {
+      "epoch": 0.71936,
+      "grad_norm": 2.3490636479097105,
+      "learning_rate": 0.00017786132085820407,
+      "loss": 1.6566,
+      "step": 4496
+    },
+    {
+      "epoch": 0.71952,
+      "grad_norm": 2.120257125541239,
+      "learning_rate": 0.00017785048029232033,
+      "loss": 1.7482,
+      "step": 4497
+    },
+    {
+      "epoch": 0.71968,
+      "grad_norm": 2.298709508139598,
+      "learning_rate": 0.00017783963740349175,
+      "loss": 1.7204,
+      "step": 4498
+    },
+    {
+      "epoch": 0.71984,
+      "grad_norm": 2.194595536766262,
+      "learning_rate": 0.0001778287921920418,
+      "loss": 1.7668,
+      "step": 4499
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 2.5062896924954225,
+      "learning_rate": 0.00017781794465829412,
+      "loss": 1.7874,
+      "step": 4500
+    },
+    {
+      "epoch": 0.72016,
+      "grad_norm": 2.7333239781570278,
+      "learning_rate": 0.00017780709480257233,
+      "loss": 1.7722,
+      "step": 4501
+    },
+    {
+      "epoch": 0.72032,
+      "grad_norm": 2.4683455306783397,
+      "learning_rate": 0.00017779624262520018,
+      "loss": 1.7325,
+      "step": 4502
+    },
+    {
+      "epoch": 0.72048,
+      "grad_norm": 2.3997977702703457,
+      "learning_rate": 0.00017778538812650155,
+      "loss": 1.6593,
+      "step": 4503
+    },
+    {
+      "epoch": 0.72064,
+      "grad_norm": 2.52693999439342,
+      "learning_rate": 0.0001777745313068003,
+      "loss": 1.6451,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7208,
+      "grad_norm": 2.782987707631361,
+      "learning_rate": 0.00017776367216642034,
+      "loss": 1.7839,
+      "step": 4505
+    },
+    {
+      "epoch": 0.72096,
+      "grad_norm": 3.3253900646322725,
+      "learning_rate": 0.00017775281070568573,
+      "loss": 1.7065,
+      "step": 4506
+    },
+    {
+      "epoch": 0.72112,
+      "grad_norm": 2.7287345801279335,
+      "learning_rate": 0.00017774194692492053,
+      "loss": 1.6866,
+      "step": 4507
+    },
+    {
+      "epoch": 0.72128,
+      "grad_norm": 2.724550521122648,
+      "learning_rate": 0.00017773108082444892,
+      "loss": 1.6229,
+      "step": 4508
+    },
+    {
+      "epoch": 0.72144,
+      "grad_norm": 2.303728671102178,
+      "learning_rate": 0.00017772021240459513,
+      "loss": 1.6214,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7216,
+      "grad_norm": 2.8209505269241704,
+      "learning_rate": 0.00017770934166568345,
+      "loss": 1.5432,
+      "step": 4510
+    },
+    {
+      "epoch": 0.72176,
+      "grad_norm": 2.2248835700552427,
+      "learning_rate": 0.00017769846860803826,
+      "loss": 1.6917,
+      "step": 4511
+    },
+    {
+      "epoch": 0.72192,
+      "grad_norm": 2.18403443718879,
+      "learning_rate": 0.00017768759323198396,
+      "loss": 1.6929,
+      "step": 4512
+    },
+    {
+      "epoch": 0.72208,
+      "grad_norm": 2.4447553204167254,
+      "learning_rate": 0.0001776767155378451,
+      "loss": 1.6382,
+      "step": 4513
+    },
+    {
+      "epoch": 0.72224,
+      "grad_norm": 2.179925333384405,
+      "learning_rate": 0.00017766583552594623,
+      "loss": 1.6988,
+      "step": 4514
+    },
+    {
+      "epoch": 0.7224,
+      "grad_norm": 2.8806145859931864,
+      "learning_rate": 0.000177654953196612,
+      "loss": 1.7229,
+      "step": 4515
+    },
+    {
+      "epoch": 0.72256,
+      "grad_norm": 2.3142163887452787,
+      "learning_rate": 0.0001776440685501671,
+      "loss": 1.7016,
+      "step": 4516
+    },
+    {
+      "epoch": 0.72272,
+      "grad_norm": 2.4347489553756194,
+      "learning_rate": 0.0001776331815869364,
+      "loss": 1.6707,
+      "step": 4517
+    },
+    {
+      "epoch": 0.72288,
+      "grad_norm": 2.4536451092774167,
+      "learning_rate": 0.00017762229230724463,
+      "loss": 1.6879,
+      "step": 4518
+    },
+    {
+      "epoch": 0.72304,
+      "grad_norm": 2.3354423183089588,
+      "learning_rate": 0.00017761140071141678,
+      "loss": 1.7368,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7232,
+      "grad_norm": 2.330869920124116,
+      "learning_rate": 0.0001776005067997778,
+      "loss": 1.6509,
+      "step": 4520
+    },
+    {
+      "epoch": 0.72336,
+      "grad_norm": 2.2624355038197814,
+      "learning_rate": 0.0001775896105726528,
+      "loss": 1.7133,
+      "step": 4521
+    },
+    {
+      "epoch": 0.72352,
+      "grad_norm": 2.3376975353856713,
+      "learning_rate": 0.00017757871203036686,
+      "loss": 1.8338,
+      "step": 4522
+    },
+    {
+      "epoch": 0.72368,
+      "grad_norm": 2.2538077837909394,
+      "learning_rate": 0.00017756781117324522,
+      "loss": 1.6666,
+      "step": 4523
+    },
+    {
+      "epoch": 0.72384,
+      "grad_norm": 2.633529201330951,
+      "learning_rate": 0.0001775569080016131,
+      "loss": 1.6971,
+      "step": 4524
+    },
+    {
+      "epoch": 0.724,
+      "grad_norm": 2.500479122114177,
+      "learning_rate": 0.00017754600251579586,
+      "loss": 1.743,
+      "step": 4525
+    },
+    {
+      "epoch": 0.72416,
+      "grad_norm": 2.300197166553934,
+      "learning_rate": 0.00017753509471611897,
+      "loss": 1.7043,
+      "step": 4526
+    },
+    {
+      "epoch": 0.72432,
+      "grad_norm": 2.247044019172769,
+      "learning_rate": 0.00017752418460290775,
+      "loss": 1.7649,
+      "step": 4527
+    },
+    {
+      "epoch": 0.72448,
+      "grad_norm": 2.247872664218221,
+      "learning_rate": 0.00017751327217648786,
+      "loss": 1.697,
+      "step": 4528
+    },
+    {
+      "epoch": 0.72464,
+      "grad_norm": 2.6493668043612515,
+      "learning_rate": 0.00017750235743718487,
+      "loss": 1.7414,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7248,
+      "grad_norm": 2.9607791118423266,
+      "learning_rate": 0.00017749144038532445,
+      "loss": 1.657,
+      "step": 4530
+    },
+    {
+      "epoch": 0.72496,
+      "grad_norm": 2.667673064338121,
+      "learning_rate": 0.00017748052102123239,
+      "loss": 1.7616,
+      "step": 4531
+    },
+    {
+      "epoch": 0.72512,
+      "grad_norm": 2.47916606949969,
+      "learning_rate": 0.00017746959934523447,
+      "loss": 1.5832,
+      "step": 4532
+    },
+    {
+      "epoch": 0.72528,
+      "grad_norm": 2.078889046297985,
+      "learning_rate": 0.0001774586753576566,
+      "loss": 1.7642,
+      "step": 4533
+    },
+    {
+      "epoch": 0.72544,
+      "grad_norm": 2.4767230223866528,
+      "learning_rate": 0.00017744774905882474,
+      "loss": 1.6886,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7256,
+      "grad_norm": 2.0355113930379862,
+      "learning_rate": 0.0001774368204490649,
+      "loss": 1.6386,
+      "step": 4535
+    },
+    {
+      "epoch": 0.72576,
+      "grad_norm": 2.34849578529315,
+      "learning_rate": 0.00017742588952870313,
+      "loss": 1.7486,
+      "step": 4536
+    },
+    {
+      "epoch": 0.72592,
+      "grad_norm": 2.488520637978939,
+      "learning_rate": 0.00017741495629806567,
+      "loss": 1.6834,
+      "step": 4537
+    },
+    {
+      "epoch": 0.72608,
+      "grad_norm": 3.0100342957394925,
+      "learning_rate": 0.0001774040207574787,
+      "loss": 1.7099,
+      "step": 4538
+    },
+    {
+      "epoch": 0.72624,
+      "grad_norm": 2.2283510441252896,
+      "learning_rate": 0.00017739308290726856,
+      "loss": 1.5948,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7264,
+      "grad_norm": 2.544387845006648,
+      "learning_rate": 0.00017738214274776163,
+      "loss": 1.7179,
+      "step": 4540
+    },
+    {
+      "epoch": 0.72656,
+      "grad_norm": 3.0872371068066946,
+      "learning_rate": 0.00017737120027928427,
+      "loss": 1.6581,
+      "step": 4541
+    },
+    {
+      "epoch": 0.72672,
+      "grad_norm": 2.2668623176999625,
+      "learning_rate": 0.00017736025550216304,
+      "loss": 1.6277,
+      "step": 4542
+    },
+    {
+      "epoch": 0.72688,
+      "grad_norm": 3.302121057574618,
+      "learning_rate": 0.00017734930841672454,
+      "loss": 1.6627,
+      "step": 4543
+    },
+    {
+      "epoch": 0.72704,
+      "grad_norm": 2.114704779290404,
+      "learning_rate": 0.00017733835902329536,
+      "loss": 1.6556,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7272,
+      "grad_norm": 2.5623031376156598,
+      "learning_rate": 0.00017732740732220224,
+      "loss": 1.7668,
+      "step": 4545
+    },
+    {
+      "epoch": 0.72736,
+      "grad_norm": 2.436549792739658,
+      "learning_rate": 0.00017731645331377197,
+      "loss": 1.6998,
+      "step": 4546
+    },
+    {
+      "epoch": 0.72752,
+      "grad_norm": 2.484687664159665,
+      "learning_rate": 0.00017730549699833143,
+      "loss": 1.7545,
+      "step": 4547
+    },
+    {
+      "epoch": 0.72768,
+      "grad_norm": 3.232121756363029,
+      "learning_rate": 0.00017729453837620745,
+      "loss": 1.7169,
+      "step": 4548
+    },
+    {
+      "epoch": 0.72784,
+      "grad_norm": 2.7490955296324104,
+      "learning_rate": 0.0001772835774477271,
+      "loss": 1.7392,
+      "step": 4549
+    },
+    {
+      "epoch": 0.728,
+      "grad_norm": 2.6734715809217744,
+      "learning_rate": 0.00017727261421321742,
+      "loss": 1.7647,
+      "step": 4550
+    },
+    {
+      "epoch": 0.72816,
+      "grad_norm": 2.5868381006764114,
+      "learning_rate": 0.0001772616486730055,
+      "loss": 1.6516,
+      "step": 4551
+    },
+    {
+      "epoch": 0.72832,
+      "grad_norm": 2.8289695662208088,
+      "learning_rate": 0.00017725068082741864,
+      "loss": 1.7145,
+      "step": 4552
+    },
+    {
+      "epoch": 0.72848,
+      "grad_norm": 2.3713726803177972,
+      "learning_rate": 0.00017723971067678396,
+      "loss": 1.7515,
+      "step": 4553
+    },
+    {
+      "epoch": 0.72864,
+      "grad_norm": 2.780344523033464,
+      "learning_rate": 0.00017722873822142887,
+      "loss": 1.729,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7288,
+      "grad_norm": 2.435957329703587,
+      "learning_rate": 0.00017721776346168078,
+      "loss": 1.8169,
+      "step": 4555
+    },
+    {
+      "epoch": 0.72896,
+      "grad_norm": 1.9586041188026218,
+      "learning_rate": 0.00017720678639786714,
+      "loss": 1.6528,
+      "step": 4556
+    },
+    {
+      "epoch": 0.72912,
+      "grad_norm": 2.2137106465489595,
+      "learning_rate": 0.0001771958070303155,
+      "loss": 1.7465,
+      "step": 4557
+    },
+    {
+      "epoch": 0.72928,
+      "grad_norm": 2.1282088216237343,
+      "learning_rate": 0.00017718482535935347,
+      "loss": 1.7265,
+      "step": 4558
+    },
+    {
+      "epoch": 0.72944,
+      "grad_norm": 2.288394047087613,
+      "learning_rate": 0.00017717384138530873,
+      "loss": 1.6606,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7296,
+      "grad_norm": 2.36725156758537,
+      "learning_rate": 0.00017716285510850898,
+      "loss": 1.6945,
+      "step": 4560
+    },
+    {
+      "epoch": 0.72976,
+      "grad_norm": 2.6178150527977757,
+      "learning_rate": 0.0001771518665292821,
+      "loss": 1.6366,
+      "step": 4561
+    },
+    {
+      "epoch": 0.72992,
+      "grad_norm": 2.232414368869648,
+      "learning_rate": 0.00017714087564795595,
+      "loss": 1.703,
+      "step": 4562
+    },
+    {
+      "epoch": 0.73008,
+      "grad_norm": 2.3960932661963663,
+      "learning_rate": 0.00017712988246485845,
+      "loss": 1.7743,
+      "step": 4563
+    },
+    {
+      "epoch": 0.73024,
+      "grad_norm": 2.60822793453697,
+      "learning_rate": 0.00017711888698031768,
+      "loss": 1.6821,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7304,
+      "grad_norm": 2.265536190685023,
+      "learning_rate": 0.0001771078891946617,
+      "loss": 1.6843,
+      "step": 4565
+    },
+    {
+      "epoch": 0.73056,
+      "grad_norm": 2.0182862175211898,
+      "learning_rate": 0.00017709688910821862,
+      "loss": 1.6946,
+      "step": 4566
+    },
+    {
+      "epoch": 0.73072,
+      "grad_norm": 2.819680308523656,
+      "learning_rate": 0.00017708588672131674,
+      "loss": 1.7267,
+      "step": 4567
+    },
+    {
+      "epoch": 0.73088,
+      "grad_norm": 2.412748202125236,
+      "learning_rate": 0.00017707488203428433,
+      "loss": 1.6899,
+      "step": 4568
+    },
+    {
+      "epoch": 0.73104,
+      "grad_norm": 2.4658188743988414,
+      "learning_rate": 0.0001770638750474497,
+      "loss": 1.6882,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7312,
+      "grad_norm": 2.338760438072754,
+      "learning_rate": 0.0001770528657611414,
+      "loss": 1.7271,
+      "step": 4570
+    },
+    {
+      "epoch": 0.73136,
+      "grad_norm": 2.367581377177549,
+      "learning_rate": 0.00017704185417568785,
+      "loss": 1.7037,
+      "step": 4571
+    },
+    {
+      "epoch": 0.73152,
+      "grad_norm": 2.390409869317122,
+      "learning_rate": 0.00017703084029141761,
+      "loss": 1.7179,
+      "step": 4572
+    },
+    {
+      "epoch": 0.73168,
+      "grad_norm": 2.5054836427970004,
+      "learning_rate": 0.00017701982410865937,
+      "loss": 1.6971,
+      "step": 4573
+    },
+    {
+      "epoch": 0.73184,
+      "grad_norm": 2.253349995763219,
+      "learning_rate": 0.00017700880562774177,
+      "loss": 1.6691,
+      "step": 4574
+    },
+    {
+      "epoch": 0.732,
+      "grad_norm": 2.5892694741678994,
+      "learning_rate": 0.00017699778484899366,
+      "loss": 1.6712,
+      "step": 4575
+    },
+    {
+      "epoch": 0.73216,
+      "grad_norm": 2.4497002632799614,
+      "learning_rate": 0.00017698676177274383,
+      "loss": 1.7281,
+      "step": 4576
+    },
+    {
+      "epoch": 0.73232,
+      "grad_norm": 2.2451608156278944,
+      "learning_rate": 0.00017697573639932122,
+      "loss": 1.7905,
+      "step": 4577
+    },
+    {
+      "epoch": 0.73248,
+      "grad_norm": 2.059750022590638,
+      "learning_rate": 0.00017696470872905477,
+      "loss": 1.6481,
+      "step": 4578
+    },
+    {
+      "epoch": 0.73264,
+      "grad_norm": 2.4835284484802895,
+      "learning_rate": 0.0001769536787622736,
+      "loss": 1.7205,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7328,
+      "grad_norm": 1.9694702592832145,
+      "learning_rate": 0.00017694264649930678,
+      "loss": 1.7643,
+      "step": 4580
+    },
+    {
+      "epoch": 0.73296,
+      "grad_norm": 2.152290162512616,
+      "learning_rate": 0.0001769316119404835,
+      "loss": 1.7192,
+      "step": 4581
+    },
+    {
+      "epoch": 0.73312,
+      "grad_norm": 2.2655673168707082,
+      "learning_rate": 0.00017692057508613302,
+      "loss": 1.6931,
+      "step": 4582
+    },
+    {
+      "epoch": 0.73328,
+      "grad_norm": 2.6072613775557834,
+      "learning_rate": 0.00017690953593658466,
+      "loss": 1.8223,
+      "step": 4583
+    },
+    {
+      "epoch": 0.73344,
+      "grad_norm": 2.648238400123474,
+      "learning_rate": 0.00017689849449216782,
+      "loss": 1.6724,
+      "step": 4584
+    },
+    {
+      "epoch": 0.7336,
+      "grad_norm": 2.090148930404113,
+      "learning_rate": 0.000176887450753212,
+      "loss": 1.723,
+      "step": 4585
+    },
+    {
+      "epoch": 0.73376,
+      "grad_norm": 2.241860527094521,
+      "learning_rate": 0.00017687640472004664,
+      "loss": 1.723,
+      "step": 4586
+    },
+    {
+      "epoch": 0.73392,
+      "grad_norm": 2.166445431759497,
+      "learning_rate": 0.00017686535639300138,
+      "loss": 1.5917,
+      "step": 4587
+    },
+    {
+      "epoch": 0.73408,
+      "grad_norm": 2.1487286589173036,
+      "learning_rate": 0.00017685430577240587,
+      "loss": 1.7298,
+      "step": 4588
+    },
+    {
+      "epoch": 0.73424,
+      "grad_norm": 2.066957122332422,
+      "learning_rate": 0.0001768432528585899,
+      "loss": 1.6372,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7344,
+      "grad_norm": 2.8130057195533635,
+      "learning_rate": 0.00017683219765188323,
+      "loss": 1.7079,
+      "step": 4590
+    },
+    {
+      "epoch": 0.73456,
+      "grad_norm": 2.2461160370379236,
+      "learning_rate": 0.00017682114015261574,
+      "loss": 1.7219,
+      "step": 4591
+    },
+    {
+      "epoch": 0.73472,
+      "grad_norm": 2.144280350714752,
+      "learning_rate": 0.00017681008036111737,
+      "loss": 1.73,
+      "step": 4592
+    },
+    {
+      "epoch": 0.73488,
+      "grad_norm": 2.088096157249695,
+      "learning_rate": 0.00017679901827771812,
+      "loss": 1.6906,
+      "step": 4593
+    },
+    {
+      "epoch": 0.73504,
+      "grad_norm": 2.548632330883097,
+      "learning_rate": 0.00017678795390274808,
+      "loss": 1.771,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7352,
+      "grad_norm": 2.2841171061360197,
+      "learning_rate": 0.0001767768872365374,
+      "loss": 1.7264,
+      "step": 4595
+    },
+    {
+      "epoch": 0.73536,
+      "grad_norm": 2.400951523659184,
+      "learning_rate": 0.00017676581827941626,
+      "loss": 1.5889,
+      "step": 4596
+    },
+    {
+      "epoch": 0.73552,
+      "grad_norm": 2.7683446952214266,
+      "learning_rate": 0.00017675474703171497,
+      "loss": 1.72,
+      "step": 4597
+    },
+    {
+      "epoch": 0.73568,
+      "grad_norm": 3.084394377591071,
+      "learning_rate": 0.00017674367349376386,
+      "loss": 1.7287,
+      "step": 4598
+    },
+    {
+      "epoch": 0.73584,
+      "grad_norm": 2.7696643856382708,
+      "learning_rate": 0.0001767325976658934,
+      "loss": 1.7487,
+      "step": 4599
+    },
+    {
+      "epoch": 0.736,
+      "grad_norm": 2.5398972852118575,
+      "learning_rate": 0.00017672151954843397,
+      "loss": 1.6624,
+      "step": 4600
+    },
+    {
+      "epoch": 0.73616,
+      "grad_norm": 2.481805495008607,
+      "learning_rate": 0.00017671043914171624,
+      "loss": 1.6317,
+      "step": 4601
+    },
+    {
+      "epoch": 0.73632,
+      "grad_norm": 2.1487932476371245,
+      "learning_rate": 0.00017669935644607076,
+      "loss": 1.7425,
+      "step": 4602
+    },
+    {
+      "epoch": 0.73648,
+      "grad_norm": 2.563670154663882,
+      "learning_rate": 0.00017668827146182826,
+      "loss": 1.6694,
+      "step": 4603
+    },
+    {
+      "epoch": 0.73664,
+      "grad_norm": 2.6302078638274553,
+      "learning_rate": 0.0001766771841893195,
+      "loss": 1.66,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7368,
+      "grad_norm": 2.1212564679514627,
+      "learning_rate": 0.00017666609462887526,
+      "loss": 1.671,
+      "step": 4605
+    },
+    {
+      "epoch": 0.73696,
+      "grad_norm": 2.697219233477335,
+      "learning_rate": 0.00017665500278082652,
+      "loss": 1.6949,
+      "step": 4606
+    },
+    {
+      "epoch": 0.73712,
+      "grad_norm": 2.311134790796858,
+      "learning_rate": 0.00017664390864550414,
+      "loss": 1.6271,
+      "step": 4607
+    },
+    {
+      "epoch": 0.73728,
+      "grad_norm": 2.6615025511618833,
+      "learning_rate": 0.00017663281222323924,
+      "loss": 1.6133,
+      "step": 4608
+    },
+    {
+      "epoch": 0.73744,
+      "grad_norm": 2.1161342717136122,
+      "learning_rate": 0.0001766217135143629,
+      "loss": 1.6461,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7376,
+      "grad_norm": 2.7314949712417724,
+      "learning_rate": 0.00017661061251920623,
+      "loss": 1.7271,
+      "step": 4610
+    },
+    {
+      "epoch": 0.73776,
+      "grad_norm": 2.5620338113940777,
+      "learning_rate": 0.00017659950923810058,
+      "loss": 1.6282,
+      "step": 4611
+    },
+    {
+      "epoch": 0.73792,
+      "grad_norm": 2.4158277922839195,
+      "learning_rate": 0.00017658840367137715,
+      "loss": 1.7306,
+      "step": 4612
+    },
+    {
+      "epoch": 0.73808,
+      "grad_norm": 2.292186576275612,
+      "learning_rate": 0.00017657729581936735,
+      "loss": 1.6701,
+      "step": 4613
+    },
+    {
+      "epoch": 0.73824,
+      "grad_norm": 2.239567672308328,
+      "learning_rate": 0.00017656618568240266,
+      "loss": 1.8609,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7384,
+      "grad_norm": 2.4757022028581726,
+      "learning_rate": 0.00017655507326081452,
+      "loss": 1.7169,
+      "step": 4615
+    },
+    {
+      "epoch": 0.73856,
+      "grad_norm": 2.5692207652653813,
+      "learning_rate": 0.00017654395855493457,
+      "loss": 1.7046,
+      "step": 4616
+    },
+    {
+      "epoch": 0.73872,
+      "grad_norm": 2.74723591184169,
+      "learning_rate": 0.0001765328415650944,
+      "loss": 1.7551,
+      "step": 4617
+    },
+    {
+      "epoch": 0.73888,
+      "grad_norm": 2.149240413073301,
+      "learning_rate": 0.00017652172229162583,
+      "loss": 1.6527,
+      "step": 4618
+    },
+    {
+      "epoch": 0.73904,
+      "grad_norm": 2.2294789780220423,
+      "learning_rate": 0.0001765106007348605,
+      "loss": 1.6906,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7392,
+      "grad_norm": 2.378662066844825,
+      "learning_rate": 0.00017649947689513037,
+      "loss": 1.6638,
+      "step": 4620
+    },
+    {
+      "epoch": 0.73936,
+      "grad_norm": 2.757241560403113,
+      "learning_rate": 0.0001764883507727673,
+      "loss": 1.7068,
+      "step": 4621
+    },
+    {
+      "epoch": 0.73952,
+      "grad_norm": 2.1553754685682325,
+      "learning_rate": 0.0001764772223681033,
+      "loss": 1.6893,
+      "step": 4622
+    },
+    {
+      "epoch": 0.73968,
+      "grad_norm": 2.1527664836704736,
+      "learning_rate": 0.00017646609168147045,
+      "loss": 1.6813,
+      "step": 4623
+    },
+    {
+      "epoch": 0.73984,
+      "grad_norm": 2.217526635145077,
+      "learning_rate": 0.00017645495871320083,
+      "loss": 1.7565,
+      "step": 4624
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 2.4104401877013886,
+      "learning_rate": 0.00017644382346362666,
+      "loss": 1.6218,
+      "step": 4625
+    },
+    {
+      "epoch": 0.74016,
+      "grad_norm": 2.473233928305038,
+      "learning_rate": 0.00017643268593308019,
+      "loss": 1.7168,
+      "step": 4626
+    },
+    {
+      "epoch": 0.74032,
+      "grad_norm": 2.438283966522188,
+      "learning_rate": 0.0001764215461218937,
+      "loss": 1.6571,
+      "step": 4627
+    },
+    {
+      "epoch": 0.74048,
+      "grad_norm": 2.672864463814738,
+      "learning_rate": 0.00017641040403039969,
+      "loss": 1.8004,
+      "step": 4628
+    },
+    {
+      "epoch": 0.74064,
+      "grad_norm": 2.2698125376128133,
+      "learning_rate": 0.00017639925965893052,
+      "loss": 1.6498,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7408,
+      "grad_norm": 2.121492992827488,
+      "learning_rate": 0.0001763881130078188,
+      "loss": 1.6366,
+      "step": 4630
+    },
+    {
+      "epoch": 0.74096,
+      "grad_norm": 2.5466836414694836,
+      "learning_rate": 0.00017637696407739708,
+      "loss": 1.7488,
+      "step": 4631
+    },
+    {
+      "epoch": 0.74112,
+      "grad_norm": 3.0752359319737748,
+      "learning_rate": 0.00017636581286799806,
+      "loss": 1.709,
+      "step": 4632
+    },
+    {
+      "epoch": 0.74128,
+      "grad_norm": 2.5634073479254713,
+      "learning_rate": 0.00017635465937995443,
+      "loss": 1.7899,
+      "step": 4633
+    },
+    {
+      "epoch": 0.74144,
+      "grad_norm": 2.6234794211329104,
+      "learning_rate": 0.00017634350361359907,
+      "loss": 1.6676,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7416,
+      "grad_norm": 2.4130780800102323,
+      "learning_rate": 0.00017633234556926474,
+      "loss": 1.7252,
+      "step": 4635
+    },
+    {
+      "epoch": 0.74176,
+      "grad_norm": 2.5918852142482227,
+      "learning_rate": 0.00017632118524728448,
+      "loss": 1.6599,
+      "step": 4636
+    },
+    {
+      "epoch": 0.74192,
+      "grad_norm": 2.4859509354978098,
+      "learning_rate": 0.0001763100226479913,
+      "loss": 1.6397,
+      "step": 4637
+    },
+    {
+      "epoch": 0.74208,
+      "grad_norm": 2.5904980768282333,
+      "learning_rate": 0.0001762988577717182,
+      "loss": 1.6984,
+      "step": 4638
+    },
+    {
+      "epoch": 0.74224,
+      "grad_norm": 3.035927557063506,
+      "learning_rate": 0.00017628769061879838,
+      "loss": 1.596,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7424,
+      "grad_norm": 2.2103504708051362,
+      "learning_rate": 0.00017627652118956503,
+      "loss": 1.763,
+      "step": 4640
+    },
+    {
+      "epoch": 0.74256,
+      "grad_norm": 2.4705063774354143,
+      "learning_rate": 0.00017626534948435145,
+      "loss": 1.6751,
+      "step": 4641
+    },
+    {
+      "epoch": 0.74272,
+      "grad_norm": 2.705128175213492,
+      "learning_rate": 0.00017625417550349095,
+      "loss": 1.7825,
+      "step": 4642
+    },
+    {
+      "epoch": 0.74288,
+      "grad_norm": 2.4555377900817525,
+      "learning_rate": 0.00017624299924731695,
+      "loss": 1.7574,
+      "step": 4643
+    },
+    {
+      "epoch": 0.74304,
+      "grad_norm": 2.868866982810291,
+      "learning_rate": 0.00017623182071616297,
+      "loss": 1.8111,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7432,
+      "grad_norm": 2.340902308517237,
+      "learning_rate": 0.00017622063991036256,
+      "loss": 1.6731,
+      "step": 4645
+    },
+    {
+      "epoch": 0.74336,
+      "grad_norm": 2.642059075016513,
+      "learning_rate": 0.0001762094568302493,
+      "loss": 1.7516,
+      "step": 4646
+    },
+    {
+      "epoch": 0.74352,
+      "grad_norm": 1.9622439259178042,
+      "learning_rate": 0.00017619827147615687,
+      "loss": 1.6092,
+      "step": 4647
+    },
+    {
+      "epoch": 0.74368,
+      "grad_norm": 2.3712810222886334,
+      "learning_rate": 0.0001761870838484191,
+      "loss": 1.7275,
+      "step": 4648
+    },
+    {
+      "epoch": 0.74384,
+      "grad_norm": 2.1888544769515654,
+      "learning_rate": 0.00017617589394736976,
+      "loss": 1.6801,
+      "step": 4649
+    },
+    {
+      "epoch": 0.744,
+      "grad_norm": 2.2736442370299126,
+      "learning_rate": 0.00017616470177334273,
+      "loss": 1.6824,
+      "step": 4650
+    },
+    {
+      "epoch": 0.74416,
+      "grad_norm": 2.513617689394156,
+      "learning_rate": 0.00017615350732667198,
+      "loss": 1.6443,
+      "step": 4651
+    },
+    {
+      "epoch": 0.74432,
+      "grad_norm": 2.374004275884025,
+      "learning_rate": 0.00017614231060769154,
+      "loss": 1.8012,
+      "step": 4652
+    },
+    {
+      "epoch": 0.74448,
+      "grad_norm": 2.8585474462665634,
+      "learning_rate": 0.0001761311116167355,
+      "loss": 1.7384,
+      "step": 4653
+    },
+    {
+      "epoch": 0.74464,
+      "grad_norm": 2.1718941410669217,
+      "learning_rate": 0.00017611991035413805,
+      "loss": 1.6807,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7448,
+      "grad_norm": 1.8033216802488725,
+      "learning_rate": 0.00017610870682023335,
+      "loss": 1.6899,
+      "step": 4655
+    },
+    {
+      "epoch": 0.74496,
+      "grad_norm": 2.325331052996999,
+      "learning_rate": 0.00017609750101535578,
+      "loss": 1.9096,
+      "step": 4656
+    },
+    {
+      "epoch": 0.74512,
+      "grad_norm": 2.851843771714418,
+      "learning_rate": 0.00017608629293983964,
+      "loss": 1.7119,
+      "step": 4657
+    },
+    {
+      "epoch": 0.74528,
+      "grad_norm": 1.9674069583064469,
+      "learning_rate": 0.0001760750825940194,
+      "loss": 1.6577,
+      "step": 4658
+    },
+    {
+      "epoch": 0.74544,
+      "grad_norm": 3.276923883385714,
+      "learning_rate": 0.00017606386997822955,
+      "loss": 1.6878,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7456,
+      "grad_norm": 3.298517839275938,
+      "learning_rate": 0.00017605265509280463,
+      "loss": 1.7063,
+      "step": 4660
+    },
+    {
+      "epoch": 0.74576,
+      "grad_norm": 2.3987413994428604,
+      "learning_rate": 0.00017604143793807935,
+      "loss": 1.7032,
+      "step": 4661
+    },
+    {
+      "epoch": 0.74592,
+      "grad_norm": 2.4461762282576314,
+      "learning_rate": 0.0001760302185143883,
+      "loss": 1.6587,
+      "step": 4662
+    },
+    {
+      "epoch": 0.74608,
+      "grad_norm": 2.4574330097183696,
+      "learning_rate": 0.00017601899682206639,
+      "loss": 1.6263,
+      "step": 4663
+    },
+    {
+      "epoch": 0.74624,
+      "grad_norm": 2.0479846898723646,
+      "learning_rate": 0.00017600777286144837,
+      "loss": 1.6503,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7464,
+      "grad_norm": 2.411078877201147,
+      "learning_rate": 0.00017599654663286914,
+      "loss": 1.7271,
+      "step": 4665
+    },
+    {
+      "epoch": 0.74656,
+      "grad_norm": 2.782128504311011,
+      "learning_rate": 0.0001759853181366637,
+      "loss": 1.6797,
+      "step": 4666
+    },
+    {
+      "epoch": 0.74672,
+      "grad_norm": 2.3380893892646197,
+      "learning_rate": 0.0001759740873731671,
+      "loss": 1.725,
+      "step": 4667
+    },
+    {
+      "epoch": 0.74688,
+      "grad_norm": 2.1783861182398843,
+      "learning_rate": 0.00017596285434271443,
+      "loss": 1.6563,
+      "step": 4668
+    },
+    {
+      "epoch": 0.74704,
+      "grad_norm": 2.3731827891000803,
+      "learning_rate": 0.00017595161904564085,
+      "loss": 1.7298,
+      "step": 4669
+    },
+    {
+      "epoch": 0.7472,
+      "grad_norm": 2.0723028013362406,
+      "learning_rate": 0.00017594038148228166,
+      "loss": 1.6516,
+      "step": 4670
+    },
+    {
+      "epoch": 0.74736,
+      "grad_norm": 2.1501110105069596,
+      "learning_rate": 0.00017592914165297216,
+      "loss": 1.6973,
+      "step": 4671
+    },
+    {
+      "epoch": 0.74752,
+      "grad_norm": 2.598237185483074,
+      "learning_rate": 0.0001759178995580477,
+      "loss": 1.6879,
+      "step": 4672
+    },
+    {
+      "epoch": 0.74768,
+      "grad_norm": 2.2217391130207806,
+      "learning_rate": 0.00017590665519784374,
+      "loss": 1.6759,
+      "step": 4673
+    },
+    {
+      "epoch": 0.74784,
+      "grad_norm": 2.782579079694914,
+      "learning_rate": 0.00017589540857269577,
+      "loss": 1.742,
+      "step": 4674
+    },
+    {
+      "epoch": 0.748,
+      "grad_norm": 2.4712998358024296,
+      "learning_rate": 0.00017588415968293944,
+      "loss": 1.6397,
+      "step": 4675
+    },
+    {
+      "epoch": 0.74816,
+      "grad_norm": 2.976650805376667,
+      "learning_rate": 0.00017587290852891035,
+      "loss": 1.6266,
+      "step": 4676
+    },
+    {
+      "epoch": 0.74832,
+      "grad_norm": 2.4765597911601414,
+      "learning_rate": 0.0001758616551109442,
+      "loss": 1.6496,
+      "step": 4677
+    },
+    {
+      "epoch": 0.74848,
+      "grad_norm": 3.3647076232686075,
+      "learning_rate": 0.00017585039942937685,
+      "loss": 1.6508,
+      "step": 4678
+    },
+    {
+      "epoch": 0.74864,
+      "grad_norm": 3.088196888250998,
+      "learning_rate": 0.0001758391414845441,
+      "loss": 1.7342,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7488,
+      "grad_norm": 2.1587760509180964,
+      "learning_rate": 0.00017582788127678184,
+      "loss": 1.6716,
+      "step": 4680
+    },
+    {
+      "epoch": 0.74896,
+      "grad_norm": 2.3236065072707506,
+      "learning_rate": 0.0001758166188064261,
+      "loss": 1.7372,
+      "step": 4681
+    },
+    {
+      "epoch": 0.74912,
+      "grad_norm": 2.65456286906577,
+      "learning_rate": 0.00017580535407381293,
+      "loss": 1.7364,
+      "step": 4682
+    },
+    {
+      "epoch": 0.74928,
+      "grad_norm": 2.514809083818637,
+      "learning_rate": 0.00017579408707927847,
+      "loss": 1.7171,
+      "step": 4683
+    },
+    {
+      "epoch": 0.74944,
+      "grad_norm": 2.395209188484907,
+      "learning_rate": 0.00017578281782315892,
+      "loss": 1.7471,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7496,
+      "grad_norm": 2.2011738923941717,
+      "learning_rate": 0.00017577154630579048,
+      "loss": 1.6263,
+      "step": 4685
+    },
+    {
+      "epoch": 0.74976,
+      "grad_norm": 2.2899702023918183,
+      "learning_rate": 0.0001757602725275095,
+      "loss": 1.6907,
+      "step": 4686
+    },
+    {
+      "epoch": 0.74992,
+      "grad_norm": 2.5251462155288205,
+      "learning_rate": 0.00017574899648865243,
+      "loss": 1.607,
+      "step": 4687
+    },
+    {
+      "epoch": 0.75008,
+      "grad_norm": 2.4311098968277154,
+      "learning_rate": 0.00017573771818955565,
+      "loss": 1.6696,
+      "step": 4688
+    },
+    {
+      "epoch": 0.75024,
+      "grad_norm": 2.4311098968277154,
+      "learning_rate": 0.00017573771818955565,
+      "loss": 2.0592,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7504,
+      "grad_norm": 2.54109230929588,
+      "learning_rate": 0.00017572643763055573,
+      "loss": 1.719,
+      "step": 4690
+    },
+    {
+      "epoch": 0.75056,
+      "grad_norm": 2.7077134720779608,
+      "learning_rate": 0.00017571515481198924,
+      "loss": 1.63,
+      "step": 4691
+    },
+    {
+      "epoch": 0.75072,
+      "grad_norm": 2.1434120578009064,
+      "learning_rate": 0.00017570386973419285,
+      "loss": 1.6416,
+      "step": 4692
+    },
+    {
+      "epoch": 0.75088,
+      "grad_norm": 3.2676263404019132,
+      "learning_rate": 0.00017569258239750335,
+      "loss": 1.6383,
+      "step": 4693
+    },
+    {
+      "epoch": 0.75104,
+      "grad_norm": 2.185432578094176,
+      "learning_rate": 0.00017568129280225744,
+      "loss": 1.6132,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7512,
+      "grad_norm": 3.6561240056456157,
+      "learning_rate": 0.00017567000094879205,
+      "loss": 1.7289,
+      "step": 4695
+    },
+    {
+      "epoch": 0.75136,
+      "grad_norm": 2.304436359623227,
+      "learning_rate": 0.0001756587068374441,
+      "loss": 1.6708,
+      "step": 4696
+    },
+    {
+      "epoch": 0.75152,
+      "grad_norm": 3.0612942538727705,
+      "learning_rate": 0.00017564741046855055,
+      "loss": 1.7411,
+      "step": 4697
+    },
+    {
+      "epoch": 0.75168,
+      "grad_norm": 3.18825395779705,
+      "learning_rate": 0.00017563611184244853,
+      "loss": 1.6684,
+      "step": 4698
+    },
+    {
+      "epoch": 0.75184,
+      "grad_norm": 3.2504178509868784,
+      "learning_rate": 0.00017562481095947515,
+      "loss": 1.6547,
+      "step": 4699
+    },
+    {
+      "epoch": 0.752,
+      "grad_norm": 3.425065722526836,
+      "learning_rate": 0.0001756135078199676,
+      "loss": 1.7935,
+      "step": 4700
+    },
+    {
+      "epoch": 0.75216,
+      "grad_norm": 3.049375433079451,
+      "learning_rate": 0.0001756022024242632,
+      "loss": 1.7042,
+      "step": 4701
+    },
+    {
+      "epoch": 0.75232,
+      "grad_norm": 3.0331495707133995,
+      "learning_rate": 0.00017559089477269918,
+      "loss": 1.6662,
+      "step": 4702
+    },
+    {
+      "epoch": 0.75248,
+      "grad_norm": 2.694574753124913,
+      "learning_rate": 0.00017557958486561303,
+      "loss": 1.5702,
+      "step": 4703
+    },
+    {
+      "epoch": 0.75264,
+      "grad_norm": 2.729525760713284,
+      "learning_rate": 0.00017556827270334216,
+      "loss": 1.7026,
+      "step": 4704
+    },
+    {
+      "epoch": 0.7528,
+      "grad_norm": 2.429255685269803,
+      "learning_rate": 0.00017555695828622422,
+      "loss": 1.8164,
+      "step": 4705
+    },
+    {
+      "epoch": 0.75296,
+      "grad_norm": 3.3180760150500497,
+      "learning_rate": 0.0001755456416145967,
+      "loss": 1.7676,
+      "step": 4706
+    },
+    {
+      "epoch": 0.75312,
+      "grad_norm": 2.233499023676405,
+      "learning_rate": 0.00017553432268879732,
+      "loss": 1.5953,
+      "step": 4707
+    },
+    {
+      "epoch": 0.75328,
+      "grad_norm": 3.1605477542748384,
+      "learning_rate": 0.00017552300150916383,
+      "loss": 1.7195,
+      "step": 4708
+    },
+    {
+      "epoch": 0.75344,
+      "grad_norm": 2.342545801644279,
+      "learning_rate": 0.000175511678076034,
+      "loss": 1.7186,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7536,
+      "grad_norm": 3.110404266772865,
+      "learning_rate": 0.00017550035238974573,
+      "loss": 1.6518,
+      "step": 4710
+    },
+    {
+      "epoch": 0.75376,
+      "grad_norm": 2.871871311303952,
+      "learning_rate": 0.000175489024450637,
+      "loss": 1.7895,
+      "step": 4711
+    },
+    {
+      "epoch": 0.75392,
+      "grad_norm": 2.5376380343882885,
+      "learning_rate": 0.00017547769425904573,
+      "loss": 1.8296,
+      "step": 4712
+    },
+    {
+      "epoch": 0.75408,
+      "grad_norm": 2.1269463892402913,
+      "learning_rate": 0.0001754663618153101,
+      "loss": 1.7206,
+      "step": 4713
+    },
+    {
+      "epoch": 0.75424,
+      "grad_norm": 2.2949261815012947,
+      "learning_rate": 0.00017545502711976816,
+      "loss": 1.6577,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7544,
+      "grad_norm": 2.435064094761272,
+      "learning_rate": 0.00017544369017275817,
+      "loss": 1.6855,
+      "step": 4715
+    },
+    {
+      "epoch": 0.75456,
+      "grad_norm": 2.377782917976421,
+      "learning_rate": 0.0001754323509746184,
+      "loss": 1.667,
+      "step": 4716
+    },
+    {
+      "epoch": 0.75472,
+      "grad_norm": 2.451457457379235,
+      "learning_rate": 0.0001754210095256872,
+      "loss": 1.6819,
+      "step": 4717
+    },
+    {
+      "epoch": 0.75488,
+      "grad_norm": 2.42158903036211,
+      "learning_rate": 0.00017540966582630296,
+      "loss": 1.7442,
+      "step": 4718
+    },
+    {
+      "epoch": 0.75504,
+      "grad_norm": 2.880563993363573,
+      "learning_rate": 0.00017539831987680417,
+      "loss": 1.5794,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7552,
+      "grad_norm": 2.167280271584904,
+      "learning_rate": 0.0001753869716775294,
+      "loss": 1.6596,
+      "step": 4720
+    },
+    {
+      "epoch": 0.75536,
+      "grad_norm": 3.329472549270715,
+      "learning_rate": 0.00017537562122881724,
+      "loss": 1.7188,
+      "step": 4721
+    },
+    {
+      "epoch": 0.75552,
+      "grad_norm": 2.2537267561904906,
+      "learning_rate": 0.00017536426853100636,
+      "loss": 1.7061,
+      "step": 4722
+    },
+    {
+      "epoch": 0.75568,
+      "grad_norm": 1.8605091193924415,
+      "learning_rate": 0.00017535291358443552,
+      "loss": 1.6299,
+      "step": 4723
+    },
+    {
+      "epoch": 0.75584,
+      "grad_norm": 2.2796793217700024,
+      "learning_rate": 0.00017534155638944358,
+      "loss": 1.6443,
+      "step": 4724
+    },
+    {
+      "epoch": 0.756,
+      "grad_norm": 2.225370029666905,
+      "learning_rate": 0.00017533019694636936,
+      "loss": 1.7022,
+      "step": 4725
+    },
+    {
+      "epoch": 0.75616,
+      "grad_norm": 2.44845243909884,
+      "learning_rate": 0.00017531883525555183,
+      "loss": 1.7393,
+      "step": 4726
+    },
+    {
+      "epoch": 0.75632,
+      "grad_norm": 2.750503517215255,
+      "learning_rate": 0.00017530747131733003,
+      "loss": 1.6196,
+      "step": 4727
+    },
+    {
+      "epoch": 0.75648,
+      "grad_norm": 2.077744321454504,
+      "learning_rate": 0.000175296105132043,
+      "loss": 1.6768,
+      "step": 4728
+    },
+    {
+      "epoch": 0.75664,
+      "grad_norm": 2.5160871273177223,
+      "learning_rate": 0.0001752847367000299,
+      "loss": 1.7304,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7568,
+      "grad_norm": 2.3472959980191463,
+      "learning_rate": 0.00017527336602162996,
+      "loss": 1.6111,
+      "step": 4730
+    },
+    {
+      "epoch": 0.75696,
+      "grad_norm": 2.2553422997482038,
+      "learning_rate": 0.00017526199309718247,
+      "loss": 1.6431,
+      "step": 4731
+    },
+    {
+      "epoch": 0.75712,
+      "grad_norm": 2.5572306327382996,
+      "learning_rate": 0.00017525061792702677,
+      "loss": 1.7012,
+      "step": 4732
+    },
+    {
+      "epoch": 0.75728,
+      "grad_norm": 2.3675493349721797,
+      "learning_rate": 0.00017523924051150228,
+      "loss": 1.7843,
+      "step": 4733
+    },
+    {
+      "epoch": 0.75744,
+      "grad_norm": 2.7296349507948894,
+      "learning_rate": 0.00017522786085094852,
+      "loss": 1.7581,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7576,
+      "grad_norm": 2.3907138928679186,
+      "learning_rate": 0.00017521647894570497,
+      "loss": 1.5969,
+      "step": 4735
+    },
+    {
+      "epoch": 0.75776,
+      "grad_norm": 1.994537031256357,
+      "learning_rate": 0.0001752050947961113,
+      "loss": 1.7437,
+      "step": 4736
+    },
+    {
+      "epoch": 0.75792,
+      "grad_norm": 2.8043889316336785,
+      "learning_rate": 0.00017519370840250723,
+      "loss": 1.621,
+      "step": 4737
+    },
+    {
+      "epoch": 0.75808,
+      "grad_norm": 2.6745339325695388,
+      "learning_rate": 0.0001751823197652324,
+      "loss": 1.6844,
+      "step": 4738
+    },
+    {
+      "epoch": 0.75824,
+      "grad_norm": 2.4380477921384087,
+      "learning_rate": 0.00017517092888462673,
+      "loss": 1.7277,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7584,
+      "grad_norm": 2.383256158940255,
+      "learning_rate": 0.0001751595357610301,
+      "loss": 1.6769,
+      "step": 4740
+    },
+    {
+      "epoch": 0.75856,
+      "grad_norm": 2.7296844862943535,
+      "learning_rate": 0.0001751481403947824,
+      "loss": 1.5928,
+      "step": 4741
+    },
+    {
+      "epoch": 0.75872,
+      "grad_norm": 2.488895800913595,
+      "learning_rate": 0.00017513674278622375,
+      "loss": 1.7289,
+      "step": 4742
+    },
+    {
+      "epoch": 0.75888,
+      "grad_norm": 2.460837258473872,
+      "learning_rate": 0.00017512534293569413,
+      "loss": 1.6278,
+      "step": 4743
+    },
+    {
+      "epoch": 0.75904,
+      "grad_norm": 2.528153785256028,
+      "learning_rate": 0.0001751139408435338,
+      "loss": 1.6856,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7592,
+      "grad_norm": 2.144640454857809,
+      "learning_rate": 0.00017510253651008287,
+      "loss": 1.6903,
+      "step": 4745
+    },
+    {
+      "epoch": 0.75936,
+      "grad_norm": 2.340422024400357,
+      "learning_rate": 0.0001750911299356817,
+      "loss": 1.5993,
+      "step": 4746
+    },
+    {
+      "epoch": 0.75952,
+      "grad_norm": 2.3690366294176433,
+      "learning_rate": 0.00017507972112067063,
+      "loss": 1.6433,
+      "step": 4747
+    },
+    {
+      "epoch": 0.75968,
+      "grad_norm": 2.450390391567942,
+      "learning_rate": 0.0001750683100653901,
+      "loss": 1.7565,
+      "step": 4748
+    },
+    {
+      "epoch": 0.75984,
+      "grad_norm": 2.2105446755748743,
+      "learning_rate": 0.0001750568967701806,
+      "loss": 1.6883,
+      "step": 4749
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 2.1956242286007046,
+      "learning_rate": 0.00017504548123538265,
+      "loss": 1.6976,
+      "step": 4750
+    },
+    {
+      "epoch": 0.76016,
+      "grad_norm": 2.564320763637079,
+      "learning_rate": 0.00017503406346133687,
+      "loss": 1.622,
+      "step": 4751
+    },
+    {
+      "epoch": 0.76032,
+      "grad_norm": 2.2902268872831173,
+      "learning_rate": 0.000175022643448384,
+      "loss": 1.7316,
+      "step": 4752
+    },
+    {
+      "epoch": 0.76048,
+      "grad_norm": 2.539634134439368,
+      "learning_rate": 0.00017501122119686476,
+      "loss": 1.7434,
+      "step": 4753
+    },
+    {
+      "epoch": 0.76064,
+      "grad_norm": 2.258374254913028,
+      "learning_rate": 0.00017499979670712,
+      "loss": 1.7109,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7608,
+      "grad_norm": 2.5238633854279264,
+      "learning_rate": 0.00017498836997949057,
+      "loss": 1.6994,
+      "step": 4755
+    },
+    {
+      "epoch": 0.76096,
+      "grad_norm": 2.5107783573826974,
+      "learning_rate": 0.00017497694101431744,
+      "loss": 1.681,
+      "step": 4756
+    },
+    {
+      "epoch": 0.76112,
+      "grad_norm": 2.298928606182818,
+      "learning_rate": 0.00017496550981194168,
+      "loss": 1.72,
+      "step": 4757
+    },
+    {
+      "epoch": 0.76128,
+      "grad_norm": 2.502280959417011,
+      "learning_rate": 0.0001749540763727043,
+      "loss": 1.684,
+      "step": 4758
+    },
+    {
+      "epoch": 0.76144,
+      "grad_norm": 2.8031567706131475,
+      "learning_rate": 0.00017494264069694654,
+      "loss": 1.7094,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7616,
+      "grad_norm": 2.7012901899559982,
+      "learning_rate": 0.0001749312027850096,
+      "loss": 1.6181,
+      "step": 4760
+    },
+    {
+      "epoch": 0.76176,
+      "grad_norm": 2.6809403261070615,
+      "learning_rate": 0.00017491976263723472,
+      "loss": 1.5405,
+      "step": 4761
+    },
+    {
+      "epoch": 0.76192,
+      "grad_norm": 2.7865609960895985,
+      "learning_rate": 0.00017490832025396328,
+      "loss": 1.6312,
+      "step": 4762
+    },
+    {
+      "epoch": 0.76208,
+      "grad_norm": 2.806765466868787,
+      "learning_rate": 0.00017489687563553675,
+      "loss": 1.6436,
+      "step": 4763
+    },
+    {
+      "epoch": 0.76224,
+      "grad_norm": 2.2004943438675295,
+      "learning_rate": 0.00017488542878229656,
+      "loss": 1.7387,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7624,
+      "grad_norm": 2.363680127295453,
+      "learning_rate": 0.00017487397969458428,
+      "loss": 1.6088,
+      "step": 4765
+    },
+    {
+      "epoch": 0.76256,
+      "grad_norm": 3.578762022448568,
+      "learning_rate": 0.0001748625283727416,
+      "loss": 1.7351,
+      "step": 4766
+    },
+    {
+      "epoch": 0.76272,
+      "grad_norm": 4.1800043347520495,
+      "learning_rate": 0.00017485107481711012,
+      "loss": 1.7254,
+      "step": 4767
+    },
+    {
+      "epoch": 0.76288,
+      "grad_norm": 1.9764192174840256,
+      "learning_rate": 0.00017483961902803164,
+      "loss": 1.6664,
+      "step": 4768
+    },
+    {
+      "epoch": 0.76304,
+      "grad_norm": 3.634942454227848,
+      "learning_rate": 0.000174828161005848,
+      "loss": 1.7155,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7632,
+      "grad_norm": 2.620673001086279,
+      "learning_rate": 0.00017481670075090107,
+      "loss": 1.7624,
+      "step": 4770
+    },
+    {
+      "epoch": 0.76336,
+      "grad_norm": 2.7282559886594724,
+      "learning_rate": 0.00017480523826353278,
+      "loss": 1.7094,
+      "step": 4771
+    },
+    {
+      "epoch": 0.76352,
+      "grad_norm": 2.517188645175714,
+      "learning_rate": 0.00017479377354408522,
+      "loss": 1.6282,
+      "step": 4772
+    },
+    {
+      "epoch": 0.76368,
+      "grad_norm": 2.2193627411629264,
+      "learning_rate": 0.00017478230659290044,
+      "loss": 1.7241,
+      "step": 4773
+    },
+    {
+      "epoch": 0.76384,
+      "grad_norm": 3.002729578357935,
+      "learning_rate": 0.00017477083741032057,
+      "loss": 1.6341,
+      "step": 4774
+    },
+    {
+      "epoch": 0.764,
+      "grad_norm": 2.3709933070496727,
+      "learning_rate": 0.0001747593659966879,
+      "loss": 1.7163,
+      "step": 4775
+    },
+    {
+      "epoch": 0.76416,
+      "grad_norm": 3.6453092284610467,
+      "learning_rate": 0.00017474789235234469,
+      "loss": 1.7511,
+      "step": 4776
+    },
+    {
+      "epoch": 0.76432,
+      "grad_norm": 2.524443209959926,
+      "learning_rate": 0.00017473641647763324,
+      "loss": 1.6133,
+      "step": 4777
+    },
+    {
+      "epoch": 0.76448,
+      "grad_norm": 3.502253952245711,
+      "learning_rate": 0.00017472493837289609,
+      "loss": 1.726,
+      "step": 4778
+    },
+    {
+      "epoch": 0.76464,
+      "grad_norm": 2.5449063158643788,
+      "learning_rate": 0.0001747134580384756,
+      "loss": 1.5994,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7648,
+      "grad_norm": 3.7874279528993284,
+      "learning_rate": 0.00017470197547471444,
+      "loss": 1.6966,
+      "step": 4780
+    },
+    {
+      "epoch": 0.76496,
+      "grad_norm": 3.40734299624516,
+      "learning_rate": 0.00017469049068195516,
+      "loss": 1.7142,
+      "step": 4781
+    },
+    {
+      "epoch": 0.76512,
+      "grad_norm": 3.7540060606823005,
+      "learning_rate": 0.00017467900366054047,
+      "loss": 1.705,
+      "step": 4782
+    },
+    {
+      "epoch": 0.76528,
+      "grad_norm": 3.1271562177135297,
+      "learning_rate": 0.0001746675144108131,
+      "loss": 1.6984,
+      "step": 4783
+    },
+    {
+      "epoch": 0.76544,
+      "grad_norm": 2.7017351470521334,
+      "learning_rate": 0.00017465602293311597,
+      "loss": 1.6165,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7656,
+      "grad_norm": 2.5061539506807673,
+      "learning_rate": 0.00017464452922779188,
+      "loss": 1.5946,
+      "step": 4785
+    },
+    {
+      "epoch": 0.76576,
+      "grad_norm": 2.5448693118747485,
+      "learning_rate": 0.0001746330332951838,
+      "loss": 1.7371,
+      "step": 4786
+    },
+    {
+      "epoch": 0.76592,
+      "grad_norm": 2.1441887660183765,
+      "learning_rate": 0.00017462153513563475,
+      "loss": 1.6571,
+      "step": 4787
+    },
+    {
+      "epoch": 0.76608,
+      "grad_norm": 2.3161365572059824,
+      "learning_rate": 0.00017461003474948783,
+      "loss": 1.7402,
+      "step": 4788
+    },
+    {
+      "epoch": 0.76624,
+      "grad_norm": 2.363978934853967,
+      "learning_rate": 0.00017459853213708622,
+      "loss": 1.5351,
+      "step": 4789
+    },
+    {
+      "epoch": 0.7664,
+      "grad_norm": 2.265058041693749,
+      "learning_rate": 0.00017458702729877305,
+      "loss": 1.7546,
+      "step": 4790
+    },
+    {
+      "epoch": 0.76656,
+      "grad_norm": 2.397863073647429,
+      "learning_rate": 0.00017457552023489173,
+      "loss": 1.656,
+      "step": 4791
+    },
+    {
+      "epoch": 0.76672,
+      "grad_norm": 2.3405264172906524,
+      "learning_rate": 0.00017456401094578553,
+      "loss": 1.6974,
+      "step": 4792
+    },
+    {
+      "epoch": 0.76688,
+      "grad_norm": 2.312513156391609,
+      "learning_rate": 0.00017455249943179788,
+      "loss": 1.6918,
+      "step": 4793
+    },
+    {
+      "epoch": 0.76704,
+      "grad_norm": 2.361101176261733,
+      "learning_rate": 0.0001745409856932723,
+      "loss": 1.7569,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7672,
+      "grad_norm": 2.7645160839235534,
+      "learning_rate": 0.00017452946973055232,
+      "loss": 1.6332,
+      "step": 4795
+    },
+    {
+      "epoch": 0.76736,
+      "grad_norm": 2.3765486017224866,
+      "learning_rate": 0.00017451795154398156,
+      "loss": 1.7029,
+      "step": 4796
+    },
+    {
+      "epoch": 0.76752,
+      "grad_norm": 2.396562504570784,
+      "learning_rate": 0.0001745064311339037,
+      "loss": 1.6503,
+      "step": 4797
+    },
+    {
+      "epoch": 0.76768,
+      "grad_norm": 2.224998881980454,
+      "learning_rate": 0.00017449490850066256,
+      "loss": 1.6664,
+      "step": 4798
+    },
+    {
+      "epoch": 0.76784,
+      "grad_norm": 2.6859857196203336,
+      "learning_rate": 0.0001744833836446018,
+      "loss": 1.6729,
+      "step": 4799
+    },
+    {
+      "epoch": 0.768,
+      "grad_norm": 2.2358996557619175,
+      "learning_rate": 0.00017447185656606552,
+      "loss": 1.734,
+      "step": 4800
+    },
+    {
+      "epoch": 0.76816,
+      "grad_norm": 2.492462675762857,
+      "learning_rate": 0.00017446032726539746,
+      "loss": 1.6902,
+      "step": 4801
+    },
+    {
+      "epoch": 0.76832,
+      "grad_norm": 2.2770901113834845,
+      "learning_rate": 0.0001744487957429418,
+      "loss": 1.6887,
+      "step": 4802
+    },
+    {
+      "epoch": 0.76848,
+      "grad_norm": 2.259108169154462,
+      "learning_rate": 0.00017443726199904252,
+      "loss": 1.6884,
+      "step": 4803
+    },
+    {
+      "epoch": 0.76864,
+      "grad_norm": 1.960765553884255,
+      "learning_rate": 0.00017442572603404384,
+      "loss": 1.6354,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7688,
+      "grad_norm": 2.9099744981115747,
+      "learning_rate": 0.00017441418784828994,
+      "loss": 1.7311,
+      "step": 4805
+    },
+    {
+      "epoch": 0.76896,
+      "grad_norm": 2.3720901476478136,
+      "learning_rate": 0.00017440264744212507,
+      "loss": 1.6456,
+      "step": 4806
+    },
+    {
+      "epoch": 0.76912,
+      "grad_norm": 2.208809208752232,
+      "learning_rate": 0.00017439110481589365,
+      "loss": 1.5097,
+      "step": 4807
+    },
+    {
+      "epoch": 0.76928,
+      "grad_norm": 2.5292205294226195,
+      "learning_rate": 0.00017437955996994005,
+      "loss": 1.6743,
+      "step": 4808
+    },
+    {
+      "epoch": 0.76944,
+      "grad_norm": 2.7966296321486563,
+      "learning_rate": 0.00017436801290460875,
+      "loss": 1.6268,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7696,
+      "grad_norm": 2.2711316263042827,
+      "learning_rate": 0.00017435646362024435,
+      "loss": 1.6923,
+      "step": 4810
+    },
+    {
+      "epoch": 0.76976,
+      "grad_norm": 2.359621594682316,
+      "learning_rate": 0.0001743449121171914,
+      "loss": 1.6483,
+      "step": 4811
+    },
+    {
+      "epoch": 0.76992,
+      "grad_norm": 2.177714370025709,
+      "learning_rate": 0.0001743333583957946,
+      "loss": 1.6889,
+      "step": 4812
+    },
+    {
+      "epoch": 0.77008,
+      "grad_norm": 2.405584348144695,
+      "learning_rate": 0.0001743218024563987,
+      "loss": 1.6341,
+      "step": 4813
+    },
+    {
+      "epoch": 0.77024,
+      "grad_norm": 2.2580553200810987,
+      "learning_rate": 0.00017431024429934852,
+      "loss": 1.7492,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7704,
+      "grad_norm": 2.4087553841536535,
+      "learning_rate": 0.00017429868392498894,
+      "loss": 1.6017,
+      "step": 4815
+    },
+    {
+      "epoch": 0.77056,
+      "grad_norm": 2.355667432332493,
+      "learning_rate": 0.0001742871213336649,
+      "loss": 1.7194,
+      "step": 4816
+    },
+    {
+      "epoch": 0.77072,
+      "grad_norm": 2.261971510501707,
+      "learning_rate": 0.0001742755565257214,
+      "loss": 1.6774,
+      "step": 4817
+    },
+    {
+      "epoch": 0.77088,
+      "grad_norm": 2.1866156245486246,
+      "learning_rate": 0.00017426398950150352,
+      "loss": 1.7588,
+      "step": 4818
+    },
+    {
+      "epoch": 0.77104,
+      "grad_norm": 2.5278177171315783,
+      "learning_rate": 0.00017425242026135638,
+      "loss": 1.6597,
+      "step": 4819
+    },
+    {
+      "epoch": 0.7712,
+      "grad_norm": 2.13957188101691,
+      "learning_rate": 0.00017424084880562525,
+      "loss": 1.6255,
+      "step": 4820
+    },
+    {
+      "epoch": 0.77136,
+      "grad_norm": 2.3232046661482846,
+      "learning_rate": 0.00017422927513465538,
+      "loss": 1.6452,
+      "step": 4821
+    },
+    {
+      "epoch": 0.77152,
+      "grad_norm": 2.652459766685712,
+      "learning_rate": 0.00017421769924879206,
+      "loss": 1.7164,
+      "step": 4822
+    },
+    {
+      "epoch": 0.77168,
+      "grad_norm": 2.30980214651283,
+      "learning_rate": 0.0001742061211483808,
+      "loss": 1.6146,
+      "step": 4823
+    },
+    {
+      "epoch": 0.77184,
+      "grad_norm": 2.2174161949548603,
+      "learning_rate": 0.00017419454083376698,
+      "loss": 1.6236,
+      "step": 4824
+    },
+    {
+      "epoch": 0.772,
+      "grad_norm": 1.9980431003851862,
+      "learning_rate": 0.00017418295830529617,
+      "loss": 1.6186,
+      "step": 4825
+    },
+    {
+      "epoch": 0.77216,
+      "grad_norm": 2.6891225437833617,
+      "learning_rate": 0.000174171373563314,
+      "loss": 1.7426,
+      "step": 4826
+    },
+    {
+      "epoch": 0.77232,
+      "grad_norm": 2.6134172589589046,
+      "learning_rate": 0.00017415978660816607,
+      "loss": 1.6816,
+      "step": 4827
+    },
+    {
+      "epoch": 0.77248,
+      "grad_norm": 2.514262274609114,
+      "learning_rate": 0.00017414819744019822,
+      "loss": 1.6231,
+      "step": 4828
+    },
+    {
+      "epoch": 0.77264,
+      "grad_norm": 2.389779352118256,
+      "learning_rate": 0.0001741366060597562,
+      "loss": 1.6737,
+      "step": 4829
+    },
+    {
+      "epoch": 0.7728,
+      "grad_norm": 2.8544590613280403,
+      "learning_rate": 0.00017412501246718588,
+      "loss": 1.7121,
+      "step": 4830
+    },
+    {
+      "epoch": 0.77296,
+      "grad_norm": 2.5901078129163184,
+      "learning_rate": 0.0001741134166628332,
+      "loss": 1.6525,
+      "step": 4831
+    },
+    {
+      "epoch": 0.77312,
+      "grad_norm": 2.4850754885355566,
+      "learning_rate": 0.00017410181864704416,
+      "loss": 1.7405,
+      "step": 4832
+    },
+    {
+      "epoch": 0.77328,
+      "grad_norm": 2.732671554287679,
+      "learning_rate": 0.00017409021842016483,
+      "loss": 1.6953,
+      "step": 4833
+    },
+    {
+      "epoch": 0.77344,
+      "grad_norm": 2.3229508232292058,
+      "learning_rate": 0.00017407861598254135,
+      "loss": 1.6048,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7736,
+      "grad_norm": 2.1582815194011,
+      "learning_rate": 0.00017406701133451988,
+      "loss": 1.5779,
+      "step": 4835
+    },
+    {
+      "epoch": 0.77376,
+      "grad_norm": 2.10695966422521,
+      "learning_rate": 0.00017405540447644677,
+      "loss": 1.6237,
+      "step": 4836
+    },
+    {
+      "epoch": 0.77392,
+      "grad_norm": 2.4480722172315854,
+      "learning_rate": 0.00017404379540866827,
+      "loss": 1.7744,
+      "step": 4837
+    },
+    {
+      "epoch": 0.77408,
+      "grad_norm": 2.939112814729998,
+      "learning_rate": 0.00017403218413153079,
+      "loss": 1.8239,
+      "step": 4838
+    },
+    {
+      "epoch": 0.77424,
+      "grad_norm": 2.437824849350452,
+      "learning_rate": 0.00017402057064538085,
+      "loss": 1.6055,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7744,
+      "grad_norm": 2.296321249008966,
+      "learning_rate": 0.00017400895495056493,
+      "loss": 1.6527,
+      "step": 4840
+    },
+    {
+      "epoch": 0.77456,
+      "grad_norm": 3.0682164235312275,
+      "learning_rate": 0.00017399733704742963,
+      "loss": 1.6324,
+      "step": 4841
+    },
+    {
+      "epoch": 0.77472,
+      "grad_norm": 2.173305358861671,
+      "learning_rate": 0.00017398571693632162,
+      "loss": 1.6647,
+      "step": 4842
+    },
+    {
+      "epoch": 0.77488,
+      "grad_norm": 3.3594117829165135,
+      "learning_rate": 0.00017397409461758763,
+      "loss": 1.7019,
+      "step": 4843
+    },
+    {
+      "epoch": 0.77504,
+      "grad_norm": 2.3315433281378946,
+      "learning_rate": 0.00017396247009157442,
+      "loss": 1.6882,
+      "step": 4844
+    },
+    {
+      "epoch": 0.7752,
+      "grad_norm": 3.0360723091566593,
+      "learning_rate": 0.0001739508433586289,
+      "loss": 1.7127,
+      "step": 4845
+    },
+    {
+      "epoch": 0.77536,
+      "grad_norm": 2.4953809013131827,
+      "learning_rate": 0.00017393921441909797,
+      "loss": 1.7371,
+      "step": 4846
+    },
+    {
+      "epoch": 0.77552,
+      "grad_norm": 2.256789833661684,
+      "learning_rate": 0.00017392758327332863,
+      "loss": 1.5409,
+      "step": 4847
+    },
+    {
+      "epoch": 0.77568,
+      "grad_norm": 2.870771218652211,
+      "learning_rate": 0.00017391594992166793,
+      "loss": 1.6664,
+      "step": 4848
+    },
+    {
+      "epoch": 0.77584,
+      "grad_norm": 2.313620216581122,
+      "learning_rate": 0.000173904314364463,
+      "loss": 1.5861,
+      "step": 4849
+    },
+    {
+      "epoch": 0.776,
+      "grad_norm": 3.073238479622931,
+      "learning_rate": 0.00017389267660206098,
+      "loss": 1.6998,
+      "step": 4850
+    },
+    {
+      "epoch": 0.77616,
+      "grad_norm": 2.419064689223655,
+      "learning_rate": 0.0001738810366348092,
+      "loss": 1.6388,
+      "step": 4851
+    },
+    {
+      "epoch": 0.77632,
+      "grad_norm": 2.371943268373913,
+      "learning_rate": 0.00017386939446305494,
+      "loss": 1.7538,
+      "step": 4852
+    },
+    {
+      "epoch": 0.77648,
+      "grad_norm": 2.740322670275376,
+      "learning_rate": 0.00017385775008714558,
+      "loss": 1.6347,
+      "step": 4853
+    },
+    {
+      "epoch": 0.77664,
+      "grad_norm": 2.418496517566747,
+      "learning_rate": 0.00017384610350742856,
+      "loss": 1.7762,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7768,
+      "grad_norm": 2.0398899720542354,
+      "learning_rate": 0.00017383445472425143,
+      "loss": 1.7163,
+      "step": 4855
+    },
+    {
+      "epoch": 0.77696,
+      "grad_norm": 2.587382958748631,
+      "learning_rate": 0.00017382280373796175,
+      "loss": 1.6528,
+      "step": 4856
+    },
+    {
+      "epoch": 0.77712,
+      "grad_norm": 2.1208269861114046,
+      "learning_rate": 0.0001738111505489072,
+      "loss": 1.6225,
+      "step": 4857
+    },
+    {
+      "epoch": 0.77728,
+      "grad_norm": 3.5218790684258594,
+      "learning_rate": 0.0001737994951574355,
+      "loss": 1.664,
+      "step": 4858
+    },
+    {
+      "epoch": 0.77744,
+      "grad_norm": 1.9305157206330998,
+      "learning_rate": 0.00017378783756389433,
+      "loss": 1.6395,
+      "step": 4859
+    },
+    {
+      "epoch": 0.7776,
+      "grad_norm": 2.9717379944518956,
+      "learning_rate": 0.00017377617776863165,
+      "loss": 1.7305,
+      "step": 4860
+    },
+    {
+      "epoch": 0.77776,
+      "grad_norm": 2.1471693807243915,
+      "learning_rate": 0.00017376451577199532,
+      "loss": 1.6912,
+      "step": 4861
+    },
+    {
+      "epoch": 0.77792,
+      "grad_norm": 2.6345828759687966,
+      "learning_rate": 0.00017375285157433332,
+      "loss": 1.7004,
+      "step": 4862
+    },
+    {
+      "epoch": 0.77808,
+      "grad_norm": 2.6118683564241123,
+      "learning_rate": 0.00017374118517599372,
+      "loss": 1.8601,
+      "step": 4863
+    },
+    {
+      "epoch": 0.77824,
+      "grad_norm": 2.064764616661512,
+      "learning_rate": 0.00017372951657732458,
+      "loss": 1.6517,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7784,
+      "grad_norm": 2.9261884081642164,
+      "learning_rate": 0.0001737178457786741,
+      "loss": 1.6773,
+      "step": 4865
+    },
+    {
+      "epoch": 0.77856,
+      "grad_norm": 2.182094678284376,
+      "learning_rate": 0.00017370617278039053,
+      "loss": 1.7686,
+      "step": 4866
+    },
+    {
+      "epoch": 0.77872,
+      "grad_norm": 2.377294327606811,
+      "learning_rate": 0.0001736944975828222,
+      "loss": 1.6811,
+      "step": 4867
+    },
+    {
+      "epoch": 0.77888,
+      "grad_norm": 2.2976891441404463,
+      "learning_rate": 0.00017368282018631742,
+      "loss": 1.6036,
+      "step": 4868
+    },
+    {
+      "epoch": 0.77904,
+      "grad_norm": 2.6675993096796367,
+      "learning_rate": 0.00017367114059122461,
+      "loss": 1.594,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7792,
+      "grad_norm": 2.1965838543654073,
+      "learning_rate": 0.0001736594587978924,
+      "loss": 1.5626,
+      "step": 4870
+    },
+    {
+      "epoch": 0.77936,
+      "grad_norm": 2.3698519582071444,
+      "learning_rate": 0.0001736477748066692,
+      "loss": 1.6931,
+      "step": 4871
+    },
+    {
+      "epoch": 0.77952,
+      "grad_norm": 3.130050679142918,
+      "learning_rate": 0.00017363608861790372,
+      "loss": 1.7123,
+      "step": 4872
+    },
+    {
+      "epoch": 0.77968,
+      "grad_norm": 2.3768036184784047,
+      "learning_rate": 0.00017362440023194472,
+      "loss": 1.746,
+      "step": 4873
+    },
+    {
+      "epoch": 0.77984,
+      "grad_norm": 2.0838180917967177,
+      "learning_rate": 0.00017361270964914085,
+      "loss": 1.6377,
+      "step": 4874
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 2.537801441646549,
+      "learning_rate": 0.00017360101686984098,
+      "loss": 1.7152,
+      "step": 4875
+    },
+    {
+      "epoch": 0.78016,
+      "grad_norm": 2.674401129662021,
+      "learning_rate": 0.00017358932189439404,
+      "loss": 1.6945,
+      "step": 4876
+    },
+    {
+      "epoch": 0.78032,
+      "grad_norm": 2.1249073246719505,
+      "learning_rate": 0.00017357762472314893,
+      "loss": 1.7219,
+      "step": 4877
+    },
+    {
+      "epoch": 0.78048,
+      "grad_norm": 2.1238838673300995,
+      "learning_rate": 0.00017356592535645475,
+      "loss": 1.5795,
+      "step": 4878
+    },
+    {
+      "epoch": 0.78064,
+      "grad_norm": 2.269716676732443,
+      "learning_rate": 0.00017355422379466054,
+      "loss": 1.7553,
+      "step": 4879
+    },
+    {
+      "epoch": 0.7808,
+      "grad_norm": 2.126401225659228,
+      "learning_rate": 0.00017354252003811546,
+      "loss": 1.6484,
+      "step": 4880
+    },
+    {
+      "epoch": 0.78096,
+      "grad_norm": 2.1248531023845345,
+      "learning_rate": 0.00017353081408716877,
+      "loss": 1.643,
+      "step": 4881
+    },
+    {
+      "epoch": 0.78112,
+      "grad_norm": 2.930557655283955,
+      "learning_rate": 0.0001735191059421697,
+      "loss": 1.5738,
+      "step": 4882
+    },
+    {
+      "epoch": 0.78128,
+      "grad_norm": 2.354420958340543,
+      "learning_rate": 0.00017350739560346767,
+      "loss": 1.7116,
+      "step": 4883
+    },
+    {
+      "epoch": 0.78144,
+      "grad_norm": 2.3867642274351226,
+      "learning_rate": 0.00017349568307141206,
+      "loss": 1.8401,
+      "step": 4884
+    },
+    {
+      "epoch": 0.7816,
+      "grad_norm": 2.0819070236722754,
+      "learning_rate": 0.00017348396834635234,
+      "loss": 1.6177,
+      "step": 4885
+    },
+    {
+      "epoch": 0.78176,
+      "grad_norm": 2.0990981846303756,
+      "learning_rate": 0.00017347225142863812,
+      "loss": 1.6246,
+      "step": 4886
+    },
+    {
+      "epoch": 0.78192,
+      "grad_norm": 2.1915369821952155,
+      "learning_rate": 0.00017346053231861896,
+      "loss": 1.6694,
+      "step": 4887
+    },
+    {
+      "epoch": 0.78208,
+      "grad_norm": 2.1328536450796824,
+      "learning_rate": 0.00017344881101664453,
+      "loss": 1.6754,
+      "step": 4888
+    },
+    {
+      "epoch": 0.78224,
+      "grad_norm": 2.3014126487063513,
+      "learning_rate": 0.00017343708752306464,
+      "loss": 1.714,
+      "step": 4889
+    },
+    {
+      "epoch": 0.7824,
+      "grad_norm": 2.2605784178968076,
+      "learning_rate": 0.00017342536183822906,
+      "loss": 1.606,
+      "step": 4890
+    },
+    {
+      "epoch": 0.78256,
+      "grad_norm": 2.8700933874836645,
+      "learning_rate": 0.00017341363396248768,
+      "loss": 1.632,
+      "step": 4891
+    },
+    {
+      "epoch": 0.78272,
+      "grad_norm": 2.4337096852626887,
+      "learning_rate": 0.00017340190389619043,
+      "loss": 1.5859,
+      "step": 4892
+    },
+    {
+      "epoch": 0.78288,
+      "grad_norm": 2.5522429172549197,
+      "learning_rate": 0.00017339017163968732,
+      "loss": 1.7176,
+      "step": 4893
+    },
+    {
+      "epoch": 0.78304,
+      "grad_norm": 2.183681818578341,
+      "learning_rate": 0.00017337843719332843,
+      "loss": 1.5946,
+      "step": 4894
+    },
+    {
+      "epoch": 0.7832,
+      "grad_norm": 1.9888042185093049,
+      "learning_rate": 0.00017336670055746392,
+      "loss": 1.6342,
+      "step": 4895
+    },
+    {
+      "epoch": 0.78336,
+      "grad_norm": 2.538862034398798,
+      "learning_rate": 0.00017335496173244392,
+      "loss": 1.5324,
+      "step": 4896
+    },
+    {
+      "epoch": 0.78352,
+      "grad_norm": 2.3079729131032796,
+      "learning_rate": 0.0001733432207186188,
+      "loss": 1.7149,
+      "step": 4897
+    },
+    {
+      "epoch": 0.78368,
+      "grad_norm": 2.2310077179214334,
+      "learning_rate": 0.00017333147751633882,
+      "loss": 1.5781,
+      "step": 4898
+    },
+    {
+      "epoch": 0.78384,
+      "grad_norm": 2.1854523691132886,
+      "learning_rate": 0.00017331973212595444,
+      "loss": 1.5855,
+      "step": 4899
+    },
+    {
+      "epoch": 0.784,
+      "grad_norm": 2.432985145204953,
+      "learning_rate": 0.00017330798454781604,
+      "loss": 1.6696,
+      "step": 4900
+    },
+    {
+      "epoch": 0.78416,
+      "grad_norm": 2.091237660807033,
+      "learning_rate": 0.00017329623478227425,
+      "loss": 1.6163,
+      "step": 4901
+    },
+    {
+      "epoch": 0.78432,
+      "grad_norm": 2.492082284883915,
+      "learning_rate": 0.00017328448282967962,
+      "loss": 1.6783,
+      "step": 4902
+    },
+    {
+      "epoch": 0.78448,
+      "grad_norm": 2.3087893887566215,
+      "learning_rate": 0.00017327272869038278,
+      "loss": 1.719,
+      "step": 4903
+    },
+    {
+      "epoch": 0.78464,
+      "grad_norm": 2.1538715356392273,
+      "learning_rate": 0.0001732609723647345,
+      "loss": 1.6234,
+      "step": 4904
+    },
+    {
+      "epoch": 0.7848,
+      "grad_norm": 1.8534253754431484,
+      "learning_rate": 0.00017324921385308554,
+      "loss": 1.6103,
+      "step": 4905
+    },
+    {
+      "epoch": 0.78496,
+      "grad_norm": 2.4541555778618194,
+      "learning_rate": 0.0001732374531557868,
+      "loss": 1.7821,
+      "step": 4906
+    },
+    {
+      "epoch": 0.78512,
+      "grad_norm": 2.0775210234034684,
+      "learning_rate": 0.00017322569027318916,
+      "loss": 1.6304,
+      "step": 4907
+    },
+    {
+      "epoch": 0.78528,
+      "grad_norm": 2.250944910931081,
+      "learning_rate": 0.00017321392520564362,
+      "loss": 1.664,
+      "step": 4908
+    },
+    {
+      "epoch": 0.78544,
+      "grad_norm": 2.392084040545199,
+      "learning_rate": 0.00017320215795350126,
+      "loss": 1.7325,
+      "step": 4909
+    },
+    {
+      "epoch": 0.7856,
+      "grad_norm": 2.2740752168056586,
+      "learning_rate": 0.00017319038851711313,
+      "loss": 1.6118,
+      "step": 4910
+    },
+    {
+      "epoch": 0.78576,
+      "grad_norm": 2.270716375544516,
+      "learning_rate": 0.00017317861689683048,
+      "loss": 1.6922,
+      "step": 4911
+    },
+    {
+      "epoch": 0.78592,
+      "grad_norm": 2.016605865156996,
+      "learning_rate": 0.00017316684309300455,
+      "loss": 1.6476,
+      "step": 4912
+    },
+    {
+      "epoch": 0.78608,
+      "grad_norm": 2.511223735108175,
+      "learning_rate": 0.0001731550671059866,
+      "loss": 1.6469,
+      "step": 4913
+    },
+    {
+      "epoch": 0.78624,
+      "grad_norm": 1.9841106009238274,
+      "learning_rate": 0.00017314328893612807,
+      "loss": 1.6178,
+      "step": 4914
+    },
+    {
+      "epoch": 0.7864,
+      "grad_norm": 2.1183531026781535,
+      "learning_rate": 0.0001731315085837804,
+      "loss": 1.6086,
+      "step": 4915
+    },
+    {
+      "epoch": 0.78656,
+      "grad_norm": 2.0315273028572274,
+      "learning_rate": 0.00017311972604929502,
+      "loss": 1.6555,
+      "step": 4916
+    },
+    {
+      "epoch": 0.78672,
+      "grad_norm": 2.498100351968872,
+      "learning_rate": 0.0001731079413330236,
+      "loss": 1.6374,
+      "step": 4917
+    },
+    {
+      "epoch": 0.78688,
+      "grad_norm": 2.440125763536732,
+      "learning_rate": 0.00017309615443531775,
+      "loss": 1.6306,
+      "step": 4918
+    },
+    {
+      "epoch": 0.78704,
+      "grad_norm": 2.559731639905946,
+      "learning_rate": 0.00017308436535652915,
+      "loss": 1.6249,
+      "step": 4919
+    },
+    {
+      "epoch": 0.7872,
+      "grad_norm": 2.122949004384937,
+      "learning_rate": 0.0001730725740970096,
+      "loss": 1.6395,
+      "step": 4920
+    },
+    {
+      "epoch": 0.78736,
+      "grad_norm": 2.1843596853703637,
+      "learning_rate": 0.0001730607806571109,
+      "loss": 1.6823,
+      "step": 4921
+    },
+    {
+      "epoch": 0.78752,
+      "grad_norm": 1.9618187829405263,
+      "learning_rate": 0.00017304898503718493,
+      "loss": 1.5204,
+      "step": 4922
+    },
+    {
+      "epoch": 0.78768,
+      "grad_norm": 2.33613276870272,
+      "learning_rate": 0.00017303718723758374,
+      "loss": 1.6826,
+      "step": 4923
+    },
+    {
+      "epoch": 0.78784,
+      "grad_norm": 2.393953505879918,
+      "learning_rate": 0.00017302538725865932,
+      "loss": 1.619,
+      "step": 4924
+    },
+    {
+      "epoch": 0.788,
+      "grad_norm": 2.1799767077190366,
+      "learning_rate": 0.0001730135851007637,
+      "loss": 1.6125,
+      "step": 4925
+    },
+    {
+      "epoch": 0.78816,
+      "grad_norm": 2.262952625552394,
+      "learning_rate": 0.00017300178076424914,
+      "loss": 1.6676,
+      "step": 4926
+    },
+    {
+      "epoch": 0.78832,
+      "grad_norm": 1.9975217124418565,
+      "learning_rate": 0.0001729899742494678,
+      "loss": 1.6197,
+      "step": 4927
+    },
+    {
+      "epoch": 0.78848,
+      "grad_norm": 2.4124177278959813,
+      "learning_rate": 0.00017297816555677197,
+      "loss": 1.6192,
+      "step": 4928
+    },
+    {
+      "epoch": 0.78864,
+      "grad_norm": 2.3703281787598827,
+      "learning_rate": 0.00017296635468651403,
+      "loss": 1.736,
+      "step": 4929
+    },
+    {
+      "epoch": 0.7888,
+      "grad_norm": 2.570368144070199,
+      "learning_rate": 0.0001729545416390464,
+      "loss": 1.6446,
+      "step": 4930
+    },
+    {
+      "epoch": 0.78896,
+      "grad_norm": 2.1082975722192363,
+      "learning_rate": 0.00017294272641472155,
+      "loss": 1.7086,
+      "step": 4931
+    },
+    {
+      "epoch": 0.78912,
+      "grad_norm": 2.0519848958074656,
+      "learning_rate": 0.00017293090901389206,
+      "loss": 1.65,
+      "step": 4932
+    },
+    {
+      "epoch": 0.78928,
+      "grad_norm": 2.362033050460854,
+      "learning_rate": 0.00017291908943691046,
+      "loss": 1.6603,
+      "step": 4933
+    },
+    {
+      "epoch": 0.78944,
+      "grad_norm": 2.208921083064596,
+      "learning_rate": 0.00017290726768412953,
+      "loss": 1.6502,
+      "step": 4934
+    },
+    {
+      "epoch": 0.7896,
+      "grad_norm": 2.1288137392572666,
+      "learning_rate": 0.00017289544375590196,
+      "loss": 1.5155,
+      "step": 4935
+    },
+    {
+      "epoch": 0.78976,
+      "grad_norm": 2.1935961243115396,
+      "learning_rate": 0.00017288361765258053,
+      "loss": 1.7354,
+      "step": 4936
+    },
+    {
+      "epoch": 0.78992,
+      "grad_norm": 2.5949675259358504,
+      "learning_rate": 0.0001728717893745182,
+      "loss": 1.5927,
+      "step": 4937
+    },
+    {
+      "epoch": 0.79008,
+      "grad_norm": 2.1389256693377576,
+      "learning_rate": 0.00017285995892206784,
+      "loss": 1.7208,
+      "step": 4938
+    },
+    {
+      "epoch": 0.79024,
+      "grad_norm": 2.089985904979491,
+      "learning_rate": 0.00017284812629558247,
+      "loss": 1.6704,
+      "step": 4939
+    },
+    {
+      "epoch": 0.7904,
+      "grad_norm": 2.3144601748103604,
+      "learning_rate": 0.00017283629149541518,
+      "loss": 1.7291,
+      "step": 4940
+    },
+    {
+      "epoch": 0.79056,
+      "grad_norm": 2.1974670583727605,
+      "learning_rate": 0.00017282445452191907,
+      "loss": 1.5197,
+      "step": 4941
+    },
+    {
+      "epoch": 0.79072,
+      "grad_norm": 2.4747985027429156,
+      "learning_rate": 0.00017281261537544735,
+      "loss": 1.6288,
+      "step": 4942
+    },
+    {
+      "epoch": 0.79088,
+      "grad_norm": 2.353723181653014,
+      "learning_rate": 0.0001728007740563533,
+      "loss": 1.6546,
+      "step": 4943
+    },
+    {
+      "epoch": 0.79104,
+      "grad_norm": 2.2276333661313594,
+      "learning_rate": 0.00017278893056499022,
+      "loss": 1.5016,
+      "step": 4944
+    },
+    {
+      "epoch": 0.7912,
+      "grad_norm": 1.7824934347759496,
+      "learning_rate": 0.00017277708490171155,
+      "loss": 1.5995,
+      "step": 4945
+    },
+    {
+      "epoch": 0.79136,
+      "grad_norm": 2.1100213109365544,
+      "learning_rate": 0.00017276523706687065,
+      "loss": 1.6149,
+      "step": 4946
+    },
+    {
+      "epoch": 0.79152,
+      "grad_norm": 2.415097186946197,
+      "learning_rate": 0.00017275338706082116,
+      "loss": 1.764,
+      "step": 4947
+    },
+    {
+      "epoch": 0.79168,
+      "grad_norm": 2.4808301086694606,
+      "learning_rate": 0.0001727415348839166,
+      "loss": 1.8647,
+      "step": 4948
+    },
+    {
+      "epoch": 0.79184,
+      "grad_norm": 2.2321657657719647,
+      "learning_rate": 0.00017272968053651063,
+      "loss": 1.6494,
+      "step": 4949
+    },
+    {
+      "epoch": 0.792,
+      "grad_norm": 2.5951214856539018,
+      "learning_rate": 0.00017271782401895698,
+      "loss": 1.6891,
+      "step": 4950
+    },
+    {
+      "epoch": 0.79216,
+      "grad_norm": 2.339456726278217,
+      "learning_rate": 0.00017270596533160943,
+      "loss": 1.702,
+      "step": 4951
+    },
+    {
+      "epoch": 0.79232,
+      "grad_norm": 2.574749882043691,
+      "learning_rate": 0.0001726941044748218,
+      "loss": 1.626,
+      "step": 4952
+    },
+    {
+      "epoch": 0.79248,
+      "grad_norm": 2.5091330732812436,
+      "learning_rate": 0.00017268224144894803,
+      "loss": 1.6734,
+      "step": 4953
+    },
+    {
+      "epoch": 0.79264,
+      "grad_norm": 2.8634431206010245,
+      "learning_rate": 0.00017267037625434207,
+      "loss": 1.6882,
+      "step": 4954
+    },
+    {
+      "epoch": 0.7928,
+      "grad_norm": 2.3132804505475315,
+      "learning_rate": 0.00017265850889135797,
+      "loss": 1.648,
+      "step": 4955
+    },
+    {
+      "epoch": 0.79296,
+      "grad_norm": 2.285452109590507,
+      "learning_rate": 0.00017264663936034986,
+      "loss": 1.6686,
+      "step": 4956
+    },
+    {
+      "epoch": 0.79312,
+      "grad_norm": 2.235219447608813,
+      "learning_rate": 0.00017263476766167187,
+      "loss": 1.6493,
+      "step": 4957
+    },
+    {
+      "epoch": 0.79328,
+      "grad_norm": 2.1662598678523706,
+      "learning_rate": 0.0001726228937956783,
+      "loss": 1.7488,
+      "step": 4958
+    },
+    {
+      "epoch": 0.79344,
+      "grad_norm": 2.1938682815490473,
+      "learning_rate": 0.00017261101776272332,
+      "loss": 1.6375,
+      "step": 4959
+    },
+    {
+      "epoch": 0.7936,
+      "grad_norm": 2.2939431794507525,
+      "learning_rate": 0.00017259913956316143,
+      "loss": 1.6495,
+      "step": 4960
+    },
+    {
+      "epoch": 0.79376,
+      "grad_norm": 1.9950098293306175,
+      "learning_rate": 0.00017258725919734702,
+      "loss": 1.5928,
+      "step": 4961
+    },
+    {
+      "epoch": 0.79392,
+      "grad_norm": 2.4619928926541683,
+      "learning_rate": 0.00017257537666563454,
+      "loss": 1.6493,
+      "step": 4962
+    },
+    {
+      "epoch": 0.79408,
+      "grad_norm": 2.3864643815375084,
+      "learning_rate": 0.00017256349196837856,
+      "loss": 1.6664,
+      "step": 4963
+    },
+    {
+      "epoch": 0.79424,
+      "grad_norm": 2.3604370606455607,
+      "learning_rate": 0.00017255160510593374,
+      "loss": 1.7348,
+      "step": 4964
+    },
+    {
+      "epoch": 0.7944,
+      "grad_norm": 2.191790218202995,
+      "learning_rate": 0.00017253971607865474,
+      "loss": 1.6099,
+      "step": 4965
+    },
+    {
+      "epoch": 0.79456,
+      "grad_norm": 2.1764929224866254,
+      "learning_rate": 0.0001725278248868963,
+      "loss": 1.5888,
+      "step": 4966
+    },
+    {
+      "epoch": 0.79472,
+      "grad_norm": 2.3376538777794713,
+      "learning_rate": 0.00017251593153101325,
+      "loss": 1.602,
+      "step": 4967
+    },
+    {
+      "epoch": 0.79488,
+      "grad_norm": 2.237089840749634,
+      "learning_rate": 0.00017250403601136048,
+      "loss": 1.5611,
+      "step": 4968
+    },
+    {
+      "epoch": 0.79504,
+      "grad_norm": 2.068145633632154,
+      "learning_rate": 0.0001724921383282929,
+      "loss": 1.547,
+      "step": 4969
+    },
+    {
+      "epoch": 0.7952,
+      "grad_norm": 2.495834367039555,
+      "learning_rate": 0.0001724802384821656,
+      "loss": 1.4696,
+      "step": 4970
+    },
+    {
+      "epoch": 0.79536,
+      "grad_norm": 2.2006461036035154,
+      "learning_rate": 0.00017246833647333355,
+      "loss": 1.6453,
+      "step": 4971
+    },
+    {
+      "epoch": 0.79552,
+      "grad_norm": 2.235213901334051,
+      "learning_rate": 0.00017245643230215192,
+      "loss": 1.6777,
+      "step": 4972
+    },
+    {
+      "epoch": 0.79568,
+      "grad_norm": 2.368297563186785,
+      "learning_rate": 0.00017244452596897598,
+      "loss": 1.6453,
+      "step": 4973
+    },
+    {
+      "epoch": 0.79584,
+      "grad_norm": 2.219385549867405,
+      "learning_rate": 0.00017243261747416092,
+      "loss": 1.5787,
+      "step": 4974
+    },
+    {
+      "epoch": 0.796,
+      "grad_norm": 2.013190058602802,
+      "learning_rate": 0.00017242070681806212,
+      "loss": 1.6031,
+      "step": 4975
+    },
+    {
+      "epoch": 0.79616,
+      "grad_norm": 2.5306209925145646,
+      "learning_rate": 0.00017240879400103496,
+      "loss": 1.6529,
+      "step": 4976
+    },
+    {
+      "epoch": 0.79632,
+      "grad_norm": 2.506382220353849,
+      "learning_rate": 0.00017239687902343489,
+      "loss": 1.6782,
+      "step": 4977
+    },
+    {
+      "epoch": 0.79648,
+      "grad_norm": 1.957111543713513,
+      "learning_rate": 0.00017238496188561744,
+      "loss": 1.6156,
+      "step": 4978
+    },
+    {
+      "epoch": 0.79664,
+      "grad_norm": 2.443080675265804,
+      "learning_rate": 0.0001723730425879382,
+      "loss": 1.7132,
+      "step": 4979
+    },
+    {
+      "epoch": 0.7968,
+      "grad_norm": 2.197551225677818,
+      "learning_rate": 0.0001723611211307528,
+      "loss": 1.6591,
+      "step": 4980
+    },
+    {
+      "epoch": 0.79696,
+      "grad_norm": 2.2759695896814067,
+      "learning_rate": 0.00017234919751441704,
+      "loss": 1.6686,
+      "step": 4981
+    },
+    {
+      "epoch": 0.79712,
+      "grad_norm": 2.386328161799941,
+      "learning_rate": 0.0001723372717392866,
+      "loss": 1.6758,
+      "step": 4982
+    },
+    {
+      "epoch": 0.79728,
+      "grad_norm": 2.658134003581571,
+      "learning_rate": 0.00017232534380571743,
+      "loss": 1.6662,
+      "step": 4983
+    },
+    {
+      "epoch": 0.79744,
+      "grad_norm": 2.157807559709922,
+      "learning_rate": 0.00017231341371406534,
+      "loss": 1.6811,
+      "step": 4984
+    },
+    {
+      "epoch": 0.7976,
+      "grad_norm": 1.975453017049782,
+      "learning_rate": 0.0001723014814646864,
+      "loss": 1.6496,
+      "step": 4985
+    },
+    {
+      "epoch": 0.79776,
+      "grad_norm": 2.005688326527374,
+      "learning_rate": 0.00017228954705793655,
+      "loss": 1.6834,
+      "step": 4986
+    },
+    {
+      "epoch": 0.79792,
+      "grad_norm": 2.6499979684240182,
+      "learning_rate": 0.000172277610494172,
+      "loss": 1.6374,
+      "step": 4987
+    },
+    {
+      "epoch": 0.79808,
+      "grad_norm": 2.5375815757444182,
+      "learning_rate": 0.00017226567177374888,
+      "loss": 1.6777,
+      "step": 4988
+    },
+    {
+      "epoch": 0.79824,
+      "grad_norm": 2.3552067120593616,
+      "learning_rate": 0.00017225373089702338,
+      "loss": 1.5772,
+      "step": 4989
+    },
+    {
+      "epoch": 0.7984,
+      "grad_norm": 2.33382744516553,
+      "learning_rate": 0.0001722417878643519,
+      "loss": 1.6137,
+      "step": 4990
+    },
+    {
+      "epoch": 0.79856,
+      "grad_norm": 2.509859475652075,
+      "learning_rate": 0.00017222984267609067,
+      "loss": 1.6548,
+      "step": 4991
+    },
+    {
+      "epoch": 0.79872,
+      "grad_norm": 2.17892528731286,
+      "learning_rate": 0.0001722178953325962,
+      "loss": 1.6231,
+      "step": 4992
+    },
+    {
+      "epoch": 0.79888,
+      "grad_norm": 2.637713091958338,
+      "learning_rate": 0.00017220594583422497,
+      "loss": 1.6864,
+      "step": 4993
+    },
+    {
+      "epoch": 0.79904,
+      "grad_norm": 1.9265469953570362,
+      "learning_rate": 0.00017219399418133354,
+      "loss": 1.8028,
+      "step": 4994
+    },
+    {
+      "epoch": 0.7992,
+      "grad_norm": 2.4081825130950834,
+      "learning_rate": 0.00017218204037427852,
+      "loss": 1.6507,
+      "step": 4995
+    },
+    {
+      "epoch": 0.79936,
+      "grad_norm": 2.259352097681028,
+      "learning_rate": 0.0001721700844134166,
+      "loss": 1.6728,
+      "step": 4996
+    },
+    {
+      "epoch": 0.79952,
+      "grad_norm": 2.5057440130946134,
+      "learning_rate": 0.0001721581262991045,
+      "loss": 1.6156,
+      "step": 4997
+    },
+    {
+      "epoch": 0.79968,
+      "grad_norm": 2.620590209916514,
+      "learning_rate": 0.00017214616603169905,
+      "loss": 1.7848,
+      "step": 4998
+    },
+    {
+      "epoch": 0.79984,
+      "grad_norm": 2.389455337199496,
+      "learning_rate": 0.0001721342036115572,
+      "loss": 1.6446,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 2.428927335200629,
+      "learning_rate": 0.00017212223903903575,
+      "loss": 1.5502,
+      "step": 5000
+    },
+    {
+      "epoch": 0.80016,
+      "grad_norm": 2.139247471776036,
+      "learning_rate": 0.00017211027231449181,
+      "loss": 1.6509,
+      "step": 5001
+    },
+    {
+      "epoch": 0.80032,
+      "grad_norm": 2.209136248369239,
+      "learning_rate": 0.00017209830343828243,
+      "loss": 1.7874,
+      "step": 5002
+    },
+    {
+      "epoch": 0.80048,
+      "grad_norm": 2.539959765262854,
+      "learning_rate": 0.00017208633241076472,
+      "loss": 1.6502,
+      "step": 5003
+    },
+    {
+      "epoch": 0.80064,
+      "grad_norm": 2.5442582337818753,
+      "learning_rate": 0.00017207435923229588,
+      "loss": 1.6643,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8008,
+      "grad_norm": 2.8678244440871197,
+      "learning_rate": 0.0001720623839032332,
+      "loss": 1.7185,
+      "step": 5005
+    },
+    {
+      "epoch": 0.80096,
+      "grad_norm": 2.3019448104561495,
+      "learning_rate": 0.000172050406423934,
+      "loss": 1.6977,
+      "step": 5006
+    },
+    {
+      "epoch": 0.80112,
+      "grad_norm": 2.3145557532462897,
+      "learning_rate": 0.00017203842679475563,
+      "loss": 1.6597,
+      "step": 5007
+    },
+    {
+      "epoch": 0.80128,
+      "grad_norm": 2.737574057905894,
+      "learning_rate": 0.00017202644501605562,
+      "loss": 1.667,
+      "step": 5008
+    },
+    {
+      "epoch": 0.80144,
+      "grad_norm": 2.513553672157852,
+      "learning_rate": 0.0001720144610881914,
+      "loss": 1.6138,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8016,
+      "grad_norm": 1.971577555336728,
+      "learning_rate": 0.00017200247501152063,
+      "loss": 1.6711,
+      "step": 5010
+    },
+    {
+      "epoch": 0.80176,
+      "grad_norm": 3.010366508968169,
+      "learning_rate": 0.00017199048678640094,
+      "loss": 1.6622,
+      "step": 5011
+    },
+    {
+      "epoch": 0.80192,
+      "grad_norm": 1.9348163088931019,
+      "learning_rate": 0.00017197849641318997,
+      "loss": 1.7276,
+      "step": 5012
+    },
+    {
+      "epoch": 0.80208,
+      "grad_norm": 3.1686841453186205,
+      "learning_rate": 0.0001719665038922456,
+      "loss": 1.6895,
+      "step": 5013
+    },
+    {
+      "epoch": 0.80224,
+      "grad_norm": 2.25534219944233,
+      "learning_rate": 0.0001719545092239256,
+      "loss": 1.6552,
+      "step": 5014
+    },
+    {
+      "epoch": 0.8024,
+      "grad_norm": 2.755178090784194,
+      "learning_rate": 0.00017194251240858788,
+      "loss": 1.6878,
+      "step": 5015
+    },
+    {
+      "epoch": 0.80256,
+      "grad_norm": 2.4240926915926715,
+      "learning_rate": 0.0001719305134465904,
+      "loss": 1.6508,
+      "step": 5016
+    },
+    {
+      "epoch": 0.80272,
+      "grad_norm": 2.8888373652760926,
+      "learning_rate": 0.00017191851233829127,
+      "loss": 1.7862,
+      "step": 5017
+    },
+    {
+      "epoch": 0.80288,
+      "grad_norm": 2.193790107694882,
+      "learning_rate": 0.0001719065090840485,
+      "loss": 1.7678,
+      "step": 5018
+    },
+    {
+      "epoch": 0.80304,
+      "grad_norm": 2.1124778060233944,
+      "learning_rate": 0.0001718945036842203,
+      "loss": 1.6253,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8032,
+      "grad_norm": 2.117170435469216,
+      "learning_rate": 0.00017188249613916482,
+      "loss": 1.6794,
+      "step": 5020
+    },
+    {
+      "epoch": 0.80336,
+      "grad_norm": 2.3945487424165255,
+      "learning_rate": 0.00017187048644924047,
+      "loss": 1.5917,
+      "step": 5021
+    },
+    {
+      "epoch": 0.80352,
+      "grad_norm": 2.5852826553360546,
+      "learning_rate": 0.00017185847461480549,
+      "loss": 1.6165,
+      "step": 5022
+    },
+    {
+      "epoch": 0.80368,
+      "grad_norm": 2.4847017246183283,
+      "learning_rate": 0.00017184646063621832,
+      "loss": 1.6066,
+      "step": 5023
+    },
+    {
+      "epoch": 0.80384,
+      "grad_norm": 2.450383775471315,
+      "learning_rate": 0.0001718344445138375,
+      "loss": 1.6354,
+      "step": 5024
+    },
+    {
+      "epoch": 0.804,
+      "grad_norm": 3.5501604206456694,
+      "learning_rate": 0.0001718224262480215,
+      "loss": 1.6319,
+      "step": 5025
+    },
+    {
+      "epoch": 0.80416,
+      "grad_norm": 2.424145048082857,
+      "learning_rate": 0.00017181040583912896,
+      "loss": 1.6697,
+      "step": 5026
+    },
+    {
+      "epoch": 0.80432,
+      "grad_norm": 2.225945951552796,
+      "learning_rate": 0.00017179838328751853,
+      "loss": 1.608,
+      "step": 5027
+    },
+    {
+      "epoch": 0.80448,
+      "grad_norm": 2.4161901082854857,
+      "learning_rate": 0.00017178635859354898,
+      "loss": 1.6513,
+      "step": 5028
+    },
+    {
+      "epoch": 0.80464,
+      "grad_norm": 2.4964100499205024,
+      "learning_rate": 0.00017177433175757912,
+      "loss": 1.6511,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8048,
+      "grad_norm": 2.5820005398654793,
+      "learning_rate": 0.00017176230277996774,
+      "loss": 1.6891,
+      "step": 5030
+    },
+    {
+      "epoch": 0.80496,
+      "grad_norm": 2.5171370183781803,
+      "learning_rate": 0.00017175027166107385,
+      "loss": 1.6638,
+      "step": 5031
+    },
+    {
+      "epoch": 0.80512,
+      "grad_norm": 2.1657378782482284,
+      "learning_rate": 0.0001717382384012564,
+      "loss": 1.6426,
+      "step": 5032
+    },
+    {
+      "epoch": 0.80528,
+      "grad_norm": 2.4499526919563115,
+      "learning_rate": 0.00017172620300087444,
+      "loss": 1.6179,
+      "step": 5033
+    },
+    {
+      "epoch": 0.80544,
+      "grad_norm": 2.7108455753359446,
+      "learning_rate": 0.0001717141654602871,
+      "loss": 1.6171,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8056,
+      "grad_norm": 2.5416778730505682,
+      "learning_rate": 0.00017170212577985355,
+      "loss": 1.6604,
+      "step": 5035
+    },
+    {
+      "epoch": 0.80576,
+      "grad_norm": 2.220590485274986,
+      "learning_rate": 0.00017169008395993306,
+      "loss": 1.6035,
+      "step": 5036
+    },
+    {
+      "epoch": 0.80592,
+      "grad_norm": 1.9579578365781907,
+      "learning_rate": 0.00017167804000088493,
+      "loss": 1.6539,
+      "step": 5037
+    },
+    {
+      "epoch": 0.80608,
+      "grad_norm": 2.490849526640705,
+      "learning_rate": 0.00017166599390306853,
+      "loss": 1.6758,
+      "step": 5038
+    },
+    {
+      "epoch": 0.80624,
+      "grad_norm": 2.354418488429506,
+      "learning_rate": 0.00017165394566684333,
+      "loss": 1.5965,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8064,
+      "grad_norm": 2.4050815568441255,
+      "learning_rate": 0.00017164189529256875,
+      "loss": 1.6951,
+      "step": 5040
+    },
+    {
+      "epoch": 0.80656,
+      "grad_norm": 2.594730532902109,
+      "learning_rate": 0.00017162984278060444,
+      "loss": 1.6977,
+      "step": 5041
+    },
+    {
+      "epoch": 0.80672,
+      "grad_norm": 2.3864466354555005,
+      "learning_rate": 0.00017161778813131,
+      "loss": 1.664,
+      "step": 5042
+    },
+    {
+      "epoch": 0.80688,
+      "grad_norm": 2.719709326301658,
+      "learning_rate": 0.0001716057313450451,
+      "loss": 1.7244,
+      "step": 5043
+    },
+    {
+      "epoch": 0.80704,
+      "grad_norm": 2.4818077777264635,
+      "learning_rate": 0.00017159367242216954,
+      "loss": 1.5833,
+      "step": 5044
+    },
+    {
+      "epoch": 0.8072,
+      "grad_norm": 2.798923927906584,
+      "learning_rate": 0.00017158161136304306,
+      "loss": 1.5757,
+      "step": 5045
+    },
+    {
+      "epoch": 0.80736,
+      "grad_norm": 2.796092611780878,
+      "learning_rate": 0.0001715695481680257,
+      "loss": 1.7235,
+      "step": 5046
+    },
+    {
+      "epoch": 0.80752,
+      "grad_norm": 2.261475201393247,
+      "learning_rate": 0.00017155748283747725,
+      "loss": 1.615,
+      "step": 5047
+    },
+    {
+      "epoch": 0.80768,
+      "grad_norm": 2.6367227669636715,
+      "learning_rate": 0.00017154541537175777,
+      "loss": 1.6464,
+      "step": 5048
+    },
+    {
+      "epoch": 0.80784,
+      "grad_norm": 2.3581536489868005,
+      "learning_rate": 0.00017153334577122738,
+      "loss": 1.6696,
+      "step": 5049
+    },
+    {
+      "epoch": 0.808,
+      "grad_norm": 2.359141366195366,
+      "learning_rate": 0.00017152127403624617,
+      "loss": 1.5864,
+      "step": 5050
+    },
+    {
+      "epoch": 0.80816,
+      "grad_norm": 2.4389624467471376,
+      "learning_rate": 0.00017150920016717437,
+      "loss": 1.6544,
+      "step": 5051
+    },
+    {
+      "epoch": 0.80832,
+      "grad_norm": 2.435113428007412,
+      "learning_rate": 0.00017149712416437223,
+      "loss": 1.6846,
+      "step": 5052
+    },
+    {
+      "epoch": 0.80848,
+      "grad_norm": 2.5260077681703845,
+      "learning_rate": 0.00017148504602820008,
+      "loss": 1.6779,
+      "step": 5053
+    },
+    {
+      "epoch": 0.80864,
+      "grad_norm": 2.4440158398603335,
+      "learning_rate": 0.00017147296575901832,
+      "loss": 1.6375,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8088,
+      "grad_norm": 2.5998947013191116,
+      "learning_rate": 0.00017146088335718743,
+      "loss": 1.6134,
+      "step": 5055
+    },
+    {
+      "epoch": 0.80896,
+      "grad_norm": 2.3004236722406146,
+      "learning_rate": 0.00017144879882306788,
+      "loss": 1.6313,
+      "step": 5056
+    },
+    {
+      "epoch": 0.80912,
+      "grad_norm": 2.7162338655228364,
+      "learning_rate": 0.0001714367121570203,
+      "loss": 1.6205,
+      "step": 5057
+    },
+    {
+      "epoch": 0.80928,
+      "grad_norm": 3.479740487425207,
+      "learning_rate": 0.00017142462335940534,
+      "loss": 1.7008,
+      "step": 5058
+    },
+    {
+      "epoch": 0.80944,
+      "grad_norm": 2.5420232857755787,
+      "learning_rate": 0.0001714125324305837,
+      "loss": 1.7692,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8096,
+      "grad_norm": 3.222890295503114,
+      "learning_rate": 0.0001714004393709161,
+      "loss": 1.601,
+      "step": 5060
+    },
+    {
+      "epoch": 0.80976,
+      "grad_norm": 2.28566061977434,
+      "learning_rate": 0.00017138834418076347,
+      "loss": 1.6198,
+      "step": 5061
+    },
+    {
+      "epoch": 0.80992,
+      "grad_norm": 2.2176535887168214,
+      "learning_rate": 0.0001713762468604867,
+      "loss": 1.5539,
+      "step": 5062
+    },
+    {
+      "epoch": 0.81008,
+      "grad_norm": 2.3288615565303776,
+      "learning_rate": 0.0001713641474104467,
+      "loss": 1.6719,
+      "step": 5063
+    },
+    {
+      "epoch": 0.81024,
+      "grad_norm": 2.772286196431696,
+      "learning_rate": 0.00017135204583100451,
+      "loss": 1.6448,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8104,
+      "grad_norm": 2.353422373713884,
+      "learning_rate": 0.00017133994212252125,
+      "loss": 1.6496,
+      "step": 5065
+    },
+    {
+      "epoch": 0.81056,
+      "grad_norm": 2.7211693349230996,
+      "learning_rate": 0.00017132783628535808,
+      "loss": 1.598,
+      "step": 5066
+    },
+    {
+      "epoch": 0.81072,
+      "grad_norm": 2.5428892709312088,
+      "learning_rate": 0.0001713157283198762,
+      "loss": 1.6866,
+      "step": 5067
+    },
+    {
+      "epoch": 0.81088,
+      "grad_norm": 2.5506313983004922,
+      "learning_rate": 0.00017130361822643693,
+      "loss": 1.6668,
+      "step": 5068
+    },
+    {
+      "epoch": 0.81104,
+      "grad_norm": 2.4363381314979082,
+      "learning_rate": 0.00017129150600540157,
+      "loss": 1.7388,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8112,
+      "grad_norm": 2.2928357895315954,
+      "learning_rate": 0.00017127939165713157,
+      "loss": 1.5702,
+      "step": 5070
+    },
+    {
+      "epoch": 0.81136,
+      "grad_norm": 2.3718485879378712,
+      "learning_rate": 0.0001712672751819884,
+      "loss": 1.5471,
+      "step": 5071
+    },
+    {
+      "epoch": 0.81152,
+      "grad_norm": 2.2938222745562746,
+      "learning_rate": 0.00017125515658033356,
+      "loss": 1.6559,
+      "step": 5072
+    },
+    {
+      "epoch": 0.81168,
+      "grad_norm": 2.2738673308811794,
+      "learning_rate": 0.00017124303585252867,
+      "loss": 1.6362,
+      "step": 5073
+    },
+    {
+      "epoch": 0.81184,
+      "grad_norm": 2.3303962223192998,
+      "learning_rate": 0.00017123091299893543,
+      "loss": 1.5655,
+      "step": 5074
+    },
+    {
+      "epoch": 0.812,
+      "grad_norm": 2.3000454112787665,
+      "learning_rate": 0.00017121878801991554,
+      "loss": 1.5443,
+      "step": 5075
+    },
+    {
+      "epoch": 0.81216,
+      "grad_norm": 2.908841236781188,
+      "learning_rate": 0.00017120666091583077,
+      "loss": 1.5943,
+      "step": 5076
+    },
+    {
+      "epoch": 0.81232,
+      "grad_norm": 2.2416340894145232,
+      "learning_rate": 0.000171194531687043,
+      "loss": 1.5804,
+      "step": 5077
+    },
+    {
+      "epoch": 0.81248,
+      "grad_norm": 2.40663282463411,
+      "learning_rate": 0.00017118240033391416,
+      "loss": 1.6308,
+      "step": 5078
+    },
+    {
+      "epoch": 0.81264,
+      "grad_norm": 2.091461470541317,
+      "learning_rate": 0.00017117026685680621,
+      "loss": 1.695,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8128,
+      "grad_norm": 3.23977938427778,
+      "learning_rate": 0.00017115813125608123,
+      "loss": 1.6877,
+      "step": 5080
+    },
+    {
+      "epoch": 0.81296,
+      "grad_norm": 2.2947240863571468,
+      "learning_rate": 0.00017114599353210125,
+      "loss": 1.6344,
+      "step": 5081
+    },
+    {
+      "epoch": 0.81312,
+      "grad_norm": 2.229418477488605,
+      "learning_rate": 0.00017113385368522854,
+      "loss": 1.6611,
+      "step": 5082
+    },
+    {
+      "epoch": 0.81328,
+      "grad_norm": 2.493678557342841,
+      "learning_rate": 0.00017112171171582526,
+      "loss": 1.6856,
+      "step": 5083
+    },
+    {
+      "epoch": 0.81344,
+      "grad_norm": 2.6256252371464366,
+      "learning_rate": 0.00017110956762425374,
+      "loss": 1.6838,
+      "step": 5084
+    },
+    {
+      "epoch": 0.8136,
+      "grad_norm": 2.011213374827839,
+      "learning_rate": 0.00017109742141087635,
+      "loss": 1.5602,
+      "step": 5085
+    },
+    {
+      "epoch": 0.81376,
+      "grad_norm": 2.424827393426642,
+      "learning_rate": 0.0001710852730760555,
+      "loss": 1.6423,
+      "step": 5086
+    },
+    {
+      "epoch": 0.81392,
+      "grad_norm": 3.2848689176207047,
+      "learning_rate": 0.00017107312262015366,
+      "loss": 1.6616,
+      "step": 5087
+    },
+    {
+      "epoch": 0.81408,
+      "grad_norm": 2.5513941058289906,
+      "learning_rate": 0.00017106097004353346,
+      "loss": 1.6097,
+      "step": 5088
+    },
+    {
+      "epoch": 0.81424,
+      "grad_norm": 2.4187528182098283,
+      "learning_rate": 0.00017104881534655742,
+      "loss": 1.5716,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8144,
+      "grad_norm": 2.018384044978108,
+      "learning_rate": 0.00017103665852958826,
+      "loss": 1.6687,
+      "step": 5090
+    },
+    {
+      "epoch": 0.81456,
+      "grad_norm": 2.2121539481873276,
+      "learning_rate": 0.00017102449959298873,
+      "loss": 1.6372,
+      "step": 5091
+    },
+    {
+      "epoch": 0.81472,
+      "grad_norm": 3.0608212633327834,
+      "learning_rate": 0.00017101233853712163,
+      "loss": 1.7222,
+      "step": 5092
+    },
+    {
+      "epoch": 0.81488,
+      "grad_norm": 2.517192549227739,
+      "learning_rate": 0.0001710001753623498,
+      "loss": 1.6752,
+      "step": 5093
+    },
+    {
+      "epoch": 0.81504,
+      "grad_norm": 2.318057105187797,
+      "learning_rate": 0.00017098801006903625,
+      "loss": 1.5897,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8152,
+      "grad_norm": 2.0877476878693897,
+      "learning_rate": 0.00017097584265754386,
+      "loss": 1.6422,
+      "step": 5095
+    },
+    {
+      "epoch": 0.81536,
+      "grad_norm": 2.3524931139217644,
+      "learning_rate": 0.0001709636731282358,
+      "loss": 1.5764,
+      "step": 5096
+    },
+    {
+      "epoch": 0.81552,
+      "grad_norm": 2.255650380338999,
+      "learning_rate": 0.0001709515014814751,
+      "loss": 1.6952,
+      "step": 5097
+    },
+    {
+      "epoch": 0.81568,
+      "grad_norm": 2.310990013847917,
+      "learning_rate": 0.000170939327717625,
+      "loss": 1.6075,
+      "step": 5098
+    },
+    {
+      "epoch": 0.81584,
+      "grad_norm": 2.0990855274722326,
+      "learning_rate": 0.00017092715183704874,
+      "loss": 1.6096,
+      "step": 5099
+    },
+    {
+      "epoch": 0.816,
+      "grad_norm": 2.000972197596777,
+      "learning_rate": 0.00017091497384010958,
+      "loss": 1.6509,
+      "step": 5100
+    },
+    {
+      "epoch": 0.81616,
+      "grad_norm": 2.4575358493085244,
+      "learning_rate": 0.00017090279372717094,
+      "loss": 1.7724,
+      "step": 5101
+    },
+    {
+      "epoch": 0.81632,
+      "grad_norm": 2.8578204563592533,
+      "learning_rate": 0.0001708906114985963,
+      "loss": 1.5945,
+      "step": 5102
+    },
+    {
+      "epoch": 0.81648,
+      "grad_norm": 2.070268048909114,
+      "learning_rate": 0.00017087842715474908,
+      "loss": 1.6637,
+      "step": 5103
+    },
+    {
+      "epoch": 0.81664,
+      "grad_norm": 2.13815752247315,
+      "learning_rate": 0.00017086624069599288,
+      "loss": 1.6521,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8168,
+      "grad_norm": 2.858461450888837,
+      "learning_rate": 0.00017085405212269132,
+      "loss": 1.6448,
+      "step": 5105
+    },
+    {
+      "epoch": 0.81696,
+      "grad_norm": 2.3313998199128005,
+      "learning_rate": 0.00017084186143520808,
+      "loss": 1.6455,
+      "step": 5106
+    },
+    {
+      "epoch": 0.81712,
+      "grad_norm": 4.4952665735516755,
+      "learning_rate": 0.00017082966863390695,
+      "loss": 1.683,
+      "step": 5107
+    },
+    {
+      "epoch": 0.81728,
+      "grad_norm": 2.370956082092176,
+      "learning_rate": 0.0001708174737191517,
+      "loss": 1.8084,
+      "step": 5108
+    },
+    {
+      "epoch": 0.81744,
+      "grad_norm": 3.0897622784942684,
+      "learning_rate": 0.0001708052766913062,
+      "loss": 1.7032,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8176,
+      "grad_norm": 2.4410312973223998,
+      "learning_rate": 0.00017079307755073445,
+      "loss": 1.6438,
+      "step": 5110
+    },
+    {
+      "epoch": 0.81776,
+      "grad_norm": 2.96055485151419,
+      "learning_rate": 0.00017078087629780045,
+      "loss": 1.6249,
+      "step": 5111
+    },
+    {
+      "epoch": 0.81792,
+      "grad_norm": 3.052620596364729,
+      "learning_rate": 0.0001707686729328682,
+      "loss": 1.7422,
+      "step": 5112
+    },
+    {
+      "epoch": 0.81808,
+      "grad_norm": 2.266431293677717,
+      "learning_rate": 0.00017075646745630185,
+      "loss": 1.6082,
+      "step": 5113
+    },
+    {
+      "epoch": 0.81824,
+      "grad_norm": 4.009739004077887,
+      "learning_rate": 0.00017074425986846566,
+      "loss": 1.6181,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8184,
+      "grad_norm": 2.620592747975133,
+      "learning_rate": 0.00017073205016972384,
+      "loss": 1.631,
+      "step": 5115
+    },
+    {
+      "epoch": 0.81856,
+      "grad_norm": 4.497656710816251,
+      "learning_rate": 0.00017071983836044064,
+      "loss": 1.781,
+      "step": 5116
+    },
+    {
+      "epoch": 0.81872,
+      "grad_norm": 2.6084906109733543,
+      "learning_rate": 0.0001707076244409806,
+      "loss": 1.7047,
+      "step": 5117
+    },
+    {
+      "epoch": 0.81888,
+      "grad_norm": 2.4880581965328172,
+      "learning_rate": 0.00017069540841170802,
+      "loss": 1.6061,
+      "step": 5118
+    },
+    {
+      "epoch": 0.81904,
+      "grad_norm": 2.4317788039342654,
+      "learning_rate": 0.0001706831902729875,
+      "loss": 1.5826,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8192,
+      "grad_norm": 2.873659179880561,
+      "learning_rate": 0.00017067097002518357,
+      "loss": 1.6868,
+      "step": 5120
+    },
+    {
+      "epoch": 0.81936,
+      "grad_norm": 2.18339454866856,
+      "learning_rate": 0.0001706587476686609,
+      "loss": 1.6435,
+      "step": 5121
+    },
+    {
+      "epoch": 0.81952,
+      "grad_norm": 2.619214430689372,
+      "learning_rate": 0.00017064652320378408,
+      "loss": 1.65,
+      "step": 5122
+    },
+    {
+      "epoch": 0.81968,
+      "grad_norm": 2.274221895392364,
+      "learning_rate": 0.000170634296630918,
+      "loss": 1.6508,
+      "step": 5123
+    },
+    {
+      "epoch": 0.81984,
+      "grad_norm": 2.3309858483563963,
+      "learning_rate": 0.00017062206795042745,
+      "loss": 1.6168,
+      "step": 5124
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 2.4014918262129297,
+      "learning_rate": 0.0001706098371626773,
+      "loss": 1.5741,
+      "step": 5125
+    },
+    {
+      "epoch": 0.82016,
+      "grad_norm": 2.6811925485345367,
+      "learning_rate": 0.00017059760426803248,
+      "loss": 1.6557,
+      "step": 5126
+    },
+    {
+      "epoch": 0.82032,
+      "grad_norm": 2.345799016392057,
+      "learning_rate": 0.000170585369266858,
+      "loss": 1.6397,
+      "step": 5127
+    },
+    {
+      "epoch": 0.82048,
+      "grad_norm": 2.2273003817138775,
+      "learning_rate": 0.000170573132159519,
+      "loss": 1.6202,
+      "step": 5128
+    },
+    {
+      "epoch": 0.82064,
+      "grad_norm": 2.6016179315501073,
+      "learning_rate": 0.00017056089294638057,
+      "loss": 1.63,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8208,
+      "grad_norm": 2.4796615354150826,
+      "learning_rate": 0.00017054865162780786,
+      "loss": 1.7144,
+      "step": 5130
+    },
+    {
+      "epoch": 0.82096,
+      "grad_norm": 2.397102228121756,
+      "learning_rate": 0.00017053640820416624,
+      "loss": 1.6773,
+      "step": 5131
+    },
+    {
+      "epoch": 0.82112,
+      "grad_norm": 2.1135370622797107,
+      "learning_rate": 0.00017052416267582094,
+      "loss": 1.6153,
+      "step": 5132
+    },
+    {
+      "epoch": 0.82128,
+      "grad_norm": 2.4158753727780033,
+      "learning_rate": 0.00017051191504313743,
+      "loss": 1.7155,
+      "step": 5133
+    },
+    {
+      "epoch": 0.82144,
+      "grad_norm": 2.4553471830848883,
+      "learning_rate": 0.00017049966530648109,
+      "loss": 1.6436,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8216,
+      "grad_norm": 2.828249696043937,
+      "learning_rate": 0.00017048741346621752,
+      "loss": 1.5871,
+      "step": 5135
+    },
+    {
+      "epoch": 0.82176,
+      "grad_norm": 2.263215458271656,
+      "learning_rate": 0.0001704751595227122,
+      "loss": 1.6441,
+      "step": 5136
+    },
+    {
+      "epoch": 0.82192,
+      "grad_norm": 2.433178894302559,
+      "learning_rate": 0.0001704629034763308,
+      "loss": 1.5518,
+      "step": 5137
+    },
+    {
+      "epoch": 0.82208,
+      "grad_norm": 2.1168005638269483,
+      "learning_rate": 0.00017045064532743905,
+      "loss": 1.5808,
+      "step": 5138
+    },
+    {
+      "epoch": 0.82224,
+      "grad_norm": 2.403596858713808,
+      "learning_rate": 0.00017043838507640267,
+      "loss": 1.5494,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8224,
+      "grad_norm": 2.5619653771622635,
+      "learning_rate": 0.00017042612272358758,
+      "loss": 1.5503,
+      "step": 5140
+    },
+    {
+      "epoch": 0.82256,
+      "grad_norm": 2.0644447050654846,
+      "learning_rate": 0.0001704138582693596,
+      "loss": 1.676,
+      "step": 5141
+    },
+    {
+      "epoch": 0.82272,
+      "grad_norm": 2.3160846582431556,
+      "learning_rate": 0.00017040159171408464,
+      "loss": 1.62,
+      "step": 5142
+    },
+    {
+      "epoch": 0.82288,
+      "grad_norm": 3.507769755277463,
+      "learning_rate": 0.00017038932305812882,
+      "loss": 1.7187,
+      "step": 5143
+    },
+    {
+      "epoch": 0.82304,
+      "grad_norm": 2.5873427514219816,
+      "learning_rate": 0.00017037705230185814,
+      "loss": 1.7222,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8232,
+      "grad_norm": 2.771205816766397,
+      "learning_rate": 0.00017036477944563878,
+      "loss": 1.6693,
+      "step": 5145
+    },
+    {
+      "epoch": 0.82336,
+      "grad_norm": 2.091846598136946,
+      "learning_rate": 0.00017035250448983695,
+      "loss": 1.6365,
+      "step": 5146
+    },
+    {
+      "epoch": 0.82352,
+      "grad_norm": 2.863433816932537,
+      "learning_rate": 0.0001703402274348189,
+      "loss": 1.6494,
+      "step": 5147
+    },
+    {
+      "epoch": 0.82368,
+      "grad_norm": 2.2644895274967003,
+      "learning_rate": 0.0001703279482809509,
+      "loss": 1.6214,
+      "step": 5148
+    },
+    {
+      "epoch": 0.82384,
+      "grad_norm": 3.4268852803949374,
+      "learning_rate": 0.00017031566702859947,
+      "loss": 1.75,
+      "step": 5149
+    },
+    {
+      "epoch": 0.824,
+      "grad_norm": 2.5522158967886965,
+      "learning_rate": 0.00017030338367813097,
+      "loss": 1.5936,
+      "step": 5150
+    },
+    {
+      "epoch": 0.82416,
+      "grad_norm": 3.1539043609492947,
+      "learning_rate": 0.00017029109822991193,
+      "loss": 1.5905,
+      "step": 5151
+    },
+    {
+      "epoch": 0.82432,
+      "grad_norm": 2.5865804842934605,
+      "learning_rate": 0.000170278810684309,
+      "loss": 1.6583,
+      "step": 5152
+    },
+    {
+      "epoch": 0.82448,
+      "grad_norm": 2.320663974350077,
+      "learning_rate": 0.0001702665210416887,
+      "loss": 1.6619,
+      "step": 5153
+    },
+    {
+      "epoch": 0.82464,
+      "grad_norm": 2.2096292782892815,
+      "learning_rate": 0.00017025422930241782,
+      "loss": 1.6368,
+      "step": 5154
+    },
+    {
+      "epoch": 0.8248,
+      "grad_norm": 2.6039967571587264,
+      "learning_rate": 0.00017024193546686314,
+      "loss": 1.7659,
+      "step": 5155
+    },
+    {
+      "epoch": 0.82496,
+      "grad_norm": 2.3611673320028115,
+      "learning_rate": 0.00017022963953539143,
+      "loss": 1.6199,
+      "step": 5156
+    },
+    {
+      "epoch": 0.82512,
+      "grad_norm": 2.2828637553159226,
+      "learning_rate": 0.0001702173415083696,
+      "loss": 1.6657,
+      "step": 5157
+    },
+    {
+      "epoch": 0.82528,
+      "grad_norm": 2.400179829070947,
+      "learning_rate": 0.00017020504138616464,
+      "loss": 1.6308,
+      "step": 5158
+    },
+    {
+      "epoch": 0.82544,
+      "grad_norm": 2.4733245292795427,
+      "learning_rate": 0.00017019273916914354,
+      "loss": 1.6725,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8256,
+      "grad_norm": 2.4914587799298955,
+      "learning_rate": 0.0001701804348576734,
+      "loss": 1.6516,
+      "step": 5160
+    },
+    {
+      "epoch": 0.82576,
+      "grad_norm": 2.1918559334039305,
+      "learning_rate": 0.0001701681284521213,
+      "loss": 1.6085,
+      "step": 5161
+    },
+    {
+      "epoch": 0.82592,
+      "grad_norm": 2.9530332149895213,
+      "learning_rate": 0.00017015581995285451,
+      "loss": 1.5824,
+      "step": 5162
+    },
+    {
+      "epoch": 0.82608,
+      "grad_norm": 2.2703134340739184,
+      "learning_rate": 0.0001701435093602403,
+      "loss": 1.6111,
+      "step": 5163
+    },
+    {
+      "epoch": 0.82624,
+      "grad_norm": 2.5283564645190832,
+      "learning_rate": 0.00017013119667464599,
+      "loss": 1.5249,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8264,
+      "grad_norm": 2.070891888423819,
+      "learning_rate": 0.00017011888189643893,
+      "loss": 1.6463,
+      "step": 5165
+    },
+    {
+      "epoch": 0.82656,
+      "grad_norm": 2.344192124751423,
+      "learning_rate": 0.00017010656502598664,
+      "loss": 1.718,
+      "step": 5166
+    },
+    {
+      "epoch": 0.82672,
+      "grad_norm": 2.2658855457213485,
+      "learning_rate": 0.0001700942460636566,
+      "loss": 1.6781,
+      "step": 5167
+    },
+    {
+      "epoch": 0.82688,
+      "grad_norm": 2.3331287790051958,
+      "learning_rate": 0.00017008192500981637,
+      "loss": 1.656,
+      "step": 5168
+    },
+    {
+      "epoch": 0.82704,
+      "grad_norm": 2.7765618913734507,
+      "learning_rate": 0.00017006960186483365,
+      "loss": 1.6213,
+      "step": 5169
+    },
+    {
+      "epoch": 0.8272,
+      "grad_norm": 2.3290170088526043,
+      "learning_rate": 0.00017005727662907608,
+      "loss": 1.6789,
+      "step": 5170
+    },
+    {
+      "epoch": 0.82736,
+      "grad_norm": 1.9381577977431035,
+      "learning_rate": 0.00017004494930291148,
+      "loss": 1.6433,
+      "step": 5171
+    },
+    {
+      "epoch": 0.82752,
+      "grad_norm": 2.0219486049994626,
+      "learning_rate": 0.00017003261988670767,
+      "loss": 1.6084,
+      "step": 5172
+    },
+    {
+      "epoch": 0.82768,
+      "grad_norm": 2.8189250319286976,
+      "learning_rate": 0.00017002028838083253,
+      "loss": 1.6847,
+      "step": 5173
+    },
+    {
+      "epoch": 0.82784,
+      "grad_norm": 2.527883177493222,
+      "learning_rate": 0.00017000795478565398,
+      "loss": 1.6712,
+      "step": 5174
+    },
+    {
+      "epoch": 0.828,
+      "grad_norm": 2.426599982507322,
+      "learning_rate": 0.00016999561910154012,
+      "loss": 1.6183,
+      "step": 5175
+    },
+    {
+      "epoch": 0.82816,
+      "grad_norm": 2.3241566145078427,
+      "learning_rate": 0.00016998328132885893,
+      "loss": 1.7068,
+      "step": 5176
+    },
+    {
+      "epoch": 0.82832,
+      "grad_norm": 2.502321937430721,
+      "learning_rate": 0.00016997094146797864,
+      "loss": 1.5889,
+      "step": 5177
+    },
+    {
+      "epoch": 0.82848,
+      "grad_norm": 2.5167133867135987,
+      "learning_rate": 0.0001699585995192674,
+      "loss": 1.6637,
+      "step": 5178
+    },
+    {
+      "epoch": 0.82864,
+      "grad_norm": 2.3220640096654894,
+      "learning_rate": 0.0001699462554830935,
+      "loss": 1.6304,
+      "step": 5179
+    },
+    {
+      "epoch": 0.8288,
+      "grad_norm": 2.9949520897830695,
+      "learning_rate": 0.00016993390935982525,
+      "loss": 1.7048,
+      "step": 5180
+    },
+    {
+      "epoch": 0.82896,
+      "grad_norm": 2.3487671214688355,
+      "learning_rate": 0.00016992156114983104,
+      "loss": 1.7097,
+      "step": 5181
+    },
+    {
+      "epoch": 0.82912,
+      "grad_norm": 2.20580340590374,
+      "learning_rate": 0.00016990921085347936,
+      "loss": 1.6481,
+      "step": 5182
+    },
+    {
+      "epoch": 0.82928,
+      "grad_norm": 2.9121669603128706,
+      "learning_rate": 0.0001698968584711387,
+      "loss": 1.6808,
+      "step": 5183
+    },
+    {
+      "epoch": 0.82944,
+      "grad_norm": 2.685951924382391,
+      "learning_rate": 0.00016988450400317757,
+      "loss": 1.6318,
+      "step": 5184
+    },
+    {
+      "epoch": 0.8296,
+      "grad_norm": 2.5652657413876763,
+      "learning_rate": 0.0001698721474499647,
+      "loss": 1.5875,
+      "step": 5185
+    },
+    {
+      "epoch": 0.82976,
+      "grad_norm": 2.2458844701629,
+      "learning_rate": 0.0001698597888118688,
+      "loss": 1.6006,
+      "step": 5186
+    },
+    {
+      "epoch": 0.82992,
+      "grad_norm": 2.3985609549878486,
+      "learning_rate": 0.00016984742808925858,
+      "loss": 1.6712,
+      "step": 5187
+    },
+    {
+      "epoch": 0.83008,
+      "grad_norm": 1.947510673121757,
+      "learning_rate": 0.0001698350652825029,
+      "loss": 1.6156,
+      "step": 5188
+    },
+    {
+      "epoch": 0.83024,
+      "grad_norm": 2.295887481843522,
+      "learning_rate": 0.0001698227003919706,
+      "loss": 1.515,
+      "step": 5189
+    },
+    {
+      "epoch": 0.8304,
+      "grad_norm": 2.2291055871306646,
+      "learning_rate": 0.00016981033341803068,
+      "loss": 1.632,
+      "step": 5190
+    },
+    {
+      "epoch": 0.83056,
+      "grad_norm": 2.3829742366246065,
+      "learning_rate": 0.00016979796436105213,
+      "loss": 1.6471,
+      "step": 5191
+    },
+    {
+      "epoch": 0.83072,
+      "grad_norm": 2.5687032077246705,
+      "learning_rate": 0.00016978559322140403,
+      "loss": 1.6772,
+      "step": 5192
+    },
+    {
+      "epoch": 0.83088,
+      "grad_norm": 2.867945235486024,
+      "learning_rate": 0.00016977321999945552,
+      "loss": 1.7133,
+      "step": 5193
+    },
+    {
+      "epoch": 0.83104,
+      "grad_norm": 2.8375322683052926,
+      "learning_rate": 0.0001697608446955758,
+      "loss": 1.6483,
+      "step": 5194
+    },
+    {
+      "epoch": 0.8312,
+      "grad_norm": 2.348204641386847,
+      "learning_rate": 0.0001697484673101341,
+      "loss": 1.6273,
+      "step": 5195
+    },
+    {
+      "epoch": 0.83136,
+      "grad_norm": 2.7655541708641618,
+      "learning_rate": 0.0001697360878434998,
+      "loss": 1.6675,
+      "step": 5196
+    },
+    {
+      "epoch": 0.83152,
+      "grad_norm": 2.5440768773591023,
+      "learning_rate": 0.00016972370629604226,
+      "loss": 1.6699,
+      "step": 5197
+    },
+    {
+      "epoch": 0.83168,
+      "grad_norm": 2.6338892606872037,
+      "learning_rate": 0.00016971132266813092,
+      "loss": 1.6647,
+      "step": 5198
+    },
+    {
+      "epoch": 0.83184,
+      "grad_norm": 2.434853037099858,
+      "learning_rate": 0.00016969893696013527,
+      "loss": 1.704,
+      "step": 5199
+    },
+    {
+      "epoch": 0.832,
+      "grad_norm": 2.534012125482912,
+      "learning_rate": 0.00016968654917242493,
+      "loss": 1.6074,
+      "step": 5200
+    },
+    {
+      "epoch": 0.83216,
+      "grad_norm": 1.9623798368685068,
+      "learning_rate": 0.00016967415930536946,
+      "loss": 1.6717,
+      "step": 5201
+    },
+    {
+      "epoch": 0.83232,
+      "grad_norm": 2.3530226523749276,
+      "learning_rate": 0.00016966176735933866,
+      "loss": 1.6908,
+      "step": 5202
+    },
+    {
+      "epoch": 0.83248,
+      "grad_norm": 2.448319640965319,
+      "learning_rate": 0.0001696493733347022,
+      "loss": 1.6824,
+      "step": 5203
+    },
+    {
+      "epoch": 0.83264,
+      "grad_norm": 2.4120664631592184,
+      "learning_rate": 0.0001696369772318299,
+      "loss": 1.6192,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8328,
+      "grad_norm": 2.1835690463140476,
+      "learning_rate": 0.00016962457905109172,
+      "loss": 1.6484,
+      "step": 5205
+    },
+    {
+      "epoch": 0.83296,
+      "grad_norm": 2.3796792361621963,
+      "learning_rate": 0.00016961217879285756,
+      "loss": 1.5852,
+      "step": 5206
+    },
+    {
+      "epoch": 0.83312,
+      "grad_norm": 1.8576008328895524,
+      "learning_rate": 0.00016959977645749738,
+      "loss": 1.5462,
+      "step": 5207
+    },
+    {
+      "epoch": 0.83328,
+      "grad_norm": 2.327476365741461,
+      "learning_rate": 0.0001695873720453813,
+      "loss": 1.6528,
+      "step": 5208
+    },
+    {
+      "epoch": 0.83344,
+      "grad_norm": 2.1294001832621654,
+      "learning_rate": 0.00016957496555687946,
+      "loss": 1.5792,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8336,
+      "grad_norm": 2.3411251366422565,
+      "learning_rate": 0.000169562556992362,
+      "loss": 1.6646,
+      "step": 5210
+    },
+    {
+      "epoch": 0.83376,
+      "grad_norm": 2.560835885033415,
+      "learning_rate": 0.00016955014635219921,
+      "loss": 1.6223,
+      "step": 5211
+    },
+    {
+      "epoch": 0.83392,
+      "grad_norm": 2.166442925393673,
+      "learning_rate": 0.0001695377336367614,
+      "loss": 1.6611,
+      "step": 5212
+    },
+    {
+      "epoch": 0.83408,
+      "grad_norm": 2.283608353141747,
+      "learning_rate": 0.00016952531884641898,
+      "loss": 1.6676,
+      "step": 5213
+    },
+    {
+      "epoch": 0.83424,
+      "grad_norm": 2.085545769305915,
+      "learning_rate": 0.00016951290198154233,
+      "loss": 1.6229,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8344,
+      "grad_norm": 2.3709304355291736,
+      "learning_rate": 0.00016950048304250194,
+      "loss": 1.6933,
+      "step": 5215
+    },
+    {
+      "epoch": 0.83456,
+      "grad_norm": 2.1889433572935917,
+      "learning_rate": 0.00016948806202966846,
+      "loss": 1.6708,
+      "step": 5216
+    },
+    {
+      "epoch": 0.83472,
+      "grad_norm": 2.381859974607401,
+      "learning_rate": 0.00016947563894341246,
+      "loss": 1.6699,
+      "step": 5217
+    },
+    {
+      "epoch": 0.83488,
+      "grad_norm": 2.234145221175795,
+      "learning_rate": 0.00016946321378410458,
+      "loss": 1.7147,
+      "step": 5218
+    },
+    {
+      "epoch": 0.83504,
+      "grad_norm": 2.854766954288137,
+      "learning_rate": 0.00016945078655211565,
+      "loss": 1.644,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8352,
+      "grad_norm": 2.032196000983913,
+      "learning_rate": 0.00016943835724781645,
+      "loss": 1.6783,
+      "step": 5220
+    },
+    {
+      "epoch": 0.83536,
+      "grad_norm": 2.441536750548125,
+      "learning_rate": 0.00016942592587157785,
+      "loss": 1.5519,
+      "step": 5221
+    },
+    {
+      "epoch": 0.83552,
+      "grad_norm": 2.4422268843338095,
+      "learning_rate": 0.0001694134924237708,
+      "loss": 1.5229,
+      "step": 5222
+    },
+    {
+      "epoch": 0.83568,
+      "grad_norm": 2.5582874549167958,
+      "learning_rate": 0.00016940105690476629,
+      "loss": 1.6459,
+      "step": 5223
+    },
+    {
+      "epoch": 0.83584,
+      "grad_norm": 2.2853546532903715,
+      "learning_rate": 0.00016938861931493535,
+      "loss": 1.5474,
+      "step": 5224
+    },
+    {
+      "epoch": 0.836,
+      "grad_norm": 2.3178042108927848,
+      "learning_rate": 0.0001693761796546491,
+      "loss": 1.6089,
+      "step": 5225
+    },
+    {
+      "epoch": 0.83616,
+      "grad_norm": 2.8103642244552987,
+      "learning_rate": 0.00016936373792427877,
+      "loss": 1.643,
+      "step": 5226
+    },
+    {
+      "epoch": 0.83632,
+      "grad_norm": 1.785164906233104,
+      "learning_rate": 0.00016935129412419558,
+      "loss": 1.5554,
+      "step": 5227
+    },
+    {
+      "epoch": 0.83648,
+      "grad_norm": 2.174144634646613,
+      "learning_rate": 0.0001693388482547708,
+      "loss": 1.6241,
+      "step": 5228
+    },
+    {
+      "epoch": 0.83664,
+      "grad_norm": 2.321969644080683,
+      "learning_rate": 0.00016932640031637583,
+      "loss": 1.5136,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8368,
+      "grad_norm": 2.0666158477160295,
+      "learning_rate": 0.00016931395030938214,
+      "loss": 1.5728,
+      "step": 5230
+    },
+    {
+      "epoch": 0.83696,
+      "grad_norm": 2.110774606616366,
+      "learning_rate": 0.00016930149823416113,
+      "loss": 1.5368,
+      "step": 5231
+    },
+    {
+      "epoch": 0.83712,
+      "grad_norm": 2.5220640036933424,
+      "learning_rate": 0.00016928904409108442,
+      "loss": 1.5912,
+      "step": 5232
+    },
+    {
+      "epoch": 0.83728,
+      "grad_norm": 2.217232272471307,
+      "learning_rate": 0.00016927658788052356,
+      "loss": 1.6344,
+      "step": 5233
+    },
+    {
+      "epoch": 0.83744,
+      "grad_norm": 2.187820066046703,
+      "learning_rate": 0.00016926412960285029,
+      "loss": 1.6483,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8376,
+      "grad_norm": 2.036314327896362,
+      "learning_rate": 0.00016925166925843633,
+      "loss": 1.5743,
+      "step": 5235
+    },
+    {
+      "epoch": 0.83776,
+      "grad_norm": 2.045073417218554,
+      "learning_rate": 0.00016923920684765346,
+      "loss": 1.6139,
+      "step": 5236
+    },
+    {
+      "epoch": 0.83792,
+      "grad_norm": 2.233314094761698,
+      "learning_rate": 0.00016922674237087354,
+      "loss": 1.6292,
+      "step": 5237
+    },
+    {
+      "epoch": 0.83808,
+      "grad_norm": 2.3258551357168633,
+      "learning_rate": 0.00016921427582846851,
+      "loss": 1.5964,
+      "step": 5238
+    },
+    {
+      "epoch": 0.83824,
+      "grad_norm": 2.097349188384737,
+      "learning_rate": 0.00016920180722081036,
+      "loss": 1.5355,
+      "step": 5239
+    },
+    {
+      "epoch": 0.8384,
+      "grad_norm": 2.3500648204001373,
+      "learning_rate": 0.00016918933654827109,
+      "loss": 1.6521,
+      "step": 5240
+    },
+    {
+      "epoch": 0.83856,
+      "grad_norm": 2.263803893554943,
+      "learning_rate": 0.00016917686381122288,
+      "loss": 1.6141,
+      "step": 5241
+    },
+    {
+      "epoch": 0.83872,
+      "grad_norm": 2.5696223074281344,
+      "learning_rate": 0.00016916438901003783,
+      "loss": 1.5656,
+      "step": 5242
+    },
+    {
+      "epoch": 0.83888,
+      "grad_norm": 2.0798484088589158,
+      "learning_rate": 0.00016915191214508817,
+      "loss": 1.6959,
+      "step": 5243
+    },
+    {
+      "epoch": 0.83904,
+      "grad_norm": 2.436214846533726,
+      "learning_rate": 0.00016913943321674623,
+      "loss": 1.6027,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8392,
+      "grad_norm": 2.336921718857263,
+      "learning_rate": 0.00016912695222538437,
+      "loss": 1.6534,
+      "step": 5245
+    },
+    {
+      "epoch": 0.83936,
+      "grad_norm": 2.5126450539770944,
+      "learning_rate": 0.000169114469171375,
+      "loss": 1.6467,
+      "step": 5246
+    },
+    {
+      "epoch": 0.83952,
+      "grad_norm": 2.4693501955041586,
+      "learning_rate": 0.0001691019840550905,
+      "loss": 1.7101,
+      "step": 5247
+    },
+    {
+      "epoch": 0.83968,
+      "grad_norm": 1.8948597664515991,
+      "learning_rate": 0.0001690894968769036,
+      "loss": 1.6265,
+      "step": 5248
+    },
+    {
+      "epoch": 0.83984,
+      "grad_norm": 2.161521213129398,
+      "learning_rate": 0.00016907700763718673,
+      "loss": 1.5283,
+      "step": 5249
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 2.1986035225332157,
+      "learning_rate": 0.00016906451633631257,
+      "loss": 1.6362,
+      "step": 5250
+    },
+    {
+      "epoch": 0.84016,
+      "grad_norm": 2.5622020606062783,
+      "learning_rate": 0.00016905202297465394,
+      "loss": 1.6591,
+      "step": 5251
+    },
+    {
+      "epoch": 0.84032,
+      "grad_norm": 1.9445506551414962,
+      "learning_rate": 0.00016903952755258352,
+      "loss": 1.6997,
+      "step": 5252
+    },
+    {
+      "epoch": 0.84048,
+      "grad_norm": 1.97550794807075,
+      "learning_rate": 0.00016902703007047425,
+      "loss": 1.5788,
+      "step": 5253
+    },
+    {
+      "epoch": 0.84064,
+      "grad_norm": 2.1640378326815934,
+      "learning_rate": 0.00016901453052869893,
+      "loss": 1.59,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8408,
+      "grad_norm": 2.1936915517961135,
+      "learning_rate": 0.00016900202892763063,
+      "loss": 1.5651,
+      "step": 5255
+    },
+    {
+      "epoch": 0.84096,
+      "grad_norm": 2.571434218916266,
+      "learning_rate": 0.00016898952526764232,
+      "loss": 1.6176,
+      "step": 5256
+    },
+    {
+      "epoch": 0.84112,
+      "grad_norm": 2.448913078568561,
+      "learning_rate": 0.00016897701954910708,
+      "loss": 1.6578,
+      "step": 5257
+    },
+    {
+      "epoch": 0.84128,
+      "grad_norm": 2.9816185595630977,
+      "learning_rate": 0.0001689645117723981,
+      "loss": 1.667,
+      "step": 5258
+    },
+    {
+      "epoch": 0.84144,
+      "grad_norm": 2.323805108810936,
+      "learning_rate": 0.00016895200193788862,
+      "loss": 1.6612,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8416,
+      "grad_norm": 2.556086289404926,
+      "learning_rate": 0.00016893949004595182,
+      "loss": 1.5788,
+      "step": 5260
+    },
+    {
+      "epoch": 0.84176,
+      "grad_norm": 1.9838100475410065,
+      "learning_rate": 0.0001689269760969611,
+      "loss": 1.647,
+      "step": 5261
+    },
+    {
+      "epoch": 0.84192,
+      "grad_norm": 2.6820189085108725,
+      "learning_rate": 0.00016891446009128987,
+      "loss": 1.5992,
+      "step": 5262
+    },
+    {
+      "epoch": 0.84208,
+      "grad_norm": 3.1567236129046736,
+      "learning_rate": 0.00016890194202931157,
+      "loss": 1.7338,
+      "step": 5263
+    },
+    {
+      "epoch": 0.84224,
+      "grad_norm": 1.994410039205978,
+      "learning_rate": 0.00016888942191139973,
+      "loss": 1.6313,
+      "step": 5264
+    },
+    {
+      "epoch": 0.8424,
+      "grad_norm": 2.1861949366905122,
+      "learning_rate": 0.0001688768997379279,
+      "loss": 1.5271,
+      "step": 5265
+    },
+    {
+      "epoch": 0.84256,
+      "grad_norm": 2.6737942828119805,
+      "learning_rate": 0.00016886437550926977,
+      "loss": 1.584,
+      "step": 5266
+    },
+    {
+      "epoch": 0.84272,
+      "grad_norm": 2.325261365712011,
+      "learning_rate": 0.00016885184922579898,
+      "loss": 1.7119,
+      "step": 5267
+    },
+    {
+      "epoch": 0.84288,
+      "grad_norm": 2.065623136024515,
+      "learning_rate": 0.00016883932088788937,
+      "loss": 1.5022,
+      "step": 5268
+    },
+    {
+      "epoch": 0.84304,
+      "grad_norm": 2.3725710371284467,
+      "learning_rate": 0.00016882679049591472,
+      "loss": 1.6109,
+      "step": 5269
+    },
+    {
+      "epoch": 0.8432,
+      "grad_norm": 2.31983678530963,
+      "learning_rate": 0.00016881425805024893,
+      "loss": 1.6808,
+      "step": 5270
+    },
+    {
+      "epoch": 0.84336,
+      "grad_norm": 2.1771112966539796,
+      "learning_rate": 0.00016880172355126597,
+      "loss": 1.5122,
+      "step": 5271
+    },
+    {
+      "epoch": 0.84352,
+      "grad_norm": 2.6866769287687795,
+      "learning_rate": 0.00016878918699933984,
+      "loss": 1.6777,
+      "step": 5272
+    },
+    {
+      "epoch": 0.84368,
+      "grad_norm": 2.483316695369138,
+      "learning_rate": 0.00016877664839484458,
+      "loss": 1.6125,
+      "step": 5273
+    },
+    {
+      "epoch": 0.84384,
+      "grad_norm": 2.6266111415754083,
+      "learning_rate": 0.00016876410773815438,
+      "loss": 1.6701,
+      "step": 5274
+    },
+    {
+      "epoch": 0.844,
+      "grad_norm": 2.8747246353008755,
+      "learning_rate": 0.00016875156502964335,
+      "loss": 1.6412,
+      "step": 5275
+    },
+    {
+      "epoch": 0.84416,
+      "grad_norm": 2.109588096764879,
+      "learning_rate": 0.00016873902026968586,
+      "loss": 1.6445,
+      "step": 5276
+    },
+    {
+      "epoch": 0.84432,
+      "grad_norm": 2.809210308975478,
+      "learning_rate": 0.00016872647345865613,
+      "loss": 1.6074,
+      "step": 5277
+    },
+    {
+      "epoch": 0.84448,
+      "grad_norm": 2.816876011615814,
+      "learning_rate": 0.00016871392459692858,
+      "loss": 1.5647,
+      "step": 5278
+    },
+    {
+      "epoch": 0.84464,
+      "grad_norm": 2.1693923643572357,
+      "learning_rate": 0.00016870137368487766,
+      "loss": 1.5969,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8448,
+      "grad_norm": 3.0614085012745456,
+      "learning_rate": 0.00016868882072287788,
+      "loss": 1.6445,
+      "step": 5280
+    },
+    {
+      "epoch": 0.84496,
+      "grad_norm": 2.1284867395369216,
+      "learning_rate": 0.00016867626571130374,
+      "loss": 1.6735,
+      "step": 5281
+    },
+    {
+      "epoch": 0.84512,
+      "grad_norm": 2.482211351131838,
+      "learning_rate": 0.0001686637086505299,
+      "loss": 1.6408,
+      "step": 5282
+    },
+    {
+      "epoch": 0.84528,
+      "grad_norm": 2.614499807312381,
+      "learning_rate": 0.0001686511495409311,
+      "loss": 1.697,
+      "step": 5283
+    },
+    {
+      "epoch": 0.84544,
+      "grad_norm": 2.275463320208867,
+      "learning_rate": 0.000168638588382882,
+      "loss": 1.5506,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8456,
+      "grad_norm": 3.0492860685279415,
+      "learning_rate": 0.00016862602517675744,
+      "loss": 1.734,
+      "step": 5285
+    },
+    {
+      "epoch": 0.84576,
+      "grad_norm": 2.0487737238173658,
+      "learning_rate": 0.00016861345992293228,
+      "loss": 1.6427,
+      "step": 5286
+    },
+    {
+      "epoch": 0.84592,
+      "grad_norm": 2.1702918982779407,
+      "learning_rate": 0.00016860089262178148,
+      "loss": 1.5906,
+      "step": 5287
+    },
+    {
+      "epoch": 0.84608,
+      "grad_norm": 2.148388673257317,
+      "learning_rate": 0.00016858832327367997,
+      "loss": 1.58,
+      "step": 5288
+    },
+    {
+      "epoch": 0.84624,
+      "grad_norm": 2.225495173761792,
+      "learning_rate": 0.00016857575187900287,
+      "loss": 1.6854,
+      "step": 5289
+    },
+    {
+      "epoch": 0.8464,
+      "grad_norm": 2.261545327664942,
+      "learning_rate": 0.00016856317843812524,
+      "loss": 1.606,
+      "step": 5290
+    },
+    {
+      "epoch": 0.84656,
+      "grad_norm": 2.074909422332182,
+      "learning_rate": 0.00016855060295142227,
+      "loss": 1.5603,
+      "step": 5291
+    },
+    {
+      "epoch": 0.84672,
+      "grad_norm": 2.359817166672592,
+      "learning_rate": 0.0001685380254192692,
+      "loss": 1.8034,
+      "step": 5292
+    },
+    {
+      "epoch": 0.84688,
+      "grad_norm": 2.090235818980222,
+      "learning_rate": 0.00016852544584204136,
+      "loss": 1.5898,
+      "step": 5293
+    },
+    {
+      "epoch": 0.84704,
+      "grad_norm": 2.2435389970511435,
+      "learning_rate": 0.00016851286422011404,
+      "loss": 1.6571,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8472,
+      "grad_norm": 2.446792077412395,
+      "learning_rate": 0.00016850028055386267,
+      "loss": 1.6373,
+      "step": 5295
+    },
+    {
+      "epoch": 0.84736,
+      "grad_norm": 1.917470662205096,
+      "learning_rate": 0.00016848769484366277,
+      "loss": 1.4965,
+      "step": 5296
+    },
+    {
+      "epoch": 0.84752,
+      "grad_norm": 2.2688526460873852,
+      "learning_rate": 0.00016847510708988981,
+      "loss": 1.6645,
+      "step": 5297
+    },
+    {
+      "epoch": 0.84768,
+      "grad_norm": 2.1939429006670044,
+      "learning_rate": 0.00016846251729291947,
+      "loss": 1.682,
+      "step": 5298
+    },
+    {
+      "epoch": 0.84784,
+      "grad_norm": 2.6043368117261516,
+      "learning_rate": 0.00016844992545312737,
+      "loss": 1.5796,
+      "step": 5299
+    },
+    {
+      "epoch": 0.848,
+      "grad_norm": 2.5552093883442373,
+      "learning_rate": 0.00016843733157088922,
+      "loss": 1.5107,
+      "step": 5300
+    },
+    {
+      "epoch": 0.84816,
+      "grad_norm": 2.3421140999131533,
+      "learning_rate": 0.00016842473564658087,
+      "loss": 1.6836,
+      "step": 5301
+    },
+    {
+      "epoch": 0.84832,
+      "grad_norm": 2.206685857336457,
+      "learning_rate": 0.00016841213768057807,
+      "loss": 1.6026,
+      "step": 5302
+    },
+    {
+      "epoch": 0.84848,
+      "grad_norm": 2.120495883719221,
+      "learning_rate": 0.00016839953767325676,
+      "loss": 1.6995,
+      "step": 5303
+    },
+    {
+      "epoch": 0.84864,
+      "grad_norm": 2.538888336779819,
+      "learning_rate": 0.00016838693562499292,
+      "loss": 1.6193,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8488,
+      "grad_norm": 2.277301895665782,
+      "learning_rate": 0.0001683743315361626,
+      "loss": 1.6264,
+      "step": 5305
+    },
+    {
+      "epoch": 0.84896,
+      "grad_norm": 2.219504325369114,
+      "learning_rate": 0.0001683617254071418,
+      "loss": 1.5792,
+      "step": 5306
+    },
+    {
+      "epoch": 0.84912,
+      "grad_norm": 2.3322751521564937,
+      "learning_rate": 0.0001683491172383068,
+      "loss": 1.5712,
+      "step": 5307
+    },
+    {
+      "epoch": 0.84928,
+      "grad_norm": 2.047614685235662,
+      "learning_rate": 0.0001683365070300337,
+      "loss": 1.6251,
+      "step": 5308
+    },
+    {
+      "epoch": 0.84944,
+      "grad_norm": 2.0216078383640936,
+      "learning_rate": 0.00016832389478269882,
+      "loss": 1.5565,
+      "step": 5309
+    },
+    {
+      "epoch": 0.8496,
+      "grad_norm": 2.8891559848083186,
+      "learning_rate": 0.00016831128049667846,
+      "loss": 1.6607,
+      "step": 5310
+    },
+    {
+      "epoch": 0.84976,
+      "grad_norm": 2.31320817961106,
+      "learning_rate": 0.00016829866417234906,
+      "loss": 1.6407,
+      "step": 5311
+    },
+    {
+      "epoch": 0.84992,
+      "grad_norm": 2.149524906218383,
+      "learning_rate": 0.000168286045810087,
+      "loss": 1.6818,
+      "step": 5312
+    },
+    {
+      "epoch": 0.85008,
+      "grad_norm": 2.064093189132955,
+      "learning_rate": 0.00016827342541026887,
+      "loss": 1.5604,
+      "step": 5313
+    },
+    {
+      "epoch": 0.85024,
+      "grad_norm": 2.359944769246429,
+      "learning_rate": 0.0001682608029732712,
+      "loss": 1.6519,
+      "step": 5314
+    },
+    {
+      "epoch": 0.8504,
+      "grad_norm": 2.276869275217186,
+      "learning_rate": 0.00016824817849947062,
+      "loss": 1.6883,
+      "step": 5315
+    },
+    {
+      "epoch": 0.85056,
+      "grad_norm": 2.373263048122089,
+      "learning_rate": 0.00016823555198924387,
+      "loss": 1.5195,
+      "step": 5316
+    },
+    {
+      "epoch": 0.85072,
+      "grad_norm": 2.2606785217882033,
+      "learning_rate": 0.00016822292344296767,
+      "loss": 1.5975,
+      "step": 5317
+    },
+    {
+      "epoch": 0.85088,
+      "grad_norm": 2.921735468579773,
+      "learning_rate": 0.00016821029286101885,
+      "loss": 1.6646,
+      "step": 5318
+    },
+    {
+      "epoch": 0.85104,
+      "grad_norm": 2.0497529034862567,
+      "learning_rate": 0.00016819766024377426,
+      "loss": 1.5641,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8512,
+      "grad_norm": 2.2749335479876014,
+      "learning_rate": 0.00016818502559161087,
+      "loss": 1.5748,
+      "step": 5320
+    },
+    {
+      "epoch": 0.85136,
+      "grad_norm": 2.291981667541761,
+      "learning_rate": 0.00016817238890490566,
+      "loss": 1.551,
+      "step": 5321
+    },
+    {
+      "epoch": 0.85152,
+      "grad_norm": 2.486447865225999,
+      "learning_rate": 0.00016815975018403574,
+      "loss": 1.5362,
+      "step": 5322
+    },
+    {
+      "epoch": 0.85168,
+      "grad_norm": 2.010785527273751,
+      "learning_rate": 0.00016814710942937816,
+      "loss": 1.6071,
+      "step": 5323
+    },
+    {
+      "epoch": 0.85184,
+      "grad_norm": 2.1554649579028236,
+      "learning_rate": 0.0001681344666413102,
+      "loss": 1.656,
+      "step": 5324
+    },
+    {
+      "epoch": 0.852,
+      "grad_norm": 2.165458638314646,
+      "learning_rate": 0.00016812182182020894,
+      "loss": 1.5846,
+      "step": 5325
+    },
+    {
+      "epoch": 0.85216,
+      "grad_norm": 1.9691525926321594,
+      "learning_rate": 0.00016810917496645184,
+      "loss": 1.5861,
+      "step": 5326
+    },
+    {
+      "epoch": 0.85232,
+      "grad_norm": 2.4913700381983945,
+      "learning_rate": 0.0001680965260804162,
+      "loss": 1.5582,
+      "step": 5327
+    },
+    {
+      "epoch": 0.85248,
+      "grad_norm": 2.5594625542554783,
+      "learning_rate": 0.00016808387516247946,
+      "loss": 1.5454,
+      "step": 5328
+    },
+    {
+      "epoch": 0.85264,
+      "grad_norm": 2.271862789404084,
+      "learning_rate": 0.00016807122221301906,
+      "loss": 1.5898,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8528,
+      "grad_norm": 2.3702167125561293,
+      "learning_rate": 0.0001680585672324126,
+      "loss": 1.5523,
+      "step": 5330
+    },
+    {
+      "epoch": 0.85296,
+      "grad_norm": 2.2902679445900898,
+      "learning_rate": 0.00016804591022103767,
+      "loss": 1.6269,
+      "step": 5331
+    },
+    {
+      "epoch": 0.85312,
+      "grad_norm": 2.2603126093542074,
+      "learning_rate": 0.00016803325117927193,
+      "loss": 1.5719,
+      "step": 5332
+    },
+    {
+      "epoch": 0.85328,
+      "grad_norm": 2.310526516220869,
+      "learning_rate": 0.00016802059010749307,
+      "loss": 1.5613,
+      "step": 5333
+    },
+    {
+      "epoch": 0.85344,
+      "grad_norm": 1.9904421581118712,
+      "learning_rate": 0.00016800792700607896,
+      "loss": 1.5915,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8536,
+      "grad_norm": 2.8209404937806206,
+      "learning_rate": 0.0001679952618754074,
+      "loss": 1.647,
+      "step": 5335
+    },
+    {
+      "epoch": 0.85376,
+      "grad_norm": 1.996160482947565,
+      "learning_rate": 0.00016798259471585632,
+      "loss": 1.731,
+      "step": 5336
+    },
+    {
+      "epoch": 0.85392,
+      "grad_norm": 3.0195107395325564,
+      "learning_rate": 0.00016796992552780364,
+      "loss": 1.604,
+      "step": 5337
+    },
+    {
+      "epoch": 0.85408,
+      "grad_norm": 2.188423601456657,
+      "learning_rate": 0.00016795725431162747,
+      "loss": 1.6231,
+      "step": 5338
+    },
+    {
+      "epoch": 0.85424,
+      "grad_norm": 2.8717059067326227,
+      "learning_rate": 0.00016794458106770578,
+      "loss": 1.6158,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8544,
+      "grad_norm": 2.045501170674159,
+      "learning_rate": 0.00016793190579641686,
+      "loss": 1.6248,
+      "step": 5340
+    },
+    {
+      "epoch": 0.85456,
+      "grad_norm": 2.3914673702861475,
+      "learning_rate": 0.00016791922849813883,
+      "loss": 1.5404,
+      "step": 5341
+    },
+    {
+      "epoch": 0.85472,
+      "grad_norm": 2.484894496588511,
+      "learning_rate": 0.00016790654917325,
+      "loss": 1.6872,
+      "step": 5342
+    },
+    {
+      "epoch": 0.85488,
+      "grad_norm": 2.1699783670547226,
+      "learning_rate": 0.00016789386782212873,
+      "loss": 1.6123,
+      "step": 5343
+    },
+    {
+      "epoch": 0.85504,
+      "grad_norm": 2.22327670937196,
+      "learning_rate": 0.00016788118444515334,
+      "loss": 1.6183,
+      "step": 5344
+    },
+    {
+      "epoch": 0.8552,
+      "grad_norm": 2.4841149303619154,
+      "learning_rate": 0.00016786849904270232,
+      "loss": 1.6096,
+      "step": 5345
+    },
+    {
+      "epoch": 0.85536,
+      "grad_norm": 2.234371870532524,
+      "learning_rate": 0.0001678558116151542,
+      "loss": 1.6734,
+      "step": 5346
+    },
+    {
+      "epoch": 0.85552,
+      "grad_norm": 2.807669586781103,
+      "learning_rate": 0.0001678431221628875,
+      "loss": 1.5274,
+      "step": 5347
+    },
+    {
+      "epoch": 0.85568,
+      "grad_norm": 1.9637319718276205,
+      "learning_rate": 0.00016783043068628093,
+      "loss": 1.6239,
+      "step": 5348
+    },
+    {
+      "epoch": 0.85584,
+      "grad_norm": 2.4779325614734415,
+      "learning_rate": 0.00016781773718571312,
+      "loss": 1.5939,
+      "step": 5349
+    },
+    {
+      "epoch": 0.856,
+      "grad_norm": 2.0451910228687176,
+      "learning_rate": 0.00016780504166156288,
+      "loss": 1.5751,
+      "step": 5350
+    },
+    {
+      "epoch": 0.85616,
+      "grad_norm": 2.2790090917427124,
+      "learning_rate": 0.000167792344114209,
+      "loss": 1.5726,
+      "step": 5351
+    },
+    {
+      "epoch": 0.85632,
+      "grad_norm": 2.704621324955993,
+      "learning_rate": 0.00016777964454403032,
+      "loss": 1.6299,
+      "step": 5352
+    },
+    {
+      "epoch": 0.85648,
+      "grad_norm": 2.133032726550781,
+      "learning_rate": 0.00016776694295140586,
+      "loss": 1.5541,
+      "step": 5353
+    },
+    {
+      "epoch": 0.85664,
+      "grad_norm": 2.5697849979602307,
+      "learning_rate": 0.00016775423933671453,
+      "loss": 1.6899,
+      "step": 5354
+    },
+    {
+      "epoch": 0.8568,
+      "grad_norm": 2.3097612884838385,
+      "learning_rate": 0.00016774153370033542,
+      "loss": 1.6578,
+      "step": 5355
+    },
+    {
+      "epoch": 0.85696,
+      "grad_norm": 2.45033113647219,
+      "learning_rate": 0.00016772882604264769,
+      "loss": 1.5908,
+      "step": 5356
+    },
+    {
+      "epoch": 0.85712,
+      "grad_norm": 2.1429116300234536,
+      "learning_rate": 0.00016771611636403046,
+      "loss": 1.6363,
+      "step": 5357
+    },
+    {
+      "epoch": 0.85728,
+      "grad_norm": 2.2434669851574123,
+      "learning_rate": 0.000167703404664863,
+      "loss": 1.6403,
+      "step": 5358
+    },
+    {
+      "epoch": 0.85744,
+      "grad_norm": 2.2710754579533705,
+      "learning_rate": 0.0001676906909455246,
+      "loss": 1.5599,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8576,
+      "grad_norm": 2.424919408460148,
+      "learning_rate": 0.00016767797520639462,
+      "loss": 1.6821,
+      "step": 5360
+    },
+    {
+      "epoch": 0.85776,
+      "grad_norm": 2.6870874139038183,
+      "learning_rate": 0.00016766525744785245,
+      "loss": 1.5701,
+      "step": 5361
+    },
+    {
+      "epoch": 0.85792,
+      "grad_norm": 2.0780736572678116,
+      "learning_rate": 0.00016765253767027764,
+      "loss": 1.5384,
+      "step": 5362
+    },
+    {
+      "epoch": 0.85808,
+      "grad_norm": 2.4120958915108686,
+      "learning_rate": 0.00016763981587404966,
+      "loss": 1.6634,
+      "step": 5363
+    },
+    {
+      "epoch": 0.85824,
+      "grad_norm": 1.963067130911596,
+      "learning_rate": 0.00016762709205954814,
+      "loss": 1.5971,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8584,
+      "grad_norm": 2.063333383339477,
+      "learning_rate": 0.00016761436622715273,
+      "loss": 1.6994,
+      "step": 5365
+    },
+    {
+      "epoch": 0.85856,
+      "grad_norm": 2.657377938586446,
+      "learning_rate": 0.00016760163837724315,
+      "loss": 1.6365,
+      "step": 5366
+    },
+    {
+      "epoch": 0.85872,
+      "grad_norm": 2.3743393674970577,
+      "learning_rate": 0.00016758890851019918,
+      "loss": 1.6213,
+      "step": 5367
+    },
+    {
+      "epoch": 0.85888,
+      "grad_norm": 2.036587654448071,
+      "learning_rate": 0.00016757617662640072,
+      "loss": 1.6485,
+      "step": 5368
+    },
+    {
+      "epoch": 0.85904,
+      "grad_norm": 2.639944945773534,
+      "learning_rate": 0.00016756344272622758,
+      "loss": 1.6802,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8592,
+      "grad_norm": 2.2716259529779355,
+      "learning_rate": 0.0001675507068100598,
+      "loss": 1.6938,
+      "step": 5370
+    },
+    {
+      "epoch": 0.85936,
+      "grad_norm": 2.2663797606842824,
+      "learning_rate": 0.0001675379688782773,
+      "loss": 1.5932,
+      "step": 5371
+    },
+    {
+      "epoch": 0.85952,
+      "grad_norm": 2.4115591191841776,
+      "learning_rate": 0.00016752522893126023,
+      "loss": 1.5947,
+      "step": 5372
+    },
+    {
+      "epoch": 0.85968,
+      "grad_norm": 2.2270809938175042,
+      "learning_rate": 0.00016751248696938875,
+      "loss": 1.5745,
+      "step": 5373
+    },
+    {
+      "epoch": 0.85984,
+      "grad_norm": 2.551662509029407,
+      "learning_rate": 0.00016749974299304303,
+      "loss": 1.669,
+      "step": 5374
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 2.1466167182882634,
+      "learning_rate": 0.0001674869970026033,
+      "loss": 1.5956,
+      "step": 5375
+    },
+    {
+      "epoch": 0.86016,
+      "grad_norm": 2.35724739193287,
+      "learning_rate": 0.00016747424899844995,
+      "loss": 1.5874,
+      "step": 5376
+    },
+    {
+      "epoch": 0.86032,
+      "grad_norm": 2.2976175950498714,
+      "learning_rate": 0.00016746149898096336,
+      "loss": 1.5959,
+      "step": 5377
+    },
+    {
+      "epoch": 0.86048,
+      "grad_norm": 2.188284641790266,
+      "learning_rate": 0.0001674487469505239,
+      "loss": 1.6296,
+      "step": 5378
+    },
+    {
+      "epoch": 0.86064,
+      "grad_norm": 2.1457274370906685,
+      "learning_rate": 0.00016743599290751214,
+      "loss": 1.5861,
+      "step": 5379
+    },
+    {
+      "epoch": 0.8608,
+      "grad_norm": 2.368883107514302,
+      "learning_rate": 0.00016742323685230859,
+      "loss": 1.6699,
+      "step": 5380
+    },
+    {
+      "epoch": 0.86096,
+      "grad_norm": 2.306194274220493,
+      "learning_rate": 0.00016741047878529393,
+      "loss": 1.5945,
+      "step": 5381
+    },
+    {
+      "epoch": 0.86112,
+      "grad_norm": 2.173477708543016,
+      "learning_rate": 0.00016739771870684876,
+      "loss": 1.614,
+      "step": 5382
+    },
+    {
+      "epoch": 0.86128,
+      "grad_norm": 2.0367657026170414,
+      "learning_rate": 0.00016738495661735393,
+      "loss": 1.512,
+      "step": 5383
+    },
+    {
+      "epoch": 0.86144,
+      "grad_norm": 2.6933905840532377,
+      "learning_rate": 0.00016737219251719015,
+      "loss": 1.6079,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8616,
+      "grad_norm": 2.676451189863215,
+      "learning_rate": 0.00016735942640673832,
+      "loss": 1.5797,
+      "step": 5385
+    },
+    {
+      "epoch": 0.86176,
+      "grad_norm": 2.106849357290249,
+      "learning_rate": 0.00016734665828637937,
+      "loss": 1.6848,
+      "step": 5386
+    },
+    {
+      "epoch": 0.86192,
+      "grad_norm": 2.755090864122201,
+      "learning_rate": 0.00016733388815649426,
+      "loss": 1.5972,
+      "step": 5387
+    },
+    {
+      "epoch": 0.86208,
+      "grad_norm": 2.237846773032639,
+      "learning_rate": 0.00016732111601746406,
+      "loss": 1.6532,
+      "step": 5388
+    },
+    {
+      "epoch": 0.86224,
+      "grad_norm": 2.2883048541626163,
+      "learning_rate": 0.00016730834186966983,
+      "loss": 1.6542,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8624,
+      "grad_norm": 2.48739014243049,
+      "learning_rate": 0.0001672955657134928,
+      "loss": 1.5943,
+      "step": 5390
+    },
+    {
+      "epoch": 0.86256,
+      "grad_norm": 2.325483485578436,
+      "learning_rate": 0.0001672827875493141,
+      "loss": 1.5391,
+      "step": 5391
+    },
+    {
+      "epoch": 0.86272,
+      "grad_norm": 2.2078065743712942,
+      "learning_rate": 0.0001672700073775151,
+      "loss": 1.6509,
+      "step": 5392
+    },
+    {
+      "epoch": 0.86288,
+      "grad_norm": 2.7734948769468017,
+      "learning_rate": 0.00016725722519847706,
+      "loss": 1.6187,
+      "step": 5393
+    },
+    {
+      "epoch": 0.86304,
+      "grad_norm": 2.375695944936667,
+      "learning_rate": 0.00016724444101258147,
+      "loss": 1.5845,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8632,
+      "grad_norm": 2.2731242425424254,
+      "learning_rate": 0.0001672316548202097,
+      "loss": 1.5535,
+      "step": 5395
+    },
+    {
+      "epoch": 0.86336,
+      "grad_norm": 3.0080129607315524,
+      "learning_rate": 0.00016721886662174335,
+      "loss": 1.6128,
+      "step": 5396
+    },
+    {
+      "epoch": 0.86352,
+      "grad_norm": 2.637763437599702,
+      "learning_rate": 0.00016720607641756394,
+      "loss": 1.6521,
+      "step": 5397
+    },
+    {
+      "epoch": 0.86368,
+      "grad_norm": 2.1519776023829253,
+      "learning_rate": 0.00016719328420805317,
+      "loss": 1.6582,
+      "step": 5398
+    },
+    {
+      "epoch": 0.86384,
+      "grad_norm": 2.2187209646613204,
+      "learning_rate": 0.0001671804899935927,
+      "loss": 1.6727,
+      "step": 5399
+    },
+    {
+      "epoch": 0.864,
+      "grad_norm": 2.0643854660147505,
+      "learning_rate": 0.0001671676937745643,
+      "loss": 1.5835,
+      "step": 5400
+    },
+    {
+      "epoch": 0.86416,
+      "grad_norm": 2.16045678768387,
+      "learning_rate": 0.00016715489555134977,
+      "loss": 1.6541,
+      "step": 5401
+    },
+    {
+      "epoch": 0.86432,
+      "grad_norm": 2.1994528008170393,
+      "learning_rate": 0.00016714209532433104,
+      "loss": 1.5735,
+      "step": 5402
+    },
+    {
+      "epoch": 0.86448,
+      "grad_norm": 2.3645304629286485,
+      "learning_rate": 0.00016712929309389005,
+      "loss": 1.6792,
+      "step": 5403
+    },
+    {
+      "epoch": 0.86464,
+      "grad_norm": 3.047024905069727,
+      "learning_rate": 0.00016711648886040873,
+      "loss": 1.5978,
+      "step": 5404
+    },
+    {
+      "epoch": 0.8648,
+      "grad_norm": 2.463087792841452,
+      "learning_rate": 0.00016710368262426917,
+      "loss": 1.5082,
+      "step": 5405
+    },
+    {
+      "epoch": 0.86496,
+      "grad_norm": 2.190971251712137,
+      "learning_rate": 0.00016709087438585352,
+      "loss": 1.5443,
+      "step": 5406
+    },
+    {
+      "epoch": 0.86512,
+      "grad_norm": 2.2301449357836147,
+      "learning_rate": 0.00016707806414554394,
+      "loss": 1.6065,
+      "step": 5407
+    },
+    {
+      "epoch": 0.86528,
+      "grad_norm": 2.229293277013421,
+      "learning_rate": 0.00016706525190372267,
+      "loss": 1.5954,
+      "step": 5408
+    },
+    {
+      "epoch": 0.86544,
+      "grad_norm": 2.421760662073147,
+      "learning_rate": 0.00016705243766077202,
+      "loss": 1.5925,
+      "step": 5409
+    },
+    {
+      "epoch": 0.8656,
+      "grad_norm": 2.3757717786839936,
+      "learning_rate": 0.00016703962141707433,
+      "loss": 1.5823,
+      "step": 5410
+    },
+    {
+      "epoch": 0.86576,
+      "grad_norm": 2.5454704133629282,
+      "learning_rate": 0.000167026803173012,
+      "loss": 1.6689,
+      "step": 5411
+    },
+    {
+      "epoch": 0.86592,
+      "grad_norm": 3.1382685070734557,
+      "learning_rate": 0.00016701398292896756,
+      "loss": 1.6339,
+      "step": 5412
+    },
+    {
+      "epoch": 0.86608,
+      "grad_norm": 1.947146290079529,
+      "learning_rate": 0.0001670011606853235,
+      "loss": 1.5578,
+      "step": 5413
+    },
+    {
+      "epoch": 0.86624,
+      "grad_norm": 2.432479209912977,
+      "learning_rate": 0.00016698833644246243,
+      "loss": 1.5981,
+      "step": 5414
+    },
+    {
+      "epoch": 0.8664,
+      "grad_norm": 2.0831620483111384,
+      "learning_rate": 0.00016697551020076703,
+      "loss": 1.6423,
+      "step": 5415
+    },
+    {
+      "epoch": 0.86656,
+      "grad_norm": 2.020920515281718,
+      "learning_rate": 0.00016696268196062,
+      "loss": 1.5678,
+      "step": 5416
+    },
+    {
+      "epoch": 0.86672,
+      "grad_norm": 2.6711624044597424,
+      "learning_rate": 0.0001669498517224041,
+      "loss": 1.5719,
+      "step": 5417
+    },
+    {
+      "epoch": 0.86688,
+      "grad_norm": 2.197770808610881,
+      "learning_rate": 0.0001669370194865022,
+      "loss": 1.6711,
+      "step": 5418
+    },
+    {
+      "epoch": 0.86704,
+      "grad_norm": 2.0706626054278803,
+      "learning_rate": 0.00016692418525329718,
+      "loss": 1.5632,
+      "step": 5419
+    },
+    {
+      "epoch": 0.8672,
+      "grad_norm": 2.340031400460364,
+      "learning_rate": 0.00016691134902317196,
+      "loss": 1.5737,
+      "step": 5420
+    },
+    {
+      "epoch": 0.86736,
+      "grad_norm": 2.351089950424219,
+      "learning_rate": 0.00016689851079650962,
+      "loss": 1.6418,
+      "step": 5421
+    },
+    {
+      "epoch": 0.86752,
+      "grad_norm": 2.5030928878362944,
+      "learning_rate": 0.00016688567057369314,
+      "loss": 1.5935,
+      "step": 5422
+    },
+    {
+      "epoch": 0.86768,
+      "grad_norm": 2.1274292792147422,
+      "learning_rate": 0.0001668728283551058,
+      "loss": 1.6548,
+      "step": 5423
+    },
+    {
+      "epoch": 0.86784,
+      "grad_norm": 2.0207720681270454,
+      "learning_rate": 0.00016685998414113065,
+      "loss": 1.4849,
+      "step": 5424
+    },
+    {
+      "epoch": 0.868,
+      "grad_norm": 2.2785659341655067,
+      "learning_rate": 0.00016684713793215102,
+      "loss": 1.6945,
+      "step": 5425
+    },
+    {
+      "epoch": 0.86816,
+      "grad_norm": 2.0639522881562544,
+      "learning_rate": 0.0001668342897285502,
+      "loss": 1.5757,
+      "step": 5426
+    },
+    {
+      "epoch": 0.86832,
+      "grad_norm": 2.009087734253604,
+      "learning_rate": 0.00016682143953071154,
+      "loss": 1.6513,
+      "step": 5427
+    },
+    {
+      "epoch": 0.86848,
+      "grad_norm": 2.0154335505929506,
+      "learning_rate": 0.0001668085873390185,
+      "loss": 1.5661,
+      "step": 5428
+    },
+    {
+      "epoch": 0.86864,
+      "grad_norm": 2.529041123404973,
+      "learning_rate": 0.00016679573315385454,
+      "loss": 1.5955,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8688,
+      "grad_norm": 3.0936861198874763,
+      "learning_rate": 0.00016678287697560328,
+      "loss": 1.6745,
+      "step": 5430
+    },
+    {
+      "epoch": 0.86896,
+      "grad_norm": 2.2698985082224405,
+      "learning_rate": 0.0001667700188046483,
+      "loss": 1.6102,
+      "step": 5431
+    },
+    {
+      "epoch": 0.86912,
+      "grad_norm": 2.5427803496886656,
+      "learning_rate": 0.0001667571586413732,
+      "loss": 1.6626,
+      "step": 5432
+    },
+    {
+      "epoch": 0.86928,
+      "grad_norm": 2.065175845831632,
+      "learning_rate": 0.0001667442964861618,
+      "loss": 1.678,
+      "step": 5433
+    },
+    {
+      "epoch": 0.86944,
+      "grad_norm": 3.654980179498889,
+      "learning_rate": 0.0001667314323393978,
+      "loss": 1.5223,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8696,
+      "grad_norm": 2.3955121671341195,
+      "learning_rate": 0.00016671856620146516,
+      "loss": 1.6997,
+      "step": 5435
+    },
+    {
+      "epoch": 0.86976,
+      "grad_norm": 3.110086277805187,
+      "learning_rate": 0.00016670569807274768,
+      "loss": 1.5298,
+      "step": 5436
+    },
+    {
+      "epoch": 0.86992,
+      "grad_norm": 2.3747934398950274,
+      "learning_rate": 0.00016669282795362937,
+      "loss": 1.6132,
+      "step": 5437
+    },
+    {
+      "epoch": 0.87008,
+      "grad_norm": 2.6355008775944104,
+      "learning_rate": 0.0001666799558444943,
+      "loss": 1.5653,
+      "step": 5438
+    },
+    {
+      "epoch": 0.87024,
+      "grad_norm": 2.2987323168928206,
+      "learning_rate": 0.00016666708174572646,
+      "loss": 1.6364,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8704,
+      "grad_norm": 2.786262661441503,
+      "learning_rate": 0.00016665420565771007,
+      "loss": 1.6059,
+      "step": 5440
+    },
+    {
+      "epoch": 0.87056,
+      "grad_norm": 2.1033148042664447,
+      "learning_rate": 0.0001666413275808293,
+      "loss": 1.6018,
+      "step": 5441
+    },
+    {
+      "epoch": 0.87072,
+      "grad_norm": 2.455871499739894,
+      "learning_rate": 0.0001666284475154684,
+      "loss": 1.6893,
+      "step": 5442
+    },
+    {
+      "epoch": 0.87088,
+      "grad_norm": 2.6160697882916306,
+      "learning_rate": 0.00016661556546201174,
+      "loss": 1.6377,
+      "step": 5443
+    },
+    {
+      "epoch": 0.87104,
+      "grad_norm": 2.230893892254758,
+      "learning_rate": 0.0001666026814208437,
+      "loss": 1.6455,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8712,
+      "grad_norm": 2.9445599994108487,
+      "learning_rate": 0.00016658979539234866,
+      "loss": 1.6859,
+      "step": 5445
+    },
+    {
+      "epoch": 0.87136,
+      "grad_norm": 2.3891086986758987,
+      "learning_rate": 0.0001665769073769112,
+      "loss": 1.5645,
+      "step": 5446
+    },
+    {
+      "epoch": 0.87152,
+      "grad_norm": 2.3540339112545507,
+      "learning_rate": 0.0001665640173749158,
+      "loss": 1.5104,
+      "step": 5447
+    },
+    {
+      "epoch": 0.87168,
+      "grad_norm": 2.3024001840327015,
+      "learning_rate": 0.0001665511253867471,
+      "loss": 1.5774,
+      "step": 5448
+    },
+    {
+      "epoch": 0.87184,
+      "grad_norm": 2.4269924579623994,
+      "learning_rate": 0.00016653823141278982,
+      "loss": 1.5305,
+      "step": 5449
+    },
+    {
+      "epoch": 0.872,
+      "grad_norm": 2.502100862615296,
+      "learning_rate": 0.00016652533545342868,
+      "loss": 1.5431,
+      "step": 5450
+    },
+    {
+      "epoch": 0.87216,
+      "grad_norm": 2.2880249119346807,
+      "learning_rate": 0.00016651243750904845,
+      "loss": 1.5967,
+      "step": 5451
+    },
+    {
+      "epoch": 0.87232,
+      "grad_norm": 2.338000328988719,
+      "learning_rate": 0.000166499537580034,
+      "loss": 1.6443,
+      "step": 5452
+    },
+    {
+      "epoch": 0.87248,
+      "grad_norm": 2.1045921459046455,
+      "learning_rate": 0.00016648663566677026,
+      "loss": 1.5799,
+      "step": 5453
+    },
+    {
+      "epoch": 0.87264,
+      "grad_norm": 2.0790730022161563,
+      "learning_rate": 0.00016647373176964219,
+      "loss": 1.6908,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8728,
+      "grad_norm": 2.462565193477771,
+      "learning_rate": 0.00016646082588903482,
+      "loss": 1.6298,
+      "step": 5455
+    },
+    {
+      "epoch": 0.87296,
+      "grad_norm": 2.140818587404612,
+      "learning_rate": 0.00016644791802533323,
+      "loss": 1.6232,
+      "step": 5456
+    },
+    {
+      "epoch": 0.87312,
+      "grad_norm": 2.729365008633602,
+      "learning_rate": 0.00016643500817892264,
+      "loss": 1.6206,
+      "step": 5457
+    },
+    {
+      "epoch": 0.87328,
+      "grad_norm": 1.9425054615464334,
+      "learning_rate": 0.00016642209635018816,
+      "loss": 1.6424,
+      "step": 5458
+    },
+    {
+      "epoch": 0.87344,
+      "grad_norm": 2.129409323551611,
+      "learning_rate": 0.00016640918253951517,
+      "loss": 1.5542,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8736,
+      "grad_norm": 2.2684034407908675,
+      "learning_rate": 0.0001663962667472889,
+      "loss": 1.6296,
+      "step": 5460
+    },
+    {
+      "epoch": 0.87376,
+      "grad_norm": 2.294722655163877,
+      "learning_rate": 0.0001663833489738948,
+      "loss": 1.6231,
+      "step": 5461
+    },
+    {
+      "epoch": 0.87392,
+      "grad_norm": 2.6134872643541938,
+      "learning_rate": 0.00016637042921971827,
+      "loss": 1.5637,
+      "step": 5462
+    },
+    {
+      "epoch": 0.87408,
+      "grad_norm": 2.5310331250500586,
+      "learning_rate": 0.00016635750748514487,
+      "loss": 1.6369,
+      "step": 5463
+    },
+    {
+      "epoch": 0.87424,
+      "grad_norm": 2.6509660569086946,
+      "learning_rate": 0.00016634458377056012,
+      "loss": 1.5868,
+      "step": 5464
+    },
+    {
+      "epoch": 0.8744,
+      "grad_norm": 2.842334490507482,
+      "learning_rate": 0.00016633165807634968,
+      "loss": 1.5954,
+      "step": 5465
+    },
+    {
+      "epoch": 0.87456,
+      "grad_norm": 2.4179434119471903,
+      "learning_rate": 0.00016631873040289923,
+      "loss": 1.4945,
+      "step": 5466
+    },
+    {
+      "epoch": 0.87472,
+      "grad_norm": 2.4860912765942733,
+      "learning_rate": 0.00016630580075059447,
+      "loss": 1.5774,
+      "step": 5467
+    },
+    {
+      "epoch": 0.87488,
+      "grad_norm": 2.8652730117643164,
+      "learning_rate": 0.00016629286911982124,
+      "loss": 1.5342,
+      "step": 5468
+    },
+    {
+      "epoch": 0.87504,
+      "grad_norm": 3.238878957742209,
+      "learning_rate": 0.00016627993551096543,
+      "loss": 1.6853,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8752,
+      "grad_norm": 2.069465098984593,
+      "learning_rate": 0.0001662669999244129,
+      "loss": 1.5941,
+      "step": 5470
+    },
+    {
+      "epoch": 0.87536,
+      "grad_norm": 2.5898124722175937,
+      "learning_rate": 0.00016625406236054965,
+      "loss": 1.6936,
+      "step": 5471
+    },
+    {
+      "epoch": 0.87552,
+      "grad_norm": 2.4971960170469716,
+      "learning_rate": 0.0001662411228197617,
+      "loss": 1.7596,
+      "step": 5472
+    },
+    {
+      "epoch": 0.87568,
+      "grad_norm": 2.448267472018756,
+      "learning_rate": 0.0001662281813024352,
+      "loss": 1.6179,
+      "step": 5473
+    },
+    {
+      "epoch": 0.87584,
+      "grad_norm": 2.016820611996509,
+      "learning_rate": 0.00016621523780895628,
+      "loss": 1.59,
+      "step": 5474
+    },
+    {
+      "epoch": 0.876,
+      "grad_norm": 2.226080149378107,
+      "learning_rate": 0.00016620229233971115,
+      "loss": 1.6336,
+      "step": 5475
+    },
+    {
+      "epoch": 0.87616,
+      "grad_norm": 2.237969388813178,
+      "learning_rate": 0.00016618934489508606,
+      "loss": 1.6731,
+      "step": 5476
+    },
+    {
+      "epoch": 0.87632,
+      "grad_norm": 2.3202962803052976,
+      "learning_rate": 0.00016617639547546737,
+      "loss": 1.6503,
+      "step": 5477
+    },
+    {
+      "epoch": 0.87648,
+      "grad_norm": 2.0255532359652184,
+      "learning_rate": 0.00016616344408124147,
+      "loss": 1.495,
+      "step": 5478
+    },
+    {
+      "epoch": 0.87664,
+      "grad_norm": 2.0996681112706703,
+      "learning_rate": 0.00016615049071279483,
+      "loss": 1.6681,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8768,
+      "grad_norm": 2.195881677325629,
+      "learning_rate": 0.0001661375353705139,
+      "loss": 1.6335,
+      "step": 5480
+    },
+    {
+      "epoch": 0.87696,
+      "grad_norm": 2.185633012287947,
+      "learning_rate": 0.00016612457805478532,
+      "loss": 1.6067,
+      "step": 5481
+    },
+    {
+      "epoch": 0.87712,
+      "grad_norm": 2.7130301721265275,
+      "learning_rate": 0.00016611161876599567,
+      "loss": 1.6475,
+      "step": 5482
+    },
+    {
+      "epoch": 0.87728,
+      "grad_norm": 2.2167506348360324,
+      "learning_rate": 0.00016609865750453166,
+      "loss": 1.5997,
+      "step": 5483
+    },
+    {
+      "epoch": 0.87744,
+      "grad_norm": 2.0507929379692116,
+      "learning_rate": 0.00016608569427078003,
+      "loss": 1.6352,
+      "step": 5484
+    },
+    {
+      "epoch": 0.8776,
+      "grad_norm": 2.947593147458215,
+      "learning_rate": 0.00016607272906512757,
+      "loss": 1.5813,
+      "step": 5485
+    },
+    {
+      "epoch": 0.87776,
+      "grad_norm": 3.0486871362219454,
+      "learning_rate": 0.00016605976188796112,
+      "loss": 1.6699,
+      "step": 5486
+    },
+    {
+      "epoch": 0.87792,
+      "grad_norm": 2.3574236350894417,
+      "learning_rate": 0.0001660467927396677,
+      "loss": 1.5593,
+      "step": 5487
+    },
+    {
+      "epoch": 0.87808,
+      "grad_norm": 2.444376039749942,
+      "learning_rate": 0.0001660338216206342,
+      "loss": 1.5594,
+      "step": 5488
+    },
+    {
+      "epoch": 0.87824,
+      "grad_norm": 2.7482135362569666,
+      "learning_rate": 0.00016602084853124764,
+      "loss": 1.6536,
+      "step": 5489
+    },
+    {
+      "epoch": 0.8784,
+      "grad_norm": 2.471595822059403,
+      "learning_rate": 0.0001660078734718952,
+      "loss": 1.5933,
+      "step": 5490
+    },
+    {
+      "epoch": 0.87856,
+      "grad_norm": 2.451313323236936,
+      "learning_rate": 0.00016599489644296403,
+      "loss": 1.5729,
+      "step": 5491
+    },
+    {
+      "epoch": 0.87872,
+      "grad_norm": 2.479889676175891,
+      "learning_rate": 0.00016598191744484128,
+      "loss": 1.5704,
+      "step": 5492
+    },
+    {
+      "epoch": 0.87888,
+      "grad_norm": 2.132037984913436,
+      "learning_rate": 0.00016596893647791424,
+      "loss": 1.6488,
+      "step": 5493
+    },
+    {
+      "epoch": 0.87904,
+      "grad_norm": 2.4843267761762178,
+      "learning_rate": 0.0001659559535425703,
+      "loss": 1.5942,
+      "step": 5494
+    },
+    {
+      "epoch": 0.8792,
+      "grad_norm": 2.5827209354474943,
+      "learning_rate": 0.0001659429686391968,
+      "loss": 1.6437,
+      "step": 5495
+    },
+    {
+      "epoch": 0.87936,
+      "grad_norm": 2.1002797970208777,
+      "learning_rate": 0.0001659299817681812,
+      "loss": 1.5602,
+      "step": 5496
+    },
+    {
+      "epoch": 0.87952,
+      "grad_norm": 3.7962863454704987,
+      "learning_rate": 0.000165916992929911,
+      "loss": 1.6146,
+      "step": 5497
+    },
+    {
+      "epoch": 0.87968,
+      "grad_norm": 2.4564214033308205,
+      "learning_rate": 0.00016590400212477383,
+      "loss": 1.6381,
+      "step": 5498
+    },
+    {
+      "epoch": 0.87984,
+      "grad_norm": 2.6850391065698895,
+      "learning_rate": 0.00016589100935315723,
+      "loss": 1.5981,
+      "step": 5499
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 2.200005269023122,
+      "learning_rate": 0.00016587801461544893,
+      "loss": 1.7364,
+      "step": 5500
+    },
+    {
+      "epoch": 0.88016,
+      "grad_norm": 2.5719741497948143,
+      "learning_rate": 0.0001658650179120367,
+      "loss": 1.5175,
+      "step": 5501
+    },
+    {
+      "epoch": 0.88032,
+      "grad_norm": 2.3534892496114033,
+      "learning_rate": 0.00016585201924330828,
+      "loss": 1.5386,
+      "step": 5502
+    },
+    {
+      "epoch": 0.88048,
+      "grad_norm": 3.5433435524395978,
+      "learning_rate": 0.00016583901860965154,
+      "loss": 1.5948,
+      "step": 5503
+    },
+    {
+      "epoch": 0.88064,
+      "grad_norm": 2.269052339422226,
+      "learning_rate": 0.00016582601601145447,
+      "loss": 1.5832,
+      "step": 5504
+    },
+    {
+      "epoch": 0.8808,
+      "grad_norm": 2.4060301265020154,
+      "learning_rate": 0.00016581301144910496,
+      "loss": 1.6069,
+      "step": 5505
+    },
+    {
+      "epoch": 0.88096,
+      "grad_norm": 2.1783482030085652,
+      "learning_rate": 0.00016580000492299112,
+      "loss": 1.6071,
+      "step": 5506
+    },
+    {
+      "epoch": 0.88112,
+      "grad_norm": 2.188793362008775,
+      "learning_rate": 0.000165786996433501,
+      "loss": 1.5674,
+      "step": 5507
+    },
+    {
+      "epoch": 0.88128,
+      "grad_norm": 3.1243775378493854,
+      "learning_rate": 0.00016577398598102277,
+      "loss": 1.5745,
+      "step": 5508
+    },
+    {
+      "epoch": 0.88144,
+      "grad_norm": 2.2144315958080507,
+      "learning_rate": 0.00016576097356594462,
+      "loss": 1.5985,
+      "step": 5509
+    },
+    {
+      "epoch": 0.8816,
+      "grad_norm": 2.4984173659241717,
+      "learning_rate": 0.00016574795918865488,
+      "loss": 1.6042,
+      "step": 5510
+    },
+    {
+      "epoch": 0.88176,
+      "grad_norm": 2.6976221258894166,
+      "learning_rate": 0.00016573494284954183,
+      "loss": 1.6705,
+      "step": 5511
+    },
+    {
+      "epoch": 0.88192,
+      "grad_norm": 2.40582311964227,
+      "learning_rate": 0.00016572192454899384,
+      "loss": 1.6613,
+      "step": 5512
+    },
+    {
+      "epoch": 0.88208,
+      "grad_norm": 2.681551827597092,
+      "learning_rate": 0.00016570890428739942,
+      "loss": 1.6443,
+      "step": 5513
+    },
+    {
+      "epoch": 0.88224,
+      "grad_norm": 2.0656748594168315,
+      "learning_rate": 0.00016569588206514704,
+      "loss": 1.594,
+      "step": 5514
+    },
+    {
+      "epoch": 0.8824,
+      "grad_norm": 2.559234993793897,
+      "learning_rate": 0.00016568285788262527,
+      "loss": 1.6017,
+      "step": 5515
+    },
+    {
+      "epoch": 0.88256,
+      "grad_norm": 2.318447016224023,
+      "learning_rate": 0.00016566983174022272,
+      "loss": 1.6258,
+      "step": 5516
+    },
+    {
+      "epoch": 0.88272,
+      "grad_norm": 2.108398578042999,
+      "learning_rate": 0.00016565680363832809,
+      "loss": 1.6116,
+      "step": 5517
+    },
+    {
+      "epoch": 0.88288,
+      "grad_norm": 2.195030279724741,
+      "learning_rate": 0.0001656437735773301,
+      "loss": 1.5562,
+      "step": 5518
+    },
+    {
+      "epoch": 0.88304,
+      "grad_norm": 2.104905963035737,
+      "learning_rate": 0.00016563074155761758,
+      "loss": 1.5761,
+      "step": 5519
+    },
+    {
+      "epoch": 0.8832,
+      "grad_norm": 2.178041404104921,
+      "learning_rate": 0.00016561770757957938,
+      "loss": 1.6105,
+      "step": 5520
+    },
+    {
+      "epoch": 0.88336,
+      "grad_norm": 2.0236197125728173,
+      "learning_rate": 0.00016560467164360436,
+      "loss": 1.5764,
+      "step": 5521
+    },
+    {
+      "epoch": 0.88352,
+      "grad_norm": 2.164761948794186,
+      "learning_rate": 0.00016559163375008156,
+      "loss": 1.6237,
+      "step": 5522
+    },
+    {
+      "epoch": 0.88368,
+      "grad_norm": 2.3636991071961533,
+      "learning_rate": 0.00016557859389939997,
+      "loss": 1.5617,
+      "step": 5523
+    },
+    {
+      "epoch": 0.88384,
+      "grad_norm": 2.2492613005690263,
+      "learning_rate": 0.00016556555209194871,
+      "loss": 1.61,
+      "step": 5524
+    },
+    {
+      "epoch": 0.884,
+      "grad_norm": 1.937776097815702,
+      "learning_rate": 0.0001655525083281169,
+      "loss": 1.5329,
+      "step": 5525
+    },
+    {
+      "epoch": 0.88416,
+      "grad_norm": 2.201208560639164,
+      "learning_rate": 0.0001655394626082938,
+      "loss": 1.573,
+      "step": 5526
+    },
+    {
+      "epoch": 0.88432,
+      "grad_norm": 2.4698223580568377,
+      "learning_rate": 0.00016552641493286857,
+      "loss": 1.666,
+      "step": 5527
+    },
+    {
+      "epoch": 0.88448,
+      "grad_norm": 2.4750136202466178,
+      "learning_rate": 0.00016551336530223066,
+      "loss": 1.532,
+      "step": 5528
+    },
+    {
+      "epoch": 0.88464,
+      "grad_norm": 2.3662319411192985,
+      "learning_rate": 0.00016550031371676934,
+      "loss": 1.5713,
+      "step": 5529
+    },
+    {
+      "epoch": 0.8848,
+      "grad_norm": 2.8253507687979154,
+      "learning_rate": 0.00016548726017687418,
+      "loss": 1.6379,
+      "step": 5530
+    },
+    {
+      "epoch": 0.88496,
+      "grad_norm": 3.0725836365598123,
+      "learning_rate": 0.0001654742046829345,
+      "loss": 1.6447,
+      "step": 5531
+    },
+    {
+      "epoch": 0.88512,
+      "grad_norm": 2.0831342935026447,
+      "learning_rate": 0.00016546114723534004,
+      "loss": 1.5816,
+      "step": 5532
+    },
+    {
+      "epoch": 0.88528,
+      "grad_norm": 2.074882116721647,
+      "learning_rate": 0.0001654480878344803,
+      "loss": 1.6082,
+      "step": 5533
+    },
+    {
+      "epoch": 0.88544,
+      "grad_norm": 2.262037783821123,
+      "learning_rate": 0.000165435026480745,
+      "loss": 1.5186,
+      "step": 5534
+    },
+    {
+      "epoch": 0.8856,
+      "grad_norm": 2.4888897472548495,
+      "learning_rate": 0.00016542196317452382,
+      "loss": 1.5955,
+      "step": 5535
+    },
+    {
+      "epoch": 0.88576,
+      "grad_norm": 2.267947497685959,
+      "learning_rate": 0.00016540889791620664,
+      "loss": 1.6227,
+      "step": 5536
+    },
+    {
+      "epoch": 0.88592,
+      "grad_norm": 2.3001539682899854,
+      "learning_rate": 0.00016539583070618324,
+      "loss": 1.5594,
+      "step": 5537
+    },
+    {
+      "epoch": 0.88608,
+      "grad_norm": 2.0300263640519622,
+      "learning_rate": 0.00016538276154484357,
+      "loss": 1.613,
+      "step": 5538
+    },
+    {
+      "epoch": 0.88624,
+      "grad_norm": 2.11687287294229,
+      "learning_rate": 0.00016536969043257752,
+      "loss": 1.584,
+      "step": 5539
+    },
+    {
+      "epoch": 0.8864,
+      "grad_norm": 2.516956254252007,
+      "learning_rate": 0.0001653566173697752,
+      "loss": 1.5928,
+      "step": 5540
+    },
+    {
+      "epoch": 0.88656,
+      "grad_norm": 2.094462703587449,
+      "learning_rate": 0.00016534354235682664,
+      "loss": 1.574,
+      "step": 5541
+    },
+    {
+      "epoch": 0.88672,
+      "grad_norm": 2.4076194996093347,
+      "learning_rate": 0.00016533046539412197,
+      "loss": 1.6092,
+      "step": 5542
+    },
+    {
+      "epoch": 0.88688,
+      "grad_norm": 1.9805528656213256,
+      "learning_rate": 0.00016531738648205142,
+      "loss": 1.5969,
+      "step": 5543
+    },
+    {
+      "epoch": 0.88704,
+      "grad_norm": 2.448012387483452,
+      "learning_rate": 0.00016530430562100522,
+      "loss": 1.7169,
+      "step": 5544
+    },
+    {
+      "epoch": 0.8872,
+      "grad_norm": 2.2988136223516453,
+      "learning_rate": 0.00016529122281137374,
+      "loss": 1.5495,
+      "step": 5545
+    },
+    {
+      "epoch": 0.88736,
+      "grad_norm": 2.8045141893568757,
+      "learning_rate": 0.00016527813805354728,
+      "loss": 1.667,
+      "step": 5546
+    },
+    {
+      "epoch": 0.88752,
+      "grad_norm": 2.164759110782868,
+      "learning_rate": 0.00016526505134791629,
+      "loss": 1.6122,
+      "step": 5547
+    },
+    {
+      "epoch": 0.88768,
+      "grad_norm": 2.4023796512568976,
+      "learning_rate": 0.0001652519626948713,
+      "loss": 1.5725,
+      "step": 5548
+    },
+    {
+      "epoch": 0.88784,
+      "grad_norm": 3.01798047267947,
+      "learning_rate": 0.00016523887209480282,
+      "loss": 1.5911,
+      "step": 5549
+    },
+    {
+      "epoch": 0.888,
+      "grad_norm": 2.1542366138883238,
+      "learning_rate": 0.00016522577954810146,
+      "loss": 1.6185,
+      "step": 5550
+    },
+    {
+      "epoch": 0.88816,
+      "grad_norm": 3.060598958651162,
+      "learning_rate": 0.00016521268505515787,
+      "loss": 1.7537,
+      "step": 5551
+    },
+    {
+      "epoch": 0.88832,
+      "grad_norm": 2.623939349937115,
+      "learning_rate": 0.00016519958861636278,
+      "loss": 1.624,
+      "step": 5552
+    },
+    {
+      "epoch": 0.88848,
+      "grad_norm": 2.3483711388483,
+      "learning_rate": 0.00016518649023210697,
+      "loss": 1.7252,
+      "step": 5553
+    },
+    {
+      "epoch": 0.88864,
+      "grad_norm": 3.153303465033247,
+      "learning_rate": 0.0001651733899027813,
+      "loss": 1.5522,
+      "step": 5554
+    },
+    {
+      "epoch": 0.8888,
+      "grad_norm": 2.2450037312818143,
+      "learning_rate": 0.00016516028762877664,
+      "loss": 1.5944,
+      "step": 5555
+    },
+    {
+      "epoch": 0.88896,
+      "grad_norm": 3.499609941817572,
+      "learning_rate": 0.00016514718341048394,
+      "loss": 1.5345,
+      "step": 5556
+    },
+    {
+      "epoch": 0.88912,
+      "grad_norm": 2.3175036411015837,
+      "learning_rate": 0.00016513407724829418,
+      "loss": 1.545,
+      "step": 5557
+    },
+    {
+      "epoch": 0.88928,
+      "grad_norm": 2.5535500230245973,
+      "learning_rate": 0.00016512096914259852,
+      "loss": 1.5709,
+      "step": 5558
+    },
+    {
+      "epoch": 0.88944,
+      "grad_norm": 2.681736013059736,
+      "learning_rate": 0.000165107859093788,
+      "loss": 1.7039,
+      "step": 5559
+    },
+    {
+      "epoch": 0.8896,
+      "grad_norm": 2.7719928928087283,
+      "learning_rate": 0.00016509474710225386,
+      "loss": 1.579,
+      "step": 5560
+    },
+    {
+      "epoch": 0.88976,
+      "grad_norm": 2.2297118055572307,
+      "learning_rate": 0.00016508163316838732,
+      "loss": 1.539,
+      "step": 5561
+    },
+    {
+      "epoch": 0.88992,
+      "grad_norm": 2.6986327766513667,
+      "learning_rate": 0.00016506851729257966,
+      "loss": 1.7292,
+      "step": 5562
+    },
+    {
+      "epoch": 0.89008,
+      "grad_norm": 2.7127663075354054,
+      "learning_rate": 0.00016505539947522224,
+      "loss": 1.6173,
+      "step": 5563
+    },
+    {
+      "epoch": 0.89024,
+      "grad_norm": 2.2775117452330558,
+      "learning_rate": 0.00016504227971670654,
+      "loss": 1.5124,
+      "step": 5564
+    },
+    {
+      "epoch": 0.8904,
+      "grad_norm": 2.5821710611388653,
+      "learning_rate": 0.00016502915801742396,
+      "loss": 1.7248,
+      "step": 5565
+    },
+    {
+      "epoch": 0.89056,
+      "grad_norm": 2.494551504818838,
+      "learning_rate": 0.00016501603437776608,
+      "loss": 1.6216,
+      "step": 5566
+    },
+    {
+      "epoch": 0.89072,
+      "grad_norm": 2.2022083083425343,
+      "learning_rate": 0.00016500290879812446,
+      "loss": 1.5433,
+      "step": 5567
+    },
+    {
+      "epoch": 0.89088,
+      "grad_norm": 2.068929205930144,
+      "learning_rate": 0.00016498978127889077,
+      "loss": 1.5644,
+      "step": 5568
+    },
+    {
+      "epoch": 0.89104,
+      "grad_norm": 2.647569859499225,
+      "learning_rate": 0.0001649766518204567,
+      "loss": 1.6338,
+      "step": 5569
+    },
+    {
+      "epoch": 0.8912,
+      "grad_norm": 2.540190109538804,
+      "learning_rate": 0.000164963520423214,
+      "loss": 1.6413,
+      "step": 5570
+    },
+    {
+      "epoch": 0.89136,
+      "grad_norm": 2.4683337245440886,
+      "learning_rate": 0.00016495038708755458,
+      "loss": 1.6507,
+      "step": 5571
+    },
+    {
+      "epoch": 0.89152,
+      "grad_norm": 2.0526479448008508,
+      "learning_rate": 0.0001649372518138702,
+      "loss": 1.5378,
+      "step": 5572
+    },
+    {
+      "epoch": 0.89168,
+      "grad_norm": 2.159747499940862,
+      "learning_rate": 0.00016492411460255284,
+      "loss": 1.6056,
+      "step": 5573
+    },
+    {
+      "epoch": 0.89184,
+      "grad_norm": 2.5269410907479553,
+      "learning_rate": 0.00016491097545399453,
+      "loss": 1.5171,
+      "step": 5574
+    },
+    {
+      "epoch": 0.892,
+      "grad_norm": 2.557853728029394,
+      "learning_rate": 0.00016489783436858727,
+      "loss": 1.6416,
+      "step": 5575
+    },
+    {
+      "epoch": 0.89216,
+      "grad_norm": 2.545820711628101,
+      "learning_rate": 0.00016488469134672324,
+      "loss": 1.6713,
+      "step": 5576
+    },
+    {
+      "epoch": 0.89232,
+      "grad_norm": 2.2698098218456795,
+      "learning_rate": 0.00016487154638879453,
+      "loss": 1.6021,
+      "step": 5577
+    },
+    {
+      "epoch": 0.89248,
+      "grad_norm": 2.284028544665764,
+      "learning_rate": 0.00016485839949519342,
+      "loss": 1.616,
+      "step": 5578
+    },
+    {
+      "epoch": 0.89264,
+      "grad_norm": 2.3224324430598497,
+      "learning_rate": 0.0001648452506663122,
+      "loss": 1.6657,
+      "step": 5579
+    },
+    {
+      "epoch": 0.8928,
+      "grad_norm": 2.697820579893762,
+      "learning_rate": 0.00016483209990254314,
+      "loss": 1.5477,
+      "step": 5580
+    },
+    {
+      "epoch": 0.89296,
+      "grad_norm": 2.8687373031453274,
+      "learning_rate": 0.00016481894720427872,
+      "loss": 1.5935,
+      "step": 5581
+    },
+    {
+      "epoch": 0.89312,
+      "grad_norm": 2.0332165511166203,
+      "learning_rate": 0.00016480579257191139,
+      "loss": 1.6285,
+      "step": 5582
+    },
+    {
+      "epoch": 0.89328,
+      "grad_norm": 3.094932146019159,
+      "learning_rate": 0.00016479263600583361,
+      "loss": 1.5561,
+      "step": 5583
+    },
+    {
+      "epoch": 0.89344,
+      "grad_norm": 2.481352583415718,
+      "learning_rate": 0.00016477947750643797,
+      "loss": 1.5809,
+      "step": 5584
+    },
+    {
+      "epoch": 0.8936,
+      "grad_norm": 1.958713429500679,
+      "learning_rate": 0.00016476631707411713,
+      "loss": 1.6185,
+      "step": 5585
+    },
+    {
+      "epoch": 0.89376,
+      "grad_norm": 2.29743135027879,
+      "learning_rate": 0.00016475315470926378,
+      "loss": 1.6316,
+      "step": 5586
+    },
+    {
+      "epoch": 0.89392,
+      "grad_norm": 2.1708905769791214,
+      "learning_rate": 0.00016473999041227065,
+      "loss": 1.5892,
+      "step": 5587
+    },
+    {
+      "epoch": 0.89408,
+      "grad_norm": 3.141598594606022,
+      "learning_rate": 0.00016472682418353053,
+      "loss": 1.6149,
+      "step": 5588
+    },
+    {
+      "epoch": 0.89424,
+      "grad_norm": 2.0783531256011307,
+      "learning_rate": 0.00016471365602343626,
+      "loss": 1.5605,
+      "step": 5589
+    },
+    {
+      "epoch": 0.8944,
+      "grad_norm": 2.352513281320985,
+      "learning_rate": 0.00016470048593238086,
+      "loss": 1.5205,
+      "step": 5590
+    },
+    {
+      "epoch": 0.89456,
+      "grad_norm": 2.232467061538314,
+      "learning_rate": 0.00016468731391075717,
+      "loss": 1.544,
+      "step": 5591
+    },
+    {
+      "epoch": 0.89472,
+      "grad_norm": 2.5399330950787693,
+      "learning_rate": 0.00016467413995895834,
+      "loss": 1.5997,
+      "step": 5592
+    },
+    {
+      "epoch": 0.89488,
+      "grad_norm": 2.3214212821226665,
+      "learning_rate": 0.00016466096407737738,
+      "loss": 1.5941,
+      "step": 5593
+    },
+    {
+      "epoch": 0.89504,
+      "grad_norm": 2.4424446224435243,
+      "learning_rate": 0.00016464778626640748,
+      "loss": 1.4931,
+      "step": 5594
+    },
+    {
+      "epoch": 0.8952,
+      "grad_norm": 1.9831797798505502,
+      "learning_rate": 0.00016463460652644187,
+      "loss": 1.6285,
+      "step": 5595
+    },
+    {
+      "epoch": 0.89536,
+      "grad_norm": 2.818433912927114,
+      "learning_rate": 0.00016462142485787378,
+      "loss": 1.5848,
+      "step": 5596
+    },
+    {
+      "epoch": 0.89552,
+      "grad_norm": 2.743902573195639,
+      "learning_rate": 0.0001646082412610965,
+      "loss": 1.4987,
+      "step": 5597
+    },
+    {
+      "epoch": 0.89568,
+      "grad_norm": 2.2171011018455666,
+      "learning_rate": 0.00016459505573650346,
+      "loss": 1.5965,
+      "step": 5598
+    },
+    {
+      "epoch": 0.89584,
+      "grad_norm": 2.5761076180191202,
+      "learning_rate": 0.0001645818682844881,
+      "loss": 1.6427,
+      "step": 5599
+    },
+    {
+      "epoch": 0.896,
+      "grad_norm": 3.1122842603021494,
+      "learning_rate": 0.00016456867890544388,
+      "loss": 1.5679,
+      "step": 5600
+    },
+    {
+      "epoch": 0.89616,
+      "grad_norm": 2.19318064687952,
+      "learning_rate": 0.00016455548759976437,
+      "loss": 1.6521,
+      "step": 5601
+    },
+    {
+      "epoch": 0.89632,
+      "grad_norm": 3.342446749844203,
+      "learning_rate": 0.0001645422943678432,
+      "loss": 1.5842,
+      "step": 5602
+    },
+    {
+      "epoch": 0.89648,
+      "grad_norm": 2.3217021069535706,
+      "learning_rate": 0.00016452909921007398,
+      "loss": 1.5536,
+      "step": 5603
+    },
+    {
+      "epoch": 0.89664,
+      "grad_norm": 3.415953574007582,
+      "learning_rate": 0.0001645159021268505,
+      "loss": 1.6077,
+      "step": 5604
+    },
+    {
+      "epoch": 0.8968,
+      "grad_norm": 3.046962837401589,
+      "learning_rate": 0.00016450270311856645,
+      "loss": 1.6352,
+      "step": 5605
+    },
+    {
+      "epoch": 0.89696,
+      "grad_norm": 2.5032972282308106,
+      "learning_rate": 0.0001644895021856158,
+      "loss": 1.6502,
+      "step": 5606
+    },
+    {
+      "epoch": 0.89712,
+      "grad_norm": 2.509511437408954,
+      "learning_rate": 0.00016447629932839233,
+      "loss": 1.4899,
+      "step": 5607
+    },
+    {
+      "epoch": 0.89728,
+      "grad_norm": 2.2459081156873184,
+      "learning_rate": 0.00016446309454729006,
+      "loss": 1.6272,
+      "step": 5608
+    },
+    {
+      "epoch": 0.89744,
+      "grad_norm": 2.6469316965140064,
+      "learning_rate": 0.00016444988784270297,
+      "loss": 1.5934,
+      "step": 5609
+    },
+    {
+      "epoch": 0.8976,
+      "grad_norm": 3.3902116585162294,
+      "learning_rate": 0.00016443667921502513,
+      "loss": 1.6077,
+      "step": 5610
+    },
+    {
+      "epoch": 0.89776,
+      "grad_norm": 2.555526704972532,
+      "learning_rate": 0.00016442346866465073,
+      "loss": 1.6373,
+      "step": 5611
+    },
+    {
+      "epoch": 0.89792,
+      "grad_norm": 2.3557995545673625,
+      "learning_rate": 0.00016441025619197386,
+      "loss": 1.5264,
+      "step": 5612
+    },
+    {
+      "epoch": 0.89808,
+      "grad_norm": 2.385816470077708,
+      "learning_rate": 0.00016439704179738878,
+      "loss": 1.657,
+      "step": 5613
+    },
+    {
+      "epoch": 0.89824,
+      "grad_norm": 2.0774276353773202,
+      "learning_rate": 0.00016438382548128987,
+      "loss": 1.5568,
+      "step": 5614
+    },
+    {
+      "epoch": 0.8984,
+      "grad_norm": 2.1548374200293128,
+      "learning_rate": 0.00016437060724407135,
+      "loss": 1.4949,
+      "step": 5615
+    },
+    {
+      "epoch": 0.89856,
+      "grad_norm": 2.4798223402259305,
+      "learning_rate": 0.00016435738708612776,
+      "loss": 1.6618,
+      "step": 5616
+    },
+    {
+      "epoch": 0.89872,
+      "grad_norm": 2.161298918672694,
+      "learning_rate": 0.0001643441650078535,
+      "loss": 1.4485,
+      "step": 5617
+    },
+    {
+      "epoch": 0.89888,
+      "grad_norm": 2.6013603470876197,
+      "learning_rate": 0.0001643309410096431,
+      "loss": 1.5994,
+      "step": 5618
+    },
+    {
+      "epoch": 0.89904,
+      "grad_norm": 2.253894966977784,
+      "learning_rate": 0.00016431771509189122,
+      "loss": 1.5689,
+      "step": 5619
+    },
+    {
+      "epoch": 0.8992,
+      "grad_norm": 2.151175926186,
+      "learning_rate": 0.00016430448725499237,
+      "loss": 1.5747,
+      "step": 5620
+    },
+    {
+      "epoch": 0.89936,
+      "grad_norm": 2.6259292630730418,
+      "learning_rate": 0.00016429125749934138,
+      "loss": 1.5987,
+      "step": 5621
+    },
+    {
+      "epoch": 0.89952,
+      "grad_norm": 2.5394754620490785,
+      "learning_rate": 0.0001642780258253329,
+      "loss": 1.5944,
+      "step": 5622
+    },
+    {
+      "epoch": 0.89968,
+      "grad_norm": 4.630861471343401,
+      "learning_rate": 0.00016426479223336182,
+      "loss": 1.7991,
+      "step": 5623
+    },
+    {
+      "epoch": 0.89984,
+      "grad_norm": 3.298387551271487,
+      "learning_rate": 0.00016425155672382296,
+      "loss": 1.6328,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.9980560507877123,
+      "learning_rate": 0.0001642383192971113,
+      "loss": 1.5127,
+      "step": 5625
+    },
+    {
+      "epoch": 0.90016,
+      "grad_norm": 2.221013178355228,
+      "learning_rate": 0.0001642250799536218,
+      "loss": 1.5956,
+      "step": 5626
+    },
+    {
+      "epoch": 0.90032,
+      "grad_norm": 2.1595678090557606,
+      "learning_rate": 0.0001642118386937495,
+      "loss": 1.592,
+      "step": 5627
+    },
+    {
+      "epoch": 0.90048,
+      "grad_norm": 2.150904488194907,
+      "learning_rate": 0.00016419859551788948,
+      "loss": 1.6099,
+      "step": 5628
+    },
+    {
+      "epoch": 0.90064,
+      "grad_norm": 3.1220682628415752,
+      "learning_rate": 0.00016418535042643693,
+      "loss": 1.6264,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9008,
+      "grad_norm": 2.452326831789603,
+      "learning_rate": 0.00016417210341978703,
+      "loss": 1.5651,
+      "step": 5630
+    },
+    {
+      "epoch": 0.90096,
+      "grad_norm": 2.5354837154127603,
+      "learning_rate": 0.0001641588544983351,
+      "loss": 1.6098,
+      "step": 5631
+    },
+    {
+      "epoch": 0.90112,
+      "grad_norm": 2.4420645669587535,
+      "learning_rate": 0.00016414560366247643,
+      "loss": 1.5943,
+      "step": 5632
+    },
+    {
+      "epoch": 0.90128,
+      "grad_norm": 2.2792831616129083,
+      "learning_rate": 0.0001641323509126064,
+      "loss": 1.5658,
+      "step": 5633
+    },
+    {
+      "epoch": 0.90144,
+      "grad_norm": 2.51747736210609,
+      "learning_rate": 0.0001641190962491205,
+      "loss": 1.6329,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9016,
+      "grad_norm": 3.184287332624619,
+      "learning_rate": 0.00016410583967241417,
+      "loss": 1.6932,
+      "step": 5635
+    },
+    {
+      "epoch": 0.90176,
+      "grad_norm": 2.9092868187638885,
+      "learning_rate": 0.00016409258118288303,
+      "loss": 1.6087,
+      "step": 5636
+    },
+    {
+      "epoch": 0.90192,
+      "grad_norm": 2.374691363650859,
+      "learning_rate": 0.00016407932078092263,
+      "loss": 1.5748,
+      "step": 5637
+    },
+    {
+      "epoch": 0.90208,
+      "grad_norm": 3.3529119698793037,
+      "learning_rate": 0.00016406605846692868,
+      "loss": 1.6779,
+      "step": 5638
+    },
+    {
+      "epoch": 0.90224,
+      "grad_norm": 2.310066078473659,
+      "learning_rate": 0.00016405279424129692,
+      "loss": 1.5102,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9024,
+      "grad_norm": 2.9323790955864237,
+      "learning_rate": 0.00016403952810442309,
+      "loss": 1.6646,
+      "step": 5640
+    },
+    {
+      "epoch": 0.90256,
+      "grad_norm": 2.1804333071243907,
+      "learning_rate": 0.00016402626005670306,
+      "loss": 1.5176,
+      "step": 5641
+    },
+    {
+      "epoch": 0.90272,
+      "grad_norm": 3.6179185358968065,
+      "learning_rate": 0.00016401299009853273,
+      "loss": 1.6276,
+      "step": 5642
+    },
+    {
+      "epoch": 0.90288,
+      "grad_norm": 2.362971315671023,
+      "learning_rate": 0.00016399971823030803,
+      "loss": 1.5867,
+      "step": 5643
+    },
+    {
+      "epoch": 0.90304,
+      "grad_norm": 3.1900397063614747,
+      "learning_rate": 0.00016398644445242503,
+      "loss": 1.5533,
+      "step": 5644
+    },
+    {
+      "epoch": 0.9032,
+      "grad_norm": 2.431238378665304,
+      "learning_rate": 0.00016397316876527976,
+      "loss": 1.5645,
+      "step": 5645
+    },
+    {
+      "epoch": 0.90336,
+      "grad_norm": 3.3098419754695847,
+      "learning_rate": 0.00016395989116926835,
+      "loss": 1.5873,
+      "step": 5646
+    },
+    {
+      "epoch": 0.90352,
+      "grad_norm": 2.1233071422490832,
+      "learning_rate": 0.00016394661166478698,
+      "loss": 1.4975,
+      "step": 5647
+    },
+    {
+      "epoch": 0.90368,
+      "grad_norm": 2.165000994826845,
+      "learning_rate": 0.0001639333302522319,
+      "loss": 1.5604,
+      "step": 5648
+    },
+    {
+      "epoch": 0.90384,
+      "grad_norm": 2.2381565540387784,
+      "learning_rate": 0.00016392004693199937,
+      "loss": 1.5595,
+      "step": 5649
+    },
+    {
+      "epoch": 0.904,
+      "grad_norm": 2.1225609750824272,
+      "learning_rate": 0.00016390676170448585,
+      "loss": 1.5541,
+      "step": 5650
+    },
+    {
+      "epoch": 0.90416,
+      "grad_norm": 2.4931069904344176,
+      "learning_rate": 0.00016389347457008765,
+      "loss": 1.5933,
+      "step": 5651
+    },
+    {
+      "epoch": 0.90432,
+      "grad_norm": 2.7529583412319467,
+      "learning_rate": 0.00016388018552920128,
+      "loss": 1.5702,
+      "step": 5652
+    },
+    {
+      "epoch": 0.90448,
+      "grad_norm": 2.287165762869328,
+      "learning_rate": 0.00016386689458222325,
+      "loss": 1.5792,
+      "step": 5653
+    },
+    {
+      "epoch": 0.90464,
+      "grad_norm": 2.330287511842668,
+      "learning_rate": 0.00016385360172955013,
+      "loss": 1.5612,
+      "step": 5654
+    },
+    {
+      "epoch": 0.9048,
+      "grad_norm": 2.3201525226676902,
+      "learning_rate": 0.00016384030697157859,
+      "loss": 1.5534,
+      "step": 5655
+    },
+    {
+      "epoch": 0.90496,
+      "grad_norm": 2.3761236178581058,
+      "learning_rate": 0.00016382701030870535,
+      "loss": 1.5673,
+      "step": 5656
+    },
+    {
+      "epoch": 0.90512,
+      "grad_norm": 2.5132373556960155,
+      "learning_rate": 0.0001638137117413271,
+      "loss": 1.7248,
+      "step": 5657
+    },
+    {
+      "epoch": 0.90528,
+      "grad_norm": 2.3620740390855532,
+      "learning_rate": 0.00016380041126984066,
+      "loss": 1.5621,
+      "step": 5658
+    },
+    {
+      "epoch": 0.90544,
+      "grad_norm": 3.0998991393329396,
+      "learning_rate": 0.00016378710889464296,
+      "loss": 1.5635,
+      "step": 5659
+    },
+    {
+      "epoch": 0.9056,
+      "grad_norm": 2.8700934307019526,
+      "learning_rate": 0.00016377380461613084,
+      "loss": 1.4802,
+      "step": 5660
+    },
+    {
+      "epoch": 0.90576,
+      "grad_norm": 2.766560750385336,
+      "learning_rate": 0.00016376049843470133,
+      "loss": 1.6225,
+      "step": 5661
+    },
+    {
+      "epoch": 0.90592,
+      "grad_norm": 2.4125807268670707,
+      "learning_rate": 0.00016374719035075147,
+      "loss": 1.5302,
+      "step": 5662
+    },
+    {
+      "epoch": 0.90608,
+      "grad_norm": 2.3842914732803036,
+      "learning_rate": 0.00016373388036467832,
+      "loss": 1.5073,
+      "step": 5663
+    },
+    {
+      "epoch": 0.90624,
+      "grad_norm": 2.533095201069794,
+      "learning_rate": 0.00016372056847687905,
+      "loss": 1.4736,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9064,
+      "grad_norm": 3.245903663859437,
+      "learning_rate": 0.00016370725468775087,
+      "loss": 1.6139,
+      "step": 5665
+    },
+    {
+      "epoch": 0.90656,
+      "grad_norm": 2.58177179336514,
+      "learning_rate": 0.00016369393899769104,
+      "loss": 1.5399,
+      "step": 5666
+    },
+    {
+      "epoch": 0.90672,
+      "grad_norm": 2.3625227676257903,
+      "learning_rate": 0.0001636806214070969,
+      "loss": 1.5405,
+      "step": 5667
+    },
+    {
+      "epoch": 0.90688,
+      "grad_norm": 2.3721207398921305,
+      "learning_rate": 0.00016366730191636579,
+      "loss": 1.5192,
+      "step": 5668
+    },
+    {
+      "epoch": 0.90704,
+      "grad_norm": 2.5256337868525884,
+      "learning_rate": 0.00016365398052589516,
+      "loss": 1.6302,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9072,
+      "grad_norm": 2.5543960497112006,
+      "learning_rate": 0.0001636406572360825,
+      "loss": 1.5307,
+      "step": 5670
+    },
+    {
+      "epoch": 0.90736,
+      "grad_norm": 3.095077761756798,
+      "learning_rate": 0.00016362733204732538,
+      "loss": 1.5554,
+      "step": 5671
+    },
+    {
+      "epoch": 0.90752,
+      "grad_norm": 2.3369305157027442,
+      "learning_rate": 0.0001636140049600214,
+      "loss": 1.5514,
+      "step": 5672
+    },
+    {
+      "epoch": 0.90768,
+      "grad_norm": 2.522784336140468,
+      "learning_rate": 0.0001636006759745682,
+      "loss": 1.6712,
+      "step": 5673
+    },
+    {
+      "epoch": 0.90784,
+      "grad_norm": 2.135404046611837,
+      "learning_rate": 0.00016358734509136348,
+      "loss": 1.6106,
+      "step": 5674
+    },
+    {
+      "epoch": 0.908,
+      "grad_norm": 2.95096354184905,
+      "learning_rate": 0.00016357401231080506,
+      "loss": 1.6398,
+      "step": 5675
+    },
+    {
+      "epoch": 0.90816,
+      "grad_norm": 2.3072258582381924,
+      "learning_rate": 0.00016356067763329075,
+      "loss": 1.6187,
+      "step": 5676
+    },
+    {
+      "epoch": 0.90832,
+      "grad_norm": 3.1235204699552535,
+      "learning_rate": 0.0001635473410592184,
+      "loss": 1.5766,
+      "step": 5677
+    },
+    {
+      "epoch": 0.90848,
+      "grad_norm": 2.208789251130852,
+      "learning_rate": 0.00016353400258898603,
+      "loss": 1.564,
+      "step": 5678
+    },
+    {
+      "epoch": 0.90864,
+      "grad_norm": 2.3612571921641425,
+      "learning_rate": 0.00016352066222299162,
+      "loss": 1.5874,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9088,
+      "grad_norm": 2.702905205913385,
+      "learning_rate": 0.0001635073199616332,
+      "loss": 1.6036,
+      "step": 5680
+    },
+    {
+      "epoch": 0.90896,
+      "grad_norm": 2.1083569437420118,
+      "learning_rate": 0.00016349397580530886,
+      "loss": 1.5585,
+      "step": 5681
+    },
+    {
+      "epoch": 0.90912,
+      "grad_norm": 2.215679670601436,
+      "learning_rate": 0.0001634806297544168,
+      "loss": 1.602,
+      "step": 5682
+    },
+    {
+      "epoch": 0.90928,
+      "grad_norm": 2.869227316982539,
+      "learning_rate": 0.00016346728180935527,
+      "loss": 1.5802,
+      "step": 5683
+    },
+    {
+      "epoch": 0.90944,
+      "grad_norm": 2.2596473386760763,
+      "learning_rate": 0.00016345393197052252,
+      "loss": 1.491,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9096,
+      "grad_norm": 2.711683900891032,
+      "learning_rate": 0.00016344058023831691,
+      "loss": 1.5818,
+      "step": 5685
+    },
+    {
+      "epoch": 0.90976,
+      "grad_norm": 2.4812491052409413,
+      "learning_rate": 0.00016342722661313682,
+      "loss": 1.4742,
+      "step": 5686
+    },
+    {
+      "epoch": 0.90992,
+      "grad_norm": 2.5219403535107254,
+      "learning_rate": 0.00016341387109538073,
+      "loss": 1.6425,
+      "step": 5687
+    },
+    {
+      "epoch": 0.91008,
+      "grad_norm": 2.5998343739837946,
+      "learning_rate": 0.0001634005136854471,
+      "loss": 1.665,
+      "step": 5688
+    },
+    {
+      "epoch": 0.91024,
+      "grad_norm": 2.475072077211948,
+      "learning_rate": 0.00016338715438373458,
+      "loss": 1.6691,
+      "step": 5689
+    },
+    {
+      "epoch": 0.9104,
+      "grad_norm": 2.822843897827985,
+      "learning_rate": 0.0001633737931906417,
+      "loss": 1.6511,
+      "step": 5690
+    },
+    {
+      "epoch": 0.91056,
+      "grad_norm": 2.0252895382878275,
+      "learning_rate": 0.00016336043010656715,
+      "loss": 1.5847,
+      "step": 5691
+    },
+    {
+      "epoch": 0.91072,
+      "grad_norm": 2.185614103357862,
+      "learning_rate": 0.00016334706513190973,
+      "loss": 1.546,
+      "step": 5692
+    },
+    {
+      "epoch": 0.91088,
+      "grad_norm": 2.23858348485231,
+      "learning_rate": 0.0001633336982670682,
+      "loss": 1.6104,
+      "step": 5693
+    },
+    {
+      "epoch": 0.91104,
+      "grad_norm": 2.187656842004312,
+      "learning_rate": 0.0001633203295124414,
+      "loss": 1.6028,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9112,
+      "grad_norm": 2.3669366483435827,
+      "learning_rate": 0.0001633069588684282,
+      "loss": 1.6616,
+      "step": 5695
+    },
+    {
+      "epoch": 0.91136,
+      "grad_norm": 2.238588708421677,
+      "learning_rate": 0.00016329358633542765,
+      "loss": 1.5684,
+      "step": 5696
+    },
+    {
+      "epoch": 0.91152,
+      "grad_norm": 2.2900222098537695,
+      "learning_rate": 0.00016328021191383866,
+      "loss": 1.6166,
+      "step": 5697
+    },
+    {
+      "epoch": 0.91168,
+      "grad_norm": 2.4436978617511356,
+      "learning_rate": 0.0001632668356040604,
+      "loss": 1.6777,
+      "step": 5698
+    },
+    {
+      "epoch": 0.91184,
+      "grad_norm": 2.1688346202270834,
+      "learning_rate": 0.0001632534574064919,
+      "loss": 1.5747,
+      "step": 5699
+    },
+    {
+      "epoch": 0.912,
+      "grad_norm": 2.3453722715827334,
+      "learning_rate": 0.00016324007732153244,
+      "loss": 1.4929,
+      "step": 5700
+    },
+    {
+      "epoch": 0.91216,
+      "grad_norm": 2.3278479421630713,
+      "learning_rate": 0.0001632266953495812,
+      "loss": 1.5675,
+      "step": 5701
+    },
+    {
+      "epoch": 0.91232,
+      "grad_norm": 2.4921924197314604,
+      "learning_rate": 0.00016321331149103752,
+      "loss": 1.6315,
+      "step": 5702
+    },
+    {
+      "epoch": 0.91248,
+      "grad_norm": 2.34217230961308,
+      "learning_rate": 0.00016319992574630074,
+      "loss": 1.6526,
+      "step": 5703
+    },
+    {
+      "epoch": 0.91264,
+      "grad_norm": 2.3518158848275017,
+      "learning_rate": 0.00016318653811577027,
+      "loss": 1.575,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9128,
+      "grad_norm": 2.4985579292207434,
+      "learning_rate": 0.00016317314859984555,
+      "loss": 1.6314,
+      "step": 5705
+    },
+    {
+      "epoch": 0.91296,
+      "grad_norm": 2.157340777382908,
+      "learning_rate": 0.00016315975719892616,
+      "loss": 1.6033,
+      "step": 5706
+    },
+    {
+      "epoch": 0.91312,
+      "grad_norm": 2.356334979187611,
+      "learning_rate": 0.00016314636391341164,
+      "loss": 1.5091,
+      "step": 5707
+    },
+    {
+      "epoch": 0.91328,
+      "grad_norm": 2.6841630903260847,
+      "learning_rate": 0.0001631329687437016,
+      "loss": 1.5929,
+      "step": 5708
+    },
+    {
+      "epoch": 0.91344,
+      "grad_norm": 3.261882361984843,
+      "learning_rate": 0.00016311957169019582,
+      "loss": 1.6469,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9136,
+      "grad_norm": 2.485442397510712,
+      "learning_rate": 0.00016310617275329395,
+      "loss": 1.5418,
+      "step": 5710
+    },
+    {
+      "epoch": 0.91376,
+      "grad_norm": 2.884050086836712,
+      "learning_rate": 0.00016309277193339588,
+      "loss": 1.5165,
+      "step": 5711
+    },
+    {
+      "epoch": 0.91392,
+      "grad_norm": 2.7363547359141496,
+      "learning_rate": 0.0001630793692309014,
+      "loss": 1.578,
+      "step": 5712
+    },
+    {
+      "epoch": 0.91408,
+      "grad_norm": 2.455739052175341,
+      "learning_rate": 0.00016306596464621048,
+      "loss": 1.5295,
+      "step": 5713
+    },
+    {
+      "epoch": 0.91424,
+      "grad_norm": 2.254593820879646,
+      "learning_rate": 0.00016305255817972305,
+      "loss": 1.7132,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9144,
+      "grad_norm": 2.424147604821289,
+      "learning_rate": 0.00016303914983183918,
+      "loss": 1.6582,
+      "step": 5715
+    },
+    {
+      "epoch": 0.91456,
+      "grad_norm": 2.3441895315306387,
+      "learning_rate": 0.00016302573960295892,
+      "loss": 1.6133,
+      "step": 5716
+    },
+    {
+      "epoch": 0.91472,
+      "grad_norm": 2.7671520774198903,
+      "learning_rate": 0.00016301232749348245,
+      "loss": 1.6566,
+      "step": 5717
+    },
+    {
+      "epoch": 0.91488,
+      "grad_norm": 2.227864274744576,
+      "learning_rate": 0.00016299891350380992,
+      "loss": 1.6194,
+      "step": 5718
+    },
+    {
+      "epoch": 0.91504,
+      "grad_norm": 2.137455977290692,
+      "learning_rate": 0.00016298549763434162,
+      "loss": 1.6386,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9152,
+      "grad_norm": 2.1178108118380683,
+      "learning_rate": 0.00016297207988547787,
+      "loss": 1.6541,
+      "step": 5720
+    },
+    {
+      "epoch": 0.91536,
+      "grad_norm": 2.3924070083664133,
+      "learning_rate": 0.000162958660257619,
+      "loss": 1.6891,
+      "step": 5721
+    },
+    {
+      "epoch": 0.91552,
+      "grad_norm": 2.486324592029327,
+      "learning_rate": 0.00016294523875116546,
+      "loss": 1.5993,
+      "step": 5722
+    },
+    {
+      "epoch": 0.91568,
+      "grad_norm": 2.602532329872684,
+      "learning_rate": 0.00016293181536651771,
+      "loss": 1.598,
+      "step": 5723
+    },
+    {
+      "epoch": 0.91584,
+      "grad_norm": 2.2232154887337328,
+      "learning_rate": 0.00016291839010407628,
+      "loss": 1.6317,
+      "step": 5724
+    },
+    {
+      "epoch": 0.916,
+      "grad_norm": 2.380060696941587,
+      "learning_rate": 0.00016290496296424178,
+      "loss": 1.6153,
+      "step": 5725
+    },
+    {
+      "epoch": 0.91616,
+      "grad_norm": 2.714682827537324,
+      "learning_rate": 0.00016289153394741484,
+      "loss": 1.5747,
+      "step": 5726
+    },
+    {
+      "epoch": 0.91632,
+      "grad_norm": 2.256117422059784,
+      "learning_rate": 0.00016287810305399617,
+      "loss": 1.5734,
+      "step": 5727
+    },
+    {
+      "epoch": 0.91648,
+      "grad_norm": 3.4453375442162577,
+      "learning_rate": 0.00016286467028438654,
+      "loss": 1.5863,
+      "step": 5728
+    },
+    {
+      "epoch": 0.91664,
+      "grad_norm": 2.2622717408153354,
+      "learning_rate": 0.00016285123563898677,
+      "loss": 1.599,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9168,
+      "grad_norm": 2.7653836309185538,
+      "learning_rate": 0.00016283779911819765,
+      "loss": 1.5909,
+      "step": 5730
+    },
+    {
+      "epoch": 0.91696,
+      "grad_norm": 2.4148347684113123,
+      "learning_rate": 0.00016282436072242025,
+      "loss": 1.6833,
+      "step": 5731
+    },
+    {
+      "epoch": 0.91712,
+      "grad_norm": 2.798633886407107,
+      "learning_rate": 0.00016281092045205542,
+      "loss": 1.6301,
+      "step": 5732
+    },
+    {
+      "epoch": 0.91728,
+      "grad_norm": 2.215133001664802,
+      "learning_rate": 0.00016279747830750426,
+      "loss": 1.5143,
+      "step": 5733
+    },
+    {
+      "epoch": 0.91744,
+      "grad_norm": 2.9554492854798173,
+      "learning_rate": 0.00016278403428916782,
+      "loss": 1.6619,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9176,
+      "grad_norm": 2.535082318234247,
+      "learning_rate": 0.00016277058839744732,
+      "loss": 1.6504,
+      "step": 5735
+    },
+    {
+      "epoch": 0.91776,
+      "grad_norm": 2.182266984363563,
+      "learning_rate": 0.00016275714063274395,
+      "loss": 1.5003,
+      "step": 5736
+    },
+    {
+      "epoch": 0.91792,
+      "grad_norm": 3.8906976778797375,
+      "learning_rate": 0.0001627436909954589,
+      "loss": 1.6587,
+      "step": 5737
+    },
+    {
+      "epoch": 0.91808,
+      "grad_norm": 2.3890169355052566,
+      "learning_rate": 0.00016273023948599355,
+      "loss": 1.6509,
+      "step": 5738
+    },
+    {
+      "epoch": 0.91824,
+      "grad_norm": 2.4632898322145547,
+      "learning_rate": 0.00016271678610474928,
+      "loss": 1.5931,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9184,
+      "grad_norm": 2.3710978336388417,
+      "learning_rate": 0.00016270333085212748,
+      "loss": 1.6238,
+      "step": 5740
+    },
+    {
+      "epoch": 0.91856,
+      "grad_norm": 2.487710728095529,
+      "learning_rate": 0.00016268987372852967,
+      "loss": 1.5909,
+      "step": 5741
+    },
+    {
+      "epoch": 0.91872,
+      "grad_norm": 2.8092589512553716,
+      "learning_rate": 0.0001626764147343574,
+      "loss": 1.573,
+      "step": 5742
+    },
+    {
+      "epoch": 0.91888,
+      "grad_norm": 2.6013241675292806,
+      "learning_rate": 0.00016266295387001221,
+      "loss": 1.5695,
+      "step": 5743
+    },
+    {
+      "epoch": 0.91904,
+      "grad_norm": 3.1035815121527004,
+      "learning_rate": 0.0001626494911358958,
+      "loss": 1.5856,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9192,
+      "grad_norm": 2.264160229550729,
+      "learning_rate": 0.00016263602653240983,
+      "loss": 1.5396,
+      "step": 5745
+    },
+    {
+      "epoch": 0.91936,
+      "grad_norm": 2.4541982142245167,
+      "learning_rate": 0.00016262256005995614,
+      "loss": 1.5858,
+      "step": 5746
+    },
+    {
+      "epoch": 0.91952,
+      "grad_norm": 2.5659003500102138,
+      "learning_rate": 0.00016260909171893648,
+      "loss": 1.6382,
+      "step": 5747
+    },
+    {
+      "epoch": 0.91968,
+      "grad_norm": 2.60571119697687,
+      "learning_rate": 0.00016259562150975279,
+      "loss": 1.6159,
+      "step": 5748
+    },
+    {
+      "epoch": 0.91984,
+      "grad_norm": 2.4321324943608107,
+      "learning_rate": 0.00016258214943280693,
+      "loss": 1.5563,
+      "step": 5749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 2.3568668782542246,
+      "learning_rate": 0.00016256867548850092,
+      "loss": 1.5603,
+      "step": 5750
+    },
+    {
+      "epoch": 0.92016,
+      "grad_norm": 2.7620804187368773,
+      "learning_rate": 0.00016255519967723684,
+      "loss": 1.6345,
+      "step": 5751
+    },
+    {
+      "epoch": 0.92032,
+      "grad_norm": 1.9842574603833678,
+      "learning_rate": 0.00016254172199941673,
+      "loss": 1.6258,
+      "step": 5752
+    },
+    {
+      "epoch": 0.92048,
+      "grad_norm": 2.3352157665313173,
+      "learning_rate": 0.0001625282424554428,
+      "loss": 1.583,
+      "step": 5753
+    },
+    {
+      "epoch": 0.92064,
+      "grad_norm": 2.4704546444356015,
+      "learning_rate": 0.0001625147610457172,
+      "loss": 1.5608,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9208,
+      "grad_norm": 2.3885456704908954,
+      "learning_rate": 0.00016250127777064222,
+      "loss": 1.6633,
+      "step": 5755
+    },
+    {
+      "epoch": 0.92096,
+      "grad_norm": 2.373840028437264,
+      "learning_rate": 0.00016248779263062016,
+      "loss": 1.6025,
+      "step": 5756
+    },
+    {
+      "epoch": 0.92112,
+      "grad_norm": 2.2396006747726527,
+      "learning_rate": 0.00016247430562605348,
+      "loss": 1.5798,
+      "step": 5757
+    },
+    {
+      "epoch": 0.92128,
+      "grad_norm": 2.487119000727088,
+      "learning_rate": 0.00016246081675734452,
+      "loss": 1.5857,
+      "step": 5758
+    },
+    {
+      "epoch": 0.92144,
+      "grad_norm": 2.2003089418659387,
+      "learning_rate": 0.00016244732602489583,
+      "loss": 1.6001,
+      "step": 5759
+    },
+    {
+      "epoch": 0.9216,
+      "grad_norm": 2.279446235132398,
+      "learning_rate": 0.00016243383342910993,
+      "loss": 1.6,
+      "step": 5760
+    },
+    {
+      "epoch": 0.92176,
+      "grad_norm": 2.4173774731187407,
+      "learning_rate": 0.00016242033897038936,
+      "loss": 1.7278,
+      "step": 5761
+    },
+    {
+      "epoch": 0.92192,
+      "grad_norm": 1.9784134814492864,
+      "learning_rate": 0.00016240684264913689,
+      "loss": 1.5836,
+      "step": 5762
+    },
+    {
+      "epoch": 0.92208,
+      "grad_norm": 2.528224029964821,
+      "learning_rate": 0.00016239334446575514,
+      "loss": 1.6147,
+      "step": 5763
+    },
+    {
+      "epoch": 0.92224,
+      "grad_norm": 2.4663421030506805,
+      "learning_rate": 0.00016237984442064694,
+      "loss": 1.5838,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9224,
+      "grad_norm": 2.966457144957844,
+      "learning_rate": 0.00016236634251421504,
+      "loss": 1.5642,
+      "step": 5765
+    },
+    {
+      "epoch": 0.92256,
+      "grad_norm": 2.5979690402652116,
+      "learning_rate": 0.00016235283874686238,
+      "loss": 1.6119,
+      "step": 5766
+    },
+    {
+      "epoch": 0.92272,
+      "grad_norm": 2.892908386359373,
+      "learning_rate": 0.00016233933311899184,
+      "loss": 1.6671,
+      "step": 5767
+    },
+    {
+      "epoch": 0.92288,
+      "grad_norm": 2.8741359682992105,
+      "learning_rate": 0.00016232582563100646,
+      "loss": 1.5804,
+      "step": 5768
+    },
+    {
+      "epoch": 0.92304,
+      "grad_norm": 3.0580782091013625,
+      "learning_rate": 0.00016231231628330926,
+      "loss": 1.6468,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9232,
+      "grad_norm": 2.4659188422497658,
+      "learning_rate": 0.00016229880507630338,
+      "loss": 1.5582,
+      "step": 5770
+    },
+    {
+      "epoch": 0.92336,
+      "grad_norm": 2.4287774290581905,
+      "learning_rate": 0.00016228529201039188,
+      "loss": 1.6612,
+      "step": 5771
+    },
+    {
+      "epoch": 0.92352,
+      "grad_norm": 2.3382365885860894,
+      "learning_rate": 0.00016227177708597802,
+      "loss": 1.5243,
+      "step": 5772
+    },
+    {
+      "epoch": 0.92368,
+      "grad_norm": 2.3100832711632053,
+      "learning_rate": 0.00016225826030346512,
+      "loss": 1.5955,
+      "step": 5773
+    },
+    {
+      "epoch": 0.92384,
+      "grad_norm": 2.7146006682028747,
+      "learning_rate": 0.00016224474166325642,
+      "loss": 1.6307,
+      "step": 5774
+    },
+    {
+      "epoch": 0.924,
+      "grad_norm": 2.241115696629725,
+      "learning_rate": 0.00016223122116575535,
+      "loss": 1.5171,
+      "step": 5775
+    },
+    {
+      "epoch": 0.92416,
+      "grad_norm": 3.6293123122667215,
+      "learning_rate": 0.0001622176988113653,
+      "loss": 1.5669,
+      "step": 5776
+    },
+    {
+      "epoch": 0.92432,
+      "grad_norm": 2.4773953744415507,
+      "learning_rate": 0.00016220417460048977,
+      "loss": 1.5436,
+      "step": 5777
+    },
+    {
+      "epoch": 0.92448,
+      "grad_norm": 2.770708380216715,
+      "learning_rate": 0.00016219064853353237,
+      "loss": 1.5636,
+      "step": 5778
+    },
+    {
+      "epoch": 0.92464,
+      "grad_norm": 2.473679714416167,
+      "learning_rate": 0.00016217712061089657,
+      "loss": 1.5774,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9248,
+      "grad_norm": 2.2593950356013326,
+      "learning_rate": 0.0001621635908329861,
+      "loss": 1.6407,
+      "step": 5780
+    },
+    {
+      "epoch": 0.92496,
+      "grad_norm": 2.325618580692458,
+      "learning_rate": 0.00016215005920020473,
+      "loss": 1.5854,
+      "step": 5781
+    },
+    {
+      "epoch": 0.92512,
+      "grad_norm": 2.165255518341222,
+      "learning_rate": 0.00016213652571295609,
+      "loss": 1.5515,
+      "step": 5782
+    },
+    {
+      "epoch": 0.92528,
+      "grad_norm": 2.770801023305391,
+      "learning_rate": 0.0001621229903716441,
+      "loss": 1.6392,
+      "step": 5783
+    },
+    {
+      "epoch": 0.92544,
+      "grad_norm": 2.274865235692356,
+      "learning_rate": 0.00016210945317667255,
+      "loss": 1.6809,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9256,
+      "grad_norm": 2.711157416362567,
+      "learning_rate": 0.00016209591412844545,
+      "loss": 1.6672,
+      "step": 5785
+    },
+    {
+      "epoch": 0.92576,
+      "grad_norm": 2.2137803168110644,
+      "learning_rate": 0.00016208237322736676,
+      "loss": 1.6273,
+      "step": 5786
+    },
+    {
+      "epoch": 0.92592,
+      "grad_norm": 2.617543854345977,
+      "learning_rate": 0.00016206883047384051,
+      "loss": 1.6549,
+      "step": 5787
+    },
+    {
+      "epoch": 0.92608,
+      "grad_norm": 1.9943303011937026,
+      "learning_rate": 0.00016205528586827082,
+      "loss": 1.5251,
+      "step": 5788
+    },
+    {
+      "epoch": 0.92624,
+      "grad_norm": 2.105720756675015,
+      "learning_rate": 0.0001620417394110618,
+      "loss": 1.5759,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9264,
+      "grad_norm": 2.5325412335435122,
+      "learning_rate": 0.00016202819110261769,
+      "loss": 1.6114,
+      "step": 5790
+    },
+    {
+      "epoch": 0.92656,
+      "grad_norm": 2.618185594305925,
+      "learning_rate": 0.00016201464094334274,
+      "loss": 1.5148,
+      "step": 5791
+    },
+    {
+      "epoch": 0.92672,
+      "grad_norm": 2.5867739476488176,
+      "learning_rate": 0.00016200108893364126,
+      "loss": 1.5443,
+      "step": 5792
+    },
+    {
+      "epoch": 0.92688,
+      "grad_norm": 3.414961874891047,
+      "learning_rate": 0.00016198753507391767,
+      "loss": 1.5646,
+      "step": 5793
+    },
+    {
+      "epoch": 0.92704,
+      "grad_norm": 3.4637629337272697,
+      "learning_rate": 0.00016197397936457632,
+      "loss": 1.5852,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9272,
+      "grad_norm": 2.6403571866738313,
+      "learning_rate": 0.00016196042180602173,
+      "loss": 1.5867,
+      "step": 5795
+    },
+    {
+      "epoch": 0.92736,
+      "grad_norm": 2.8767793324950075,
+      "learning_rate": 0.00016194686239865845,
+      "loss": 1.5905,
+      "step": 5796
+    },
+    {
+      "epoch": 0.92752,
+      "grad_norm": 2.9584864867523923,
+      "learning_rate": 0.0001619333011428911,
+      "loss": 1.5537,
+      "step": 5797
+    },
+    {
+      "epoch": 0.92768,
+      "grad_norm": 2.3755293196919576,
+      "learning_rate": 0.00016191973803912423,
+      "loss": 1.6948,
+      "step": 5798
+    },
+    {
+      "epoch": 0.92784,
+      "grad_norm": 3.646070567089244,
+      "learning_rate": 0.00016190617308776265,
+      "loss": 1.627,
+      "step": 5799
+    },
+    {
+      "epoch": 0.928,
+      "grad_norm": 2.351058913675742,
+      "learning_rate": 0.00016189260628921104,
+      "loss": 1.5398,
+      "step": 5800
+    },
+    {
+      "epoch": 0.92816,
+      "grad_norm": 3.604812283075802,
+      "learning_rate": 0.00016187903764387426,
+      "loss": 1.625,
+      "step": 5801
+    },
+    {
+      "epoch": 0.92832,
+      "grad_norm": 2.4882106765500436,
+      "learning_rate": 0.00016186546715215715,
+      "loss": 1.5949,
+      "step": 5802
+    },
+    {
+      "epoch": 0.92848,
+      "grad_norm": 3.325259459520293,
+      "learning_rate": 0.00016185189481446465,
+      "loss": 1.5366,
+      "step": 5803
+    },
+    {
+      "epoch": 0.92864,
+      "grad_norm": 2.4633875686150106,
+      "learning_rate": 0.00016183832063120174,
+      "loss": 1.5283,
+      "step": 5804
+    },
+    {
+      "epoch": 0.9288,
+      "grad_norm": 2.124069088083541,
+      "learning_rate": 0.00016182474460277346,
+      "loss": 1.5779,
+      "step": 5805
+    },
+    {
+      "epoch": 0.92896,
+      "grad_norm": 3.231258575767572,
+      "learning_rate": 0.00016181116672958488,
+      "loss": 1.5398,
+      "step": 5806
+    },
+    {
+      "epoch": 0.92912,
+      "grad_norm": 2.261463306239277,
+      "learning_rate": 0.00016179758701204114,
+      "loss": 1.5395,
+      "step": 5807
+    },
+    {
+      "epoch": 0.92928,
+      "grad_norm": 2.5343867261927087,
+      "learning_rate": 0.00016178400545054748,
+      "loss": 1.5567,
+      "step": 5808
+    },
+    {
+      "epoch": 0.92944,
+      "grad_norm": 2.345310691694251,
+      "learning_rate": 0.0001617704220455091,
+      "loss": 1.5379,
+      "step": 5809
+    },
+    {
+      "epoch": 0.9296,
+      "grad_norm": 2.5305410508830524,
+      "learning_rate": 0.00016175683679733137,
+      "loss": 1.5999,
+      "step": 5810
+    },
+    {
+      "epoch": 0.92976,
+      "grad_norm": 2.5877346825174445,
+      "learning_rate": 0.0001617432497064196,
+      "loss": 1.5044,
+      "step": 5811
+    },
+    {
+      "epoch": 0.92992,
+      "grad_norm": 2.4022686368574875,
+      "learning_rate": 0.00016172966077317922,
+      "loss": 1.611,
+      "step": 5812
+    },
+    {
+      "epoch": 0.93008,
+      "grad_norm": 2.6031914907500604,
+      "learning_rate": 0.00016171606999801572,
+      "loss": 1.5835,
+      "step": 5813
+    },
+    {
+      "epoch": 0.93024,
+      "grad_norm": 2.4687663681050322,
+      "learning_rate": 0.0001617024773813346,
+      "loss": 1.5575,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9304,
+      "grad_norm": 2.1746192716001493,
+      "learning_rate": 0.0001616888829235415,
+      "loss": 1.5405,
+      "step": 5815
+    },
+    {
+      "epoch": 0.93056,
+      "grad_norm": 2.174920030917632,
+      "learning_rate": 0.00016167528662504205,
+      "loss": 1.5337,
+      "step": 5816
+    },
+    {
+      "epoch": 0.93072,
+      "grad_norm": 2.5486232271908755,
+      "learning_rate": 0.00016166168848624182,
+      "loss": 1.6383,
+      "step": 5817
+    },
+    {
+      "epoch": 0.93088,
+      "grad_norm": 2.1813497782156825,
+      "learning_rate": 0.00016164808850754675,
+      "loss": 1.6311,
+      "step": 5818
+    },
+    {
+      "epoch": 0.93104,
+      "grad_norm": 2.03487712084022,
+      "learning_rate": 0.0001616344866893625,
+      "loss": 1.5092,
+      "step": 5819
+    },
+    {
+      "epoch": 0.9312,
+      "grad_norm": 2.222284409076628,
+      "learning_rate": 0.000161620883032095,
+      "loss": 1.6095,
+      "step": 5820
+    },
+    {
+      "epoch": 0.93136,
+      "grad_norm": 2.399542847481476,
+      "learning_rate": 0.00016160727753615012,
+      "loss": 1.6224,
+      "step": 5821
+    },
+    {
+      "epoch": 0.93152,
+      "grad_norm": 1.9558775705581246,
+      "learning_rate": 0.00016159367020193385,
+      "loss": 1.6008,
+      "step": 5822
+    },
+    {
+      "epoch": 0.93168,
+      "grad_norm": 2.2684961250394955,
+      "learning_rate": 0.00016158006102985224,
+      "loss": 1.5694,
+      "step": 5823
+    },
+    {
+      "epoch": 0.93184,
+      "grad_norm": 2.1362833484576167,
+      "learning_rate": 0.00016156645002031133,
+      "loss": 1.5558,
+      "step": 5824
+    },
+    {
+      "epoch": 0.932,
+      "grad_norm": 2.1598314032051102,
+      "learning_rate": 0.00016155283717371726,
+      "loss": 1.5238,
+      "step": 5825
+    },
+    {
+      "epoch": 0.93216,
+      "grad_norm": 2.1821826514190947,
+      "learning_rate": 0.0001615392224904762,
+      "loss": 1.5773,
+      "step": 5826
+    },
+    {
+      "epoch": 0.93232,
+      "grad_norm": 2.365216753049997,
+      "learning_rate": 0.00016152560597099445,
+      "loss": 1.5931,
+      "step": 5827
+    },
+    {
+      "epoch": 0.93248,
+      "grad_norm": 2.4959773285351727,
+      "learning_rate": 0.00016151198761567826,
+      "loss": 1.6942,
+      "step": 5828
+    },
+    {
+      "epoch": 0.93264,
+      "grad_norm": 2.973321687206325,
+      "learning_rate": 0.00016149836742493395,
+      "loss": 1.6383,
+      "step": 5829
+    },
+    {
+      "epoch": 0.9328,
+      "grad_norm": 2.638697038749973,
+      "learning_rate": 0.000161484745399168,
+      "loss": 1.5076,
+      "step": 5830
+    },
+    {
+      "epoch": 0.93296,
+      "grad_norm": 2.4185758781503077,
+      "learning_rate": 0.00016147112153878687,
+      "loss": 1.5291,
+      "step": 5831
+    },
+    {
+      "epoch": 0.93312,
+      "grad_norm": 2.0650688262530115,
+      "learning_rate": 0.000161457495844197,
+      "loss": 1.4968,
+      "step": 5832
+    },
+    {
+      "epoch": 0.93328,
+      "grad_norm": 2.0650688262530115,
+      "learning_rate": 0.000161457495844197,
+      "loss": 1.61,
+      "step": 5833
+    },
+    {
+      "epoch": 0.93344,
+      "grad_norm": 2.8062334047405235,
+      "learning_rate": 0.00016144386831580505,
+      "loss": 1.5984,
+      "step": 5834
+    },
+    {
+      "epoch": 0.9336,
+      "grad_norm": 2.286104732550117,
+      "learning_rate": 0.00016143023895401757,
+      "loss": 1.5878,
+      "step": 5835
+    },
+    {
+      "epoch": 0.93376,
+      "grad_norm": 2.5072608203995665,
+      "learning_rate": 0.0001614166077592413,
+      "loss": 1.5213,
+      "step": 5836
+    },
+    {
+      "epoch": 0.93392,
+      "grad_norm": 2.2878668500183217,
+      "learning_rate": 0.0001614029747318829,
+      "loss": 1.4707,
+      "step": 5837
+    },
+    {
+      "epoch": 0.93408,
+      "grad_norm": 2.2330786912313303,
+      "learning_rate": 0.00016138933987234927,
+      "loss": 1.5186,
+      "step": 5838
+    },
+    {
+      "epoch": 0.93424,
+      "grad_norm": 2.6589725165918265,
+      "learning_rate": 0.00016137570318104714,
+      "loss": 1.5205,
+      "step": 5839
+    },
+    {
+      "epoch": 0.9344,
+      "grad_norm": 1.9887255079765063,
+      "learning_rate": 0.00016136206465838346,
+      "loss": 1.5297,
+      "step": 5840
+    },
+    {
+      "epoch": 0.93456,
+      "grad_norm": 2.2015445590684197,
+      "learning_rate": 0.00016134842430476522,
+      "loss": 1.5242,
+      "step": 5841
+    },
+    {
+      "epoch": 0.93472,
+      "grad_norm": 2.3309919873023626,
+      "learning_rate": 0.00016133478212059936,
+      "loss": 1.4766,
+      "step": 5842
+    },
+    {
+      "epoch": 0.93488,
+      "grad_norm": 2.204912346702524,
+      "learning_rate": 0.000161321138106293,
+      "loss": 1.5581,
+      "step": 5843
+    },
+    {
+      "epoch": 0.93504,
+      "grad_norm": 2.1182972598469214,
+      "learning_rate": 0.00016130749226225322,
+      "loss": 1.6345,
+      "step": 5844
+    },
+    {
+      "epoch": 0.9352,
+      "grad_norm": 2.206076572151033,
+      "learning_rate": 0.00016129384458888723,
+      "loss": 1.5249,
+      "step": 5845
+    },
+    {
+      "epoch": 0.93536,
+      "grad_norm": 2.6066164413866755,
+      "learning_rate": 0.00016128019508660223,
+      "loss": 1.5646,
+      "step": 5846
+    },
+    {
+      "epoch": 0.93552,
+      "grad_norm": 2.0322765987087257,
+      "learning_rate": 0.00016126654375580546,
+      "loss": 1.5426,
+      "step": 5847
+    },
+    {
+      "epoch": 0.93568,
+      "grad_norm": 2.252371463112468,
+      "learning_rate": 0.00016125289059690432,
+      "loss": 1.5019,
+      "step": 5848
+    },
+    {
+      "epoch": 0.93584,
+      "grad_norm": 2.7384906669397773,
+      "learning_rate": 0.0001612392356103062,
+      "loss": 1.6122,
+      "step": 5849
+    },
+    {
+      "epoch": 0.936,
+      "grad_norm": 2.335693660880905,
+      "learning_rate": 0.00016122557879641848,
+      "loss": 1.6849,
+      "step": 5850
+    },
+    {
+      "epoch": 0.93616,
+      "grad_norm": 1.9772222829013977,
+      "learning_rate": 0.00016121192015564875,
+      "loss": 1.607,
+      "step": 5851
+    },
+    {
+      "epoch": 0.93632,
+      "grad_norm": 2.108013917726775,
+      "learning_rate": 0.00016119825968840453,
+      "loss": 1.5878,
+      "step": 5852
+    },
+    {
+      "epoch": 0.93648,
+      "grad_norm": 3.622502053173447,
+      "learning_rate": 0.00016118459739509336,
+      "loss": 1.5603,
+      "step": 5853
+    },
+    {
+      "epoch": 0.93664,
+      "grad_norm": 2.3849045508253233,
+      "learning_rate": 0.000161170933276123,
+      "loss": 1.6471,
+      "step": 5854
+    },
+    {
+      "epoch": 0.9368,
+      "grad_norm": 2.8982942765411948,
+      "learning_rate": 0.0001611572673319011,
+      "loss": 1.6105,
+      "step": 5855
+    },
+    {
+      "epoch": 0.93696,
+      "grad_norm": 2.245808719711051,
+      "learning_rate": 0.00016114359956283548,
+      "loss": 1.4821,
+      "step": 5856
+    },
+    {
+      "epoch": 0.93712,
+      "grad_norm": 2.473890757270457,
+      "learning_rate": 0.00016112992996933394,
+      "loss": 1.6328,
+      "step": 5857
+    },
+    {
+      "epoch": 0.93728,
+      "grad_norm": 2.10888426787845,
+      "learning_rate": 0.00016111625855180436,
+      "loss": 1.6393,
+      "step": 5858
+    },
+    {
+      "epoch": 0.93744,
+      "grad_norm": 2.086038347030604,
+      "learning_rate": 0.00016110258531065467,
+      "loss": 1.578,
+      "step": 5859
+    },
+    {
+      "epoch": 0.9376,
+      "grad_norm": 2.3676206481739364,
+      "learning_rate": 0.00016108891024629288,
+      "loss": 1.6034,
+      "step": 5860
+    },
+    {
+      "epoch": 0.93776,
+      "grad_norm": 2.410176351969036,
+      "learning_rate": 0.00016107523335912705,
+      "loss": 1.5466,
+      "step": 5861
+    },
+    {
+      "epoch": 0.93792,
+      "grad_norm": 1.9731909093135744,
+      "learning_rate": 0.0001610615546495652,
+      "loss": 1.6045,
+      "step": 5862
+    },
+    {
+      "epoch": 0.93808,
+      "grad_norm": 2.141119808098176,
+      "learning_rate": 0.0001610478741180156,
+      "loss": 1.5054,
+      "step": 5863
+    },
+    {
+      "epoch": 0.93824,
+      "grad_norm": 2.386144927404318,
+      "learning_rate": 0.0001610341917648863,
+      "loss": 1.5772,
+      "step": 5864
+    },
+    {
+      "epoch": 0.9384,
+      "grad_norm": 2.331902573943386,
+      "learning_rate": 0.00016102050759058575,
+      "loss": 1.5393,
+      "step": 5865
+    },
+    {
+      "epoch": 0.93856,
+      "grad_norm": 2.996315348872658,
+      "learning_rate": 0.0001610068215955221,
+      "loss": 1.6234,
+      "step": 5866
+    },
+    {
+      "epoch": 0.93872,
+      "grad_norm": 2.4687791348637402,
+      "learning_rate": 0.00016099313378010384,
+      "loss": 1.636,
+      "step": 5867
+    },
+    {
+      "epoch": 0.93888,
+      "grad_norm": 2.1746353719752074,
+      "learning_rate": 0.00016097944414473933,
+      "loss": 1.6166,
+      "step": 5868
+    },
+    {
+      "epoch": 0.93904,
+      "grad_norm": 2.0258676840822334,
+      "learning_rate": 0.00016096575268983708,
+      "loss": 1.5652,
+      "step": 5869
+    },
+    {
+      "epoch": 0.9392,
+      "grad_norm": 2.3531311428071935,
+      "learning_rate": 0.00016095205941580557,
+      "loss": 1.5626,
+      "step": 5870
+    },
+    {
+      "epoch": 0.93936,
+      "grad_norm": 2.3258421360879797,
+      "learning_rate": 0.00016093836432305342,
+      "loss": 1.6377,
+      "step": 5871
+    },
+    {
+      "epoch": 0.93952,
+      "grad_norm": 2.2844362206659015,
+      "learning_rate": 0.00016092466741198931,
+      "loss": 1.5539,
+      "step": 5872
+    },
+    {
+      "epoch": 0.93968,
+      "grad_norm": 2.2141444509453385,
+      "learning_rate": 0.00016091096868302187,
+      "loss": 1.5835,
+      "step": 5873
+    },
+    {
+      "epoch": 0.93984,
+      "grad_norm": 2.3009847801707695,
+      "learning_rate": 0.0001608972681365599,
+      "loss": 1.6597,
+      "step": 5874
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 2.6393215313055185,
+      "learning_rate": 0.0001608835657730122,
+      "loss": 1.5476,
+      "step": 5875
+    },
+    {
+      "epoch": 0.94016,
+      "grad_norm": 2.339516639556887,
+      "learning_rate": 0.00016086986159278757,
+      "loss": 1.5515,
+      "step": 5876
+    },
+    {
+      "epoch": 0.94032,
+      "grad_norm": 2.0982070486414868,
+      "learning_rate": 0.00016085615559629498,
+      "loss": 1.6009,
+      "step": 5877
+    },
+    {
+      "epoch": 0.94048,
+      "grad_norm": 2.647939625374388,
+      "learning_rate": 0.0001608424477839434,
+      "loss": 1.6203,
+      "step": 5878
+    },
+    {
+      "epoch": 0.94064,
+      "grad_norm": 2.2011952625325195,
+      "learning_rate": 0.00016082873815614183,
+      "loss": 1.597,
+      "step": 5879
+    },
+    {
+      "epoch": 0.9408,
+      "grad_norm": 2.321715987961197,
+      "learning_rate": 0.00016081502671329938,
+      "loss": 1.6018,
+      "step": 5880
+    },
+    {
+      "epoch": 0.94096,
+      "grad_norm": 3.0644319833799063,
+      "learning_rate": 0.0001608013134558251,
+      "loss": 1.7048,
+      "step": 5881
+    },
+    {
+      "epoch": 0.94112,
+      "grad_norm": 2.314962909019194,
+      "learning_rate": 0.00016078759838412824,
+      "loss": 1.6376,
+      "step": 5882
+    },
+    {
+      "epoch": 0.94128,
+      "grad_norm": 2.562820112543314,
+      "learning_rate": 0.00016077388149861804,
+      "loss": 1.5616,
+      "step": 5883
+    },
+    {
+      "epoch": 0.94144,
+      "grad_norm": 2.0731251263230854,
+      "learning_rate": 0.00016076016279970373,
+      "loss": 1.5068,
+      "step": 5884
+    },
+    {
+      "epoch": 0.9416,
+      "grad_norm": 2.627677622817891,
+      "learning_rate": 0.00016074644228779476,
+      "loss": 1.6727,
+      "step": 5885
+    },
+    {
+      "epoch": 0.94176,
+      "grad_norm": 2.3609336607309412,
+      "learning_rate": 0.00016073271996330042,
+      "loss": 1.6355,
+      "step": 5886
+    },
+    {
+      "epoch": 0.94192,
+      "grad_norm": 2.419520357923835,
+      "learning_rate": 0.00016071899582663024,
+      "loss": 1.4938,
+      "step": 5887
+    },
+    {
+      "epoch": 0.94208,
+      "grad_norm": 3.957445587132795,
+      "learning_rate": 0.00016070526987819369,
+      "loss": 1.6338,
+      "step": 5888
+    },
+    {
+      "epoch": 0.94224,
+      "grad_norm": 2.855076202276163,
+      "learning_rate": 0.00016069154211840035,
+      "loss": 1.6281,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9424,
+      "grad_norm": 1.9738079096789352,
+      "learning_rate": 0.00016067781254765985,
+      "loss": 1.523,
+      "step": 5890
+    },
+    {
+      "epoch": 0.94256,
+      "grad_norm": 2.0926123573345197,
+      "learning_rate": 0.00016066408116638175,
+      "loss": 1.537,
+      "step": 5891
+    },
+    {
+      "epoch": 0.94272,
+      "grad_norm": 2.2924383713103955,
+      "learning_rate": 0.00016065034797497596,
+      "loss": 1.5718,
+      "step": 5892
+    },
+    {
+      "epoch": 0.94288,
+      "grad_norm": 2.1496804619226473,
+      "learning_rate": 0.00016063661297385212,
+      "loss": 1.4952,
+      "step": 5893
+    },
+    {
+      "epoch": 0.94304,
+      "grad_norm": 2.111678443547665,
+      "learning_rate": 0.00016062287616342013,
+      "loss": 1.4942,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9432,
+      "grad_norm": 2.4684554530556464,
+      "learning_rate": 0.00016060913754408984,
+      "loss": 1.5876,
+      "step": 5895
+    },
+    {
+      "epoch": 0.94336,
+      "grad_norm": 2.2787558402318013,
+      "learning_rate": 0.0001605953971162712,
+      "loss": 1.5181,
+      "step": 5896
+    },
+    {
+      "epoch": 0.94352,
+      "grad_norm": 2.4535324611922773,
+      "learning_rate": 0.0001605816548803742,
+      "loss": 1.5451,
+      "step": 5897
+    },
+    {
+      "epoch": 0.94368,
+      "grad_norm": 2.3923125310957425,
+      "learning_rate": 0.00016056791083680894,
+      "loss": 1.5534,
+      "step": 5898
+    },
+    {
+      "epoch": 0.94384,
+      "grad_norm": 2.5014072100600835,
+      "learning_rate": 0.0001605541649859854,
+      "loss": 1.5102,
+      "step": 5899
+    },
+    {
+      "epoch": 0.944,
+      "grad_norm": 2.256468050305789,
+      "learning_rate": 0.0001605404173283139,
+      "loss": 1.5677,
+      "step": 5900
+    },
+    {
+      "epoch": 0.94416,
+      "grad_norm": 2.5414545952211745,
+      "learning_rate": 0.0001605266678642045,
+      "loss": 1.5966,
+      "step": 5901
+    },
+    {
+      "epoch": 0.94432,
+      "grad_norm": 2.646216814248704,
+      "learning_rate": 0.00016051291659406754,
+      "loss": 1.561,
+      "step": 5902
+    },
+    {
+      "epoch": 0.94448,
+      "grad_norm": 2.5465338541823157,
+      "learning_rate": 0.00016049916351831335,
+      "loss": 1.6006,
+      "step": 5903
+    },
+    {
+      "epoch": 0.94464,
+      "grad_norm": 3.2934706057755863,
+      "learning_rate": 0.00016048540863735226,
+      "loss": 1.5418,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9448,
+      "grad_norm": 2.1155941143440953,
+      "learning_rate": 0.0001604716519515947,
+      "loss": 1.5813,
+      "step": 5905
+    },
+    {
+      "epoch": 0.94496,
+      "grad_norm": 3.7228580575231356,
+      "learning_rate": 0.0001604578934614512,
+      "loss": 1.5587,
+      "step": 5906
+    },
+    {
+      "epoch": 0.94512,
+      "grad_norm": 2.7368711671318975,
+      "learning_rate": 0.00016044413316733223,
+      "loss": 1.583,
+      "step": 5907
+    },
+    {
+      "epoch": 0.94528,
+      "grad_norm": 2.4431803035066664,
+      "learning_rate": 0.00016043037106964838,
+      "loss": 1.5983,
+      "step": 5908
+    },
+    {
+      "epoch": 0.94544,
+      "grad_norm": 3.19242111995109,
+      "learning_rate": 0.00016041660716881035,
+      "loss": 1.5185,
+      "step": 5909
+    },
+    {
+      "epoch": 0.9456,
+      "grad_norm": 2.2931302116859884,
+      "learning_rate": 0.0001604028414652288,
+      "loss": 1.5857,
+      "step": 5910
+    },
+    {
+      "epoch": 0.94576,
+      "grad_norm": 2.6618115554968234,
+      "learning_rate": 0.00016038907395931443,
+      "loss": 1.4505,
+      "step": 5911
+    },
+    {
+      "epoch": 0.94592,
+      "grad_norm": 2.320574512468053,
+      "learning_rate": 0.00016037530465147815,
+      "loss": 1.5267,
+      "step": 5912
+    },
+    {
+      "epoch": 0.94608,
+      "grad_norm": 2.6649239161790543,
+      "learning_rate": 0.00016036153354213073,
+      "loss": 1.517,
+      "step": 5913
+    },
+    {
+      "epoch": 0.94624,
+      "grad_norm": 2.527785419101503,
+      "learning_rate": 0.00016034776063168307,
+      "loss": 1.5069,
+      "step": 5914
+    },
+    {
+      "epoch": 0.9464,
+      "grad_norm": 2.946040428450927,
+      "learning_rate": 0.00016033398592054624,
+      "loss": 1.5143,
+      "step": 5915
+    },
+    {
+      "epoch": 0.94656,
+      "grad_norm": 2.902555719339784,
+      "learning_rate": 0.00016032020940913113,
+      "loss": 1.6024,
+      "step": 5916
+    },
+    {
+      "epoch": 0.94672,
+      "grad_norm": 2.3528610342470704,
+      "learning_rate": 0.00016030643109784889,
+      "loss": 1.672,
+      "step": 5917
+    },
+    {
+      "epoch": 0.94688,
+      "grad_norm": 2.895659040601594,
+      "learning_rate": 0.00016029265098711063,
+      "loss": 1.5336,
+      "step": 5918
+    },
+    {
+      "epoch": 0.94704,
+      "grad_norm": 2.3938705738142474,
+      "learning_rate": 0.0001602788690773275,
+      "loss": 1.5237,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9472,
+      "grad_norm": 2.595983701773291,
+      "learning_rate": 0.00016026508536891076,
+      "loss": 1.579,
+      "step": 5920
+    },
+    {
+      "epoch": 0.94736,
+      "grad_norm": 2.25179974001111,
+      "learning_rate": 0.00016025129986227169,
+      "loss": 1.6624,
+      "step": 5921
+    },
+    {
+      "epoch": 0.94752,
+      "grad_norm": 2.581228750106199,
+      "learning_rate": 0.0001602375125578216,
+      "loss": 1.5656,
+      "step": 5922
+    },
+    {
+      "epoch": 0.94768,
+      "grad_norm": 2.4580275836032164,
+      "learning_rate": 0.00016022372345597194,
+      "loss": 1.4678,
+      "step": 5923
+    },
+    {
+      "epoch": 0.94784,
+      "grad_norm": 2.765223274370002,
+      "learning_rate": 0.00016020993255713413,
+      "loss": 1.6229,
+      "step": 5924
+    },
+    {
+      "epoch": 0.948,
+      "grad_norm": 2.1142831808004297,
+      "learning_rate": 0.00016019613986171965,
+      "loss": 1.5809,
+      "step": 5925
+    },
+    {
+      "epoch": 0.94816,
+      "grad_norm": 2.815393126087753,
+      "learning_rate": 0.00016018234537014007,
+      "loss": 1.5576,
+      "step": 5926
+    },
+    {
+      "epoch": 0.94832,
+      "grad_norm": 2.1568703431453367,
+      "learning_rate": 0.000160168549082807,
+      "loss": 1.5554,
+      "step": 5927
+    },
+    {
+      "epoch": 0.94848,
+      "grad_norm": 2.281852020480492,
+      "learning_rate": 0.00016015475100013214,
+      "loss": 1.6755,
+      "step": 5928
+    },
+    {
+      "epoch": 0.94864,
+      "grad_norm": 2.1243491956475173,
+      "learning_rate": 0.00016014095112252711,
+      "loss": 1.5416,
+      "step": 5929
+    },
+    {
+      "epoch": 0.9488,
+      "grad_norm": 2.0243352372586276,
+      "learning_rate": 0.00016012714945040377,
+      "loss": 1.5846,
+      "step": 5930
+    },
+    {
+      "epoch": 0.94896,
+      "grad_norm": 2.144371473716262,
+      "learning_rate": 0.00016011334598417388,
+      "loss": 1.5945,
+      "step": 5931
+    },
+    {
+      "epoch": 0.94912,
+      "grad_norm": 2.4049330935564246,
+      "learning_rate": 0.00016009954072424937,
+      "loss": 1.5662,
+      "step": 5932
+    },
+    {
+      "epoch": 0.94928,
+      "grad_norm": 2.1699716706087195,
+      "learning_rate": 0.00016008573367104213,
+      "loss": 1.6134,
+      "step": 5933
+    },
+    {
+      "epoch": 0.94944,
+      "grad_norm": 1.996157804558877,
+      "learning_rate": 0.00016007192482496413,
+      "loss": 1.5984,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9496,
+      "grad_norm": 2.1847614964371647,
+      "learning_rate": 0.00016005811418642746,
+      "loss": 1.5479,
+      "step": 5935
+    },
+    {
+      "epoch": 0.94976,
+      "grad_norm": 2.793369673052034,
+      "learning_rate": 0.00016004430175584416,
+      "loss": 1.5934,
+      "step": 5936
+    },
+    {
+      "epoch": 0.94992,
+      "grad_norm": 2.5420546082531312,
+      "learning_rate": 0.00016003048753362637,
+      "loss": 1.5693,
+      "step": 5937
+    },
+    {
+      "epoch": 0.95008,
+      "grad_norm": 2.3603216445517847,
+      "learning_rate": 0.00016001667152018632,
+      "loss": 1.5607,
+      "step": 5938
+    },
+    {
+      "epoch": 0.95024,
+      "grad_norm": 2.3853061638820887,
+      "learning_rate": 0.00016000285371593626,
+      "loss": 1.6054,
+      "step": 5939
+    },
+    {
+      "epoch": 0.9504,
+      "grad_norm": 2.534233200203288,
+      "learning_rate": 0.00015998903412128848,
+      "loss": 1.695,
+      "step": 5940
+    },
+    {
+      "epoch": 0.95056,
+      "grad_norm": 2.4278283820404503,
+      "learning_rate": 0.0001599752127366553,
+      "loss": 1.5398,
+      "step": 5941
+    },
+    {
+      "epoch": 0.95072,
+      "grad_norm": 2.208829681393267,
+      "learning_rate": 0.00015996138956244922,
+      "loss": 1.6556,
+      "step": 5942
+    },
+    {
+      "epoch": 0.95088,
+      "grad_norm": 2.346788231472129,
+      "learning_rate": 0.0001599475645990826,
+      "loss": 1.5309,
+      "step": 5943
+    },
+    {
+      "epoch": 0.95104,
+      "grad_norm": 2.022347574519781,
+      "learning_rate": 0.00015993373784696803,
+      "loss": 1.4775,
+      "step": 5944
+    },
+    {
+      "epoch": 0.9512,
+      "grad_norm": 2.184083640961253,
+      "learning_rate": 0.00015991990930651803,
+      "loss": 1.6428,
+      "step": 5945
+    },
+    {
+      "epoch": 0.95136,
+      "grad_norm": 2.6057357459824884,
+      "learning_rate": 0.00015990607897814526,
+      "loss": 1.5254,
+      "step": 5946
+    },
+    {
+      "epoch": 0.95152,
+      "grad_norm": 1.9385268427648157,
+      "learning_rate": 0.0001598922468622624,
+      "loss": 1.6343,
+      "step": 5947
+    },
+    {
+      "epoch": 0.95168,
+      "grad_norm": 2.728360638681898,
+      "learning_rate": 0.00015987841295928216,
+      "loss": 1.4363,
+      "step": 5948
+    },
+    {
+      "epoch": 0.95184,
+      "grad_norm": 2.3604092219530326,
+      "learning_rate": 0.00015986457726961732,
+      "loss": 1.5718,
+      "step": 5949
+    },
+    {
+      "epoch": 0.952,
+      "grad_norm": 2.14929360343853,
+      "learning_rate": 0.00015985073979368072,
+      "loss": 1.5611,
+      "step": 5950
+    },
+    {
+      "epoch": 0.95216,
+      "grad_norm": 2.456075729413646,
+      "learning_rate": 0.0001598369005318853,
+      "loss": 1.5701,
+      "step": 5951
+    },
+    {
+      "epoch": 0.95232,
+      "grad_norm": 2.357749367726095,
+      "learning_rate": 0.00015982305948464393,
+      "loss": 1.521,
+      "step": 5952
+    },
+    {
+      "epoch": 0.95248,
+      "grad_norm": 2.2587253747478613,
+      "learning_rate": 0.00015980921665236962,
+      "loss": 1.5747,
+      "step": 5953
+    },
+    {
+      "epoch": 0.95264,
+      "grad_norm": 2.1707798016900193,
+      "learning_rate": 0.00015979537203547546,
+      "loss": 1.5816,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9528,
+      "grad_norm": 2.27120755953106,
+      "learning_rate": 0.00015978152563437452,
+      "loss": 1.5013,
+      "step": 5955
+    },
+    {
+      "epoch": 0.95296,
+      "grad_norm": 2.2871785332847994,
+      "learning_rate": 0.00015976767744947995,
+      "loss": 1.6197,
+      "step": 5956
+    },
+    {
+      "epoch": 0.95312,
+      "grad_norm": 2.4007689335860065,
+      "learning_rate": 0.000159753827481205,
+      "loss": 1.5819,
+      "step": 5957
+    },
+    {
+      "epoch": 0.95328,
+      "grad_norm": 2.933789028783556,
+      "learning_rate": 0.00015973997572996291,
+      "loss": 1.5806,
+      "step": 5958
+    },
+    {
+      "epoch": 0.95344,
+      "grad_norm": 2.7139237432036456,
+      "learning_rate": 0.00015972612219616698,
+      "loss": 1.6465,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9536,
+      "grad_norm": 2.2247393008016245,
+      "learning_rate": 0.0001597122668802306,
+      "loss": 1.4996,
+      "step": 5960
+    },
+    {
+      "epoch": 0.95376,
+      "grad_norm": 2.458145237047727,
+      "learning_rate": 0.00015969840978256717,
+      "loss": 1.483,
+      "step": 5961
+    },
+    {
+      "epoch": 0.95392,
+      "grad_norm": 2.3062020145180173,
+      "learning_rate": 0.00015968455090359021,
+      "loss": 1.559,
+      "step": 5962
+    },
+    {
+      "epoch": 0.95408,
+      "grad_norm": 2.3154445816560627,
+      "learning_rate": 0.0001596706902437132,
+      "loss": 1.5245,
+      "step": 5963
+    },
+    {
+      "epoch": 0.95424,
+      "grad_norm": 2.0877370772244053,
+      "learning_rate": 0.00015965682780334976,
+      "loss": 1.5932,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9544,
+      "grad_norm": 2.7397196105379185,
+      "learning_rate": 0.00015964296358291348,
+      "loss": 1.5117,
+      "step": 5965
+    },
+    {
+      "epoch": 0.95456,
+      "grad_norm": 2.406729546118435,
+      "learning_rate": 0.00015962909758281813,
+      "loss": 1.6112,
+      "step": 5966
+    },
+    {
+      "epoch": 0.95472,
+      "grad_norm": 2.35981366534359,
+      "learning_rate": 0.00015961522980347735,
+      "loss": 1.5834,
+      "step": 5967
+    },
+    {
+      "epoch": 0.95488,
+      "grad_norm": 2.475043176029115,
+      "learning_rate": 0.000159601360245305,
+      "loss": 1.5832,
+      "step": 5968
+    },
+    {
+      "epoch": 0.95504,
+      "grad_norm": 2.4080873445691804,
+      "learning_rate": 0.00015958748890871488,
+      "loss": 1.5824,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9552,
+      "grad_norm": 1.9966438945777525,
+      "learning_rate": 0.00015957361579412095,
+      "loss": 1.5062,
+      "step": 5970
+    },
+    {
+      "epoch": 0.95536,
+      "grad_norm": 2.996489907820325,
+      "learning_rate": 0.00015955974090193711,
+      "loss": 1.5337,
+      "step": 5971
+    },
+    {
+      "epoch": 0.95552,
+      "grad_norm": 2.2184736591865093,
+      "learning_rate": 0.00015954586423257743,
+      "loss": 1.561,
+      "step": 5972
+    },
+    {
+      "epoch": 0.95568,
+      "grad_norm": 2.7064793944505525,
+      "learning_rate": 0.00015953198578645588,
+      "loss": 1.521,
+      "step": 5973
+    },
+    {
+      "epoch": 0.95584,
+      "grad_norm": 1.992355306179698,
+      "learning_rate": 0.00015951810556398663,
+      "loss": 1.5279,
+      "step": 5974
+    },
+    {
+      "epoch": 0.956,
+      "grad_norm": 3.597378259473022,
+      "learning_rate": 0.00015950422356558386,
+      "loss": 1.694,
+      "step": 5975
+    },
+    {
+      "epoch": 0.95616,
+      "grad_norm": 2.1112497173557436,
+      "learning_rate": 0.00015949033979166175,
+      "loss": 1.5251,
+      "step": 5976
+    },
+    {
+      "epoch": 0.95632,
+      "grad_norm": 2.285141066825282,
+      "learning_rate": 0.00015947645424263458,
+      "loss": 1.49,
+      "step": 5977
+    },
+    {
+      "epoch": 0.95648,
+      "grad_norm": 2.5574040569437364,
+      "learning_rate": 0.0001594625669189167,
+      "loss": 1.5498,
+      "step": 5978
+    },
+    {
+      "epoch": 0.95664,
+      "grad_norm": 2.7330425394004987,
+      "learning_rate": 0.00015944867782092246,
+      "loss": 1.5245,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9568,
+      "grad_norm": 2.273294368599326,
+      "learning_rate": 0.00015943478694906631,
+      "loss": 1.483,
+      "step": 5980
+    },
+    {
+      "epoch": 0.95696,
+      "grad_norm": 2.171766864525958,
+      "learning_rate": 0.0001594208943037627,
+      "loss": 1.6126,
+      "step": 5981
+    },
+    {
+      "epoch": 0.95712,
+      "grad_norm": 2.3031214233515676,
+      "learning_rate": 0.0001594069998854262,
+      "loss": 1.6038,
+      "step": 5982
+    },
+    {
+      "epoch": 0.95728,
+      "grad_norm": 2.1613220164287927,
+      "learning_rate": 0.0001593931036944714,
+      "loss": 1.6187,
+      "step": 5983
+    },
+    {
+      "epoch": 0.95744,
+      "grad_norm": 2.372825611636189,
+      "learning_rate": 0.00015937920573131292,
+      "loss": 1.4763,
+      "step": 5984
+    },
+    {
+      "epoch": 0.9576,
+      "grad_norm": 2.7536724043076273,
+      "learning_rate": 0.00015936530599636547,
+      "loss": 1.6705,
+      "step": 5985
+    },
+    {
+      "epoch": 0.95776,
+      "grad_norm": 2.172219478382437,
+      "learning_rate": 0.0001593514044900438,
+      "loss": 1.621,
+      "step": 5986
+    },
+    {
+      "epoch": 0.95792,
+      "grad_norm": 2.8525305591430064,
+      "learning_rate": 0.0001593375012127627,
+      "loss": 1.6242,
+      "step": 5987
+    },
+    {
+      "epoch": 0.95808,
+      "grad_norm": 2.654977631734892,
+      "learning_rate": 0.00015932359616493703,
+      "loss": 1.6501,
+      "step": 5988
+    },
+    {
+      "epoch": 0.95824,
+      "grad_norm": 3.1408824716330304,
+      "learning_rate": 0.0001593096893469817,
+      "loss": 1.5862,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9584,
+      "grad_norm": 2.133385847496723,
+      "learning_rate": 0.0001592957807593116,
+      "loss": 1.6458,
+      "step": 5990
+    },
+    {
+      "epoch": 0.95856,
+      "grad_norm": 2.779149709958451,
+      "learning_rate": 0.00015928187040234188,
+      "loss": 1.5859,
+      "step": 5991
+    },
+    {
+      "epoch": 0.95872,
+      "grad_norm": 2.213660024388477,
+      "learning_rate": 0.0001592679582764875,
+      "loss": 1.6188,
+      "step": 5992
+    },
+    {
+      "epoch": 0.95888,
+      "grad_norm": 2.9166795842573485,
+      "learning_rate": 0.00015925404438216364,
+      "loss": 1.7073,
+      "step": 5993
+    },
+    {
+      "epoch": 0.95904,
+      "grad_norm": 2.6145420048018573,
+      "learning_rate": 0.0001592401287197854,
+      "loss": 1.6124,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9592,
+      "grad_norm": 2.1173692435457343,
+      "learning_rate": 0.00015922621128976804,
+      "loss": 1.4731,
+      "step": 5995
+    },
+    {
+      "epoch": 0.95936,
+      "grad_norm": 2.6779146357355468,
+      "learning_rate": 0.0001592122920925269,
+      "loss": 1.5866,
+      "step": 5996
+    },
+    {
+      "epoch": 0.95952,
+      "grad_norm": 2.416452752319544,
+      "learning_rate": 0.00015919837112847714,
+      "loss": 1.57,
+      "step": 5997
+    },
+    {
+      "epoch": 0.95968,
+      "grad_norm": 2.400889092197906,
+      "learning_rate": 0.00015918444839803432,
+      "loss": 1.484,
+      "step": 5998
+    },
+    {
+      "epoch": 0.95984,
+      "grad_norm": 2.7191341324203844,
+      "learning_rate": 0.00015917052390161378,
+      "loss": 1.5657,
+      "step": 5999
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 2.3010511295879987,
+      "learning_rate": 0.000159156597639631,
+      "loss": 1.4872,
+      "step": 6000
+    },
+    {
+      "epoch": 0.96016,
+      "grad_norm": 2.7928713695265617,
+      "learning_rate": 0.00015914266961250163,
+      "loss": 1.5789,
+      "step": 6001
+    },
+    {
+      "epoch": 0.96032,
+      "grad_norm": 2.466375992760688,
+      "learning_rate": 0.00015912873982064108,
+      "loss": 1.5456,
+      "step": 6002
+    },
+    {
+      "epoch": 0.96048,
+      "grad_norm": 2.2038918622617705,
+      "learning_rate": 0.00015911480826446517,
+      "loss": 1.5422,
+      "step": 6003
+    },
+    {
+      "epoch": 0.96064,
+      "grad_norm": 2.8099598225606566,
+      "learning_rate": 0.0001591008749443895,
+      "loss": 1.4979,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9608,
+      "grad_norm": 2.4482404003823417,
+      "learning_rate": 0.0001590869398608298,
+      "loss": 1.5671,
+      "step": 6005
+    },
+    {
+      "epoch": 0.96096,
+      "grad_norm": 2.5118168207266827,
+      "learning_rate": 0.00015907300301420195,
+      "loss": 1.5679,
+      "step": 6006
+    },
+    {
+      "epoch": 0.96112,
+      "grad_norm": 2.450680228169462,
+      "learning_rate": 0.00015905906440492177,
+      "loss": 1.5607,
+      "step": 6007
+    },
+    {
+      "epoch": 0.96128,
+      "grad_norm": 2.1701647368974593,
+      "learning_rate": 0.00015904512403340518,
+      "loss": 1.5773,
+      "step": 6008
+    },
+    {
+      "epoch": 0.96144,
+      "grad_norm": 2.550527618097877,
+      "learning_rate": 0.0001590311819000681,
+      "loss": 1.5903,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9616,
+      "grad_norm": 2.03563525024665,
+      "learning_rate": 0.0001590172380053266,
+      "loss": 1.5355,
+      "step": 6010
+    },
+    {
+      "epoch": 0.96176,
+      "grad_norm": 1.9508427296313395,
+      "learning_rate": 0.00015900329234959666,
+      "loss": 1.4787,
+      "step": 6011
+    },
+    {
+      "epoch": 0.96192,
+      "grad_norm": 2.1154764472017997,
+      "learning_rate": 0.00015898934493329447,
+      "loss": 1.7008,
+      "step": 6012
+    },
+    {
+      "epoch": 0.96208,
+      "grad_norm": 2.182270747423967,
+      "learning_rate": 0.0001589753957568362,
+      "loss": 1.5802,
+      "step": 6013
+    },
+    {
+      "epoch": 0.96224,
+      "grad_norm": 1.9845771442393492,
+      "learning_rate": 0.00015896144482063806,
+      "loss": 1.5582,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9624,
+      "grad_norm": 2.203330025854076,
+      "learning_rate": 0.00015894749212511631,
+      "loss": 1.5969,
+      "step": 6015
+    },
+    {
+      "epoch": 0.96256,
+      "grad_norm": 2.245912732923343,
+      "learning_rate": 0.00015893353767068731,
+      "loss": 1.5156,
+      "step": 6016
+    },
+    {
+      "epoch": 0.96272,
+      "grad_norm": 1.8984328723406363,
+      "learning_rate": 0.0001589195814577674,
+      "loss": 1.4268,
+      "step": 6017
+    },
+    {
+      "epoch": 0.96288,
+      "grad_norm": 2.0884063981574883,
+      "learning_rate": 0.00015890562348677304,
+      "loss": 1.5817,
+      "step": 6018
+    },
+    {
+      "epoch": 0.96304,
+      "grad_norm": 2.1616012825805555,
+      "learning_rate": 0.0001588916637581207,
+      "loss": 1.5524,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9632,
+      "grad_norm": 2.0897386942078384,
+      "learning_rate": 0.00015887770227222697,
+      "loss": 1.6051,
+      "step": 6020
+    },
+    {
+      "epoch": 0.96336,
+      "grad_norm": 2.668812537293966,
+      "learning_rate": 0.00015886373902950838,
+      "loss": 1.5693,
+      "step": 6021
+    },
+    {
+      "epoch": 0.96352,
+      "grad_norm": 2.3217529998507382,
+      "learning_rate": 0.00015884977403038156,
+      "loss": 1.5314,
+      "step": 6022
+    },
+    {
+      "epoch": 0.96368,
+      "grad_norm": 2.168384017699723,
+      "learning_rate": 0.00015883580727526326,
+      "loss": 1.4982,
+      "step": 6023
+    },
+    {
+      "epoch": 0.96384,
+      "grad_norm": 2.162378243207734,
+      "learning_rate": 0.00015882183876457019,
+      "loss": 1.5683,
+      "step": 6024
+    },
+    {
+      "epoch": 0.964,
+      "grad_norm": 2.0722996386226136,
+      "learning_rate": 0.0001588078684987192,
+      "loss": 1.5982,
+      "step": 6025
+    },
+    {
+      "epoch": 0.96416,
+      "grad_norm": 2.1957600845259178,
+      "learning_rate": 0.00015879389647812707,
+      "loss": 1.5813,
+      "step": 6026
+    },
+    {
+      "epoch": 0.96432,
+      "grad_norm": 2.396560674422256,
+      "learning_rate": 0.00015877992270321075,
+      "loss": 1.5381,
+      "step": 6027
+    },
+    {
+      "epoch": 0.96448,
+      "grad_norm": 1.971745520661029,
+      "learning_rate": 0.00015876594717438718,
+      "loss": 1.479,
+      "step": 6028
+    },
+    {
+      "epoch": 0.96464,
+      "grad_norm": 2.170011852676434,
+      "learning_rate": 0.00015875196989207336,
+      "loss": 1.5385,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9648,
+      "grad_norm": 2.3187690751871703,
+      "learning_rate": 0.0001587379908566864,
+      "loss": 1.4893,
+      "step": 6030
+    },
+    {
+      "epoch": 0.96496,
+      "grad_norm": 2.16015476032124,
+      "learning_rate": 0.00015872401006864338,
+      "loss": 1.5806,
+      "step": 6031
+    },
+    {
+      "epoch": 0.96512,
+      "grad_norm": 2.045640464880044,
+      "learning_rate": 0.00015871002752836148,
+      "loss": 1.6148,
+      "step": 6032
+    },
+    {
+      "epoch": 0.96528,
+      "grad_norm": 2.23135196395457,
+      "learning_rate": 0.00015869604323625792,
+      "loss": 1.5653,
+      "step": 6033
+    },
+    {
+      "epoch": 0.96544,
+      "grad_norm": 2.1460050732828346,
+      "learning_rate": 0.0001586820571927499,
+      "loss": 1.5266,
+      "step": 6034
+    },
+    {
+      "epoch": 0.9656,
+      "grad_norm": 2.5902711439347255,
+      "learning_rate": 0.00015866806939825485,
+      "loss": 1.6344,
+      "step": 6035
+    },
+    {
+      "epoch": 0.96576,
+      "grad_norm": 2.3810905998568663,
+      "learning_rate": 0.00015865407985319008,
+      "loss": 1.5182,
+      "step": 6036
+    },
+    {
+      "epoch": 0.96592,
+      "grad_norm": 2.514666896712966,
+      "learning_rate": 0.00015864008855797303,
+      "loss": 1.5224,
+      "step": 6037
+    },
+    {
+      "epoch": 0.96608,
+      "grad_norm": 2.286473789229095,
+      "learning_rate": 0.00015862609551302122,
+      "loss": 1.5044,
+      "step": 6038
+    },
+    {
+      "epoch": 0.96624,
+      "grad_norm": 2.290127951780573,
+      "learning_rate": 0.00015861210071875208,
+      "loss": 1.5835,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9664,
+      "grad_norm": 2.2975329043596737,
+      "learning_rate": 0.00015859810417558328,
+      "loss": 1.5891,
+      "step": 6040
+    },
+    {
+      "epoch": 0.96656,
+      "grad_norm": 3.0577281571417125,
+      "learning_rate": 0.00015858410588393248,
+      "loss": 1.6476,
+      "step": 6041
+    },
+    {
+      "epoch": 0.96672,
+      "grad_norm": 2.427618312553719,
+      "learning_rate": 0.0001585701058442173,
+      "loss": 1.5362,
+      "step": 6042
+    },
+    {
+      "epoch": 0.96688,
+      "grad_norm": 2.4867732157325726,
+      "learning_rate": 0.00015855610405685552,
+      "loss": 1.5224,
+      "step": 6043
+    },
+    {
+      "epoch": 0.96704,
+      "grad_norm": 2.399323972293842,
+      "learning_rate": 0.00015854210052226492,
+      "loss": 1.6487,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9672,
+      "grad_norm": 1.857680449888427,
+      "learning_rate": 0.00015852809524086333,
+      "loss": 1.5496,
+      "step": 6045
+    },
+    {
+      "epoch": 0.96736,
+      "grad_norm": 2.0203227288497008,
+      "learning_rate": 0.00015851408821306867,
+      "loss": 1.5263,
+      "step": 6046
+    },
+    {
+      "epoch": 0.96752,
+      "grad_norm": 2.464130779985364,
+      "learning_rate": 0.0001585000794392989,
+      "loss": 1.5281,
+      "step": 6047
+    },
+    {
+      "epoch": 0.96768,
+      "grad_norm": 2.1842998629911534,
+      "learning_rate": 0.000158486068919972,
+      "loss": 1.5701,
+      "step": 6048
+    },
+    {
+      "epoch": 0.96784,
+      "grad_norm": 2.089129862489139,
+      "learning_rate": 0.000158472056655506,
+      "loss": 1.5607,
+      "step": 6049
+    },
+    {
+      "epoch": 0.968,
+      "grad_norm": 2.3189466748088625,
+      "learning_rate": 0.00015845804264631906,
+      "loss": 1.5112,
+      "step": 6050
+    },
+    {
+      "epoch": 0.96816,
+      "grad_norm": 2.463402496344029,
+      "learning_rate": 0.0001584440268928293,
+      "loss": 1.5076,
+      "step": 6051
+    },
+    {
+      "epoch": 0.96832,
+      "grad_norm": 2.6357041871394413,
+      "learning_rate": 0.0001584300093954549,
+      "loss": 1.6313,
+      "step": 6052
+    },
+    {
+      "epoch": 0.96848,
+      "grad_norm": 2.133538952958352,
+      "learning_rate": 0.00015841599015461422,
+      "loss": 1.5443,
+      "step": 6053
+    },
+    {
+      "epoch": 0.96864,
+      "grad_norm": 2.021753071149972,
+      "learning_rate": 0.00015840196917072546,
+      "loss": 1.6112,
+      "step": 6054
+    },
+    {
+      "epoch": 0.9688,
+      "grad_norm": 2.46700887580078,
+      "learning_rate": 0.0001583879464442071,
+      "loss": 1.6115,
+      "step": 6055
+    },
+    {
+      "epoch": 0.96896,
+      "grad_norm": 1.982406201632441,
+      "learning_rate": 0.00015837392197547747,
+      "loss": 1.5568,
+      "step": 6056
+    },
+    {
+      "epoch": 0.96912,
+      "grad_norm": 2.551221205947265,
+      "learning_rate": 0.00015835989576495506,
+      "loss": 1.5648,
+      "step": 6057
+    },
+    {
+      "epoch": 0.96928,
+      "grad_norm": 2.30044359769247,
+      "learning_rate": 0.0001583458678130584,
+      "loss": 1.5342,
+      "step": 6058
+    },
+    {
+      "epoch": 0.96944,
+      "grad_norm": 2.2067888763727472,
+      "learning_rate": 0.00015833183812020608,
+      "loss": 1.5161,
+      "step": 6059
+    },
+    {
+      "epoch": 0.9696,
+      "grad_norm": 2.5671980942461934,
+      "learning_rate": 0.00015831780668681668,
+      "loss": 1.6057,
+      "step": 6060
+    },
+    {
+      "epoch": 0.96976,
+      "grad_norm": 2.2321748103098957,
+      "learning_rate": 0.00015830377351330895,
+      "loss": 1.616,
+      "step": 6061
+    },
+    {
+      "epoch": 0.96992,
+      "grad_norm": 2.0168684105486028,
+      "learning_rate": 0.00015828973860010156,
+      "loss": 1.5115,
+      "step": 6062
+    },
+    {
+      "epoch": 0.97008,
+      "grad_norm": 2.6482295787768844,
+      "learning_rate": 0.00015827570194761327,
+      "loss": 1.5426,
+      "step": 6063
+    },
+    {
+      "epoch": 0.97024,
+      "grad_norm": 2.7485130237565674,
+      "learning_rate": 0.000158261663556263,
+      "loss": 1.5415,
+      "step": 6064
+    },
+    {
+      "epoch": 0.9704,
+      "grad_norm": 2.0809326863200757,
+      "learning_rate": 0.00015824762342646958,
+      "loss": 1.4638,
+      "step": 6065
+    },
+    {
+      "epoch": 0.97056,
+      "grad_norm": 2.2799569906161152,
+      "learning_rate": 0.00015823358155865193,
+      "loss": 1.5368,
+      "step": 6066
+    },
+    {
+      "epoch": 0.97072,
+      "grad_norm": 2.1628099325179386,
+      "learning_rate": 0.00015821953795322905,
+      "loss": 1.5638,
+      "step": 6067
+    },
+    {
+      "epoch": 0.97088,
+      "grad_norm": 2.271556426282503,
+      "learning_rate": 0.00015820549261062004,
+      "loss": 1.5596,
+      "step": 6068
+    },
+    {
+      "epoch": 0.97104,
+      "grad_norm": 2.3417234714130815,
+      "learning_rate": 0.00015819144553124391,
+      "loss": 1.5468,
+      "step": 6069
+    },
+    {
+      "epoch": 0.9712,
+      "grad_norm": 2.4780586576362444,
+      "learning_rate": 0.00015817739671551982,
+      "loss": 1.5548,
+      "step": 6070
+    },
+    {
+      "epoch": 0.97136,
+      "grad_norm": 2.391679404254315,
+      "learning_rate": 0.00015816334616386703,
+      "loss": 1.5634,
+      "step": 6071
+    },
+    {
+      "epoch": 0.97152,
+      "grad_norm": 2.5123325058858077,
+      "learning_rate": 0.00015814929387670472,
+      "loss": 1.6884,
+      "step": 6072
+    },
+    {
+      "epoch": 0.97168,
+      "grad_norm": 2.385437298506703,
+      "learning_rate": 0.00015813523985445222,
+      "loss": 1.584,
+      "step": 6073
+    },
+    {
+      "epoch": 0.97184,
+      "grad_norm": 2.137081411634058,
+      "learning_rate": 0.00015812118409752886,
+      "loss": 1.5582,
+      "step": 6074
+    },
+    {
+      "epoch": 0.972,
+      "grad_norm": 2.3274104380164937,
+      "learning_rate": 0.00015810712660635409,
+      "loss": 1.6228,
+      "step": 6075
+    },
+    {
+      "epoch": 0.97216,
+      "grad_norm": 2.139158822199404,
+      "learning_rate": 0.0001580930673813473,
+      "loss": 1.5407,
+      "step": 6076
+    },
+    {
+      "epoch": 0.97232,
+      "grad_norm": 2.139158822199404,
+      "learning_rate": 0.0001580930673813473,
+      "loss": 1.5759,
+      "step": 6077
+    },
+    {
+      "epoch": 0.97248,
+      "grad_norm": 2.400905568640275,
+      "learning_rate": 0.00015807900642292807,
+      "loss": 1.4921,
+      "step": 6078
+    },
+    {
+      "epoch": 0.97264,
+      "grad_norm": 2.6335201451458756,
+      "learning_rate": 0.0001580649437315159,
+      "loss": 1.5614,
+      "step": 6079
+    },
+    {
+      "epoch": 0.9728,
+      "grad_norm": 2.3736730306147797,
+      "learning_rate": 0.0001580508793075304,
+      "loss": 1.5181,
+      "step": 6080
+    },
+    {
+      "epoch": 0.97296,
+      "grad_norm": 2.4034846145286006,
+      "learning_rate": 0.00015803681315139126,
+      "loss": 1.5445,
+      "step": 6081
+    },
+    {
+      "epoch": 0.97312,
+      "grad_norm": 2.584092326673941,
+      "learning_rate": 0.0001580227452635182,
+      "loss": 1.6438,
+      "step": 6082
+    },
+    {
+      "epoch": 0.97328,
+      "grad_norm": 2.070818991323528,
+      "learning_rate": 0.00015800867564433094,
+      "loss": 1.5704,
+      "step": 6083
+    },
+    {
+      "epoch": 0.97344,
+      "grad_norm": 2.4653163823780253,
+      "learning_rate": 0.00015799460429424933,
+      "loss": 1.5473,
+      "step": 6084
+    },
+    {
+      "epoch": 0.9736,
+      "grad_norm": 2.6957881316054215,
+      "learning_rate": 0.00015798053121369326,
+      "loss": 1.6131,
+      "step": 6085
+    },
+    {
+      "epoch": 0.97376,
+      "grad_norm": 3.007611346040186,
+      "learning_rate": 0.00015796645640308264,
+      "loss": 1.5828,
+      "step": 6086
+    },
+    {
+      "epoch": 0.97392,
+      "grad_norm": 1.976867929496173,
+      "learning_rate": 0.0001579523798628374,
+      "loss": 1.5344,
+      "step": 6087
+    },
+    {
+      "epoch": 0.97408,
+      "grad_norm": 2.21319681598483,
+      "learning_rate": 0.00015793830159337757,
+      "loss": 1.5597,
+      "step": 6088
+    },
+    {
+      "epoch": 0.97424,
+      "grad_norm": 2.2694239140130166,
+      "learning_rate": 0.00015792422159512326,
+      "loss": 1.513,
+      "step": 6089
+    },
+    {
+      "epoch": 0.9744,
+      "grad_norm": 2.3684609008846786,
+      "learning_rate": 0.0001579101398684946,
+      "loss": 1.5049,
+      "step": 6090
+    },
+    {
+      "epoch": 0.97456,
+      "grad_norm": 2.128468260706377,
+      "learning_rate": 0.00015789605641391172,
+      "loss": 1.5799,
+      "step": 6091
+    },
+    {
+      "epoch": 0.97472,
+      "grad_norm": 2.4052338235469284,
+      "learning_rate": 0.0001578819712317949,
+      "loss": 1.6015,
+      "step": 6092
+    },
+    {
+      "epoch": 0.97488,
+      "grad_norm": 2.1907700062021607,
+      "learning_rate": 0.0001578678843225644,
+      "loss": 1.534,
+      "step": 6093
+    },
+    {
+      "epoch": 0.97504,
+      "grad_norm": 2.137591463614977,
+      "learning_rate": 0.00015785379568664055,
+      "loss": 1.5415,
+      "step": 6094
+    },
+    {
+      "epoch": 0.9752,
+      "grad_norm": 2.4249532111547154,
+      "learning_rate": 0.00015783970532444374,
+      "loss": 1.5505,
+      "step": 6095
+    },
+    {
+      "epoch": 0.97536,
+      "grad_norm": 2.7306168781269977,
+      "learning_rate": 0.0001578256132363944,
+      "loss": 1.5561,
+      "step": 6096
+    },
+    {
+      "epoch": 0.97552,
+      "grad_norm": 2.5515256601083376,
+      "learning_rate": 0.00015781151942291304,
+      "loss": 1.5629,
+      "step": 6097
+    },
+    {
+      "epoch": 0.97568,
+      "grad_norm": 3.002108136381675,
+      "learning_rate": 0.00015779742388442015,
+      "loss": 1.5186,
+      "step": 6098
+    },
+    {
+      "epoch": 0.97584,
+      "grad_norm": 1.945605570374376,
+      "learning_rate": 0.00015778332662133636,
+      "loss": 1.4654,
+      "step": 6099
+    },
+    {
+      "epoch": 0.976,
+      "grad_norm": 2.891947917572989,
+      "learning_rate": 0.00015776922763408233,
+      "loss": 1.5251,
+      "step": 6100
+    },
+    {
+      "epoch": 0.97616,
+      "grad_norm": 2.147091766952127,
+      "learning_rate": 0.00015775512692307872,
+      "loss": 1.5461,
+      "step": 6101
+    },
+    {
+      "epoch": 0.97632,
+      "grad_norm": 2.600776481622256,
+      "learning_rate": 0.00015774102448874624,
+      "loss": 1.6282,
+      "step": 6102
+    },
+    {
+      "epoch": 0.97648,
+      "grad_norm": 2.4739434904133746,
+      "learning_rate": 0.00015772692033150576,
+      "loss": 1.662,
+      "step": 6103
+    },
+    {
+      "epoch": 0.97664,
+      "grad_norm": 2.2452536170838675,
+      "learning_rate": 0.0001577128144517781,
+      "loss": 1.551,
+      "step": 6104
+    },
+    {
+      "epoch": 0.9768,
+      "grad_norm": 2.5715631424662098,
+      "learning_rate": 0.00015769870684998412,
+      "loss": 1.5358,
+      "step": 6105
+    },
+    {
+      "epoch": 0.97696,
+      "grad_norm": 2.4642755905765514,
+      "learning_rate": 0.00015768459752654485,
+      "loss": 1.6253,
+      "step": 6106
+    },
+    {
+      "epoch": 0.97712,
+      "grad_norm": 2.2407047437747094,
+      "learning_rate": 0.0001576704864818812,
+      "loss": 1.4708,
+      "step": 6107
+    },
+    {
+      "epoch": 0.97728,
+      "grad_norm": 2.604068552435161,
+      "learning_rate": 0.00015765637371641434,
+      "loss": 1.5289,
+      "step": 6108
+    },
+    {
+      "epoch": 0.97744,
+      "grad_norm": 2.705087374097285,
+      "learning_rate": 0.0001576422592305652,
+      "loss": 1.5588,
+      "step": 6109
+    },
+    {
+      "epoch": 0.9776,
+      "grad_norm": 2.2063593933560637,
+      "learning_rate": 0.00015762814302475511,
+      "loss": 1.5275,
+      "step": 6110
+    },
+    {
+      "epoch": 0.97776,
+      "grad_norm": 2.1307777451349312,
+      "learning_rate": 0.0001576140250994052,
+      "loss": 1.6419,
+      "step": 6111
+    },
+    {
+      "epoch": 0.97792,
+      "grad_norm": 2.3802375503762208,
+      "learning_rate": 0.00015759990545493672,
+      "loss": 1.4647,
+      "step": 6112
+    },
+    {
+      "epoch": 0.97808,
+      "grad_norm": 2.431722251807969,
+      "learning_rate": 0.000157585784091771,
+      "loss": 1.5119,
+      "step": 6113
+    },
+    {
+      "epoch": 0.97824,
+      "grad_norm": 2.3011091241728616,
+      "learning_rate": 0.00015757166101032936,
+      "loss": 1.5927,
+      "step": 6114
+    },
+    {
+      "epoch": 0.9784,
+      "grad_norm": 2.2253150194956133,
+      "learning_rate": 0.0001575575362110333,
+      "loss": 1.5445,
+      "step": 6115
+    },
+    {
+      "epoch": 0.97856,
+      "grad_norm": 2.229021696357095,
+      "learning_rate": 0.00015754340969430418,
+      "loss": 1.5249,
+      "step": 6116
+    },
+    {
+      "epoch": 0.97872,
+      "grad_norm": 2.316088986339486,
+      "learning_rate": 0.0001575292814605636,
+      "loss": 1.5567,
+      "step": 6117
+    },
+    {
+      "epoch": 0.97888,
+      "grad_norm": 2.1667456771004843,
+      "learning_rate": 0.00015751515151023307,
+      "loss": 1.552,
+      "step": 6118
+    },
+    {
+      "epoch": 0.97904,
+      "grad_norm": 2.5558201719139384,
+      "learning_rate": 0.00015750101984373424,
+      "loss": 1.5722,
+      "step": 6119
+    },
+    {
+      "epoch": 0.9792,
+      "grad_norm": 2.2918456913120115,
+      "learning_rate": 0.00015748688646148877,
+      "loss": 1.5645,
+      "step": 6120
+    },
+    {
+      "epoch": 0.97936,
+      "grad_norm": 2.1443899123884917,
+      "learning_rate": 0.00015747275136391835,
+      "loss": 1.5458,
+      "step": 6121
+    },
+    {
+      "epoch": 0.97952,
+      "grad_norm": 2.4832761575647853,
+      "learning_rate": 0.0001574586145514448,
+      "loss": 1.5228,
+      "step": 6122
+    },
+    {
+      "epoch": 0.97968,
+      "grad_norm": 2.232810320637114,
+      "learning_rate": 0.00015744447602448992,
+      "loss": 1.5247,
+      "step": 6123
+    },
+    {
+      "epoch": 0.97984,
+      "grad_norm": 2.5884647950056814,
+      "learning_rate": 0.00015743033578347556,
+      "loss": 1.5595,
+      "step": 6124
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.456196718133416,
+      "learning_rate": 0.00015741619382882366,
+      "loss": 1.6302,
+      "step": 6125
+    },
+    {
+      "epoch": 0.98016,
+      "grad_norm": 2.7766733614705306,
+      "learning_rate": 0.00015740205016095623,
+      "loss": 1.5626,
+      "step": 6126
+    },
+    {
+      "epoch": 0.98032,
+      "grad_norm": 2.243815756845935,
+      "learning_rate": 0.00015738790478029527,
+      "loss": 1.5252,
+      "step": 6127
+    },
+    {
+      "epoch": 0.98048,
+      "grad_norm": 2.71072232305137,
+      "learning_rate": 0.00015737375768726282,
+      "loss": 1.63,
+      "step": 6128
+    },
+    {
+      "epoch": 0.98064,
+      "grad_norm": 2.728228284634612,
+      "learning_rate": 0.00015735960888228108,
+      "loss": 1.7035,
+      "step": 6129
+    },
+    {
+      "epoch": 0.9808,
+      "grad_norm": 2.9879237270240604,
+      "learning_rate": 0.00015734545836577218,
+      "loss": 1.6466,
+      "step": 6130
+    },
+    {
+      "epoch": 0.98096,
+      "grad_norm": 2.3848623147832244,
+      "learning_rate": 0.0001573313061381583,
+      "loss": 1.5426,
+      "step": 6131
+    },
+    {
+      "epoch": 0.98112,
+      "grad_norm": 2.631545545531388,
+      "learning_rate": 0.0001573171521998619,
+      "loss": 1.4952,
+      "step": 6132
+    },
+    {
+      "epoch": 0.98128,
+      "grad_norm": 2.293082024763275,
+      "learning_rate": 0.00015730299655130513,
+      "loss": 1.5208,
+      "step": 6133
+    },
+    {
+      "epoch": 0.98144,
+      "grad_norm": 2.9916146793451963,
+      "learning_rate": 0.00015728883919291045,
+      "loss": 1.4765,
+      "step": 6134
+    },
+    {
+      "epoch": 0.9816,
+      "grad_norm": 3.0312270578760994,
+      "learning_rate": 0.00015727468012510028,
+      "loss": 1.5495,
+      "step": 6135
+    },
+    {
+      "epoch": 0.98176,
+      "grad_norm": 3.3745799591771872,
+      "learning_rate": 0.00015726051934829714,
+      "loss": 1.6828,
+      "step": 6136
+    },
+    {
+      "epoch": 0.98192,
+      "grad_norm": 2.8855630791416544,
+      "learning_rate": 0.0001572463568629235,
+      "loss": 1.5155,
+      "step": 6137
+    },
+    {
+      "epoch": 0.98208,
+      "grad_norm": 2.787432446980218,
+      "learning_rate": 0.00015723219266940201,
+      "loss": 1.5989,
+      "step": 6138
+    },
+    {
+      "epoch": 0.98224,
+      "grad_norm": 3.007645467065774,
+      "learning_rate": 0.0001572180267681553,
+      "loss": 1.5354,
+      "step": 6139
+    },
+    {
+      "epoch": 0.9824,
+      "grad_norm": 3.3988280409171012,
+      "learning_rate": 0.00015720385915960603,
+      "loss": 1.6292,
+      "step": 6140
+    },
+    {
+      "epoch": 0.98256,
+      "grad_norm": 2.7302356101967162,
+      "learning_rate": 0.00015718968984417697,
+      "loss": 1.5807,
+      "step": 6141
+    },
+    {
+      "epoch": 0.98272,
+      "grad_norm": 2.487018389680164,
+      "learning_rate": 0.0001571755188222909,
+      "loss": 1.5806,
+      "step": 6142
+    },
+    {
+      "epoch": 0.98288,
+      "grad_norm": 3.614145810544895,
+      "learning_rate": 0.00015716134609437062,
+      "loss": 1.5812,
+      "step": 6143
+    },
+    {
+      "epoch": 0.98304,
+      "grad_norm": 2.6284117395977202,
+      "learning_rate": 0.0001571471716608391,
+      "loss": 1.5915,
+      "step": 6144
+    },
+    {
+      "epoch": 0.9832,
+      "grad_norm": 2.9703468351512385,
+      "learning_rate": 0.00015713299552211923,
+      "loss": 1.5603,
+      "step": 6145
+    },
+    {
+      "epoch": 0.98336,
+      "grad_norm": 2.442813248347867,
+      "learning_rate": 0.00015711881767863407,
+      "loss": 1.5698,
+      "step": 6146
+    },
+    {
+      "epoch": 0.98352,
+      "grad_norm": 2.7782798304587732,
+      "learning_rate": 0.0001571046381308066,
+      "loss": 1.4291,
+      "step": 6147
+    },
+    {
+      "epoch": 0.98368,
+      "grad_norm": 2.576425425629686,
+      "learning_rate": 0.00015709045687905992,
+      "loss": 1.5974,
+      "step": 6148
+    },
+    {
+      "epoch": 0.98384,
+      "grad_norm": 2.2965441582173938,
+      "learning_rate": 0.00015707627392381722,
+      "loss": 1.5289,
+      "step": 6149
+    },
+    {
+      "epoch": 0.984,
+      "grad_norm": 2.4997578931272644,
+      "learning_rate": 0.00015706208926550164,
+      "loss": 1.5555,
+      "step": 6150
+    },
+    {
+      "epoch": 0.98416,
+      "grad_norm": 2.52188977708189,
+      "learning_rate": 0.0001570479029045365,
+      "loss": 1.5786,
+      "step": 6151
+    },
+    {
+      "epoch": 0.98432,
+      "grad_norm": 2.4339189671847046,
+      "learning_rate": 0.00015703371484134508,
+      "loss": 1.5074,
+      "step": 6152
+    },
+    {
+      "epoch": 0.98448,
+      "grad_norm": 2.246954354875616,
+      "learning_rate": 0.00015701952507635067,
+      "loss": 1.5118,
+      "step": 6153
+    },
+    {
+      "epoch": 0.98464,
+      "grad_norm": 2.287913693429862,
+      "learning_rate": 0.00015700533360997674,
+      "loss": 1.5799,
+      "step": 6154
+    },
+    {
+      "epoch": 0.9848,
+      "grad_norm": 2.2306116268259766,
+      "learning_rate": 0.00015699114044264673,
+      "loss": 1.5843,
+      "step": 6155
+    },
+    {
+      "epoch": 0.98496,
+      "grad_norm": 2.4322616736776306,
+      "learning_rate": 0.00015697694557478413,
+      "loss": 1.5786,
+      "step": 6156
+    },
+    {
+      "epoch": 0.98512,
+      "grad_norm": 2.3835072224052385,
+      "learning_rate": 0.00015696274900681248,
+      "loss": 1.5195,
+      "step": 6157
+    },
+    {
+      "epoch": 0.98528,
+      "grad_norm": 2.252727026597292,
+      "learning_rate": 0.0001569485507391554,
+      "loss": 1.4195,
+      "step": 6158
+    },
+    {
+      "epoch": 0.98544,
+      "grad_norm": 2.4629036314341604,
+      "learning_rate": 0.00015693435077223655,
+      "loss": 1.4906,
+      "step": 6159
+    },
+    {
+      "epoch": 0.9856,
+      "grad_norm": 2.7989112002103926,
+      "learning_rate": 0.00015692014910647964,
+      "loss": 1.6329,
+      "step": 6160
+    },
+    {
+      "epoch": 0.98576,
+      "grad_norm": 2.5514353971772894,
+      "learning_rate": 0.00015690594574230846,
+      "loss": 1.5355,
+      "step": 6161
+    },
+    {
+      "epoch": 0.98592,
+      "grad_norm": 2.3915703737403344,
+      "learning_rate": 0.00015689174068014673,
+      "loss": 1.538,
+      "step": 6162
+    },
+    {
+      "epoch": 0.98608,
+      "grad_norm": 2.4530844110129073,
+      "learning_rate": 0.00015687753392041838,
+      "loss": 1.6533,
+      "step": 6163
+    },
+    {
+      "epoch": 0.98624,
+      "grad_norm": 2.38879719345123,
+      "learning_rate": 0.00015686332546354729,
+      "loss": 1.5577,
+      "step": 6164
+    },
+    {
+      "epoch": 0.9864,
+      "grad_norm": 2.280172309072704,
+      "learning_rate": 0.00015684911530995744,
+      "loss": 1.537,
+      "step": 6165
+    },
+    {
+      "epoch": 0.98656,
+      "grad_norm": 2.4249228080601646,
+      "learning_rate": 0.00015683490346007282,
+      "loss": 1.524,
+      "step": 6166
+    },
+    {
+      "epoch": 0.98672,
+      "grad_norm": 2.272171235029926,
+      "learning_rate": 0.00015682068991431752,
+      "loss": 1.5753,
+      "step": 6167
+    },
+    {
+      "epoch": 0.98688,
+      "grad_norm": 2.492660131858362,
+      "learning_rate": 0.00015680647467311557,
+      "loss": 1.5131,
+      "step": 6168
+    },
+    {
+      "epoch": 0.98704,
+      "grad_norm": 2.116336534484752,
+      "learning_rate": 0.00015679225773689126,
+      "loss": 1.4696,
+      "step": 6169
+    },
+    {
+      "epoch": 0.9872,
+      "grad_norm": 2.2938168043185616,
+      "learning_rate": 0.0001567780391060687,
+      "loss": 1.481,
+      "step": 6170
+    },
+    {
+      "epoch": 0.98736,
+      "grad_norm": 2.2835388036197175,
+      "learning_rate": 0.0001567638187810722,
+      "loss": 1.547,
+      "step": 6171
+    },
+    {
+      "epoch": 0.98752,
+      "grad_norm": 2.01906935860248,
+      "learning_rate": 0.00015674959676232602,
+      "loss": 1.4933,
+      "step": 6172
+    },
+    {
+      "epoch": 0.98768,
+      "grad_norm": 2.335878053808254,
+      "learning_rate": 0.0001567353730502546,
+      "loss": 1.4913,
+      "step": 6173
+    },
+    {
+      "epoch": 0.98784,
+      "grad_norm": 2.0712252609697592,
+      "learning_rate": 0.0001567211476452823,
+      "loss": 1.6189,
+      "step": 6174
+    },
+    {
+      "epoch": 0.988,
+      "grad_norm": 2.274381634219793,
+      "learning_rate": 0.00015670692054783363,
+      "loss": 1.6777,
+      "step": 6175
+    },
+    {
+      "epoch": 0.98816,
+      "grad_norm": 2.4769627088105066,
+      "learning_rate": 0.00015669269175833307,
+      "loss": 1.5599,
+      "step": 6176
+    },
+    {
+      "epoch": 0.98832,
+      "grad_norm": 2.387335304570233,
+      "learning_rate": 0.00015667846127720523,
+      "loss": 1.5462,
+      "step": 6177
+    },
+    {
+      "epoch": 0.98848,
+      "grad_norm": 2.1769255415838944,
+      "learning_rate": 0.00015666422910487466,
+      "loss": 1.5479,
+      "step": 6178
+    },
+    {
+      "epoch": 0.98864,
+      "grad_norm": 2.3455943232301735,
+      "learning_rate": 0.00015664999524176606,
+      "loss": 1.6117,
+      "step": 6179
+    },
+    {
+      "epoch": 0.9888,
+      "grad_norm": 2.4421056039916826,
+      "learning_rate": 0.00015663575968830417,
+      "loss": 1.4907,
+      "step": 6180
+    },
+    {
+      "epoch": 0.98896,
+      "grad_norm": 2.1366309301427564,
+      "learning_rate": 0.00015662152244491374,
+      "loss": 1.6422,
+      "step": 6181
+    },
+    {
+      "epoch": 0.98912,
+      "grad_norm": 1.8738439898009291,
+      "learning_rate": 0.00015660728351201954,
+      "loss": 1.5464,
+      "step": 6182
+    },
+    {
+      "epoch": 0.98928,
+      "grad_norm": 2.4353351761036204,
+      "learning_rate": 0.00015659304289004653,
+      "loss": 1.4444,
+      "step": 6183
+    },
+    {
+      "epoch": 0.98944,
+      "grad_norm": 2.323803220127518,
+      "learning_rate": 0.0001565788005794196,
+      "loss": 1.5317,
+      "step": 6184
+    },
+    {
+      "epoch": 0.9896,
+      "grad_norm": 2.5925336103639642,
+      "learning_rate": 0.00015656455658056366,
+      "loss": 1.6018,
+      "step": 6185
+    },
+    {
+      "epoch": 0.98976,
+      "grad_norm": 2.1738785733200454,
+      "learning_rate": 0.00015655031089390378,
+      "loss": 1.546,
+      "step": 6186
+    },
+    {
+      "epoch": 0.98992,
+      "grad_norm": 2.3789160408502084,
+      "learning_rate": 0.00015653606351986502,
+      "loss": 1.5253,
+      "step": 6187
+    },
+    {
+      "epoch": 0.99008,
+      "grad_norm": 2.3451306825166767,
+      "learning_rate": 0.00015652181445887254,
+      "loss": 1.6205,
+      "step": 6188
+    },
+    {
+      "epoch": 0.99024,
+      "grad_norm": 2.6040696898908515,
+      "learning_rate": 0.00015650756371135148,
+      "loss": 1.5599,
+      "step": 6189
+    },
+    {
+      "epoch": 0.9904,
+      "grad_norm": 2.9580878185418324,
+      "learning_rate": 0.00015649331127772702,
+      "loss": 1.6344,
+      "step": 6190
+    },
+    {
+      "epoch": 0.99056,
+      "grad_norm": 1.9599628368068769,
+      "learning_rate": 0.0001564790571584245,
+      "loss": 1.4578,
+      "step": 6191
+    },
+    {
+      "epoch": 0.99072,
+      "grad_norm": 3.1000084883154333,
+      "learning_rate": 0.0001564648013538692,
+      "loss": 1.5176,
+      "step": 6192
+    },
+    {
+      "epoch": 0.99088,
+      "grad_norm": 2.3253594220649974,
+      "learning_rate": 0.00015645054386448652,
+      "loss": 1.5636,
+      "step": 6193
+    },
+    {
+      "epoch": 0.99104,
+      "grad_norm": 2.821168526521178,
+      "learning_rate": 0.00015643628469070187,
+      "loss": 1.6144,
+      "step": 6194
+    },
+    {
+      "epoch": 0.9912,
+      "grad_norm": 2.0296991933408317,
+      "learning_rate": 0.0001564220238329407,
+      "loss": 1.496,
+      "step": 6195
+    },
+    {
+      "epoch": 0.99136,
+      "grad_norm": 2.273276749320594,
+      "learning_rate": 0.00015640776129162854,
+      "loss": 1.5836,
+      "step": 6196
+    },
+    {
+      "epoch": 0.99152,
+      "grad_norm": 2.8624080985823275,
+      "learning_rate": 0.00015639349706719102,
+      "loss": 1.6068,
+      "step": 6197
+    },
+    {
+      "epoch": 0.99168,
+      "grad_norm": 2.1136252908813504,
+      "learning_rate": 0.00015637923116005372,
+      "loss": 1.5956,
+      "step": 6198
+    },
+    {
+      "epoch": 0.99184,
+      "grad_norm": 2.3654216384777325,
+      "learning_rate": 0.00015636496357064228,
+      "loss": 1.4729,
+      "step": 6199
+    },
+    {
+      "epoch": 0.992,
+      "grad_norm": 2.7836914945539504,
+      "learning_rate": 0.00015635069429938248,
+      "loss": 1.5409,
+      "step": 6200
+    },
+    {
+      "epoch": 0.99216,
+      "grad_norm": 2.8330871315496435,
+      "learning_rate": 0.0001563364233467001,
+      "loss": 1.5298,
+      "step": 6201
+    },
+    {
+      "epoch": 0.99232,
+      "grad_norm": 2.2454602909441546,
+      "learning_rate": 0.0001563221507130209,
+      "loss": 1.5374,
+      "step": 6202
+    },
+    {
+      "epoch": 0.99248,
+      "grad_norm": 2.3118100307201286,
+      "learning_rate": 0.00015630787639877078,
+      "loss": 1.5291,
+      "step": 6203
+    },
+    {
+      "epoch": 0.99264,
+      "grad_norm": 2.8347393877923737,
+      "learning_rate": 0.00015629360040437571,
+      "loss": 1.5302,
+      "step": 6204
+    },
+    {
+      "epoch": 0.9928,
+      "grad_norm": 2.627467059361297,
+      "learning_rate": 0.00015627932273026162,
+      "loss": 1.5086,
+      "step": 6205
+    },
+    {
+      "epoch": 0.99296,
+      "grad_norm": 2.0279402160616917,
+      "learning_rate": 0.00015626504337685457,
+      "loss": 1.5747,
+      "step": 6206
+    },
+    {
+      "epoch": 0.99312,
+      "grad_norm": 2.0075036987685304,
+      "learning_rate": 0.0001562507623445806,
+      "loss": 1.4459,
+      "step": 6207
+    },
+    {
+      "epoch": 0.99328,
+      "grad_norm": 3.0846020095626296,
+      "learning_rate": 0.00015623647963386583,
+      "loss": 1.5973,
+      "step": 6208
+    },
+    {
+      "epoch": 0.99344,
+      "grad_norm": 2.2247547354610675,
+      "learning_rate": 0.00015622219524513646,
+      "loss": 1.5597,
+      "step": 6209
+    },
+    {
+      "epoch": 0.9936,
+      "grad_norm": 3.239058349882341,
+      "learning_rate": 0.00015620790917881871,
+      "loss": 1.4534,
+      "step": 6210
+    },
+    {
+      "epoch": 0.99376,
+      "grad_norm": 2.155146930976846,
+      "learning_rate": 0.00015619362143533884,
+      "loss": 1.52,
+      "step": 6211
+    },
+    {
+      "epoch": 0.99392,
+      "grad_norm": 3.0844167827065934,
+      "learning_rate": 0.0001561793320151232,
+      "loss": 1.5111,
+      "step": 6212
+    },
+    {
+      "epoch": 0.99408,
+      "grad_norm": 2.066975361699536,
+      "learning_rate": 0.00015616504091859817,
+      "loss": 1.5752,
+      "step": 6213
+    },
+    {
+      "epoch": 0.99424,
+      "grad_norm": 3.2925771942912663,
+      "learning_rate": 0.00015615074814619018,
+      "loss": 1.4997,
+      "step": 6214
+    },
+    {
+      "epoch": 0.9944,
+      "grad_norm": 2.2287668685625444,
+      "learning_rate": 0.00015613645369832562,
+      "loss": 1.536,
+      "step": 6215
+    },
+    {
+      "epoch": 0.99456,
+      "grad_norm": 2.1738164586380613,
+      "learning_rate": 0.0001561221575754311,
+      "loss": 1.4997,
+      "step": 6216
+    },
+    {
+      "epoch": 0.99472,
+      "grad_norm": 2.5024541944549195,
+      "learning_rate": 0.0001561078597779332,
+      "loss": 1.5422,
+      "step": 6217
+    },
+    {
+      "epoch": 0.99488,
+      "grad_norm": 2.203389313346578,
+      "learning_rate": 0.00015609356030625853,
+      "loss": 1.6859,
+      "step": 6218
+    },
+    {
+      "epoch": 0.99504,
+      "grad_norm": 2.553503000259611,
+      "learning_rate": 0.0001560792591608337,
+      "loss": 1.479,
+      "step": 6219
+    },
+    {
+      "epoch": 0.9952,
+      "grad_norm": 2.5889754418748154,
+      "learning_rate": 0.00015606495634208555,
+      "loss": 1.5231,
+      "step": 6220
+    },
+    {
+      "epoch": 0.99536,
+      "grad_norm": 2.240495924893959,
+      "learning_rate": 0.0001560506518504408,
+      "loss": 1.5066,
+      "step": 6221
+    },
+    {
+      "epoch": 0.99552,
+      "grad_norm": 2.748280944597783,
+      "learning_rate": 0.00015603634568632627,
+      "loss": 1.5673,
+      "step": 6222
+    },
+    {
+      "epoch": 0.99568,
+      "grad_norm": 2.123454825767384,
+      "learning_rate": 0.0001560220378501688,
+      "loss": 1.459,
+      "step": 6223
+    },
+    {
+      "epoch": 0.99584,
+      "grad_norm": 2.3193610472449264,
+      "learning_rate": 0.0001560077283423954,
+      "loss": 1.5801,
+      "step": 6224
+    },
+    {
+      "epoch": 0.996,
+      "grad_norm": 2.4543816978765816,
+      "learning_rate": 0.00015599341716343297,
+      "loss": 1.5191,
+      "step": 6225
+    },
+    {
+      "epoch": 0.99616,
+      "grad_norm": 2.1511015782991314,
+      "learning_rate": 0.00015597910431370857,
+      "loss": 1.482,
+      "step": 6226
+    },
+    {
+      "epoch": 0.99632,
+      "grad_norm": 2.1475985853941726,
+      "learning_rate": 0.0001559647897936493,
+      "loss": 1.5585,
+      "step": 6227
+    },
+    {
+      "epoch": 0.99648,
+      "grad_norm": 1.7043971373749531,
+      "learning_rate": 0.00015595047360368217,
+      "loss": 1.5061,
+      "step": 6228
+    },
+    {
+      "epoch": 0.99664,
+      "grad_norm": 2.5347907777409038,
+      "learning_rate": 0.00015593615574423448,
+      "loss": 1.5408,
+      "step": 6229
+    },
+    {
+      "epoch": 0.9968,
+      "grad_norm": 2.301902615217009,
+      "learning_rate": 0.00015592183621573343,
+      "loss": 1.5801,
+      "step": 6230
+    },
+    {
+      "epoch": 0.99696,
+      "grad_norm": 2.0850258023833295,
+      "learning_rate": 0.00015590751501860626,
+      "loss": 1.5151,
+      "step": 6231
+    },
+    {
+      "epoch": 0.99712,
+      "grad_norm": 2.444200084952057,
+      "learning_rate": 0.0001558931921532803,
+      "loss": 1.5919,
+      "step": 6232
+    },
+    {
+      "epoch": 0.99728,
+      "grad_norm": 2.762056736182959,
+      "learning_rate": 0.00015587886762018295,
+      "loss": 1.6259,
+      "step": 6233
+    },
+    {
+      "epoch": 0.99744,
+      "grad_norm": 2.4072292969249998,
+      "learning_rate": 0.00015586454141974158,
+      "loss": 1.6345,
+      "step": 6234
+    },
+    {
+      "epoch": 0.9976,
+      "grad_norm": 2.7573030785786634,
+      "learning_rate": 0.00015585021355238375,
+      "loss": 1.5595,
+      "step": 6235
+    },
+    {
+      "epoch": 0.99776,
+      "grad_norm": 2.7700047423241796,
+      "learning_rate": 0.00015583588401853687,
+      "loss": 1.5545,
+      "step": 6236
+    },
+    {
+      "epoch": 0.99792,
+      "grad_norm": 2.13574934685889,
+      "learning_rate": 0.0001558215528186286,
+      "loss": 1.5675,
+      "step": 6237
+    },
+    {
+      "epoch": 0.99808,
+      "grad_norm": 2.6981891186201117,
+      "learning_rate": 0.00015580721995308653,
+      "loss": 1.4974,
+      "step": 6238
+    },
+    {
+      "epoch": 0.99824,
+      "grad_norm": 2.135459014852435,
+      "learning_rate": 0.00015579288542233837,
+      "loss": 1.5734,
+      "step": 6239
+    },
+    {
+      "epoch": 0.9984,
+      "grad_norm": 2.782192985540685,
+      "learning_rate": 0.00015577854922681176,
+      "loss": 1.5113,
+      "step": 6240
+    },
+    {
+      "epoch": 0.99856,
+      "grad_norm": 2.158554410266676,
+      "learning_rate": 0.00015576421136693456,
+      "loss": 1.5748,
+      "step": 6241
+    },
+    {
+      "epoch": 0.99872,
+      "grad_norm": 2.121944028512453,
+      "learning_rate": 0.00015574987184313453,
+      "loss": 1.4494,
+      "step": 6242
+    },
+    {
+      "epoch": 0.99888,
+      "grad_norm": 2.4503322527796314,
+      "learning_rate": 0.00015573553065583955,
+      "loss": 1.5489,
+      "step": 6243
+    },
+    {
+      "epoch": 0.99904,
+      "grad_norm": 2.2450985661954403,
+      "learning_rate": 0.00015572118780547754,
+      "loss": 1.5481,
+      "step": 6244
+    },
+    {
+      "epoch": 0.9992,
+      "grad_norm": 2.5577053675242603,
+      "learning_rate": 0.00015570684329247651,
+      "loss": 1.5149,
+      "step": 6245
+    },
+    {
+      "epoch": 0.99936,
+      "grad_norm": 2.0879553155491535,
+      "learning_rate": 0.00015569249711726445,
+      "loss": 1.4964,
+      "step": 6246
+    },
+    {
+      "epoch": 0.99952,
+      "grad_norm": 2.2871267457304616,
+      "learning_rate": 0.00015567814928026943,
+      "loss": 1.5585,
+      "step": 6247
+    },
+    {
+      "epoch": 0.99968,
+      "grad_norm": 2.207119832091538,
+      "learning_rate": 0.00015566379978191954,
+      "loss": 1.4671,
+      "step": 6248
+    },
+    {
+      "epoch": 0.99984,
+      "grad_norm": 2.73420754494684,
+      "learning_rate": 0.000155649448622643,
+      "loss": 1.5641,
+      "step": 6249
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 2.147333025483879,
+      "learning_rate": 0.00015563509580286797,
+      "loss": 1.6018,
+      "step": 6250
+    },
+    {
+      "epoch": 1.00016,
+      "grad_norm": 2.6815254706190546,
+      "learning_rate": 0.0001556207413230228,
+      "loss": 1.4299,
+      "step": 6251
+    },
+    {
+      "epoch": 1.00032,
+      "grad_norm": 2.094587859219373,
+      "learning_rate": 0.00015560638518353573,
+      "loss": 1.4606,
+      "step": 6252
+    },
+    {
+      "epoch": 1.00048,
+      "grad_norm": 2.058074731486955,
+      "learning_rate": 0.00015559202738483515,
+      "loss": 1.4042,
+      "step": 6253
+    },
+    {
+      "epoch": 1.00064,
+      "grad_norm": 2.450237276376254,
+      "learning_rate": 0.00015557766792734947,
+      "loss": 1.3955,
+      "step": 6254
+    },
+    {
+      "epoch": 1.0008,
+      "grad_norm": 2.5334330669856953,
+      "learning_rate": 0.00015556330681150716,
+      "loss": 1.4838,
+      "step": 6255
+    },
+    {
+      "epoch": 1.00096,
+      "grad_norm": 2.465848352667038,
+      "learning_rate": 0.00015554894403773677,
+      "loss": 1.5024,
+      "step": 6256
+    },
+    {
+      "epoch": 1.00112,
+      "grad_norm": 2.0827347523753303,
+      "learning_rate": 0.00015553457960646683,
+      "loss": 1.4754,
+      "step": 6257
+    },
+    {
+      "epoch": 1.00128,
+      "grad_norm": 2.931736699183744,
+      "learning_rate": 0.00015552021351812595,
+      "loss": 1.5494,
+      "step": 6258
+    },
+    {
+      "epoch": 1.00144,
+      "grad_norm": 2.2755709676495823,
+      "learning_rate": 0.00015550584577314277,
+      "loss": 1.4466,
+      "step": 6259
+    },
+    {
+      "epoch": 1.0016,
+      "grad_norm": 2.292525868364926,
+      "learning_rate": 0.00015549147637194607,
+      "loss": 1.4587,
+      "step": 6260
+    },
+    {
+      "epoch": 1.00176,
+      "grad_norm": 1.9871546696146343,
+      "learning_rate": 0.00015547710531496457,
+      "loss": 1.4041,
+      "step": 6261
+    },
+    {
+      "epoch": 1.00192,
+      "grad_norm": 2.33097715586223,
+      "learning_rate": 0.0001554627326026271,
+      "loss": 1.5563,
+      "step": 6262
+    },
+    {
+      "epoch": 1.00208,
+      "grad_norm": 2.3546640797230416,
+      "learning_rate": 0.00015544835823536247,
+      "loss": 1.4988,
+      "step": 6263
+    },
+    {
+      "epoch": 1.00224,
+      "grad_norm": 2.1555199385594546,
+      "learning_rate": 0.00015543398221359964,
+      "loss": 1.4595,
+      "step": 6264
+    },
+    {
+      "epoch": 1.0024,
+      "grad_norm": 2.4885584908353366,
+      "learning_rate": 0.00015541960453776757,
+      "loss": 1.4827,
+      "step": 6265
+    },
+    {
+      "epoch": 1.00256,
+      "grad_norm": 2.2401963432956964,
+      "learning_rate": 0.00015540522520829524,
+      "loss": 1.4597,
+      "step": 6266
+    },
+    {
+      "epoch": 1.00272,
+      "grad_norm": 2.687279322275193,
+      "learning_rate": 0.00015539084422561174,
+      "loss": 1.5598,
+      "step": 6267
+    },
+    {
+      "epoch": 1.00288,
+      "grad_norm": 2.034576277711459,
+      "learning_rate": 0.00015537646159014614,
+      "loss": 1.4245,
+      "step": 6268
+    },
+    {
+      "epoch": 1.00304,
+      "grad_norm": 2.5141098793857024,
+      "learning_rate": 0.0001553620773023276,
+      "loss": 1.4368,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0032,
+      "grad_norm": 2.2358599500674714,
+      "learning_rate": 0.00015534769136258541,
+      "loss": 1.4247,
+      "step": 6270
+    },
+    {
+      "epoch": 1.00336,
+      "grad_norm": 2.489496717988073,
+      "learning_rate": 0.00015533330377134872,
+      "loss": 1.4865,
+      "step": 6271
+    },
+    {
+      "epoch": 1.00352,
+      "grad_norm": 2.2015077000697807,
+      "learning_rate": 0.00015531891452904689,
+      "loss": 1.4459,
+      "step": 6272
+    },
+    {
+      "epoch": 1.00368,
+      "grad_norm": 2.54874153935591,
+      "learning_rate": 0.0001553045236361092,
+      "loss": 1.4707,
+      "step": 6273
+    },
+    {
+      "epoch": 1.00384,
+      "grad_norm": 2.6117733597083013,
+      "learning_rate": 0.00015529013109296516,
+      "loss": 1.451,
+      "step": 6274
+    },
+    {
+      "epoch": 1.004,
+      "grad_norm": 2.683625957517069,
+      "learning_rate": 0.0001552757369000442,
+      "loss": 1.406,
+      "step": 6275
+    },
+    {
+      "epoch": 1.00416,
+      "grad_norm": 2.8697939551726583,
+      "learning_rate": 0.00015526134105777576,
+      "loss": 1.535,
+      "step": 6276
+    },
+    {
+      "epoch": 1.00432,
+      "grad_norm": 2.0897275589276347,
+      "learning_rate": 0.00015524694356658944,
+      "loss": 1.4836,
+      "step": 6277
+    },
+    {
+      "epoch": 1.00448,
+      "grad_norm": 3.0990493559734706,
+      "learning_rate": 0.00015523254442691479,
+      "loss": 1.4802,
+      "step": 6278
+    },
+    {
+      "epoch": 1.00464,
+      "grad_norm": 1.97695542395046,
+      "learning_rate": 0.00015521814363918157,
+      "loss": 1.3969,
+      "step": 6279
+    },
+    {
+      "epoch": 1.0048,
+      "grad_norm": 2.322254429745649,
+      "learning_rate": 0.00015520374120381935,
+      "loss": 1.518,
+      "step": 6280
+    },
+    {
+      "epoch": 1.00496,
+      "grad_norm": 2.484217026315027,
+      "learning_rate": 0.00015518933712125796,
+      "loss": 1.4602,
+      "step": 6281
+    },
+    {
+      "epoch": 1.00512,
+      "grad_norm": 2.235164483319628,
+      "learning_rate": 0.00015517493139192713,
+      "loss": 1.5643,
+      "step": 6282
+    },
+    {
+      "epoch": 1.00528,
+      "grad_norm": 2.6026830207224068,
+      "learning_rate": 0.0001551605240162568,
+      "loss": 1.4361,
+      "step": 6283
+    },
+    {
+      "epoch": 1.0054400000000001,
+      "grad_norm": 2.2894492636899377,
+      "learning_rate": 0.00015514611499467678,
+      "loss": 1.4951,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0056,
+      "grad_norm": 2.285068088667653,
+      "learning_rate": 0.00015513170432761704,
+      "loss": 1.4458,
+      "step": 6285
+    },
+    {
+      "epoch": 1.00576,
+      "grad_norm": 3.0812747166602046,
+      "learning_rate": 0.00015511729201550759,
+      "loss": 1.5042,
+      "step": 6286
+    },
+    {
+      "epoch": 1.00592,
+      "grad_norm": 2.528049080287711,
+      "learning_rate": 0.00015510287805877844,
+      "loss": 1.4518,
+      "step": 6287
+    },
+    {
+      "epoch": 1.00608,
+      "grad_norm": 2.0710113272499258,
+      "learning_rate": 0.00015508846245785972,
+      "loss": 1.437,
+      "step": 6288
+    },
+    {
+      "epoch": 1.00624,
+      "grad_norm": 3.1221855001358434,
+      "learning_rate": 0.0001550740452131816,
+      "loss": 1.543,
+      "step": 6289
+    },
+    {
+      "epoch": 1.0064,
+      "grad_norm": 2.010572551466955,
+      "learning_rate": 0.00015505962632517413,
+      "loss": 1.4701,
+      "step": 6290
+    },
+    {
+      "epoch": 1.00656,
+      "grad_norm": 2.800521289734866,
+      "learning_rate": 0.0001550452057942677,
+      "loss": 1.4867,
+      "step": 6291
+    },
+    {
+      "epoch": 1.00672,
+      "grad_norm": 2.306317981847957,
+      "learning_rate": 0.00015503078362089255,
+      "loss": 1.4795,
+      "step": 6292
+    },
+    {
+      "epoch": 1.00688,
+      "grad_norm": 2.9575476658665187,
+      "learning_rate": 0.00015501635980547898,
+      "loss": 1.4696,
+      "step": 6293
+    },
+    {
+      "epoch": 1.00704,
+      "grad_norm": 2.083002084577977,
+      "learning_rate": 0.00015500193434845734,
+      "loss": 1.5166,
+      "step": 6294
+    },
+    {
+      "epoch": 1.0072,
+      "grad_norm": 2.112672940643797,
+      "learning_rate": 0.00015498750725025822,
+      "loss": 1.4722,
+      "step": 6295
+    },
+    {
+      "epoch": 1.00736,
+      "grad_norm": 2.5205904836871715,
+      "learning_rate": 0.00015497307851131196,
+      "loss": 1.4801,
+      "step": 6296
+    },
+    {
+      "epoch": 1.00752,
+      "grad_norm": 2.1264358873563727,
+      "learning_rate": 0.00015495864813204916,
+      "loss": 1.4955,
+      "step": 6297
+    },
+    {
+      "epoch": 1.00768,
+      "grad_norm": 1.944150071948575,
+      "learning_rate": 0.0001549442161129004,
+      "loss": 1.4028,
+      "step": 6298
+    },
+    {
+      "epoch": 1.00784,
+      "grad_norm": 2.438551171908332,
+      "learning_rate": 0.00015492978245429625,
+      "loss": 1.4663,
+      "step": 6299
+    },
+    {
+      "epoch": 1.008,
+      "grad_norm": 2.4204732161687383,
+      "learning_rate": 0.00015491534715666747,
+      "loss": 1.4993,
+      "step": 6300
+    },
+    {
+      "epoch": 1.00816,
+      "grad_norm": 2.1638156975347043,
+      "learning_rate": 0.00015490091022044472,
+      "loss": 1.4631,
+      "step": 6301
+    },
+    {
+      "epoch": 1.00832,
+      "grad_norm": 3.4643308873544463,
+      "learning_rate": 0.00015488647164605882,
+      "loss": 1.7231,
+      "step": 6302
+    },
+    {
+      "epoch": 1.00848,
+      "grad_norm": 2.9865840659863028,
+      "learning_rate": 0.0001548720314339406,
+      "loss": 1.3951,
+      "step": 6303
+    },
+    {
+      "epoch": 1.00864,
+      "grad_norm": 2.427588244354809,
+      "learning_rate": 0.00015485758958452093,
+      "loss": 1.4993,
+      "step": 6304
+    },
+    {
+      "epoch": 1.0088,
+      "grad_norm": 2.2647845431159546,
+      "learning_rate": 0.00015484314609823068,
+      "loss": 1.4553,
+      "step": 6305
+    },
+    {
+      "epoch": 1.00896,
+      "grad_norm": 2.652117567733315,
+      "learning_rate": 0.00015482870097550087,
+      "loss": 1.3965,
+      "step": 6306
+    },
+    {
+      "epoch": 1.00912,
+      "grad_norm": 2.541836656872451,
+      "learning_rate": 0.00015481425421676255,
+      "loss": 1.5728,
+      "step": 6307
+    },
+    {
+      "epoch": 1.00928,
+      "grad_norm": 2.4679940297558245,
+      "learning_rate": 0.00015479980582244673,
+      "loss": 1.5672,
+      "step": 6308
+    },
+    {
+      "epoch": 1.00944,
+      "grad_norm": 2.412173960714897,
+      "learning_rate": 0.00015478535579298454,
+      "loss": 1.5095,
+      "step": 6309
+    },
+    {
+      "epoch": 1.0096,
+      "grad_norm": 2.47833611624656,
+      "learning_rate": 0.0001547709041288072,
+      "loss": 1.4274,
+      "step": 6310
+    },
+    {
+      "epoch": 1.00976,
+      "grad_norm": 2.4948151797469906,
+      "learning_rate": 0.00015475645083034588,
+      "loss": 1.5349,
+      "step": 6311
+    },
+    {
+      "epoch": 1.00992,
+      "grad_norm": 2.3176276020576863,
+      "learning_rate": 0.00015474199589803186,
+      "loss": 1.4868,
+      "step": 6312
+    },
+    {
+      "epoch": 1.01008,
+      "grad_norm": 2.2832867857244827,
+      "learning_rate": 0.00015472753933229645,
+      "loss": 1.5406,
+      "step": 6313
+    },
+    {
+      "epoch": 1.01024,
+      "grad_norm": 3.0156266866102346,
+      "learning_rate": 0.00015471308113357098,
+      "loss": 1.4864,
+      "step": 6314
+    },
+    {
+      "epoch": 1.0104,
+      "grad_norm": 2.5039148065645658,
+      "learning_rate": 0.00015469862130228695,
+      "loss": 1.4062,
+      "step": 6315
+    },
+    {
+      "epoch": 1.01056,
+      "grad_norm": 2.366635484072869,
+      "learning_rate": 0.00015468415983887574,
+      "loss": 1.4653,
+      "step": 6316
+    },
+    {
+      "epoch": 1.01072,
+      "grad_norm": 3.162751240300714,
+      "learning_rate": 0.00015466969674376885,
+      "loss": 1.4775,
+      "step": 6317
+    },
+    {
+      "epoch": 1.01088,
+      "grad_norm": 1.9429616311483309,
+      "learning_rate": 0.0001546552320173979,
+      "loss": 1.4381,
+      "step": 6318
+    },
+    {
+      "epoch": 1.01104,
+      "grad_norm": 2.933027217027786,
+      "learning_rate": 0.0001546407656601945,
+      "loss": 1.4378,
+      "step": 6319
+    },
+    {
+      "epoch": 1.0112,
+      "grad_norm": 2.2577238288217805,
+      "learning_rate": 0.00015462629767259024,
+      "loss": 1.4786,
+      "step": 6320
+    },
+    {
+      "epoch": 1.01136,
+      "grad_norm": 2.0101588353897646,
+      "learning_rate": 0.00015461182805501684,
+      "loss": 1.4682,
+      "step": 6321
+    },
+    {
+      "epoch": 1.01152,
+      "grad_norm": 2.7467359517625463,
+      "learning_rate": 0.0001545973568079061,
+      "loss": 1.5305,
+      "step": 6322
+    },
+    {
+      "epoch": 1.01168,
+      "grad_norm": 2.615395428557289,
+      "learning_rate": 0.00015458288393168978,
+      "loss": 1.4631,
+      "step": 6323
+    },
+    {
+      "epoch": 1.01184,
+      "grad_norm": 2.0895234597006143,
+      "learning_rate": 0.00015456840942679976,
+      "loss": 1.5044,
+      "step": 6324
+    },
+    {
+      "epoch": 1.012,
+      "grad_norm": 3.3631690438209576,
+      "learning_rate": 0.00015455393329366786,
+      "loss": 1.4421,
+      "step": 6325
+    },
+    {
+      "epoch": 1.01216,
+      "grad_norm": 2.2486210239545574,
+      "learning_rate": 0.0001545394555327261,
+      "loss": 1.493,
+      "step": 6326
+    },
+    {
+      "epoch": 1.01232,
+      "grad_norm": 2.199183820024461,
+      "learning_rate": 0.00015452497614440648,
+      "loss": 1.5044,
+      "step": 6327
+    },
+    {
+      "epoch": 1.01248,
+      "grad_norm": 2.7517194023650937,
+      "learning_rate": 0.00015451049512914104,
+      "loss": 1.4393,
+      "step": 6328
+    },
+    {
+      "epoch": 1.01264,
+      "grad_norm": 2.21860210356445,
+      "learning_rate": 0.00015449601248736182,
+      "loss": 1.5382,
+      "step": 6329
+    },
+    {
+      "epoch": 1.0128,
+      "grad_norm": 3.073397635844067,
+      "learning_rate": 0.00015448152821950103,
+      "loss": 1.5324,
+      "step": 6330
+    },
+    {
+      "epoch": 1.01296,
+      "grad_norm": 2.6038328506131174,
+      "learning_rate": 0.0001544670423259908,
+      "loss": 1.5066,
+      "step": 6331
+    },
+    {
+      "epoch": 1.01312,
+      "grad_norm": 2.4949849865294924,
+      "learning_rate": 0.0001544525548072634,
+      "loss": 1.481,
+      "step": 6332
+    },
+    {
+      "epoch": 1.01328,
+      "grad_norm": 3.104514368560895,
+      "learning_rate": 0.0001544380656637511,
+      "loss": 1.4279,
+      "step": 6333
+    },
+    {
+      "epoch": 1.01344,
+      "grad_norm": 2.160129579747832,
+      "learning_rate": 0.00015442357489588626,
+      "loss": 1.4387,
+      "step": 6334
+    },
+    {
+      "epoch": 1.0136,
+      "grad_norm": 3.5936681480746184,
+      "learning_rate": 0.00015440908250410125,
+      "loss": 1.395,
+      "step": 6335
+    },
+    {
+      "epoch": 1.01376,
+      "grad_norm": 2.4071022396937005,
+      "learning_rate": 0.0001543945884888285,
+      "loss": 1.416,
+      "step": 6336
+    },
+    {
+      "epoch": 1.01392,
+      "grad_norm": 2.619847072479535,
+      "learning_rate": 0.00015438009285050048,
+      "loss": 1.4237,
+      "step": 6337
+    },
+    {
+      "epoch": 1.01408,
+      "grad_norm": 2.669716097763651,
+      "learning_rate": 0.00015436559558954971,
+      "loss": 1.4449,
+      "step": 6338
+    },
+    {
+      "epoch": 1.01424,
+      "grad_norm": 2.323352983629063,
+      "learning_rate": 0.00015435109670640883,
+      "loss": 1.4467,
+      "step": 6339
+    },
+    {
+      "epoch": 1.0144,
+      "grad_norm": 2.2135945923931635,
+      "learning_rate": 0.00015433659620151042,
+      "loss": 1.5125,
+      "step": 6340
+    },
+    {
+      "epoch": 1.01456,
+      "grad_norm": 2.489692872721047,
+      "learning_rate": 0.00015432209407528712,
+      "loss": 1.4666,
+      "step": 6341
+    },
+    {
+      "epoch": 1.01472,
+      "grad_norm": 2.0844035761982522,
+      "learning_rate": 0.00015430759032817172,
+      "loss": 1.4114,
+      "step": 6342
+    },
+    {
+      "epoch": 1.01488,
+      "grad_norm": 2.3395820037026516,
+      "learning_rate": 0.00015429308496059694,
+      "loss": 1.4729,
+      "step": 6343
+    },
+    {
+      "epoch": 1.01504,
+      "grad_norm": 2.4557533978881465,
+      "learning_rate": 0.00015427857797299562,
+      "loss": 1.5123,
+      "step": 6344
+    },
+    {
+      "epoch": 1.0152,
+      "grad_norm": 2.366250592023503,
+      "learning_rate": 0.00015426406936580063,
+      "loss": 1.5296,
+      "step": 6345
+    },
+    {
+      "epoch": 1.01536,
+      "grad_norm": 2.7837495762207793,
+      "learning_rate": 0.00015424955913944492,
+      "loss": 1.4961,
+      "step": 6346
+    },
+    {
+      "epoch": 1.01552,
+      "grad_norm": 2.2621682675845722,
+      "learning_rate": 0.00015423504729436137,
+      "loss": 1.4659,
+      "step": 6347
+    },
+    {
+      "epoch": 1.01568,
+      "grad_norm": 2.6399561718107307,
+      "learning_rate": 0.00015422053383098305,
+      "loss": 1.5179,
+      "step": 6348
+    },
+    {
+      "epoch": 1.01584,
+      "grad_norm": 2.35356968745512,
+      "learning_rate": 0.000154206018749743,
+      "loss": 1.4802,
+      "step": 6349
+    },
+    {
+      "epoch": 1.016,
+      "grad_norm": 2.078435959823474,
+      "learning_rate": 0.00015419150205107433,
+      "loss": 1.4619,
+      "step": 6350
+    },
+    {
+      "epoch": 1.01616,
+      "grad_norm": 1.9543468468860494,
+      "learning_rate": 0.0001541769837354102,
+      "loss": 1.4128,
+      "step": 6351
+    },
+    {
+      "epoch": 1.01632,
+      "grad_norm": 2.2173361518671646,
+      "learning_rate": 0.00015416246380318385,
+      "loss": 1.4292,
+      "step": 6352
+    },
+    {
+      "epoch": 1.01648,
+      "grad_norm": 2.2934699420974676,
+      "learning_rate": 0.00015414794225482848,
+      "loss": 1.5087,
+      "step": 6353
+    },
+    {
+      "epoch": 1.01664,
+      "grad_norm": 1.754182901160818,
+      "learning_rate": 0.00015413341909077745,
+      "loss": 1.4886,
+      "step": 6354
+    },
+    {
+      "epoch": 1.0168,
+      "grad_norm": 2.125111810605863,
+      "learning_rate": 0.000154118894311464,
+      "loss": 1.4248,
+      "step": 6355
+    },
+    {
+      "epoch": 1.01696,
+      "grad_norm": 2.304177388164402,
+      "learning_rate": 0.00015410436791732168,
+      "loss": 1.4367,
+      "step": 6356
+    },
+    {
+      "epoch": 1.01712,
+      "grad_norm": 2.163472448178267,
+      "learning_rate": 0.00015408983990878382,
+      "loss": 1.4003,
+      "step": 6357
+    },
+    {
+      "epoch": 1.01728,
+      "grad_norm": 2.3074745791258877,
+      "learning_rate": 0.000154075310286284,
+      "loss": 1.4447,
+      "step": 6358
+    },
+    {
+      "epoch": 1.01744,
+      "grad_norm": 2.0193121394233158,
+      "learning_rate": 0.00015406077905025567,
+      "loss": 1.4238,
+      "step": 6359
+    },
+    {
+      "epoch": 1.0176,
+      "grad_norm": 2.055949920602047,
+      "learning_rate": 0.00015404624620113245,
+      "loss": 1.4116,
+      "step": 6360
+    },
+    {
+      "epoch": 1.01776,
+      "grad_norm": 2.374068962667004,
+      "learning_rate": 0.00015403171173934807,
+      "loss": 1.5449,
+      "step": 6361
+    },
+    {
+      "epoch": 1.01792,
+      "grad_norm": 2.2891715627322697,
+      "learning_rate": 0.0001540171756653361,
+      "loss": 1.4887,
+      "step": 6362
+    },
+    {
+      "epoch": 1.01808,
+      "grad_norm": 2.4984117388669227,
+      "learning_rate": 0.0001540026379795303,
+      "loss": 1.4952,
+      "step": 6363
+    },
+    {
+      "epoch": 1.01824,
+      "grad_norm": 2.9507271180662538,
+      "learning_rate": 0.0001539880986823645,
+      "loss": 1.4728,
+      "step": 6364
+    },
+    {
+      "epoch": 1.0184,
+      "grad_norm": 1.9013152231192498,
+      "learning_rate": 0.00015397355777427253,
+      "loss": 1.4194,
+      "step": 6365
+    },
+    {
+      "epoch": 1.01856,
+      "grad_norm": 1.8258502773370184,
+      "learning_rate": 0.00015395901525568819,
+      "loss": 1.4131,
+      "step": 6366
+    },
+    {
+      "epoch": 1.01872,
+      "grad_norm": 2.469315765341725,
+      "learning_rate": 0.00015394447112704549,
+      "loss": 1.3934,
+      "step": 6367
+    },
+    {
+      "epoch": 1.01888,
+      "grad_norm": 2.4962973777332955,
+      "learning_rate": 0.00015392992538877838,
+      "loss": 1.4807,
+      "step": 6368
+    },
+    {
+      "epoch": 1.01904,
+      "grad_norm": 2.37110684242603,
+      "learning_rate": 0.00015391537804132087,
+      "loss": 1.4257,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0192,
+      "grad_norm": 3.2638840324425034,
+      "learning_rate": 0.00015390082908510705,
+      "loss": 1.5156,
+      "step": 6370
+    },
+    {
+      "epoch": 1.01936,
+      "grad_norm": 2.263885652289546,
+      "learning_rate": 0.000153886278520571,
+      "loss": 1.4334,
+      "step": 6371
+    },
+    {
+      "epoch": 1.01952,
+      "grad_norm": 3.6362158679118357,
+      "learning_rate": 0.00015387172634814694,
+      "loss": 1.53,
+      "step": 6372
+    },
+    {
+      "epoch": 1.01968,
+      "grad_norm": 2.142177540862391,
+      "learning_rate": 0.00015385717256826908,
+      "loss": 1.5495,
+      "step": 6373
+    },
+    {
+      "epoch": 1.01984,
+      "grad_norm": 3.2855357042811417,
+      "learning_rate": 0.00015384261718137164,
+      "loss": 1.4467,
+      "step": 6374
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 2.6087599688648604,
+      "learning_rate": 0.000153828060187889,
+      "loss": 1.4273,
+      "step": 6375
+    },
+    {
+      "epoch": 1.02016,
+      "grad_norm": 2.4939587868061093,
+      "learning_rate": 0.00015381350158825545,
+      "loss": 1.4596,
+      "step": 6376
+    },
+    {
+      "epoch": 1.02032,
+      "grad_norm": 3.8156147398886877,
+      "learning_rate": 0.00015379894138290543,
+      "loss": 1.4654,
+      "step": 6377
+    },
+    {
+      "epoch": 1.02048,
+      "grad_norm": 2.304290142875409,
+      "learning_rate": 0.0001537843795722734,
+      "loss": 1.43,
+      "step": 6378
+    },
+    {
+      "epoch": 1.02064,
+      "grad_norm": 4.202542007640451,
+      "learning_rate": 0.00015376981615679387,
+      "loss": 1.5179,
+      "step": 6379
+    },
+    {
+      "epoch": 1.0208,
+      "grad_norm": 2.4700829343715283,
+      "learning_rate": 0.00015375525113690136,
+      "loss": 1.4239,
+      "step": 6380
+    },
+    {
+      "epoch": 1.02096,
+      "grad_norm": 2.845879532375009,
+      "learning_rate": 0.0001537406845130305,
+      "loss": 1.4293,
+      "step": 6381
+    },
+    {
+      "epoch": 1.02112,
+      "grad_norm": 3.091875983490248,
+      "learning_rate": 0.00015372611628561587,
+      "loss": 1.4341,
+      "step": 6382
+    },
+    {
+      "epoch": 1.02128,
+      "grad_norm": 2.0973941346851297,
+      "learning_rate": 0.00015371154645509228,
+      "loss": 1.4438,
+      "step": 6383
+    },
+    {
+      "epoch": 1.02144,
+      "grad_norm": 2.1551858528120014,
+      "learning_rate": 0.0001536969750218944,
+      "loss": 1.4213,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0216,
+      "grad_norm": 2.528324037856621,
+      "learning_rate": 0.000153682401986457,
+      "loss": 1.5101,
+      "step": 6385
+    },
+    {
+      "epoch": 1.02176,
+      "grad_norm": 2.2332381534121915,
+      "learning_rate": 0.00015366782734921495,
+      "loss": 1.4887,
+      "step": 6386
+    },
+    {
+      "epoch": 1.02192,
+      "grad_norm": 2.855632082766148,
+      "learning_rate": 0.00015365325111060316,
+      "loss": 1.4789,
+      "step": 6387
+    },
+    {
+      "epoch": 1.02208,
+      "grad_norm": 2.165441970394895,
+      "learning_rate": 0.00015363867327105655,
+      "loss": 1.4511,
+      "step": 6388
+    },
+    {
+      "epoch": 1.02224,
+      "grad_norm": 2.1665344071573287,
+      "learning_rate": 0.00015362409383101006,
+      "loss": 1.4681,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0224,
+      "grad_norm": 2.3719160523627187,
+      "learning_rate": 0.00015360951279089875,
+      "loss": 1.4318,
+      "step": 6390
+    },
+    {
+      "epoch": 1.02256,
+      "grad_norm": 2.356296458388781,
+      "learning_rate": 0.00015359493015115768,
+      "loss": 1.4797,
+      "step": 6391
+    },
+    {
+      "epoch": 1.02272,
+      "grad_norm": 2.1643709670187814,
+      "learning_rate": 0.00015358034591222205,
+      "loss": 1.4741,
+      "step": 6392
+    },
+    {
+      "epoch": 1.02288,
+      "grad_norm": 2.3548784385516783,
+      "learning_rate": 0.00015356576007452693,
+      "loss": 1.4862,
+      "step": 6393
+    },
+    {
+      "epoch": 1.02304,
+      "grad_norm": 2.174559469520946,
+      "learning_rate": 0.00015355117263850753,
+      "loss": 1.4169,
+      "step": 6394
+    },
+    {
+      "epoch": 1.0232,
+      "grad_norm": 2.1317549670592437,
+      "learning_rate": 0.00015353658360459926,
+      "loss": 1.4209,
+      "step": 6395
+    },
+    {
+      "epoch": 1.02336,
+      "grad_norm": 2.257949672075298,
+      "learning_rate": 0.0001535219929732373,
+      "loss": 1.4952,
+      "step": 6396
+    },
+    {
+      "epoch": 1.02352,
+      "grad_norm": 1.964471371811003,
+      "learning_rate": 0.00015350740074485708,
+      "loss": 1.4299,
+      "step": 6397
+    },
+    {
+      "epoch": 1.02368,
+      "grad_norm": 2.2811250659962217,
+      "learning_rate": 0.000153492806919894,
+      "loss": 1.4416,
+      "step": 6398
+    },
+    {
+      "epoch": 1.02384,
+      "grad_norm": 1.8198648146897092,
+      "learning_rate": 0.00015347821149878346,
+      "loss": 1.4105,
+      "step": 6399
+    },
+    {
+      "epoch": 1.024,
+      "grad_norm": 2.012655148170196,
+      "learning_rate": 0.00015346361448196106,
+      "loss": 1.445,
+      "step": 6400
+    },
+    {
+      "epoch": 1.02416,
+      "grad_norm": 2.1788012499431386,
+      "learning_rate": 0.0001534490158698623,
+      "loss": 1.4605,
+      "step": 6401
+    },
+    {
+      "epoch": 1.02432,
+      "grad_norm": 2.0699400167124966,
+      "learning_rate": 0.0001534344156629228,
+      "loss": 1.5009,
+      "step": 6402
+    },
+    {
+      "epoch": 1.02448,
+      "grad_norm": 2.2654669580517917,
+      "learning_rate": 0.00015341981386157816,
+      "loss": 1.4811,
+      "step": 6403
+    },
+    {
+      "epoch": 1.02464,
+      "grad_norm": 2.051830829670822,
+      "learning_rate": 0.00015340521046626414,
+      "loss": 1.49,
+      "step": 6404
+    },
+    {
+      "epoch": 1.0248,
+      "grad_norm": 2.5558737347716662,
+      "learning_rate": 0.0001533906054774165,
+      "loss": 1.4694,
+      "step": 6405
+    },
+    {
+      "epoch": 1.02496,
+      "grad_norm": 2.0804418629185792,
+      "learning_rate": 0.00015337599889547092,
+      "loss": 1.4204,
+      "step": 6406
+    },
+    {
+      "epoch": 1.02512,
+      "grad_norm": 3.3551991000340684,
+      "learning_rate": 0.00015336139072086335,
+      "loss": 1.4401,
+      "step": 6407
+    },
+    {
+      "epoch": 1.02528,
+      "grad_norm": 2.060047971282045,
+      "learning_rate": 0.0001533467809540296,
+      "loss": 1.4194,
+      "step": 6408
+    },
+    {
+      "epoch": 1.02544,
+      "grad_norm": 3.046504253670365,
+      "learning_rate": 0.0001533321695954057,
+      "loss": 1.4442,
+      "step": 6409
+    },
+    {
+      "epoch": 1.0256,
+      "grad_norm": 2.269350703972266,
+      "learning_rate": 0.00015331755664542752,
+      "loss": 1.5326,
+      "step": 6410
+    },
+    {
+      "epoch": 1.02576,
+      "grad_norm": 2.618680809664444,
+      "learning_rate": 0.00015330294210453115,
+      "loss": 1.4865,
+      "step": 6411
+    },
+    {
+      "epoch": 1.02592,
+      "grad_norm": 2.382600378000578,
+      "learning_rate": 0.00015328832597315267,
+      "loss": 1.4381,
+      "step": 6412
+    },
+    {
+      "epoch": 1.02608,
+      "grad_norm": 2.2953238589825236,
+      "learning_rate": 0.0001532737082517282,
+      "loss": 1.4526,
+      "step": 6413
+    },
+    {
+      "epoch": 1.02624,
+      "grad_norm": 2.8465821774771682,
+      "learning_rate": 0.0001532590889406939,
+      "loss": 1.5845,
+      "step": 6414
+    },
+    {
+      "epoch": 1.0264,
+      "grad_norm": 2.6665439038973835,
+      "learning_rate": 0.000153244468040486,
+      "loss": 1.4523,
+      "step": 6415
+    },
+    {
+      "epoch": 1.02656,
+      "grad_norm": 1.9341196408433612,
+      "learning_rate": 0.00015322984555154076,
+      "loss": 1.4983,
+      "step": 6416
+    },
+    {
+      "epoch": 1.02672,
+      "grad_norm": 2.511807416740225,
+      "learning_rate": 0.00015321522147429447,
+      "loss": 1.4511,
+      "step": 6417
+    },
+    {
+      "epoch": 1.02688,
+      "grad_norm": 2.035840025047668,
+      "learning_rate": 0.00015320059580918352,
+      "loss": 1.4671,
+      "step": 6418
+    },
+    {
+      "epoch": 1.02704,
+      "grad_norm": 2.2983096741551474,
+      "learning_rate": 0.0001531859685566443,
+      "loss": 1.3824,
+      "step": 6419
+    },
+    {
+      "epoch": 1.0272,
+      "grad_norm": 2.140914298745621,
+      "learning_rate": 0.00015317133971711331,
+      "loss": 1.4416,
+      "step": 6420
+    },
+    {
+      "epoch": 1.02736,
+      "grad_norm": 2.3360626140749474,
+      "learning_rate": 0.000153156709291027,
+      "loss": 1.401,
+      "step": 6421
+    },
+    {
+      "epoch": 1.02752,
+      "grad_norm": 2.509186752635013,
+      "learning_rate": 0.00015314207727882195,
+      "loss": 1.4216,
+      "step": 6422
+    },
+    {
+      "epoch": 1.02768,
+      "grad_norm": 2.3571682588287057,
+      "learning_rate": 0.00015312744368093476,
+      "loss": 1.4461,
+      "step": 6423
+    },
+    {
+      "epoch": 1.02784,
+      "grad_norm": 2.402588403875387,
+      "learning_rate": 0.00015311280849780205,
+      "loss": 1.466,
+      "step": 6424
+    },
+    {
+      "epoch": 1.028,
+      "grad_norm": 2.2423750757331558,
+      "learning_rate": 0.00015309817172986053,
+      "loss": 1.5457,
+      "step": 6425
+    },
+    {
+      "epoch": 1.02816,
+      "grad_norm": 2.307066568915049,
+      "learning_rate": 0.00015308353337754692,
+      "loss": 1.4849,
+      "step": 6426
+    },
+    {
+      "epoch": 1.02832,
+      "grad_norm": 1.96761174401391,
+      "learning_rate": 0.00015306889344129807,
+      "loss": 1.4927,
+      "step": 6427
+    },
+    {
+      "epoch": 1.02848,
+      "grad_norm": 2.7819513121146393,
+      "learning_rate": 0.00015305425192155075,
+      "loss": 1.4885,
+      "step": 6428
+    },
+    {
+      "epoch": 1.02864,
+      "grad_norm": 2.2841020888341204,
+      "learning_rate": 0.00015303960881874184,
+      "loss": 1.4999,
+      "step": 6429
+    },
+    {
+      "epoch": 1.0288,
+      "grad_norm": 2.574339217947863,
+      "learning_rate": 0.00015302496413330832,
+      "loss": 1.4876,
+      "step": 6430
+    },
+    {
+      "epoch": 1.02896,
+      "grad_norm": 1.8412185242028096,
+      "learning_rate": 0.00015301031786568711,
+      "loss": 1.5041,
+      "step": 6431
+    },
+    {
+      "epoch": 1.02912,
+      "grad_norm": 2.239904751835625,
+      "learning_rate": 0.00015299567001631528,
+      "loss": 1.4571,
+      "step": 6432
+    },
+    {
+      "epoch": 1.02928,
+      "grad_norm": 2.7468733677739543,
+      "learning_rate": 0.00015298102058562986,
+      "loss": 1.4432,
+      "step": 6433
+    },
+    {
+      "epoch": 1.02944,
+      "grad_norm": 2.015726191700404,
+      "learning_rate": 0.000152966369574068,
+      "loss": 1.4097,
+      "step": 6434
+    },
+    {
+      "epoch": 1.0296,
+      "grad_norm": 2.1968975236170993,
+      "learning_rate": 0.00015295171698206686,
+      "loss": 1.4419,
+      "step": 6435
+    },
+    {
+      "epoch": 1.02976,
+      "grad_norm": 2.19338634013079,
+      "learning_rate": 0.00015293706281006362,
+      "loss": 1.4636,
+      "step": 6436
+    },
+    {
+      "epoch": 1.02992,
+      "grad_norm": 2.2317687945539046,
+      "learning_rate": 0.00015292240705849558,
+      "loss": 1.5009,
+      "step": 6437
+    },
+    {
+      "epoch": 1.03008,
+      "grad_norm": 2.470704590113539,
+      "learning_rate": 0.00015290774972780003,
+      "loss": 1.4741,
+      "step": 6438
+    },
+    {
+      "epoch": 1.03024,
+      "grad_norm": 2.3220136842816244,
+      "learning_rate": 0.00015289309081841431,
+      "loss": 1.4187,
+      "step": 6439
+    },
+    {
+      "epoch": 1.0304,
+      "grad_norm": 2.3084621445509983,
+      "learning_rate": 0.00015287843033077585,
+      "loss": 1.4013,
+      "step": 6440
+    },
+    {
+      "epoch": 1.03056,
+      "grad_norm": 2.625422380410288,
+      "learning_rate": 0.0001528637682653221,
+      "loss": 1.5191,
+      "step": 6441
+    },
+    {
+      "epoch": 1.03072,
+      "grad_norm": 2.2999321409461815,
+      "learning_rate": 0.00015284910462249048,
+      "loss": 1.3967,
+      "step": 6442
+    },
+    {
+      "epoch": 1.03088,
+      "grad_norm": 1.8238760857331993,
+      "learning_rate": 0.00015283443940271863,
+      "loss": 1.4008,
+      "step": 6443
+    },
+    {
+      "epoch": 1.03104,
+      "grad_norm": 2.264990313986975,
+      "learning_rate": 0.00015281977260644407,
+      "loss": 1.5037,
+      "step": 6444
+    },
+    {
+      "epoch": 1.0312,
+      "grad_norm": 2.2630154623550194,
+      "learning_rate": 0.00015280510423410447,
+      "loss": 1.4837,
+      "step": 6445
+    },
+    {
+      "epoch": 1.03136,
+      "grad_norm": 2.6238355229917185,
+      "learning_rate": 0.00015279043428613752,
+      "loss": 1.5495,
+      "step": 6446
+    },
+    {
+      "epoch": 1.03152,
+      "grad_norm": 2.1685375204268307,
+      "learning_rate": 0.00015277576276298094,
+      "loss": 1.4816,
+      "step": 6447
+    },
+    {
+      "epoch": 1.03168,
+      "grad_norm": 2.1831615895544974,
+      "learning_rate": 0.00015276108966507247,
+      "loss": 1.6002,
+      "step": 6448
+    },
+    {
+      "epoch": 1.03184,
+      "grad_norm": 2.661170279081632,
+      "learning_rate": 0.00015274641499285,
+      "loss": 1.5541,
+      "step": 6449
+    },
+    {
+      "epoch": 1.032,
+      "grad_norm": 2.25833144799881,
+      "learning_rate": 0.00015273173874675132,
+      "loss": 1.4216,
+      "step": 6450
+    },
+    {
+      "epoch": 1.03216,
+      "grad_norm": 2.1260071320158316,
+      "learning_rate": 0.00015271706092721444,
+      "loss": 1.4972,
+      "step": 6451
+    },
+    {
+      "epoch": 1.03232,
+      "grad_norm": 2.369330556284753,
+      "learning_rate": 0.00015270238153467723,
+      "loss": 1.4027,
+      "step": 6452
+    },
+    {
+      "epoch": 1.03248,
+      "grad_norm": 2.174756622674578,
+      "learning_rate": 0.00015268770056957778,
+      "loss": 1.3947,
+      "step": 6453
+    },
+    {
+      "epoch": 1.03264,
+      "grad_norm": 2.1286644695371497,
+      "learning_rate": 0.00015267301803235413,
+      "loss": 1.5001,
+      "step": 6454
+    },
+    {
+      "epoch": 1.0328,
+      "grad_norm": 2.271424704828945,
+      "learning_rate": 0.00015265833392344436,
+      "loss": 1.51,
+      "step": 6455
+    },
+    {
+      "epoch": 1.03296,
+      "grad_norm": 2.2594472308377496,
+      "learning_rate": 0.00015264364824328665,
+      "loss": 1.4928,
+      "step": 6456
+    },
+    {
+      "epoch": 1.03312,
+      "grad_norm": 2.091279111272941,
+      "learning_rate": 0.00015262896099231917,
+      "loss": 1.4283,
+      "step": 6457
+    },
+    {
+      "epoch": 1.03328,
+      "grad_norm": 2.3209351558120392,
+      "learning_rate": 0.0001526142721709802,
+      "loss": 1.4251,
+      "step": 6458
+    },
+    {
+      "epoch": 1.03344,
+      "grad_norm": 2.194676660018619,
+      "learning_rate": 0.000152599581779708,
+      "loss": 1.4826,
+      "step": 6459
+    },
+    {
+      "epoch": 1.0336,
+      "grad_norm": 2.3407186255941435,
+      "learning_rate": 0.0001525848898189409,
+      "loss": 1.4961,
+      "step": 6460
+    },
+    {
+      "epoch": 1.03376,
+      "grad_norm": 2.162173879710814,
+      "learning_rate": 0.00015257019628911738,
+      "loss": 1.5083,
+      "step": 6461
+    },
+    {
+      "epoch": 1.03392,
+      "grad_norm": 1.9079082164825567,
+      "learning_rate": 0.00015255550119067576,
+      "loss": 1.4684,
+      "step": 6462
+    },
+    {
+      "epoch": 1.0340799999999999,
+      "grad_norm": 2.5873225479964823,
+      "learning_rate": 0.00015254080452405458,
+      "loss": 1.4162,
+      "step": 6463
+    },
+    {
+      "epoch": 1.03424,
+      "grad_norm": 2.692601168255951,
+      "learning_rate": 0.00015252610628969232,
+      "loss": 1.4556,
+      "step": 6464
+    },
+    {
+      "epoch": 1.0344,
+      "grad_norm": 2.372458386952236,
+      "learning_rate": 0.00015251140648802762,
+      "loss": 1.4663,
+      "step": 6465
+    },
+    {
+      "epoch": 1.03456,
+      "grad_norm": 3.2575150350589595,
+      "learning_rate": 0.00015249670511949907,
+      "loss": 1.4024,
+      "step": 6466
+    },
+    {
+      "epoch": 1.03472,
+      "grad_norm": 2.172980254205086,
+      "learning_rate": 0.0001524820021845453,
+      "loss": 1.4573,
+      "step": 6467
+    },
+    {
+      "epoch": 1.03488,
+      "grad_norm": 2.7068712774803543,
+      "learning_rate": 0.0001524672976836051,
+      "loss": 1.5249,
+      "step": 6468
+    },
+    {
+      "epoch": 1.03504,
+      "grad_norm": 2.7077366235702143,
+      "learning_rate": 0.0001524525916171172,
+      "loss": 1.451,
+      "step": 6469
+    },
+    {
+      "epoch": 1.0352,
+      "grad_norm": 2.6456824908510748,
+      "learning_rate": 0.00015243788398552037,
+      "loss": 1.3536,
+      "step": 6470
+    },
+    {
+      "epoch": 1.03536,
+      "grad_norm": 3.039471502591546,
+      "learning_rate": 0.0001524231747892535,
+      "loss": 1.523,
+      "step": 6471
+    },
+    {
+      "epoch": 1.03552,
+      "grad_norm": 1.8628713252975606,
+      "learning_rate": 0.00015240846402875548,
+      "loss": 1.4427,
+      "step": 6472
+    },
+    {
+      "epoch": 1.03568,
+      "grad_norm": 2.4391999328640255,
+      "learning_rate": 0.00015239375170446526,
+      "loss": 1.4637,
+      "step": 6473
+    },
+    {
+      "epoch": 1.03584,
+      "grad_norm": 2.1508186401912592,
+      "learning_rate": 0.00015237903781682182,
+      "loss": 1.4971,
+      "step": 6474
+    },
+    {
+      "epoch": 1.036,
+      "grad_norm": 2.2966788161682783,
+      "learning_rate": 0.0001523643223662642,
+      "loss": 1.469,
+      "step": 6475
+    },
+    {
+      "epoch": 1.03616,
+      "grad_norm": 1.898557667776354,
+      "learning_rate": 0.00015234960535323157,
+      "loss": 1.4784,
+      "step": 6476
+    },
+    {
+      "epoch": 1.03632,
+      "grad_norm": 2.1589705282563236,
+      "learning_rate": 0.00015233488677816295,
+      "loss": 1.457,
+      "step": 6477
+    },
+    {
+      "epoch": 1.03648,
+      "grad_norm": 2.2514087987349063,
+      "learning_rate": 0.0001523201666414976,
+      "loss": 1.4443,
+      "step": 6478
+    },
+    {
+      "epoch": 1.03664,
+      "grad_norm": 2.4349960390729954,
+      "learning_rate": 0.00015230544494367466,
+      "loss": 1.433,
+      "step": 6479
+    },
+    {
+      "epoch": 1.0368,
+      "grad_norm": 2.480212000405424,
+      "learning_rate": 0.00015229072168513352,
+      "loss": 1.5894,
+      "step": 6480
+    },
+    {
+      "epoch": 1.03696,
+      "grad_norm": 2.563960909204764,
+      "learning_rate": 0.00015227599686631343,
+      "loss": 1.4223,
+      "step": 6481
+    },
+    {
+      "epoch": 1.03712,
+      "grad_norm": 2.226953378034301,
+      "learning_rate": 0.00015226127048765376,
+      "loss": 1.4554,
+      "step": 6482
+    },
+    {
+      "epoch": 1.03728,
+      "grad_norm": 2.143952523506342,
+      "learning_rate": 0.0001522465425495939,
+      "loss": 1.4748,
+      "step": 6483
+    },
+    {
+      "epoch": 1.03744,
+      "grad_norm": 1.9673446570737703,
+      "learning_rate": 0.00015223181305257336,
+      "loss": 1.462,
+      "step": 6484
+    },
+    {
+      "epoch": 1.0376,
+      "grad_norm": 2.4153770511495845,
+      "learning_rate": 0.00015221708199703165,
+      "loss": 1.4224,
+      "step": 6485
+    },
+    {
+      "epoch": 1.03776,
+      "grad_norm": 2.133265729222153,
+      "learning_rate": 0.0001522023493834083,
+      "loss": 1.4403,
+      "step": 6486
+    },
+    {
+      "epoch": 1.03792,
+      "grad_norm": 2.141300417105501,
+      "learning_rate": 0.00015218761521214286,
+      "loss": 1.4182,
+      "step": 6487
+    },
+    {
+      "epoch": 1.03808,
+      "grad_norm": 2.6755681439123657,
+      "learning_rate": 0.0001521728794836751,
+      "loss": 1.4248,
+      "step": 6488
+    },
+    {
+      "epoch": 1.03824,
+      "grad_norm": 2.138781991011542,
+      "learning_rate": 0.0001521581421984446,
+      "loss": 1.422,
+      "step": 6489
+    },
+    {
+      "epoch": 1.0384,
+      "grad_norm": 2.1110751606633404,
+      "learning_rate": 0.00015214340335689114,
+      "loss": 1.4243,
+      "step": 6490
+    },
+    {
+      "epoch": 1.03856,
+      "grad_norm": 2.773232337148975,
+      "learning_rate": 0.0001521286629594545,
+      "loss": 1.4938,
+      "step": 6491
+    },
+    {
+      "epoch": 1.03872,
+      "grad_norm": 2.1968134268523056,
+      "learning_rate": 0.0001521139210065745,
+      "loss": 1.3597,
+      "step": 6492
+    },
+    {
+      "epoch": 1.03888,
+      "grad_norm": 2.2757993673210795,
+      "learning_rate": 0.0001520991774986911,
+      "loss": 1.4165,
+      "step": 6493
+    },
+    {
+      "epoch": 1.03904,
+      "grad_norm": 2.5868816257988425,
+      "learning_rate": 0.00015208443243624412,
+      "loss": 1.4373,
+      "step": 6494
+    },
+    {
+      "epoch": 1.0392,
+      "grad_norm": 2.3583116671455464,
+      "learning_rate": 0.00015206968581967355,
+      "loss": 1.4529,
+      "step": 6495
+    },
+    {
+      "epoch": 1.03936,
+      "grad_norm": 3.0196676545293104,
+      "learning_rate": 0.00015205493764941947,
+      "loss": 1.5947,
+      "step": 6496
+    },
+    {
+      "epoch": 1.03952,
+      "grad_norm": 2.0161082524825646,
+      "learning_rate": 0.00015204018792592188,
+      "loss": 1.47,
+      "step": 6497
+    },
+    {
+      "epoch": 1.03968,
+      "grad_norm": 2.113100093542495,
+      "learning_rate": 0.00015202543664962094,
+      "loss": 1.4728,
+      "step": 6498
+    },
+    {
+      "epoch": 1.03984,
+      "grad_norm": 2.8800924059201782,
+      "learning_rate": 0.00015201068382095673,
+      "loss": 1.524,
+      "step": 6499
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 2.035598765834644,
+      "learning_rate": 0.00015199592944036954,
+      "loss": 1.4291,
+      "step": 6500
+    },
+    {
+      "epoch": 1.04016,
+      "grad_norm": 2.4195953793810063,
+      "learning_rate": 0.00015198117350829956,
+      "loss": 1.4545,
+      "step": 6501
+    },
+    {
+      "epoch": 1.04032,
+      "grad_norm": 2.066627634917977,
+      "learning_rate": 0.0001519664160251871,
+      "loss": 1.4165,
+      "step": 6502
+    },
+    {
+      "epoch": 1.04048,
+      "grad_norm": 2.0392711627869566,
+      "learning_rate": 0.00015195165699147256,
+      "loss": 1.4453,
+      "step": 6503
+    },
+    {
+      "epoch": 1.04064,
+      "grad_norm": 2.3521337181289863,
+      "learning_rate": 0.00015193689640759624,
+      "loss": 1.4526,
+      "step": 6504
+    },
+    {
+      "epoch": 1.0408,
+      "grad_norm": 2.2283372316253445,
+      "learning_rate": 0.00015192213427399863,
+      "loss": 1.4334,
+      "step": 6505
+    },
+    {
+      "epoch": 1.04096,
+      "grad_norm": 2.792275806216547,
+      "learning_rate": 0.00015190737059112016,
+      "loss": 1.5177,
+      "step": 6506
+    },
+    {
+      "epoch": 1.04112,
+      "grad_norm": 2.1801274816496234,
+      "learning_rate": 0.00015189260535940138,
+      "loss": 1.5142,
+      "step": 6507
+    },
+    {
+      "epoch": 1.04128,
+      "grad_norm": 2.4810306237002058,
+      "learning_rate": 0.00015187783857928291,
+      "loss": 1.4158,
+      "step": 6508
+    },
+    {
+      "epoch": 1.04144,
+      "grad_norm": 2.3925304211269163,
+      "learning_rate": 0.0001518630702512053,
+      "loss": 1.4734,
+      "step": 6509
+    },
+    {
+      "epoch": 1.0416,
+      "grad_norm": 2.3725424405909408,
+      "learning_rate": 0.00015184830037560926,
+      "loss": 1.4274,
+      "step": 6510
+    },
+    {
+      "epoch": 1.04176,
+      "grad_norm": 2.307147372918676,
+      "learning_rate": 0.0001518335289529355,
+      "loss": 1.4379,
+      "step": 6511
+    },
+    {
+      "epoch": 1.04192,
+      "grad_norm": 2.2104055553950768,
+      "learning_rate": 0.00015181875598362476,
+      "loss": 1.5039,
+      "step": 6512
+    },
+    {
+      "epoch": 1.04208,
+      "grad_norm": 2.3542774346062294,
+      "learning_rate": 0.00015180398146811785,
+      "loss": 1.4276,
+      "step": 6513
+    },
+    {
+      "epoch": 1.04224,
+      "grad_norm": 2.1170702214514807,
+      "learning_rate": 0.0001517892054068556,
+      "loss": 1.48,
+      "step": 6514
+    },
+    {
+      "epoch": 1.0424,
+      "grad_norm": 2.1670787073137556,
+      "learning_rate": 0.00015177442780027898,
+      "loss": 1.4751,
+      "step": 6515
+    },
+    {
+      "epoch": 1.04256,
+      "grad_norm": 2.1754569840932105,
+      "learning_rate": 0.0001517596486488288,
+      "loss": 1.4428,
+      "step": 6516
+    },
+    {
+      "epoch": 1.04272,
+      "grad_norm": 2.4926147252286945,
+      "learning_rate": 0.00015174486795294616,
+      "loss": 1.4926,
+      "step": 6517
+    },
+    {
+      "epoch": 1.04288,
+      "grad_norm": 1.9461610478273492,
+      "learning_rate": 0.00015173008571307207,
+      "loss": 1.3992,
+      "step": 6518
+    },
+    {
+      "epoch": 1.04304,
+      "grad_norm": 2.1608730034264982,
+      "learning_rate": 0.0001517153019296476,
+      "loss": 1.5736,
+      "step": 6519
+    },
+    {
+      "epoch": 1.0432,
+      "grad_norm": 2.049055167386426,
+      "learning_rate": 0.00015170051660311389,
+      "loss": 1.4613,
+      "step": 6520
+    },
+    {
+      "epoch": 1.04336,
+      "grad_norm": 2.0420152579232433,
+      "learning_rate": 0.00015168572973391209,
+      "loss": 1.4191,
+      "step": 6521
+    },
+    {
+      "epoch": 1.04352,
+      "grad_norm": 2.4144518124047867,
+      "learning_rate": 0.0001516709413224834,
+      "loss": 1.5308,
+      "step": 6522
+    },
+    {
+      "epoch": 1.04368,
+      "grad_norm": 1.9397585004373885,
+      "learning_rate": 0.00015165615136926913,
+      "loss": 1.3889,
+      "step": 6523
+    },
+    {
+      "epoch": 1.04384,
+      "grad_norm": 2.162310000243197,
+      "learning_rate": 0.00015164135987471062,
+      "loss": 1.4927,
+      "step": 6524
+    },
+    {
+      "epoch": 1.044,
+      "grad_norm": 1.9911149208281327,
+      "learning_rate": 0.0001516265668392491,
+      "loss": 1.4245,
+      "step": 6525
+    },
+    {
+      "epoch": 1.04416,
+      "grad_norm": 2.459612872931315,
+      "learning_rate": 0.0001516117722633261,
+      "loss": 1.4588,
+      "step": 6526
+    },
+    {
+      "epoch": 1.04432,
+      "grad_norm": 2.0734850926880077,
+      "learning_rate": 0.00015159697614738302,
+      "loss": 1.489,
+      "step": 6527
+    },
+    {
+      "epoch": 1.04448,
+      "grad_norm": 1.9715303771415396,
+      "learning_rate": 0.00015158217849186136,
+      "loss": 1.4879,
+      "step": 6528
+    },
+    {
+      "epoch": 1.04464,
+      "grad_norm": 1.9329599855460708,
+      "learning_rate": 0.00015156737929720264,
+      "loss": 1.4779,
+      "step": 6529
+    },
+    {
+      "epoch": 1.0448,
+      "grad_norm": 2.4674842343377326,
+      "learning_rate": 0.00015155257856384847,
+      "loss": 1.4694,
+      "step": 6530
+    },
+    {
+      "epoch": 1.04496,
+      "grad_norm": 2.3829502978992285,
+      "learning_rate": 0.0001515377762922405,
+      "loss": 1.482,
+      "step": 6531
+    },
+    {
+      "epoch": 1.04512,
+      "grad_norm": 2.5414446020398724,
+      "learning_rate": 0.00015152297248282036,
+      "loss": 1.5401,
+      "step": 6532
+    },
+    {
+      "epoch": 1.04528,
+      "grad_norm": 2.7270922700574163,
+      "learning_rate": 0.0001515081671360298,
+      "loss": 1.4805,
+      "step": 6533
+    },
+    {
+      "epoch": 1.04544,
+      "grad_norm": 2.0190516297980303,
+      "learning_rate": 0.0001514933602523106,
+      "loss": 1.4944,
+      "step": 6534
+    },
+    {
+      "epoch": 1.0456,
+      "grad_norm": 2.3802944442316902,
+      "learning_rate": 0.00015147855183210454,
+      "loss": 1.464,
+      "step": 6535
+    },
+    {
+      "epoch": 1.04576,
+      "grad_norm": 3.0255700345095766,
+      "learning_rate": 0.0001514637418758535,
+      "loss": 1.416,
+      "step": 6536
+    },
+    {
+      "epoch": 1.04592,
+      "grad_norm": 2.2763157348262344,
+      "learning_rate": 0.00015144893038399943,
+      "loss": 1.4586,
+      "step": 6537
+    },
+    {
+      "epoch": 1.04608,
+      "grad_norm": 2.93525542385918,
+      "learning_rate": 0.00015143411735698422,
+      "loss": 1.497,
+      "step": 6538
+    },
+    {
+      "epoch": 1.04624,
+      "grad_norm": 2.567568892421156,
+      "learning_rate": 0.00015141930279524991,
+      "loss": 1.4682,
+      "step": 6539
+    },
+    {
+      "epoch": 1.0464,
+      "grad_norm": 2.262229953410342,
+      "learning_rate": 0.00015140448669923853,
+      "loss": 1.3715,
+      "step": 6540
+    },
+    {
+      "epoch": 1.04656,
+      "grad_norm": 2.934516414708315,
+      "learning_rate": 0.00015138966906939216,
+      "loss": 1.4543,
+      "step": 6541
+    },
+    {
+      "epoch": 1.04672,
+      "grad_norm": 2.195473464244021,
+      "learning_rate": 0.00015137484990615296,
+      "loss": 1.4571,
+      "step": 6542
+    },
+    {
+      "epoch": 1.04688,
+      "grad_norm": 2.1313718054573285,
+      "learning_rate": 0.00015136002920996308,
+      "loss": 1.4752,
+      "step": 6543
+    },
+    {
+      "epoch": 1.04704,
+      "grad_norm": 2.3212868867915306,
+      "learning_rate": 0.00015134520698126478,
+      "loss": 1.4613,
+      "step": 6544
+    },
+    {
+      "epoch": 1.0472,
+      "grad_norm": 2.3679073133296655,
+      "learning_rate": 0.00015133038322050034,
+      "loss": 1.5306,
+      "step": 6545
+    },
+    {
+      "epoch": 1.04736,
+      "grad_norm": 3.3092805023799117,
+      "learning_rate": 0.00015131555792811203,
+      "loss": 1.4368,
+      "step": 6546
+    },
+    {
+      "epoch": 1.04752,
+      "grad_norm": 2.444529665159795,
+      "learning_rate": 0.00015130073110454227,
+      "loss": 1.4934,
+      "step": 6547
+    },
+    {
+      "epoch": 1.04768,
+      "grad_norm": 2.525381673039518,
+      "learning_rate": 0.00015128590275023345,
+      "loss": 1.4547,
+      "step": 6548
+    },
+    {
+      "epoch": 1.04784,
+      "grad_norm": 2.6807006456488005,
+      "learning_rate": 0.00015127107286562797,
+      "loss": 1.493,
+      "step": 6549
+    },
+    {
+      "epoch": 1.048,
+      "grad_norm": 2.362763255986489,
+      "learning_rate": 0.00015125624145116844,
+      "loss": 1.4369,
+      "step": 6550
+    },
+    {
+      "epoch": 1.04816,
+      "grad_norm": 2.951583175418506,
+      "learning_rate": 0.00015124140850729734,
+      "loss": 1.4686,
+      "step": 6551
+    },
+    {
+      "epoch": 1.04832,
+      "grad_norm": 2.1648845771207785,
+      "learning_rate": 0.00015122657403445725,
+      "loss": 1.4473,
+      "step": 6552
+    },
+    {
+      "epoch": 1.04848,
+      "grad_norm": 2.113171489107443,
+      "learning_rate": 0.00015121173803309089,
+      "loss": 1.5067,
+      "step": 6553
+    },
+    {
+      "epoch": 1.04864,
+      "grad_norm": 2.406285135514398,
+      "learning_rate": 0.00015119690050364085,
+      "loss": 1.5683,
+      "step": 6554
+    },
+    {
+      "epoch": 1.0488,
+      "grad_norm": 2.3170423510530345,
+      "learning_rate": 0.00015118206144654992,
+      "loss": 1.4607,
+      "step": 6555
+    },
+    {
+      "epoch": 1.04896,
+      "grad_norm": 1.9236884329184984,
+      "learning_rate": 0.00015116722086226084,
+      "loss": 1.429,
+      "step": 6556
+    },
+    {
+      "epoch": 1.04912,
+      "grad_norm": 2.035528564100824,
+      "learning_rate": 0.00015115237875121644,
+      "loss": 1.4756,
+      "step": 6557
+    },
+    {
+      "epoch": 1.04928,
+      "grad_norm": 1.9792279083283322,
+      "learning_rate": 0.0001511375351138596,
+      "loss": 1.4425,
+      "step": 6558
+    },
+    {
+      "epoch": 1.04944,
+      "grad_norm": 2.1669841512801375,
+      "learning_rate": 0.00015112268995063328,
+      "loss": 1.4233,
+      "step": 6559
+    },
+    {
+      "epoch": 1.0496,
+      "grad_norm": 2.7141951403442452,
+      "learning_rate": 0.00015110784326198033,
+      "loss": 1.4875,
+      "step": 6560
+    },
+    {
+      "epoch": 1.04976,
+      "grad_norm": 1.97001058158236,
+      "learning_rate": 0.00015109299504834382,
+      "loss": 1.4658,
+      "step": 6561
+    },
+    {
+      "epoch": 1.04992,
+      "grad_norm": 2.94004792818516,
+      "learning_rate": 0.0001510781453101668,
+      "loss": 1.3932,
+      "step": 6562
+    },
+    {
+      "epoch": 1.05008,
+      "grad_norm": 1.9917835162506392,
+      "learning_rate": 0.00015106329404789235,
+      "loss": 1.41,
+      "step": 6563
+    },
+    {
+      "epoch": 1.05024,
+      "grad_norm": 2.4396265636859646,
+      "learning_rate": 0.00015104844126196358,
+      "loss": 1.4856,
+      "step": 6564
+    },
+    {
+      "epoch": 1.0504,
+      "grad_norm": 2.157129030002828,
+      "learning_rate": 0.00015103358695282374,
+      "loss": 1.4735,
+      "step": 6565
+    },
+    {
+      "epoch": 1.05056,
+      "grad_norm": 2.1889513680995036,
+      "learning_rate": 0.00015101873112091606,
+      "loss": 1.5088,
+      "step": 6566
+    },
+    {
+      "epoch": 1.05072,
+      "grad_norm": 1.922119911669561,
+      "learning_rate": 0.0001510038737666837,
+      "loss": 1.4855,
+      "step": 6567
+    },
+    {
+      "epoch": 1.05088,
+      "grad_norm": 2.339879029327738,
+      "learning_rate": 0.00015098901489057011,
+      "loss": 1.494,
+      "step": 6568
+    },
+    {
+      "epoch": 1.05104,
+      "grad_norm": 2.1430837970805583,
+      "learning_rate": 0.00015097415449301865,
+      "loss": 1.3927,
+      "step": 6569
+    },
+    {
+      "epoch": 1.0512,
+      "grad_norm": 2.1896748137048716,
+      "learning_rate": 0.00015095929257447267,
+      "loss": 1.4903,
+      "step": 6570
+    },
+    {
+      "epoch": 1.05136,
+      "grad_norm": 2.061781067483645,
+      "learning_rate": 0.00015094442913537564,
+      "loss": 1.4553,
+      "step": 6571
+    },
+    {
+      "epoch": 1.05152,
+      "grad_norm": 1.9731238261899184,
+      "learning_rate": 0.0001509295641761711,
+      "loss": 1.415,
+      "step": 6572
+    },
+    {
+      "epoch": 1.05168,
+      "grad_norm": 2.5858447327777045,
+      "learning_rate": 0.00015091469769730256,
+      "loss": 1.5184,
+      "step": 6573
+    },
+    {
+      "epoch": 1.0518399999999999,
+      "grad_norm": 2.2455312499038573,
+      "learning_rate": 0.00015089982969921362,
+      "loss": 1.4899,
+      "step": 6574
+    },
+    {
+      "epoch": 1.052,
+      "grad_norm": 2.517177346391644,
+      "learning_rate": 0.00015088496018234797,
+      "loss": 1.4922,
+      "step": 6575
+    },
+    {
+      "epoch": 1.05216,
+      "grad_norm": 2.419642691533618,
+      "learning_rate": 0.00015087008914714924,
+      "loss": 1.4647,
+      "step": 6576
+    },
+    {
+      "epoch": 1.05232,
+      "grad_norm": 1.992894631426513,
+      "learning_rate": 0.00015085521659406116,
+      "loss": 1.5313,
+      "step": 6577
+    },
+    {
+      "epoch": 1.05248,
+      "grad_norm": 2.465842915421936,
+      "learning_rate": 0.00015084034252352754,
+      "loss": 1.5401,
+      "step": 6578
+    },
+    {
+      "epoch": 1.05264,
+      "grad_norm": 1.9982805587000347,
+      "learning_rate": 0.00015082546693599217,
+      "loss": 1.4433,
+      "step": 6579
+    },
+    {
+      "epoch": 1.0528,
+      "grad_norm": 1.9279984709786708,
+      "learning_rate": 0.00015081058983189895,
+      "loss": 1.4654,
+      "step": 6580
+    },
+    {
+      "epoch": 1.05296,
+      "grad_norm": 2.3551031908609654,
+      "learning_rate": 0.00015079571121169173,
+      "loss": 1.4929,
+      "step": 6581
+    },
+    {
+      "epoch": 1.05312,
+      "grad_norm": 2.3202053529842783,
+      "learning_rate": 0.00015078083107581452,
+      "loss": 1.4974,
+      "step": 6582
+    },
+    {
+      "epoch": 1.05328,
+      "grad_norm": 2.363043956564883,
+      "learning_rate": 0.00015076594942471132,
+      "loss": 1.4499,
+      "step": 6583
+    },
+    {
+      "epoch": 1.05344,
+      "grad_norm": 2.340387405759736,
+      "learning_rate": 0.00015075106625882616,
+      "loss": 1.4797,
+      "step": 6584
+    },
+    {
+      "epoch": 1.0536,
+      "grad_norm": 2.1356890110324693,
+      "learning_rate": 0.00015073618157860312,
+      "loss": 1.4181,
+      "step": 6585
+    },
+    {
+      "epoch": 1.05376,
+      "grad_norm": 2.2934008090348796,
+      "learning_rate": 0.00015072129538448636,
+      "loss": 1.4006,
+      "step": 6586
+    },
+    {
+      "epoch": 1.05392,
+      "grad_norm": 2.0271987877855278,
+      "learning_rate": 0.00015070640767692007,
+      "loss": 1.4652,
+      "step": 6587
+    },
+    {
+      "epoch": 1.05408,
+      "grad_norm": 2.2062092612493225,
+      "learning_rate": 0.00015069151845634843,
+      "loss": 1.4434,
+      "step": 6588
+    },
+    {
+      "epoch": 1.05424,
+      "grad_norm": 2.3057179200826914,
+      "learning_rate": 0.00015067662772321578,
+      "loss": 1.5286,
+      "step": 6589
+    },
+    {
+      "epoch": 1.0544,
+      "grad_norm": 2.7215709585537384,
+      "learning_rate": 0.00015066173547796636,
+      "loss": 1.5298,
+      "step": 6590
+    },
+    {
+      "epoch": 1.05456,
+      "grad_norm": 1.9887345252569648,
+      "learning_rate": 0.0001506468417210446,
+      "loss": 1.414,
+      "step": 6591
+    },
+    {
+      "epoch": 1.05472,
+      "grad_norm": 2.357388960703901,
+      "learning_rate": 0.0001506319464528949,
+      "loss": 1.4728,
+      "step": 6592
+    },
+    {
+      "epoch": 1.05488,
+      "grad_norm": 2.21683241053176,
+      "learning_rate": 0.00015061704967396167,
+      "loss": 1.4546,
+      "step": 6593
+    },
+    {
+      "epoch": 1.05504,
+      "grad_norm": 1.8850584688766747,
+      "learning_rate": 0.00015060215138468946,
+      "loss": 1.3567,
+      "step": 6594
+    },
+    {
+      "epoch": 1.0552,
+      "grad_norm": 2.1061553736400946,
+      "learning_rate": 0.00015058725158552275,
+      "loss": 1.4584,
+      "step": 6595
+    },
+    {
+      "epoch": 1.05536,
+      "grad_norm": 2.154711281502855,
+      "learning_rate": 0.00015057235027690622,
+      "loss": 1.4529,
+      "step": 6596
+    },
+    {
+      "epoch": 1.05552,
+      "grad_norm": 2.22966613495105,
+      "learning_rate": 0.0001505574474592844,
+      "loss": 1.4131,
+      "step": 6597
+    },
+    {
+      "epoch": 1.05568,
+      "grad_norm": 2.3141590532285217,
+      "learning_rate": 0.000150542543133102,
+      "loss": 1.5415,
+      "step": 6598
+    },
+    {
+      "epoch": 1.05584,
+      "grad_norm": 2.48066718777215,
+      "learning_rate": 0.0001505276372988038,
+      "loss": 1.4696,
+      "step": 6599
+    },
+    {
+      "epoch": 1.056,
+      "grad_norm": 2.544180788295923,
+      "learning_rate": 0.00015051272995683453,
+      "loss": 1.4646,
+      "step": 6600
+    },
+    {
+      "epoch": 1.05616,
+      "grad_norm": 2.711167217561504,
+      "learning_rate": 0.00015049782110763903,
+      "loss": 1.4257,
+      "step": 6601
+    },
+    {
+      "epoch": 1.05632,
+      "grad_norm": 2.907851962418562,
+      "learning_rate": 0.0001504829107516621,
+      "loss": 1.4359,
+      "step": 6602
+    },
+    {
+      "epoch": 1.05648,
+      "grad_norm": 2.0097984209500575,
+      "learning_rate": 0.00015046799888934867,
+      "loss": 1.4805,
+      "step": 6603
+    },
+    {
+      "epoch": 1.05664,
+      "grad_norm": 3.118433394588277,
+      "learning_rate": 0.00015045308552114368,
+      "loss": 1.55,
+      "step": 6604
+    },
+    {
+      "epoch": 1.0568,
+      "grad_norm": 2.1650864114334962,
+      "learning_rate": 0.00015043817064749218,
+      "loss": 1.5509,
+      "step": 6605
+    },
+    {
+      "epoch": 1.05696,
+      "grad_norm": 2.3360268298064066,
+      "learning_rate": 0.00015042325426883912,
+      "loss": 1.4171,
+      "step": 6606
+    },
+    {
+      "epoch": 1.05712,
+      "grad_norm": 2.319127408020637,
+      "learning_rate": 0.00015040833638562962,
+      "loss": 1.4536,
+      "step": 6607
+    },
+    {
+      "epoch": 1.05728,
+      "grad_norm": 1.9897832735660121,
+      "learning_rate": 0.00015039341699830887,
+      "loss": 1.4013,
+      "step": 6608
+    },
+    {
+      "epoch": 1.05744,
+      "grad_norm": 3.0562838691152727,
+      "learning_rate": 0.00015037849610732192,
+      "loss": 1.5453,
+      "step": 6609
+    },
+    {
+      "epoch": 1.0576,
+      "grad_norm": 2.552623992381644,
+      "learning_rate": 0.0001503635737131141,
+      "loss": 1.6202,
+      "step": 6610
+    },
+    {
+      "epoch": 1.05776,
+      "grad_norm": 2.3789566041242876,
+      "learning_rate": 0.00015034864981613064,
+      "loss": 1.5129,
+      "step": 6611
+    },
+    {
+      "epoch": 1.05792,
+      "grad_norm": 2.7219259948163277,
+      "learning_rate": 0.00015033372441681681,
+      "loss": 1.4249,
+      "step": 6612
+    },
+    {
+      "epoch": 1.05808,
+      "grad_norm": 3.034552901388941,
+      "learning_rate": 0.000150318797515618,
+      "loss": 1.5386,
+      "step": 6613
+    },
+    {
+      "epoch": 1.05824,
+      "grad_norm": 2.1576072821675614,
+      "learning_rate": 0.00015030386911297957,
+      "loss": 1.4159,
+      "step": 6614
+    },
+    {
+      "epoch": 1.0584,
+      "grad_norm": 2.5134055199181016,
+      "learning_rate": 0.00015028893920934698,
+      "loss": 1.5093,
+      "step": 6615
+    },
+    {
+      "epoch": 1.05856,
+      "grad_norm": 2.229889681406055,
+      "learning_rate": 0.0001502740078051658,
+      "loss": 1.4646,
+      "step": 6616
+    },
+    {
+      "epoch": 1.05872,
+      "grad_norm": 2.1037633994508145,
+      "learning_rate": 0.0001502590749008814,
+      "loss": 1.4456,
+      "step": 6617
+    },
+    {
+      "epoch": 1.05888,
+      "grad_norm": 2.751412835196577,
+      "learning_rate": 0.00015024414049693947,
+      "loss": 1.4857,
+      "step": 6618
+    },
+    {
+      "epoch": 1.05904,
+      "grad_norm": 2.3587926601469,
+      "learning_rate": 0.0001502292045937856,
+      "loss": 1.4142,
+      "step": 6619
+    },
+    {
+      "epoch": 1.0592,
+      "grad_norm": 2.282745228669717,
+      "learning_rate": 0.00015021426719186547,
+      "loss": 1.4322,
+      "step": 6620
+    },
+    {
+      "epoch": 1.05936,
+      "grad_norm": 2.540808493635697,
+      "learning_rate": 0.00015019932829162476,
+      "loss": 1.4556,
+      "step": 6621
+    },
+    {
+      "epoch": 1.05952,
+      "grad_norm": 2.2225031222259597,
+      "learning_rate": 0.00015018438789350923,
+      "loss": 1.4472,
+      "step": 6622
+    },
+    {
+      "epoch": 1.05968,
+      "grad_norm": 2.353072662225334,
+      "learning_rate": 0.0001501694459979647,
+      "loss": 1.484,
+      "step": 6623
+    },
+    {
+      "epoch": 1.05984,
+      "grad_norm": 2.393134287668441,
+      "learning_rate": 0.00015015450260543703,
+      "loss": 1.459,
+      "step": 6624
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 2.5175289688493208,
+      "learning_rate": 0.00015013955771637206,
+      "loss": 1.4356,
+      "step": 6625
+    },
+    {
+      "epoch": 1.06016,
+      "grad_norm": 2.4164534859755236,
+      "learning_rate": 0.00015012461133121575,
+      "loss": 1.3977,
+      "step": 6626
+    },
+    {
+      "epoch": 1.06032,
+      "grad_norm": 2.098518518826564,
+      "learning_rate": 0.0001501096634504141,
+      "loss": 1.5389,
+      "step": 6627
+    },
+    {
+      "epoch": 1.06048,
+      "grad_norm": 2.035526777102743,
+      "learning_rate": 0.0001500947140744131,
+      "loss": 1.3512,
+      "step": 6628
+    },
+    {
+      "epoch": 1.06064,
+      "grad_norm": 2.2578281296655995,
+      "learning_rate": 0.0001500797632036588,
+      "loss": 1.4314,
+      "step": 6629
+    },
+    {
+      "epoch": 1.0608,
+      "grad_norm": 1.9663489311163347,
+      "learning_rate": 0.00015006481083859737,
+      "loss": 1.3823,
+      "step": 6630
+    },
+    {
+      "epoch": 1.06096,
+      "grad_norm": 2.6759725855438505,
+      "learning_rate": 0.00015004985697967494,
+      "loss": 1.4599,
+      "step": 6631
+    },
+    {
+      "epoch": 1.06112,
+      "grad_norm": 2.7541564057579255,
+      "learning_rate": 0.00015003490162733768,
+      "loss": 1.4591,
+      "step": 6632
+    },
+    {
+      "epoch": 1.06128,
+      "grad_norm": 3.1889483801763268,
+      "learning_rate": 0.0001500199447820319,
+      "loss": 1.44,
+      "step": 6633
+    },
+    {
+      "epoch": 1.06144,
+      "grad_norm": 3.165867905727056,
+      "learning_rate": 0.00015000498644420382,
+      "loss": 1.4249,
+      "step": 6634
+    },
+    {
+      "epoch": 1.0616,
+      "grad_norm": 2.2738581808154397,
+      "learning_rate": 0.00014999002661429984,
+      "loss": 1.4279,
+      "step": 6635
+    },
+    {
+      "epoch": 1.06176,
+      "grad_norm": 3.2802340813053434,
+      "learning_rate": 0.00014997506529276628,
+      "loss": 1.5544,
+      "step": 6636
+    },
+    {
+      "epoch": 1.06192,
+      "grad_norm": 2.337303940828805,
+      "learning_rate": 0.0001499601024800496,
+      "loss": 1.5037,
+      "step": 6637
+    },
+    {
+      "epoch": 1.06208,
+      "grad_norm": 2.509687659927974,
+      "learning_rate": 0.0001499451381765963,
+      "loss": 1.4354,
+      "step": 6638
+    },
+    {
+      "epoch": 1.06224,
+      "grad_norm": 2.2615356698575537,
+      "learning_rate": 0.0001499301723828528,
+      "loss": 1.3744,
+      "step": 6639
+    },
+    {
+      "epoch": 1.0624,
+      "grad_norm": 1.8365658227999053,
+      "learning_rate": 0.00014991520509926573,
+      "loss": 1.3342,
+      "step": 6640
+    },
+    {
+      "epoch": 1.06256,
+      "grad_norm": 2.1543131065994654,
+      "learning_rate": 0.0001499002363262817,
+      "loss": 1.4141,
+      "step": 6641
+    },
+    {
+      "epoch": 1.06272,
+      "grad_norm": 2.763940242993138,
+      "learning_rate": 0.00014988526606434733,
+      "loss": 1.4817,
+      "step": 6642
+    },
+    {
+      "epoch": 1.06288,
+      "grad_norm": 1.867681969448509,
+      "learning_rate": 0.00014987029431390926,
+      "loss": 1.435,
+      "step": 6643
+    },
+    {
+      "epoch": 1.06304,
+      "grad_norm": 3.4928345811126085,
+      "learning_rate": 0.0001498553210754143,
+      "loss": 1.4697,
+      "step": 6644
+    },
+    {
+      "epoch": 1.0632,
+      "grad_norm": 2.206041112579542,
+      "learning_rate": 0.00014984034634930925,
+      "loss": 1.4983,
+      "step": 6645
+    },
+    {
+      "epoch": 1.06336,
+      "grad_norm": 2.130496376461168,
+      "learning_rate": 0.00014982537013604085,
+      "loss": 1.4381,
+      "step": 6646
+    },
+    {
+      "epoch": 1.06352,
+      "grad_norm": 2.700480007107328,
+      "learning_rate": 0.00014981039243605599,
+      "loss": 1.4371,
+      "step": 6647
+    },
+    {
+      "epoch": 1.06368,
+      "grad_norm": 2.1072389627659605,
+      "learning_rate": 0.00014979541324980163,
+      "loss": 1.3897,
+      "step": 6648
+    },
+    {
+      "epoch": 1.06384,
+      "grad_norm": 2.7251730426835157,
+      "learning_rate": 0.0001497804325777247,
+      "loss": 1.4497,
+      "step": 6649
+    },
+    {
+      "epoch": 1.064,
+      "grad_norm": 2.7118009950684563,
+      "learning_rate": 0.00014976545042027218,
+      "loss": 1.4564,
+      "step": 6650
+    },
+    {
+      "epoch": 1.06416,
+      "grad_norm": 2.058979370580706,
+      "learning_rate": 0.00014975046677789114,
+      "loss": 1.3907,
+      "step": 6651
+    },
+    {
+      "epoch": 1.06432,
+      "grad_norm": 2.8145400886738545,
+      "learning_rate": 0.00014973548165102867,
+      "loss": 1.4317,
+      "step": 6652
+    },
+    {
+      "epoch": 1.06448,
+      "grad_norm": 2.5157911076398713,
+      "learning_rate": 0.00014972049504013193,
+      "loss": 1.4207,
+      "step": 6653
+    },
+    {
+      "epoch": 1.06464,
+      "grad_norm": 3.1272073784912675,
+      "learning_rate": 0.00014970550694564803,
+      "loss": 1.4978,
+      "step": 6654
+    },
+    {
+      "epoch": 1.0648,
+      "grad_norm": 2.4706350148466583,
+      "learning_rate": 0.00014969051736802422,
+      "loss": 1.5508,
+      "step": 6655
+    },
+    {
+      "epoch": 1.06496,
+      "grad_norm": 2.085161696298772,
+      "learning_rate": 0.0001496755263077078,
+      "loss": 1.4977,
+      "step": 6656
+    },
+    {
+      "epoch": 1.06512,
+      "grad_norm": 2.5272203540704092,
+      "learning_rate": 0.00014966053376514607,
+      "loss": 1.4229,
+      "step": 6657
+    },
+    {
+      "epoch": 1.06528,
+      "grad_norm": 2.1723592315199003,
+      "learning_rate": 0.00014964553974078636,
+      "loss": 1.5204,
+      "step": 6658
+    },
+    {
+      "epoch": 1.06544,
+      "grad_norm": 2.136900558776639,
+      "learning_rate": 0.0001496305442350761,
+      "loss": 1.4485,
+      "step": 6659
+    },
+    {
+      "epoch": 1.0656,
+      "grad_norm": 3.3960352397296987,
+      "learning_rate": 0.00014961554724846272,
+      "loss": 1.4463,
+      "step": 6660
+    },
+    {
+      "epoch": 1.06576,
+      "grad_norm": 2.3782903097834307,
+      "learning_rate": 0.00014960054878139368,
+      "loss": 1.4925,
+      "step": 6661
+    },
+    {
+      "epoch": 1.06592,
+      "grad_norm": 2.379535192684025,
+      "learning_rate": 0.00014958554883431657,
+      "loss": 1.4217,
+      "step": 6662
+    },
+    {
+      "epoch": 1.06608,
+      "grad_norm": 2.0248779234274275,
+      "learning_rate": 0.0001495705474076789,
+      "loss": 1.3774,
+      "step": 6663
+    },
+    {
+      "epoch": 1.06624,
+      "grad_norm": 2.3731830488927574,
+      "learning_rate": 0.00014955554450192836,
+      "loss": 1.4838,
+      "step": 6664
+    },
+    {
+      "epoch": 1.0664,
+      "grad_norm": 2.29930101371983,
+      "learning_rate": 0.00014954054011751257,
+      "loss": 1.4343,
+      "step": 6665
+    },
+    {
+      "epoch": 1.06656,
+      "grad_norm": 2.3288340462894057,
+      "learning_rate": 0.00014952553425487928,
+      "loss": 1.4833,
+      "step": 6666
+    },
+    {
+      "epoch": 1.0667200000000001,
+      "grad_norm": 2.2999385721117744,
+      "learning_rate": 0.00014951052691447619,
+      "loss": 1.3856,
+      "step": 6667
+    },
+    {
+      "epoch": 1.06688,
+      "grad_norm": 2.9569022313176383,
+      "learning_rate": 0.0001494955180967511,
+      "loss": 1.4607,
+      "step": 6668
+    },
+    {
+      "epoch": 1.06704,
+      "grad_norm": 2.754838414003039,
+      "learning_rate": 0.0001494805078021519,
+      "loss": 1.4612,
+      "step": 6669
+    },
+    {
+      "epoch": 1.0672,
+      "grad_norm": 2.176092537956234,
+      "learning_rate": 0.00014946549603112646,
+      "loss": 1.4458,
+      "step": 6670
+    },
+    {
+      "epoch": 1.06736,
+      "grad_norm": 2.0997102576792943,
+      "learning_rate": 0.00014945048278412263,
+      "loss": 1.4797,
+      "step": 6671
+    },
+    {
+      "epoch": 1.06752,
+      "grad_norm": 2.3554364946156627,
+      "learning_rate": 0.00014943546806158852,
+      "loss": 1.4858,
+      "step": 6672
+    },
+    {
+      "epoch": 1.06768,
+      "grad_norm": 2.277341300419031,
+      "learning_rate": 0.00014942045186397204,
+      "loss": 1.3838,
+      "step": 6673
+    },
+    {
+      "epoch": 1.06784,
+      "grad_norm": 2.1199110321292856,
+      "learning_rate": 0.00014940543419172134,
+      "loss": 1.441,
+      "step": 6674
+    },
+    {
+      "epoch": 1.068,
+      "grad_norm": 2.5963256597804767,
+      "learning_rate": 0.00014939041504528444,
+      "loss": 1.4695,
+      "step": 6675
+    },
+    {
+      "epoch": 1.06816,
+      "grad_norm": 2.0527520899230565,
+      "learning_rate": 0.0001493753944251095,
+      "loss": 1.5186,
+      "step": 6676
+    },
+    {
+      "epoch": 1.06832,
+      "grad_norm": 2.0668255996944724,
+      "learning_rate": 0.00014936037233164478,
+      "loss": 1.4655,
+      "step": 6677
+    },
+    {
+      "epoch": 1.06848,
+      "grad_norm": 2.00606602307537,
+      "learning_rate": 0.00014934534876533847,
+      "loss": 1.4307,
+      "step": 6678
+    },
+    {
+      "epoch": 1.06864,
+      "grad_norm": 2.037406321370364,
+      "learning_rate": 0.00014933032372663882,
+      "loss": 1.5091,
+      "step": 6679
+    },
+    {
+      "epoch": 1.0688,
+      "grad_norm": 2.380078397386266,
+      "learning_rate": 0.0001493152972159942,
+      "loss": 1.4538,
+      "step": 6680
+    },
+    {
+      "epoch": 1.06896,
+      "grad_norm": 2.4056293442129393,
+      "learning_rate": 0.000149300269233853,
+      "loss": 1.4497,
+      "step": 6681
+    },
+    {
+      "epoch": 1.06912,
+      "grad_norm": 2.067790122344086,
+      "learning_rate": 0.00014928523978066362,
+      "loss": 1.4518,
+      "step": 6682
+    },
+    {
+      "epoch": 1.06928,
+      "grad_norm": 2.2378535855666266,
+      "learning_rate": 0.00014927020885687446,
+      "loss": 1.5162,
+      "step": 6683
+    },
+    {
+      "epoch": 1.06944,
+      "grad_norm": 2.144168724930481,
+      "learning_rate": 0.00014925517646293405,
+      "loss": 1.5097,
+      "step": 6684
+    },
+    {
+      "epoch": 1.0695999999999999,
+      "grad_norm": 1.8515826479107587,
+      "learning_rate": 0.000149240142599291,
+      "loss": 1.4591,
+      "step": 6685
+    },
+    {
+      "epoch": 1.06976,
+      "grad_norm": 2.588539819025708,
+      "learning_rate": 0.0001492251072663938,
+      "loss": 1.4418,
+      "step": 6686
+    },
+    {
+      "epoch": 1.06992,
+      "grad_norm": 2.1442303578086217,
+      "learning_rate": 0.00014921007046469115,
+      "loss": 1.5219,
+      "step": 6687
+    },
+    {
+      "epoch": 1.07008,
+      "grad_norm": 2.386006133142145,
+      "learning_rate": 0.00014919503219463172,
+      "loss": 1.4943,
+      "step": 6688
+    },
+    {
+      "epoch": 1.07024,
+      "grad_norm": 2.402459467522887,
+      "learning_rate": 0.00014917999245666418,
+      "loss": 1.4266,
+      "step": 6689
+    },
+    {
+      "epoch": 1.0704,
+      "grad_norm": 2.207083312915663,
+      "learning_rate": 0.00014916495125123734,
+      "loss": 1.4245,
+      "step": 6690
+    },
+    {
+      "epoch": 1.07056,
+      "grad_norm": 2.3040021103014525,
+      "learning_rate": 0.0001491499085788,
+      "loss": 1.3629,
+      "step": 6691
+    },
+    {
+      "epoch": 1.0707200000000001,
+      "grad_norm": 2.1615354077695894,
+      "learning_rate": 0.000149134864439801,
+      "loss": 1.4271,
+      "step": 6692
+    },
+    {
+      "epoch": 1.07088,
+      "grad_norm": 1.8654436208159488,
+      "learning_rate": 0.00014911981883468923,
+      "loss": 1.4886,
+      "step": 6693
+    },
+    {
+      "epoch": 1.07104,
+      "grad_norm": 2.2645860175582517,
+      "learning_rate": 0.00014910477176391365,
+      "loss": 1.4643,
+      "step": 6694
+    },
+    {
+      "epoch": 1.0712,
+      "grad_norm": 2.1140867185040797,
+      "learning_rate": 0.00014908972322792328,
+      "loss": 1.4335,
+      "step": 6695
+    },
+    {
+      "epoch": 1.07136,
+      "grad_norm": 2.5531385841729235,
+      "learning_rate": 0.00014907467322716706,
+      "loss": 1.5088,
+      "step": 6696
+    },
+    {
+      "epoch": 1.07152,
+      "grad_norm": 1.9147594884059653,
+      "learning_rate": 0.00014905962176209406,
+      "loss": 1.4184,
+      "step": 6697
+    },
+    {
+      "epoch": 1.07168,
+      "grad_norm": 2.077196560575393,
+      "learning_rate": 0.0001490445688331535,
+      "loss": 1.5184,
+      "step": 6698
+    },
+    {
+      "epoch": 1.07184,
+      "grad_norm": 3.0017541912028656,
+      "learning_rate": 0.00014902951444079445,
+      "loss": 1.63,
+      "step": 6699
+    },
+    {
+      "epoch": 1.072,
+      "grad_norm": 1.8945071852706596,
+      "learning_rate": 0.00014901445858546612,
+      "loss": 1.4644,
+      "step": 6700
+    },
+    {
+      "epoch": 1.07216,
+      "grad_norm": 2.460593879234495,
+      "learning_rate": 0.00014899940126761777,
+      "loss": 1.4449,
+      "step": 6701
+    },
+    {
+      "epoch": 1.07232,
+      "grad_norm": 2.2548728849554176,
+      "learning_rate": 0.0001489843424876987,
+      "loss": 1.4815,
+      "step": 6702
+    },
+    {
+      "epoch": 1.07248,
+      "grad_norm": 2.2966650279729137,
+      "learning_rate": 0.00014896928224615824,
+      "loss": 1.4413,
+      "step": 6703
+    },
+    {
+      "epoch": 1.07264,
+      "grad_norm": 2.2288820913309344,
+      "learning_rate": 0.00014895422054344573,
+      "loss": 1.502,
+      "step": 6704
+    },
+    {
+      "epoch": 1.0728,
+      "grad_norm": 2.790851692188078,
+      "learning_rate": 0.0001489391573800106,
+      "loss": 1.5543,
+      "step": 6705
+    },
+    {
+      "epoch": 1.07296,
+      "grad_norm": 2.3923046415321982,
+      "learning_rate": 0.00014892409275630235,
+      "loss": 1.5205,
+      "step": 6706
+    },
+    {
+      "epoch": 1.07312,
+      "grad_norm": 3.060978671766119,
+      "learning_rate": 0.00014890902667277046,
+      "loss": 1.5636,
+      "step": 6707
+    },
+    {
+      "epoch": 1.07328,
+      "grad_norm": 2.7731958836009096,
+      "learning_rate": 0.0001488939591298645,
+      "loss": 1.4986,
+      "step": 6708
+    },
+    {
+      "epoch": 1.07344,
+      "grad_norm": 3.259817737720769,
+      "learning_rate": 0.000148878890128034,
+      "loss": 1.4641,
+      "step": 6709
+    },
+    {
+      "epoch": 1.0735999999999999,
+      "grad_norm": 2.588929556005782,
+      "learning_rate": 0.00014886381966772866,
+      "loss": 1.5707,
+      "step": 6710
+    },
+    {
+      "epoch": 1.07376,
+      "grad_norm": 2.3756292596940445,
+      "learning_rate": 0.00014884874774939816,
+      "loss": 1.4074,
+      "step": 6711
+    },
+    {
+      "epoch": 1.07392,
+      "grad_norm": 3.6815040022002306,
+      "learning_rate": 0.0001488336743734922,
+      "loss": 1.4612,
+      "step": 6712
+    },
+    {
+      "epoch": 1.07408,
+      "grad_norm": 2.360332101515584,
+      "learning_rate": 0.00014881859954046054,
+      "loss": 1.3895,
+      "step": 6713
+    },
+    {
+      "epoch": 1.07424,
+      "grad_norm": 4.080541009899241,
+      "learning_rate": 0.000148803523250753,
+      "loss": 1.4645,
+      "step": 6714
+    },
+    {
+      "epoch": 1.0744,
+      "grad_norm": 2.347270487020927,
+      "learning_rate": 0.00014878844550481946,
+      "loss": 1.4593,
+      "step": 6715
+    },
+    {
+      "epoch": 1.07456,
+      "grad_norm": 2.7722387484727427,
+      "learning_rate": 0.0001487733663031098,
+      "loss": 1.4906,
+      "step": 6716
+    },
+    {
+      "epoch": 1.07472,
+      "grad_norm": 3.81633277462574,
+      "learning_rate": 0.00014875828564607396,
+      "loss": 1.4244,
+      "step": 6717
+    },
+    {
+      "epoch": 1.07488,
+      "grad_norm": 2.8117784499559964,
+      "learning_rate": 0.00014874320353416195,
+      "loss": 1.5034,
+      "step": 6718
+    },
+    {
+      "epoch": 1.07504,
+      "grad_norm": 3.054403626663226,
+      "learning_rate": 0.00014872811996782372,
+      "loss": 1.4976,
+      "step": 6719
+    },
+    {
+      "epoch": 1.0752,
+      "grad_norm": 2.6327587607129974,
+      "learning_rate": 0.00014871303494750943,
+      "loss": 1.5046,
+      "step": 6720
+    },
+    {
+      "epoch": 1.07536,
+      "grad_norm": 3.019740283386349,
+      "learning_rate": 0.00014869794847366911,
+      "loss": 1.4274,
+      "step": 6721
+    },
+    {
+      "epoch": 1.07552,
+      "grad_norm": 2.8820943470124636,
+      "learning_rate": 0.00014868286054675302,
+      "loss": 1.3642,
+      "step": 6722
+    },
+    {
+      "epoch": 1.07568,
+      "grad_norm": 2.20516294694183,
+      "learning_rate": 0.00014866777116721128,
+      "loss": 1.4979,
+      "step": 6723
+    },
+    {
+      "epoch": 1.07584,
+      "grad_norm": 3.4157738457933946,
+      "learning_rate": 0.0001486526803354942,
+      "loss": 1.5078,
+      "step": 6724
+    },
+    {
+      "epoch": 1.076,
+      "grad_norm": 2.144281005220912,
+      "learning_rate": 0.000148637588052052,
+      "loss": 1.4795,
+      "step": 6725
+    },
+    {
+      "epoch": 1.07616,
+      "grad_norm": 3.3614879723020032,
+      "learning_rate": 0.00014862249431733509,
+      "loss": 1.4874,
+      "step": 6726
+    },
+    {
+      "epoch": 1.07632,
+      "grad_norm": 2.9663242002162975,
+      "learning_rate": 0.00014860739913179376,
+      "loss": 1.5534,
+      "step": 6727
+    },
+    {
+      "epoch": 1.07648,
+      "grad_norm": 2.2318201619376303,
+      "learning_rate": 0.00014859230249587846,
+      "loss": 1.4,
+      "step": 6728
+    },
+    {
+      "epoch": 1.07664,
+      "grad_norm": 2.8736895981562225,
+      "learning_rate": 0.00014857720441003967,
+      "loss": 1.52,
+      "step": 6729
+    },
+    {
+      "epoch": 1.0768,
+      "grad_norm": 3.1695680885247723,
+      "learning_rate": 0.0001485621048747279,
+      "loss": 1.4041,
+      "step": 6730
+    },
+    {
+      "epoch": 1.07696,
+      "grad_norm": 3.783888731553723,
+      "learning_rate": 0.00014854700389039366,
+      "loss": 1.533,
+      "step": 6731
+    },
+    {
+      "epoch": 1.07712,
+      "grad_norm": 2.2866354512076104,
+      "learning_rate": 0.0001485319014574876,
+      "loss": 1.4664,
+      "step": 6732
+    },
+    {
+      "epoch": 1.07728,
+      "grad_norm": 2.308364635092833,
+      "learning_rate": 0.0001485167975764603,
+      "loss": 1.4231,
+      "step": 6733
+    },
+    {
+      "epoch": 1.07744,
+      "grad_norm": 2.4116276360598947,
+      "learning_rate": 0.00014850169224776244,
+      "loss": 1.4307,
+      "step": 6734
+    },
+    {
+      "epoch": 1.0776,
+      "grad_norm": 2.050937303221302,
+      "learning_rate": 0.00014848658547184478,
+      "loss": 1.4947,
+      "step": 6735
+    },
+    {
+      "epoch": 1.07776,
+      "grad_norm": 1.9692112888843665,
+      "learning_rate": 0.00014847147724915803,
+      "loss": 1.4566,
+      "step": 6736
+    },
+    {
+      "epoch": 1.07792,
+      "grad_norm": 2.1626328114943187,
+      "learning_rate": 0.00014845636758015307,
+      "loss": 1.4018,
+      "step": 6737
+    },
+    {
+      "epoch": 1.07808,
+      "grad_norm": 2.2571798424776213,
+      "learning_rate": 0.00014844125646528068,
+      "loss": 1.3992,
+      "step": 6738
+    },
+    {
+      "epoch": 1.07824,
+      "grad_norm": 2.26271326798071,
+      "learning_rate": 0.00014842614390499182,
+      "loss": 1.4075,
+      "step": 6739
+    },
+    {
+      "epoch": 1.0784,
+      "grad_norm": 2.526226920986057,
+      "learning_rate": 0.00014841102989973738,
+      "loss": 1.4738,
+      "step": 6740
+    },
+    {
+      "epoch": 1.07856,
+      "grad_norm": 2.262147992921642,
+      "learning_rate": 0.0001483959144499683,
+      "loss": 1.4669,
+      "step": 6741
+    },
+    {
+      "epoch": 1.07872,
+      "grad_norm": 2.140278129225632,
+      "learning_rate": 0.0001483807975561357,
+      "loss": 1.4915,
+      "step": 6742
+    },
+    {
+      "epoch": 1.07888,
+      "grad_norm": 2.227207704335657,
+      "learning_rate": 0.0001483656792186906,
+      "loss": 1.354,
+      "step": 6743
+    },
+    {
+      "epoch": 1.07904,
+      "grad_norm": 2.141423876631719,
+      "learning_rate": 0.00014835055943808407,
+      "loss": 1.4768,
+      "step": 6744
+    },
+    {
+      "epoch": 1.0792,
+      "grad_norm": 2.57731080151344,
+      "learning_rate": 0.00014833543821476735,
+      "loss": 1.4826,
+      "step": 6745
+    },
+    {
+      "epoch": 1.07936,
+      "grad_norm": 2.162947197864746,
+      "learning_rate": 0.00014832031554919156,
+      "loss": 1.4542,
+      "step": 6746
+    },
+    {
+      "epoch": 1.07952,
+      "grad_norm": 2.799099343606009,
+      "learning_rate": 0.00014830519144180793,
+      "loss": 1.5037,
+      "step": 6747
+    },
+    {
+      "epoch": 1.07968,
+      "grad_norm": 1.8996735348316882,
+      "learning_rate": 0.00014829006589306784,
+      "loss": 1.421,
+      "step": 6748
+    },
+    {
+      "epoch": 1.07984,
+      "grad_norm": 2.01765632599734,
+      "learning_rate": 0.00014827493890342254,
+      "loss": 1.4419,
+      "step": 6749
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 2.230411570719251,
+      "learning_rate": 0.0001482598104733234,
+      "loss": 1.4451,
+      "step": 6750
+    },
+    {
+      "epoch": 1.08016,
+      "grad_norm": 2.0781086028039693,
+      "learning_rate": 0.0001482446806032218,
+      "loss": 1.3741,
+      "step": 6751
+    },
+    {
+      "epoch": 1.08032,
+      "grad_norm": 2.0879917737160207,
+      "learning_rate": 0.00014822954929356928,
+      "loss": 1.525,
+      "step": 6752
+    },
+    {
+      "epoch": 1.08048,
+      "grad_norm": 2.488927661582233,
+      "learning_rate": 0.00014821441654481725,
+      "loss": 1.4952,
+      "step": 6753
+    },
+    {
+      "epoch": 1.08064,
+      "grad_norm": 2.507375419859088,
+      "learning_rate": 0.0001481992823574173,
+      "loss": 1.4761,
+      "step": 6754
+    },
+    {
+      "epoch": 1.0808,
+      "grad_norm": 1.8363531227173777,
+      "learning_rate": 0.000148184146731821,
+      "loss": 1.392,
+      "step": 6755
+    },
+    {
+      "epoch": 1.08096,
+      "grad_norm": 2.4146436366788167,
+      "learning_rate": 0.00014816900966847996,
+      "loss": 1.4571,
+      "step": 6756
+    },
+    {
+      "epoch": 1.08112,
+      "grad_norm": 1.9498316234795492,
+      "learning_rate": 0.00014815387116784589,
+      "loss": 1.4232,
+      "step": 6757
+    },
+    {
+      "epoch": 1.08128,
+      "grad_norm": 2.279537491153731,
+      "learning_rate": 0.00014813873123037044,
+      "loss": 1.4789,
+      "step": 6758
+    },
+    {
+      "epoch": 1.08144,
+      "grad_norm": 2.1912337098071877,
+      "learning_rate": 0.00014812358985650544,
+      "loss": 1.5332,
+      "step": 6759
+    },
+    {
+      "epoch": 1.0816,
+      "grad_norm": 2.1631091574074888,
+      "learning_rate": 0.0001481084470467026,
+      "loss": 1.3839,
+      "step": 6760
+    },
+    {
+      "epoch": 1.08176,
+      "grad_norm": 2.1220514404828754,
+      "learning_rate": 0.00014809330280141382,
+      "loss": 1.4603,
+      "step": 6761
+    },
+    {
+      "epoch": 1.08192,
+      "grad_norm": 2.862200718944194,
+      "learning_rate": 0.00014807815712109093,
+      "loss": 1.4391,
+      "step": 6762
+    },
+    {
+      "epoch": 1.08208,
+      "grad_norm": 2.0947733989033304,
+      "learning_rate": 0.00014806301000618594,
+      "loss": 1.4332,
+      "step": 6763
+    },
+    {
+      "epoch": 1.08224,
+      "grad_norm": 3.1820231229204525,
+      "learning_rate": 0.00014804786145715073,
+      "loss": 1.4085,
+      "step": 6764
+    },
+    {
+      "epoch": 1.0824,
+      "grad_norm": 2.3587598615531755,
+      "learning_rate": 0.00014803271147443737,
+      "loss": 1.4274,
+      "step": 6765
+    },
+    {
+      "epoch": 1.08256,
+      "grad_norm": 3.3692105588358188,
+      "learning_rate": 0.00014801756005849789,
+      "loss": 1.4814,
+      "step": 6766
+    },
+    {
+      "epoch": 1.08272,
+      "grad_norm": 2.15938961721387,
+      "learning_rate": 0.0001480024072097844,
+      "loss": 1.467,
+      "step": 6767
+    },
+    {
+      "epoch": 1.08288,
+      "grad_norm": 2.010537496887891,
+      "learning_rate": 0.00014798725292874897,
+      "loss": 1.4581,
+      "step": 6768
+    },
+    {
+      "epoch": 1.08304,
+      "grad_norm": 3.1803509364395794,
+      "learning_rate": 0.0001479720972158439,
+      "loss": 1.4634,
+      "step": 6769
+    },
+    {
+      "epoch": 1.0832,
+      "grad_norm": 1.861497522940795,
+      "learning_rate": 0.0001479569400715213,
+      "loss": 1.4404,
+      "step": 6770
+    },
+    {
+      "epoch": 1.08336,
+      "grad_norm": 4.020156415746338,
+      "learning_rate": 0.00014794178149623352,
+      "loss": 1.4894,
+      "step": 6771
+    },
+    {
+      "epoch": 1.08352,
+      "grad_norm": 2.8305661163392446,
+      "learning_rate": 0.00014792662149043284,
+      "loss": 1.4353,
+      "step": 6772
+    },
+    {
+      "epoch": 1.08368,
+      "grad_norm": 2.7719148440868837,
+      "learning_rate": 0.0001479114600545716,
+      "loss": 1.3761,
+      "step": 6773
+    },
+    {
+      "epoch": 1.08384,
+      "grad_norm": 3.7451053681989075,
+      "learning_rate": 0.0001478962971891022,
+      "loss": 1.5135,
+      "step": 6774
+    },
+    {
+      "epoch": 1.084,
+      "grad_norm": 2.235864935823904,
+      "learning_rate": 0.0001478811328944771,
+      "loss": 1.4925,
+      "step": 6775
+    },
+    {
+      "epoch": 1.08416,
+      "grad_norm": 2.246653434614251,
+      "learning_rate": 0.0001478659671711488,
+      "loss": 1.4346,
+      "step": 6776
+    },
+    {
+      "epoch": 1.08432,
+      "grad_norm": 2.9534733509817355,
+      "learning_rate": 0.0001478508000195697,
+      "loss": 1.4967,
+      "step": 6777
+    },
+    {
+      "epoch": 1.08448,
+      "grad_norm": 2.1173640532121567,
+      "learning_rate": 0.0001478356314401925,
+      "loss": 1.3933,
+      "step": 6778
+    },
+    {
+      "epoch": 1.08464,
+      "grad_norm": 2.5799868322273416,
+      "learning_rate": 0.00014782046143346978,
+      "loss": 1.455,
+      "step": 6779
+    },
+    {
+      "epoch": 1.0848,
+      "grad_norm": 2.4034670006368293,
+      "learning_rate": 0.00014780528999985414,
+      "loss": 1.4868,
+      "step": 6780
+    },
+    {
+      "epoch": 1.08496,
+      "grad_norm": 2.021376129433465,
+      "learning_rate": 0.00014779011713979832,
+      "loss": 1.3722,
+      "step": 6781
+    },
+    {
+      "epoch": 1.08512,
+      "grad_norm": 2.62003366411055,
+      "learning_rate": 0.00014777494285375504,
+      "loss": 1.4408,
+      "step": 6782
+    },
+    {
+      "epoch": 1.08528,
+      "grad_norm": 2.8889168369530123,
+      "learning_rate": 0.0001477597671421771,
+      "loss": 1.4832,
+      "step": 6783
+    },
+    {
+      "epoch": 1.08544,
+      "grad_norm": 2.148392500944601,
+      "learning_rate": 0.0001477445900055173,
+      "loss": 1.5201,
+      "step": 6784
+    },
+    {
+      "epoch": 1.0856,
+      "grad_norm": 3.138707855654998,
+      "learning_rate": 0.00014772941144422847,
+      "loss": 1.4759,
+      "step": 6785
+    },
+    {
+      "epoch": 1.08576,
+      "grad_norm": 2.221938790150311,
+      "learning_rate": 0.00014771423145876355,
+      "loss": 1.3715,
+      "step": 6786
+    },
+    {
+      "epoch": 1.08592,
+      "grad_norm": 2.1032172563802427,
+      "learning_rate": 0.0001476990500495755,
+      "loss": 1.3705,
+      "step": 6787
+    },
+    {
+      "epoch": 1.08608,
+      "grad_norm": 2.0513416282635673,
+      "learning_rate": 0.00014768386721711733,
+      "loss": 1.52,
+      "step": 6788
+    },
+    {
+      "epoch": 1.08624,
+      "grad_norm": 2.121061747065683,
+      "learning_rate": 0.00014766868296184202,
+      "loss": 1.4403,
+      "step": 6789
+    },
+    {
+      "epoch": 1.0864,
+      "grad_norm": 3.033392432348546,
+      "learning_rate": 0.0001476534972842027,
+      "loss": 1.5484,
+      "step": 6790
+    },
+    {
+      "epoch": 1.08656,
+      "grad_norm": 2.444128580600597,
+      "learning_rate": 0.00014763831018465244,
+      "loss": 1.4162,
+      "step": 6791
+    },
+    {
+      "epoch": 1.08672,
+      "grad_norm": 2.154283578070564,
+      "learning_rate": 0.00014762312166364443,
+      "loss": 1.4397,
+      "step": 6792
+    },
+    {
+      "epoch": 1.08688,
+      "grad_norm": 2.144864599994824,
+      "learning_rate": 0.00014760793172163185,
+      "loss": 1.4068,
+      "step": 6793
+    },
+    {
+      "epoch": 1.08704,
+      "grad_norm": 2.2466954646790422,
+      "learning_rate": 0.00014759274035906794,
+      "loss": 1.4761,
+      "step": 6794
+    },
+    {
+      "epoch": 1.0872,
+      "grad_norm": 2.2275071113329803,
+      "learning_rate": 0.00014757754757640604,
+      "loss": 1.4395,
+      "step": 6795
+    },
+    {
+      "epoch": 1.0873599999999999,
+      "grad_norm": 1.8361758644653878,
+      "learning_rate": 0.00014756235337409944,
+      "loss": 1.3498,
+      "step": 6796
+    },
+    {
+      "epoch": 1.08752,
+      "grad_norm": 1.819116438536979,
+      "learning_rate": 0.0001475471577526015,
+      "loss": 1.3489,
+      "step": 6797
+    },
+    {
+      "epoch": 1.08768,
+      "grad_norm": 2.3524991807658835,
+      "learning_rate": 0.0001475319607123657,
+      "loss": 1.3398,
+      "step": 6798
+    },
+    {
+      "epoch": 1.08784,
+      "grad_norm": 2.214849118307639,
+      "learning_rate": 0.0001475167622538454,
+      "loss": 1.4804,
+      "step": 6799
+    },
+    {
+      "epoch": 1.088,
+      "grad_norm": 2.178003730856199,
+      "learning_rate": 0.0001475015623774942,
+      "loss": 1.4477,
+      "step": 6800
+    },
+    {
+      "epoch": 1.08816,
+      "grad_norm": 2.0551679813252175,
+      "learning_rate": 0.00014748636108376557,
+      "loss": 1.4767,
+      "step": 6801
+    },
+    {
+      "epoch": 1.08832,
+      "grad_norm": 2.494020044234289,
+      "learning_rate": 0.0001474711583731131,
+      "loss": 1.5623,
+      "step": 6802
+    },
+    {
+      "epoch": 1.0884800000000001,
+      "grad_norm": 2.346790061011265,
+      "learning_rate": 0.00014745595424599047,
+      "loss": 1.474,
+      "step": 6803
+    },
+    {
+      "epoch": 1.08864,
+      "grad_norm": 2.1347843541164524,
+      "learning_rate": 0.00014744074870285133,
+      "loss": 1.4656,
+      "step": 6804
+    },
+    {
+      "epoch": 1.0888,
+      "grad_norm": 2.413141008427547,
+      "learning_rate": 0.00014742554174414934,
+      "loss": 1.4658,
+      "step": 6805
+    },
+    {
+      "epoch": 1.08896,
+      "grad_norm": 2.148717819527095,
+      "learning_rate": 0.0001474103333703383,
+      "loss": 1.5061,
+      "step": 6806
+    },
+    {
+      "epoch": 1.08912,
+      "grad_norm": 2.311497432252624,
+      "learning_rate": 0.00014739512358187204,
+      "loss": 1.5208,
+      "step": 6807
+    },
+    {
+      "epoch": 1.08928,
+      "grad_norm": 2.3167524820050502,
+      "learning_rate": 0.00014737991237920433,
+      "loss": 1.4783,
+      "step": 6808
+    },
+    {
+      "epoch": 1.08944,
+      "grad_norm": 2.192729238076999,
+      "learning_rate": 0.00014736469976278906,
+      "loss": 1.5432,
+      "step": 6809
+    },
+    {
+      "epoch": 1.0896,
+      "grad_norm": 2.0321398789496823,
+      "learning_rate": 0.00014734948573308016,
+      "loss": 1.3949,
+      "step": 6810
+    },
+    {
+      "epoch": 1.08976,
+      "grad_norm": 2.248116592529448,
+      "learning_rate": 0.0001473342702905316,
+      "loss": 1.5223,
+      "step": 6811
+    },
+    {
+      "epoch": 1.08992,
+      "grad_norm": 2.036559500885551,
+      "learning_rate": 0.0001473190534355974,
+      "loss": 1.5143,
+      "step": 6812
+    },
+    {
+      "epoch": 1.09008,
+      "grad_norm": 2.345891400663143,
+      "learning_rate": 0.0001473038351687316,
+      "loss": 1.5056,
+      "step": 6813
+    },
+    {
+      "epoch": 1.09024,
+      "grad_norm": 2.4513986731764326,
+      "learning_rate": 0.00014728861549038831,
+      "loss": 1.4758,
+      "step": 6814
+    },
+    {
+      "epoch": 1.0904,
+      "grad_norm": 2.507957423276148,
+      "learning_rate": 0.0001472733944010216,
+      "loss": 1.4725,
+      "step": 6815
+    },
+    {
+      "epoch": 1.09056,
+      "grad_norm": 2.195935186737089,
+      "learning_rate": 0.0001472581719010857,
+      "loss": 1.425,
+      "step": 6816
+    },
+    {
+      "epoch": 1.09072,
+      "grad_norm": 2.2361447190107184,
+      "learning_rate": 0.00014724294799103483,
+      "loss": 1.4241,
+      "step": 6817
+    },
+    {
+      "epoch": 1.09088,
+      "grad_norm": 2.7038035963806206,
+      "learning_rate": 0.00014722772267132322,
+      "loss": 1.3838,
+      "step": 6818
+    },
+    {
+      "epoch": 1.09104,
+      "grad_norm": 2.404915735177806,
+      "learning_rate": 0.00014721249594240514,
+      "loss": 1.4143,
+      "step": 6819
+    },
+    {
+      "epoch": 1.0912,
+      "grad_norm": 2.2077969354587,
+      "learning_rate": 0.00014719726780473505,
+      "loss": 1.5179,
+      "step": 6820
+    },
+    {
+      "epoch": 1.0913599999999999,
+      "grad_norm": 1.8145157110753012,
+      "learning_rate": 0.0001471820382587672,
+      "loss": 1.5031,
+      "step": 6821
+    },
+    {
+      "epoch": 1.09152,
+      "grad_norm": 2.0948486403564246,
+      "learning_rate": 0.0001471668073049561,
+      "loss": 1.4523,
+      "step": 6822
+    },
+    {
+      "epoch": 1.09168,
+      "grad_norm": 2.1775693807475056,
+      "learning_rate": 0.00014715157494375625,
+      "loss": 1.4641,
+      "step": 6823
+    },
+    {
+      "epoch": 1.09184,
+      "grad_norm": 2.118333461167682,
+      "learning_rate": 0.00014713634117562204,
+      "loss": 1.5065,
+      "step": 6824
+    },
+    {
+      "epoch": 1.092,
+      "grad_norm": 2.176594868119968,
+      "learning_rate": 0.00014712110600100813,
+      "loss": 1.4945,
+      "step": 6825
+    },
+    {
+      "epoch": 1.09216,
+      "grad_norm": 2.540799971741811,
+      "learning_rate": 0.00014710586942036908,
+      "loss": 1.3888,
+      "step": 6826
+    },
+    {
+      "epoch": 1.09232,
+      "grad_norm": 2.531218164262258,
+      "learning_rate": 0.00014709063143415952,
+      "loss": 1.455,
+      "step": 6827
+    },
+    {
+      "epoch": 1.0924800000000001,
+      "grad_norm": 2.926978022097877,
+      "learning_rate": 0.0001470753920428341,
+      "loss": 1.5282,
+      "step": 6828
+    },
+    {
+      "epoch": 1.09264,
+      "grad_norm": 2.5273049026275523,
+      "learning_rate": 0.00014706015124684764,
+      "loss": 1.4995,
+      "step": 6829
+    },
+    {
+      "epoch": 1.0928,
+      "grad_norm": 2.2336464026694913,
+      "learning_rate": 0.00014704490904665483,
+      "loss": 1.5033,
+      "step": 6830
+    },
+    {
+      "epoch": 1.09296,
+      "grad_norm": 2.250115979332348,
+      "learning_rate": 0.00014702966544271047,
+      "loss": 1.4374,
+      "step": 6831
+    },
+    {
+      "epoch": 1.09312,
+      "grad_norm": 2.1999433296730806,
+      "learning_rate": 0.0001470144204354694,
+      "loss": 1.5182,
+      "step": 6832
+    },
+    {
+      "epoch": 1.09328,
+      "grad_norm": 2.6378598760941627,
+      "learning_rate": 0.00014699917402538656,
+      "loss": 1.5431,
+      "step": 6833
+    },
+    {
+      "epoch": 1.09344,
+      "grad_norm": 2.962943178218371,
+      "learning_rate": 0.00014698392621291685,
+      "loss": 1.5009,
+      "step": 6834
+    },
+    {
+      "epoch": 1.0936,
+      "grad_norm": 2.036265478033125,
+      "learning_rate": 0.00014696867699851525,
+      "loss": 1.3632,
+      "step": 6835
+    },
+    {
+      "epoch": 1.09376,
+      "grad_norm": 2.8745065047613867,
+      "learning_rate": 0.00014695342638263673,
+      "loss": 1.5338,
+      "step": 6836
+    },
+    {
+      "epoch": 1.09392,
+      "grad_norm": 3.1162880456702955,
+      "learning_rate": 0.00014693817436573642,
+      "loss": 1.5251,
+      "step": 6837
+    },
+    {
+      "epoch": 1.09408,
+      "grad_norm": 1.9303904971284813,
+      "learning_rate": 0.00014692292094826937,
+      "loss": 1.4737,
+      "step": 6838
+    },
+    {
+      "epoch": 1.09424,
+      "grad_norm": 3.3308355476857616,
+      "learning_rate": 0.00014690766613069075,
+      "loss": 1.5077,
+      "step": 6839
+    },
+    {
+      "epoch": 1.0944,
+      "grad_norm": 2.1668668594962295,
+      "learning_rate": 0.00014689240991345575,
+      "loss": 1.3543,
+      "step": 6840
+    },
+    {
+      "epoch": 1.09456,
+      "grad_norm": 2.464787694252338,
+      "learning_rate": 0.00014687715229701952,
+      "loss": 1.572,
+      "step": 6841
+    },
+    {
+      "epoch": 1.09472,
+      "grad_norm": 3.012266427770219,
+      "learning_rate": 0.0001468618932818374,
+      "loss": 1.5125,
+      "step": 6842
+    },
+    {
+      "epoch": 1.09488,
+      "grad_norm": 2.3590237839087744,
+      "learning_rate": 0.00014684663286836463,
+      "loss": 1.4883,
+      "step": 6843
+    },
+    {
+      "epoch": 1.09504,
+      "grad_norm": 2.523201908114162,
+      "learning_rate": 0.00014683137105705663,
+      "loss": 1.4388,
+      "step": 6844
+    },
+    {
+      "epoch": 1.0952,
+      "grad_norm": 2.0385437676017095,
+      "learning_rate": 0.00014681610784836874,
+      "loss": 1.4499,
+      "step": 6845
+    },
+    {
+      "epoch": 1.09536,
+      "grad_norm": 2.3689645174283984,
+      "learning_rate": 0.00014680084324275643,
+      "loss": 1.5107,
+      "step": 6846
+    },
+    {
+      "epoch": 1.09552,
+      "grad_norm": 2.410394858694244,
+      "learning_rate": 0.00014678557724067515,
+      "loss": 1.4469,
+      "step": 6847
+    },
+    {
+      "epoch": 1.09568,
+      "grad_norm": 2.244835508262431,
+      "learning_rate": 0.0001467703098425804,
+      "loss": 1.5044,
+      "step": 6848
+    },
+    {
+      "epoch": 1.09584,
+      "grad_norm": 2.563179139900949,
+      "learning_rate": 0.00014675504104892778,
+      "loss": 1.4711,
+      "step": 6849
+    },
+    {
+      "epoch": 1.096,
+      "grad_norm": 2.2033562749595363,
+      "learning_rate": 0.00014673977086017285,
+      "loss": 1.4185,
+      "step": 6850
+    },
+    {
+      "epoch": 1.09616,
+      "grad_norm": 2.216044785540582,
+      "learning_rate": 0.00014672449927677125,
+      "loss": 1.4665,
+      "step": 6851
+    },
+    {
+      "epoch": 1.09632,
+      "grad_norm": 2.1226029255819094,
+      "learning_rate": 0.00014670922629917872,
+      "loss": 1.4152,
+      "step": 6852
+    },
+    {
+      "epoch": 1.09648,
+      "grad_norm": 2.104764624515667,
+      "learning_rate": 0.0001466939519278509,
+      "loss": 1.55,
+      "step": 6853
+    },
+    {
+      "epoch": 1.09664,
+      "grad_norm": 2.1733189456664936,
+      "learning_rate": 0.00014667867616324364,
+      "loss": 1.4604,
+      "step": 6854
+    },
+    {
+      "epoch": 1.0968,
+      "grad_norm": 2.0819287201826358,
+      "learning_rate": 0.00014666339900581266,
+      "loss": 1.4058,
+      "step": 6855
+    },
+    {
+      "epoch": 1.09696,
+      "grad_norm": 2.449668657468445,
+      "learning_rate": 0.00014664812045601385,
+      "loss": 1.49,
+      "step": 6856
+    },
+    {
+      "epoch": 1.09712,
+      "grad_norm": 2.010218275068059,
+      "learning_rate": 0.0001466328405143031,
+      "loss": 1.3733,
+      "step": 6857
+    },
+    {
+      "epoch": 1.09728,
+      "grad_norm": 1.9033089898468945,
+      "learning_rate": 0.00014661755918113635,
+      "loss": 1.4217,
+      "step": 6858
+    },
+    {
+      "epoch": 1.09744,
+      "grad_norm": 2.301241925123,
+      "learning_rate": 0.00014660227645696956,
+      "loss": 1.4828,
+      "step": 6859
+    },
+    {
+      "epoch": 1.0976,
+      "grad_norm": 2.1528600491711942,
+      "learning_rate": 0.00014658699234225875,
+      "loss": 1.4515,
+      "step": 6860
+    },
+    {
+      "epoch": 1.09776,
+      "grad_norm": 2.1894591921834783,
+      "learning_rate": 0.00014657170683745994,
+      "loss": 1.5318,
+      "step": 6861
+    },
+    {
+      "epoch": 1.09792,
+      "grad_norm": 2.380143521605643,
+      "learning_rate": 0.00014655641994302932,
+      "loss": 1.4497,
+      "step": 6862
+    },
+    {
+      "epoch": 1.09808,
+      "grad_norm": 2.342713781758468,
+      "learning_rate": 0.00014654113165942293,
+      "loss": 1.5223,
+      "step": 6863
+    },
+    {
+      "epoch": 1.09824,
+      "grad_norm": 2.479739563956182,
+      "learning_rate": 0.000146525841987097,
+      "loss": 1.4004,
+      "step": 6864
+    },
+    {
+      "epoch": 1.0984,
+      "grad_norm": 2.2044464801223826,
+      "learning_rate": 0.00014651055092650776,
+      "loss": 1.456,
+      "step": 6865
+    },
+    {
+      "epoch": 1.09856,
+      "grad_norm": 2.5240110725795533,
+      "learning_rate": 0.00014649525847811143,
+      "loss": 1.4623,
+      "step": 6866
+    },
+    {
+      "epoch": 1.09872,
+      "grad_norm": 2.26177650802916,
+      "learning_rate": 0.00014647996464236434,
+      "loss": 1.4267,
+      "step": 6867
+    },
+    {
+      "epoch": 1.09888,
+      "grad_norm": 2.447510032314895,
+      "learning_rate": 0.00014646466941972285,
+      "loss": 1.436,
+      "step": 6868
+    },
+    {
+      "epoch": 1.09904,
+      "grad_norm": 1.970736201305463,
+      "learning_rate": 0.00014644937281064327,
+      "loss": 1.4415,
+      "step": 6869
+    },
+    {
+      "epoch": 1.0992,
+      "grad_norm": 2.2667908277093827,
+      "learning_rate": 0.00014643407481558214,
+      "loss": 1.477,
+      "step": 6870
+    },
+    {
+      "epoch": 1.09936,
+      "grad_norm": 2.6325254480920957,
+      "learning_rate": 0.00014641877543499588,
+      "loss": 1.4445,
+      "step": 6871
+    },
+    {
+      "epoch": 1.09952,
+      "grad_norm": 2.6025325078427666,
+      "learning_rate": 0.00014640347466934102,
+      "loss": 1.4781,
+      "step": 6872
+    },
+    {
+      "epoch": 1.09968,
+      "grad_norm": 2.3752130227469794,
+      "learning_rate": 0.00014638817251907407,
+      "loss": 1.4253,
+      "step": 6873
+    },
+    {
+      "epoch": 1.09984,
+      "grad_norm": 2.1614504059534045,
+      "learning_rate": 0.00014637286898465163,
+      "loss": 1.419,
+      "step": 6874
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 2.191891113374399,
+      "learning_rate": 0.0001463575640665304,
+      "loss": 1.4274,
+      "step": 6875
+    },
+    {
+      "epoch": 1.10016,
+      "grad_norm": 2.138649755139249,
+      "learning_rate": 0.00014634225776516695,
+      "loss": 1.4442,
+      "step": 6876
+    },
+    {
+      "epoch": 1.10032,
+      "grad_norm": 2.635810056308528,
+      "learning_rate": 0.00014632695008101807,
+      "loss": 1.5397,
+      "step": 6877
+    },
+    {
+      "epoch": 1.10048,
+      "grad_norm": 2.384195607481012,
+      "learning_rate": 0.0001463116410145405,
+      "loss": 1.4383,
+      "step": 6878
+    },
+    {
+      "epoch": 1.10064,
+      "grad_norm": 2.3367920926463737,
+      "learning_rate": 0.0001462963305661911,
+      "loss": 1.515,
+      "step": 6879
+    },
+    {
+      "epoch": 1.1008,
+      "grad_norm": 2.4996903280903044,
+      "learning_rate": 0.0001462810187364266,
+      "loss": 1.4402,
+      "step": 6880
+    },
+    {
+      "epoch": 1.10096,
+      "grad_norm": 2.2702118822897415,
+      "learning_rate": 0.00014626570552570396,
+      "loss": 1.4541,
+      "step": 6881
+    },
+    {
+      "epoch": 1.10112,
+      "grad_norm": 2.0100543129463984,
+      "learning_rate": 0.00014625039093448008,
+      "loss": 1.4798,
+      "step": 6882
+    },
+    {
+      "epoch": 1.10128,
+      "grad_norm": 2.067304982840222,
+      "learning_rate": 0.00014623507496321195,
+      "loss": 1.4975,
+      "step": 6883
+    },
+    {
+      "epoch": 1.10144,
+      "grad_norm": 2.2345197374287737,
+      "learning_rate": 0.0001462197576123565,
+      "loss": 1.4804,
+      "step": 6884
+    },
+    {
+      "epoch": 1.1016,
+      "grad_norm": 3.4019890916117563,
+      "learning_rate": 0.0001462044388823709,
+      "loss": 1.4847,
+      "step": 6885
+    },
+    {
+      "epoch": 1.10176,
+      "grad_norm": 2.1702798262849896,
+      "learning_rate": 0.0001461891187737121,
+      "loss": 1.4298,
+      "step": 6886
+    },
+    {
+      "epoch": 1.10192,
+      "grad_norm": 2.5274561816660217,
+      "learning_rate": 0.00014617379728683737,
+      "loss": 1.5217,
+      "step": 6887
+    },
+    {
+      "epoch": 1.10208,
+      "grad_norm": 2.5111780140286544,
+      "learning_rate": 0.00014615847442220377,
+      "loss": 1.4836,
+      "step": 6888
+    },
+    {
+      "epoch": 1.10224,
+      "grad_norm": 2.0570168577986547,
+      "learning_rate": 0.00014614315018026856,
+      "loss": 1.453,
+      "step": 6889
+    },
+    {
+      "epoch": 1.1024,
+      "grad_norm": 3.416548315425475,
+      "learning_rate": 0.00014612782456148899,
+      "loss": 1.4596,
+      "step": 6890
+    },
+    {
+      "epoch": 1.10256,
+      "grad_norm": 2.070973447063012,
+      "learning_rate": 0.00014611249756632234,
+      "loss": 1.4494,
+      "step": 6891
+    },
+    {
+      "epoch": 1.10272,
+      "grad_norm": 2.108063949856207,
+      "learning_rate": 0.00014609716919522595,
+      "loss": 1.4607,
+      "step": 6892
+    },
+    {
+      "epoch": 1.10288,
+      "grad_norm": 2.288020654447933,
+      "learning_rate": 0.0001460818394486572,
+      "loss": 1.4494,
+      "step": 6893
+    },
+    {
+      "epoch": 1.10304,
+      "grad_norm": 2.2817025684898216,
+      "learning_rate": 0.00014606650832707353,
+      "loss": 1.4715,
+      "step": 6894
+    },
+    {
+      "epoch": 1.1032,
+      "grad_norm": 2.4000746088641915,
+      "learning_rate": 0.0001460511758309324,
+      "loss": 1.5194,
+      "step": 6895
+    },
+    {
+      "epoch": 1.10336,
+      "grad_norm": 2.1053504483274996,
+      "learning_rate": 0.00014603584196069128,
+      "loss": 1.4242,
+      "step": 6896
+    },
+    {
+      "epoch": 1.10352,
+      "grad_norm": 2.2520107954128177,
+      "learning_rate": 0.00014602050671680766,
+      "loss": 1.4012,
+      "step": 6897
+    },
+    {
+      "epoch": 1.10368,
+      "grad_norm": 2.2600159054681312,
+      "learning_rate": 0.00014600517009973925,
+      "loss": 1.4958,
+      "step": 6898
+    },
+    {
+      "epoch": 1.10384,
+      "grad_norm": 2.204222153433765,
+      "learning_rate": 0.00014598983210994355,
+      "loss": 1.4514,
+      "step": 6899
+    },
+    {
+      "epoch": 1.104,
+      "grad_norm": 3.1141886879982787,
+      "learning_rate": 0.00014597449274787828,
+      "loss": 1.5057,
+      "step": 6900
+    },
+    {
+      "epoch": 1.10416,
+      "grad_norm": 2.083751745854243,
+      "learning_rate": 0.00014595915201400113,
+      "loss": 1.5033,
+      "step": 6901
+    },
+    {
+      "epoch": 1.10432,
+      "grad_norm": 3.405739787665208,
+      "learning_rate": 0.0001459438099087699,
+      "loss": 1.5122,
+      "step": 6902
+    },
+    {
+      "epoch": 1.10448,
+      "grad_norm": 2.131533231096978,
+      "learning_rate": 0.0001459284664326423,
+      "loss": 1.4884,
+      "step": 6903
+    },
+    {
+      "epoch": 1.10464,
+      "grad_norm": 2.166179821047362,
+      "learning_rate": 0.0001459131215860762,
+      "loss": 1.4471,
+      "step": 6904
+    },
+    {
+      "epoch": 1.1048,
+      "grad_norm": 2.6287719662287166,
+      "learning_rate": 0.0001458977753695294,
+      "loss": 1.468,
+      "step": 6905
+    },
+    {
+      "epoch": 1.10496,
+      "grad_norm": 2.2497051201533047,
+      "learning_rate": 0.00014588242778345992,
+      "loss": 1.3662,
+      "step": 6906
+    },
+    {
+      "epoch": 1.10512,
+      "grad_norm": 2.213321541367971,
+      "learning_rate": 0.00014586707882832567,
+      "loss": 1.4088,
+      "step": 6907
+    },
+    {
+      "epoch": 1.10528,
+      "grad_norm": 3.2329090584116575,
+      "learning_rate": 0.00014585172850458458,
+      "loss": 1.4276,
+      "step": 6908
+    },
+    {
+      "epoch": 1.10544,
+      "grad_norm": 2.1546943843519477,
+      "learning_rate": 0.00014583637681269473,
+      "loss": 1.4436,
+      "step": 6909
+    },
+    {
+      "epoch": 1.1056,
+      "grad_norm": 3.3378583265637687,
+      "learning_rate": 0.0001458210237531142,
+      "loss": 1.3444,
+      "step": 6910
+    },
+    {
+      "epoch": 1.10576,
+      "grad_norm": 2.1878155733607203,
+      "learning_rate": 0.00014580566932630105,
+      "loss": 1.4239,
+      "step": 6911
+    },
+    {
+      "epoch": 1.10592,
+      "grad_norm": 1.9989727623761813,
+      "learning_rate": 0.00014579031353271352,
+      "loss": 1.5372,
+      "step": 6912
+    },
+    {
+      "epoch": 1.10608,
+      "grad_norm": 2.917122499758635,
+      "learning_rate": 0.00014577495637280977,
+      "loss": 1.5458,
+      "step": 6913
+    },
+    {
+      "epoch": 1.1062400000000001,
+      "grad_norm": 2.2195188208893386,
+      "learning_rate": 0.000145759597847048,
+      "loss": 1.4556,
+      "step": 6914
+    },
+    {
+      "epoch": 1.1064,
+      "grad_norm": 2.2300046528804627,
+      "learning_rate": 0.0001457442379558865,
+      "loss": 1.4049,
+      "step": 6915
+    },
+    {
+      "epoch": 1.10656,
+      "grad_norm": 2.309567599793992,
+      "learning_rate": 0.00014572887669978357,
+      "loss": 1.4458,
+      "step": 6916
+    },
+    {
+      "epoch": 1.10672,
+      "grad_norm": 2.2607719345921535,
+      "learning_rate": 0.00014571351407919765,
+      "loss": 1.48,
+      "step": 6917
+    },
+    {
+      "epoch": 1.10688,
+      "grad_norm": 2.669209648288565,
+      "learning_rate": 0.00014569815009458703,
+      "loss": 1.4247,
+      "step": 6918
+    },
+    {
+      "epoch": 1.10704,
+      "grad_norm": 2.0863240628373916,
+      "learning_rate": 0.00014568278474641022,
+      "loss": 1.4781,
+      "step": 6919
+    },
+    {
+      "epoch": 1.1072,
+      "grad_norm": 2.3623682719011034,
+      "learning_rate": 0.00014566741803512568,
+      "loss": 1.4039,
+      "step": 6920
+    },
+    {
+      "epoch": 1.10736,
+      "grad_norm": 2.1558370249939998,
+      "learning_rate": 0.00014565204996119194,
+      "loss": 1.4442,
+      "step": 6921
+    },
+    {
+      "epoch": 1.10752,
+      "grad_norm": 2.465021684412463,
+      "learning_rate": 0.00014563668052506757,
+      "loss": 1.4752,
+      "step": 6922
+    },
+    {
+      "epoch": 1.10768,
+      "grad_norm": 2.323298920319639,
+      "learning_rate": 0.00014562130972721113,
+      "loss": 1.4393,
+      "step": 6923
+    },
+    {
+      "epoch": 1.10784,
+      "grad_norm": 2.2006402765176287,
+      "learning_rate": 0.00014560593756808126,
+      "loss": 1.4462,
+      "step": 6924
+    },
+    {
+      "epoch": 1.108,
+      "grad_norm": 2.464417660434194,
+      "learning_rate": 0.00014559056404813672,
+      "loss": 1.4044,
+      "step": 6925
+    },
+    {
+      "epoch": 1.10816,
+      "grad_norm": 2.223737324131621,
+      "learning_rate": 0.00014557518916783613,
+      "loss": 1.4188,
+      "step": 6926
+    },
+    {
+      "epoch": 1.10832,
+      "grad_norm": 2.206517533234633,
+      "learning_rate": 0.00014555981292763834,
+      "loss": 1.4487,
+      "step": 6927
+    },
+    {
+      "epoch": 1.10848,
+      "grad_norm": 2.0598601603305764,
+      "learning_rate": 0.0001455444353280021,
+      "loss": 1.4138,
+      "step": 6928
+    },
+    {
+      "epoch": 1.10864,
+      "grad_norm": 2.3724622110571416,
+      "learning_rate": 0.0001455290563693863,
+      "loss": 1.4798,
+      "step": 6929
+    },
+    {
+      "epoch": 1.1088,
+      "grad_norm": 2.1295460811973315,
+      "learning_rate": 0.0001455136760522498,
+      "loss": 1.3991,
+      "step": 6930
+    },
+    {
+      "epoch": 1.10896,
+      "grad_norm": 1.8717408074089503,
+      "learning_rate": 0.0001454982943770515,
+      "loss": 1.4462,
+      "step": 6931
+    },
+    {
+      "epoch": 1.1091199999999999,
+      "grad_norm": 2.3891546124877943,
+      "learning_rate": 0.00014548291134425042,
+      "loss": 1.4275,
+      "step": 6932
+    },
+    {
+      "epoch": 1.10928,
+      "grad_norm": 1.7587376871874678,
+      "learning_rate": 0.00014546752695430553,
+      "loss": 1.3969,
+      "step": 6933
+    },
+    {
+      "epoch": 1.10944,
+      "grad_norm": 2.2249493470379007,
+      "learning_rate": 0.00014545214120767586,
+      "loss": 1.5259,
+      "step": 6934
+    },
+    {
+      "epoch": 1.1096,
+      "grad_norm": 1.958235279745152,
+      "learning_rate": 0.00014543675410482053,
+      "loss": 1.4182,
+      "step": 6935
+    },
+    {
+      "epoch": 1.10976,
+      "grad_norm": 2.1773140816976797,
+      "learning_rate": 0.0001454213656461987,
+      "loss": 1.441,
+      "step": 6936
+    },
+    {
+      "epoch": 1.10992,
+      "grad_norm": 2.4719165484827488,
+      "learning_rate": 0.00014540597583226952,
+      "loss": 1.4292,
+      "step": 6937
+    },
+    {
+      "epoch": 1.11008,
+      "grad_norm": 2.2008140503661533,
+      "learning_rate": 0.00014539058466349217,
+      "loss": 1.4761,
+      "step": 6938
+    },
+    {
+      "epoch": 1.1102400000000001,
+      "grad_norm": 2.027973662074377,
+      "learning_rate": 0.0001453751921403259,
+      "loss": 1.4248,
+      "step": 6939
+    },
+    {
+      "epoch": 1.1104,
+      "grad_norm": 2.474215560603183,
+      "learning_rate": 0.00014535979826323,
+      "loss": 1.5306,
+      "step": 6940
+    },
+    {
+      "epoch": 1.11056,
+      "grad_norm": 1.9092245742242002,
+      "learning_rate": 0.00014534440303266383,
+      "loss": 1.4667,
+      "step": 6941
+    },
+    {
+      "epoch": 1.11072,
+      "grad_norm": 2.2590406508654324,
+      "learning_rate": 0.00014532900644908677,
+      "loss": 1.4775,
+      "step": 6942
+    },
+    {
+      "epoch": 1.11088,
+      "grad_norm": 1.9626447072527031,
+      "learning_rate": 0.00014531360851295817,
+      "loss": 1.4423,
+      "step": 6943
+    },
+    {
+      "epoch": 1.11104,
+      "grad_norm": 3.6654769557156377,
+      "learning_rate": 0.00014529820922473756,
+      "loss": 1.4338,
+      "step": 6944
+    },
+    {
+      "epoch": 1.1112,
+      "grad_norm": 3.305468098947037,
+      "learning_rate": 0.00014528280858488437,
+      "loss": 1.6515,
+      "step": 6945
+    },
+    {
+      "epoch": 1.11136,
+      "grad_norm": 2.6416727516207623,
+      "learning_rate": 0.00014526740659385817,
+      "loss": 1.409,
+      "step": 6946
+    },
+    {
+      "epoch": 1.11152,
+      "grad_norm": 2.2120199645766028,
+      "learning_rate": 0.0001452520032521185,
+      "loss": 1.4742,
+      "step": 6947
+    },
+    {
+      "epoch": 1.11168,
+      "grad_norm": 2.7397447861867357,
+      "learning_rate": 0.00014523659856012502,
+      "loss": 1.4812,
+      "step": 6948
+    },
+    {
+      "epoch": 1.11184,
+      "grad_norm": 2.1410539471779297,
+      "learning_rate": 0.00014522119251833733,
+      "loss": 1.4979,
+      "step": 6949
+    },
+    {
+      "epoch": 1.112,
+      "grad_norm": 2.730024504031361,
+      "learning_rate": 0.00014520578512721514,
+      "loss": 1.4664,
+      "step": 6950
+    },
+    {
+      "epoch": 1.11216,
+      "grad_norm": 2.170880069793852,
+      "learning_rate": 0.00014519037638721822,
+      "loss": 1.4792,
+      "step": 6951
+    },
+    {
+      "epoch": 1.11232,
+      "grad_norm": 2.31053173985486,
+      "learning_rate": 0.00014517496629880633,
+      "loss": 1.4288,
+      "step": 6952
+    },
+    {
+      "epoch": 1.11248,
+      "grad_norm": 2.334487889195176,
+      "learning_rate": 0.0001451595548624392,
+      "loss": 1.4421,
+      "step": 6953
+    },
+    {
+      "epoch": 1.11264,
+      "grad_norm": 2.132838376466778,
+      "learning_rate": 0.00014514414207857684,
+      "loss": 1.4609,
+      "step": 6954
+    },
+    {
+      "epoch": 1.1128,
+      "grad_norm": 2.602658384752055,
+      "learning_rate": 0.00014512872794767903,
+      "loss": 1.5065,
+      "step": 6955
+    },
+    {
+      "epoch": 1.11296,
+      "grad_norm": 1.9193897922819492,
+      "learning_rate": 0.00014511331247020573,
+      "loss": 1.436,
+      "step": 6956
+    },
+    {
+      "epoch": 1.1131199999999999,
+      "grad_norm": 2.9681589656823286,
+      "learning_rate": 0.00014509789564661697,
+      "loss": 1.4913,
+      "step": 6957
+    },
+    {
+      "epoch": 1.11328,
+      "grad_norm": 2.2983450768912945,
+      "learning_rate": 0.00014508247747737265,
+      "loss": 1.5315,
+      "step": 6958
+    },
+    {
+      "epoch": 1.11344,
+      "grad_norm": 2.1313221945640057,
+      "learning_rate": 0.00014506705796293293,
+      "loss": 1.4432,
+      "step": 6959
+    },
+    {
+      "epoch": 1.1136,
+      "grad_norm": 2.4370417000959996,
+      "learning_rate": 0.00014505163710375784,
+      "loss": 1.437,
+      "step": 6960
+    },
+    {
+      "epoch": 1.11376,
+      "grad_norm": 2.576686576559526,
+      "learning_rate": 0.0001450362149003076,
+      "loss": 1.5099,
+      "step": 6961
+    },
+    {
+      "epoch": 1.11392,
+      "grad_norm": 2.103791976421539,
+      "learning_rate": 0.00014502079135304228,
+      "loss": 1.4487,
+      "step": 6962
+    },
+    {
+      "epoch": 1.11408,
+      "grad_norm": 2.1782965717567095,
+      "learning_rate": 0.00014500536646242217,
+      "loss": 1.3767,
+      "step": 6963
+    },
+    {
+      "epoch": 1.11424,
+      "grad_norm": 2.0460373487846852,
+      "learning_rate": 0.00014498994022890753,
+      "loss": 1.5232,
+      "step": 6964
+    },
+    {
+      "epoch": 1.1144,
+      "grad_norm": 2.2525599847412083,
+      "learning_rate": 0.0001449745126529586,
+      "loss": 1.5441,
+      "step": 6965
+    },
+    {
+      "epoch": 1.11456,
+      "grad_norm": 2.3233775624280235,
+      "learning_rate": 0.00014495908373503573,
+      "loss": 1.5464,
+      "step": 6966
+    },
+    {
+      "epoch": 1.11472,
+      "grad_norm": 2.2133609411228274,
+      "learning_rate": 0.00014494365347559934,
+      "loss": 1.551,
+      "step": 6967
+    },
+    {
+      "epoch": 1.11488,
+      "grad_norm": 2.2124544625175098,
+      "learning_rate": 0.00014492822187510982,
+      "loss": 1.4508,
+      "step": 6968
+    },
+    {
+      "epoch": 1.11504,
+      "grad_norm": 1.9981070864660262,
+      "learning_rate": 0.00014491278893402762,
+      "loss": 1.4796,
+      "step": 6969
+    },
+    {
+      "epoch": 1.1152,
+      "grad_norm": 2.341226765808234,
+      "learning_rate": 0.00014489735465281325,
+      "loss": 1.4966,
+      "step": 6970
+    },
+    {
+      "epoch": 1.11536,
+      "grad_norm": 1.9090540015805784,
+      "learning_rate": 0.00014488191903192724,
+      "loss": 1.4532,
+      "step": 6971
+    },
+    {
+      "epoch": 1.11552,
+      "grad_norm": 2.435303859782138,
+      "learning_rate": 0.0001448664820718302,
+      "loss": 1.5167,
+      "step": 6972
+    },
+    {
+      "epoch": 1.11568,
+      "grad_norm": 2.1199924995442774,
+      "learning_rate": 0.00014485104377298265,
+      "loss": 1.4202,
+      "step": 6973
+    },
+    {
+      "epoch": 1.11584,
+      "grad_norm": 2.6832354433735417,
+      "learning_rate": 0.0001448356041358453,
+      "loss": 1.4532,
+      "step": 6974
+    },
+    {
+      "epoch": 1.116,
+      "grad_norm": 2.1872860984804197,
+      "learning_rate": 0.0001448201631608789,
+      "loss": 1.5206,
+      "step": 6975
+    },
+    {
+      "epoch": 1.11616,
+      "grad_norm": 2.3296653498302824,
+      "learning_rate": 0.0001448047208485441,
+      "loss": 1.4973,
+      "step": 6976
+    },
+    {
+      "epoch": 1.11632,
+      "grad_norm": 2.375503681876374,
+      "learning_rate": 0.00014478927719930175,
+      "loss": 1.5194,
+      "step": 6977
+    },
+    {
+      "epoch": 1.11648,
+      "grad_norm": 2.877100800589203,
+      "learning_rate": 0.00014477383221361264,
+      "loss": 1.4896,
+      "step": 6978
+    },
+    {
+      "epoch": 1.11664,
+      "grad_norm": 2.6499594257687327,
+      "learning_rate": 0.0001447583858919376,
+      "loss": 1.4179,
+      "step": 6979
+    },
+    {
+      "epoch": 1.1168,
+      "grad_norm": 2.929456130040946,
+      "learning_rate": 0.00014474293823473757,
+      "loss": 1.3698,
+      "step": 6980
+    },
+    {
+      "epoch": 1.11696,
+      "grad_norm": 2.335736605318734,
+      "learning_rate": 0.00014472748924247342,
+      "loss": 1.4182,
+      "step": 6981
+    },
+    {
+      "epoch": 1.11712,
+      "grad_norm": 2.8304716858928725,
+      "learning_rate": 0.00014471203891560622,
+      "loss": 1.4378,
+      "step": 6982
+    },
+    {
+      "epoch": 1.11728,
+      "grad_norm": 3.0543825884549203,
+      "learning_rate": 0.0001446965872545969,
+      "loss": 1.4619,
+      "step": 6983
+    },
+    {
+      "epoch": 1.11744,
+      "grad_norm": 2.2689388702682565,
+      "learning_rate": 0.00014468113425990652,
+      "loss": 1.3664,
+      "step": 6984
+    },
+    {
+      "epoch": 1.1176,
+      "grad_norm": 2.3234298690076223,
+      "learning_rate": 0.00014466567993199624,
+      "loss": 1.43,
+      "step": 6985
+    },
+    {
+      "epoch": 1.11776,
+      "grad_norm": 2.5261287662184757,
+      "learning_rate": 0.00014465022427132714,
+      "loss": 1.418,
+      "step": 6986
+    },
+    {
+      "epoch": 1.11792,
+      "grad_norm": 2.0965155839246323,
+      "learning_rate": 0.00014463476727836043,
+      "loss": 1.4396,
+      "step": 6987
+    },
+    {
+      "epoch": 1.11808,
+      "grad_norm": 3.083731982714114,
+      "learning_rate": 0.00014461930895355728,
+      "loss": 1.472,
+      "step": 6988
+    },
+    {
+      "epoch": 1.11824,
+      "grad_norm": 2.482075933100123,
+      "learning_rate": 0.000144603849297379,
+      "loss": 1.4357,
+      "step": 6989
+    },
+    {
+      "epoch": 1.1184,
+      "grad_norm": 3.041961303784929,
+      "learning_rate": 0.0001445883883102868,
+      "loss": 1.5335,
+      "step": 6990
+    },
+    {
+      "epoch": 1.11856,
+      "grad_norm": 2.513877112332657,
+      "learning_rate": 0.00014457292599274208,
+      "loss": 1.4976,
+      "step": 6991
+    },
+    {
+      "epoch": 1.11872,
+      "grad_norm": 2.365546200772168,
+      "learning_rate": 0.00014455746234520622,
+      "loss": 1.4587,
+      "step": 6992
+    },
+    {
+      "epoch": 1.11888,
+      "grad_norm": 3.0427506396254476,
+      "learning_rate": 0.00014454199736814062,
+      "loss": 1.4744,
+      "step": 6993
+    },
+    {
+      "epoch": 1.11904,
+      "grad_norm": 2.3901672441941333,
+      "learning_rate": 0.0001445265310620067,
+      "loss": 1.4138,
+      "step": 6994
+    },
+    {
+      "epoch": 1.1192,
+      "grad_norm": 2.2483887366788537,
+      "learning_rate": 0.00014451106342726596,
+      "loss": 1.4444,
+      "step": 6995
+    },
+    {
+      "epoch": 1.11936,
+      "grad_norm": 3.0110977299241792,
+      "learning_rate": 0.00014449559446438,
+      "loss": 1.5069,
+      "step": 6996
+    },
+    {
+      "epoch": 1.11952,
+      "grad_norm": 2.0662056066867267,
+      "learning_rate": 0.00014448012417381028,
+      "loss": 1.4279,
+      "step": 6997
+    },
+    {
+      "epoch": 1.11968,
+      "grad_norm": 2.1658202062331773,
+      "learning_rate": 0.00014446465255601852,
+      "loss": 1.4182,
+      "step": 6998
+    },
+    {
+      "epoch": 1.11984,
+      "grad_norm": 3.5314261109820793,
+      "learning_rate": 0.00014444917961146624,
+      "loss": 1.4351,
+      "step": 6999
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 2.248355731706868,
+      "learning_rate": 0.00014443370534061528,
+      "loss": 1.5134,
+      "step": 7000
+    },
+    {
+      "epoch": 1.12016,
+      "grad_norm": 3.1504921232942187,
+      "learning_rate": 0.0001444182297439273,
+      "loss": 1.454,
+      "step": 7001
+    },
+    {
+      "epoch": 1.12032,
+      "grad_norm": 2.0778935272177783,
+      "learning_rate": 0.00014440275282186403,
+      "loss": 1.4311,
+      "step": 7002
+    },
+    {
+      "epoch": 1.12048,
+      "grad_norm": 1.9846159175428277,
+      "learning_rate": 0.00014438727457488737,
+      "loss": 1.4424,
+      "step": 7003
+    },
+    {
+      "epoch": 1.12064,
+      "grad_norm": 2.6548745764451875,
+      "learning_rate": 0.00014437179500345908,
+      "loss": 1.4893,
+      "step": 7004
+    },
+    {
+      "epoch": 1.1208,
+      "grad_norm": 2.388582776822154,
+      "learning_rate": 0.00014435631410804106,
+      "loss": 1.4601,
+      "step": 7005
+    },
+    {
+      "epoch": 1.12096,
+      "grad_norm": 2.3210168150207826,
+      "learning_rate": 0.0001443408318890953,
+      "loss": 1.4616,
+      "step": 7006
+    },
+    {
+      "epoch": 1.12112,
+      "grad_norm": 1.9382144138534743,
+      "learning_rate": 0.0001443253483470837,
+      "loss": 1.4424,
+      "step": 7007
+    },
+    {
+      "epoch": 1.12128,
+      "grad_norm": 2.4382947393312433,
+      "learning_rate": 0.0001443098634824683,
+      "loss": 1.4912,
+      "step": 7008
+    },
+    {
+      "epoch": 1.12144,
+      "grad_norm": 2.2372680474105033,
+      "learning_rate": 0.0001442943772957112,
+      "loss": 1.3783,
+      "step": 7009
+    },
+    {
+      "epoch": 1.1216,
+      "grad_norm": 2.078455823015235,
+      "learning_rate": 0.00014427888978727434,
+      "loss": 1.4424,
+      "step": 7010
+    },
+    {
+      "epoch": 1.12176,
+      "grad_norm": 2.1786429899372948,
+      "learning_rate": 0.00014426340095762,
+      "loss": 1.46,
+      "step": 7011
+    },
+    {
+      "epoch": 1.12192,
+      "grad_norm": 2.087183474065499,
+      "learning_rate": 0.00014424791080721024,
+      "loss": 1.508,
+      "step": 7012
+    },
+    {
+      "epoch": 1.12208,
+      "grad_norm": 2.2499708180855174,
+      "learning_rate": 0.00014423241933650727,
+      "loss": 1.3806,
+      "step": 7013
+    },
+    {
+      "epoch": 1.12224,
+      "grad_norm": 2.3190569973653945,
+      "learning_rate": 0.0001442169265459734,
+      "loss": 1.5214,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1224,
+      "grad_norm": 2.458799084518842,
+      "learning_rate": 0.00014420143243607084,
+      "loss": 1.4936,
+      "step": 7015
+    },
+    {
+      "epoch": 1.12256,
+      "grad_norm": 1.934151750461296,
+      "learning_rate": 0.00014418593700726197,
+      "loss": 1.4313,
+      "step": 7016
+    },
+    {
+      "epoch": 1.12272,
+      "grad_norm": 2.1700939025161627,
+      "learning_rate": 0.0001441704402600091,
+      "loss": 1.4065,
+      "step": 7017
+    },
+    {
+      "epoch": 1.12288,
+      "grad_norm": 1.9134059327915829,
+      "learning_rate": 0.0001441549421947747,
+      "loss": 1.4368,
+      "step": 7018
+    },
+    {
+      "epoch": 1.12304,
+      "grad_norm": 2.1350308905789244,
+      "learning_rate": 0.00014413944281202115,
+      "loss": 1.3454,
+      "step": 7019
+    },
+    {
+      "epoch": 1.1232,
+      "grad_norm": 2.3682888225746406,
+      "learning_rate": 0.00014412394211221093,
+      "loss": 1.5667,
+      "step": 7020
+    },
+    {
+      "epoch": 1.12336,
+      "grad_norm": 2.114548860242049,
+      "learning_rate": 0.0001441084400958066,
+      "loss": 1.4,
+      "step": 7021
+    },
+    {
+      "epoch": 1.12352,
+      "grad_norm": 2.0208040484900907,
+      "learning_rate": 0.00014409293676327067,
+      "loss": 1.3779,
+      "step": 7022
+    },
+    {
+      "epoch": 1.12368,
+      "grad_norm": 2.648379897759166,
+      "learning_rate": 0.00014407743211506573,
+      "loss": 1.4118,
+      "step": 7023
+    },
+    {
+      "epoch": 1.12384,
+      "grad_norm": 2.3928625760386875,
+      "learning_rate": 0.0001440619261516545,
+      "loss": 1.4399,
+      "step": 7024
+    },
+    {
+      "epoch": 1.124,
+      "grad_norm": 2.602738136748125,
+      "learning_rate": 0.00014404641887349955,
+      "loss": 1.456,
+      "step": 7025
+    },
+    {
+      "epoch": 1.12416,
+      "grad_norm": 2.315514376848535,
+      "learning_rate": 0.00014403091028106365,
+      "loss": 1.4758,
+      "step": 7026
+    },
+    {
+      "epoch": 1.12432,
+      "grad_norm": 2.188873577536733,
+      "learning_rate": 0.00014401540037480956,
+      "loss": 1.4062,
+      "step": 7027
+    },
+    {
+      "epoch": 1.12448,
+      "grad_norm": 2.2139479281700845,
+      "learning_rate": 0.00014399988915520006,
+      "loss": 1.4663,
+      "step": 7028
+    },
+    {
+      "epoch": 1.12464,
+      "grad_norm": 2.2593220844057873,
+      "learning_rate": 0.000143984376622698,
+      "loss": 1.4883,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1248,
+      "grad_norm": 2.4072832528136017,
+      "learning_rate": 0.0001439688627777662,
+      "loss": 1.5401,
+      "step": 7030
+    },
+    {
+      "epoch": 1.12496,
+      "grad_norm": 2.7517272417919667,
+      "learning_rate": 0.0001439533476208676,
+      "loss": 1.4096,
+      "step": 7031
+    },
+    {
+      "epoch": 1.12512,
+      "grad_norm": 2.232294286907889,
+      "learning_rate": 0.00014393783115246517,
+      "loss": 1.49,
+      "step": 7032
+    },
+    {
+      "epoch": 1.12528,
+      "grad_norm": 2.9602156897089205,
+      "learning_rate": 0.0001439223133730219,
+      "loss": 1.5278,
+      "step": 7033
+    },
+    {
+      "epoch": 1.12544,
+      "grad_norm": 2.1101826119154357,
+      "learning_rate": 0.00014390679428300077,
+      "loss": 1.4204,
+      "step": 7034
+    },
+    {
+      "epoch": 1.1256,
+      "grad_norm": 2.867714910219544,
+      "learning_rate": 0.0001438912738828649,
+      "loss": 1.449,
+      "step": 7035
+    },
+    {
+      "epoch": 1.12576,
+      "grad_norm": 2.8903920519768604,
+      "learning_rate": 0.00014387575217307736,
+      "loss": 1.5311,
+      "step": 7036
+    },
+    {
+      "epoch": 1.12592,
+      "grad_norm": 2.384401470969599,
+      "learning_rate": 0.00014386022915410132,
+      "loss": 1.4616,
+      "step": 7037
+    },
+    {
+      "epoch": 1.12608,
+      "grad_norm": 3.0499617862388466,
+      "learning_rate": 0.00014384470482639995,
+      "loss": 1.4482,
+      "step": 7038
+    },
+    {
+      "epoch": 1.12624,
+      "grad_norm": 2.1895053914419087,
+      "learning_rate": 0.0001438291791904365,
+      "loss": 1.468,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1264,
+      "grad_norm": 2.780508886729483,
+      "learning_rate": 0.00014381365224667417,
+      "loss": 1.5058,
+      "step": 7040
+    },
+    {
+      "epoch": 1.12656,
+      "grad_norm": 2.2650140662039644,
+      "learning_rate": 0.00014379812399557628,
+      "loss": 1.4528,
+      "step": 7041
+    },
+    {
+      "epoch": 1.12672,
+      "grad_norm": 2.1398422886498585,
+      "learning_rate": 0.00014378259443760625,
+      "loss": 1.464,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1268799999999999,
+      "grad_norm": 2.3286020469284545,
+      "learning_rate": 0.00014376706357322735,
+      "loss": 1.4579,
+      "step": 7043
+    },
+    {
+      "epoch": 1.12704,
+      "grad_norm": 2.2808952918335543,
+      "learning_rate": 0.00014375153140290309,
+      "loss": 1.4306,
+      "step": 7044
+    },
+    {
+      "epoch": 1.1272,
+      "grad_norm": 2.272042398250336,
+      "learning_rate": 0.00014373599792709687,
+      "loss": 1.439,
+      "step": 7045
+    },
+    {
+      "epoch": 1.12736,
+      "grad_norm": 2.6463053467789064,
+      "learning_rate": 0.00014372046314627221,
+      "loss": 1.5028,
+      "step": 7046
+    },
+    {
+      "epoch": 1.12752,
+      "grad_norm": 2.617236590469936,
+      "learning_rate": 0.00014370492706089263,
+      "loss": 1.5202,
+      "step": 7047
+    },
+    {
+      "epoch": 1.12768,
+      "grad_norm": 1.9680239230628518,
+      "learning_rate": 0.00014368938967142166,
+      "loss": 1.4533,
+      "step": 7048
+    },
+    {
+      "epoch": 1.12784,
+      "grad_norm": 3.1353731239573945,
+      "learning_rate": 0.00014367385097832305,
+      "loss": 1.522,
+      "step": 7049
+    },
+    {
+      "epoch": 1.1280000000000001,
+      "grad_norm": 2.0890680356951967,
+      "learning_rate": 0.00014365831098206032,
+      "loss": 1.4425,
+      "step": 7050
+    },
+    {
+      "epoch": 1.12816,
+      "grad_norm": 2.8108825127067156,
+      "learning_rate": 0.0001436427696830972,
+      "loss": 1.4767,
+      "step": 7051
+    },
+    {
+      "epoch": 1.12832,
+      "grad_norm": 2.5973459390287625,
+      "learning_rate": 0.00014362722708189746,
+      "loss": 1.4322,
+      "step": 7052
+    },
+    {
+      "epoch": 1.12848,
+      "grad_norm": 2.2042855294838684,
+      "learning_rate": 0.00014361168317892484,
+      "loss": 1.4291,
+      "step": 7053
+    },
+    {
+      "epoch": 1.12864,
+      "grad_norm": 2.258557557141651,
+      "learning_rate": 0.00014359613797464312,
+      "loss": 1.4215,
+      "step": 7054
+    },
+    {
+      "epoch": 1.1288,
+      "grad_norm": 2.181141444659206,
+      "learning_rate": 0.00014358059146951616,
+      "loss": 1.4052,
+      "step": 7055
+    },
+    {
+      "epoch": 1.12896,
+      "grad_norm": 2.2033271356063198,
+      "learning_rate": 0.00014356504366400786,
+      "loss": 1.5012,
+      "step": 7056
+    },
+    {
+      "epoch": 1.12912,
+      "grad_norm": 2.271715745167451,
+      "learning_rate": 0.00014354949455858213,
+      "loss": 1.3845,
+      "step": 7057
+    },
+    {
+      "epoch": 1.12928,
+      "grad_norm": 2.0529991362972813,
+      "learning_rate": 0.00014353394415370297,
+      "loss": 1.403,
+      "step": 7058
+    },
+    {
+      "epoch": 1.12944,
+      "grad_norm": 2.081621673272986,
+      "learning_rate": 0.00014351839244983433,
+      "loss": 1.4557,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1296,
+      "grad_norm": 2.360132605174742,
+      "learning_rate": 0.0001435028394474403,
+      "loss": 1.4785,
+      "step": 7060
+    },
+    {
+      "epoch": 1.12976,
+      "grad_norm": 2.102779614331327,
+      "learning_rate": 0.00014348728514698489,
+      "loss": 1.5289,
+      "step": 7061
+    },
+    {
+      "epoch": 1.12992,
+      "grad_norm": 2.019918980437422,
+      "learning_rate": 0.00014347172954893227,
+      "loss": 1.4425,
+      "step": 7062
+    },
+    {
+      "epoch": 1.13008,
+      "grad_norm": 2.2621033569705924,
+      "learning_rate": 0.00014345617265374664,
+      "loss": 1.4414,
+      "step": 7063
+    },
+    {
+      "epoch": 1.13024,
+      "grad_norm": 2.110625384298827,
+      "learning_rate": 0.00014344061446189206,
+      "loss": 1.5264,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1304,
+      "grad_norm": 2.3204235429911653,
+      "learning_rate": 0.00014342505497383286,
+      "loss": 1.4231,
+      "step": 7065
+    },
+    {
+      "epoch": 1.13056,
+      "grad_norm": 2.4620037968217763,
+      "learning_rate": 0.00014340949419003335,
+      "loss": 1.4215,
+      "step": 7066
+    },
+    {
+      "epoch": 1.13072,
+      "grad_norm": 2.3228832444024103,
+      "learning_rate": 0.00014339393211095774,
+      "loss": 1.4713,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1308799999999999,
+      "grad_norm": 2.4218691847788114,
+      "learning_rate": 0.00014337836873707044,
+      "loss": 1.4712,
+      "step": 7068
+    },
+    {
+      "epoch": 1.13104,
+      "grad_norm": 2.187492429791224,
+      "learning_rate": 0.00014336280406883583,
+      "loss": 1.4028,
+      "step": 7069
+    },
+    {
+      "epoch": 1.1312,
+      "grad_norm": 2.239560489982189,
+      "learning_rate": 0.00014334723810671834,
+      "loss": 1.3823,
+      "step": 7070
+    },
+    {
+      "epoch": 1.13136,
+      "grad_norm": 2.387290693437035,
+      "learning_rate": 0.0001433316708511824,
+      "loss": 1.4881,
+      "step": 7071
+    },
+    {
+      "epoch": 1.13152,
+      "grad_norm": 2.087186751637269,
+      "learning_rate": 0.00014331610230269254,
+      "loss": 1.3886,
+      "step": 7072
+    },
+    {
+      "epoch": 1.13168,
+      "grad_norm": 2.35635478776315,
+      "learning_rate": 0.00014330053246171332,
+      "loss": 1.4114,
+      "step": 7073
+    },
+    {
+      "epoch": 1.13184,
+      "grad_norm": 2.3139629418592085,
+      "learning_rate": 0.00014328496132870932,
+      "loss": 1.3799,
+      "step": 7074
+    },
+    {
+      "epoch": 1.1320000000000001,
+      "grad_norm": 2.1378364710322377,
+      "learning_rate": 0.00014326938890414518,
+      "loss": 1.5298,
+      "step": 7075
+    },
+    {
+      "epoch": 1.13216,
+      "grad_norm": 2.300208817244198,
+      "learning_rate": 0.00014325381518848546,
+      "loss": 1.3972,
+      "step": 7076
+    },
+    {
+      "epoch": 1.13232,
+      "grad_norm": 2.3080884789679224,
+      "learning_rate": 0.00014323824018219495,
+      "loss": 1.4859,
+      "step": 7077
+    },
+    {
+      "epoch": 1.13248,
+      "grad_norm": 2.9029961321548563,
+      "learning_rate": 0.00014322266388573835,
+      "loss": 1.4607,
+      "step": 7078
+    },
+    {
+      "epoch": 1.13264,
+      "grad_norm": 2.401918500773859,
+      "learning_rate": 0.00014320708629958045,
+      "loss": 1.4348,
+      "step": 7079
+    },
+    {
+      "epoch": 1.1328,
+      "grad_norm": 2.5473376250515125,
+      "learning_rate": 0.00014319150742418603,
+      "loss": 1.4233,
+      "step": 7080
+    },
+    {
+      "epoch": 1.13296,
+      "grad_norm": 2.500378593017387,
+      "learning_rate": 0.00014317592726001999,
+      "loss": 1.4577,
+      "step": 7081
+    },
+    {
+      "epoch": 1.13312,
+      "grad_norm": 2.619056465456393,
+      "learning_rate": 0.0001431603458075472,
+      "loss": 1.4504,
+      "step": 7082
+    },
+    {
+      "epoch": 1.13328,
+      "grad_norm": 2.881398245634495,
+      "learning_rate": 0.00014314476306723258,
+      "loss": 1.4914,
+      "step": 7083
+    },
+    {
+      "epoch": 1.13344,
+      "grad_norm": 2.415153081951359,
+      "learning_rate": 0.0001431291790395411,
+      "loss": 1.4055,
+      "step": 7084
+    },
+    {
+      "epoch": 1.1336,
+      "grad_norm": 2.377399093871713,
+      "learning_rate": 0.00014311359372493773,
+      "loss": 1.4417,
+      "step": 7085
+    },
+    {
+      "epoch": 1.13376,
+      "grad_norm": 2.249657149414559,
+      "learning_rate": 0.00014309800712388758,
+      "loss": 1.5788,
+      "step": 7086
+    },
+    {
+      "epoch": 1.13392,
+      "grad_norm": 3.279744936196465,
+      "learning_rate": 0.0001430824192368557,
+      "loss": 1.4043,
+      "step": 7087
+    },
+    {
+      "epoch": 1.13408,
+      "grad_norm": 2.2833123849653707,
+      "learning_rate": 0.00014306683006430715,
+      "loss": 1.4556,
+      "step": 7088
+    },
+    {
+      "epoch": 1.13424,
+      "grad_norm": 2.4422538118678325,
+      "learning_rate": 0.00014305123960670722,
+      "loss": 1.5163,
+      "step": 7089
+    },
+    {
+      "epoch": 1.1344,
+      "grad_norm": 2.4721945922226243,
+      "learning_rate": 0.000143035647864521,
+      "loss": 1.5536,
+      "step": 7090
+    },
+    {
+      "epoch": 1.13456,
+      "grad_norm": 2.5930252784881884,
+      "learning_rate": 0.00014302005483821377,
+      "loss": 1.5412,
+      "step": 7091
+    },
+    {
+      "epoch": 1.13472,
+      "grad_norm": 2.1437454244472214,
+      "learning_rate": 0.00014300446052825078,
+      "loss": 1.4942,
+      "step": 7092
+    },
+    {
+      "epoch": 1.1348799999999999,
+      "grad_norm": 1.8950466005652182,
+      "learning_rate": 0.00014298886493509732,
+      "loss": 1.4181,
+      "step": 7093
+    },
+    {
+      "epoch": 1.13504,
+      "grad_norm": 2.195231651916534,
+      "learning_rate": 0.00014297326805921883,
+      "loss": 1.4495,
+      "step": 7094
+    },
+    {
+      "epoch": 1.1352,
+      "grad_norm": 2.094338387547407,
+      "learning_rate": 0.0001429576699010806,
+      "loss": 1.4612,
+      "step": 7095
+    },
+    {
+      "epoch": 1.13536,
+      "grad_norm": 2.0572253883730456,
+      "learning_rate": 0.00014294207046114813,
+      "loss": 1.4247,
+      "step": 7096
+    },
+    {
+      "epoch": 1.13552,
+      "grad_norm": 2.4897147657069683,
+      "learning_rate": 0.00014292646973988683,
+      "loss": 1.5595,
+      "step": 7097
+    },
+    {
+      "epoch": 1.13568,
+      "grad_norm": 2.2137927684435628,
+      "learning_rate": 0.0001429108677377622,
+      "loss": 1.3794,
+      "step": 7098
+    },
+    {
+      "epoch": 1.13584,
+      "grad_norm": 2.9481770799405473,
+      "learning_rate": 0.00014289526445523983,
+      "loss": 1.5016,
+      "step": 7099
+    },
+    {
+      "epoch": 1.1360000000000001,
+      "grad_norm": 2.4648200336133725,
+      "learning_rate": 0.0001428796598927853,
+      "loss": 1.4238,
+      "step": 7100
+    },
+    {
+      "epoch": 1.13616,
+      "grad_norm": 2.757496565170845,
+      "learning_rate": 0.00014286405405086416,
+      "loss": 1.5303,
+      "step": 7101
+    },
+    {
+      "epoch": 1.13632,
+      "grad_norm": 2.2072992324464447,
+      "learning_rate": 0.00014284844692994214,
+      "loss": 1.3686,
+      "step": 7102
+    },
+    {
+      "epoch": 1.13648,
+      "grad_norm": 2.2488778108089766,
+      "learning_rate": 0.0001428328385304849,
+      "loss": 1.525,
+      "step": 7103
+    },
+    {
+      "epoch": 1.13664,
+      "grad_norm": 2.21774405909199,
+      "learning_rate": 0.00014281722885295815,
+      "loss": 1.4145,
+      "step": 7104
+    },
+    {
+      "epoch": 1.1368,
+      "grad_norm": 2.2052015347190315,
+      "learning_rate": 0.00014280161789782767,
+      "loss": 1.4754,
+      "step": 7105
+    },
+    {
+      "epoch": 1.13696,
+      "grad_norm": 2.357335758048635,
+      "learning_rate": 0.00014278600566555927,
+      "loss": 1.3919,
+      "step": 7106
+    },
+    {
+      "epoch": 1.13712,
+      "grad_norm": 2.192247391131239,
+      "learning_rate": 0.00014277039215661885,
+      "loss": 1.3878,
+      "step": 7107
+    },
+    {
+      "epoch": 1.13728,
+      "grad_norm": 2.990690486335954,
+      "learning_rate": 0.00014275477737147224,
+      "loss": 1.5016,
+      "step": 7108
+    },
+    {
+      "epoch": 1.13744,
+      "grad_norm": 2.762692856562104,
+      "learning_rate": 0.00014273916131058538,
+      "loss": 1.511,
+      "step": 7109
+    },
+    {
+      "epoch": 1.1376,
+      "grad_norm": 2.7023439415787815,
+      "learning_rate": 0.0001427235439744242,
+      "loss": 1.4974,
+      "step": 7110
+    },
+    {
+      "epoch": 1.13776,
+      "grad_norm": 2.03384222387162,
+      "learning_rate": 0.00014270792536345475,
+      "loss": 1.4699,
+      "step": 7111
+    },
+    {
+      "epoch": 1.13792,
+      "grad_norm": 2.484705853005669,
+      "learning_rate": 0.00014269230547814303,
+      "loss": 1.4873,
+      "step": 7112
+    },
+    {
+      "epoch": 1.13808,
+      "grad_norm": 2.2326666693271955,
+      "learning_rate": 0.0001426766843189551,
+      "loss": 1.4501,
+      "step": 7113
+    },
+    {
+      "epoch": 1.13824,
+      "grad_norm": 2.1324036938729605,
+      "learning_rate": 0.0001426610618863571,
+      "loss": 1.4878,
+      "step": 7114
+    },
+    {
+      "epoch": 1.1384,
+      "grad_norm": 1.947182406693196,
+      "learning_rate": 0.0001426454381808152,
+      "loss": 1.4038,
+      "step": 7115
+    },
+    {
+      "epoch": 1.13856,
+      "grad_norm": 2.1593096992836824,
+      "learning_rate": 0.00014262981320279557,
+      "loss": 1.4455,
+      "step": 7116
+    },
+    {
+      "epoch": 1.13872,
+      "grad_norm": 2.4501523437786097,
+      "learning_rate": 0.0001426141869527644,
+      "loss": 1.4897,
+      "step": 7117
+    },
+    {
+      "epoch": 1.13888,
+      "grad_norm": 2.3071655042207944,
+      "learning_rate": 0.00014259855943118803,
+      "loss": 1.4904,
+      "step": 7118
+    },
+    {
+      "epoch": 1.13904,
+      "grad_norm": 2.3284133031480705,
+      "learning_rate": 0.0001425829306385327,
+      "loss": 1.4093,
+      "step": 7119
+    },
+    {
+      "epoch": 1.1392,
+      "grad_norm": 2.2500751881792636,
+      "learning_rate": 0.00014256730057526476,
+      "loss": 1.4709,
+      "step": 7120
+    },
+    {
+      "epoch": 1.13936,
+      "grad_norm": 2.5534449070765355,
+      "learning_rate": 0.00014255166924185061,
+      "loss": 1.446,
+      "step": 7121
+    },
+    {
+      "epoch": 1.13952,
+      "grad_norm": 2.486265500542126,
+      "learning_rate": 0.00014253603663875664,
+      "loss": 1.4732,
+      "step": 7122
+    },
+    {
+      "epoch": 1.13968,
+      "grad_norm": 2.138271632945608,
+      "learning_rate": 0.0001425204027664493,
+      "loss": 1.4485,
+      "step": 7123
+    },
+    {
+      "epoch": 1.13984,
+      "grad_norm": 1.9602940664855855,
+      "learning_rate": 0.00014250476762539514,
+      "loss": 1.4653,
+      "step": 7124
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 2.0728554937626558,
+      "learning_rate": 0.00014248913121606065,
+      "loss": 1.4615,
+      "step": 7125
+    },
+    {
+      "epoch": 1.14016,
+      "grad_norm": 2.034588349118574,
+      "learning_rate": 0.00014247349353891238,
+      "loss": 1.4324,
+      "step": 7126
+    },
+    {
+      "epoch": 1.14032,
+      "grad_norm": 2.0660033566333027,
+      "learning_rate": 0.00014245785459441696,
+      "loss": 1.463,
+      "step": 7127
+    },
+    {
+      "epoch": 1.14048,
+      "grad_norm": 2.642403942185432,
+      "learning_rate": 0.00014244221438304103,
+      "loss": 1.642,
+      "step": 7128
+    },
+    {
+      "epoch": 1.1406399999999999,
+      "grad_norm": 2.6035889525927836,
+      "learning_rate": 0.00014242657290525126,
+      "loss": 1.4146,
+      "step": 7129
+    },
+    {
+      "epoch": 1.1408,
+      "grad_norm": 2.41705351416977,
+      "learning_rate": 0.00014241093016151437,
+      "loss": 1.3901,
+      "step": 7130
+    },
+    {
+      "epoch": 1.14096,
+      "grad_norm": 2.011762037523282,
+      "learning_rate": 0.0001423952861522971,
+      "loss": 1.4294,
+      "step": 7131
+    },
+    {
+      "epoch": 1.14112,
+      "grad_norm": 2.2708096968849074,
+      "learning_rate": 0.00014237964087806631,
+      "loss": 1.4295,
+      "step": 7132
+    },
+    {
+      "epoch": 1.14128,
+      "grad_norm": 2.1045177584218115,
+      "learning_rate": 0.0001423639943392888,
+      "loss": 1.4393,
+      "step": 7133
+    },
+    {
+      "epoch": 1.14144,
+      "grad_norm": 2.24198551464442,
+      "learning_rate": 0.0001423483465364314,
+      "loss": 1.4413,
+      "step": 7134
+    },
+    {
+      "epoch": 1.1416,
+      "grad_norm": 2.0412240620236006,
+      "learning_rate": 0.00014233269746996108,
+      "loss": 1.424,
+      "step": 7135
+    },
+    {
+      "epoch": 1.14176,
+      "grad_norm": 2.177926807370498,
+      "learning_rate": 0.0001423170471403447,
+      "loss": 1.4712,
+      "step": 7136
+    },
+    {
+      "epoch": 1.14192,
+      "grad_norm": 2.384582280709894,
+      "learning_rate": 0.00014230139554804934,
+      "loss": 1.4346,
+      "step": 7137
+    },
+    {
+      "epoch": 1.14208,
+      "grad_norm": 2.3526212196475713,
+      "learning_rate": 0.00014228574269354194,
+      "loss": 1.4539,
+      "step": 7138
+    },
+    {
+      "epoch": 1.14224,
+      "grad_norm": 2.218354606662127,
+      "learning_rate": 0.00014227008857728963,
+      "loss": 1.4641,
+      "step": 7139
+    },
+    {
+      "epoch": 1.1424,
+      "grad_norm": 2.190970474021144,
+      "learning_rate": 0.00014225443319975943,
+      "loss": 1.466,
+      "step": 7140
+    },
+    {
+      "epoch": 1.14256,
+      "grad_norm": 2.6620597370965964,
+      "learning_rate": 0.00014223877656141854,
+      "loss": 1.4975,
+      "step": 7141
+    },
+    {
+      "epoch": 1.14272,
+      "grad_norm": 2.3858747336420403,
+      "learning_rate": 0.00014222311866273412,
+      "loss": 1.4954,
+      "step": 7142
+    },
+    {
+      "epoch": 1.14288,
+      "grad_norm": 2.0507684303064657,
+      "learning_rate": 0.00014220745950417336,
+      "loss": 1.4547,
+      "step": 7143
+    },
+    {
+      "epoch": 1.14304,
+      "grad_norm": 2.362853945812006,
+      "learning_rate": 0.0001421917990862035,
+      "loss": 1.4876,
+      "step": 7144
+    },
+    {
+      "epoch": 1.1432,
+      "grad_norm": 2.0249644830318245,
+      "learning_rate": 0.00014217613740929183,
+      "loss": 1.4739,
+      "step": 7145
+    },
+    {
+      "epoch": 1.14336,
+      "grad_norm": 2.0423009903527483,
+      "learning_rate": 0.00014216047447390568,
+      "loss": 1.3561,
+      "step": 7146
+    },
+    {
+      "epoch": 1.14352,
+      "grad_norm": 1.6890593596318084,
+      "learning_rate": 0.00014214481028051242,
+      "loss": 1.4668,
+      "step": 7147
+    },
+    {
+      "epoch": 1.14368,
+      "grad_norm": 2.2847498370589303,
+      "learning_rate": 0.0001421291448295794,
+      "loss": 1.3875,
+      "step": 7148
+    },
+    {
+      "epoch": 1.14384,
+      "grad_norm": 2.572477525108263,
+      "learning_rate": 0.00014211347812157415,
+      "loss": 1.4009,
+      "step": 7149
+    },
+    {
+      "epoch": 1.144,
+      "grad_norm": 2.2061983263237046,
+      "learning_rate": 0.00014209781015696402,
+      "loss": 1.4547,
+      "step": 7150
+    },
+    {
+      "epoch": 1.14416,
+      "grad_norm": 2.634827880487511,
+      "learning_rate": 0.00014208214093621658,
+      "loss": 1.3528,
+      "step": 7151
+    },
+    {
+      "epoch": 1.14432,
+      "grad_norm": 2.1172719809358904,
+      "learning_rate": 0.0001420664704597994,
+      "loss": 1.4473,
+      "step": 7152
+    },
+    {
+      "epoch": 1.14448,
+      "grad_norm": 2.2562914511689556,
+      "learning_rate": 0.00014205079872818004,
+      "loss": 1.4401,
+      "step": 7153
+    },
+    {
+      "epoch": 1.1446399999999999,
+      "grad_norm": 2.167384541477942,
+      "learning_rate": 0.00014203512574182612,
+      "loss": 1.3767,
+      "step": 7154
+    },
+    {
+      "epoch": 1.1448,
+      "grad_norm": 2.145681074293765,
+      "learning_rate": 0.00014201945150120525,
+      "loss": 1.4869,
+      "step": 7155
+    },
+    {
+      "epoch": 1.14496,
+      "grad_norm": 2.3472951087158433,
+      "learning_rate": 0.00014200377600678523,
+      "loss": 1.3462,
+      "step": 7156
+    },
+    {
+      "epoch": 1.14512,
+      "grad_norm": 2.4484214498353514,
+      "learning_rate": 0.00014198809925903373,
+      "loss": 1.3894,
+      "step": 7157
+    },
+    {
+      "epoch": 1.14528,
+      "grad_norm": 2.4004860771642287,
+      "learning_rate": 0.00014197242125841853,
+      "loss": 1.5503,
+      "step": 7158
+    },
+    {
+      "epoch": 1.14544,
+      "grad_norm": 2.3445908526239694,
+      "learning_rate": 0.00014195674200540744,
+      "loss": 1.3456,
+      "step": 7159
+    },
+    {
+      "epoch": 1.1456,
+      "grad_norm": 2.45134107952737,
+      "learning_rate": 0.00014194106150046832,
+      "loss": 1.3859,
+      "step": 7160
+    },
+    {
+      "epoch": 1.1457600000000001,
+      "grad_norm": 2.448318500500611,
+      "learning_rate": 0.000141925379744069,
+      "loss": 1.4091,
+      "step": 7161
+    },
+    {
+      "epoch": 1.14592,
+      "grad_norm": 2.220186084945294,
+      "learning_rate": 0.00014190969673667748,
+      "loss": 1.4733,
+      "step": 7162
+    },
+    {
+      "epoch": 1.14608,
+      "grad_norm": 2.0274145293590897,
+      "learning_rate": 0.00014189401247876164,
+      "loss": 1.4084,
+      "step": 7163
+    },
+    {
+      "epoch": 1.14624,
+      "grad_norm": 2.7816832896097377,
+      "learning_rate": 0.00014187832697078954,
+      "loss": 1.4365,
+      "step": 7164
+    },
+    {
+      "epoch": 1.1464,
+      "grad_norm": 2.2297091334721406,
+      "learning_rate": 0.00014186264021322917,
+      "loss": 1.5573,
+      "step": 7165
+    },
+    {
+      "epoch": 1.14656,
+      "grad_norm": 2.132060918554815,
+      "learning_rate": 0.00014184695220654863,
+      "loss": 1.4392,
+      "step": 7166
+    },
+    {
+      "epoch": 1.14672,
+      "grad_norm": 2.3777699391036244,
+      "learning_rate": 0.00014183126295121604,
+      "loss": 1.4207,
+      "step": 7167
+    },
+    {
+      "epoch": 1.14688,
+      "grad_norm": 1.9858600756500018,
+      "learning_rate": 0.00014181557244769948,
+      "loss": 1.4319,
+      "step": 7168
+    },
+    {
+      "epoch": 1.14704,
+      "grad_norm": 2.2485543809502766,
+      "learning_rate": 0.00014179988069646716,
+      "loss": 1.513,
+      "step": 7169
+    },
+    {
+      "epoch": 1.1472,
+      "grad_norm": 2.3563731912088364,
+      "learning_rate": 0.00014178418769798733,
+      "loss": 1.4521,
+      "step": 7170
+    },
+    {
+      "epoch": 1.14736,
+      "grad_norm": 2.1671340941797284,
+      "learning_rate": 0.0001417684934527282,
+      "loss": 1.3826,
+      "step": 7171
+    },
+    {
+      "epoch": 1.14752,
+      "grad_norm": 2.5490417687640328,
+      "learning_rate": 0.00014175279796115815,
+      "loss": 1.398,
+      "step": 7172
+    },
+    {
+      "epoch": 1.14768,
+      "grad_norm": 1.9326147063458887,
+      "learning_rate": 0.00014173710122374538,
+      "loss": 1.3754,
+      "step": 7173
+    },
+    {
+      "epoch": 1.14784,
+      "grad_norm": 2.3557650601539213,
+      "learning_rate": 0.00014172140324095837,
+      "loss": 1.4392,
+      "step": 7174
+    },
+    {
+      "epoch": 1.148,
+      "grad_norm": 2.342036602938565,
+      "learning_rate": 0.0001417057040132655,
+      "loss": 1.4737,
+      "step": 7175
+    },
+    {
+      "epoch": 1.14816,
+      "grad_norm": 2.3848545970266,
+      "learning_rate": 0.00014169000354113517,
+      "loss": 1.3646,
+      "step": 7176
+    },
+    {
+      "epoch": 1.14832,
+      "grad_norm": 3.0323896110660216,
+      "learning_rate": 0.00014167430182503586,
+      "loss": 1.3482,
+      "step": 7177
+    },
+    {
+      "epoch": 1.14848,
+      "grad_norm": 1.971814100116116,
+      "learning_rate": 0.0001416585988654361,
+      "loss": 1.4571,
+      "step": 7178
+    },
+    {
+      "epoch": 1.1486399999999999,
+      "grad_norm": 2.1589080050396317,
+      "learning_rate": 0.00014164289466280446,
+      "loss": 1.4138,
+      "step": 7179
+    },
+    {
+      "epoch": 1.1488,
+      "grad_norm": 3.2792652314727184,
+      "learning_rate": 0.0001416271892176096,
+      "loss": 1.4451,
+      "step": 7180
+    },
+    {
+      "epoch": 1.14896,
+      "grad_norm": 2.1899831078140743,
+      "learning_rate": 0.00014161148253032,
+      "loss": 1.4283,
+      "step": 7181
+    },
+    {
+      "epoch": 1.14912,
+      "grad_norm": 2.595579764571652,
+      "learning_rate": 0.0001415957746014044,
+      "loss": 1.3645,
+      "step": 7182
+    },
+    {
+      "epoch": 1.14928,
+      "grad_norm": 2.8845949439067358,
+      "learning_rate": 0.0001415800654313315,
+      "loss": 1.454,
+      "step": 7183
+    },
+    {
+      "epoch": 1.14944,
+      "grad_norm": 2.667308494368492,
+      "learning_rate": 0.00014156435502057005,
+      "loss": 1.3978,
+      "step": 7184
+    },
+    {
+      "epoch": 1.1496,
+      "grad_norm": 3.098561658896646,
+      "learning_rate": 0.0001415486433695888,
+      "loss": 1.4288,
+      "step": 7185
+    },
+    {
+      "epoch": 1.1497600000000001,
+      "grad_norm": 2.475329902489417,
+      "learning_rate": 0.00014153293047885658,
+      "loss": 1.4062,
+      "step": 7186
+    },
+    {
+      "epoch": 1.14992,
+      "grad_norm": 2.307833891880242,
+      "learning_rate": 0.0001415172163488422,
+      "loss": 1.3593,
+      "step": 7187
+    },
+    {
+      "epoch": 1.15008,
+      "grad_norm": 2.612532310974586,
+      "learning_rate": 0.00014150150098001465,
+      "loss": 1.4662,
+      "step": 7188
+    },
+    {
+      "epoch": 1.15024,
+      "grad_norm": 2.0492810000750596,
+      "learning_rate": 0.00014148578437284275,
+      "loss": 1.4933,
+      "step": 7189
+    },
+    {
+      "epoch": 1.1504,
+      "grad_norm": 2.740028738991766,
+      "learning_rate": 0.0001414700665277955,
+      "loss": 1.3293,
+      "step": 7190
+    },
+    {
+      "epoch": 1.15056,
+      "grad_norm": 2.206268204975871,
+      "learning_rate": 0.00014145434744534187,
+      "loss": 1.4617,
+      "step": 7191
+    },
+    {
+      "epoch": 1.15072,
+      "grad_norm": 2.34991541819347,
+      "learning_rate": 0.00014143862712595094,
+      "loss": 1.4291,
+      "step": 7192
+    },
+    {
+      "epoch": 1.15088,
+      "grad_norm": 3.3207238645308963,
+      "learning_rate": 0.00014142290557009175,
+      "loss": 1.4427,
+      "step": 7193
+    },
+    {
+      "epoch": 1.15104,
+      "grad_norm": 2.156656184835171,
+      "learning_rate": 0.00014140718277823344,
+      "loss": 1.4313,
+      "step": 7194
+    },
+    {
+      "epoch": 1.1512,
+      "grad_norm": 3.0258520234643402,
+      "learning_rate": 0.0001413914587508451,
+      "loss": 1.4114,
+      "step": 7195
+    },
+    {
+      "epoch": 1.15136,
+      "grad_norm": 2.49418451900046,
+      "learning_rate": 0.00014137573348839595,
+      "loss": 1.4632,
+      "step": 7196
+    },
+    {
+      "epoch": 1.15152,
+      "grad_norm": 2.7768561608429527,
+      "learning_rate": 0.0001413600069913552,
+      "loss": 1.4283,
+      "step": 7197
+    },
+    {
+      "epoch": 1.15168,
+      "grad_norm": 3.3693368582660415,
+      "learning_rate": 0.00014134427926019213,
+      "loss": 1.3948,
+      "step": 7198
+    },
+    {
+      "epoch": 1.15184,
+      "grad_norm": 2.217550813476463,
+      "learning_rate": 0.00014132855029537603,
+      "loss": 1.5019,
+      "step": 7199
+    },
+    {
+      "epoch": 1.152,
+      "grad_norm": 2.7435848399751768,
+      "learning_rate": 0.00014131282009737618,
+      "loss": 1.4374,
+      "step": 7200
+    },
+    {
+      "epoch": 1.15216,
+      "grad_norm": 2.2420449830430496,
+      "learning_rate": 0.00014129708866666198,
+      "loss": 1.4066,
+      "step": 7201
+    },
+    {
+      "epoch": 1.15232,
+      "grad_norm": 2.09958741653404,
+      "learning_rate": 0.00014128135600370283,
+      "loss": 1.3774,
+      "step": 7202
+    },
+    {
+      "epoch": 1.15248,
+      "grad_norm": 2.814815596261177,
+      "learning_rate": 0.00014126562210896814,
+      "loss": 1.45,
+      "step": 7203
+    },
+    {
+      "epoch": 1.1526399999999999,
+      "grad_norm": 2.125360956821451,
+      "learning_rate": 0.00014124988698292748,
+      "loss": 1.506,
+      "step": 7204
+    },
+    {
+      "epoch": 1.1528,
+      "grad_norm": 2.323977578541007,
+      "learning_rate": 0.00014123415062605024,
+      "loss": 1.4266,
+      "step": 7205
+    },
+    {
+      "epoch": 1.15296,
+      "grad_norm": 2.0515769957433556,
+      "learning_rate": 0.00014121841303880607,
+      "loss": 1.3851,
+      "step": 7206
+    },
+    {
+      "epoch": 1.15312,
+      "grad_norm": 2.069107173325222,
+      "learning_rate": 0.00014120267422166452,
+      "loss": 1.4618,
+      "step": 7207
+    },
+    {
+      "epoch": 1.15328,
+      "grad_norm": 3.021985073752878,
+      "learning_rate": 0.00014118693417509518,
+      "loss": 1.47,
+      "step": 7208
+    },
+    {
+      "epoch": 1.15344,
+      "grad_norm": 2.313352114215398,
+      "learning_rate": 0.00014117119289956776,
+      "loss": 1.4398,
+      "step": 7209
+    },
+    {
+      "epoch": 1.1536,
+      "grad_norm": 2.6859709969292473,
+      "learning_rate": 0.00014115545039555195,
+      "loss": 1.5208,
+      "step": 7210
+    },
+    {
+      "epoch": 1.1537600000000001,
+      "grad_norm": 2.417989557660861,
+      "learning_rate": 0.00014113970666351745,
+      "loss": 1.4548,
+      "step": 7211
+    },
+    {
+      "epoch": 1.15392,
+      "grad_norm": 2.2375856777737395,
+      "learning_rate": 0.00014112396170393401,
+      "loss": 1.4741,
+      "step": 7212
+    },
+    {
+      "epoch": 1.15408,
+      "grad_norm": 2.918403811256326,
+      "learning_rate": 0.00014110821551727151,
+      "loss": 1.4343,
+      "step": 7213
+    },
+    {
+      "epoch": 1.15424,
+      "grad_norm": 2.2695331821745297,
+      "learning_rate": 0.0001410924681039998,
+      "loss": 1.4023,
+      "step": 7214
+    },
+    {
+      "epoch": 1.1544,
+      "grad_norm": 2.180484050757881,
+      "learning_rate": 0.00014107671946458867,
+      "loss": 1.4438,
+      "step": 7215
+    },
+    {
+      "epoch": 1.15456,
+      "grad_norm": 1.9447685284411367,
+      "learning_rate": 0.0001410609695995081,
+      "loss": 1.4113,
+      "step": 7216
+    },
+    {
+      "epoch": 1.15472,
+      "grad_norm": 2.2564711583697643,
+      "learning_rate": 0.00014104521850922806,
+      "loss": 1.4633,
+      "step": 7217
+    },
+    {
+      "epoch": 1.15488,
+      "grad_norm": 1.8663396044198572,
+      "learning_rate": 0.00014102946619421846,
+      "loss": 1.4094,
+      "step": 7218
+    },
+    {
+      "epoch": 1.15504,
+      "grad_norm": 2.13916965064613,
+      "learning_rate": 0.00014101371265494943,
+      "loss": 1.4745,
+      "step": 7219
+    },
+    {
+      "epoch": 1.1552,
+      "grad_norm": 2.32550769772727,
+      "learning_rate": 0.0001409979578918909,
+      "loss": 1.4151,
+      "step": 7220
+    },
+    {
+      "epoch": 1.15536,
+      "grad_norm": 2.171145265017346,
+      "learning_rate": 0.0001409822019055131,
+      "loss": 1.4823,
+      "step": 7221
+    },
+    {
+      "epoch": 1.15552,
+      "grad_norm": 2.1542886365604796,
+      "learning_rate": 0.00014096644469628613,
+      "loss": 1.5579,
+      "step": 7222
+    },
+    {
+      "epoch": 1.15568,
+      "grad_norm": 2.1402341961936733,
+      "learning_rate": 0.0001409506862646801,
+      "loss": 1.4192,
+      "step": 7223
+    },
+    {
+      "epoch": 1.15584,
+      "grad_norm": 2.157577419233117,
+      "learning_rate": 0.0001409349266111653,
+      "loss": 1.4627,
+      "step": 7224
+    },
+    {
+      "epoch": 1.156,
+      "grad_norm": 2.1332257522940856,
+      "learning_rate": 0.0001409191657362119,
+      "loss": 1.4226,
+      "step": 7225
+    },
+    {
+      "epoch": 1.15616,
+      "grad_norm": 2.256213037472151,
+      "learning_rate": 0.00014090340364029028,
+      "loss": 1.402,
+      "step": 7226
+    },
+    {
+      "epoch": 1.15632,
+      "grad_norm": 2.4940928661927635,
+      "learning_rate": 0.00014088764032387064,
+      "loss": 1.4874,
+      "step": 7227
+    },
+    {
+      "epoch": 1.15648,
+      "grad_norm": 2.406796221962914,
+      "learning_rate": 0.0001408718757874234,
+      "loss": 1.4863,
+      "step": 7228
+    },
+    {
+      "epoch": 1.15664,
+      "grad_norm": 2.01545736511636,
+      "learning_rate": 0.00014085611003141897,
+      "loss": 1.4588,
+      "step": 7229
+    },
+    {
+      "epoch": 1.1568,
+      "grad_norm": 2.3509093615630876,
+      "learning_rate": 0.00014084034305632772,
+      "loss": 1.4099,
+      "step": 7230
+    },
+    {
+      "epoch": 1.15696,
+      "grad_norm": 2.314103622406891,
+      "learning_rate": 0.00014082457486262016,
+      "loss": 1.4064,
+      "step": 7231
+    },
+    {
+      "epoch": 1.15712,
+      "grad_norm": 2.5098451604792347,
+      "learning_rate": 0.0001408088054507668,
+      "loss": 1.4346,
+      "step": 7232
+    },
+    {
+      "epoch": 1.15728,
+      "grad_norm": 2.213645273162562,
+      "learning_rate": 0.0001407930348212381,
+      "loss": 1.5037,
+      "step": 7233
+    },
+    {
+      "epoch": 1.15744,
+      "grad_norm": 2.4210464451457914,
+      "learning_rate": 0.00014077726297450472,
+      "loss": 1.4246,
+      "step": 7234
+    },
+    {
+      "epoch": 1.1576,
+      "grad_norm": 1.8988920316246716,
+      "learning_rate": 0.0001407614899110372,
+      "loss": 1.45,
+      "step": 7235
+    },
+    {
+      "epoch": 1.1577600000000001,
+      "grad_norm": 2.368977038655147,
+      "learning_rate": 0.00014074571563130623,
+      "loss": 1.3829,
+      "step": 7236
+    },
+    {
+      "epoch": 1.15792,
+      "grad_norm": 2.10680758435386,
+      "learning_rate": 0.00014072994013578248,
+      "loss": 1.528,
+      "step": 7237
+    },
+    {
+      "epoch": 1.15808,
+      "grad_norm": 2.3980110243290427,
+      "learning_rate": 0.00014071416342493665,
+      "loss": 1.4866,
+      "step": 7238
+    },
+    {
+      "epoch": 1.15824,
+      "grad_norm": 2.2425629551767425,
+      "learning_rate": 0.00014069838549923955,
+      "loss": 1.445,
+      "step": 7239
+    },
+    {
+      "epoch": 1.1584,
+      "grad_norm": 2.374312677218643,
+      "learning_rate": 0.0001406826063591619,
+      "loss": 1.425,
+      "step": 7240
+    },
+    {
+      "epoch": 1.15856,
+      "grad_norm": 2.1069446817594755,
+      "learning_rate": 0.0001406668260051746,
+      "loss": 1.441,
+      "step": 7241
+    },
+    {
+      "epoch": 1.15872,
+      "grad_norm": 2.049448357795199,
+      "learning_rate": 0.00014065104443774842,
+      "loss": 1.3861,
+      "step": 7242
+    },
+    {
+      "epoch": 1.15888,
+      "grad_norm": 2.336860556332172,
+      "learning_rate": 0.00014063526165735432,
+      "loss": 1.4033,
+      "step": 7243
+    },
+    {
+      "epoch": 1.15904,
+      "grad_norm": 2.0084709227545208,
+      "learning_rate": 0.00014061947766446322,
+      "loss": 1.4318,
+      "step": 7244
+    },
+    {
+      "epoch": 1.1592,
+      "grad_norm": 1.9838085854831042,
+      "learning_rate": 0.0001406036924595461,
+      "loss": 1.3944,
+      "step": 7245
+    },
+    {
+      "epoch": 1.15936,
+      "grad_norm": 2.1677611164593866,
+      "learning_rate": 0.00014058790604307397,
+      "loss": 1.4127,
+      "step": 7246
+    },
+    {
+      "epoch": 1.15952,
+      "grad_norm": 2.3699109653862953,
+      "learning_rate": 0.00014057211841551786,
+      "loss": 1.371,
+      "step": 7247
+    },
+    {
+      "epoch": 1.15968,
+      "grad_norm": 2.3111526104074978,
+      "learning_rate": 0.00014055632957734888,
+      "loss": 1.468,
+      "step": 7248
+    },
+    {
+      "epoch": 1.15984,
+      "grad_norm": 2.1414197722030437,
+      "learning_rate": 0.0001405405395290381,
+      "loss": 1.4023,
+      "step": 7249
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 2.4307280218486356,
+      "learning_rate": 0.0001405247482710567,
+      "loss": 1.4048,
+      "step": 7250
+    },
+    {
+      "epoch": 1.16016,
+      "grad_norm": 2.4966817608968093,
+      "learning_rate": 0.00014050895580387587,
+      "loss": 1.4406,
+      "step": 7251
+    },
+    {
+      "epoch": 1.16032,
+      "grad_norm": 2.527028810837283,
+      "learning_rate": 0.00014049316212796676,
+      "loss": 1.5158,
+      "step": 7252
+    },
+    {
+      "epoch": 1.16048,
+      "grad_norm": 2.472862594129977,
+      "learning_rate": 0.00014047736724380076,
+      "loss": 1.4026,
+      "step": 7253
+    },
+    {
+      "epoch": 1.16064,
+      "grad_norm": 2.751899548046099,
+      "learning_rate": 0.0001404615711518491,
+      "loss": 1.4446,
+      "step": 7254
+    },
+    {
+      "epoch": 1.1608,
+      "grad_norm": 2.3292377911170314,
+      "learning_rate": 0.00014044577385258312,
+      "loss": 1.4301,
+      "step": 7255
+    },
+    {
+      "epoch": 1.16096,
+      "grad_norm": 2.1668211399026656,
+      "learning_rate": 0.00014042997534647418,
+      "loss": 1.3895,
+      "step": 7256
+    },
+    {
+      "epoch": 1.16112,
+      "grad_norm": 2.580767994559106,
+      "learning_rate": 0.00014041417563399368,
+      "loss": 1.4458,
+      "step": 7257
+    },
+    {
+      "epoch": 1.16128,
+      "grad_norm": 2.174814923228735,
+      "learning_rate": 0.00014039837471561308,
+      "loss": 1.4218,
+      "step": 7258
+    },
+    {
+      "epoch": 1.16144,
+      "grad_norm": 1.6967827310408,
+      "learning_rate": 0.00014038257259180384,
+      "loss": 1.4101,
+      "step": 7259
+    },
+    {
+      "epoch": 1.1616,
+      "grad_norm": 2.116393420653193,
+      "learning_rate": 0.00014036676926303744,
+      "loss": 1.4544,
+      "step": 7260
+    },
+    {
+      "epoch": 1.16176,
+      "grad_norm": 2.4874122941506975,
+      "learning_rate": 0.00014035096472978553,
+      "loss": 1.4827,
+      "step": 7261
+    },
+    {
+      "epoch": 1.16192,
+      "grad_norm": 2.385207913740133,
+      "learning_rate": 0.0001403351589925196,
+      "loss": 1.3876,
+      "step": 7262
+    },
+    {
+      "epoch": 1.16208,
+      "grad_norm": 2.1622767785559316,
+      "learning_rate": 0.0001403193520517113,
+      "loss": 1.4713,
+      "step": 7263
+    },
+    {
+      "epoch": 1.16224,
+      "grad_norm": 2.300901720609164,
+      "learning_rate": 0.00014030354390783228,
+      "loss": 1.4399,
+      "step": 7264
+    },
+    {
+      "epoch": 1.1623999999999999,
+      "grad_norm": 2.648940312987342,
+      "learning_rate": 0.00014028773456135424,
+      "loss": 1.4656,
+      "step": 7265
+    },
+    {
+      "epoch": 1.16256,
+      "grad_norm": 2.204095557158803,
+      "learning_rate": 0.00014027192401274894,
+      "loss": 1.4377,
+      "step": 7266
+    },
+    {
+      "epoch": 1.16272,
+      "grad_norm": 2.027514736377657,
+      "learning_rate": 0.00014025611226248808,
+      "loss": 1.4522,
+      "step": 7267
+    },
+    {
+      "epoch": 1.16288,
+      "grad_norm": 2.153683416574822,
+      "learning_rate": 0.00014024029931104345,
+      "loss": 1.4301,
+      "step": 7268
+    },
+    {
+      "epoch": 1.16304,
+      "grad_norm": 1.9153459384896445,
+      "learning_rate": 0.000140224485158887,
+      "loss": 1.4401,
+      "step": 7269
+    },
+    {
+      "epoch": 1.1632,
+      "grad_norm": 2.794435023834097,
+      "learning_rate": 0.00014020866980649045,
+      "loss": 1.4653,
+      "step": 7270
+    },
+    {
+      "epoch": 1.16336,
+      "grad_norm": 2.0391045141975743,
+      "learning_rate": 0.00014019285325432584,
+      "loss": 1.4227,
+      "step": 7271
+    },
+    {
+      "epoch": 1.16352,
+      "grad_norm": 2.8260532655909367,
+      "learning_rate": 0.00014017703550286503,
+      "loss": 1.3671,
+      "step": 7272
+    },
+    {
+      "epoch": 1.16368,
+      "grad_norm": 2.594497718164233,
+      "learning_rate": 0.00014016121655258004,
+      "loss": 1.5319,
+      "step": 7273
+    },
+    {
+      "epoch": 1.16384,
+      "grad_norm": 3.271367516165277,
+      "learning_rate": 0.00014014539640394287,
+      "loss": 1.6062,
+      "step": 7274
+    },
+    {
+      "epoch": 1.164,
+      "grad_norm": 2.6169853148031885,
+      "learning_rate": 0.00014012957505742557,
+      "loss": 1.4708,
+      "step": 7275
+    },
+    {
+      "epoch": 1.16416,
+      "grad_norm": 2.652276620867743,
+      "learning_rate": 0.0001401137525135002,
+      "loss": 1.463,
+      "step": 7276
+    },
+    {
+      "epoch": 1.16432,
+      "grad_norm": 2.533961222167817,
+      "learning_rate": 0.00014009792877263895,
+      "loss": 1.4559,
+      "step": 7277
+    },
+    {
+      "epoch": 1.16448,
+      "grad_norm": 1.9794209105737848,
+      "learning_rate": 0.0001400821038353139,
+      "loss": 1.4302,
+      "step": 7278
+    },
+    {
+      "epoch": 1.16464,
+      "grad_norm": 2.733082967803239,
+      "learning_rate": 0.00014006627770199727,
+      "loss": 1.3944,
+      "step": 7279
+    },
+    {
+      "epoch": 1.1648,
+      "grad_norm": 2.259154441413576,
+      "learning_rate": 0.0001400504503731613,
+      "loss": 1.459,
+      "step": 7280
+    },
+    {
+      "epoch": 1.16496,
+      "grad_norm": 2.582554380556922,
+      "learning_rate": 0.0001400346218492783,
+      "loss": 1.491,
+      "step": 7281
+    },
+    {
+      "epoch": 1.16512,
+      "grad_norm": 2.264493786040898,
+      "learning_rate": 0.00014001879213082046,
+      "loss": 1.4375,
+      "step": 7282
+    },
+    {
+      "epoch": 1.16528,
+      "grad_norm": 2.0562751258575416,
+      "learning_rate": 0.00014000296121826023,
+      "loss": 1.4039,
+      "step": 7283
+    },
+    {
+      "epoch": 1.16544,
+      "grad_norm": 2.4757565859002137,
+      "learning_rate": 0.00013998712911206989,
+      "loss": 1.4204,
+      "step": 7284
+    },
+    {
+      "epoch": 1.1656,
+      "grad_norm": 2.1891639195958144,
+      "learning_rate": 0.0001399712958127219,
+      "loss": 1.4241,
+      "step": 7285
+    },
+    {
+      "epoch": 1.16576,
+      "grad_norm": 2.1656385517917682,
+      "learning_rate": 0.00013995546132068866,
+      "loss": 1.4346,
+      "step": 7286
+    },
+    {
+      "epoch": 1.16592,
+      "grad_norm": 2.240479380096775,
+      "learning_rate": 0.0001399396256364427,
+      "loss": 1.4919,
+      "step": 7287
+    },
+    {
+      "epoch": 1.16608,
+      "grad_norm": 2.192612525985976,
+      "learning_rate": 0.00013992378876045652,
+      "loss": 1.4463,
+      "step": 7288
+    },
+    {
+      "epoch": 1.16624,
+      "grad_norm": 1.9910237591264661,
+      "learning_rate": 0.00013990795069320264,
+      "loss": 1.3495,
+      "step": 7289
+    },
+    {
+      "epoch": 1.1663999999999999,
+      "grad_norm": 2.3397813274443613,
+      "learning_rate": 0.00013989211143515366,
+      "loss": 1.4195,
+      "step": 7290
+    },
+    {
+      "epoch": 1.16656,
+      "grad_norm": 2.1307486590159677,
+      "learning_rate": 0.0001398762709867822,
+      "loss": 1.4762,
+      "step": 7291
+    },
+    {
+      "epoch": 1.16672,
+      "grad_norm": 2.408278047175154,
+      "learning_rate": 0.00013986042934856094,
+      "loss": 1.457,
+      "step": 7292
+    },
+    {
+      "epoch": 1.16688,
+      "grad_norm": 2.214187932662291,
+      "learning_rate": 0.00013984458652096255,
+      "loss": 1.4332,
+      "step": 7293
+    },
+    {
+      "epoch": 1.16704,
+      "grad_norm": 1.9873603704993812,
+      "learning_rate": 0.00013982874250445973,
+      "loss": 1.3992,
+      "step": 7294
+    },
+    {
+      "epoch": 1.1672,
+      "grad_norm": 2.1495916772331065,
+      "learning_rate": 0.00013981289729952533,
+      "loss": 1.4086,
+      "step": 7295
+    },
+    {
+      "epoch": 1.16736,
+      "grad_norm": 2.586555762402404,
+      "learning_rate": 0.000139797050906632,
+      "loss": 1.4227,
+      "step": 7296
+    },
+    {
+      "epoch": 1.1675200000000001,
+      "grad_norm": 2.0673061140544924,
+      "learning_rate": 0.00013978120332625273,
+      "loss": 1.4503,
+      "step": 7297
+    },
+    {
+      "epoch": 1.16768,
+      "grad_norm": 2.4274800262593335,
+      "learning_rate": 0.0001397653545588603,
+      "loss": 1.3624,
+      "step": 7298
+    },
+    {
+      "epoch": 1.16784,
+      "grad_norm": 2.1815426071486366,
+      "learning_rate": 0.00013974950460492763,
+      "loss": 1.3669,
+      "step": 7299
+    },
+    {
+      "epoch": 1.168,
+      "grad_norm": 2.204545551443369,
+      "learning_rate": 0.0001397336534649277,
+      "loss": 1.3808,
+      "step": 7300
+    },
+    {
+      "epoch": 1.16816,
+      "grad_norm": 2.014195468000615,
+      "learning_rate": 0.00013971780113933337,
+      "loss": 1.3883,
+      "step": 7301
+    },
+    {
+      "epoch": 1.16832,
+      "grad_norm": 1.977034651812557,
+      "learning_rate": 0.0001397019476286178,
+      "loss": 1.389,
+      "step": 7302
+    },
+    {
+      "epoch": 1.16848,
+      "grad_norm": 2.160037045693847,
+      "learning_rate": 0.00013968609293325392,
+      "loss": 1.4336,
+      "step": 7303
+    },
+    {
+      "epoch": 1.16864,
+      "grad_norm": 2.1704743211230255,
+      "learning_rate": 0.0001396702370537149,
+      "loss": 1.6134,
+      "step": 7304
+    },
+    {
+      "epoch": 1.1688,
+      "grad_norm": 2.48833987557712,
+      "learning_rate": 0.00013965437999047378,
+      "loss": 1.4001,
+      "step": 7305
+    },
+    {
+      "epoch": 1.16896,
+      "grad_norm": 2.1169268679270683,
+      "learning_rate": 0.00013963852174400375,
+      "loss": 1.4546,
+      "step": 7306
+    },
+    {
+      "epoch": 1.16912,
+      "grad_norm": 2.1044052607316757,
+      "learning_rate": 0.000139622662314778,
+      "loss": 1.4264,
+      "step": 7307
+    },
+    {
+      "epoch": 1.16928,
+      "grad_norm": 2.835620268318194,
+      "learning_rate": 0.0001396068017032697,
+      "loss": 1.5335,
+      "step": 7308
+    },
+    {
+      "epoch": 1.16944,
+      "grad_norm": 2.7983971422285263,
+      "learning_rate": 0.0001395909399099522,
+      "loss": 1.4013,
+      "step": 7309
+    },
+    {
+      "epoch": 1.1696,
+      "grad_norm": 1.8830559743634983,
+      "learning_rate": 0.00013957507693529873,
+      "loss": 1.4875,
+      "step": 7310
+    },
+    {
+      "epoch": 1.16976,
+      "grad_norm": 3.5093939017698683,
+      "learning_rate": 0.00013955921277978263,
+      "loss": 1.4456,
+      "step": 7311
+    },
+    {
+      "epoch": 1.16992,
+      "grad_norm": 2.3777363583339355,
+      "learning_rate": 0.00013954334744387727,
+      "loss": 1.4442,
+      "step": 7312
+    },
+    {
+      "epoch": 1.17008,
+      "grad_norm": 4.010913929781316,
+      "learning_rate": 0.00013952748092805603,
+      "loss": 1.4679,
+      "step": 7313
+    },
+    {
+      "epoch": 1.17024,
+      "grad_norm": 2.2690171155145245,
+      "learning_rate": 0.00013951161323279236,
+      "loss": 1.3618,
+      "step": 7314
+    },
+    {
+      "epoch": 1.1703999999999999,
+      "grad_norm": 2.7929207839068986,
+      "learning_rate": 0.00013949574435855974,
+      "loss": 1.4384,
+      "step": 7315
+    },
+    {
+      "epoch": 1.17056,
+      "grad_norm": 2.9424686998877685,
+      "learning_rate": 0.00013947987430583164,
+      "loss": 1.4768,
+      "step": 7316
+    },
+    {
+      "epoch": 1.17072,
+      "grad_norm": 2.0689240964724007,
+      "learning_rate": 0.00013946400307508162,
+      "loss": 1.3804,
+      "step": 7317
+    },
+    {
+      "epoch": 1.17088,
+      "grad_norm": 3.8605678860463057,
+      "learning_rate": 0.00013944813066678327,
+      "loss": 1.4341,
+      "step": 7318
+    },
+    {
+      "epoch": 1.17104,
+      "grad_norm": 1.9359110844095622,
+      "learning_rate": 0.00013943225708141016,
+      "loss": 1.4011,
+      "step": 7319
+    },
+    {
+      "epoch": 1.1712,
+      "grad_norm": 3.8304904583604045,
+      "learning_rate": 0.000139416382319436,
+      "loss": 1.4746,
+      "step": 7320
+    },
+    {
+      "epoch": 1.17136,
+      "grad_norm": 2.8143937737316604,
+      "learning_rate": 0.0001394005063813344,
+      "loss": 1.4213,
+      "step": 7321
+    },
+    {
+      "epoch": 1.1715200000000001,
+      "grad_norm": 2.1584102634846154,
+      "learning_rate": 0.0001393846292675791,
+      "loss": 1.4561,
+      "step": 7322
+    },
+    {
+      "epoch": 1.17168,
+      "grad_norm": 3.651326912073475,
+      "learning_rate": 0.00013936875097864387,
+      "loss": 1.4543,
+      "step": 7323
+    },
+    {
+      "epoch": 1.17184,
+      "grad_norm": 2.392292968704192,
+      "learning_rate": 0.00013935287151500245,
+      "loss": 1.4495,
+      "step": 7324
+    },
+    {
+      "epoch": 1.172,
+      "grad_norm": 3.279168946293212,
+      "learning_rate": 0.00013933699087712866,
+      "loss": 1.4594,
+      "step": 7325
+    },
+    {
+      "epoch": 1.17216,
+      "grad_norm": 3.494651766642704,
+      "learning_rate": 0.00013932110906549644,
+      "loss": 1.4789,
+      "step": 7326
+    },
+    {
+      "epoch": 1.17232,
+      "grad_norm": 2.0815180251123384,
+      "learning_rate": 0.0001393052260805796,
+      "loss": 1.4271,
+      "step": 7327
+    },
+    {
+      "epoch": 1.17248,
+      "grad_norm": 3.357296630198489,
+      "learning_rate": 0.00013928934192285208,
+      "loss": 1.4102,
+      "step": 7328
+    },
+    {
+      "epoch": 1.17264,
+      "grad_norm": 2.6428002052840047,
+      "learning_rate": 0.00013927345659278786,
+      "loss": 1.4143,
+      "step": 7329
+    },
+    {
+      "epoch": 1.1728,
+      "grad_norm": 3.0059618883803814,
+      "learning_rate": 0.0001392575700908609,
+      "loss": 1.3633,
+      "step": 7330
+    },
+    {
+      "epoch": 1.17296,
+      "grad_norm": 3.144616439395455,
+      "learning_rate": 0.0001392416824175453,
+      "loss": 1.4221,
+      "step": 7331
+    },
+    {
+      "epoch": 1.17312,
+      "grad_norm": 2.501461962923107,
+      "learning_rate": 0.00013922579357331504,
+      "loss": 1.4952,
+      "step": 7332
+    },
+    {
+      "epoch": 1.17328,
+      "grad_norm": 3.439596820402793,
+      "learning_rate": 0.00013920990355864428,
+      "loss": 1.4276,
+      "step": 7333
+    },
+    {
+      "epoch": 1.17344,
+      "grad_norm": 3.1560965555277094,
+      "learning_rate": 0.0001391940123740071,
+      "loss": 1.5156,
+      "step": 7334
+    },
+    {
+      "epoch": 1.1736,
+      "grad_norm": 2.16983739823021,
+      "learning_rate": 0.00013917812001987772,
+      "loss": 1.5529,
+      "step": 7335
+    },
+    {
+      "epoch": 1.17376,
+      "grad_norm": 2.55325140745831,
+      "learning_rate": 0.00013916222649673035,
+      "loss": 1.4527,
+      "step": 7336
+    },
+    {
+      "epoch": 1.17392,
+      "grad_norm": 2.142637597982546,
+      "learning_rate": 0.00013914633180503919,
+      "loss": 1.3869,
+      "step": 7337
+    },
+    {
+      "epoch": 1.17408,
+      "grad_norm": 2.2786780968827176,
+      "learning_rate": 0.00013913043594527852,
+      "loss": 1.4728,
+      "step": 7338
+    },
+    {
+      "epoch": 1.17424,
+      "grad_norm": 2.5885555608890773,
+      "learning_rate": 0.00013911453891792268,
+      "loss": 1.4088,
+      "step": 7339
+    },
+    {
+      "epoch": 1.1743999999999999,
+      "grad_norm": 2.320957632601235,
+      "learning_rate": 0.000139098640723446,
+      "loss": 1.4471,
+      "step": 7340
+    },
+    {
+      "epoch": 1.17456,
+      "grad_norm": 2.519480905796864,
+      "learning_rate": 0.00013908274136232283,
+      "loss": 1.397,
+      "step": 7341
+    },
+    {
+      "epoch": 1.17472,
+      "grad_norm": 2.3409222780114156,
+      "learning_rate": 0.0001390668408350276,
+      "loss": 1.3788,
+      "step": 7342
+    },
+    {
+      "epoch": 1.17488,
+      "grad_norm": 2.1641605317726778,
+      "learning_rate": 0.00013905093914203477,
+      "loss": 1.4183,
+      "step": 7343
+    },
+    {
+      "epoch": 1.17504,
+      "grad_norm": 2.3637630205324793,
+      "learning_rate": 0.00013903503628381883,
+      "loss": 1.3549,
+      "step": 7344
+    },
+    {
+      "epoch": 1.1752,
+      "grad_norm": 2.83964769520336,
+      "learning_rate": 0.00013901913226085428,
+      "loss": 1.3821,
+      "step": 7345
+    },
+    {
+      "epoch": 1.17536,
+      "grad_norm": 2.8218327540621564,
+      "learning_rate": 0.0001390032270736157,
+      "loss": 1.4721,
+      "step": 7346
+    },
+    {
+      "epoch": 1.1755200000000001,
+      "grad_norm": 2.4559275076251903,
+      "learning_rate": 0.00013898732072257765,
+      "loss": 1.4352,
+      "step": 7347
+    },
+    {
+      "epoch": 1.17568,
+      "grad_norm": 1.925654342761668,
+      "learning_rate": 0.00013897141320821476,
+      "loss": 1.3423,
+      "step": 7348
+    },
+    {
+      "epoch": 1.17584,
+      "grad_norm": 2.293830848756551,
+      "learning_rate": 0.00013895550453100168,
+      "loss": 1.4628,
+      "step": 7349
+    },
+    {
+      "epoch": 1.176,
+      "grad_norm": 2.0348041083060417,
+      "learning_rate": 0.00013893959469141312,
+      "loss": 1.3652,
+      "step": 7350
+    },
+    {
+      "epoch": 1.17616,
+      "grad_norm": 2.052281518826001,
+      "learning_rate": 0.00013892368368992376,
+      "loss": 1.3537,
+      "step": 7351
+    },
+    {
+      "epoch": 1.17632,
+      "grad_norm": 2.9548504904964656,
+      "learning_rate": 0.00013890777152700842,
+      "loss": 1.543,
+      "step": 7352
+    },
+    {
+      "epoch": 1.17648,
+      "grad_norm": 2.351439596577609,
+      "learning_rate": 0.0001388918582031419,
+      "loss": 1.4174,
+      "step": 7353
+    },
+    {
+      "epoch": 1.17664,
+      "grad_norm": 2.48665548792425,
+      "learning_rate": 0.00013887594371879896,
+      "loss": 1.4378,
+      "step": 7354
+    },
+    {
+      "epoch": 1.1768,
+      "grad_norm": 2.8599820857888467,
+      "learning_rate": 0.00013886002807445453,
+      "loss": 1.4489,
+      "step": 7355
+    },
+    {
+      "epoch": 1.17696,
+      "grad_norm": 2.4216282381633314,
+      "learning_rate": 0.0001388441112705835,
+      "loss": 1.4743,
+      "step": 7356
+    },
+    {
+      "epoch": 1.17712,
+      "grad_norm": 2.424913894342917,
+      "learning_rate": 0.00013882819330766078,
+      "loss": 1.4251,
+      "step": 7357
+    },
+    {
+      "epoch": 1.17728,
+      "grad_norm": 3.13102624888944,
+      "learning_rate": 0.00013881227418616134,
+      "loss": 1.4247,
+      "step": 7358
+    },
+    {
+      "epoch": 1.17744,
+      "grad_norm": 2.2489775275502057,
+      "learning_rate": 0.00013879635390656018,
+      "loss": 1.4451,
+      "step": 7359
+    },
+    {
+      "epoch": 1.1776,
+      "grad_norm": 2.054359595029863,
+      "learning_rate": 0.00013878043246933237,
+      "loss": 1.4283,
+      "step": 7360
+    },
+    {
+      "epoch": 1.17776,
+      "grad_norm": 2.630438328766721,
+      "learning_rate": 0.00013876450987495298,
+      "loss": 1.3439,
+      "step": 7361
+    },
+    {
+      "epoch": 1.17792,
+      "grad_norm": 2.25239564431588,
+      "learning_rate": 0.00013874858612389706,
+      "loss": 1.4377,
+      "step": 7362
+    },
+    {
+      "epoch": 1.17808,
+      "grad_norm": 2.3966130964687884,
+      "learning_rate": 0.00013873266121663984,
+      "loss": 1.4582,
+      "step": 7363
+    },
+    {
+      "epoch": 1.17824,
+      "grad_norm": 2.2808228355589812,
+      "learning_rate": 0.00013871673515365642,
+      "loss": 1.4344,
+      "step": 7364
+    },
+    {
+      "epoch": 1.1784,
+      "grad_norm": 2.125683500110162,
+      "learning_rate": 0.00013870080793542205,
+      "loss": 1.4561,
+      "step": 7365
+    },
+    {
+      "epoch": 1.17856,
+      "grad_norm": 2.214215991197151,
+      "learning_rate": 0.00013868487956241197,
+      "loss": 1.5121,
+      "step": 7366
+    },
+    {
+      "epoch": 1.17872,
+      "grad_norm": 2.068968155347034,
+      "learning_rate": 0.00013866895003510142,
+      "loss": 1.4482,
+      "step": 7367
+    },
+    {
+      "epoch": 1.17888,
+      "grad_norm": 2.2494508556458537,
+      "learning_rate": 0.00013865301935396578,
+      "loss": 1.4467,
+      "step": 7368
+    },
+    {
+      "epoch": 1.17904,
+      "grad_norm": 2.2027750332947633,
+      "learning_rate": 0.00013863708751948037,
+      "loss": 1.4027,
+      "step": 7369
+    },
+    {
+      "epoch": 1.1792,
+      "grad_norm": 2.460236396709215,
+      "learning_rate": 0.00013862115453212054,
+      "loss": 1.4346,
+      "step": 7370
+    },
+    {
+      "epoch": 1.17936,
+      "grad_norm": 2.141327221039809,
+      "learning_rate": 0.00013860522039236176,
+      "loss": 1.4486,
+      "step": 7371
+    },
+    {
+      "epoch": 1.1795200000000001,
+      "grad_norm": 1.9991183887781063,
+      "learning_rate": 0.00013858928510067944,
+      "loss": 1.4244,
+      "step": 7372
+    },
+    {
+      "epoch": 1.17968,
+      "grad_norm": 2.2498455579136447,
+      "learning_rate": 0.0001385733486575491,
+      "loss": 1.4975,
+      "step": 7373
+    },
+    {
+      "epoch": 1.17984,
+      "grad_norm": 2.2100833188035396,
+      "learning_rate": 0.00013855741106344626,
+      "loss": 1.3841,
+      "step": 7374
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 2.348055006236241,
+      "learning_rate": 0.00013854147231884642,
+      "loss": 1.4291,
+      "step": 7375
+    },
+    {
+      "epoch": 1.1801599999999999,
+      "grad_norm": 2.37889896716866,
+      "learning_rate": 0.00013852553242422522,
+      "loss": 1.4298,
+      "step": 7376
+    },
+    {
+      "epoch": 1.18032,
+      "grad_norm": 2.1089889738766923,
+      "learning_rate": 0.00013850959138005824,
+      "loss": 1.2363,
+      "step": 7377
+    },
+    {
+      "epoch": 1.18048,
+      "grad_norm": 2.7495531101118273,
+      "learning_rate": 0.00013849364918682125,
+      "loss": 1.5278,
+      "step": 7378
+    },
+    {
+      "epoch": 1.18064,
+      "grad_norm": 2.0684445014644646,
+      "learning_rate": 0.0001384777058449898,
+      "loss": 1.3765,
+      "step": 7379
+    },
+    {
+      "epoch": 1.1808,
+      "grad_norm": 2.6106107957600435,
+      "learning_rate": 0.00013846176135503968,
+      "loss": 1.4612,
+      "step": 7380
+    },
+    {
+      "epoch": 1.18096,
+      "grad_norm": 2.409982196449824,
+      "learning_rate": 0.00013844581571744667,
+      "loss": 1.4214,
+      "step": 7381
+    },
+    {
+      "epoch": 1.18112,
+      "grad_norm": 2.1097946209272718,
+      "learning_rate": 0.0001384298689326865,
+      "loss": 1.4927,
+      "step": 7382
+    },
+    {
+      "epoch": 1.18128,
+      "grad_norm": 2.745166113306601,
+      "learning_rate": 0.00013841392100123504,
+      "loss": 1.4423,
+      "step": 7383
+    },
+    {
+      "epoch": 1.18144,
+      "grad_norm": 2.455773657609783,
+      "learning_rate": 0.0001383979719235682,
+      "loss": 1.4327,
+      "step": 7384
+    },
+    {
+      "epoch": 1.1816,
+      "grad_norm": 2.5795126349031445,
+      "learning_rate": 0.00013838202170016176,
+      "loss": 1.4131,
+      "step": 7385
+    },
+    {
+      "epoch": 1.18176,
+      "grad_norm": 2.0149490886313077,
+      "learning_rate": 0.00013836607033149175,
+      "loss": 1.3908,
+      "step": 7386
+    },
+    {
+      "epoch": 1.18192,
+      "grad_norm": 2.217282289305423,
+      "learning_rate": 0.00013835011781803414,
+      "loss": 1.3954,
+      "step": 7387
+    },
+    {
+      "epoch": 1.18208,
+      "grad_norm": 2.471309464614619,
+      "learning_rate": 0.00013833416416026488,
+      "loss": 1.4549,
+      "step": 7388
+    },
+    {
+      "epoch": 1.18224,
+      "grad_norm": 2.0271615881267726,
+      "learning_rate": 0.00013831820935866003,
+      "loss": 1.4218,
+      "step": 7389
+    },
+    {
+      "epoch": 1.1824,
+      "grad_norm": 2.5160297263075764,
+      "learning_rate": 0.00013830225341369563,
+      "loss": 1.469,
+      "step": 7390
+    },
+    {
+      "epoch": 1.18256,
+      "grad_norm": 1.7858427244466448,
+      "learning_rate": 0.0001382862963258478,
+      "loss": 1.4043,
+      "step": 7391
+    },
+    {
+      "epoch": 1.18272,
+      "grad_norm": 2.588785366319812,
+      "learning_rate": 0.00013827033809559266,
+      "loss": 1.4393,
+      "step": 7392
+    },
+    {
+      "epoch": 1.18288,
+      "grad_norm": 2.0753566066384592,
+      "learning_rate": 0.00013825437872340643,
+      "loss": 1.4818,
+      "step": 7393
+    },
+    {
+      "epoch": 1.18304,
+      "grad_norm": 2.2674262753632286,
+      "learning_rate": 0.00013823841820976527,
+      "loss": 1.4621,
+      "step": 7394
+    },
+    {
+      "epoch": 1.1832,
+      "grad_norm": 2.1675507520339745,
+      "learning_rate": 0.00013822245655514543,
+      "loss": 1.3965,
+      "step": 7395
+    },
+    {
+      "epoch": 1.18336,
+      "grad_norm": 2.186033248754086,
+      "learning_rate": 0.00013820649376002317,
+      "loss": 1.4793,
+      "step": 7396
+    },
+    {
+      "epoch": 1.18352,
+      "grad_norm": 2.2592773540168904,
+      "learning_rate": 0.00013819052982487484,
+      "loss": 1.4681,
+      "step": 7397
+    },
+    {
+      "epoch": 1.18368,
+      "grad_norm": 2.121026512349721,
+      "learning_rate": 0.0001381745647501767,
+      "loss": 1.4693,
+      "step": 7398
+    },
+    {
+      "epoch": 1.18384,
+      "grad_norm": 2.508245534387749,
+      "learning_rate": 0.0001381585985364052,
+      "loss": 1.5265,
+      "step": 7399
+    },
+    {
+      "epoch": 1.184,
+      "grad_norm": 1.977845468956297,
+      "learning_rate": 0.00013814263118403674,
+      "loss": 1.4299,
+      "step": 7400
+    },
+    {
+      "epoch": 1.1841599999999999,
+      "grad_norm": 2.3188998690548717,
+      "learning_rate": 0.00013812666269354771,
+      "loss": 1.4481,
+      "step": 7401
+    },
+    {
+      "epoch": 1.18432,
+      "grad_norm": 2.0543644885984396,
+      "learning_rate": 0.00013811069306541467,
+      "loss": 1.4,
+      "step": 7402
+    },
+    {
+      "epoch": 1.18448,
+      "grad_norm": 2.3199362615731425,
+      "learning_rate": 0.00013809472230011407,
+      "loss": 1.4767,
+      "step": 7403
+    },
+    {
+      "epoch": 1.18464,
+      "grad_norm": 2.7021763749379297,
+      "learning_rate": 0.00013807875039812244,
+      "loss": 1.4985,
+      "step": 7404
+    },
+    {
+      "epoch": 1.1848,
+      "grad_norm": 2.2292700477932157,
+      "learning_rate": 0.0001380627773599164,
+      "loss": 1.3923,
+      "step": 7405
+    },
+    {
+      "epoch": 1.18496,
+      "grad_norm": 2.5984868651408752,
+      "learning_rate": 0.00013804680318597255,
+      "loss": 1.3927,
+      "step": 7406
+    },
+    {
+      "epoch": 1.18512,
+      "grad_norm": 2.2453273469403716,
+      "learning_rate": 0.00013803082787676752,
+      "loss": 1.4701,
+      "step": 7407
+    },
+    {
+      "epoch": 1.1852800000000001,
+      "grad_norm": 2.3431872424943245,
+      "learning_rate": 0.00013801485143277798,
+      "loss": 1.4339,
+      "step": 7408
+    },
+    {
+      "epoch": 1.18544,
+      "grad_norm": 2.1964455997419927,
+      "learning_rate": 0.0001379988738544807,
+      "loss": 1.3726,
+      "step": 7409
+    },
+    {
+      "epoch": 1.1856,
+      "grad_norm": 2.5193202440023836,
+      "learning_rate": 0.00013798289514235242,
+      "loss": 1.4438,
+      "step": 7410
+    },
+    {
+      "epoch": 1.18576,
+      "grad_norm": 2.306670923542775,
+      "learning_rate": 0.00013796691529686985,
+      "loss": 1.4027,
+      "step": 7411
+    },
+    {
+      "epoch": 1.18592,
+      "grad_norm": 2.1443014607654485,
+      "learning_rate": 0.00013795093431850987,
+      "loss": 1.4772,
+      "step": 7412
+    },
+    {
+      "epoch": 1.18608,
+      "grad_norm": 2.841311419668413,
+      "learning_rate": 0.00013793495220774932,
+      "loss": 1.397,
+      "step": 7413
+    },
+    {
+      "epoch": 1.18624,
+      "grad_norm": 2.246557018975698,
+      "learning_rate": 0.00013791896896506506,
+      "loss": 1.458,
+      "step": 7414
+    },
+    {
+      "epoch": 1.1864,
+      "grad_norm": 2.1742475696394092,
+      "learning_rate": 0.00013790298459093403,
+      "loss": 1.4232,
+      "step": 7415
+    },
+    {
+      "epoch": 1.18656,
+      "grad_norm": 2.1156264832972727,
+      "learning_rate": 0.00013788699908583312,
+      "loss": 1.3424,
+      "step": 7416
+    },
+    {
+      "epoch": 1.18672,
+      "grad_norm": 2.4478900898720735,
+      "learning_rate": 0.00013787101245023943,
+      "loss": 1.3595,
+      "step": 7417
+    },
+    {
+      "epoch": 1.18688,
+      "grad_norm": 2.0749711167247913,
+      "learning_rate": 0.00013785502468462987,
+      "loss": 1.3594,
+      "step": 7418
+    },
+    {
+      "epoch": 1.18704,
+      "grad_norm": 2.2765805460457056,
+      "learning_rate": 0.00013783903578948157,
+      "loss": 1.3663,
+      "step": 7419
+    },
+    {
+      "epoch": 1.1872,
+      "grad_norm": 1.7597989930722122,
+      "learning_rate": 0.00013782304576527156,
+      "loss": 1.4019,
+      "step": 7420
+    },
+    {
+      "epoch": 1.18736,
+      "grad_norm": 2.295153430901277,
+      "learning_rate": 0.000137807054612477,
+      "loss": 1.4445,
+      "step": 7421
+    },
+    {
+      "epoch": 1.18752,
+      "grad_norm": 2.0000103688893858,
+      "learning_rate": 0.000137791062331575,
+      "loss": 1.4362,
+      "step": 7422
+    },
+    {
+      "epoch": 1.18768,
+      "grad_norm": 2.112075170136014,
+      "learning_rate": 0.00013777506892304278,
+      "loss": 1.4155,
+      "step": 7423
+    },
+    {
+      "epoch": 1.18784,
+      "grad_norm": 2.4008067203881995,
+      "learning_rate": 0.0001377590743873575,
+      "loss": 1.4146,
+      "step": 7424
+    },
+    {
+      "epoch": 1.188,
+      "grad_norm": 2.2499383487700055,
+      "learning_rate": 0.0001377430787249965,
+      "loss": 1.5256,
+      "step": 7425
+    },
+    {
+      "epoch": 1.1881599999999999,
+      "grad_norm": 2.18052839088768,
+      "learning_rate": 0.00013772708193643705,
+      "loss": 1.4709,
+      "step": 7426
+    },
+    {
+      "epoch": 1.18832,
+      "grad_norm": 2.540430368704883,
+      "learning_rate": 0.0001377110840221564,
+      "loss": 1.5601,
+      "step": 7427
+    },
+    {
+      "epoch": 1.18848,
+      "grad_norm": 2.1568067802435857,
+      "learning_rate": 0.000137695084982632,
+      "loss": 1.4476,
+      "step": 7428
+    },
+    {
+      "epoch": 1.18864,
+      "grad_norm": 2.403513051951451,
+      "learning_rate": 0.00013767908481834116,
+      "loss": 1.3806,
+      "step": 7429
+    },
+    {
+      "epoch": 1.1888,
+      "grad_norm": 2.086001353693486,
+      "learning_rate": 0.00013766308352976134,
+      "loss": 1.443,
+      "step": 7430
+    },
+    {
+      "epoch": 1.18896,
+      "grad_norm": 1.9621116076326215,
+      "learning_rate": 0.00013764708111737,
+      "loss": 1.4021,
+      "step": 7431
+    },
+    {
+      "epoch": 1.18912,
+      "grad_norm": 2.565260877349093,
+      "learning_rate": 0.0001376310775816446,
+      "loss": 1.431,
+      "step": 7432
+    },
+    {
+      "epoch": 1.1892800000000001,
+      "grad_norm": 2.0655040575325043,
+      "learning_rate": 0.0001376150729230627,
+      "loss": 1.4134,
+      "step": 7433
+    },
+    {
+      "epoch": 1.18944,
+      "grad_norm": 2.2392652382568436,
+      "learning_rate": 0.0001375990671421018,
+      "loss": 1.4667,
+      "step": 7434
+    },
+    {
+      "epoch": 1.1896,
+      "grad_norm": 2.4028159752481613,
+      "learning_rate": 0.00013758306023923955,
+      "loss": 1.4824,
+      "step": 7435
+    },
+    {
+      "epoch": 1.18976,
+      "grad_norm": 2.1465201578810964,
+      "learning_rate": 0.00013756705221495356,
+      "loss": 1.5413,
+      "step": 7436
+    },
+    {
+      "epoch": 1.18992,
+      "grad_norm": 2.2094730408415275,
+      "learning_rate": 0.00013755104306972145,
+      "loss": 1.4379,
+      "step": 7437
+    },
+    {
+      "epoch": 1.19008,
+      "grad_norm": 2.4205943035846915,
+      "learning_rate": 0.00013753503280402094,
+      "loss": 1.5045,
+      "step": 7438
+    },
+    {
+      "epoch": 1.19024,
+      "grad_norm": 2.2174938179812558,
+      "learning_rate": 0.00013751902141832977,
+      "loss": 1.4162,
+      "step": 7439
+    },
+    {
+      "epoch": 1.1904,
+      "grad_norm": 2.2589101551865998,
+      "learning_rate": 0.0001375030089131256,
+      "loss": 1.3566,
+      "step": 7440
+    },
+    {
+      "epoch": 1.19056,
+      "grad_norm": 2.0168461307040433,
+      "learning_rate": 0.00013748699528888633,
+      "loss": 1.4087,
+      "step": 7441
+    },
+    {
+      "epoch": 1.19072,
+      "grad_norm": 2.030459214441904,
+      "learning_rate": 0.00013747098054608977,
+      "loss": 1.4199,
+      "step": 7442
+    },
+    {
+      "epoch": 1.19088,
+      "grad_norm": 2.0800746420150573,
+      "learning_rate": 0.00013745496468521373,
+      "loss": 1.3736,
+      "step": 7443
+    },
+    {
+      "epoch": 1.19104,
+      "grad_norm": 2.392238013217656,
+      "learning_rate": 0.00013743894770673612,
+      "loss": 1.4025,
+      "step": 7444
+    },
+    {
+      "epoch": 1.1912,
+      "grad_norm": 2.154033614987352,
+      "learning_rate": 0.00013742292961113484,
+      "loss": 1.4516,
+      "step": 7445
+    },
+    {
+      "epoch": 1.19136,
+      "grad_norm": 2.1172679334831366,
+      "learning_rate": 0.00013740691039888794,
+      "loss": 1.3701,
+      "step": 7446
+    },
+    {
+      "epoch": 1.19152,
+      "grad_norm": 2.056848655561998,
+      "learning_rate": 0.0001373908900704733,
+      "loss": 1.4376,
+      "step": 7447
+    },
+    {
+      "epoch": 1.19168,
+      "grad_norm": 2.4163839741278483,
+      "learning_rate": 0.000137374868626369,
+      "loss": 1.4568,
+      "step": 7448
+    },
+    {
+      "epoch": 1.19184,
+      "grad_norm": 2.4163839741278483,
+      "learning_rate": 0.000137374868626369,
+      "loss": 1.4255,
+      "step": 7449
+    },
+    {
+      "epoch": 1.192,
+      "grad_norm": 2.097276103414867,
+      "learning_rate": 0.00013735884606705306,
+      "loss": 1.4498,
+      "step": 7450
+    },
+    {
+      "epoch": 1.1921599999999999,
+      "grad_norm": 2.1518696442283334,
+      "learning_rate": 0.0001373428223930036,
+      "loss": 1.4307,
+      "step": 7451
+    },
+    {
+      "epoch": 1.19232,
+      "grad_norm": 2.3605515129287262,
+      "learning_rate": 0.0001373267976046987,
+      "loss": 1.391,
+      "step": 7452
+    },
+    {
+      "epoch": 1.19248,
+      "grad_norm": 2.0802831097964822,
+      "learning_rate": 0.0001373107717026166,
+      "loss": 1.4034,
+      "step": 7453
+    },
+    {
+      "epoch": 1.19264,
+      "grad_norm": 2.181725365724318,
+      "learning_rate": 0.00013729474468723542,
+      "loss": 1.4255,
+      "step": 7454
+    },
+    {
+      "epoch": 1.1928,
+      "grad_norm": 2.1571526337865823,
+      "learning_rate": 0.00013727871655903342,
+      "loss": 1.4175,
+      "step": 7455
+    },
+    {
+      "epoch": 1.19296,
+      "grad_norm": 2.4523119172060723,
+      "learning_rate": 0.00013726268731848883,
+      "loss": 1.5069,
+      "step": 7456
+    },
+    {
+      "epoch": 1.19312,
+      "grad_norm": 2.184088238632522,
+      "learning_rate": 0.00013724665696607995,
+      "loss": 1.4165,
+      "step": 7457
+    },
+    {
+      "epoch": 1.1932800000000001,
+      "grad_norm": 2.6579716560721844,
+      "learning_rate": 0.0001372306255022851,
+      "loss": 1.4023,
+      "step": 7458
+    },
+    {
+      "epoch": 1.19344,
+      "grad_norm": 2.3384509103900246,
+      "learning_rate": 0.00013721459292758258,
+      "loss": 1.3911,
+      "step": 7459
+    },
+    {
+      "epoch": 1.1936,
+      "grad_norm": 2.34669421385482,
+      "learning_rate": 0.0001371985592424509,
+      "loss": 1.4939,
+      "step": 7460
+    },
+    {
+      "epoch": 1.19376,
+      "grad_norm": 2.25202648543968,
+      "learning_rate": 0.00013718252444736842,
+      "loss": 1.4157,
+      "step": 7461
+    },
+    {
+      "epoch": 1.19392,
+      "grad_norm": 1.9780843585229877,
+      "learning_rate": 0.00013716648854281358,
+      "loss": 1.5033,
+      "step": 7462
+    },
+    {
+      "epoch": 1.19408,
+      "grad_norm": 2.0292733251198323,
+      "learning_rate": 0.00013715045152926485,
+      "loss": 1.3622,
+      "step": 7463
+    },
+    {
+      "epoch": 1.19424,
+      "grad_norm": 2.126409279423644,
+      "learning_rate": 0.00013713441340720083,
+      "loss": 1.4622,
+      "step": 7464
+    },
+    {
+      "epoch": 1.1944,
+      "grad_norm": 2.271278482999961,
+      "learning_rate": 0.0001371183741771,
+      "loss": 1.4245,
+      "step": 7465
+    },
+    {
+      "epoch": 1.19456,
+      "grad_norm": 2.1297724212495273,
+      "learning_rate": 0.00013710233383944093,
+      "loss": 1.4108,
+      "step": 7466
+    },
+    {
+      "epoch": 1.19472,
+      "grad_norm": 2.6754469675330297,
+      "learning_rate": 0.00013708629239470233,
+      "loss": 1.4495,
+      "step": 7467
+    },
+    {
+      "epoch": 1.19488,
+      "grad_norm": 2.2408700436100553,
+      "learning_rate": 0.00013707024984336278,
+      "loss": 1.4728,
+      "step": 7468
+    },
+    {
+      "epoch": 1.19504,
+      "grad_norm": 2.269532555864553,
+      "learning_rate": 0.00013705420618590101,
+      "loss": 1.4759,
+      "step": 7469
+    },
+    {
+      "epoch": 1.1952,
+      "grad_norm": 2.0829118895117724,
+      "learning_rate": 0.00013703816142279573,
+      "loss": 1.4546,
+      "step": 7470
+    },
+    {
+      "epoch": 1.19536,
+      "grad_norm": 2.3834958341880834,
+      "learning_rate": 0.00013702211555452566,
+      "loss": 1.3968,
+      "step": 7471
+    },
+    {
+      "epoch": 1.19552,
+      "grad_norm": 2.547345171701889,
+      "learning_rate": 0.00013700606858156963,
+      "loss": 1.4171,
+      "step": 7472
+    },
+    {
+      "epoch": 1.19568,
+      "grad_norm": 2.046836277537389,
+      "learning_rate": 0.00013699002050440644,
+      "loss": 1.4082,
+      "step": 7473
+    },
+    {
+      "epoch": 1.19584,
+      "grad_norm": 2.03005321243298,
+      "learning_rate": 0.00013697397132351487,
+      "loss": 1.4445,
+      "step": 7474
+    },
+    {
+      "epoch": 1.196,
+      "grad_norm": 2.3594803067464953,
+      "learning_rate": 0.00013695792103937394,
+      "loss": 1.3642,
+      "step": 7475
+    },
+    {
+      "epoch": 1.19616,
+      "grad_norm": 2.6194699044070466,
+      "learning_rate": 0.0001369418696524625,
+      "loss": 1.4995,
+      "step": 7476
+    },
+    {
+      "epoch": 1.19632,
+      "grad_norm": 2.2737395695536176,
+      "learning_rate": 0.00013692581716325948,
+      "loss": 1.3945,
+      "step": 7477
+    },
+    {
+      "epoch": 1.19648,
+      "grad_norm": 1.9795877352328772,
+      "learning_rate": 0.0001369097635722439,
+      "loss": 1.4413,
+      "step": 7478
+    },
+    {
+      "epoch": 1.19664,
+      "grad_norm": 2.29014732940139,
+      "learning_rate": 0.00013689370887989474,
+      "loss": 1.4367,
+      "step": 7479
+    },
+    {
+      "epoch": 1.1968,
+      "grad_norm": 2.215742223535974,
+      "learning_rate": 0.0001368776530866911,
+      "loss": 1.4301,
+      "step": 7480
+    },
+    {
+      "epoch": 1.19696,
+      "grad_norm": 2.2707106038638245,
+      "learning_rate": 0.00013686159619311197,
+      "loss": 1.4785,
+      "step": 7481
+    },
+    {
+      "epoch": 1.19712,
+      "grad_norm": 2.5037466074574932,
+      "learning_rate": 0.00013684553819963657,
+      "loss": 1.4371,
+      "step": 7482
+    },
+    {
+      "epoch": 1.1972800000000001,
+      "grad_norm": 1.9914608350689629,
+      "learning_rate": 0.00013682947910674398,
+      "loss": 1.3799,
+      "step": 7483
+    },
+    {
+      "epoch": 1.19744,
+      "grad_norm": 2.2369013455888447,
+      "learning_rate": 0.00013681341891491343,
+      "loss": 1.4417,
+      "step": 7484
+    },
+    {
+      "epoch": 1.1976,
+      "grad_norm": 2.2787056472873672,
+      "learning_rate": 0.00013679735762462403,
+      "loss": 1.4656,
+      "step": 7485
+    },
+    {
+      "epoch": 1.19776,
+      "grad_norm": 2.3164902559340117,
+      "learning_rate": 0.00013678129523635516,
+      "loss": 1.4125,
+      "step": 7486
+    },
+    {
+      "epoch": 1.19792,
+      "grad_norm": 2.6945802774227916,
+      "learning_rate": 0.00013676523175058604,
+      "loss": 1.4082,
+      "step": 7487
+    },
+    {
+      "epoch": 1.19808,
+      "grad_norm": 2.1631204401941986,
+      "learning_rate": 0.00013674916716779596,
+      "loss": 1.3737,
+      "step": 7488
+    },
+    {
+      "epoch": 1.19824,
+      "grad_norm": 2.5549741098430108,
+      "learning_rate": 0.00013673310148846427,
+      "loss": 1.3924,
+      "step": 7489
+    },
+    {
+      "epoch": 1.1984,
+      "grad_norm": 2.8147798773499018,
+      "learning_rate": 0.00013671703471307035,
+      "loss": 1.4215,
+      "step": 7490
+    },
+    {
+      "epoch": 1.19856,
+      "grad_norm": 1.9388280252791474,
+      "learning_rate": 0.00013670096684209363,
+      "loss": 1.3493,
+      "step": 7491
+    },
+    {
+      "epoch": 1.19872,
+      "grad_norm": 3.2163660288411497,
+      "learning_rate": 0.00013668489787601356,
+      "loss": 1.4914,
+      "step": 7492
+    },
+    {
+      "epoch": 1.19888,
+      "grad_norm": 2.037269017514059,
+      "learning_rate": 0.00013666882781530957,
+      "loss": 1.4867,
+      "step": 7493
+    },
+    {
+      "epoch": 1.19904,
+      "grad_norm": 2.6176648690743924,
+      "learning_rate": 0.0001366527566604612,
+      "loss": 1.3642,
+      "step": 7494
+    },
+    {
+      "epoch": 1.1992,
+      "grad_norm": 2.4663551537307193,
+      "learning_rate": 0.00013663668441194797,
+      "loss": 1.4052,
+      "step": 7495
+    },
+    {
+      "epoch": 1.19936,
+      "grad_norm": 1.9158653416854674,
+      "learning_rate": 0.00013662061107024947,
+      "loss": 1.3894,
+      "step": 7496
+    },
+    {
+      "epoch": 1.19952,
+      "grad_norm": 2.575964686910809,
+      "learning_rate": 0.0001366045366358453,
+      "loss": 1.4386,
+      "step": 7497
+    },
+    {
+      "epoch": 1.19968,
+      "grad_norm": 2.3814873220306105,
+      "learning_rate": 0.00013658846110921506,
+      "loss": 1.426,
+      "step": 7498
+    },
+    {
+      "epoch": 1.19984,
+      "grad_norm": 3.3771154302182036,
+      "learning_rate": 0.00013657238449083848,
+      "loss": 1.4349,
+      "step": 7499
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 2.6227283380013664,
+      "learning_rate": 0.00013655630678119525,
+      "loss": 1.465,
+      "step": 7500
+    },
+    {
+      "epoch": 1.20016,
+      "grad_norm": 2.3385748199914604,
+      "learning_rate": 0.0001365402279807651,
+      "loss": 1.3632,
+      "step": 7501
+    },
+    {
+      "epoch": 1.20032,
+      "grad_norm": 2.7552860947367437,
+      "learning_rate": 0.00013652414809002777,
+      "loss": 1.4013,
+      "step": 7502
+    },
+    {
+      "epoch": 1.20048,
+      "grad_norm": 2.22588319725911,
+      "learning_rate": 0.0001365080671094631,
+      "loss": 1.4599,
+      "step": 7503
+    },
+    {
+      "epoch": 1.20064,
+      "grad_norm": 2.2475826863414317,
+      "learning_rate": 0.00013649198503955088,
+      "loss": 1.4685,
+      "step": 7504
+    },
+    {
+      "epoch": 1.2008,
+      "grad_norm": 2.4224191167262794,
+      "learning_rate": 0.00013647590188077102,
+      "loss": 1.4332,
+      "step": 7505
+    },
+    {
+      "epoch": 1.20096,
+      "grad_norm": 2.2395432668395667,
+      "learning_rate": 0.00013645981763360342,
+      "loss": 1.4414,
+      "step": 7506
+    },
+    {
+      "epoch": 1.20112,
+      "grad_norm": 1.9311663937720673,
+      "learning_rate": 0.00013644373229852795,
+      "loss": 1.4215,
+      "step": 7507
+    },
+    {
+      "epoch": 1.20128,
+      "grad_norm": 2.6261321658340937,
+      "learning_rate": 0.00013642764587602462,
+      "loss": 1.4548,
+      "step": 7508
+    },
+    {
+      "epoch": 1.20144,
+      "grad_norm": 2.1450533353719066,
+      "learning_rate": 0.00013641155836657343,
+      "loss": 1.369,
+      "step": 7509
+    },
+    {
+      "epoch": 1.2016,
+      "grad_norm": 2.4335499273473453,
+      "learning_rate": 0.00013639546977065438,
+      "loss": 1.4064,
+      "step": 7510
+    },
+    {
+      "epoch": 1.20176,
+      "grad_norm": 2.4506387387126076,
+      "learning_rate": 0.00013637938008874756,
+      "loss": 1.4523,
+      "step": 7511
+    },
+    {
+      "epoch": 1.2019199999999999,
+      "grad_norm": 1.9521140645941901,
+      "learning_rate": 0.00013636328932133305,
+      "loss": 1.4231,
+      "step": 7512
+    },
+    {
+      "epoch": 1.20208,
+      "grad_norm": 2.306299325798555,
+      "learning_rate": 0.00013634719746889097,
+      "loss": 1.4173,
+      "step": 7513
+    },
+    {
+      "epoch": 1.20224,
+      "grad_norm": 2.759833042977578,
+      "learning_rate": 0.00013633110453190148,
+      "loss": 1.4721,
+      "step": 7514
+    },
+    {
+      "epoch": 1.2024,
+      "grad_norm": 1.9299119892618846,
+      "learning_rate": 0.0001363150105108448,
+      "loss": 1.4366,
+      "step": 7515
+    },
+    {
+      "epoch": 1.20256,
+      "grad_norm": 2.1933686076763523,
+      "learning_rate": 0.00013629891540620107,
+      "loss": 1.3769,
+      "step": 7516
+    },
+    {
+      "epoch": 1.20272,
+      "grad_norm": 2.172679412751591,
+      "learning_rate": 0.0001362828192184506,
+      "loss": 1.4857,
+      "step": 7517
+    },
+    {
+      "epoch": 1.20288,
+      "grad_norm": 2.132469561340286,
+      "learning_rate": 0.0001362667219480737,
+      "loss": 1.4312,
+      "step": 7518
+    },
+    {
+      "epoch": 1.20304,
+      "grad_norm": 2.1576202618791647,
+      "learning_rate": 0.00013625062359555066,
+      "loss": 1.4351,
+      "step": 7519
+    },
+    {
+      "epoch": 1.2032,
+      "grad_norm": 1.9103623211665883,
+      "learning_rate": 0.00013623452416136185,
+      "loss": 1.4042,
+      "step": 7520
+    },
+    {
+      "epoch": 1.20336,
+      "grad_norm": 2.4581662811581477,
+      "learning_rate": 0.00013621842364598764,
+      "loss": 1.4024,
+      "step": 7521
+    },
+    {
+      "epoch": 1.20352,
+      "grad_norm": 2.2072300070856516,
+      "learning_rate": 0.00013620232204990842,
+      "loss": 1.4126,
+      "step": 7522
+    },
+    {
+      "epoch": 1.20368,
+      "grad_norm": 1.9200547454512529,
+      "learning_rate": 0.00013618621937360465,
+      "loss": 1.4411,
+      "step": 7523
+    },
+    {
+      "epoch": 1.20384,
+      "grad_norm": 2.151648784391176,
+      "learning_rate": 0.00013617011561755683,
+      "loss": 1.3676,
+      "step": 7524
+    },
+    {
+      "epoch": 1.204,
+      "grad_norm": 2.096146259736935,
+      "learning_rate": 0.00013615401078224547,
+      "loss": 1.3381,
+      "step": 7525
+    },
+    {
+      "epoch": 1.20416,
+      "grad_norm": 2.7912234116744914,
+      "learning_rate": 0.00013613790486815113,
+      "loss": 1.3897,
+      "step": 7526
+    },
+    {
+      "epoch": 1.20432,
+      "grad_norm": 2.146341397623456,
+      "learning_rate": 0.00013612179787575435,
+      "loss": 1.4606,
+      "step": 7527
+    },
+    {
+      "epoch": 1.20448,
+      "grad_norm": 2.0721575587716585,
+      "learning_rate": 0.00013610568980553577,
+      "loss": 1.4154,
+      "step": 7528
+    },
+    {
+      "epoch": 1.20464,
+      "grad_norm": 2.405864904214309,
+      "learning_rate": 0.000136089580657976,
+      "loss": 1.3899,
+      "step": 7529
+    },
+    {
+      "epoch": 1.2048,
+      "grad_norm": 2.156846588644525,
+      "learning_rate": 0.00013607347043355573,
+      "loss": 1.4482,
+      "step": 7530
+    },
+    {
+      "epoch": 1.20496,
+      "grad_norm": 1.984180663641548,
+      "learning_rate": 0.0001360573591327557,
+      "loss": 1.4967,
+      "step": 7531
+    },
+    {
+      "epoch": 1.20512,
+      "grad_norm": 2.8173953521013946,
+      "learning_rate": 0.00013604124675605655,
+      "loss": 1.3437,
+      "step": 7532
+    },
+    {
+      "epoch": 1.20528,
+      "grad_norm": 2.3071922051485827,
+      "learning_rate": 0.0001360251333039392,
+      "loss": 1.4156,
+      "step": 7533
+    },
+    {
+      "epoch": 1.20544,
+      "grad_norm": 2.6536071191456356,
+      "learning_rate": 0.0001360090187768843,
+      "loss": 1.378,
+      "step": 7534
+    },
+    {
+      "epoch": 1.2056,
+      "grad_norm": 2.5752472890113625,
+      "learning_rate": 0.0001359929031753728,
+      "loss": 1.4085,
+      "step": 7535
+    },
+    {
+      "epoch": 1.20576,
+      "grad_norm": 2.125005960675252,
+      "learning_rate": 0.00013597678649988546,
+      "loss": 1.4898,
+      "step": 7536
+    },
+    {
+      "epoch": 1.2059199999999999,
+      "grad_norm": 2.7182923288764913,
+      "learning_rate": 0.00013596066875090328,
+      "loss": 1.412,
+      "step": 7537
+    },
+    {
+      "epoch": 1.20608,
+      "grad_norm": 2.5419652230732366,
+      "learning_rate": 0.00013594454992890714,
+      "loss": 1.3565,
+      "step": 7538
+    },
+    {
+      "epoch": 1.20624,
+      "grad_norm": 2.4201355106017086,
+      "learning_rate": 0.00013592843003437798,
+      "loss": 1.4212,
+      "step": 7539
+    },
+    {
+      "epoch": 1.2064,
+      "grad_norm": 2.198947218683021,
+      "learning_rate": 0.00013591230906779682,
+      "loss": 1.3603,
+      "step": 7540
+    },
+    {
+      "epoch": 1.20656,
+      "grad_norm": 2.316917701262154,
+      "learning_rate": 0.00013589618702964472,
+      "loss": 1.4336,
+      "step": 7541
+    },
+    {
+      "epoch": 1.20672,
+      "grad_norm": 2.175516891634375,
+      "learning_rate": 0.00013588006392040267,
+      "loss": 1.3896,
+      "step": 7542
+    },
+    {
+      "epoch": 1.20688,
+      "grad_norm": 1.9280133280461784,
+      "learning_rate": 0.00013586393974055182,
+      "loss": 1.4029,
+      "step": 7543
+    },
+    {
+      "epoch": 1.2070400000000001,
+      "grad_norm": 2.131764323541729,
+      "learning_rate": 0.00013584781449057327,
+      "loss": 1.336,
+      "step": 7544
+    },
+    {
+      "epoch": 1.2072,
+      "grad_norm": 1.8399835592937208,
+      "learning_rate": 0.00013583168817094815,
+      "loss": 1.3786,
+      "step": 7545
+    },
+    {
+      "epoch": 1.20736,
+      "grad_norm": 2.173033267082524,
+      "learning_rate": 0.0001358155607821577,
+      "loss": 1.3891,
+      "step": 7546
+    },
+    {
+      "epoch": 1.20752,
+      "grad_norm": 2.9083852085574367,
+      "learning_rate": 0.00013579943232468308,
+      "loss": 1.3897,
+      "step": 7547
+    },
+    {
+      "epoch": 1.20768,
+      "grad_norm": 2.109647572876746,
+      "learning_rate": 0.00013578330279900554,
+      "loss": 1.4232,
+      "step": 7548
+    },
+    {
+      "epoch": 1.20784,
+      "grad_norm": 2.8072609754966455,
+      "learning_rate": 0.00013576717220560643,
+      "loss": 1.4904,
+      "step": 7549
+    },
+    {
+      "epoch": 1.208,
+      "grad_norm": 1.9573474035621647,
+      "learning_rate": 0.00013575104054496702,
+      "loss": 1.4314,
+      "step": 7550
+    },
+    {
+      "epoch": 1.20816,
+      "grad_norm": 2.2368883359813885,
+      "learning_rate": 0.00013573490781756863,
+      "loss": 1.3788,
+      "step": 7551
+    },
+    {
+      "epoch": 1.20832,
+      "grad_norm": 1.9900752354301825,
+      "learning_rate": 0.00013571877402389267,
+      "loss": 1.3778,
+      "step": 7552
+    },
+    {
+      "epoch": 1.20848,
+      "grad_norm": 2.092135602767866,
+      "learning_rate": 0.00013570263916442055,
+      "loss": 1.3885,
+      "step": 7553
+    },
+    {
+      "epoch": 1.20864,
+      "grad_norm": 2.5803419945572874,
+      "learning_rate": 0.0001356865032396337,
+      "loss": 1.3979,
+      "step": 7554
+    },
+    {
+      "epoch": 1.2088,
+      "grad_norm": 3.232801042241345,
+      "learning_rate": 0.0001356703662500136,
+      "loss": 1.5013,
+      "step": 7555
+    },
+    {
+      "epoch": 1.20896,
+      "grad_norm": 2.5467214095689235,
+      "learning_rate": 0.0001356542281960417,
+      "loss": 1.4687,
+      "step": 7556
+    },
+    {
+      "epoch": 1.20912,
+      "grad_norm": 3.175686622789188,
+      "learning_rate": 0.00013563808907819964,
+      "loss": 1.42,
+      "step": 7557
+    },
+    {
+      "epoch": 1.20928,
+      "grad_norm": 2.7333982012907647,
+      "learning_rate": 0.00013562194889696892,
+      "loss": 1.4261,
+      "step": 7558
+    },
+    {
+      "epoch": 1.20944,
+      "grad_norm": 2.298865731374525,
+      "learning_rate": 0.00013560580765283113,
+      "loss": 1.4101,
+      "step": 7559
+    },
+    {
+      "epoch": 1.2096,
+      "grad_norm": 2.790213580985626,
+      "learning_rate": 0.00013558966534626797,
+      "loss": 1.3773,
+      "step": 7560
+    },
+    {
+      "epoch": 1.20976,
+      "grad_norm": 2.1449370520593107,
+      "learning_rate": 0.000135573521977761,
+      "loss": 1.3638,
+      "step": 7561
+    },
+    {
+      "epoch": 1.2099199999999999,
+      "grad_norm": 2.925941913818032,
+      "learning_rate": 0.00013555737754779198,
+      "loss": 1.3632,
+      "step": 7562
+    },
+    {
+      "epoch": 1.21008,
+      "grad_norm": 2.872529561541503,
+      "learning_rate": 0.00013554123205684264,
+      "loss": 1.4585,
+      "step": 7563
+    },
+    {
+      "epoch": 1.21024,
+      "grad_norm": 2.1886134707582308,
+      "learning_rate": 0.00013552508550539473,
+      "loss": 1.4276,
+      "step": 7564
+    },
+    {
+      "epoch": 1.2104,
+      "grad_norm": 2.6883655785040714,
+      "learning_rate": 0.00013550893789393,
+      "loss": 1.4201,
+      "step": 7565
+    },
+    {
+      "epoch": 1.21056,
+      "grad_norm": 2.8544364097318895,
+      "learning_rate": 0.00013549278922293033,
+      "loss": 1.4051,
+      "step": 7566
+    },
+    {
+      "epoch": 1.21072,
+      "grad_norm": 1.9065722777057597,
+      "learning_rate": 0.00013547663949287756,
+      "loss": 1.4263,
+      "step": 7567
+    },
+    {
+      "epoch": 1.21088,
+      "grad_norm": 3.580350634559719,
+      "learning_rate": 0.00013546048870425356,
+      "loss": 1.4139,
+      "step": 7568
+    },
+    {
+      "epoch": 1.2110400000000001,
+      "grad_norm": 2.2900822714819444,
+      "learning_rate": 0.00013544433685754026,
+      "loss": 1.343,
+      "step": 7569
+    },
+    {
+      "epoch": 1.2112,
+      "grad_norm": 1.8339466363260206,
+      "learning_rate": 0.00013542818395321956,
+      "loss": 1.3838,
+      "step": 7570
+    },
+    {
+      "epoch": 1.21136,
+      "grad_norm": 3.56776024672953,
+      "learning_rate": 0.0001354120299917735,
+      "loss": 1.3811,
+      "step": 7571
+    },
+    {
+      "epoch": 1.21152,
+      "grad_norm": 2.3139994894733578,
+      "learning_rate": 0.00013539587497368407,
+      "loss": 1.3943,
+      "step": 7572
+    },
+    {
+      "epoch": 1.21168,
+      "grad_norm": 2.061107730024072,
+      "learning_rate": 0.0001353797188994333,
+      "loss": 1.4206,
+      "step": 7573
+    },
+    {
+      "epoch": 1.21184,
+      "grad_norm": 2.4325077110566955,
+      "learning_rate": 0.00013536356176950326,
+      "loss": 1.4705,
+      "step": 7574
+    },
+    {
+      "epoch": 1.212,
+      "grad_norm": 2.4169196313867483,
+      "learning_rate": 0.0001353474035843761,
+      "loss": 1.3815,
+      "step": 7575
+    },
+    {
+      "epoch": 1.21216,
+      "grad_norm": 2.2973118150208176,
+      "learning_rate": 0.00013533124434453387,
+      "loss": 1.5274,
+      "step": 7576
+    },
+    {
+      "epoch": 1.21232,
+      "grad_norm": 2.731973100296467,
+      "learning_rate": 0.00013531508405045884,
+      "loss": 1.36,
+      "step": 7577
+    },
+    {
+      "epoch": 1.21248,
+      "grad_norm": 2.1981644693615663,
+      "learning_rate": 0.00013529892270263314,
+      "loss": 1.3935,
+      "step": 7578
+    },
+    {
+      "epoch": 1.21264,
+      "grad_norm": 1.938602062037591,
+      "learning_rate": 0.000135282760301539,
+      "loss": 1.343,
+      "step": 7579
+    },
+    {
+      "epoch": 1.2128,
+      "grad_norm": 3.8300153352625474,
+      "learning_rate": 0.00013526659684765877,
+      "loss": 1.448,
+      "step": 7580
+    },
+    {
+      "epoch": 1.21296,
+      "grad_norm": 2.3769909106773057,
+      "learning_rate": 0.0001352504323414746,
+      "loss": 1.363,
+      "step": 7581
+    },
+    {
+      "epoch": 1.21312,
+      "grad_norm": 2.4346161563843753,
+      "learning_rate": 0.00013523426678346896,
+      "loss": 1.4166,
+      "step": 7582
+    },
+    {
+      "epoch": 1.21328,
+      "grad_norm": 2.22255419765705,
+      "learning_rate": 0.00013521810017412411,
+      "loss": 1.4096,
+      "step": 7583
+    },
+    {
+      "epoch": 1.21344,
+      "grad_norm": 2.542070989229085,
+      "learning_rate": 0.0001352019325139225,
+      "loss": 1.431,
+      "step": 7584
+    },
+    {
+      "epoch": 1.2136,
+      "grad_norm": 2.8363179601545747,
+      "learning_rate": 0.00013518576380334646,
+      "loss": 1.5095,
+      "step": 7585
+    },
+    {
+      "epoch": 1.21376,
+      "grad_norm": 3.2329849625535982,
+      "learning_rate": 0.00013516959404287853,
+      "loss": 1.3933,
+      "step": 7586
+    },
+    {
+      "epoch": 1.2139199999999999,
+      "grad_norm": 2.2653009649961717,
+      "learning_rate": 0.0001351534232330012,
+      "loss": 1.4772,
+      "step": 7587
+    },
+    {
+      "epoch": 1.21408,
+      "grad_norm": 2.4401490579088665,
+      "learning_rate": 0.0001351372513741969,
+      "loss": 1.3905,
+      "step": 7588
+    },
+    {
+      "epoch": 1.21424,
+      "grad_norm": 2.965486889201235,
+      "learning_rate": 0.0001351210784669482,
+      "loss": 1.393,
+      "step": 7589
+    },
+    {
+      "epoch": 1.2144,
+      "grad_norm": 2.097377018270871,
+      "learning_rate": 0.0001351049045117377,
+      "loss": 1.4222,
+      "step": 7590
+    },
+    {
+      "epoch": 1.21456,
+      "grad_norm": 3.2434253017210293,
+      "learning_rate": 0.00013508872950904806,
+      "loss": 1.3871,
+      "step": 7591
+    },
+    {
+      "epoch": 1.21472,
+      "grad_norm": 2.7410298046152612,
+      "learning_rate": 0.00013507255345936183,
+      "loss": 1.5095,
+      "step": 7592
+    },
+    {
+      "epoch": 1.21488,
+      "grad_norm": 2.610333982136673,
+      "learning_rate": 0.0001350563763631617,
+      "loss": 1.4191,
+      "step": 7593
+    },
+    {
+      "epoch": 1.2150400000000001,
+      "grad_norm": 2.610333982136673,
+      "learning_rate": 0.0001350563763631617,
+      "loss": 1.4302,
+      "step": 7594
+    },
+    {
+      "epoch": 1.2152,
+      "grad_norm": 2.473182687801616,
+      "learning_rate": 0.0001350401982209304,
+      "loss": 1.3872,
+      "step": 7595
+    },
+    {
+      "epoch": 1.21536,
+      "grad_norm": 2.1485333108941735,
+      "learning_rate": 0.00013502401903315063,
+      "loss": 1.4606,
+      "step": 7596
+    },
+    {
+      "epoch": 1.21552,
+      "grad_norm": 2.255098876045817,
+      "learning_rate": 0.00013500783880030515,
+      "loss": 1.4007,
+      "step": 7597
+    },
+    {
+      "epoch": 1.21568,
+      "grad_norm": 2.443226782204094,
+      "learning_rate": 0.0001349916575228768,
+      "loss": 1.3992,
+      "step": 7598
+    },
+    {
+      "epoch": 1.21584,
+      "grad_norm": 2.2143758789160723,
+      "learning_rate": 0.00013497547520134836,
+      "loss": 1.3434,
+      "step": 7599
+    },
+    {
+      "epoch": 1.216,
+      "grad_norm": 2.478742479184506,
+      "learning_rate": 0.00013495929183620273,
+      "loss": 1.44,
+      "step": 7600
+    },
+    {
+      "epoch": 1.21616,
+      "grad_norm": 2.230147841682108,
+      "learning_rate": 0.00013494310742792276,
+      "loss": 1.4172,
+      "step": 7601
+    },
+    {
+      "epoch": 1.21632,
+      "grad_norm": 2.6471131008144657,
+      "learning_rate": 0.00013492692197699136,
+      "loss": 1.3885,
+      "step": 7602
+    },
+    {
+      "epoch": 1.21648,
+      "grad_norm": 2.050137226104183,
+      "learning_rate": 0.00013491073548389156,
+      "loss": 1.5287,
+      "step": 7603
+    },
+    {
+      "epoch": 1.21664,
+      "grad_norm": 2.1414633630684325,
+      "learning_rate": 0.00013489454794910624,
+      "loss": 1.3656,
+      "step": 7604
+    },
+    {
+      "epoch": 1.2168,
+      "grad_norm": 2.2125355051420366,
+      "learning_rate": 0.00013487835937311845,
+      "loss": 1.3902,
+      "step": 7605
+    },
+    {
+      "epoch": 1.21696,
+      "grad_norm": 2.212882899365729,
+      "learning_rate": 0.00013486216975641124,
+      "loss": 1.3927,
+      "step": 7606
+    },
+    {
+      "epoch": 1.21712,
+      "grad_norm": 2.281674803282582,
+      "learning_rate": 0.0001348459790994677,
+      "loss": 1.4785,
+      "step": 7607
+    },
+    {
+      "epoch": 1.21728,
+      "grad_norm": 2.367064530158535,
+      "learning_rate": 0.0001348297874027709,
+      "loss": 1.3638,
+      "step": 7608
+    },
+    {
+      "epoch": 1.21744,
+      "grad_norm": 2.492615372850982,
+      "learning_rate": 0.000134813594666804,
+      "loss": 1.3846,
+      "step": 7609
+    },
+    {
+      "epoch": 1.2176,
+      "grad_norm": 2.5576274345454584,
+      "learning_rate": 0.0001347974008920502,
+      "loss": 1.4151,
+      "step": 7610
+    },
+    {
+      "epoch": 1.21776,
+      "grad_norm": 2.0626863039578605,
+      "learning_rate": 0.00013478120607899263,
+      "loss": 1.3697,
+      "step": 7611
+    },
+    {
+      "epoch": 1.21792,
+      "grad_norm": 2.2618995588801463,
+      "learning_rate": 0.00013476501022811457,
+      "loss": 1.3818,
+      "step": 7612
+    },
+    {
+      "epoch": 1.21808,
+      "grad_norm": 3.3731305307209465,
+      "learning_rate": 0.00013474881333989927,
+      "loss": 1.3739,
+      "step": 7613
+    },
+    {
+      "epoch": 1.21824,
+      "grad_norm": 2.210757865680817,
+      "learning_rate": 0.00013473261541482997,
+      "loss": 1.4491,
+      "step": 7614
+    },
+    {
+      "epoch": 1.2184,
+      "grad_norm": 2.4183793391443618,
+      "learning_rate": 0.00013471641645339006,
+      "loss": 1.4402,
+      "step": 7615
+    },
+    {
+      "epoch": 1.21856,
+      "grad_norm": 3.339978441677294,
+      "learning_rate": 0.00013470021645606287,
+      "loss": 1.3753,
+      "step": 7616
+    },
+    {
+      "epoch": 1.21872,
+      "grad_norm": 2.2768876292168536,
+      "learning_rate": 0.0001346840154233318,
+      "loss": 1.3779,
+      "step": 7617
+    },
+    {
+      "epoch": 1.21888,
+      "grad_norm": 2.4776558020130732,
+      "learning_rate": 0.00013466781335568024,
+      "loss": 1.4376,
+      "step": 7618
+    },
+    {
+      "epoch": 1.2190400000000001,
+      "grad_norm": 2.6344344232688286,
+      "learning_rate": 0.00013465161025359165,
+      "loss": 1.5155,
+      "step": 7619
+    },
+    {
+      "epoch": 1.2192,
+      "grad_norm": 2.0610583865810437,
+      "learning_rate": 0.00013463540611754948,
+      "loss": 1.3795,
+      "step": 7620
+    },
+    {
+      "epoch": 1.21936,
+      "grad_norm": 3.1786769256167533,
+      "learning_rate": 0.00013461920094803726,
+      "loss": 1.432,
+      "step": 7621
+    },
+    {
+      "epoch": 1.21952,
+      "grad_norm": 2.771924309195937,
+      "learning_rate": 0.00013460299474553854,
+      "loss": 1.3959,
+      "step": 7622
+    },
+    {
+      "epoch": 1.2196799999999999,
+      "grad_norm": 2.4145024409773184,
+      "learning_rate": 0.00013458678751053689,
+      "loss": 1.4461,
+      "step": 7623
+    },
+    {
+      "epoch": 1.21984,
+      "grad_norm": 3.2065393924783954,
+      "learning_rate": 0.00013457057924351587,
+      "loss": 1.4832,
+      "step": 7624
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 2.155367526239326,
+      "learning_rate": 0.0001345543699449592,
+      "loss": 1.4109,
+      "step": 7625
+    },
+    {
+      "epoch": 1.22016,
+      "grad_norm": 2.163262023631329,
+      "learning_rate": 0.00013453815961535044,
+      "loss": 1.4297,
+      "step": 7626
+    },
+    {
+      "epoch": 1.22032,
+      "grad_norm": 2.066784901798655,
+      "learning_rate": 0.00013452194825517333,
+      "loss": 1.5412,
+      "step": 7627
+    },
+    {
+      "epoch": 1.22048,
+      "grad_norm": 2.3950953264156642,
+      "learning_rate": 0.0001345057358649116,
+      "loss": 1.3906,
+      "step": 7628
+    },
+    {
+      "epoch": 1.22064,
+      "grad_norm": 1.709497056559193,
+      "learning_rate": 0.00013448952244504895,
+      "loss": 1.3299,
+      "step": 7629
+    },
+    {
+      "epoch": 1.2208,
+      "grad_norm": 2.480630202558256,
+      "learning_rate": 0.00013447330799606927,
+      "loss": 1.4357,
+      "step": 7630
+    },
+    {
+      "epoch": 1.22096,
+      "grad_norm": 2.1342084331704245,
+      "learning_rate": 0.00013445709251845624,
+      "loss": 1.371,
+      "step": 7631
+    },
+    {
+      "epoch": 1.22112,
+      "grad_norm": 2.316624586073391,
+      "learning_rate": 0.00013444087601269383,
+      "loss": 1.4109,
+      "step": 7632
+    },
+    {
+      "epoch": 1.22128,
+      "grad_norm": 2.191966040812517,
+      "learning_rate": 0.00013442465847926584,
+      "loss": 1.4402,
+      "step": 7633
+    },
+    {
+      "epoch": 1.22144,
+      "grad_norm": 2.628938767121711,
+      "learning_rate": 0.00013440843991865622,
+      "loss": 1.446,
+      "step": 7634
+    },
+    {
+      "epoch": 1.2216,
+      "grad_norm": 2.592865044415441,
+      "learning_rate": 0.0001343922203313489,
+      "loss": 1.4289,
+      "step": 7635
+    },
+    {
+      "epoch": 1.22176,
+      "grad_norm": 2.2266927332289583,
+      "learning_rate": 0.0001343759997178278,
+      "loss": 1.4381,
+      "step": 7636
+    },
+    {
+      "epoch": 1.22192,
+      "grad_norm": 2.518405201467184,
+      "learning_rate": 0.00013435977807857704,
+      "loss": 1.5119,
+      "step": 7637
+    },
+    {
+      "epoch": 1.22208,
+      "grad_norm": 2.1107448785299674,
+      "learning_rate": 0.0001343435554140805,
+      "loss": 1.4098,
+      "step": 7638
+    },
+    {
+      "epoch": 1.22224,
+      "grad_norm": 2.1179415204681047,
+      "learning_rate": 0.0001343273317248223,
+      "loss": 1.3673,
+      "step": 7639
+    },
+    {
+      "epoch": 1.2224,
+      "grad_norm": 2.034256112810964,
+      "learning_rate": 0.00013431110701128657,
+      "loss": 1.4102,
+      "step": 7640
+    },
+    {
+      "epoch": 1.22256,
+      "grad_norm": 2.0789944539039364,
+      "learning_rate": 0.00013429488127395743,
+      "loss": 1.4233,
+      "step": 7641
+    },
+    {
+      "epoch": 1.22272,
+      "grad_norm": 2.326346538615457,
+      "learning_rate": 0.00013427865451331897,
+      "loss": 1.392,
+      "step": 7642
+    },
+    {
+      "epoch": 1.22288,
+      "grad_norm": 1.913878775070153,
+      "learning_rate": 0.0001342624267298554,
+      "loss": 1.4473,
+      "step": 7643
+    },
+    {
+      "epoch": 1.22304,
+      "grad_norm": 2.1105421147043715,
+      "learning_rate": 0.00013424619792405098,
+      "loss": 1.4129,
+      "step": 7644
+    },
+    {
+      "epoch": 1.2232,
+      "grad_norm": 2.2228479715023868,
+      "learning_rate": 0.00013422996809638987,
+      "loss": 1.4286,
+      "step": 7645
+    },
+    {
+      "epoch": 1.22336,
+      "grad_norm": 2.6382989668007695,
+      "learning_rate": 0.00013421373724735642,
+      "loss": 1.3539,
+      "step": 7646
+    },
+    {
+      "epoch": 1.22352,
+      "grad_norm": 2.0325308239839894,
+      "learning_rate": 0.0001341975053774349,
+      "loss": 1.4051,
+      "step": 7647
+    },
+    {
+      "epoch": 1.2236799999999999,
+      "grad_norm": 2.2901991668238266,
+      "learning_rate": 0.00013418127248710965,
+      "loss": 1.4175,
+      "step": 7648
+    },
+    {
+      "epoch": 1.22384,
+      "grad_norm": 2.4477866566727298,
+      "learning_rate": 0.00013416503857686504,
+      "loss": 1.4082,
+      "step": 7649
+    },
+    {
+      "epoch": 1.224,
+      "grad_norm": 2.362183058750347,
+      "learning_rate": 0.00013414880364718545,
+      "loss": 1.3731,
+      "step": 7650
+    },
+    {
+      "epoch": 1.22416,
+      "grad_norm": 2.05991942091306,
+      "learning_rate": 0.00013413256769855534,
+      "loss": 1.3498,
+      "step": 7651
+    },
+    {
+      "epoch": 1.22432,
+      "grad_norm": 2.078081940877021,
+      "learning_rate": 0.00013411633073145913,
+      "loss": 1.4245,
+      "step": 7652
+    },
+    {
+      "epoch": 1.22448,
+      "grad_norm": 2.535081432036881,
+      "learning_rate": 0.00013410009274638133,
+      "loss": 1.3748,
+      "step": 7653
+    },
+    {
+      "epoch": 1.22464,
+      "grad_norm": 2.2603117665348442,
+      "learning_rate": 0.00013408385374380646,
+      "loss": 1.3759,
+      "step": 7654
+    },
+    {
+      "epoch": 1.2248,
+      "grad_norm": 2.0862226519841993,
+      "learning_rate": 0.00013406761372421904,
+      "loss": 1.4005,
+      "step": 7655
+    },
+    {
+      "epoch": 1.22496,
+      "grad_norm": 2.108212876396453,
+      "learning_rate": 0.00013405137268810365,
+      "loss": 1.4167,
+      "step": 7656
+    },
+    {
+      "epoch": 1.22512,
+      "grad_norm": 2.118588709897644,
+      "learning_rate": 0.00013403513063594495,
+      "loss": 1.4767,
+      "step": 7657
+    },
+    {
+      "epoch": 1.22528,
+      "grad_norm": 2.498025551918207,
+      "learning_rate": 0.00013401888756822756,
+      "loss": 1.4383,
+      "step": 7658
+    },
+    {
+      "epoch": 1.22544,
+      "grad_norm": 2.143061850068283,
+      "learning_rate": 0.00013400264348543611,
+      "loss": 1.3967,
+      "step": 7659
+    },
+    {
+      "epoch": 1.2256,
+      "grad_norm": 2.0326682348317577,
+      "learning_rate": 0.00013398639838805535,
+      "loss": 1.4755,
+      "step": 7660
+    },
+    {
+      "epoch": 1.22576,
+      "grad_norm": 2.0842406848862423,
+      "learning_rate": 0.00013397015227656995,
+      "loss": 1.3959,
+      "step": 7661
+    },
+    {
+      "epoch": 1.22592,
+      "grad_norm": 2.2117110917602267,
+      "learning_rate": 0.00013395390515146474,
+      "loss": 1.4022,
+      "step": 7662
+    },
+    {
+      "epoch": 1.22608,
+      "grad_norm": 2.395916596458019,
+      "learning_rate": 0.00013393765701322444,
+      "loss": 1.4434,
+      "step": 7663
+    },
+    {
+      "epoch": 1.22624,
+      "grad_norm": 2.0049693100384953,
+      "learning_rate": 0.00013392140786233392,
+      "loss": 1.4233,
+      "step": 7664
+    },
+    {
+      "epoch": 1.2264,
+      "grad_norm": 2.0492917499593903,
+      "learning_rate": 0.00013390515769927803,
+      "loss": 1.4259,
+      "step": 7665
+    },
+    {
+      "epoch": 1.22656,
+      "grad_norm": 2.7228959021184567,
+      "learning_rate": 0.00013388890652454163,
+      "loss": 1.4394,
+      "step": 7666
+    },
+    {
+      "epoch": 1.22672,
+      "grad_norm": 1.9387704526319685,
+      "learning_rate": 0.00013387265433860966,
+      "loss": 1.411,
+      "step": 7667
+    },
+    {
+      "epoch": 1.22688,
+      "grad_norm": 2.2906615808358466,
+      "learning_rate": 0.00013385640114196702,
+      "loss": 1.4464,
+      "step": 7668
+    },
+    {
+      "epoch": 1.22704,
+      "grad_norm": 2.3532912598256006,
+      "learning_rate": 0.00013384014693509872,
+      "loss": 1.491,
+      "step": 7669
+    },
+    {
+      "epoch": 1.2272,
+      "grad_norm": 2.246942033788655,
+      "learning_rate": 0.00013382389171848975,
+      "loss": 1.468,
+      "step": 7670
+    },
+    {
+      "epoch": 1.22736,
+      "grad_norm": 2.049360972060328,
+      "learning_rate": 0.0001338076354926251,
+      "loss": 1.3977,
+      "step": 7671
+    },
+    {
+      "epoch": 1.22752,
+      "grad_norm": 2.0574655280125764,
+      "learning_rate": 0.0001337913782579899,
+      "loss": 1.4014,
+      "step": 7672
+    },
+    {
+      "epoch": 1.2276799999999999,
+      "grad_norm": 2.234331485567469,
+      "learning_rate": 0.0001337751200150692,
+      "loss": 1.4942,
+      "step": 7673
+    },
+    {
+      "epoch": 1.22784,
+      "grad_norm": 2.2755426920563626,
+      "learning_rate": 0.00013375886076434813,
+      "loss": 1.4417,
+      "step": 7674
+    },
+    {
+      "epoch": 1.228,
+      "grad_norm": 2.0524967066479127,
+      "learning_rate": 0.00013374260050631187,
+      "loss": 1.4698,
+      "step": 7675
+    },
+    {
+      "epoch": 1.22816,
+      "grad_norm": 2.028038422101869,
+      "learning_rate": 0.00013372633924144556,
+      "loss": 1.4265,
+      "step": 7676
+    },
+    {
+      "epoch": 1.22832,
+      "grad_norm": 1.850068165975925,
+      "learning_rate": 0.00013371007697023446,
+      "loss": 1.3764,
+      "step": 7677
+    },
+    {
+      "epoch": 1.22848,
+      "grad_norm": 2.239488030280121,
+      "learning_rate": 0.00013369381369316376,
+      "loss": 1.4469,
+      "step": 7678
+    },
+    {
+      "epoch": 1.22864,
+      "grad_norm": 2.3962449263581656,
+      "learning_rate": 0.00013367754941071873,
+      "loss": 1.3345,
+      "step": 7679
+    },
+    {
+      "epoch": 1.2288000000000001,
+      "grad_norm": 2.1171231145332143,
+      "learning_rate": 0.0001336612841233847,
+      "loss": 1.4497,
+      "step": 7680
+    },
+    {
+      "epoch": 1.22896,
+      "grad_norm": 2.272698622655557,
+      "learning_rate": 0.00013364501783164705,
+      "loss": 1.3714,
+      "step": 7681
+    },
+    {
+      "epoch": 1.22912,
+      "grad_norm": 2.1881254147698774,
+      "learning_rate": 0.00013362875053599108,
+      "loss": 1.4528,
+      "step": 7682
+    },
+    {
+      "epoch": 1.22928,
+      "grad_norm": 2.5547547508407193,
+      "learning_rate": 0.00013361248223690218,
+      "loss": 1.4451,
+      "step": 7683
+    },
+    {
+      "epoch": 1.22944,
+      "grad_norm": 2.0829264723541225,
+      "learning_rate": 0.0001335962129348658,
+      "loss": 1.4935,
+      "step": 7684
+    },
+    {
+      "epoch": 1.2296,
+      "grad_norm": 2.48335730496202,
+      "learning_rate": 0.0001335799426303674,
+      "loss": 1.3708,
+      "step": 7685
+    },
+    {
+      "epoch": 1.22976,
+      "grad_norm": 1.89686458953661,
+      "learning_rate": 0.0001335636713238924,
+      "loss": 1.4477,
+      "step": 7686
+    },
+    {
+      "epoch": 1.22992,
+      "grad_norm": 2.460507538832084,
+      "learning_rate": 0.0001335473990159264,
+      "loss": 1.4049,
+      "step": 7687
+    },
+    {
+      "epoch": 1.23008,
+      "grad_norm": 2.894717359479958,
+      "learning_rate": 0.00013353112570695487,
+      "loss": 1.4746,
+      "step": 7688
+    },
+    {
+      "epoch": 1.23024,
+      "grad_norm": 2.368293401710819,
+      "learning_rate": 0.00013351485139746338,
+      "loss": 1.4807,
+      "step": 7689
+    },
+    {
+      "epoch": 1.2304,
+      "grad_norm": 2.565542566956633,
+      "learning_rate": 0.00013349857608793762,
+      "loss": 1.3355,
+      "step": 7690
+    },
+    {
+      "epoch": 1.23056,
+      "grad_norm": 2.3075157223735867,
+      "learning_rate": 0.00013348229977886312,
+      "loss": 1.4002,
+      "step": 7691
+    },
+    {
+      "epoch": 1.23072,
+      "grad_norm": 2.071897407260616,
+      "learning_rate": 0.00013346602247072558,
+      "loss": 1.4005,
+      "step": 7692
+    },
+    {
+      "epoch": 1.23088,
+      "grad_norm": 2.7979622914774387,
+      "learning_rate": 0.00013344974416401074,
+      "loss": 1.4763,
+      "step": 7693
+    },
+    {
+      "epoch": 1.23104,
+      "grad_norm": 2.121368396665138,
+      "learning_rate": 0.00013343346485920424,
+      "loss": 1.4174,
+      "step": 7694
+    },
+    {
+      "epoch": 1.2312,
+      "grad_norm": 2.290391152808065,
+      "learning_rate": 0.00013341718455679187,
+      "loss": 1.4123,
+      "step": 7695
+    },
+    {
+      "epoch": 1.23136,
+      "grad_norm": 2.073990264295193,
+      "learning_rate": 0.00013340090325725942,
+      "loss": 1.4042,
+      "step": 7696
+    },
+    {
+      "epoch": 1.23152,
+      "grad_norm": 2.1286791723684013,
+      "learning_rate": 0.00013338462096109267,
+      "loss": 1.4183,
+      "step": 7697
+    },
+    {
+      "epoch": 1.2316799999999999,
+      "grad_norm": 2.421020991531808,
+      "learning_rate": 0.0001333683376687775,
+      "loss": 1.3676,
+      "step": 7698
+    },
+    {
+      "epoch": 1.23184,
+      "grad_norm": 2.1614322654333384,
+      "learning_rate": 0.00013335205338079975,
+      "loss": 1.4005,
+      "step": 7699
+    },
+    {
+      "epoch": 1.232,
+      "grad_norm": 2.153332357774346,
+      "learning_rate": 0.0001333357680976453,
+      "loss": 1.4476,
+      "step": 7700
+    },
+    {
+      "epoch": 1.23216,
+      "grad_norm": 2.3311454628199466,
+      "learning_rate": 0.0001333194818198001,
+      "loss": 1.387,
+      "step": 7701
+    },
+    {
+      "epoch": 1.23232,
+      "grad_norm": 2.131773687778668,
+      "learning_rate": 0.00013330319454775018,
+      "loss": 1.3798,
+      "step": 7702
+    },
+    {
+      "epoch": 1.23248,
+      "grad_norm": 1.87547017483573,
+      "learning_rate": 0.0001332869062819814,
+      "loss": 1.3485,
+      "step": 7703
+    },
+    {
+      "epoch": 1.23264,
+      "grad_norm": 2.0994379096655744,
+      "learning_rate": 0.00013327061702297986,
+      "loss": 1.4061,
+      "step": 7704
+    },
+    {
+      "epoch": 1.2328000000000001,
+      "grad_norm": 1.8780727094120675,
+      "learning_rate": 0.0001332543267712316,
+      "loss": 1.4323,
+      "step": 7705
+    },
+    {
+      "epoch": 1.23296,
+      "grad_norm": 2.356831606020507,
+      "learning_rate": 0.00013323803552722267,
+      "loss": 1.4503,
+      "step": 7706
+    },
+    {
+      "epoch": 1.23312,
+      "grad_norm": 2.465463839390504,
+      "learning_rate": 0.00013322174329143918,
+      "loss": 1.4046,
+      "step": 7707
+    },
+    {
+      "epoch": 1.23328,
+      "grad_norm": 2.3436094530626943,
+      "learning_rate": 0.00013320545006436732,
+      "loss": 1.4676,
+      "step": 7708
+    },
+    {
+      "epoch": 1.23344,
+      "grad_norm": 2.0539710065951122,
+      "learning_rate": 0.0001331891558464932,
+      "loss": 1.4427,
+      "step": 7709
+    },
+    {
+      "epoch": 1.2336,
+      "grad_norm": 2.1112651777108544,
+      "learning_rate": 0.00013317286063830303,
+      "loss": 1.3978,
+      "step": 7710
+    },
+    {
+      "epoch": 1.23376,
+      "grad_norm": 2.694314968087416,
+      "learning_rate": 0.00013315656444028303,
+      "loss": 1.3751,
+      "step": 7711
+    },
+    {
+      "epoch": 1.23392,
+      "grad_norm": 2.104914557514399,
+      "learning_rate": 0.00013314026725291946,
+      "loss": 1.4136,
+      "step": 7712
+    },
+    {
+      "epoch": 1.23408,
+      "grad_norm": 2.420962770975035,
+      "learning_rate": 0.0001331239690766986,
+      "loss": 1.4181,
+      "step": 7713
+    },
+    {
+      "epoch": 1.23424,
+      "grad_norm": 2.4149285915013077,
+      "learning_rate": 0.00013310766991210676,
+      "loss": 1.5031,
+      "step": 7714
+    },
+    {
+      "epoch": 1.2344,
+      "grad_norm": 2.3904214005143904,
+      "learning_rate": 0.00013309136975963033,
+      "loss": 1.4036,
+      "step": 7715
+    },
+    {
+      "epoch": 1.23456,
+      "grad_norm": 2.4216481056312187,
+      "learning_rate": 0.00013307506861975563,
+      "loss": 1.4905,
+      "step": 7716
+    },
+    {
+      "epoch": 1.23472,
+      "grad_norm": 2.1782839154217672,
+      "learning_rate": 0.00013305876649296908,
+      "loss": 1.4417,
+      "step": 7717
+    },
+    {
+      "epoch": 1.23488,
+      "grad_norm": 2.687129239501883,
+      "learning_rate": 0.0001330424633797571,
+      "loss": 1.4436,
+      "step": 7718
+    },
+    {
+      "epoch": 1.23504,
+      "grad_norm": 2.2392048107363607,
+      "learning_rate": 0.00013302615928060618,
+      "loss": 1.3654,
+      "step": 7719
+    },
+    {
+      "epoch": 1.2352,
+      "grad_norm": 2.0043046193656537,
+      "learning_rate": 0.00013300985419600276,
+      "loss": 1.4726,
+      "step": 7720
+    },
+    {
+      "epoch": 1.23536,
+      "grad_norm": 2.5901580747192043,
+      "learning_rate": 0.0001329935481264334,
+      "loss": 1.3851,
+      "step": 7721
+    },
+    {
+      "epoch": 1.23552,
+      "grad_norm": 2.212405374849984,
+      "learning_rate": 0.00013297724107238465,
+      "loss": 1.3821,
+      "step": 7722
+    },
+    {
+      "epoch": 1.23568,
+      "grad_norm": 2.1498699992485895,
+      "learning_rate": 0.00013296093303434307,
+      "loss": 1.4609,
+      "step": 7723
+    },
+    {
+      "epoch": 1.23584,
+      "grad_norm": 2.233007716332051,
+      "learning_rate": 0.00013294462401279531,
+      "loss": 1.489,
+      "step": 7724
+    },
+    {
+      "epoch": 1.236,
+      "grad_norm": 2.2153646151828212,
+      "learning_rate": 0.00013292831400822793,
+      "loss": 1.3752,
+      "step": 7725
+    },
+    {
+      "epoch": 1.23616,
+      "grad_norm": 2.1625852373867542,
+      "learning_rate": 0.00013291200302112768,
+      "loss": 1.5539,
+      "step": 7726
+    },
+    {
+      "epoch": 1.23632,
+      "grad_norm": 2.4673397911993753,
+      "learning_rate": 0.0001328956910519812,
+      "loss": 1.3425,
+      "step": 7727
+    },
+    {
+      "epoch": 1.23648,
+      "grad_norm": 2.0732790714343436,
+      "learning_rate": 0.0001328793781012752,
+      "loss": 1.4591,
+      "step": 7728
+    },
+    {
+      "epoch": 1.23664,
+      "grad_norm": 2.48312357219294,
+      "learning_rate": 0.0001328630641694965,
+      "loss": 1.4159,
+      "step": 7729
+    },
+    {
+      "epoch": 1.2368000000000001,
+      "grad_norm": 2.224612176769772,
+      "learning_rate": 0.00013284674925713186,
+      "loss": 1.4404,
+      "step": 7730
+    },
+    {
+      "epoch": 1.23696,
+      "grad_norm": 2.0900209327816364,
+      "learning_rate": 0.00013283043336466809,
+      "loss": 1.4419,
+      "step": 7731
+    },
+    {
+      "epoch": 1.23712,
+      "grad_norm": 1.9346359118656802,
+      "learning_rate": 0.000132814116492592,
+      "loss": 1.4464,
+      "step": 7732
+    },
+    {
+      "epoch": 1.23728,
+      "grad_norm": 2.243022317949283,
+      "learning_rate": 0.0001327977986413905,
+      "loss": 1.4553,
+      "step": 7733
+    },
+    {
+      "epoch": 1.23744,
+      "grad_norm": 2.1383135918024236,
+      "learning_rate": 0.0001327814798115505,
+      "loss": 1.4323,
+      "step": 7734
+    },
+    {
+      "epoch": 1.2376,
+      "grad_norm": 2.050856683584092,
+      "learning_rate": 0.00013276516000355887,
+      "loss": 1.3905,
+      "step": 7735
+    },
+    {
+      "epoch": 1.23776,
+      "grad_norm": 2.218489526113882,
+      "learning_rate": 0.00013274883921790263,
+      "loss": 1.3911,
+      "step": 7736
+    },
+    {
+      "epoch": 1.23792,
+      "grad_norm": 1.9637622155059127,
+      "learning_rate": 0.00013273251745506877,
+      "loss": 1.4355,
+      "step": 7737
+    },
+    {
+      "epoch": 1.23808,
+      "grad_norm": 2.2393250191076945,
+      "learning_rate": 0.00013271619471554423,
+      "loss": 1.4201,
+      "step": 7738
+    },
+    {
+      "epoch": 1.23824,
+      "grad_norm": 2.1643759582510906,
+      "learning_rate": 0.00013269987099981615,
+      "loss": 1.4573,
+      "step": 7739
+    },
+    {
+      "epoch": 1.2384,
+      "grad_norm": 2.1520055658434316,
+      "learning_rate": 0.00013268354630837156,
+      "loss": 1.3767,
+      "step": 7740
+    },
+    {
+      "epoch": 1.23856,
+      "grad_norm": 2.351433438423276,
+      "learning_rate": 0.0001326672206416976,
+      "loss": 1.4069,
+      "step": 7741
+    },
+    {
+      "epoch": 1.23872,
+      "grad_norm": 2.1600733919322104,
+      "learning_rate": 0.00013265089400028135,
+      "loss": 1.4229,
+      "step": 7742
+    },
+    {
+      "epoch": 1.23888,
+      "grad_norm": 2.45426911077364,
+      "learning_rate": 0.00013263456638461,
+      "loss": 1.5655,
+      "step": 7743
+    },
+    {
+      "epoch": 1.23904,
+      "grad_norm": 2.3351664047284237,
+      "learning_rate": 0.00013261823779517074,
+      "loss": 1.3341,
+      "step": 7744
+    },
+    {
+      "epoch": 1.2392,
+      "grad_norm": 1.9763029114694965,
+      "learning_rate": 0.00013260190823245082,
+      "loss": 1.3612,
+      "step": 7745
+    },
+    {
+      "epoch": 1.23936,
+      "grad_norm": 2.471847254473668,
+      "learning_rate": 0.00013258557769693744,
+      "loss": 1.4225,
+      "step": 7746
+    },
+    {
+      "epoch": 1.23952,
+      "grad_norm": 2.134307443508362,
+      "learning_rate": 0.0001325692461891179,
+      "loss": 1.3784,
+      "step": 7747
+    },
+    {
+      "epoch": 1.23968,
+      "grad_norm": 2.926249611375558,
+      "learning_rate": 0.00013255291370947952,
+      "loss": 1.4528,
+      "step": 7748
+    },
+    {
+      "epoch": 1.23984,
+      "grad_norm": 2.280037287974474,
+      "learning_rate": 0.00013253658025850964,
+      "loss": 1.388,
+      "step": 7749
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 3.5280329626933886,
+      "learning_rate": 0.00013252024583669563,
+      "loss": 1.4666,
+      "step": 7750
+    },
+    {
+      "epoch": 1.24016,
+      "grad_norm": 2.413796737736978,
+      "learning_rate": 0.00013250391044452483,
+      "loss": 1.3479,
+      "step": 7751
+    },
+    {
+      "epoch": 1.24032,
+      "grad_norm": 2.579068548740689,
+      "learning_rate": 0.00013248757408248474,
+      "loss": 1.4364,
+      "step": 7752
+    },
+    {
+      "epoch": 1.24048,
+      "grad_norm": 2.286534438282327,
+      "learning_rate": 0.00013247123675106276,
+      "loss": 1.4383,
+      "step": 7753
+    },
+    {
+      "epoch": 1.24064,
+      "grad_norm": 2.897176621338049,
+      "learning_rate": 0.0001324548984507464,
+      "loss": 1.4341,
+      "step": 7754
+    },
+    {
+      "epoch": 1.2408,
+      "grad_norm": 2.280565159012832,
+      "learning_rate": 0.00013243855918202314,
+      "loss": 1.4151,
+      "step": 7755
+    },
+    {
+      "epoch": 1.24096,
+      "grad_norm": 2.515176997767578,
+      "learning_rate": 0.00013242221894538056,
+      "loss": 1.4136,
+      "step": 7756
+    },
+    {
+      "epoch": 1.24112,
+      "grad_norm": 2.369625822466041,
+      "learning_rate": 0.00013240587774130622,
+      "loss": 1.3983,
+      "step": 7757
+    },
+    {
+      "epoch": 1.24128,
+      "grad_norm": 2.2065657516669974,
+      "learning_rate": 0.0001323895355702877,
+      "loss": 1.4137,
+      "step": 7758
+    },
+    {
+      "epoch": 1.2414399999999999,
+      "grad_norm": 2.3042392651448234,
+      "learning_rate": 0.00013237319243281263,
+      "loss": 1.3571,
+      "step": 7759
+    },
+    {
+      "epoch": 1.2416,
+      "grad_norm": 1.9615280339213668,
+      "learning_rate": 0.00013235684832936867,
+      "loss": 1.3821,
+      "step": 7760
+    },
+    {
+      "epoch": 1.24176,
+      "grad_norm": 1.893064575485609,
+      "learning_rate": 0.00013234050326044351,
+      "loss": 1.4643,
+      "step": 7761
+    },
+    {
+      "epoch": 1.24192,
+      "grad_norm": 2.7711178106618886,
+      "learning_rate": 0.00013232415722652484,
+      "loss": 1.4807,
+      "step": 7762
+    },
+    {
+      "epoch": 1.24208,
+      "grad_norm": 2.134341481018179,
+      "learning_rate": 0.00013230781022810042,
+      "loss": 1.4105,
+      "step": 7763
+    },
+    {
+      "epoch": 1.24224,
+      "grad_norm": 2.240002709270841,
+      "learning_rate": 0.00013229146226565808,
+      "loss": 1.3292,
+      "step": 7764
+    },
+    {
+      "epoch": 1.2424,
+      "grad_norm": 2.314567622143494,
+      "learning_rate": 0.0001322751133396855,
+      "loss": 1.3706,
+      "step": 7765
+    },
+    {
+      "epoch": 1.24256,
+      "grad_norm": 2.579411207859892,
+      "learning_rate": 0.00013225876345067058,
+      "loss": 1.4341,
+      "step": 7766
+    },
+    {
+      "epoch": 1.24272,
+      "grad_norm": 2.060276387710464,
+      "learning_rate": 0.00013224241259910118,
+      "loss": 1.4795,
+      "step": 7767
+    },
+    {
+      "epoch": 1.24288,
+      "grad_norm": 2.8072704433922038,
+      "learning_rate": 0.00013222606078546517,
+      "loss": 1.4331,
+      "step": 7768
+    },
+    {
+      "epoch": 1.24304,
+      "grad_norm": 2.8973485039027795,
+      "learning_rate": 0.00013220970801025046,
+      "loss": 1.483,
+      "step": 7769
+    },
+    {
+      "epoch": 1.2432,
+      "grad_norm": 2.864432727952955,
+      "learning_rate": 0.00013219335427394496,
+      "loss": 1.3838,
+      "step": 7770
+    },
+    {
+      "epoch": 1.24336,
+      "grad_norm": 2.5731080713261143,
+      "learning_rate": 0.00013217699957703672,
+      "loss": 1.4024,
+      "step": 7771
+    },
+    {
+      "epoch": 1.24352,
+      "grad_norm": 2.201868956634687,
+      "learning_rate": 0.00013216064392001367,
+      "loss": 1.4122,
+      "step": 7772
+    },
+    {
+      "epoch": 1.24368,
+      "grad_norm": 2.602933980785881,
+      "learning_rate": 0.0001321442873033639,
+      "loss": 1.4412,
+      "step": 7773
+    },
+    {
+      "epoch": 1.24384,
+      "grad_norm": 2.104609085287344,
+      "learning_rate": 0.00013212792972757545,
+      "loss": 1.3463,
+      "step": 7774
+    },
+    {
+      "epoch": 1.244,
+      "grad_norm": 2.0631654071667604,
+      "learning_rate": 0.00013211157119313635,
+      "loss": 1.4538,
+      "step": 7775
+    },
+    {
+      "epoch": 1.24416,
+      "grad_norm": 2.309916374459807,
+      "learning_rate": 0.00013209521170053478,
+      "loss": 1.4182,
+      "step": 7776
+    },
+    {
+      "epoch": 1.24432,
+      "grad_norm": 2.1441049687109386,
+      "learning_rate": 0.00013207885125025886,
+      "loss": 1.394,
+      "step": 7777
+    },
+    {
+      "epoch": 1.24448,
+      "grad_norm": 2.01881818655976,
+      "learning_rate": 0.00013206248984279673,
+      "loss": 1.4403,
+      "step": 7778
+    },
+    {
+      "epoch": 1.24464,
+      "grad_norm": 2.2926168020183146,
+      "learning_rate": 0.00013204612747863664,
+      "loss": 1.4146,
+      "step": 7779
+    },
+    {
+      "epoch": 1.2448,
+      "grad_norm": 2.1905962861582204,
+      "learning_rate": 0.00013202976415826683,
+      "loss": 1.4021,
+      "step": 7780
+    },
+    {
+      "epoch": 1.24496,
+      "grad_norm": 2.130558361994742,
+      "learning_rate": 0.00013201339988217552,
+      "loss": 1.3999,
+      "step": 7781
+    },
+    {
+      "epoch": 1.24512,
+      "grad_norm": 2.238201097611326,
+      "learning_rate": 0.000131997034650851,
+      "loss": 1.3651,
+      "step": 7782
+    },
+    {
+      "epoch": 1.24528,
+      "grad_norm": 2.2525736034570643,
+      "learning_rate": 0.0001319806684647816,
+      "loss": 1.4903,
+      "step": 7783
+    },
+    {
+      "epoch": 1.2454399999999999,
+      "grad_norm": 2.2912668375257716,
+      "learning_rate": 0.00013196430132445563,
+      "loss": 1.4127,
+      "step": 7784
+    },
+    {
+      "epoch": 1.2456,
+      "grad_norm": 2.7042667864892933,
+      "learning_rate": 0.0001319479332303615,
+      "loss": 1.462,
+      "step": 7785
+    },
+    {
+      "epoch": 1.24576,
+      "grad_norm": 2.070939095034588,
+      "learning_rate": 0.00013193156418298758,
+      "loss": 1.4562,
+      "step": 7786
+    },
+    {
+      "epoch": 1.24592,
+      "grad_norm": 2.302153588042742,
+      "learning_rate": 0.00013191519418282233,
+      "loss": 1.417,
+      "step": 7787
+    },
+    {
+      "epoch": 1.24608,
+      "grad_norm": 2.2861793496449025,
+      "learning_rate": 0.0001318988232303542,
+      "loss": 1.3199,
+      "step": 7788
+    },
+    {
+      "epoch": 1.24624,
+      "grad_norm": 1.864285426840154,
+      "learning_rate": 0.00013188245132607164,
+      "loss": 1.3317,
+      "step": 7789
+    },
+    {
+      "epoch": 1.2464,
+      "grad_norm": 2.141702022474719,
+      "learning_rate": 0.0001318660784704632,
+      "loss": 1.4031,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2465600000000001,
+      "grad_norm": 3.2510165266763043,
+      "learning_rate": 0.00013184970466401742,
+      "loss": 1.4986,
+      "step": 7791
+    },
+    {
+      "epoch": 1.24672,
+      "grad_norm": 1.8792915768609748,
+      "learning_rate": 0.00013183332990722286,
+      "loss": 1.4077,
+      "step": 7792
+    },
+    {
+      "epoch": 1.24688,
+      "grad_norm": 2.7748765780621882,
+      "learning_rate": 0.00013181695420056813,
+      "loss": 1.3972,
+      "step": 7793
+    },
+    {
+      "epoch": 1.24704,
+      "grad_norm": 2.079892238546015,
+      "learning_rate": 0.00013180057754454182,
+      "loss": 1.4066,
+      "step": 7794
+    },
+    {
+      "epoch": 1.2472,
+      "grad_norm": 2.8921061343388437,
+      "learning_rate": 0.00013178419993963264,
+      "loss": 1.3784,
+      "step": 7795
+    },
+    {
+      "epoch": 1.24736,
+      "grad_norm": 2.082661868182034,
+      "learning_rate": 0.00013176782138632923,
+      "loss": 1.3421,
+      "step": 7796
+    },
+    {
+      "epoch": 1.24752,
+      "grad_norm": 2.368579124928833,
+      "learning_rate": 0.0001317514418851203,
+      "loss": 1.4378,
+      "step": 7797
+    },
+    {
+      "epoch": 1.24768,
+      "grad_norm": 2.0368707522373315,
+      "learning_rate": 0.00013173506143649468,
+      "loss": 1.3599,
+      "step": 7798
+    },
+    {
+      "epoch": 1.24784,
+      "grad_norm": 2.237340962512082,
+      "learning_rate": 0.00013171868004094103,
+      "loss": 1.3541,
+      "step": 7799
+    },
+    {
+      "epoch": 1.248,
+      "grad_norm": 1.8998706555518012,
+      "learning_rate": 0.0001317022976989482,
+      "loss": 1.3888,
+      "step": 7800
+    },
+    {
+      "epoch": 1.24816,
+      "grad_norm": 1.796289726830649,
+      "learning_rate": 0.00013168591441100496,
+      "loss": 1.3566,
+      "step": 7801
+    },
+    {
+      "epoch": 1.24832,
+      "grad_norm": 2.2287412975171144,
+      "learning_rate": 0.00013166953017760026,
+      "loss": 1.4963,
+      "step": 7802
+    },
+    {
+      "epoch": 1.24848,
+      "grad_norm": 1.9976134174038456,
+      "learning_rate": 0.00013165314499922288,
+      "loss": 1.3973,
+      "step": 7803
+    },
+    {
+      "epoch": 1.24864,
+      "grad_norm": 1.8391898468369474,
+      "learning_rate": 0.0001316367588763618,
+      "loss": 1.4457,
+      "step": 7804
+    },
+    {
+      "epoch": 1.2488,
+      "grad_norm": 1.8987669455783325,
+      "learning_rate": 0.00013162037180950594,
+      "loss": 1.4506,
+      "step": 7805
+    },
+    {
+      "epoch": 1.24896,
+      "grad_norm": 2.1698989923289553,
+      "learning_rate": 0.00013160398379914428,
+      "loss": 1.3958,
+      "step": 7806
+    },
+    {
+      "epoch": 1.24912,
+      "grad_norm": 2.254099815229264,
+      "learning_rate": 0.00013158759484576576,
+      "loss": 1.3688,
+      "step": 7807
+    },
+    {
+      "epoch": 1.24928,
+      "grad_norm": 2.211478159497754,
+      "learning_rate": 0.00013157120494985947,
+      "loss": 1.4547,
+      "step": 7808
+    },
+    {
+      "epoch": 1.2494399999999999,
+      "grad_norm": 2.223249868253943,
+      "learning_rate": 0.0001315548141119144,
+      "loss": 1.4399,
+      "step": 7809
+    },
+    {
+      "epoch": 1.2496,
+      "grad_norm": 1.9968372680150779,
+      "learning_rate": 0.00013153842233241963,
+      "loss": 1.4102,
+      "step": 7810
+    },
+    {
+      "epoch": 1.24976,
+      "grad_norm": 1.8913487914843634,
+      "learning_rate": 0.00013152202961186433,
+      "loss": 1.4003,
+      "step": 7811
+    },
+    {
+      "epoch": 1.24992,
+      "grad_norm": 2.2242427866961423,
+      "learning_rate": 0.00013150563595073762,
+      "loss": 1.4571,
+      "step": 7812
+    },
+    {
+      "epoch": 1.25008,
+      "grad_norm": 2.0416361203675812,
+      "learning_rate": 0.0001314892413495286,
+      "loss": 1.396,
+      "step": 7813
+    },
+    {
+      "epoch": 1.25024,
+      "grad_norm": 1.9751201484614265,
+      "learning_rate": 0.00013147284580872651,
+      "loss": 1.3771,
+      "step": 7814
+    },
+    {
+      "epoch": 1.2504,
+      "grad_norm": 3.162906848977576,
+      "learning_rate": 0.0001314564493288206,
+      "loss": 1.3854,
+      "step": 7815
+    },
+    {
+      "epoch": 1.2505600000000001,
+      "grad_norm": 1.9117865679605972,
+      "learning_rate": 0.00013144005191030006,
+      "loss": 1.4148,
+      "step": 7816
+    },
+    {
+      "epoch": 1.25072,
+      "grad_norm": 2.5657401182057935,
+      "learning_rate": 0.00013142365355365415,
+      "loss": 1.4321,
+      "step": 7817
+    },
+    {
+      "epoch": 1.25088,
+      "grad_norm": 2.44521835886421,
+      "learning_rate": 0.00013140725425937226,
+      "loss": 1.4561,
+      "step": 7818
+    },
+    {
+      "epoch": 1.25104,
+      "grad_norm": 1.99916589882253,
+      "learning_rate": 0.00013139085402794363,
+      "loss": 1.4145,
+      "step": 7819
+    },
+    {
+      "epoch": 1.2511999999999999,
+      "grad_norm": 2.4773703381839174,
+      "learning_rate": 0.00013137445285985765,
+      "loss": 1.3798,
+      "step": 7820
+    },
+    {
+      "epoch": 1.25136,
+      "grad_norm": 2.247784039919003,
+      "learning_rate": 0.00013135805075560375,
+      "loss": 1.3902,
+      "step": 7821
+    },
+    {
+      "epoch": 1.25152,
+      "grad_norm": 1.995888790626788,
+      "learning_rate": 0.0001313416477156713,
+      "loss": 1.4541,
+      "step": 7822
+    },
+    {
+      "epoch": 1.25168,
+      "grad_norm": 2.6686984522310717,
+      "learning_rate": 0.00013132524374054975,
+      "loss": 1.5157,
+      "step": 7823
+    },
+    {
+      "epoch": 1.25184,
+      "grad_norm": 2.2980982955361577,
+      "learning_rate": 0.00013130883883072855,
+      "loss": 1.4125,
+      "step": 7824
+    },
+    {
+      "epoch": 1.252,
+      "grad_norm": 2.4560968184151837,
+      "learning_rate": 0.00013129243298669725,
+      "loss": 1.4838,
+      "step": 7825
+    },
+    {
+      "epoch": 1.25216,
+      "grad_norm": 2.3479958992098493,
+      "learning_rate": 0.00013127602620894533,
+      "loss": 1.4733,
+      "step": 7826
+    },
+    {
+      "epoch": 1.25232,
+      "grad_norm": 2.2425738657888235,
+      "learning_rate": 0.00013125961849796234,
+      "loss": 1.3737,
+      "step": 7827
+    },
+    {
+      "epoch": 1.25248,
+      "grad_norm": 2.15246451543923,
+      "learning_rate": 0.00013124320985423793,
+      "loss": 1.358,
+      "step": 7828
+    },
+    {
+      "epoch": 1.25264,
+      "grad_norm": 2.471378046701053,
+      "learning_rate": 0.00013122680027826165,
+      "loss": 1.5019,
+      "step": 7829
+    },
+    {
+      "epoch": 1.2528000000000001,
+      "grad_norm": 2.2843999759886926,
+      "learning_rate": 0.00013121038977052316,
+      "loss": 1.4659,
+      "step": 7830
+    },
+    {
+      "epoch": 1.25296,
+      "grad_norm": 2.3453953477887413,
+      "learning_rate": 0.0001311939783315121,
+      "loss": 1.3882,
+      "step": 7831
+    },
+    {
+      "epoch": 1.25312,
+      "grad_norm": 2.0306392422659343,
+      "learning_rate": 0.00013117756596171817,
+      "loss": 1.475,
+      "step": 7832
+    },
+    {
+      "epoch": 1.25328,
+      "grad_norm": 2.1517346045697967,
+      "learning_rate": 0.00013116115266163114,
+      "loss": 1.3707,
+      "step": 7833
+    },
+    {
+      "epoch": 1.2534399999999999,
+      "grad_norm": 2.3306630032530946,
+      "learning_rate": 0.0001311447384317407,
+      "loss": 1.401,
+      "step": 7834
+    },
+    {
+      "epoch": 1.2536,
+      "grad_norm": 2.5818647891391158,
+      "learning_rate": 0.0001311283232725366,
+      "loss": 1.3564,
+      "step": 7835
+    },
+    {
+      "epoch": 1.25376,
+      "grad_norm": 2.305423256803906,
+      "learning_rate": 0.00013111190718450877,
+      "loss": 1.4098,
+      "step": 7836
+    },
+    {
+      "epoch": 1.25392,
+      "grad_norm": 2.8544174767391586,
+      "learning_rate": 0.00013109549016814692,
+      "loss": 1.3864,
+      "step": 7837
+    },
+    {
+      "epoch": 1.25408,
+      "grad_norm": 2.250988250048368,
+      "learning_rate": 0.00013107907222394098,
+      "loss": 1.4498,
+      "step": 7838
+    },
+    {
+      "epoch": 1.25424,
+      "grad_norm": 2.1099932283128116,
+      "learning_rate": 0.00013106265335238078,
+      "loss": 1.4012,
+      "step": 7839
+    },
+    {
+      "epoch": 1.2544,
+      "grad_norm": 2.0299378500717777,
+      "learning_rate": 0.00013104623355395632,
+      "loss": 1.428,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2545600000000001,
+      "grad_norm": 2.342440623860813,
+      "learning_rate": 0.00013102981282915745,
+      "loss": 1.408,
+      "step": 7841
+    },
+    {
+      "epoch": 1.25472,
+      "grad_norm": 2.485270891901526,
+      "learning_rate": 0.0001310133911784742,
+      "loss": 1.4543,
+      "step": 7842
+    },
+    {
+      "epoch": 1.25488,
+      "grad_norm": 2.369631585043099,
+      "learning_rate": 0.0001309969686023965,
+      "loss": 1.442,
+      "step": 7843
+    },
+    {
+      "epoch": 1.25504,
+      "grad_norm": 2.601552257243819,
+      "learning_rate": 0.00013098054510141447,
+      "loss": 1.4102,
+      "step": 7844
+    },
+    {
+      "epoch": 1.2551999999999999,
+      "grad_norm": 2.1647282583403,
+      "learning_rate": 0.0001309641206760181,
+      "loss": 1.4576,
+      "step": 7845
+    },
+    {
+      "epoch": 1.25536,
+      "grad_norm": 2.591738304813205,
+      "learning_rate": 0.0001309476953266975,
+      "loss": 1.4088,
+      "step": 7846
+    },
+    {
+      "epoch": 1.25552,
+      "grad_norm": 2.4105492012372376,
+      "learning_rate": 0.00013093126905394276,
+      "loss": 1.4449,
+      "step": 7847
+    },
+    {
+      "epoch": 1.25568,
+      "grad_norm": 1.9163374917252693,
+      "learning_rate": 0.00013091484185824402,
+      "loss": 1.3371,
+      "step": 7848
+    },
+    {
+      "epoch": 1.25584,
+      "grad_norm": 2.017317508864005,
+      "learning_rate": 0.00013089841374009148,
+      "loss": 1.3823,
+      "step": 7849
+    },
+    {
+      "epoch": 1.256,
+      "grad_norm": 2.1174086194170765,
+      "learning_rate": 0.00013088198469997527,
+      "loss": 1.4085,
+      "step": 7850
+    },
+    {
+      "epoch": 1.25616,
+      "grad_norm": 2.2706819424460667,
+      "learning_rate": 0.00013086555473838563,
+      "loss": 1.4422,
+      "step": 7851
+    },
+    {
+      "epoch": 1.25632,
+      "grad_norm": 2.373546165721031,
+      "learning_rate": 0.0001308491238558128,
+      "loss": 1.4143,
+      "step": 7852
+    },
+    {
+      "epoch": 1.25648,
+      "grad_norm": 1.9871145664183456,
+      "learning_rate": 0.00013083269205274708,
+      "loss": 1.353,
+      "step": 7853
+    },
+    {
+      "epoch": 1.25664,
+      "grad_norm": 2.9459133050430606,
+      "learning_rate": 0.0001308162593296788,
+      "loss": 1.4073,
+      "step": 7854
+    },
+    {
+      "epoch": 1.2568,
+      "grad_norm": 2.543082381747955,
+      "learning_rate": 0.00013079982568709817,
+      "loss": 1.476,
+      "step": 7855
+    },
+    {
+      "epoch": 1.25696,
+      "grad_norm": 2.3868145611654494,
+      "learning_rate": 0.00013078339112549566,
+      "loss": 1.386,
+      "step": 7856
+    },
+    {
+      "epoch": 1.25712,
+      "grad_norm": 2.2539641374012716,
+      "learning_rate": 0.00013076695564536163,
+      "loss": 1.382,
+      "step": 7857
+    },
+    {
+      "epoch": 1.25728,
+      "grad_norm": 1.9840445407380443,
+      "learning_rate": 0.00013075051924718646,
+      "loss": 1.3274,
+      "step": 7858
+    },
+    {
+      "epoch": 1.25744,
+      "grad_norm": 2.483311935426607,
+      "learning_rate": 0.00013073408193146058,
+      "loss": 1.4266,
+      "step": 7859
+    },
+    {
+      "epoch": 1.2576,
+      "grad_norm": 2.238693914844637,
+      "learning_rate": 0.00013071764369867452,
+      "loss": 1.4542,
+      "step": 7860
+    },
+    {
+      "epoch": 1.25776,
+      "grad_norm": 2.311139461873385,
+      "learning_rate": 0.0001307012045493187,
+      "loss": 1.3869,
+      "step": 7861
+    },
+    {
+      "epoch": 1.25792,
+      "grad_norm": 2.7353461668485815,
+      "learning_rate": 0.0001306847644838837,
+      "loss": 1.3827,
+      "step": 7862
+    },
+    {
+      "epoch": 1.25808,
+      "grad_norm": 2.2216178739314176,
+      "learning_rate": 0.00013066832350286001,
+      "loss": 1.4386,
+      "step": 7863
+    },
+    {
+      "epoch": 1.25824,
+      "grad_norm": 2.435317846385179,
+      "learning_rate": 0.00013065188160673826,
+      "loss": 1.4205,
+      "step": 7864
+    },
+    {
+      "epoch": 1.2584,
+      "grad_norm": 2.4644355386307417,
+      "learning_rate": 0.00013063543879600903,
+      "loss": 1.4212,
+      "step": 7865
+    },
+    {
+      "epoch": 1.2585600000000001,
+      "grad_norm": 2.249135526359335,
+      "learning_rate": 0.00013061899507116296,
+      "loss": 1.3989,
+      "step": 7866
+    },
+    {
+      "epoch": 1.25872,
+      "grad_norm": 2.6993586744982414,
+      "learning_rate": 0.00013060255043269065,
+      "loss": 1.4187,
+      "step": 7867
+    },
+    {
+      "epoch": 1.25888,
+      "grad_norm": 2.6159332197471366,
+      "learning_rate": 0.00013058610488108288,
+      "loss": 1.361,
+      "step": 7868
+    },
+    {
+      "epoch": 1.25904,
+      "grad_norm": 2.1976584002656434,
+      "learning_rate": 0.00013056965841683026,
+      "loss": 1.4112,
+      "step": 7869
+    },
+    {
+      "epoch": 1.2591999999999999,
+      "grad_norm": 2.670785339966677,
+      "learning_rate": 0.0001305532110404236,
+      "loss": 1.4078,
+      "step": 7870
+    },
+    {
+      "epoch": 1.25936,
+      "grad_norm": 1.961242589185436,
+      "learning_rate": 0.00013053676275235363,
+      "loss": 1.3647,
+      "step": 7871
+    },
+    {
+      "epoch": 1.25952,
+      "grad_norm": 2.1246625402993358,
+      "learning_rate": 0.00013052031355311117,
+      "loss": 1.4008,
+      "step": 7872
+    },
+    {
+      "epoch": 1.25968,
+      "grad_norm": 2.2500590166007584,
+      "learning_rate": 0.000130503863443187,
+      "loss": 1.3724,
+      "step": 7873
+    },
+    {
+      "epoch": 1.25984,
+      "grad_norm": 2.1901821994181856,
+      "learning_rate": 0.00013048741242307203,
+      "loss": 1.4076,
+      "step": 7874
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 1.934268306236539,
+      "learning_rate": 0.00013047096049325708,
+      "loss": 1.3785,
+      "step": 7875
+    },
+    {
+      "epoch": 1.26016,
+      "grad_norm": 2.01330848238121,
+      "learning_rate": 0.00013045450765423308,
+      "loss": 1.4318,
+      "step": 7876
+    },
+    {
+      "epoch": 1.26032,
+      "grad_norm": 2.4400412134089717,
+      "learning_rate": 0.00013043805390649093,
+      "loss": 1.3863,
+      "step": 7877
+    },
+    {
+      "epoch": 1.26048,
+      "grad_norm": 2.5229788541160714,
+      "learning_rate": 0.00013042159925052163,
+      "loss": 1.4301,
+      "step": 7878
+    },
+    {
+      "epoch": 1.26064,
+      "grad_norm": 2.2916964720280486,
+      "learning_rate": 0.0001304051436868161,
+      "loss": 1.4286,
+      "step": 7879
+    },
+    {
+      "epoch": 1.2608,
+      "grad_norm": 2.0609603910906436,
+      "learning_rate": 0.00013038868721586543,
+      "loss": 1.364,
+      "step": 7880
+    },
+    {
+      "epoch": 1.26096,
+      "grad_norm": 2.0771985224521865,
+      "learning_rate": 0.0001303722298381606,
+      "loss": 1.4291,
+      "step": 7881
+    },
+    {
+      "epoch": 1.26112,
+      "grad_norm": 2.911214963187087,
+      "learning_rate": 0.0001303557715541927,
+      "loss": 1.4221,
+      "step": 7882
+    },
+    {
+      "epoch": 1.26128,
+      "grad_norm": 2.7524796947090278,
+      "learning_rate": 0.00013033931236445279,
+      "loss": 1.4659,
+      "step": 7883
+    },
+    {
+      "epoch": 1.26144,
+      "grad_norm": 2.802011935631012,
+      "learning_rate": 0.000130322852269432,
+      "loss": 1.4163,
+      "step": 7884
+    },
+    {
+      "epoch": 1.2616,
+      "grad_norm": 2.2144270380513524,
+      "learning_rate": 0.0001303063912696215,
+      "loss": 1.3556,
+      "step": 7885
+    },
+    {
+      "epoch": 1.26176,
+      "grad_norm": 2.2908079295271313,
+      "learning_rate": 0.00013028992936551243,
+      "loss": 1.4653,
+      "step": 7886
+    },
+    {
+      "epoch": 1.26192,
+      "grad_norm": 1.8417488462562486,
+      "learning_rate": 0.00013027346655759603,
+      "loss": 1.3915,
+      "step": 7887
+    },
+    {
+      "epoch": 1.26208,
+      "grad_norm": 2.007584007465001,
+      "learning_rate": 0.0001302570028463635,
+      "loss": 1.4168,
+      "step": 7888
+    },
+    {
+      "epoch": 1.26224,
+      "grad_norm": 2.3879939352804835,
+      "learning_rate": 0.00013024053823230605,
+      "loss": 1.4891,
+      "step": 7889
+    },
+    {
+      "epoch": 1.2624,
+      "grad_norm": 2.1301000568049937,
+      "learning_rate": 0.00013022407271591507,
+      "loss": 1.3593,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2625600000000001,
+      "grad_norm": 2.0686803871364394,
+      "learning_rate": 0.00013020760629768176,
+      "loss": 1.4144,
+      "step": 7891
+    },
+    {
+      "epoch": 1.26272,
+      "grad_norm": 2.042426057918864,
+      "learning_rate": 0.00013019113897809746,
+      "loss": 1.4023,
+      "step": 7892
+    },
+    {
+      "epoch": 1.26288,
+      "grad_norm": 2.6083026575599813,
+      "learning_rate": 0.00013017467075765362,
+      "loss": 1.4612,
+      "step": 7893
+    },
+    {
+      "epoch": 1.26304,
+      "grad_norm": 2.459678253127413,
+      "learning_rate": 0.0001301582016368416,
+      "loss": 1.4919,
+      "step": 7894
+    },
+    {
+      "epoch": 1.2631999999999999,
+      "grad_norm": 2.0064421666678913,
+      "learning_rate": 0.00013014173161615274,
+      "loss": 1.4597,
+      "step": 7895
+    },
+    {
+      "epoch": 1.26336,
+      "grad_norm": 2.0943990401746566,
+      "learning_rate": 0.00013012526069607853,
+      "loss": 1.4914,
+      "step": 7896
+    },
+    {
+      "epoch": 1.26352,
+      "grad_norm": 2.1485363102604706,
+      "learning_rate": 0.00013010878887711045,
+      "loss": 1.3769,
+      "step": 7897
+    },
+    {
+      "epoch": 1.26368,
+      "grad_norm": 1.882255885600734,
+      "learning_rate": 0.00013009231615974,
+      "loss": 1.4077,
+      "step": 7898
+    },
+    {
+      "epoch": 1.26384,
+      "grad_norm": 2.165719964788861,
+      "learning_rate": 0.00013007584254445863,
+      "loss": 1.4004,
+      "step": 7899
+    },
+    {
+      "epoch": 1.264,
+      "grad_norm": 2.2559744935960397,
+      "learning_rate": 0.00013005936803175798,
+      "loss": 1.3514,
+      "step": 7900
+    },
+    {
+      "epoch": 1.26416,
+      "grad_norm": 1.9968746785354048,
+      "learning_rate": 0.00013004289262212959,
+      "loss": 1.3739,
+      "step": 7901
+    },
+    {
+      "epoch": 1.26432,
+      "grad_norm": 1.9812680304264163,
+      "learning_rate": 0.00013002641631606508,
+      "loss": 1.3538,
+      "step": 7902
+    },
+    {
+      "epoch": 1.26448,
+      "grad_norm": 2.0476659729610014,
+      "learning_rate": 0.00013000993911405605,
+      "loss": 1.3466,
+      "step": 7903
+    },
+    {
+      "epoch": 1.26464,
+      "grad_norm": 1.971541692921764,
+      "learning_rate": 0.00012999346101659414,
+      "loss": 1.417,
+      "step": 7904
+    },
+    {
+      "epoch": 1.2648,
+      "grad_norm": 1.956920309643698,
+      "learning_rate": 0.00012997698202417106,
+      "loss": 1.4366,
+      "step": 7905
+    },
+    {
+      "epoch": 1.2649599999999999,
+      "grad_norm": 2.088922988963067,
+      "learning_rate": 0.00012996050213727854,
+      "loss": 1.417,
+      "step": 7906
+    },
+    {
+      "epoch": 1.26512,
+      "grad_norm": 2.399832998134418,
+      "learning_rate": 0.0001299440213564083,
+      "loss": 1.4156,
+      "step": 7907
+    },
+    {
+      "epoch": 1.26528,
+      "grad_norm": 2.439538587794676,
+      "learning_rate": 0.00012992753968205204,
+      "loss": 1.3866,
+      "step": 7908
+    },
+    {
+      "epoch": 1.26544,
+      "grad_norm": 2.09391802317201,
+      "learning_rate": 0.00012991105711470166,
+      "loss": 1.4408,
+      "step": 7909
+    },
+    {
+      "epoch": 1.2656,
+      "grad_norm": 2.1558263122403623,
+      "learning_rate": 0.0001298945736548489,
+      "loss": 1.4724,
+      "step": 7910
+    },
+    {
+      "epoch": 1.26576,
+      "grad_norm": 2.1800554674698875,
+      "learning_rate": 0.00012987808930298565,
+      "loss": 1.3618,
+      "step": 7911
+    },
+    {
+      "epoch": 1.26592,
+      "grad_norm": 2.2777686577390397,
+      "learning_rate": 0.00012986160405960374,
+      "loss": 1.4846,
+      "step": 7912
+    },
+    {
+      "epoch": 1.26608,
+      "grad_norm": 2.3212300355606716,
+      "learning_rate": 0.00012984511792519507,
+      "loss": 1.475,
+      "step": 7913
+    },
+    {
+      "epoch": 1.26624,
+      "grad_norm": 2.3212300355606716,
+      "learning_rate": 0.00012984511792519507,
+      "loss": 1.4954,
+      "step": 7914
+    },
+    {
+      "epoch": 1.2664,
+      "grad_norm": 1.8524952918367381,
+      "learning_rate": 0.00012982863090025155,
+      "loss": 1.3873,
+      "step": 7915
+    },
+    {
+      "epoch": 1.2665600000000001,
+      "grad_norm": 2.1955769692938194,
+      "learning_rate": 0.00012981214298526516,
+      "loss": 1.3782,
+      "step": 7916
+    },
+    {
+      "epoch": 1.26672,
+      "grad_norm": 2.1588260306224383,
+      "learning_rate": 0.00012979565418072789,
+      "loss": 1.3642,
+      "step": 7917
+    },
+    {
+      "epoch": 1.26688,
+      "grad_norm": 2.745366483357425,
+      "learning_rate": 0.0001297791644871317,
+      "loss": 1.5343,
+      "step": 7918
+    },
+    {
+      "epoch": 1.26704,
+      "grad_norm": 1.8612415637445505,
+      "learning_rate": 0.00012976267390496862,
+      "loss": 1.4415,
+      "step": 7919
+    },
+    {
+      "epoch": 1.2671999999999999,
+      "grad_norm": 1.9956439718561216,
+      "learning_rate": 0.0001297461824347307,
+      "loss": 1.378,
+      "step": 7920
+    },
+    {
+      "epoch": 1.26736,
+      "grad_norm": 2.470964036469449,
+      "learning_rate": 0.0001297296900769101,
+      "loss": 1.3527,
+      "step": 7921
+    },
+    {
+      "epoch": 1.26752,
+      "grad_norm": 2.7334469131934163,
+      "learning_rate": 0.00012971319683199886,
+      "loss": 1.4495,
+      "step": 7922
+    },
+    {
+      "epoch": 1.26768,
+      "grad_norm": 2.239749780743196,
+      "learning_rate": 0.00012969670270048908,
+      "loss": 1.3422,
+      "step": 7923
+    },
+    {
+      "epoch": 1.26784,
+      "grad_norm": 2.9751326320851796,
+      "learning_rate": 0.000129680207682873,
+      "loss": 1.4353,
+      "step": 7924
+    },
+    {
+      "epoch": 1.268,
+      "grad_norm": 2.0997725110800545,
+      "learning_rate": 0.00012966371177964276,
+      "loss": 1.3704,
+      "step": 7925
+    },
+    {
+      "epoch": 1.26816,
+      "grad_norm": 2.526248864175488,
+      "learning_rate": 0.0001296472149912906,
+      "loss": 1.4124,
+      "step": 7926
+    },
+    {
+      "epoch": 1.2683200000000001,
+      "grad_norm": 2.444038785878359,
+      "learning_rate": 0.00012963071731830872,
+      "loss": 1.4132,
+      "step": 7927
+    },
+    {
+      "epoch": 1.26848,
+      "grad_norm": 1.9880053554843655,
+      "learning_rate": 0.00012961421876118943,
+      "loss": 1.3805,
+      "step": 7928
+    },
+    {
+      "epoch": 1.26864,
+      "grad_norm": 2.906649148331151,
+      "learning_rate": 0.00012959771932042496,
+      "loss": 1.3831,
+      "step": 7929
+    },
+    {
+      "epoch": 1.2688,
+      "grad_norm": 1.9285261117827486,
+      "learning_rate": 0.00012958121899650773,
+      "loss": 1.3633,
+      "step": 7930
+    },
+    {
+      "epoch": 1.2689599999999999,
+      "grad_norm": 2.2663064946782425,
+      "learning_rate": 0.00012956471778993001,
+      "loss": 1.4883,
+      "step": 7931
+    },
+    {
+      "epoch": 1.26912,
+      "grad_norm": 2.2598881283114034,
+      "learning_rate": 0.0001295482157011842,
+      "loss": 1.4502,
+      "step": 7932
+    },
+    {
+      "epoch": 1.26928,
+      "grad_norm": 2.084217545367193,
+      "learning_rate": 0.0001295317127307627,
+      "loss": 1.3936,
+      "step": 7933
+    },
+    {
+      "epoch": 1.26944,
+      "grad_norm": 2.999848279195266,
+      "learning_rate": 0.0001295152088791579,
+      "loss": 1.3627,
+      "step": 7934
+    },
+    {
+      "epoch": 1.2696,
+      "grad_norm": 2.1506298029929813,
+      "learning_rate": 0.00012949870414686226,
+      "loss": 1.3708,
+      "step": 7935
+    },
+    {
+      "epoch": 1.26976,
+      "grad_norm": 2.2940961556809034,
+      "learning_rate": 0.0001294821985343683,
+      "loss": 1.4601,
+      "step": 7936
+    },
+    {
+      "epoch": 1.26992,
+      "grad_norm": 2.9913363807159508,
+      "learning_rate": 0.0001294656920421685,
+      "loss": 1.3733,
+      "step": 7937
+    },
+    {
+      "epoch": 1.27008,
+      "grad_norm": 2.3486164338238185,
+      "learning_rate": 0.00012944918467075537,
+      "loss": 1.3822,
+      "step": 7938
+    },
+    {
+      "epoch": 1.27024,
+      "grad_norm": 3.127221707389181,
+      "learning_rate": 0.0001294326764206215,
+      "loss": 1.3196,
+      "step": 7939
+    },
+    {
+      "epoch": 1.2704,
+      "grad_norm": 2.3060013892956777,
+      "learning_rate": 0.00012941616729225945,
+      "loss": 1.3216,
+      "step": 7940
+    },
+    {
+      "epoch": 1.2705600000000001,
+      "grad_norm": 2.3902399800626544,
+      "learning_rate": 0.00012939965728616182,
+      "loss": 1.3775,
+      "step": 7941
+    },
+    {
+      "epoch": 1.27072,
+      "grad_norm": 2.3866223339789516,
+      "learning_rate": 0.00012938314640282127,
+      "loss": 1.3564,
+      "step": 7942
+    },
+    {
+      "epoch": 1.27088,
+      "grad_norm": 2.2734974482197123,
+      "learning_rate": 0.0001293666346427304,
+      "loss": 1.456,
+      "step": 7943
+    },
+    {
+      "epoch": 1.27104,
+      "grad_norm": 2.333833349139466,
+      "learning_rate": 0.000129350122006382,
+      "loss": 1.4167,
+      "step": 7944
+    },
+    {
+      "epoch": 1.2711999999999999,
+      "grad_norm": 1.9543809381917894,
+      "learning_rate": 0.0001293336084942687,
+      "loss": 1.4283,
+      "step": 7945
+    },
+    {
+      "epoch": 1.27136,
+      "grad_norm": 2.0324441869469476,
+      "learning_rate": 0.00012931709410688325,
+      "loss": 1.4254,
+      "step": 7946
+    },
+    {
+      "epoch": 1.27152,
+      "grad_norm": 2.075728142983663,
+      "learning_rate": 0.0001293005788447185,
+      "loss": 1.4286,
+      "step": 7947
+    },
+    {
+      "epoch": 1.27168,
+      "grad_norm": 1.9538923129368297,
+      "learning_rate": 0.00012928406270826712,
+      "loss": 1.3605,
+      "step": 7948
+    },
+    {
+      "epoch": 1.27184,
+      "grad_norm": 1.8535511199640904,
+      "learning_rate": 0.00012926754569802197,
+      "loss": 1.4759,
+      "step": 7949
+    },
+    {
+      "epoch": 1.272,
+      "grad_norm": 2.108691930186439,
+      "learning_rate": 0.00012925102781447594,
+      "loss": 1.4489,
+      "step": 7950
+    },
+    {
+      "epoch": 1.27216,
+      "grad_norm": 2.05275500363998,
+      "learning_rate": 0.00012923450905812182,
+      "loss": 1.3464,
+      "step": 7951
+    },
+    {
+      "epoch": 1.2723200000000001,
+      "grad_norm": 2.3120908452746574,
+      "learning_rate": 0.0001292179894294526,
+      "loss": 1.3653,
+      "step": 7952
+    },
+    {
+      "epoch": 1.27248,
+      "grad_norm": 2.7067519975762266,
+      "learning_rate": 0.0001292014689289611,
+      "loss": 1.3882,
+      "step": 7953
+    },
+    {
+      "epoch": 1.27264,
+      "grad_norm": 1.998259737163796,
+      "learning_rate": 0.00012918494755714035,
+      "loss": 1.3665,
+      "step": 7954
+    },
+    {
+      "epoch": 1.2728,
+      "grad_norm": 1.9374035789613053,
+      "learning_rate": 0.00012916842531448327,
+      "loss": 1.3765,
+      "step": 7955
+    },
+    {
+      "epoch": 1.2729599999999999,
+      "grad_norm": 2.0951001983717825,
+      "learning_rate": 0.0001291519022014829,
+      "loss": 1.4039,
+      "step": 7956
+    },
+    {
+      "epoch": 1.27312,
+      "grad_norm": 2.455328167787536,
+      "learning_rate": 0.00012913537821863225,
+      "loss": 1.4648,
+      "step": 7957
+    },
+    {
+      "epoch": 1.27328,
+      "grad_norm": 2.027334544815486,
+      "learning_rate": 0.00012911885336642438,
+      "loss": 1.3877,
+      "step": 7958
+    },
+    {
+      "epoch": 1.27344,
+      "grad_norm": 1.9438799219183347,
+      "learning_rate": 0.00012910232764535234,
+      "loss": 1.4033,
+      "step": 7959
+    },
+    {
+      "epoch": 1.2736,
+      "grad_norm": 2.125199441169938,
+      "learning_rate": 0.00012908580105590922,
+      "loss": 1.3307,
+      "step": 7960
+    },
+    {
+      "epoch": 1.27376,
+      "grad_norm": 2.006573939116326,
+      "learning_rate": 0.0001290692735985882,
+      "loss": 1.3376,
+      "step": 7961
+    },
+    {
+      "epoch": 1.27392,
+      "grad_norm": 2.236607941462789,
+      "learning_rate": 0.00012905274527388243,
+      "loss": 1.3779,
+      "step": 7962
+    },
+    {
+      "epoch": 1.27408,
+      "grad_norm": 2.771917921720805,
+      "learning_rate": 0.00012903621608228506,
+      "loss": 1.4776,
+      "step": 7963
+    },
+    {
+      "epoch": 1.27424,
+      "grad_norm": 2.185967475832284,
+      "learning_rate": 0.00012901968602428934,
+      "loss": 1.3694,
+      "step": 7964
+    },
+    {
+      "epoch": 1.2744,
+      "grad_norm": 2.1084280762396386,
+      "learning_rate": 0.00012900315510038846,
+      "loss": 1.3372,
+      "step": 7965
+    },
+    {
+      "epoch": 1.2745600000000001,
+      "grad_norm": 2.29988538562899,
+      "learning_rate": 0.0001289866233110757,
+      "loss": 1.2926,
+      "step": 7966
+    },
+    {
+      "epoch": 1.27472,
+      "grad_norm": 1.9553111634116094,
+      "learning_rate": 0.0001289700906568443,
+      "loss": 1.3829,
+      "step": 7967
+    },
+    {
+      "epoch": 1.27488,
+      "grad_norm": 2.01496209762248,
+      "learning_rate": 0.00012895355713818762,
+      "loss": 1.3779,
+      "step": 7968
+    },
+    {
+      "epoch": 1.27504,
+      "grad_norm": 3.095001980543093,
+      "learning_rate": 0.000128937022755599,
+      "loss": 1.3156,
+      "step": 7969
+    },
+    {
+      "epoch": 1.2752,
+      "grad_norm": 2.12488987779822,
+      "learning_rate": 0.0001289204875095718,
+      "loss": 1.3852,
+      "step": 7970
+    },
+    {
+      "epoch": 1.27536,
+      "grad_norm": 2.97458482317606,
+      "learning_rate": 0.00012890395140059937,
+      "loss": 1.3187,
+      "step": 7971
+    },
+    {
+      "epoch": 1.27552,
+      "grad_norm": 2.461794483410351,
+      "learning_rate": 0.00012888741442917517,
+      "loss": 1.4149,
+      "step": 7972
+    },
+    {
+      "epoch": 1.27568,
+      "grad_norm": 2.2262233827760687,
+      "learning_rate": 0.00012887087659579257,
+      "loss": 1.5298,
+      "step": 7973
+    },
+    {
+      "epoch": 1.27584,
+      "grad_norm": 2.316210339409607,
+      "learning_rate": 0.00012885433790094512,
+      "loss": 1.3995,
+      "step": 7974
+    },
+    {
+      "epoch": 1.276,
+      "grad_norm": 2.26718187037539,
+      "learning_rate": 0.00012883779834512628,
+      "loss": 1.459,
+      "step": 7975
+    },
+    {
+      "epoch": 1.27616,
+      "grad_norm": 2.1419744350639878,
+      "learning_rate": 0.00012882125792882952,
+      "loss": 1.4747,
+      "step": 7976
+    },
+    {
+      "epoch": 1.2763200000000001,
+      "grad_norm": 2.090859247430979,
+      "learning_rate": 0.00012880471665254847,
+      "loss": 1.3219,
+      "step": 7977
+    },
+    {
+      "epoch": 1.27648,
+      "grad_norm": 2.1827959969167923,
+      "learning_rate": 0.00012878817451677661,
+      "loss": 1.3905,
+      "step": 7978
+    },
+    {
+      "epoch": 1.27664,
+      "grad_norm": 2.0475284812324688,
+      "learning_rate": 0.00012877163152200758,
+      "loss": 1.3779,
+      "step": 7979
+    },
+    {
+      "epoch": 1.2768,
+      "grad_norm": 2.2709086414539184,
+      "learning_rate": 0.00012875508766873502,
+      "loss": 1.423,
+      "step": 7980
+    },
+    {
+      "epoch": 1.2769599999999999,
+      "grad_norm": 2.448975433726529,
+      "learning_rate": 0.00012873854295745253,
+      "loss": 1.3678,
+      "step": 7981
+    },
+    {
+      "epoch": 1.27712,
+      "grad_norm": 2.6372532201522705,
+      "learning_rate": 0.00012872199738865378,
+      "loss": 1.4241,
+      "step": 7982
+    },
+    {
+      "epoch": 1.27728,
+      "grad_norm": 2.3998269912118784,
+      "learning_rate": 0.0001287054509628325,
+      "loss": 1.4434,
+      "step": 7983
+    },
+    {
+      "epoch": 1.27744,
+      "grad_norm": 2.334643557916711,
+      "learning_rate": 0.00012868890368048237,
+      "loss": 1.3665,
+      "step": 7984
+    },
+    {
+      "epoch": 1.2776,
+      "grad_norm": 2.1595427674713994,
+      "learning_rate": 0.00012867235554209715,
+      "loss": 1.387,
+      "step": 7985
+    },
+    {
+      "epoch": 1.27776,
+      "grad_norm": 2.198645861188001,
+      "learning_rate": 0.00012865580654817065,
+      "loss": 1.4396,
+      "step": 7986
+    },
+    {
+      "epoch": 1.27792,
+      "grad_norm": 1.8582289183635714,
+      "learning_rate": 0.00012863925669919662,
+      "loss": 1.3817,
+      "step": 7987
+    },
+    {
+      "epoch": 1.27808,
+      "grad_norm": 2.227114198800455,
+      "learning_rate": 0.0001286227059956689,
+      "loss": 1.4107,
+      "step": 7988
+    },
+    {
+      "epoch": 1.27824,
+      "grad_norm": 2.257401147747518,
+      "learning_rate": 0.00012860615443808135,
+      "loss": 1.3996,
+      "step": 7989
+    },
+    {
+      "epoch": 1.2784,
+      "grad_norm": 2.024301519706778,
+      "learning_rate": 0.00012858960202692785,
+      "loss": 1.3943,
+      "step": 7990
+    },
+    {
+      "epoch": 1.27856,
+      "grad_norm": 2.3322891800654575,
+      "learning_rate": 0.00012857304876270225,
+      "loss": 1.3648,
+      "step": 7991
+    },
+    {
+      "epoch": 1.27872,
+      "grad_norm": 2.1470874096181696,
+      "learning_rate": 0.0001285564946458985,
+      "loss": 1.3198,
+      "step": 7992
+    },
+    {
+      "epoch": 1.27888,
+      "grad_norm": 2.2135938637436836,
+      "learning_rate": 0.00012853993967701062,
+      "loss": 1.4598,
+      "step": 7993
+    },
+    {
+      "epoch": 1.27904,
+      "grad_norm": 2.2244891263915294,
+      "learning_rate": 0.0001285233838565325,
+      "loss": 1.3788,
+      "step": 7994
+    },
+    {
+      "epoch": 1.2792,
+      "grad_norm": 2.3980238598987564,
+      "learning_rate": 0.00012850682718495816,
+      "loss": 1.4158,
+      "step": 7995
+    },
+    {
+      "epoch": 1.27936,
+      "grad_norm": 2.597584666042465,
+      "learning_rate": 0.00012849026966278167,
+      "loss": 1.3974,
+      "step": 7996
+    },
+    {
+      "epoch": 1.27952,
+      "grad_norm": 1.996419159164114,
+      "learning_rate": 0.000128473711290497,
+      "loss": 1.419,
+      "step": 7997
+    },
+    {
+      "epoch": 1.27968,
+      "grad_norm": 3.176560879447366,
+      "learning_rate": 0.0001284571520685983,
+      "loss": 1.4249,
+      "step": 7998
+    },
+    {
+      "epoch": 1.27984,
+      "grad_norm": 2.041776585754355,
+      "learning_rate": 0.00012844059199757966,
+      "loss": 1.3157,
+      "step": 7999
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 2.6514016912517757,
+      "learning_rate": 0.00012842403107793515,
+      "loss": 1.3976,
+      "step": 8000
+    },
+    {
+      "epoch": 1.28016,
+      "grad_norm": 2.809087679336117,
+      "learning_rate": 0.000128407469310159,
+      "loss": 1.4191,
+      "step": 8001
+    },
+    {
+      "epoch": 1.2803200000000001,
+      "grad_norm": 2.4061644789744543,
+      "learning_rate": 0.00012839090669474536,
+      "loss": 1.4058,
+      "step": 8002
+    },
+    {
+      "epoch": 1.28048,
+      "grad_norm": 2.9532381125483664,
+      "learning_rate": 0.00012837434323218844,
+      "loss": 1.4167,
+      "step": 8003
+    },
+    {
+      "epoch": 1.28064,
+      "grad_norm": 2.1329652534563284,
+      "learning_rate": 0.00012835777892298246,
+      "loss": 1.3648,
+      "step": 8004
+    },
+    {
+      "epoch": 1.2808,
+      "grad_norm": 2.6759612068277123,
+      "learning_rate": 0.00012834121376762168,
+      "loss": 1.5092,
+      "step": 8005
+    },
+    {
+      "epoch": 1.2809599999999999,
+      "grad_norm": 2.398560073929075,
+      "learning_rate": 0.00012832464776660037,
+      "loss": 1.4035,
+      "step": 8006
+    },
+    {
+      "epoch": 1.28112,
+      "grad_norm": 2.213345839152753,
+      "learning_rate": 0.00012830808092041286,
+      "loss": 1.3837,
+      "step": 8007
+    },
+    {
+      "epoch": 1.28128,
+      "grad_norm": 2.4417666735559305,
+      "learning_rate": 0.00012829151322955344,
+      "loss": 1.4852,
+      "step": 8008
+    },
+    {
+      "epoch": 1.28144,
+      "grad_norm": 2.0015452495932093,
+      "learning_rate": 0.00012827494469451651,
+      "loss": 1.4135,
+      "step": 8009
+    },
+    {
+      "epoch": 1.2816,
+      "grad_norm": 2.2149553486467966,
+      "learning_rate": 0.00012825837531579642,
+      "loss": 1.4295,
+      "step": 8010
+    },
+    {
+      "epoch": 1.28176,
+      "grad_norm": 2.4005634201872783,
+      "learning_rate": 0.0001282418050938876,
+      "loss": 1.5358,
+      "step": 8011
+    },
+    {
+      "epoch": 1.28192,
+      "grad_norm": 2.122837819935902,
+      "learning_rate": 0.0001282252340292845,
+      "loss": 1.4708,
+      "step": 8012
+    },
+    {
+      "epoch": 1.28208,
+      "grad_norm": 1.9819565340264818,
+      "learning_rate": 0.00012820866212248148,
+      "loss": 1.3809,
+      "step": 8013
+    },
+    {
+      "epoch": 1.28224,
+      "grad_norm": 3.0980182223653823,
+      "learning_rate": 0.0001281920893739731,
+      "loss": 1.4303,
+      "step": 8014
+    },
+    {
+      "epoch": 1.2824,
+      "grad_norm": 2.0488056057059127,
+      "learning_rate": 0.0001281755157842539,
+      "loss": 1.4557,
+      "step": 8015
+    },
+    {
+      "epoch": 1.28256,
+      "grad_norm": 2.6830480334575095,
+      "learning_rate": 0.0001281589413538183,
+      "loss": 1.3411,
+      "step": 8016
+    },
+    {
+      "epoch": 1.28272,
+      "grad_norm": 2.3814629603976183,
+      "learning_rate": 0.00012814236608316094,
+      "loss": 1.4196,
+      "step": 8017
+    },
+    {
+      "epoch": 1.28288,
+      "grad_norm": 2.091385904701355,
+      "learning_rate": 0.00012812578997277641,
+      "loss": 1.4106,
+      "step": 8018
+    },
+    {
+      "epoch": 1.28304,
+      "grad_norm": 2.2605762584033826,
+      "learning_rate": 0.00012810921302315928,
+      "loss": 1.4407,
+      "step": 8019
+    },
+    {
+      "epoch": 1.2832,
+      "grad_norm": 2.193564623557982,
+      "learning_rate": 0.0001280926352348042,
+      "loss": 1.3969,
+      "step": 8020
+    },
+    {
+      "epoch": 1.28336,
+      "grad_norm": 2.142477624471191,
+      "learning_rate": 0.0001280760566082058,
+      "loss": 1.4159,
+      "step": 8021
+    },
+    {
+      "epoch": 1.28352,
+      "grad_norm": 2.699147411397762,
+      "learning_rate": 0.00012805947714385882,
+      "loss": 1.4229,
+      "step": 8022
+    },
+    {
+      "epoch": 1.28368,
+      "grad_norm": 2.0596716165103346,
+      "learning_rate": 0.0001280428968422579,
+      "loss": 1.4098,
+      "step": 8023
+    },
+    {
+      "epoch": 1.28384,
+      "grad_norm": 2.633869483872342,
+      "learning_rate": 0.00012802631570389778,
+      "loss": 1.3953,
+      "step": 8024
+    },
+    {
+      "epoch": 1.284,
+      "grad_norm": 2.194308919605268,
+      "learning_rate": 0.00012800973372927328,
+      "loss": 1.3655,
+      "step": 8025
+    },
+    {
+      "epoch": 1.28416,
+      "grad_norm": 2.118200130202835,
+      "learning_rate": 0.0001279931509188791,
+      "loss": 1.4254,
+      "step": 8026
+    },
+    {
+      "epoch": 1.2843200000000001,
+      "grad_norm": 2.537507714022594,
+      "learning_rate": 0.00012797656727321012,
+      "loss": 1.4269,
+      "step": 8027
+    },
+    {
+      "epoch": 1.28448,
+      "grad_norm": 2.0321819908551033,
+      "learning_rate": 0.00012795998279276113,
+      "loss": 1.4561,
+      "step": 8028
+    },
+    {
+      "epoch": 1.28464,
+      "grad_norm": 2.113645473373816,
+      "learning_rate": 0.000127943397478027,
+      "loss": 1.3891,
+      "step": 8029
+    },
+    {
+      "epoch": 1.2848,
+      "grad_norm": 2.6161787563783556,
+      "learning_rate": 0.0001279268113295026,
+      "loss": 1.484,
+      "step": 8030
+    },
+    {
+      "epoch": 1.2849599999999999,
+      "grad_norm": 1.992807423004894,
+      "learning_rate": 0.00012791022434768286,
+      "loss": 1.383,
+      "step": 8031
+    },
+    {
+      "epoch": 1.28512,
+      "grad_norm": 2.2575826112923973,
+      "learning_rate": 0.00012789363653306266,
+      "loss": 1.4309,
+      "step": 8032
+    },
+    {
+      "epoch": 1.28528,
+      "grad_norm": 2.165598563853064,
+      "learning_rate": 0.000127877047886137,
+      "loss": 1.4255,
+      "step": 8033
+    },
+    {
+      "epoch": 1.28544,
+      "grad_norm": 2.1790936224232285,
+      "learning_rate": 0.00012786045840740084,
+      "loss": 1.4173,
+      "step": 8034
+    },
+    {
+      "epoch": 1.2856,
+      "grad_norm": 2.1438315196759965,
+      "learning_rate": 0.00012784386809734923,
+      "loss": 1.3899,
+      "step": 8035
+    },
+    {
+      "epoch": 1.28576,
+      "grad_norm": 2.345569996568262,
+      "learning_rate": 0.00012782727695647716,
+      "loss": 1.3683,
+      "step": 8036
+    },
+    {
+      "epoch": 1.28592,
+      "grad_norm": 2.1046436374459057,
+      "learning_rate": 0.0001278106849852797,
+      "loss": 1.4228,
+      "step": 8037
+    },
+    {
+      "epoch": 1.2860800000000001,
+      "grad_norm": 1.8946535880706314,
+      "learning_rate": 0.0001277940921842519,
+      "loss": 1.3318,
+      "step": 8038
+    },
+    {
+      "epoch": 1.28624,
+      "grad_norm": 2.049959559473627,
+      "learning_rate": 0.0001277774985538889,
+      "loss": 1.4326,
+      "step": 8039
+    },
+    {
+      "epoch": 1.2864,
+      "grad_norm": 1.9648578109135002,
+      "learning_rate": 0.00012776090409468585,
+      "loss": 1.4003,
+      "step": 8040
+    },
+    {
+      "epoch": 1.28656,
+      "grad_norm": 1.8570247684452734,
+      "learning_rate": 0.00012774430880713785,
+      "loss": 1.3298,
+      "step": 8041
+    },
+    {
+      "epoch": 1.2867199999999999,
+      "grad_norm": 2.269618810643841,
+      "learning_rate": 0.00012772771269174007,
+      "loss": 1.4567,
+      "step": 8042
+    },
+    {
+      "epoch": 1.28688,
+      "grad_norm": 2.215588996103967,
+      "learning_rate": 0.00012771111574898778,
+      "loss": 1.4888,
+      "step": 8043
+    },
+    {
+      "epoch": 1.28704,
+      "grad_norm": 2.5124466042124487,
+      "learning_rate": 0.0001276945179793762,
+      "loss": 1.3784,
+      "step": 8044
+    },
+    {
+      "epoch": 1.2872,
+      "grad_norm": 2.476795329799695,
+      "learning_rate": 0.00012767791938340052,
+      "loss": 1.4199,
+      "step": 8045
+    },
+    {
+      "epoch": 1.28736,
+      "grad_norm": 2.3464436593812006,
+      "learning_rate": 0.00012766131996155608,
+      "loss": 1.4247,
+      "step": 8046
+    },
+    {
+      "epoch": 1.28752,
+      "grad_norm": 2.0190838009062326,
+      "learning_rate": 0.00012764471971433815,
+      "loss": 1.3692,
+      "step": 8047
+    },
+    {
+      "epoch": 1.28768,
+      "grad_norm": 2.239776456575156,
+      "learning_rate": 0.00012762811864224207,
+      "loss": 1.3554,
+      "step": 8048
+    },
+    {
+      "epoch": 1.28784,
+      "grad_norm": 2.2123339996087705,
+      "learning_rate": 0.00012761151674576317,
+      "loss": 1.3491,
+      "step": 8049
+    },
+    {
+      "epoch": 1.288,
+      "grad_norm": 2.486084297166202,
+      "learning_rate": 0.00012759491402539687,
+      "loss": 1.3166,
+      "step": 8050
+    },
+    {
+      "epoch": 1.28816,
+      "grad_norm": 2.291538977161559,
+      "learning_rate": 0.00012757831048163852,
+      "loss": 1.4672,
+      "step": 8051
+    },
+    {
+      "epoch": 1.2883200000000001,
+      "grad_norm": 2.3786806228616704,
+      "learning_rate": 0.0001275617061149836,
+      "loss": 1.3822,
+      "step": 8052
+    },
+    {
+      "epoch": 1.28848,
+      "grad_norm": 2.545075045293198,
+      "learning_rate": 0.0001275451009259275,
+      "loss": 1.4482,
+      "step": 8053
+    },
+    {
+      "epoch": 1.28864,
+      "grad_norm": 2.596043512578078,
+      "learning_rate": 0.00012752849491496577,
+      "loss": 1.3761,
+      "step": 8054
+    },
+    {
+      "epoch": 1.2888,
+      "grad_norm": 2.0960939459055514,
+      "learning_rate": 0.00012751188808259384,
+      "loss": 1.447,
+      "step": 8055
+    },
+    {
+      "epoch": 1.2889599999999999,
+      "grad_norm": 1.9157528975102716,
+      "learning_rate": 0.00012749528042930726,
+      "loss": 1.3794,
+      "step": 8056
+    },
+    {
+      "epoch": 1.28912,
+      "grad_norm": 2.2262888574883544,
+      "learning_rate": 0.00012747867195560155,
+      "loss": 1.4954,
+      "step": 8057
+    },
+    {
+      "epoch": 1.28928,
+      "grad_norm": 2.08913218877582,
+      "learning_rate": 0.00012746206266197232,
+      "loss": 1.4272,
+      "step": 8058
+    },
+    {
+      "epoch": 1.28944,
+      "grad_norm": 2.3085847581654604,
+      "learning_rate": 0.00012744545254891518,
+      "loss": 1.4785,
+      "step": 8059
+    },
+    {
+      "epoch": 1.2896,
+      "grad_norm": 2.8612119432409333,
+      "learning_rate": 0.00012742884161692571,
+      "loss": 1.5013,
+      "step": 8060
+    },
+    {
+      "epoch": 1.28976,
+      "grad_norm": 2.3074682203173302,
+      "learning_rate": 0.00012741222986649962,
+      "loss": 1.4471,
+      "step": 8061
+    },
+    {
+      "epoch": 1.28992,
+      "grad_norm": 2.839336979749249,
+      "learning_rate": 0.0001273956172981325,
+      "loss": 1.3966,
+      "step": 8062
+    },
+    {
+      "epoch": 1.2900800000000001,
+      "grad_norm": 2.6500392516210742,
+      "learning_rate": 0.00012737900391232004,
+      "loss": 1.4449,
+      "step": 8063
+    },
+    {
+      "epoch": 1.29024,
+      "grad_norm": 2.8506079430456066,
+      "learning_rate": 0.00012736238970955806,
+      "loss": 1.3967,
+      "step": 8064
+    },
+    {
+      "epoch": 1.2904,
+      "grad_norm": 2.381010736367425,
+      "learning_rate": 0.00012734577469034218,
+      "loss": 1.4526,
+      "step": 8065
+    },
+    {
+      "epoch": 1.29056,
+      "grad_norm": 2.0388810509919204,
+      "learning_rate": 0.00012732915885516828,
+      "loss": 1.3789,
+      "step": 8066
+    },
+    {
+      "epoch": 1.2907199999999999,
+      "grad_norm": 2.8564050793463434,
+      "learning_rate": 0.0001273125422045321,
+      "loss": 1.3805,
+      "step": 8067
+    },
+    {
+      "epoch": 1.29088,
+      "grad_norm": 2.33255989442505,
+      "learning_rate": 0.00012729592473892943,
+      "loss": 1.3809,
+      "step": 8068
+    },
+    {
+      "epoch": 1.29104,
+      "grad_norm": 2.252342138031035,
+      "learning_rate": 0.00012727930645885617,
+      "loss": 1.3888,
+      "step": 8069
+    },
+    {
+      "epoch": 1.2912,
+      "grad_norm": 2.369311077053226,
+      "learning_rate": 0.00012726268736480814,
+      "loss": 1.362,
+      "step": 8070
+    },
+    {
+      "epoch": 1.29136,
+      "grad_norm": 1.837032065924744,
+      "learning_rate": 0.00012724606745728124,
+      "loss": 1.4121,
+      "step": 8071
+    },
+    {
+      "epoch": 1.29152,
+      "grad_norm": 2.0891873864744266,
+      "learning_rate": 0.0001272294467367714,
+      "loss": 1.385,
+      "step": 8072
+    },
+    {
+      "epoch": 1.29168,
+      "grad_norm": 1.8862410779457637,
+      "learning_rate": 0.00012721282520377456,
+      "loss": 1.476,
+      "step": 8073
+    },
+    {
+      "epoch": 1.29184,
+      "grad_norm": 2.0473850827872466,
+      "learning_rate": 0.00012719620285878665,
+      "loss": 1.3601,
+      "step": 8074
+    },
+    {
+      "epoch": 1.292,
+      "grad_norm": 2.553386831589534,
+      "learning_rate": 0.00012717957970230368,
+      "loss": 1.3204,
+      "step": 8075
+    },
+    {
+      "epoch": 1.29216,
+      "grad_norm": 2.040398597684969,
+      "learning_rate": 0.00012716295573482163,
+      "loss": 1.5175,
+      "step": 8076
+    },
+    {
+      "epoch": 1.2923200000000001,
+      "grad_norm": 2.3758413363650446,
+      "learning_rate": 0.0001271463309568366,
+      "loss": 1.4596,
+      "step": 8077
+    },
+    {
+      "epoch": 1.29248,
+      "grad_norm": 2.1764296764184428,
+      "learning_rate": 0.0001271297053688446,
+      "loss": 1.4511,
+      "step": 8078
+    },
+    {
+      "epoch": 1.29264,
+      "grad_norm": 2.40075099088683,
+      "learning_rate": 0.00012711307897134171,
+      "loss": 1.4777,
+      "step": 8079
+    },
+    {
+      "epoch": 1.2928,
+      "grad_norm": 2.6100303404517535,
+      "learning_rate": 0.0001270964517648241,
+      "loss": 1.3975,
+      "step": 8080
+    },
+    {
+      "epoch": 1.2929599999999999,
+      "grad_norm": 2.0963009628748077,
+      "learning_rate": 0.0001270798237497878,
+      "loss": 1.4095,
+      "step": 8081
+    },
+    {
+      "epoch": 1.29312,
+      "grad_norm": 2.084675365267033,
+      "learning_rate": 0.00012706319492672901,
+      "loss": 1.3625,
+      "step": 8082
+    },
+    {
+      "epoch": 1.29328,
+      "grad_norm": 2.320411368402611,
+      "learning_rate": 0.00012704656529614395,
+      "loss": 1.3519,
+      "step": 8083
+    },
+    {
+      "epoch": 1.29344,
+      "grad_norm": 2.0555601173565465,
+      "learning_rate": 0.0001270299348585288,
+      "loss": 1.3516,
+      "step": 8084
+    },
+    {
+      "epoch": 1.2936,
+      "grad_norm": 2.0244771180135195,
+      "learning_rate": 0.00012701330361437977,
+      "loss": 1.443,
+      "step": 8085
+    },
+    {
+      "epoch": 1.29376,
+      "grad_norm": 2.0153600123856252,
+      "learning_rate": 0.00012699667156419313,
+      "loss": 1.4141,
+      "step": 8086
+    },
+    {
+      "epoch": 1.29392,
+      "grad_norm": 2.493628426615084,
+      "learning_rate": 0.00012698003870846515,
+      "loss": 1.3412,
+      "step": 8087
+    },
+    {
+      "epoch": 1.2940800000000001,
+      "grad_norm": 1.9602652693994096,
+      "learning_rate": 0.00012696340504769212,
+      "loss": 1.3839,
+      "step": 8088
+    },
+    {
+      "epoch": 1.29424,
+      "grad_norm": 1.9907808017274087,
+      "learning_rate": 0.00012694677058237033,
+      "loss": 1.3955,
+      "step": 8089
+    },
+    {
+      "epoch": 1.2944,
+      "grad_norm": 2.1440101517506225,
+      "learning_rate": 0.00012693013531299622,
+      "loss": 1.4449,
+      "step": 8090
+    },
+    {
+      "epoch": 1.29456,
+      "grad_norm": 2.1410695317821524,
+      "learning_rate": 0.0001269134992400661,
+      "loss": 1.4303,
+      "step": 8091
+    },
+    {
+      "epoch": 1.2947199999999999,
+      "grad_norm": 1.972120698549513,
+      "learning_rate": 0.00012689686236407637,
+      "loss": 1.3947,
+      "step": 8092
+    },
+    {
+      "epoch": 1.29488,
+      "grad_norm": 2.0882062405523274,
+      "learning_rate": 0.0001268802246855235,
+      "loss": 1.4061,
+      "step": 8093
+    },
+    {
+      "epoch": 1.29504,
+      "grad_norm": 2.0737851649951304,
+      "learning_rate": 0.00012686358620490385,
+      "loss": 1.3588,
+      "step": 8094
+    },
+    {
+      "epoch": 1.2952,
+      "grad_norm": 2.0058379543285683,
+      "learning_rate": 0.00012684694692271397,
+      "loss": 1.3727,
+      "step": 8095
+    },
+    {
+      "epoch": 1.29536,
+      "grad_norm": 2.1082909806731096,
+      "learning_rate": 0.0001268303068394503,
+      "loss": 1.3307,
+      "step": 8096
+    },
+    {
+      "epoch": 1.29552,
+      "grad_norm": 2.2231414362704474,
+      "learning_rate": 0.00012681366595560935,
+      "loss": 1.4143,
+      "step": 8097
+    },
+    {
+      "epoch": 1.29568,
+      "grad_norm": 2.472381198973431,
+      "learning_rate": 0.0001267970242716877,
+      "loss": 1.3929,
+      "step": 8098
+    },
+    {
+      "epoch": 1.29584,
+      "grad_norm": 1.93346732716998,
+      "learning_rate": 0.00012678038178818188,
+      "loss": 1.4236,
+      "step": 8099
+    },
+    {
+      "epoch": 1.296,
+      "grad_norm": 2.031298686496301,
+      "learning_rate": 0.0001267637385055885,
+      "loss": 1.3646,
+      "step": 8100
+    },
+    {
+      "epoch": 1.29616,
+      "grad_norm": 2.675971869327693,
+      "learning_rate": 0.00012674709442440417,
+      "loss": 1.3811,
+      "step": 8101
+    },
+    {
+      "epoch": 1.29632,
+      "grad_norm": 1.9981578155660653,
+      "learning_rate": 0.0001267304495451255,
+      "loss": 1.4051,
+      "step": 8102
+    },
+    {
+      "epoch": 1.29648,
+      "grad_norm": 2.3752479274008214,
+      "learning_rate": 0.00012671380386824918,
+      "loss": 1.4205,
+      "step": 8103
+    },
+    {
+      "epoch": 1.29664,
+      "grad_norm": 2.3157748573524763,
+      "learning_rate": 0.00012669715739427187,
+      "loss": 1.3718,
+      "step": 8104
+    },
+    {
+      "epoch": 1.2968,
+      "grad_norm": 2.304296873655637,
+      "learning_rate": 0.0001266805101236903,
+      "loss": 1.4021,
+      "step": 8105
+    },
+    {
+      "epoch": 1.29696,
+      "grad_norm": 2.085511030488217,
+      "learning_rate": 0.00012666386205700117,
+      "loss": 1.3905,
+      "step": 8106
+    },
+    {
+      "epoch": 1.29712,
+      "grad_norm": 2.509938348690184,
+      "learning_rate": 0.00012664721319470123,
+      "loss": 1.3471,
+      "step": 8107
+    },
+    {
+      "epoch": 1.29728,
+      "grad_norm": 3.1900054189616474,
+      "learning_rate": 0.0001266305635372873,
+      "loss": 1.348,
+      "step": 8108
+    },
+    {
+      "epoch": 1.29744,
+      "grad_norm": 2.6087891633325646,
+      "learning_rate": 0.00012661391308525617,
+      "loss": 1.4592,
+      "step": 8109
+    },
+    {
+      "epoch": 1.2976,
+      "grad_norm": 3.289441330101583,
+      "learning_rate": 0.00012659726183910465,
+      "loss": 1.3923,
+      "step": 8110
+    },
+    {
+      "epoch": 1.29776,
+      "grad_norm": 2.096082197616072,
+      "learning_rate": 0.00012658060979932957,
+      "loss": 1.4611,
+      "step": 8111
+    },
+    {
+      "epoch": 1.29792,
+      "grad_norm": 2.432155141569759,
+      "learning_rate": 0.00012656395696642787,
+      "loss": 1.3996,
+      "step": 8112
+    },
+    {
+      "epoch": 1.2980800000000001,
+      "grad_norm": 3.4689660397212347,
+      "learning_rate": 0.00012654730334089637,
+      "loss": 1.3919,
+      "step": 8113
+    },
+    {
+      "epoch": 1.29824,
+      "grad_norm": 2.1992702465794323,
+      "learning_rate": 0.00012653064892323204,
+      "loss": 1.426,
+      "step": 8114
+    },
+    {
+      "epoch": 1.2984,
+      "grad_norm": 2.537276901853538,
+      "learning_rate": 0.00012651399371393175,
+      "loss": 1.4069,
+      "step": 8115
+    },
+    {
+      "epoch": 1.29856,
+      "grad_norm": 2.836426328106402,
+      "learning_rate": 0.00012649733771349258,
+      "loss": 1.3696,
+      "step": 8116
+    },
+    {
+      "epoch": 1.2987199999999999,
+      "grad_norm": 1.946964856657628,
+      "learning_rate": 0.00012648068092241144,
+      "loss": 1.3446,
+      "step": 8117
+    },
+    {
+      "epoch": 1.29888,
+      "grad_norm": 2.8272908823211305,
+      "learning_rate": 0.0001264640233411854,
+      "loss": 1.3692,
+      "step": 8118
+    },
+    {
+      "epoch": 1.29904,
+      "grad_norm": 2.119161480754853,
+      "learning_rate": 0.00012644736497031142,
+      "loss": 1.3666,
+      "step": 8119
+    },
+    {
+      "epoch": 1.2992,
+      "grad_norm": 2.0131976601078927,
+      "learning_rate": 0.00012643070581028662,
+      "loss": 1.4518,
+      "step": 8120
+    },
+    {
+      "epoch": 1.29936,
+      "grad_norm": 2.142350222031552,
+      "learning_rate": 0.0001264140458616081,
+      "loss": 1.3973,
+      "step": 8121
+    },
+    {
+      "epoch": 1.29952,
+      "grad_norm": 2.5504212496461007,
+      "learning_rate": 0.00012639738512477294,
+      "loss": 1.4722,
+      "step": 8122
+    },
+    {
+      "epoch": 1.29968,
+      "grad_norm": 2.0831523132206358,
+      "learning_rate": 0.0001263807236002782,
+      "loss": 1.3885,
+      "step": 8123
+    },
+    {
+      "epoch": 1.29984,
+      "grad_norm": 3.3343970378975722,
+      "learning_rate": 0.00012636406128862116,
+      "loss": 1.4218,
+      "step": 8124
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 2.8225301571350196,
+      "learning_rate": 0.00012634739819029897,
+      "loss": 1.4089,
+      "step": 8125
+    },
+    {
+      "epoch": 1.30016,
+      "grad_norm": 2.5298423045862335,
+      "learning_rate": 0.0001263307343058088,
+      "loss": 1.3377,
+      "step": 8126
+    },
+    {
+      "epoch": 1.30032,
+      "grad_norm": 2.306911938067522,
+      "learning_rate": 0.00012631406963564787,
+      "loss": 1.3052,
+      "step": 8127
+    },
+    {
+      "epoch": 1.30048,
+      "grad_norm": 2.1039278202299276,
+      "learning_rate": 0.00012629740418031347,
+      "loss": 1.3979,
+      "step": 8128
+    },
+    {
+      "epoch": 1.30064,
+      "grad_norm": 2.280987201762579,
+      "learning_rate": 0.00012628073794030282,
+      "loss": 1.4378,
+      "step": 8129
+    },
+    {
+      "epoch": 1.3008,
+      "grad_norm": 2.0799614213024284,
+      "learning_rate": 0.00012626407091611324,
+      "loss": 1.4164,
+      "step": 8130
+    },
+    {
+      "epoch": 1.30096,
+      "grad_norm": 2.3304091999225043,
+      "learning_rate": 0.00012624740310824206,
+      "loss": 1.3436,
+      "step": 8131
+    },
+    {
+      "epoch": 1.30112,
+      "grad_norm": 2.2831424386919092,
+      "learning_rate": 0.00012623073451718665,
+      "loss": 1.3725,
+      "step": 8132
+    },
+    {
+      "epoch": 1.30128,
+      "grad_norm": 2.2244111481290108,
+      "learning_rate": 0.0001262140651434443,
+      "loss": 1.3501,
+      "step": 8133
+    },
+    {
+      "epoch": 1.30144,
+      "grad_norm": 2.5012904392265254,
+      "learning_rate": 0.00012619739498751246,
+      "loss": 1.3633,
+      "step": 8134
+    },
+    {
+      "epoch": 1.3016,
+      "grad_norm": 2.127184050735505,
+      "learning_rate": 0.00012618072404988856,
+      "loss": 1.3665,
+      "step": 8135
+    },
+    {
+      "epoch": 1.30176,
+      "grad_norm": 2.771501139695493,
+      "learning_rate": 0.00012616405233106996,
+      "loss": 1.3926,
+      "step": 8136
+    },
+    {
+      "epoch": 1.30192,
+      "grad_norm": 2.5695931948283985,
+      "learning_rate": 0.00012614737983155419,
+      "loss": 1.3461,
+      "step": 8137
+    },
+    {
+      "epoch": 1.3020800000000001,
+      "grad_norm": 2.754030292387779,
+      "learning_rate": 0.0001261307065518387,
+      "loss": 1.4175,
+      "step": 8138
+    },
+    {
+      "epoch": 1.30224,
+      "grad_norm": 2.070131673414306,
+      "learning_rate": 0.00012611403249242098,
+      "loss": 1.3815,
+      "step": 8139
+    },
+    {
+      "epoch": 1.3024,
+      "grad_norm": 2.0699289239565704,
+      "learning_rate": 0.0001260973576537986,
+      "loss": 1.4657,
+      "step": 8140
+    },
+    {
+      "epoch": 1.30256,
+      "grad_norm": 2.0377477739379777,
+      "learning_rate": 0.0001260806820364691,
+      "loss": 1.3784,
+      "step": 8141
+    },
+    {
+      "epoch": 1.3027199999999999,
+      "grad_norm": 2.158583469961574,
+      "learning_rate": 0.00012606400564093006,
+      "loss": 1.3957,
+      "step": 8142
+    },
+    {
+      "epoch": 1.30288,
+      "grad_norm": 2.228431298543062,
+      "learning_rate": 0.00012604732846767906,
+      "loss": 1.3783,
+      "step": 8143
+    },
+    {
+      "epoch": 1.30304,
+      "grad_norm": 2.3317946202595814,
+      "learning_rate": 0.00012603065051721376,
+      "loss": 1.4133,
+      "step": 8144
+    },
+    {
+      "epoch": 1.3032,
+      "grad_norm": 2.010634187496623,
+      "learning_rate": 0.00012601397179003173,
+      "loss": 1.4361,
+      "step": 8145
+    },
+    {
+      "epoch": 1.30336,
+      "grad_norm": 2.6277152222335793,
+      "learning_rate": 0.0001259972922866307,
+      "loss": 1.4326,
+      "step": 8146
+    },
+    {
+      "epoch": 1.30352,
+      "grad_norm": 2.291818737515722,
+      "learning_rate": 0.00012598061200750835,
+      "loss": 1.3798,
+      "step": 8147
+    },
+    {
+      "epoch": 1.30368,
+      "grad_norm": 2.157291062571217,
+      "learning_rate": 0.00012596393095316242,
+      "loss": 1.4092,
+      "step": 8148
+    },
+    {
+      "epoch": 1.30384,
+      "grad_norm": 2.260873710842884,
+      "learning_rate": 0.00012594724912409058,
+      "loss": 1.3755,
+      "step": 8149
+    },
+    {
+      "epoch": 1.304,
+      "grad_norm": 1.970714499596521,
+      "learning_rate": 0.00012593056652079068,
+      "loss": 1.3649,
+      "step": 8150
+    },
+    {
+      "epoch": 1.30416,
+      "grad_norm": 2.8971038955621657,
+      "learning_rate": 0.00012591388314376045,
+      "loss": 1.4043,
+      "step": 8151
+    },
+    {
+      "epoch": 1.30432,
+      "grad_norm": 2.382106564181647,
+      "learning_rate": 0.0001258971989934977,
+      "loss": 1.4245,
+      "step": 8152
+    },
+    {
+      "epoch": 1.3044799999999999,
+      "grad_norm": 2.0859658539823753,
+      "learning_rate": 0.0001258805140705003,
+      "loss": 1.4391,
+      "step": 8153
+    },
+    {
+      "epoch": 1.30464,
+      "grad_norm": 1.9739788799061355,
+      "learning_rate": 0.00012586382837526602,
+      "loss": 1.4466,
+      "step": 8154
+    },
+    {
+      "epoch": 1.3048,
+      "grad_norm": 2.013196493083364,
+      "learning_rate": 0.00012584714190829282,
+      "loss": 1.3978,
+      "step": 8155
+    },
+    {
+      "epoch": 1.30496,
+      "grad_norm": 1.8646583809941593,
+      "learning_rate": 0.00012583045467007856,
+      "loss": 1.3656,
+      "step": 8156
+    },
+    {
+      "epoch": 1.30512,
+      "grad_norm": 2.0133340079622357,
+      "learning_rate": 0.00012581376666112115,
+      "loss": 1.4039,
+      "step": 8157
+    },
+    {
+      "epoch": 1.30528,
+      "grad_norm": 2.3412925997740692,
+      "learning_rate": 0.00012579707788191858,
+      "loss": 1.4189,
+      "step": 8158
+    },
+    {
+      "epoch": 1.30544,
+      "grad_norm": 1.8184757092616466,
+      "learning_rate": 0.00012578038833296881,
+      "loss": 1.3803,
+      "step": 8159
+    },
+    {
+      "epoch": 1.3056,
+      "grad_norm": 2.0606108053428915,
+      "learning_rate": 0.0001257636980147698,
+      "loss": 1.3724,
+      "step": 8160
+    },
+    {
+      "epoch": 1.30576,
+      "grad_norm": 2.3706788901613427,
+      "learning_rate": 0.0001257470069278196,
+      "loss": 1.4156,
+      "step": 8161
+    },
+    {
+      "epoch": 1.30592,
+      "grad_norm": 2.1832658405746135,
+      "learning_rate": 0.00012573031507261622,
+      "loss": 1.3844,
+      "step": 8162
+    },
+    {
+      "epoch": 1.3060800000000001,
+      "grad_norm": 2.359649175316109,
+      "learning_rate": 0.00012571362244965772,
+      "loss": 1.3619,
+      "step": 8163
+    },
+    {
+      "epoch": 1.30624,
+      "grad_norm": 1.9612757969937409,
+      "learning_rate": 0.00012569692905944218,
+      "loss": 1.3698,
+      "step": 8164
+    },
+    {
+      "epoch": 1.3064,
+      "grad_norm": 2.175913025886537,
+      "learning_rate": 0.00012568023490246776,
+      "loss": 1.4096,
+      "step": 8165
+    },
+    {
+      "epoch": 1.30656,
+      "grad_norm": 2.5253280137279006,
+      "learning_rate": 0.00012566353997923253,
+      "loss": 1.3735,
+      "step": 8166
+    },
+    {
+      "epoch": 1.3067199999999999,
+      "grad_norm": 2.3065937624791153,
+      "learning_rate": 0.00012564684429023469,
+      "loss": 1.4507,
+      "step": 8167
+    },
+    {
+      "epoch": 1.30688,
+      "grad_norm": 2.59025636059318,
+      "learning_rate": 0.00012563014783597236,
+      "loss": 1.4202,
+      "step": 8168
+    },
+    {
+      "epoch": 1.30704,
+      "grad_norm": 2.3862070412287815,
+      "learning_rate": 0.00012561345061694377,
+      "loss": 1.4174,
+      "step": 8169
+    },
+    {
+      "epoch": 1.3072,
+      "grad_norm": 2.456393483246187,
+      "learning_rate": 0.00012559675263364715,
+      "loss": 1.4118,
+      "step": 8170
+    },
+    {
+      "epoch": 1.30736,
+      "grad_norm": 2.382762516290906,
+      "learning_rate": 0.00012558005388658072,
+      "loss": 1.3273,
+      "step": 8171
+    },
+    {
+      "epoch": 1.30752,
+      "grad_norm": 2.1919182654547034,
+      "learning_rate": 0.00012556335437624274,
+      "loss": 1.4278,
+      "step": 8172
+    },
+    {
+      "epoch": 1.30768,
+      "grad_norm": 2.127172532173687,
+      "learning_rate": 0.0001255466541031315,
+      "loss": 1.3525,
+      "step": 8173
+    },
+    {
+      "epoch": 1.3078400000000001,
+      "grad_norm": 2.067010401283253,
+      "learning_rate": 0.00012552995306774537,
+      "loss": 1.4261,
+      "step": 8174
+    },
+    {
+      "epoch": 1.308,
+      "grad_norm": 2.568614038086538,
+      "learning_rate": 0.00012551325127058263,
+      "loss": 1.3616,
+      "step": 8175
+    },
+    {
+      "epoch": 1.30816,
+      "grad_norm": 2.5975437477405645,
+      "learning_rate": 0.00012549654871214167,
+      "loss": 1.4427,
+      "step": 8176
+    },
+    {
+      "epoch": 1.30832,
+      "grad_norm": 2.3418627244555412,
+      "learning_rate": 0.00012547984539292082,
+      "loss": 1.3916,
+      "step": 8177
+    },
+    {
+      "epoch": 1.3084799999999999,
+      "grad_norm": 3.2286869091690886,
+      "learning_rate": 0.00012546314131341851,
+      "loss": 1.4251,
+      "step": 8178
+    },
+    {
+      "epoch": 1.30864,
+      "grad_norm": 2.196450953945995,
+      "learning_rate": 0.00012544643647413318,
+      "loss": 1.4289,
+      "step": 8179
+    },
+    {
+      "epoch": 1.3088,
+      "grad_norm": 2.8032628580228,
+      "learning_rate": 0.00012542973087556326,
+      "loss": 1.3947,
+      "step": 8180
+    },
+    {
+      "epoch": 1.30896,
+      "grad_norm": 2.010947958099227,
+      "learning_rate": 0.00012541302451820723,
+      "loss": 1.4484,
+      "step": 8181
+    },
+    {
+      "epoch": 1.30912,
+      "grad_norm": 2.4629565199720687,
+      "learning_rate": 0.00012539631740256357,
+      "loss": 1.579,
+      "step": 8182
+    },
+    {
+      "epoch": 1.30928,
+      "grad_norm": 2.285357517151367,
+      "learning_rate": 0.00012537960952913082,
+      "loss": 1.3629,
+      "step": 8183
+    },
+    {
+      "epoch": 1.30944,
+      "grad_norm": 1.9870403825777507,
+      "learning_rate": 0.00012536290089840748,
+      "loss": 1.352,
+      "step": 8184
+    },
+    {
+      "epoch": 1.3096,
+      "grad_norm": 1.9653746639019931,
+      "learning_rate": 0.00012534619151089216,
+      "loss": 1.339,
+      "step": 8185
+    },
+    {
+      "epoch": 1.30976,
+      "grad_norm": 2.6243486221780006,
+      "learning_rate": 0.0001253294813670834,
+      "loss": 1.4099,
+      "step": 8186
+    },
+    {
+      "epoch": 1.30992,
+      "grad_norm": 2.1482722632239453,
+      "learning_rate": 0.0001253127704674798,
+      "loss": 1.4026,
+      "step": 8187
+    },
+    {
+      "epoch": 1.3100800000000001,
+      "grad_norm": 2.253747604007526,
+      "learning_rate": 0.00012529605881258005,
+      "loss": 1.4007,
+      "step": 8188
+    },
+    {
+      "epoch": 1.31024,
+      "grad_norm": 2.267924106404915,
+      "learning_rate": 0.00012527934640288278,
+      "loss": 1.4154,
+      "step": 8189
+    },
+    {
+      "epoch": 1.3104,
+      "grad_norm": 2.143439596741203,
+      "learning_rate": 0.00012526263323888663,
+      "loss": 1.3819,
+      "step": 8190
+    },
+    {
+      "epoch": 1.31056,
+      "grad_norm": 2.1170246005692155,
+      "learning_rate": 0.00012524591932109032,
+      "loss": 1.4515,
+      "step": 8191
+    },
+    {
+      "epoch": 1.3107199999999999,
+      "grad_norm": 2.1222379450395668,
+      "learning_rate": 0.00012522920464999254,
+      "loss": 1.3849,
+      "step": 8192
+    },
+    {
+      "epoch": 1.31088,
+      "grad_norm": 2.0329326542982473,
+      "learning_rate": 0.0001252124892260921,
+      "loss": 1.3529,
+      "step": 8193
+    },
+    {
+      "epoch": 1.31104,
+      "grad_norm": 2.1902568835167306,
+      "learning_rate": 0.0001251957730498877,
+      "loss": 1.4222,
+      "step": 8194
+    },
+    {
+      "epoch": 1.3112,
+      "grad_norm": 2.11057982236774,
+      "learning_rate": 0.00012517905612187814,
+      "loss": 1.4196,
+      "step": 8195
+    },
+    {
+      "epoch": 1.31136,
+      "grad_norm": 2.284484212845412,
+      "learning_rate": 0.00012516233844256223,
+      "loss": 1.3739,
+      "step": 8196
+    },
+    {
+      "epoch": 1.31152,
+      "grad_norm": 2.408028314668167,
+      "learning_rate": 0.00012514562001243878,
+      "loss": 1.4304,
+      "step": 8197
+    },
+    {
+      "epoch": 1.31168,
+      "grad_norm": 2.296535963892679,
+      "learning_rate": 0.0001251289008320067,
+      "loss": 1.3629,
+      "step": 8198
+    },
+    {
+      "epoch": 1.3118400000000001,
+      "grad_norm": 2.286999092766501,
+      "learning_rate": 0.00012511218090176487,
+      "loss": 1.4272,
+      "step": 8199
+    },
+    {
+      "epoch": 1.312,
+      "grad_norm": 2.1028319384540946,
+      "learning_rate": 0.00012509546022221213,
+      "loss": 1.4022,
+      "step": 8200
+    },
+    {
+      "epoch": 1.31216,
+      "grad_norm": 2.0163056843795717,
+      "learning_rate": 0.00012507873879384743,
+      "loss": 1.3954,
+      "step": 8201
+    },
+    {
+      "epoch": 1.31232,
+      "grad_norm": 2.0249090275942767,
+      "learning_rate": 0.0001250620166171697,
+      "loss": 1.4484,
+      "step": 8202
+    },
+    {
+      "epoch": 1.3124799999999999,
+      "grad_norm": 2.191530593979682,
+      "learning_rate": 0.0001250452936926779,
+      "loss": 1.4037,
+      "step": 8203
+    },
+    {
+      "epoch": 1.31264,
+      "grad_norm": 2.130562445431204,
+      "learning_rate": 0.000125028570020871,
+      "loss": 1.3837,
+      "step": 8204
+    },
+    {
+      "epoch": 1.3128,
+      "grad_norm": 1.984839598042556,
+      "learning_rate": 0.0001250118456022481,
+      "loss": 1.3771,
+      "step": 8205
+    },
+    {
+      "epoch": 1.31296,
+      "grad_norm": 2.466683386856507,
+      "learning_rate": 0.00012499512043730816,
+      "loss": 1.4349,
+      "step": 8206
+    },
+    {
+      "epoch": 1.31312,
+      "grad_norm": 1.755483923109123,
+      "learning_rate": 0.00012497839452655022,
+      "loss": 1.3823,
+      "step": 8207
+    },
+    {
+      "epoch": 1.31328,
+      "grad_norm": 2.4989485113351666,
+      "learning_rate": 0.0001249616678704734,
+      "loss": 1.4214,
+      "step": 8208
+    },
+    {
+      "epoch": 1.31344,
+      "grad_norm": 2.5323934344449692,
+      "learning_rate": 0.0001249449404695768,
+      "loss": 1.3957,
+      "step": 8209
+    },
+    {
+      "epoch": 1.3136,
+      "grad_norm": 2.2112510452321654,
+      "learning_rate": 0.00012492821232435954,
+      "loss": 1.4292,
+      "step": 8210
+    },
+    {
+      "epoch": 1.31376,
+      "grad_norm": 2.3521503943776305,
+      "learning_rate": 0.0001249114834353207,
+      "loss": 1.3746,
+      "step": 8211
+    },
+    {
+      "epoch": 1.31392,
+      "grad_norm": 2.1133507087966557,
+      "learning_rate": 0.00012489475380295955,
+      "loss": 1.3784,
+      "step": 8212
+    },
+    {
+      "epoch": 1.3140800000000001,
+      "grad_norm": 2.171837613173987,
+      "learning_rate": 0.00012487802342777516,
+      "loss": 1.326,
+      "step": 8213
+    },
+    {
+      "epoch": 1.31424,
+      "grad_norm": 3.504304356658582,
+      "learning_rate": 0.00012486129231026685,
+      "loss": 1.4541,
+      "step": 8214
+    },
+    {
+      "epoch": 1.3144,
+      "grad_norm": 2.0181634628332294,
+      "learning_rate": 0.00012484456045093378,
+      "loss": 1.4202,
+      "step": 8215
+    },
+    {
+      "epoch": 1.31456,
+      "grad_norm": 2.5732636558611652,
+      "learning_rate": 0.00012482782785027525,
+      "loss": 1.3442,
+      "step": 8216
+    },
+    {
+      "epoch": 1.3147199999999999,
+      "grad_norm": 2.1575762911912957,
+      "learning_rate": 0.0001248110945087905,
+      "loss": 1.4152,
+      "step": 8217
+    },
+    {
+      "epoch": 1.31488,
+      "grad_norm": 2.320084449506193,
+      "learning_rate": 0.00012479436042697883,
+      "loss": 1.43,
+      "step": 8218
+    },
+    {
+      "epoch": 1.31504,
+      "grad_norm": 2.023183754661037,
+      "learning_rate": 0.00012477762560533957,
+      "loss": 1.3439,
+      "step": 8219
+    },
+    {
+      "epoch": 1.3152,
+      "grad_norm": 2.1429350816912027,
+      "learning_rate": 0.0001247608900443721,
+      "loss": 1.4891,
+      "step": 8220
+    },
+    {
+      "epoch": 1.31536,
+      "grad_norm": 2.1983783080608235,
+      "learning_rate": 0.00012474415374457573,
+      "loss": 1.3636,
+      "step": 8221
+    },
+    {
+      "epoch": 1.31552,
+      "grad_norm": 1.904772133048976,
+      "learning_rate": 0.00012472741670644988,
+      "loss": 1.4093,
+      "step": 8222
+    },
+    {
+      "epoch": 1.31568,
+      "grad_norm": 2.2159730138916958,
+      "learning_rate": 0.00012471067893049395,
+      "loss": 1.4548,
+      "step": 8223
+    },
+    {
+      "epoch": 1.3158400000000001,
+      "grad_norm": 2.033818039129985,
+      "learning_rate": 0.00012469394041720737,
+      "loss": 1.3815,
+      "step": 8224
+    },
+    {
+      "epoch": 1.316,
+      "grad_norm": 2.1847686658610375,
+      "learning_rate": 0.00012467720116708956,
+      "loss": 1.4375,
+      "step": 8225
+    },
+    {
+      "epoch": 1.31616,
+      "grad_norm": 1.979294472062637,
+      "learning_rate": 0.00012466046118064005,
+      "loss": 1.3599,
+      "step": 8226
+    },
+    {
+      "epoch": 1.31632,
+      "grad_norm": 1.9073100305847126,
+      "learning_rate": 0.00012464372045835833,
+      "loss": 1.3786,
+      "step": 8227
+    },
+    {
+      "epoch": 1.3164799999999999,
+      "grad_norm": 2.163325051054943,
+      "learning_rate": 0.00012462697900074386,
+      "loss": 1.3509,
+      "step": 8228
+    },
+    {
+      "epoch": 1.31664,
+      "grad_norm": 2.283155209195169,
+      "learning_rate": 0.00012461023680829624,
+      "loss": 1.4331,
+      "step": 8229
+    },
+    {
+      "epoch": 1.3168,
+      "grad_norm": 2.161739791289864,
+      "learning_rate": 0.00012459349388151502,
+      "loss": 1.3903,
+      "step": 8230
+    },
+    {
+      "epoch": 1.31696,
+      "grad_norm": 1.958362083033666,
+      "learning_rate": 0.00012457675022089978,
+      "loss": 1.4391,
+      "step": 8231
+    },
+    {
+      "epoch": 1.31712,
+      "grad_norm": 2.331441080909169,
+      "learning_rate": 0.0001245600058269501,
+      "loss": 1.4637,
+      "step": 8232
+    },
+    {
+      "epoch": 1.31728,
+      "grad_norm": 2.3198210954695293,
+      "learning_rate": 0.00012454326070016566,
+      "loss": 1.3908,
+      "step": 8233
+    },
+    {
+      "epoch": 1.31744,
+      "grad_norm": 1.957022257552395,
+      "learning_rate": 0.0001245265148410461,
+      "loss": 1.3849,
+      "step": 8234
+    },
+    {
+      "epoch": 1.3176,
+      "grad_norm": 2.0932091688378183,
+      "learning_rate": 0.00012450976825009106,
+      "loss": 1.3854,
+      "step": 8235
+    },
+    {
+      "epoch": 1.31776,
+      "grad_norm": 2.3606872054356405,
+      "learning_rate": 0.00012449302092780025,
+      "loss": 1.4175,
+      "step": 8236
+    },
+    {
+      "epoch": 1.31792,
+      "grad_norm": 2.469006042824037,
+      "learning_rate": 0.00012447627287467337,
+      "loss": 1.3743,
+      "step": 8237
+    },
+    {
+      "epoch": 1.31808,
+      "grad_norm": 2.282543215253696,
+      "learning_rate": 0.00012445952409121018,
+      "loss": 1.417,
+      "step": 8238
+    },
+    {
+      "epoch": 1.31824,
+      "grad_norm": 2.2280649010803684,
+      "learning_rate": 0.00012444277457791044,
+      "loss": 1.4289,
+      "step": 8239
+    },
+    {
+      "epoch": 1.3184,
+      "grad_norm": 1.873758928571891,
+      "learning_rate": 0.0001244260243352739,
+      "loss": 1.3437,
+      "step": 8240
+    },
+    {
+      "epoch": 1.31856,
+      "grad_norm": 2.1728905604645394,
+      "learning_rate": 0.00012440927336380042,
+      "loss": 1.4339,
+      "step": 8241
+    },
+    {
+      "epoch": 1.31872,
+      "grad_norm": 2.3204020996986086,
+      "learning_rate": 0.00012439252166398978,
+      "loss": 1.3342,
+      "step": 8242
+    },
+    {
+      "epoch": 1.31888,
+      "grad_norm": 1.7633232832147536,
+      "learning_rate": 0.00012437576923634183,
+      "loss": 1.3525,
+      "step": 8243
+    },
+    {
+      "epoch": 1.31904,
+      "grad_norm": 2.591208739973262,
+      "learning_rate": 0.00012435901608135645,
+      "loss": 1.4328,
+      "step": 8244
+    },
+    {
+      "epoch": 1.3192,
+      "grad_norm": 1.9405462600720478,
+      "learning_rate": 0.0001243422621995335,
+      "loss": 1.4802,
+      "step": 8245
+    },
+    {
+      "epoch": 1.31936,
+      "grad_norm": 2.3490750292591556,
+      "learning_rate": 0.00012432550759137296,
+      "loss": 1.3832,
+      "step": 8246
+    },
+    {
+      "epoch": 1.31952,
+      "grad_norm": 2.621227408886565,
+      "learning_rate": 0.0001243087522573747,
+      "loss": 1.4258,
+      "step": 8247
+    },
+    {
+      "epoch": 1.31968,
+      "grad_norm": 2.113709467585038,
+      "learning_rate": 0.0001242919961980387,
+      "loss": 1.396,
+      "step": 8248
+    },
+    {
+      "epoch": 1.3198400000000001,
+      "grad_norm": 2.242301807450352,
+      "learning_rate": 0.00012427523941386493,
+      "loss": 1.3156,
+      "step": 8249
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 2.02632431664991,
+      "learning_rate": 0.00012425848190535338,
+      "loss": 1.325,
+      "step": 8250
+    },
+    {
+      "epoch": 1.32016,
+      "grad_norm": 2.1968082300927616,
+      "learning_rate": 0.0001242417236730041,
+      "loss": 1.4114,
+      "step": 8251
+    },
+    {
+      "epoch": 1.32032,
+      "grad_norm": 2.1026776523868533,
+      "learning_rate": 0.00012422496471731708,
+      "loss": 1.3655,
+      "step": 8252
+    },
+    {
+      "epoch": 1.3204799999999999,
+      "grad_norm": 2.171890493092083,
+      "learning_rate": 0.0001242082050387924,
+      "loss": 1.3964,
+      "step": 8253
+    },
+    {
+      "epoch": 1.32064,
+      "grad_norm": 2.2117522243307635,
+      "learning_rate": 0.0001241914446379302,
+      "loss": 1.3133,
+      "step": 8254
+    },
+    {
+      "epoch": 1.3208,
+      "grad_norm": 2.3838655338495203,
+      "learning_rate": 0.0001241746835152305,
+      "loss": 1.3812,
+      "step": 8255
+    },
+    {
+      "epoch": 1.32096,
+      "grad_norm": 2.46502653930894,
+      "learning_rate": 0.0001241579216711935,
+      "loss": 1.4325,
+      "step": 8256
+    },
+    {
+      "epoch": 1.32112,
+      "grad_norm": 1.991807329048338,
+      "learning_rate": 0.00012414115910631935,
+      "loss": 1.3662,
+      "step": 8257
+    },
+    {
+      "epoch": 1.32128,
+      "grad_norm": 2.1708401929661676,
+      "learning_rate": 0.00012412439582110813,
+      "loss": 1.3792,
+      "step": 8258
+    },
+    {
+      "epoch": 1.32144,
+      "grad_norm": 2.0481093821925356,
+      "learning_rate": 0.00012410763181606012,
+      "loss": 1.3674,
+      "step": 8259
+    },
+    {
+      "epoch": 1.3216,
+      "grad_norm": 2.4408377750645758,
+      "learning_rate": 0.0001240908670916755,
+      "loss": 1.3726,
+      "step": 8260
+    },
+    {
+      "epoch": 1.32176,
+      "grad_norm": 2.1410718393094004,
+      "learning_rate": 0.0001240741016484545,
+      "loss": 1.3987,
+      "step": 8261
+    },
+    {
+      "epoch": 1.32192,
+      "grad_norm": 2.072085337396898,
+      "learning_rate": 0.00012405733548689739,
+      "loss": 1.3921,
+      "step": 8262
+    },
+    {
+      "epoch": 1.32208,
+      "grad_norm": 2.0261860452974503,
+      "learning_rate": 0.00012404056860750446,
+      "loss": 1.4106,
+      "step": 8263
+    },
+    {
+      "epoch": 1.32224,
+      "grad_norm": 2.1909078137926725,
+      "learning_rate": 0.00012402380101077598,
+      "loss": 1.3441,
+      "step": 8264
+    },
+    {
+      "epoch": 1.3224,
+      "grad_norm": 2.1215729081942123,
+      "learning_rate": 0.0001240070326972123,
+      "loss": 1.3824,
+      "step": 8265
+    },
+    {
+      "epoch": 1.32256,
+      "grad_norm": 2.4628790788595882,
+      "learning_rate": 0.0001239902636673137,
+      "loss": 1.4118,
+      "step": 8266
+    },
+    {
+      "epoch": 1.32272,
+      "grad_norm": 2.194874373675503,
+      "learning_rate": 0.00012397349392158065,
+      "loss": 1.3401,
+      "step": 8267
+    },
+    {
+      "epoch": 1.32288,
+      "grad_norm": 2.6459704881570603,
+      "learning_rate": 0.00012395672346051344,
+      "loss": 1.4382,
+      "step": 8268
+    },
+    {
+      "epoch": 1.32304,
+      "grad_norm": 2.104921617617967,
+      "learning_rate": 0.00012393995228461251,
+      "loss": 1.393,
+      "step": 8269
+    },
+    {
+      "epoch": 1.3232,
+      "grad_norm": 2.0960055085514586,
+      "learning_rate": 0.0001239231803943783,
+      "loss": 1.3493,
+      "step": 8270
+    },
+    {
+      "epoch": 1.32336,
+      "grad_norm": 2.208867996052485,
+      "learning_rate": 0.00012390640779031122,
+      "loss": 1.4199,
+      "step": 8271
+    },
+    {
+      "epoch": 1.32352,
+      "grad_norm": 2.2997557911767257,
+      "learning_rate": 0.0001238896344729118,
+      "loss": 1.3276,
+      "step": 8272
+    },
+    {
+      "epoch": 1.32368,
+      "grad_norm": 2.204489671162299,
+      "learning_rate": 0.00012387286044268048,
+      "loss": 1.3984,
+      "step": 8273
+    },
+    {
+      "epoch": 1.3238400000000001,
+      "grad_norm": 2.373251367885678,
+      "learning_rate": 0.0001238560857001178,
+      "loss": 1.4,
+      "step": 8274
+    },
+    {
+      "epoch": 1.324,
+      "grad_norm": 2.551268519510862,
+      "learning_rate": 0.0001238393102457243,
+      "loss": 1.4206,
+      "step": 8275
+    },
+    {
+      "epoch": 1.32416,
+      "grad_norm": 2.4734159726519898,
+      "learning_rate": 0.00012382253408000052,
+      "loss": 1.4062,
+      "step": 8276
+    },
+    {
+      "epoch": 1.32432,
+      "grad_norm": 2.519589455665964,
+      "learning_rate": 0.000123805757203447,
+      "loss": 1.3921,
+      "step": 8277
+    },
+    {
+      "epoch": 1.3244799999999999,
+      "grad_norm": 1.9076116397189822,
+      "learning_rate": 0.00012378897961656438,
+      "loss": 1.3441,
+      "step": 8278
+    },
+    {
+      "epoch": 1.32464,
+      "grad_norm": 2.5741189432192164,
+      "learning_rate": 0.0001237722013198533,
+      "loss": 1.3402,
+      "step": 8279
+    },
+    {
+      "epoch": 1.3248,
+      "grad_norm": 2.7130975928073218,
+      "learning_rate": 0.00012375542231381434,
+      "loss": 1.329,
+      "step": 8280
+    },
+    {
+      "epoch": 1.32496,
+      "grad_norm": 2.418427725419326,
+      "learning_rate": 0.00012373864259894823,
+      "loss": 1.4461,
+      "step": 8281
+    },
+    {
+      "epoch": 1.32512,
+      "grad_norm": 2.1404751621117124,
+      "learning_rate": 0.0001237218621757556,
+      "loss": 1.3381,
+      "step": 8282
+    },
+    {
+      "epoch": 1.32528,
+      "grad_norm": 2.0123008130462594,
+      "learning_rate": 0.00012370508104473716,
+      "loss": 1.3184,
+      "step": 8283
+    },
+    {
+      "epoch": 1.32544,
+      "grad_norm": 1.9403227789674355,
+      "learning_rate": 0.00012368829920639366,
+      "loss": 1.3011,
+      "step": 8284
+    },
+    {
+      "epoch": 1.3256000000000001,
+      "grad_norm": 2.541086636184375,
+      "learning_rate": 0.00012367151666122582,
+      "loss": 1.4508,
+      "step": 8285
+    },
+    {
+      "epoch": 1.32576,
+      "grad_norm": 1.8735377640174145,
+      "learning_rate": 0.00012365473340973442,
+      "loss": 1.4031,
+      "step": 8286
+    },
+    {
+      "epoch": 1.32592,
+      "grad_norm": 2.170331178900119,
+      "learning_rate": 0.00012363794945242023,
+      "loss": 1.5163,
+      "step": 8287
+    },
+    {
+      "epoch": 1.32608,
+      "grad_norm": 2.207436728561986,
+      "learning_rate": 0.00012362116478978406,
+      "loss": 1.3944,
+      "step": 8288
+    },
+    {
+      "epoch": 1.3262399999999999,
+      "grad_norm": 2.1020056408317425,
+      "learning_rate": 0.0001236043794223268,
+      "loss": 1.3733,
+      "step": 8289
+    },
+    {
+      "epoch": 1.3264,
+      "grad_norm": 2.3786474809871696,
+      "learning_rate": 0.0001235875933505492,
+      "loss": 1.4227,
+      "step": 8290
+    },
+    {
+      "epoch": 1.32656,
+      "grad_norm": 2.8460256682810527,
+      "learning_rate": 0.00012357080657495224,
+      "loss": 1.395,
+      "step": 8291
+    },
+    {
+      "epoch": 1.32672,
+      "grad_norm": 1.8651816755890158,
+      "learning_rate": 0.0001235540190960367,
+      "loss": 1.4227,
+      "step": 8292
+    },
+    {
+      "epoch": 1.32688,
+      "grad_norm": 2.598025924507524,
+      "learning_rate": 0.0001235372309143036,
+      "loss": 1.4683,
+      "step": 8293
+    },
+    {
+      "epoch": 1.32704,
+      "grad_norm": 2.214782100114404,
+      "learning_rate": 0.0001235204420302538,
+      "loss": 1.4119,
+      "step": 8294
+    },
+    {
+      "epoch": 1.3272,
+      "grad_norm": 2.560058050191506,
+      "learning_rate": 0.0001235036524443883,
+      "loss": 1.338,
+      "step": 8295
+    },
+    {
+      "epoch": 1.32736,
+      "grad_norm": 2.7759132367299344,
+      "learning_rate": 0.00012348686215720806,
+      "loss": 1.3332,
+      "step": 8296
+    },
+    {
+      "epoch": 1.32752,
+      "grad_norm": 2.2333427419987246,
+      "learning_rate": 0.00012347007116921406,
+      "loss": 1.3448,
+      "step": 8297
+    },
+    {
+      "epoch": 1.32768,
+      "grad_norm": 2.4452200800973007,
+      "learning_rate": 0.00012345327948090734,
+      "loss": 1.4136,
+      "step": 8298
+    },
+    {
+      "epoch": 1.3278400000000001,
+      "grad_norm": 2.3884552580000027,
+      "learning_rate": 0.00012343648709278894,
+      "loss": 1.3803,
+      "step": 8299
+    },
+    {
+      "epoch": 1.328,
+      "grad_norm": 2.3597226981265695,
+      "learning_rate": 0.00012341969400535993,
+      "loss": 1.4338,
+      "step": 8300
+    },
+    {
+      "epoch": 1.32816,
+      "grad_norm": 2.0653819230122075,
+      "learning_rate": 0.00012340290021912136,
+      "loss": 1.4139,
+      "step": 8301
+    },
+    {
+      "epoch": 1.32832,
+      "grad_norm": 3.099974198950367,
+      "learning_rate": 0.00012338610573457435,
+      "loss": 1.4129,
+      "step": 8302
+    },
+    {
+      "epoch": 1.3284799999999999,
+      "grad_norm": 2.272460327788203,
+      "learning_rate": 0.00012336931055222,
+      "loss": 1.327,
+      "step": 8303
+    },
+    {
+      "epoch": 1.32864,
+      "grad_norm": 2.5536401794470236,
+      "learning_rate": 0.0001233525146725595,
+      "loss": 1.3427,
+      "step": 8304
+    },
+    {
+      "epoch": 1.3288,
+      "grad_norm": 2.179672539296313,
+      "learning_rate": 0.000123335718096094,
+      "loss": 1.3451,
+      "step": 8305
+    },
+    {
+      "epoch": 1.32896,
+      "grad_norm": 2.358669695003856,
+      "learning_rate": 0.00012331892082332468,
+      "loss": 1.3427,
+      "step": 8306
+    },
+    {
+      "epoch": 1.32912,
+      "grad_norm": 2.143138201673539,
+      "learning_rate": 0.00012330212285475273,
+      "loss": 1.3291,
+      "step": 8307
+    },
+    {
+      "epoch": 1.32928,
+      "grad_norm": 2.206689451469264,
+      "learning_rate": 0.00012328532419087938,
+      "loss": 1.4013,
+      "step": 8308
+    },
+    {
+      "epoch": 1.32944,
+      "grad_norm": 2.3036450611332784,
+      "learning_rate": 0.0001232685248322059,
+      "loss": 1.4823,
+      "step": 8309
+    },
+    {
+      "epoch": 1.3296000000000001,
+      "grad_norm": 2.661868658357052,
+      "learning_rate": 0.0001232517247792335,
+      "loss": 1.4931,
+      "step": 8310
+    },
+    {
+      "epoch": 1.32976,
+      "grad_norm": 2.3666455922911602,
+      "learning_rate": 0.00012323492403246356,
+      "loss": 1.3744,
+      "step": 8311
+    },
+    {
+      "epoch": 1.32992,
+      "grad_norm": 2.1538998878144393,
+      "learning_rate": 0.00012321812259239736,
+      "loss": 1.5144,
+      "step": 8312
+    },
+    {
+      "epoch": 1.33008,
+      "grad_norm": 2.1789184318393993,
+      "learning_rate": 0.0001232013204595362,
+      "loss": 1.3352,
+      "step": 8313
+    },
+    {
+      "epoch": 1.3302399999999999,
+      "grad_norm": 2.3478521475417535,
+      "learning_rate": 0.0001231845176343814,
+      "loss": 1.4817,
+      "step": 8314
+    },
+    {
+      "epoch": 1.3304,
+      "grad_norm": 1.7819815892716582,
+      "learning_rate": 0.00012316771411743444,
+      "loss": 1.4249,
+      "step": 8315
+    },
+    {
+      "epoch": 1.33056,
+      "grad_norm": 1.9666522440132193,
+      "learning_rate": 0.0001231509099091966,
+      "loss": 1.4441,
+      "step": 8316
+    },
+    {
+      "epoch": 1.33072,
+      "grad_norm": 2.4847287734799632,
+      "learning_rate": 0.0001231341050101694,
+      "loss": 1.433,
+      "step": 8317
+    },
+    {
+      "epoch": 1.33088,
+      "grad_norm": 2.7814386003233267,
+      "learning_rate": 0.00012311729942085414,
+      "loss": 1.5224,
+      "step": 8318
+    },
+    {
+      "epoch": 1.33104,
+      "grad_norm": 2.131415526773546,
+      "learning_rate": 0.0001231004931417524,
+      "loss": 1.3873,
+      "step": 8319
+    },
+    {
+      "epoch": 1.3312,
+      "grad_norm": 2.2801944973247052,
+      "learning_rate": 0.00012308368617336562,
+      "loss": 1.4431,
+      "step": 8320
+    },
+    {
+      "epoch": 1.33136,
+      "grad_norm": 1.8236687568867058,
+      "learning_rate": 0.00012306687851619525,
+      "loss": 1.3925,
+      "step": 8321
+    },
+    {
+      "epoch": 1.33152,
+      "grad_norm": 2.1196378807737086,
+      "learning_rate": 0.00012305007017074284,
+      "loss": 1.4097,
+      "step": 8322
+    },
+    {
+      "epoch": 1.33168,
+      "grad_norm": 3.0121789970038715,
+      "learning_rate": 0.00012303326113750993,
+      "loss": 1.4457,
+      "step": 8323
+    },
+    {
+      "epoch": 1.3318400000000001,
+      "grad_norm": 2.1661665333021083,
+      "learning_rate": 0.00012301645141699806,
+      "loss": 1.4176,
+      "step": 8324
+    },
+    {
+      "epoch": 1.332,
+      "grad_norm": 2.8205588055630995,
+      "learning_rate": 0.00012299964100970882,
+      "loss": 1.4015,
+      "step": 8325
+    },
+    {
+      "epoch": 1.33216,
+      "grad_norm": 2.4232644079799086,
+      "learning_rate": 0.00012298282991614383,
+      "loss": 1.3871,
+      "step": 8326
+    },
+    {
+      "epoch": 1.33232,
+      "grad_norm": 2.7653196187263127,
+      "learning_rate": 0.00012296601813680465,
+      "loss": 1.4362,
+      "step": 8327
+    },
+    {
+      "epoch": 1.3324799999999999,
+      "grad_norm": 2.4515570619196603,
+      "learning_rate": 0.00012294920567219297,
+      "loss": 1.4438,
+      "step": 8328
+    },
+    {
+      "epoch": 1.33264,
+      "grad_norm": 2.038862572209322,
+      "learning_rate": 0.00012293239252281045,
+      "loss": 1.3498,
+      "step": 8329
+    },
+    {
+      "epoch": 1.3328,
+      "grad_norm": 2.129430753501875,
+      "learning_rate": 0.00012291557868915874,
+      "loss": 1.3907,
+      "step": 8330
+    },
+    {
+      "epoch": 1.33296,
+      "grad_norm": 2.1481596598069377,
+      "learning_rate": 0.00012289876417173956,
+      "loss": 1.4107,
+      "step": 8331
+    },
+    {
+      "epoch": 1.33312,
+      "grad_norm": 2.0137631665820637,
+      "learning_rate": 0.0001228819489710546,
+      "loss": 1.3478,
+      "step": 8332
+    },
+    {
+      "epoch": 1.33328,
+      "grad_norm": 2.3297616981219584,
+      "learning_rate": 0.00012286513308760566,
+      "loss": 1.3743,
+      "step": 8333
+    },
+    {
+      "epoch": 1.33344,
+      "grad_norm": 1.8969854674695281,
+      "learning_rate": 0.00012284831652189444,
+      "loss": 1.3467,
+      "step": 8334
+    },
+    {
+      "epoch": 1.3336000000000001,
+      "grad_norm": 2.0379062270849957,
+      "learning_rate": 0.00012283149927442278,
+      "loss": 1.3872,
+      "step": 8335
+    },
+    {
+      "epoch": 1.33376,
+      "grad_norm": 2.203806294739254,
+      "learning_rate": 0.00012281468134569238,
+      "loss": 1.4322,
+      "step": 8336
+    },
+    {
+      "epoch": 1.33392,
+      "grad_norm": 2.306370953154781,
+      "learning_rate": 0.00012279786273620519,
+      "loss": 1.4083,
+      "step": 8337
+    },
+    {
+      "epoch": 1.33408,
+      "grad_norm": 2.233486867501631,
+      "learning_rate": 0.00012278104344646298,
+      "loss": 1.4181,
+      "step": 8338
+    },
+    {
+      "epoch": 1.3342399999999999,
+      "grad_norm": 2.4133101406417103,
+      "learning_rate": 0.00012276422347696765,
+      "loss": 1.4115,
+      "step": 8339
+    },
+    {
+      "epoch": 1.3344,
+      "grad_norm": 2.2482443117907236,
+      "learning_rate": 0.00012274740282822107,
+      "loss": 1.3827,
+      "step": 8340
+    },
+    {
+      "epoch": 1.33456,
+      "grad_norm": 1.9368010517306073,
+      "learning_rate": 0.0001227305815007251,
+      "loss": 1.3652,
+      "step": 8341
+    },
+    {
+      "epoch": 1.33472,
+      "grad_norm": 2.243125218988278,
+      "learning_rate": 0.0001227137594949817,
+      "loss": 1.3185,
+      "step": 8342
+    },
+    {
+      "epoch": 1.33488,
+      "grad_norm": 1.9330231906383326,
+      "learning_rate": 0.00012269693681149285,
+      "loss": 1.3316,
+      "step": 8343
+    },
+    {
+      "epoch": 1.33504,
+      "grad_norm": 1.996319532401115,
+      "learning_rate": 0.00012268011345076043,
+      "loss": 1.4246,
+      "step": 8344
+    },
+    {
+      "epoch": 1.3352,
+      "grad_norm": 1.9539582074230941,
+      "learning_rate": 0.0001226632894132865,
+      "loss": 1.3542,
+      "step": 8345
+    },
+    {
+      "epoch": 1.33536,
+      "grad_norm": 2.374578799674046,
+      "learning_rate": 0.00012264646469957306,
+      "loss": 1.4221,
+      "step": 8346
+    },
+    {
+      "epoch": 1.33552,
+      "grad_norm": 2.3080613881219403,
+      "learning_rate": 0.0001226296393101221,
+      "loss": 1.4248,
+      "step": 8347
+    },
+    {
+      "epoch": 1.33568,
+      "grad_norm": 2.094527187159722,
+      "learning_rate": 0.00012261281324543565,
+      "loss": 1.4173,
+      "step": 8348
+    },
+    {
+      "epoch": 1.3358400000000001,
+      "grad_norm": 2.102257996379829,
+      "learning_rate": 0.00012259598650601582,
+      "loss": 1.4053,
+      "step": 8349
+    },
+    {
+      "epoch": 1.336,
+      "grad_norm": 1.965524103807404,
+      "learning_rate": 0.0001225791590923647,
+      "loss": 1.3649,
+      "step": 8350
+    },
+    {
+      "epoch": 1.33616,
+      "grad_norm": 2.6579717110393726,
+      "learning_rate": 0.00012256233100498435,
+      "loss": 1.386,
+      "step": 8351
+    },
+    {
+      "epoch": 1.33632,
+      "grad_norm": 2.048510578181698,
+      "learning_rate": 0.00012254550224437692,
+      "loss": 1.3602,
+      "step": 8352
+    },
+    {
+      "epoch": 1.33648,
+      "grad_norm": 2.4483203944753145,
+      "learning_rate": 0.00012252867281104454,
+      "loss": 1.4156,
+      "step": 8353
+    },
+    {
+      "epoch": 1.33664,
+      "grad_norm": 2.6383009437609175,
+      "learning_rate": 0.00012251184270548944,
+      "loss": 1.3485,
+      "step": 8354
+    },
+    {
+      "epoch": 1.3368,
+      "grad_norm": 2.329367605895423,
+      "learning_rate": 0.0001224950119282137,
+      "loss": 1.3761,
+      "step": 8355
+    },
+    {
+      "epoch": 1.33696,
+      "grad_norm": 1.9316517604595147,
+      "learning_rate": 0.0001224781804797196,
+      "loss": 1.3834,
+      "step": 8356
+    },
+    {
+      "epoch": 1.33712,
+      "grad_norm": 2.8207912358136027,
+      "learning_rate": 0.00012246134836050936,
+      "loss": 1.3933,
+      "step": 8357
+    },
+    {
+      "epoch": 1.33728,
+      "grad_norm": 2.0576887934977317,
+      "learning_rate": 0.00012244451557108524,
+      "loss": 1.3254,
+      "step": 8358
+    },
+    {
+      "epoch": 1.33744,
+      "grad_norm": 2.623013859078245,
+      "learning_rate": 0.00012242768211194946,
+      "loss": 1.4208,
+      "step": 8359
+    },
+    {
+      "epoch": 1.3376000000000001,
+      "grad_norm": 2.661194128010368,
+      "learning_rate": 0.0001224108479836043,
+      "loss": 1.3375,
+      "step": 8360
+    },
+    {
+      "epoch": 1.33776,
+      "grad_norm": 2.4408911058296887,
+      "learning_rate": 0.0001223940131865521,
+      "loss": 1.3467,
+      "step": 8361
+    },
+    {
+      "epoch": 1.33792,
+      "grad_norm": 2.502974401956265,
+      "learning_rate": 0.0001223771777212952,
+      "loss": 1.4122,
+      "step": 8362
+    },
+    {
+      "epoch": 1.33808,
+      "grad_norm": 2.3280845649104265,
+      "learning_rate": 0.00012236034158833592,
+      "loss": 1.3863,
+      "step": 8363
+    },
+    {
+      "epoch": 1.3382399999999999,
+      "grad_norm": 2.3443429417399466,
+      "learning_rate": 0.0001223435047881766,
+      "loss": 1.3509,
+      "step": 8364
+    },
+    {
+      "epoch": 1.3384,
+      "grad_norm": 2.3774098408026623,
+      "learning_rate": 0.0001223266673213197,
+      "loss": 1.3814,
+      "step": 8365
+    },
+    {
+      "epoch": 1.33856,
+      "grad_norm": 1.9697450363440507,
+      "learning_rate": 0.00012230982918826758,
+      "loss": 1.3725,
+      "step": 8366
+    },
+    {
+      "epoch": 1.33872,
+      "grad_norm": 2.0846686918186874,
+      "learning_rate": 0.00012229299038952264,
+      "loss": 1.4074,
+      "step": 8367
+    },
+    {
+      "epoch": 1.33888,
+      "grad_norm": 2.3433125243091473,
+      "learning_rate": 0.00012227615092558734,
+      "loss": 1.4276,
+      "step": 8368
+    },
+    {
+      "epoch": 1.33904,
+      "grad_norm": 2.0325254149275724,
+      "learning_rate": 0.00012225931079696417,
+      "loss": 1.3248,
+      "step": 8369
+    },
+    {
+      "epoch": 1.3392,
+      "grad_norm": 2.7535617809500166,
+      "learning_rate": 0.0001222424700041556,
+      "loss": 1.3396,
+      "step": 8370
+    },
+    {
+      "epoch": 1.33936,
+      "grad_norm": 2.0313188151333814,
+      "learning_rate": 0.00012222562854766414,
+      "loss": 1.3883,
+      "step": 8371
+    },
+    {
+      "epoch": 1.33952,
+      "grad_norm": 1.9478198850984538,
+      "learning_rate": 0.00012220878642799231,
+      "loss": 1.356,
+      "step": 8372
+    },
+    {
+      "epoch": 1.33968,
+      "grad_norm": 2.1604601673764106,
+      "learning_rate": 0.00012219194364564265,
+      "loss": 1.3072,
+      "step": 8373
+    },
+    {
+      "epoch": 1.33984,
+      "grad_norm": 2.0666229597728347,
+      "learning_rate": 0.00012217510020111775,
+      "loss": 1.464,
+      "step": 8374
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 1.9760788894171715,
+      "learning_rate": 0.00012215825609492015,
+      "loss": 1.3657,
+      "step": 8375
+    },
+    {
+      "epoch": 1.34016,
+      "grad_norm": 2.2822986066345265,
+      "learning_rate": 0.00012214141132755248,
+      "loss": 1.419,
+      "step": 8376
+    },
+    {
+      "epoch": 1.34032,
+      "grad_norm": 2.2954609143690616,
+      "learning_rate": 0.00012212456589951736,
+      "loss": 1.3298,
+      "step": 8377
+    },
+    {
+      "epoch": 1.34048,
+      "grad_norm": 2.058386245735638,
+      "learning_rate": 0.00012210771981131745,
+      "loss": 1.3908,
+      "step": 8378
+    },
+    {
+      "epoch": 1.34064,
+      "grad_norm": 2.392592633474648,
+      "learning_rate": 0.00012209087306345537,
+      "loss": 1.4316,
+      "step": 8379
+    },
+    {
+      "epoch": 1.3408,
+      "grad_norm": 2.559466964930928,
+      "learning_rate": 0.00012207402565643383,
+      "loss": 1.3525,
+      "step": 8380
+    },
+    {
+      "epoch": 1.34096,
+      "grad_norm": 2.201697140033207,
+      "learning_rate": 0.00012205717759075555,
+      "loss": 1.426,
+      "step": 8381
+    },
+    {
+      "epoch": 1.34112,
+      "grad_norm": 2.2523665455518236,
+      "learning_rate": 0.00012204032886692319,
+      "loss": 1.3934,
+      "step": 8382
+    },
+    {
+      "epoch": 1.34128,
+      "grad_norm": 2.3554636190031952,
+      "learning_rate": 0.00012202347948543955,
+      "loss": 1.4216,
+      "step": 8383
+    },
+    {
+      "epoch": 1.34144,
+      "grad_norm": 2.312635503049324,
+      "learning_rate": 0.00012200662944680739,
+      "loss": 1.3963,
+      "step": 8384
+    },
+    {
+      "epoch": 1.3416000000000001,
+      "grad_norm": 2.1418162979333544,
+      "learning_rate": 0.00012198977875152946,
+      "loss": 1.3016,
+      "step": 8385
+    },
+    {
+      "epoch": 1.34176,
+      "grad_norm": 2.293004176310508,
+      "learning_rate": 0.00012197292740010858,
+      "loss": 1.4381,
+      "step": 8386
+    },
+    {
+      "epoch": 1.34192,
+      "grad_norm": 2.454139682599989,
+      "learning_rate": 0.00012195607539304757,
+      "loss": 1.3315,
+      "step": 8387
+    },
+    {
+      "epoch": 1.34208,
+      "grad_norm": 2.461751839426208,
+      "learning_rate": 0.00012193922273084925,
+      "loss": 1.3786,
+      "step": 8388
+    },
+    {
+      "epoch": 1.3422399999999999,
+      "grad_norm": 1.9934716079828407,
+      "learning_rate": 0.00012192236941401651,
+      "loss": 1.3866,
+      "step": 8389
+    },
+    {
+      "epoch": 1.3424,
+      "grad_norm": 2.5156544576415474,
+      "learning_rate": 0.00012190551544305222,
+      "loss": 1.3947,
+      "step": 8390
+    },
+    {
+      "epoch": 1.34256,
+      "grad_norm": 2.2460495849341124,
+      "learning_rate": 0.00012188866081845925,
+      "loss": 1.3051,
+      "step": 8391
+    },
+    {
+      "epoch": 1.34272,
+      "grad_norm": 2.1927994682361436,
+      "learning_rate": 0.00012187180554074053,
+      "loss": 1.3782,
+      "step": 8392
+    },
+    {
+      "epoch": 1.34288,
+      "grad_norm": 2.3778987318159532,
+      "learning_rate": 0.00012185494961039901,
+      "loss": 1.4199,
+      "step": 8393
+    },
+    {
+      "epoch": 1.34304,
+      "grad_norm": 1.8755442864029275,
+      "learning_rate": 0.00012183809302793763,
+      "loss": 1.3399,
+      "step": 8394
+    },
+    {
+      "epoch": 1.3432,
+      "grad_norm": 2.2586299257639264,
+      "learning_rate": 0.0001218212357938594,
+      "loss": 1.392,
+      "step": 8395
+    },
+    {
+      "epoch": 1.34336,
+      "grad_norm": 1.9340397277412726,
+      "learning_rate": 0.00012180437790866728,
+      "loss": 1.3638,
+      "step": 8396
+    },
+    {
+      "epoch": 1.34352,
+      "grad_norm": 1.852452038868624,
+      "learning_rate": 0.0001217875193728643,
+      "loss": 1.4735,
+      "step": 8397
+    },
+    {
+      "epoch": 1.34368,
+      "grad_norm": 2.0209899617130556,
+      "learning_rate": 0.0001217706601869535,
+      "loss": 1.3274,
+      "step": 8398
+    },
+    {
+      "epoch": 1.34384,
+      "grad_norm": 1.9961327017947823,
+      "learning_rate": 0.00012175380035143789,
+      "loss": 1.3844,
+      "step": 8399
+    },
+    {
+      "epoch": 1.3439999999999999,
+      "grad_norm": 2.0231952713331527,
+      "learning_rate": 0.0001217369398668206,
+      "loss": 1.3661,
+      "step": 8400
+    },
+    {
+      "epoch": 1.34416,
+      "grad_norm": 1.969156148489672,
+      "learning_rate": 0.00012172007873360467,
+      "loss": 1.3721,
+      "step": 8401
+    },
+    {
+      "epoch": 1.34432,
+      "grad_norm": 2.0295763337549104,
+      "learning_rate": 0.00012170321695229326,
+      "loss": 1.4038,
+      "step": 8402
+    },
+    {
+      "epoch": 1.34448,
+      "grad_norm": 2.3163465412073068,
+      "learning_rate": 0.00012168635452338947,
+      "loss": 1.4318,
+      "step": 8403
+    },
+    {
+      "epoch": 1.34464,
+      "grad_norm": 2.1826929936776303,
+      "learning_rate": 0.0001216694914473965,
+      "loss": 1.4669,
+      "step": 8404
+    },
+    {
+      "epoch": 1.3448,
+      "grad_norm": 2.054490549547956,
+      "learning_rate": 0.00012165262772481745,
+      "loss": 1.3397,
+      "step": 8405
+    },
+    {
+      "epoch": 1.34496,
+      "grad_norm": 2.1825718340279283,
+      "learning_rate": 0.00012163576335615555,
+      "loss": 1.3733,
+      "step": 8406
+    },
+    {
+      "epoch": 1.34512,
+      "grad_norm": 2.097463903992322,
+      "learning_rate": 0.000121618898341914,
+      "loss": 1.4102,
+      "step": 8407
+    },
+    {
+      "epoch": 1.34528,
+      "grad_norm": 2.0824047561293324,
+      "learning_rate": 0.00012160203268259603,
+      "loss": 1.3121,
+      "step": 8408
+    },
+    {
+      "epoch": 1.34544,
+      "grad_norm": 2.520267802832061,
+      "learning_rate": 0.00012158516637870484,
+      "loss": 1.4712,
+      "step": 8409
+    },
+    {
+      "epoch": 1.3456000000000001,
+      "grad_norm": 1.9043494307619997,
+      "learning_rate": 0.00012156829943074378,
+      "loss": 1.4101,
+      "step": 8410
+    },
+    {
+      "epoch": 1.34576,
+      "grad_norm": 2.375545727769305,
+      "learning_rate": 0.00012155143183921607,
+      "loss": 1.3782,
+      "step": 8411
+    },
+    {
+      "epoch": 1.34592,
+      "grad_norm": 2.7408172761672907,
+      "learning_rate": 0.00012153456360462505,
+      "loss": 1.4587,
+      "step": 8412
+    },
+    {
+      "epoch": 1.34608,
+      "grad_norm": 2.073393217852532,
+      "learning_rate": 0.00012151769472747406,
+      "loss": 1.4487,
+      "step": 8413
+    },
+    {
+      "epoch": 1.3462399999999999,
+      "grad_norm": 2.299017652848782,
+      "learning_rate": 0.00012150082520826638,
+      "loss": 1.3768,
+      "step": 8414
+    },
+    {
+      "epoch": 1.3464,
+      "grad_norm": 2.6411750071825737,
+      "learning_rate": 0.00012148395504750541,
+      "loss": 1.3861,
+      "step": 8415
+    },
+    {
+      "epoch": 1.34656,
+      "grad_norm": 2.324166348549318,
+      "learning_rate": 0.00012146708424569453,
+      "loss": 1.3696,
+      "step": 8416
+    },
+    {
+      "epoch": 1.34672,
+      "grad_norm": 2.011852319791708,
+      "learning_rate": 0.00012145021280333711,
+      "loss": 1.3663,
+      "step": 8417
+    },
+    {
+      "epoch": 1.34688,
+      "grad_norm": 2.2818742001158734,
+      "learning_rate": 0.00012143334072093661,
+      "loss": 1.3795,
+      "step": 8418
+    },
+    {
+      "epoch": 1.34704,
+      "grad_norm": 2.1747417438567545,
+      "learning_rate": 0.00012141646799899648,
+      "loss": 1.4551,
+      "step": 8419
+    },
+    {
+      "epoch": 1.3472,
+      "grad_norm": 2.025755138337601,
+      "learning_rate": 0.00012139959463802013,
+      "loss": 1.3634,
+      "step": 8420
+    },
+    {
+      "epoch": 1.3473600000000001,
+      "grad_norm": 2.632851629318397,
+      "learning_rate": 0.00012138272063851109,
+      "loss": 1.3721,
+      "step": 8421
+    },
+    {
+      "epoch": 1.34752,
+      "grad_norm": 2.3788929243362467,
+      "learning_rate": 0.00012136584600097278,
+      "loss": 1.453,
+      "step": 8422
+    },
+    {
+      "epoch": 1.34768,
+      "grad_norm": 2.0154366900666596,
+      "learning_rate": 0.00012134897072590879,
+      "loss": 1.4289,
+      "step": 8423
+    },
+    {
+      "epoch": 1.34784,
+      "grad_norm": 2.4372452042288564,
+      "learning_rate": 0.00012133209481382262,
+      "loss": 1.3715,
+      "step": 8424
+    },
+    {
+      "epoch": 1.3479999999999999,
+      "grad_norm": 2.434020276055066,
+      "learning_rate": 0.00012131521826521779,
+      "loss": 1.3806,
+      "step": 8425
+    },
+    {
+      "epoch": 1.34816,
+      "grad_norm": 1.9465323055265384,
+      "learning_rate": 0.00012129834108059795,
+      "loss": 1.3519,
+      "step": 8426
+    },
+    {
+      "epoch": 1.34832,
+      "grad_norm": 2.272613081254378,
+      "learning_rate": 0.00012128146326046662,
+      "loss": 1.4111,
+      "step": 8427
+    },
+    {
+      "epoch": 1.34848,
+      "grad_norm": 2.0553836031038646,
+      "learning_rate": 0.00012126458480532745,
+      "loss": 1.4777,
+      "step": 8428
+    },
+    {
+      "epoch": 1.34864,
+      "grad_norm": 1.9026459250225933,
+      "learning_rate": 0.00012124770571568405,
+      "loss": 1.4416,
+      "step": 8429
+    },
+    {
+      "epoch": 1.3488,
+      "grad_norm": 1.9987600352358186,
+      "learning_rate": 0.00012123082599204008,
+      "loss": 1.3668,
+      "step": 8430
+    },
+    {
+      "epoch": 1.34896,
+      "grad_norm": 2.2778757191318917,
+      "learning_rate": 0.00012121394563489918,
+      "loss": 1.4295,
+      "step": 8431
+    },
+    {
+      "epoch": 1.34912,
+      "grad_norm": 2.1633086851110717,
+      "learning_rate": 0.00012119706464476508,
+      "loss": 1.4208,
+      "step": 8432
+    },
+    {
+      "epoch": 1.34928,
+      "grad_norm": 2.208985724325426,
+      "learning_rate": 0.00012118018302214143,
+      "loss": 1.3387,
+      "step": 8433
+    },
+    {
+      "epoch": 1.34944,
+      "grad_norm": 1.8846033416117522,
+      "learning_rate": 0.00012116330076753199,
+      "loss": 1.3116,
+      "step": 8434
+    },
+    {
+      "epoch": 1.3496000000000001,
+      "grad_norm": 2.261090822686479,
+      "learning_rate": 0.0001211464178814405,
+      "loss": 1.396,
+      "step": 8435
+    },
+    {
+      "epoch": 1.34976,
+      "grad_norm": 2.4839646887558726,
+      "learning_rate": 0.00012112953436437073,
+      "loss": 1.4455,
+      "step": 8436
+    },
+    {
+      "epoch": 1.34992,
+      "grad_norm": 2.168996275489473,
+      "learning_rate": 0.00012111265021682643,
+      "loss": 1.3642,
+      "step": 8437
+    },
+    {
+      "epoch": 1.35008,
+      "grad_norm": 2.16487124959083,
+      "learning_rate": 0.0001210957654393114,
+      "loss": 1.4238,
+      "step": 8438
+    },
+    {
+      "epoch": 1.3502399999999999,
+      "grad_norm": 2.0787887839249386,
+      "learning_rate": 0.0001210788800323295,
+      "loss": 1.3765,
+      "step": 8439
+    },
+    {
+      "epoch": 1.3504,
+      "grad_norm": 2.2625171317404766,
+      "learning_rate": 0.00012106199399638451,
+      "loss": 1.3559,
+      "step": 8440
+    },
+    {
+      "epoch": 1.35056,
+      "grad_norm": 2.2617478779795244,
+      "learning_rate": 0.00012104510733198029,
+      "loss": 1.4241,
+      "step": 8441
+    },
+    {
+      "epoch": 1.35072,
+      "grad_norm": 2.0193598221076154,
+      "learning_rate": 0.00012102822003962075,
+      "loss": 1.3782,
+      "step": 8442
+    },
+    {
+      "epoch": 1.35088,
+      "grad_norm": 2.233795571695564,
+      "learning_rate": 0.00012101133211980977,
+      "loss": 1.4507,
+      "step": 8443
+    },
+    {
+      "epoch": 1.35104,
+      "grad_norm": 1.9991334357336528,
+      "learning_rate": 0.00012099444357305126,
+      "loss": 1.3778,
+      "step": 8444
+    },
+    {
+      "epoch": 1.3512,
+      "grad_norm": 2.031544176277913,
+      "learning_rate": 0.00012097755439984915,
+      "loss": 1.3367,
+      "step": 8445
+    },
+    {
+      "epoch": 1.3513600000000001,
+      "grad_norm": 2.702058766597125,
+      "learning_rate": 0.00012096066460070736,
+      "loss": 1.4274,
+      "step": 8446
+    },
+    {
+      "epoch": 1.35152,
+      "grad_norm": 2.047505211856062,
+      "learning_rate": 0.00012094377417612992,
+      "loss": 1.335,
+      "step": 8447
+    },
+    {
+      "epoch": 1.35168,
+      "grad_norm": 1.731754722468363,
+      "learning_rate": 0.00012092688312662076,
+      "loss": 1.3314,
+      "step": 8448
+    },
+    {
+      "epoch": 1.35184,
+      "grad_norm": 2.2535984690278976,
+      "learning_rate": 0.00012090999145268389,
+      "loss": 1.3788,
+      "step": 8449
+    },
+    {
+      "epoch": 1.3519999999999999,
+      "grad_norm": 2.0542456605646615,
+      "learning_rate": 0.00012089309915482336,
+      "loss": 1.3379,
+      "step": 8450
+    },
+    {
+      "epoch": 1.35216,
+      "grad_norm": 1.9432941375797692,
+      "learning_rate": 0.00012087620623354317,
+      "loss": 1.3028,
+      "step": 8451
+    },
+    {
+      "epoch": 1.35232,
+      "grad_norm": 2.4119011319237673,
+      "learning_rate": 0.00012085931268934743,
+      "loss": 1.3964,
+      "step": 8452
+    },
+    {
+      "epoch": 1.35248,
+      "grad_norm": 2.0569300091561393,
+      "learning_rate": 0.00012084241852274019,
+      "loss": 1.3601,
+      "step": 8453
+    },
+    {
+      "epoch": 1.35264,
+      "grad_norm": 2.344110780051967,
+      "learning_rate": 0.00012082552373422555,
+      "loss": 1.3141,
+      "step": 8454
+    },
+    {
+      "epoch": 1.3528,
+      "grad_norm": 2.109376490198377,
+      "learning_rate": 0.00012080862832430764,
+      "loss": 1.3927,
+      "step": 8455
+    },
+    {
+      "epoch": 1.35296,
+      "grad_norm": 2.0876745374600585,
+      "learning_rate": 0.00012079173229349057,
+      "loss": 1.4087,
+      "step": 8456
+    },
+    {
+      "epoch": 1.35312,
+      "grad_norm": 2.5807266777116813,
+      "learning_rate": 0.0001207748356422785,
+      "loss": 1.3684,
+      "step": 8457
+    },
+    {
+      "epoch": 1.35328,
+      "grad_norm": 2.1153827973218045,
+      "learning_rate": 0.00012075793837117561,
+      "loss": 1.3878,
+      "step": 8458
+    },
+    {
+      "epoch": 1.35344,
+      "grad_norm": 2.8036448572910033,
+      "learning_rate": 0.00012074104048068607,
+      "loss": 1.4504,
+      "step": 8459
+    },
+    {
+      "epoch": 1.3536000000000001,
+      "grad_norm": 2.108495379406666,
+      "learning_rate": 0.00012072414197131412,
+      "loss": 1.4039,
+      "step": 8460
+    },
+    {
+      "epoch": 1.35376,
+      "grad_norm": 2.324731207054906,
+      "learning_rate": 0.00012070724284356398,
+      "loss": 1.4153,
+      "step": 8461
+    },
+    {
+      "epoch": 1.35392,
+      "grad_norm": 2.261750099721962,
+      "learning_rate": 0.0001206903430979399,
+      "loss": 1.4642,
+      "step": 8462
+    },
+    {
+      "epoch": 1.35408,
+      "grad_norm": 2.275580484784235,
+      "learning_rate": 0.00012067344273494609,
+      "loss": 1.4191,
+      "step": 8463
+    },
+    {
+      "epoch": 1.3542399999999999,
+      "grad_norm": 2.222605998958377,
+      "learning_rate": 0.00012065654175508692,
+      "loss": 1.3665,
+      "step": 8464
+    },
+    {
+      "epoch": 1.3544,
+      "grad_norm": 2.1546364097576642,
+      "learning_rate": 0.00012063964015886661,
+      "loss": 1.4222,
+      "step": 8465
+    },
+    {
+      "epoch": 1.35456,
+      "grad_norm": 2.0711911688383546,
+      "learning_rate": 0.00012062273794678952,
+      "loss": 1.3693,
+      "step": 8466
+    },
+    {
+      "epoch": 1.35472,
+      "grad_norm": 2.2898997446309766,
+      "learning_rate": 0.00012060583511935997,
+      "loss": 1.3179,
+      "step": 8467
+    },
+    {
+      "epoch": 1.35488,
+      "grad_norm": 1.8861768507940462,
+      "learning_rate": 0.00012058893167708234,
+      "loss": 1.321,
+      "step": 8468
+    },
+    {
+      "epoch": 1.35504,
+      "grad_norm": 1.5825017165517508,
+      "learning_rate": 0.00012057202762046101,
+      "loss": 1.3516,
+      "step": 8469
+    },
+    {
+      "epoch": 1.3552,
+      "grad_norm": 1.9942098788633003,
+      "learning_rate": 0.00012055512295000033,
+      "loss": 1.316,
+      "step": 8470
+    },
+    {
+      "epoch": 1.3553600000000001,
+      "grad_norm": 2.258042200594799,
+      "learning_rate": 0.00012053821766620473,
+      "loss": 1.4439,
+      "step": 8471
+    },
+    {
+      "epoch": 1.35552,
+      "grad_norm": 1.9630188591073172,
+      "learning_rate": 0.00012052131176957867,
+      "loss": 1.3596,
+      "step": 8472
+    },
+    {
+      "epoch": 1.35568,
+      "grad_norm": 2.044047538217333,
+      "learning_rate": 0.00012050440526062651,
+      "loss": 1.3871,
+      "step": 8473
+    },
+    {
+      "epoch": 1.35584,
+      "grad_norm": 2.0223604618562345,
+      "learning_rate": 0.0001204874981398528,
+      "loss": 1.3517,
+      "step": 8474
+    },
+    {
+      "epoch": 1.3559999999999999,
+      "grad_norm": 2.0575835343949715,
+      "learning_rate": 0.00012047059040776203,
+      "loss": 1.3224,
+      "step": 8475
+    },
+    {
+      "epoch": 1.35616,
+      "grad_norm": 2.683565061890368,
+      "learning_rate": 0.00012045368206485865,
+      "loss": 1.4077,
+      "step": 8476
+    },
+    {
+      "epoch": 1.35632,
+      "grad_norm": 2.285218844603369,
+      "learning_rate": 0.00012043677311164721,
+      "loss": 1.4435,
+      "step": 8477
+    },
+    {
+      "epoch": 1.35648,
+      "grad_norm": 2.294668742864497,
+      "learning_rate": 0.00012041986354863222,
+      "loss": 1.4375,
+      "step": 8478
+    },
+    {
+      "epoch": 1.35664,
+      "grad_norm": 2.1615875231355113,
+      "learning_rate": 0.00012040295337631828,
+      "loss": 1.4394,
+      "step": 8479
+    },
+    {
+      "epoch": 1.3568,
+      "grad_norm": 2.4301756606993936,
+      "learning_rate": 0.00012038604259520993,
+      "loss": 1.2958,
+      "step": 8480
+    },
+    {
+      "epoch": 1.35696,
+      "grad_norm": 2.0176737835566882,
+      "learning_rate": 0.00012036913120581177,
+      "loss": 1.4072,
+      "step": 8481
+    },
+    {
+      "epoch": 1.35712,
+      "grad_norm": 2.272021620649831,
+      "learning_rate": 0.0001203522192086284,
+      "loss": 1.3813,
+      "step": 8482
+    },
+    {
+      "epoch": 1.35728,
+      "grad_norm": 2.818229226822726,
+      "learning_rate": 0.0001203353066041645,
+      "loss": 1.3849,
+      "step": 8483
+    },
+    {
+      "epoch": 1.35744,
+      "grad_norm": 2.210790417368951,
+      "learning_rate": 0.00012031839339292468,
+      "loss": 1.4409,
+      "step": 8484
+    },
+    {
+      "epoch": 1.3576,
+      "grad_norm": 2.0240645308799774,
+      "learning_rate": 0.0001203014795754136,
+      "loss": 1.3657,
+      "step": 8485
+    },
+    {
+      "epoch": 1.35776,
+      "grad_norm": 2.367907017468043,
+      "learning_rate": 0.00012028456515213593,
+      "loss": 1.3835,
+      "step": 8486
+    },
+    {
+      "epoch": 1.35792,
+      "grad_norm": 2.565906429491674,
+      "learning_rate": 0.00012026765012359642,
+      "loss": 1.3286,
+      "step": 8487
+    },
+    {
+      "epoch": 1.35808,
+      "grad_norm": 2.3917574989930888,
+      "learning_rate": 0.00012025073449029975,
+      "loss": 1.3592,
+      "step": 8488
+    },
+    {
+      "epoch": 1.35824,
+      "grad_norm": 2.822633392437241,
+      "learning_rate": 0.00012023381825275068,
+      "loss": 1.4696,
+      "step": 8489
+    },
+    {
+      "epoch": 1.3584,
+      "grad_norm": 2.3365075825778128,
+      "learning_rate": 0.00012021690141145395,
+      "loss": 1.2865,
+      "step": 8490
+    },
+    {
+      "epoch": 1.35856,
+      "grad_norm": 2.6727617972033877,
+      "learning_rate": 0.00012019998396691434,
+      "loss": 1.4245,
+      "step": 8491
+    },
+    {
+      "epoch": 1.35872,
+      "grad_norm": 2.473349800021858,
+      "learning_rate": 0.00012018306591963666,
+      "loss": 1.412,
+      "step": 8492
+    },
+    {
+      "epoch": 1.35888,
+      "grad_norm": 2.280410643596212,
+      "learning_rate": 0.0001201661472701257,
+      "loss": 1.3985,
+      "step": 8493
+    },
+    {
+      "epoch": 1.35904,
+      "grad_norm": 2.9855836130423503,
+      "learning_rate": 0.00012014922801888629,
+      "loss": 1.4228,
+      "step": 8494
+    },
+    {
+      "epoch": 1.3592,
+      "grad_norm": 2.709022496586248,
+      "learning_rate": 0.00012013230816642324,
+      "loss": 1.4421,
+      "step": 8495
+    },
+    {
+      "epoch": 1.3593600000000001,
+      "grad_norm": 2.456163882962198,
+      "learning_rate": 0.00012011538771324149,
+      "loss": 1.3956,
+      "step": 8496
+    },
+    {
+      "epoch": 1.35952,
+      "grad_norm": 2.252277140790424,
+      "learning_rate": 0.00012009846665984587,
+      "loss": 1.3987,
+      "step": 8497
+    },
+    {
+      "epoch": 1.35968,
+      "grad_norm": 2.147203626424226,
+      "learning_rate": 0.00012008154500674127,
+      "loss": 1.4358,
+      "step": 8498
+    },
+    {
+      "epoch": 1.35984,
+      "grad_norm": 2.1591432607077388,
+      "learning_rate": 0.00012006462275443267,
+      "loss": 1.4057,
+      "step": 8499
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 2.0234336787331224,
+      "learning_rate": 0.00012004769990342493,
+      "loss": 1.3058,
+      "step": 8500
+    },
+    {
+      "epoch": 1.36016,
+      "grad_norm": 2.1253715041698116,
+      "learning_rate": 0.00012003077645422305,
+      "loss": 1.4033,
+      "step": 8501
+    },
+    {
+      "epoch": 1.36032,
+      "grad_norm": 2.2466048236538803,
+      "learning_rate": 0.00012001385240733198,
+      "loss": 1.3792,
+      "step": 8502
+    },
+    {
+      "epoch": 1.36048,
+      "grad_norm": 2.0909424055561483,
+      "learning_rate": 0.00011999692776325673,
+      "loss": 1.4186,
+      "step": 8503
+    },
+    {
+      "epoch": 1.36064,
+      "grad_norm": 2.0024457192758214,
+      "learning_rate": 0.0001199800025225023,
+      "loss": 1.4451,
+      "step": 8504
+    },
+    {
+      "epoch": 1.3608,
+      "grad_norm": 2.1086467238078823,
+      "learning_rate": 0.00011996307668557369,
+      "loss": 1.3955,
+      "step": 8505
+    },
+    {
+      "epoch": 1.36096,
+      "grad_norm": 2.1589369067070567,
+      "learning_rate": 0.000119946150252976,
+      "loss": 1.353,
+      "step": 8506
+    },
+    {
+      "epoch": 1.36112,
+      "grad_norm": 2.496680674591983,
+      "learning_rate": 0.0001199292232252142,
+      "loss": 1.4173,
+      "step": 8507
+    },
+    {
+      "epoch": 1.36128,
+      "grad_norm": 2.126171439305114,
+      "learning_rate": 0.00011991229560279342,
+      "loss": 1.3937,
+      "step": 8508
+    },
+    {
+      "epoch": 1.36144,
+      "grad_norm": 2.1881878490769884,
+      "learning_rate": 0.00011989536738621878,
+      "loss": 1.3485,
+      "step": 8509
+    },
+    {
+      "epoch": 1.3616,
+      "grad_norm": 2.2812165487911806,
+      "learning_rate": 0.00011987843857599539,
+      "loss": 1.3485,
+      "step": 8510
+    },
+    {
+      "epoch": 1.36176,
+      "grad_norm": 2.209019071824429,
+      "learning_rate": 0.00011986150917262833,
+      "loss": 1.3196,
+      "step": 8511
+    },
+    {
+      "epoch": 1.36192,
+      "grad_norm": 2.156378543910446,
+      "learning_rate": 0.00011984457917662277,
+      "loss": 1.3685,
+      "step": 8512
+    },
+    {
+      "epoch": 1.36208,
+      "grad_norm": 2.2516999190480043,
+      "learning_rate": 0.0001198276485884839,
+      "loss": 1.4259,
+      "step": 8513
+    },
+    {
+      "epoch": 1.36224,
+      "grad_norm": 2.337458064955549,
+      "learning_rate": 0.0001198107174087169,
+      "loss": 1.4077,
+      "step": 8514
+    },
+    {
+      "epoch": 1.3624,
+      "grad_norm": 2.3058861554161063,
+      "learning_rate": 0.00011979378563782693,
+      "loss": 1.4009,
+      "step": 8515
+    },
+    {
+      "epoch": 1.36256,
+      "grad_norm": 2.0660609213465375,
+      "learning_rate": 0.00011977685327631924,
+      "loss": 1.3492,
+      "step": 8516
+    },
+    {
+      "epoch": 1.36272,
+      "grad_norm": 2.321741375509173,
+      "learning_rate": 0.00011975992032469908,
+      "loss": 1.381,
+      "step": 8517
+    },
+    {
+      "epoch": 1.36288,
+      "grad_norm": 2.3583616863533092,
+      "learning_rate": 0.00011974298678347169,
+      "loss": 1.3684,
+      "step": 8518
+    },
+    {
+      "epoch": 1.36304,
+      "grad_norm": 2.4565522276050458,
+      "learning_rate": 0.00011972605265314234,
+      "loss": 1.4102,
+      "step": 8519
+    },
+    {
+      "epoch": 1.3632,
+      "grad_norm": 2.0680083620546394,
+      "learning_rate": 0.00011970911793421634,
+      "loss": 1.4189,
+      "step": 8520
+    },
+    {
+      "epoch": 1.3633600000000001,
+      "grad_norm": 2.7223521272201285,
+      "learning_rate": 0.00011969218262719897,
+      "loss": 1.3215,
+      "step": 8521
+    },
+    {
+      "epoch": 1.36352,
+      "grad_norm": 1.9865214759363086,
+      "learning_rate": 0.00011967524673259553,
+      "loss": 1.3729,
+      "step": 8522
+    },
+    {
+      "epoch": 1.36368,
+      "grad_norm": 2.3142383694355155,
+      "learning_rate": 0.00011965831025091145,
+      "loss": 1.3862,
+      "step": 8523
+    },
+    {
+      "epoch": 1.36384,
+      "grad_norm": 2.084074996068471,
+      "learning_rate": 0.00011964137318265198,
+      "loss": 1.3411,
+      "step": 8524
+    },
+    {
+      "epoch": 1.3639999999999999,
+      "grad_norm": 2.26912310288751,
+      "learning_rate": 0.00011962443552832261,
+      "loss": 1.3699,
+      "step": 8525
+    },
+    {
+      "epoch": 1.36416,
+      "grad_norm": 2.159962343642047,
+      "learning_rate": 0.00011960749728842867,
+      "loss": 1.4123,
+      "step": 8526
+    },
+    {
+      "epoch": 1.36432,
+      "grad_norm": 1.9967873742889688,
+      "learning_rate": 0.00011959055846347555,
+      "loss": 1.4246,
+      "step": 8527
+    },
+    {
+      "epoch": 1.36448,
+      "grad_norm": 2.174296823318793,
+      "learning_rate": 0.00011957361905396873,
+      "loss": 1.4477,
+      "step": 8528
+    },
+    {
+      "epoch": 1.36464,
+      "grad_norm": 2.7065049752913035,
+      "learning_rate": 0.00011955667906041366,
+      "loss": 1.4047,
+      "step": 8529
+    },
+    {
+      "epoch": 1.3648,
+      "grad_norm": 2.032520227430299,
+      "learning_rate": 0.00011953973848331577,
+      "loss": 1.3726,
+      "step": 8530
+    },
+    {
+      "epoch": 1.36496,
+      "grad_norm": 2.5689606060868564,
+      "learning_rate": 0.00011952279732318054,
+      "loss": 1.3931,
+      "step": 8531
+    },
+    {
+      "epoch": 1.3651200000000001,
+      "grad_norm": 1.8967168743625389,
+      "learning_rate": 0.00011950585558051348,
+      "loss": 1.3634,
+      "step": 8532
+    },
+    {
+      "epoch": 1.36528,
+      "grad_norm": 2.235225740324072,
+      "learning_rate": 0.00011948891325582012,
+      "loss": 1.3601,
+      "step": 8533
+    },
+    {
+      "epoch": 1.36544,
+      "grad_norm": 2.168510200498086,
+      "learning_rate": 0.00011947197034960603,
+      "loss": 1.4303,
+      "step": 8534
+    },
+    {
+      "epoch": 1.3656,
+      "grad_norm": 2.4643875786034033,
+      "learning_rate": 0.00011945502686237666,
+      "loss": 1.3518,
+      "step": 8535
+    },
+    {
+      "epoch": 1.3657599999999999,
+      "grad_norm": 2.4913037332641617,
+      "learning_rate": 0.00011943808279463766,
+      "loss": 1.3935,
+      "step": 8536
+    },
+    {
+      "epoch": 1.36592,
+      "grad_norm": 2.7729331406994397,
+      "learning_rate": 0.00011942113814689462,
+      "loss": 1.4857,
+      "step": 8537
+    },
+    {
+      "epoch": 1.36608,
+      "grad_norm": 1.9320265996474897,
+      "learning_rate": 0.00011940419291965308,
+      "loss": 1.4074,
+      "step": 8538
+    },
+    {
+      "epoch": 1.36624,
+      "grad_norm": 2.0636428230925303,
+      "learning_rate": 0.0001193872471134187,
+      "loss": 1.3945,
+      "step": 8539
+    },
+    {
+      "epoch": 1.3664,
+      "grad_norm": 2.2178242017435346,
+      "learning_rate": 0.00011937030072869713,
+      "loss": 1.3636,
+      "step": 8540
+    },
+    {
+      "epoch": 1.36656,
+      "grad_norm": 2.190797456453938,
+      "learning_rate": 0.00011935335376599402,
+      "loss": 1.3484,
+      "step": 8541
+    },
+    {
+      "epoch": 1.36672,
+      "grad_norm": 2.0892468097564456,
+      "learning_rate": 0.00011933640622581501,
+      "loss": 1.3452,
+      "step": 8542
+    },
+    {
+      "epoch": 1.36688,
+      "grad_norm": 2.119490585156096,
+      "learning_rate": 0.00011931945810866584,
+      "loss": 1.3861,
+      "step": 8543
+    },
+    {
+      "epoch": 1.36704,
+      "grad_norm": 2.233574885376537,
+      "learning_rate": 0.00011930250941505219,
+      "loss": 1.3199,
+      "step": 8544
+    },
+    {
+      "epoch": 1.3672,
+      "grad_norm": 2.1026776788860566,
+      "learning_rate": 0.00011928556014547978,
+      "loss": 1.3796,
+      "step": 8545
+    },
+    {
+      "epoch": 1.3673600000000001,
+      "grad_norm": 2.1859570211923707,
+      "learning_rate": 0.00011926861030045434,
+      "loss": 1.4965,
+      "step": 8546
+    },
+    {
+      "epoch": 1.36752,
+      "grad_norm": 1.7664914593888636,
+      "learning_rate": 0.00011925165988048166,
+      "loss": 1.3763,
+      "step": 8547
+    },
+    {
+      "epoch": 1.36768,
+      "grad_norm": 2.373914569896829,
+      "learning_rate": 0.00011923470888606749,
+      "loss": 1.363,
+      "step": 8548
+    },
+    {
+      "epoch": 1.36784,
+      "grad_norm": 2.215240730528736,
+      "learning_rate": 0.00011921775731771765,
+      "loss": 1.3291,
+      "step": 8549
+    },
+    {
+      "epoch": 1.3679999999999999,
+      "grad_norm": 2.315364429040891,
+      "learning_rate": 0.00011920080517593792,
+      "loss": 1.433,
+      "step": 8550
+    },
+    {
+      "epoch": 1.36816,
+      "grad_norm": 2.2648128687936606,
+      "learning_rate": 0.00011918385246123416,
+      "loss": 1.3875,
+      "step": 8551
+    },
+    {
+      "epoch": 1.36832,
+      "grad_norm": 2.2130447425012694,
+      "learning_rate": 0.00011916689917411219,
+      "loss": 1.3686,
+      "step": 8552
+    },
+    {
+      "epoch": 1.36848,
+      "grad_norm": 2.261417211683744,
+      "learning_rate": 0.00011914994531507788,
+      "loss": 1.4102,
+      "step": 8553
+    },
+    {
+      "epoch": 1.36864,
+      "grad_norm": 2.148135093625399,
+      "learning_rate": 0.0001191329908846371,
+      "loss": 1.4072,
+      "step": 8554
+    },
+    {
+      "epoch": 1.3688,
+      "grad_norm": 2.104133503104205,
+      "learning_rate": 0.00011911603588329577,
+      "loss": 1.3932,
+      "step": 8555
+    },
+    {
+      "epoch": 1.36896,
+      "grad_norm": 3.058018038716318,
+      "learning_rate": 0.00011909908031155974,
+      "loss": 1.3825,
+      "step": 8556
+    },
+    {
+      "epoch": 1.3691200000000001,
+      "grad_norm": 2.0571062712227115,
+      "learning_rate": 0.000119082124169935,
+      "loss": 1.4129,
+      "step": 8557
+    },
+    {
+      "epoch": 1.36928,
+      "grad_norm": 3.028951330788571,
+      "learning_rate": 0.00011906516745892749,
+      "loss": 1.3982,
+      "step": 8558
+    },
+    {
+      "epoch": 1.36944,
+      "grad_norm": 2.3845842883097554,
+      "learning_rate": 0.00011904821017904317,
+      "loss": 1.332,
+      "step": 8559
+    },
+    {
+      "epoch": 1.3696,
+      "grad_norm": 2.490639933140835,
+      "learning_rate": 0.00011903125233078802,
+      "loss": 1.3941,
+      "step": 8560
+    },
+    {
+      "epoch": 1.3697599999999999,
+      "grad_norm": 2.4094387242272925,
+      "learning_rate": 0.000119014293914668,
+      "loss": 1.3238,
+      "step": 8561
+    },
+    {
+      "epoch": 1.36992,
+      "grad_norm": 2.0416739710767406,
+      "learning_rate": 0.00011899733493118921,
+      "loss": 1.3883,
+      "step": 8562
+    },
+    {
+      "epoch": 1.37008,
+      "grad_norm": 2.392755734385999,
+      "learning_rate": 0.00011898037538085757,
+      "loss": 1.3807,
+      "step": 8563
+    },
+    {
+      "epoch": 1.37024,
+      "grad_norm": 2.2113140142813026,
+      "learning_rate": 0.00011896341526417922,
+      "loss": 1.3621,
+      "step": 8564
+    },
+    {
+      "epoch": 1.3704,
+      "grad_norm": 2.486607994399695,
+      "learning_rate": 0.00011894645458166018,
+      "loss": 1.3126,
+      "step": 8565
+    },
+    {
+      "epoch": 1.37056,
+      "grad_norm": 1.9666221749872645,
+      "learning_rate": 0.00011892949333380654,
+      "loss": 1.304,
+      "step": 8566
+    },
+    {
+      "epoch": 1.37072,
+      "grad_norm": 1.973572665873962,
+      "learning_rate": 0.00011891253152112443,
+      "loss": 1.3962,
+      "step": 8567
+    },
+    {
+      "epoch": 1.37088,
+      "grad_norm": 1.9949163331352018,
+      "learning_rate": 0.0001188955691441199,
+      "loss": 1.4574,
+      "step": 8568
+    },
+    {
+      "epoch": 1.37104,
+      "grad_norm": 2.068040993314823,
+      "learning_rate": 0.00011887860620329916,
+      "loss": 1.3248,
+      "step": 8569
+    },
+    {
+      "epoch": 1.3712,
+      "grad_norm": 2.00525745890479,
+      "learning_rate": 0.00011886164269916828,
+      "loss": 1.4198,
+      "step": 8570
+    },
+    {
+      "epoch": 1.3713600000000001,
+      "grad_norm": 2.187593696614687,
+      "learning_rate": 0.0001188446786322335,
+      "loss": 1.3963,
+      "step": 8571
+    },
+    {
+      "epoch": 1.37152,
+      "grad_norm": 2.407046414817546,
+      "learning_rate": 0.00011882771400300095,
+      "loss": 1.322,
+      "step": 8572
+    },
+    {
+      "epoch": 1.37168,
+      "grad_norm": 2.3520305367243073,
+      "learning_rate": 0.00011881074881197683,
+      "loss": 1.4193,
+      "step": 8573
+    },
+    {
+      "epoch": 1.37184,
+      "grad_norm": 1.9523490600853728,
+      "learning_rate": 0.0001187937830596674,
+      "loss": 1.4625,
+      "step": 8574
+    },
+    {
+      "epoch": 1.3719999999999999,
+      "grad_norm": 2.464212426253066,
+      "learning_rate": 0.00011877681674657887,
+      "loss": 1.3866,
+      "step": 8575
+    },
+    {
+      "epoch": 1.37216,
+      "grad_norm": 2.351162516935294,
+      "learning_rate": 0.00011875984987321749,
+      "loss": 1.4602,
+      "step": 8576
+    },
+    {
+      "epoch": 1.37232,
+      "grad_norm": 1.9571356541566076,
+      "learning_rate": 0.00011874288244008953,
+      "loss": 1.3753,
+      "step": 8577
+    },
+    {
+      "epoch": 1.37248,
+      "grad_norm": 2.0679357788936663,
+      "learning_rate": 0.00011872591444770128,
+      "loss": 1.3944,
+      "step": 8578
+    },
+    {
+      "epoch": 1.37264,
+      "grad_norm": 2.1208571642005243,
+      "learning_rate": 0.00011870894589655898,
+      "loss": 1.3762,
+      "step": 8579
+    },
+    {
+      "epoch": 1.3728,
+      "grad_norm": 2.0756654114494433,
+      "learning_rate": 0.00011869197678716903,
+      "loss": 1.3674,
+      "step": 8580
+    },
+    {
+      "epoch": 1.37296,
+      "grad_norm": 2.352312981672795,
+      "learning_rate": 0.00011867500712003771,
+      "loss": 1.3896,
+      "step": 8581
+    },
+    {
+      "epoch": 1.3731200000000001,
+      "grad_norm": 2.141144401743876,
+      "learning_rate": 0.00011865803689567142,
+      "loss": 1.3458,
+      "step": 8582
+    },
+    {
+      "epoch": 1.37328,
+      "grad_norm": 2.279267755429567,
+      "learning_rate": 0.00011864106611457648,
+      "loss": 1.3398,
+      "step": 8583
+    },
+    {
+      "epoch": 1.37344,
+      "grad_norm": 2.0875131554428514,
+      "learning_rate": 0.00011862409477725929,
+      "loss": 1.3283,
+      "step": 8584
+    },
+    {
+      "epoch": 1.3736,
+      "grad_norm": 2.577053473151467,
+      "learning_rate": 0.00011860712288422626,
+      "loss": 1.3901,
+      "step": 8585
+    },
+    {
+      "epoch": 1.3737599999999999,
+      "grad_norm": 2.4178727380465674,
+      "learning_rate": 0.00011859015043598378,
+      "loss": 1.3707,
+      "step": 8586
+    },
+    {
+      "epoch": 1.37392,
+      "grad_norm": 2.942113646087527,
+      "learning_rate": 0.00011857317743303833,
+      "loss": 1.4177,
+      "step": 8587
+    },
+    {
+      "epoch": 1.37408,
+      "grad_norm": 2.4908004608943797,
+      "learning_rate": 0.0001185562038758963,
+      "loss": 1.3497,
+      "step": 8588
+    },
+    {
+      "epoch": 1.37424,
+      "grad_norm": 2.0637822851356806,
+      "learning_rate": 0.00011853922976506418,
+      "loss": 1.3649,
+      "step": 8589
+    },
+    {
+      "epoch": 1.3744,
+      "grad_norm": 2.1623147035476955,
+      "learning_rate": 0.00011852225510104847,
+      "loss": 1.395,
+      "step": 8590
+    },
+    {
+      "epoch": 1.37456,
+      "grad_norm": 2.485164961087627,
+      "learning_rate": 0.00011850527988435567,
+      "loss": 1.408,
+      "step": 8591
+    },
+    {
+      "epoch": 1.37472,
+      "grad_norm": 2.3102812409522313,
+      "learning_rate": 0.00011848830411549228,
+      "loss": 1.2898,
+      "step": 8592
+    },
+    {
+      "epoch": 1.37488,
+      "grad_norm": 2.488905018485394,
+      "learning_rate": 0.00011847132779496482,
+      "loss": 1.4209,
+      "step": 8593
+    },
+    {
+      "epoch": 1.37504,
+      "grad_norm": 2.148353154398898,
+      "learning_rate": 0.00011845435092327988,
+      "loss": 1.3671,
+      "step": 8594
+    },
+    {
+      "epoch": 1.3752,
+      "grad_norm": 2.089450304517505,
+      "learning_rate": 0.00011843737350094398,
+      "loss": 1.4044,
+      "step": 8595
+    },
+    {
+      "epoch": 1.3753600000000001,
+      "grad_norm": 2.941271266900995,
+      "learning_rate": 0.00011842039552846371,
+      "loss": 1.3782,
+      "step": 8596
+    },
+    {
+      "epoch": 1.37552,
+      "grad_norm": 2.0806834369093714,
+      "learning_rate": 0.00011840341700634567,
+      "loss": 1.3734,
+      "step": 8597
+    },
+    {
+      "epoch": 1.37568,
+      "grad_norm": 2.219733539772564,
+      "learning_rate": 0.00011838643793509651,
+      "loss": 1.4028,
+      "step": 8598
+    },
+    {
+      "epoch": 1.37584,
+      "grad_norm": 2.0141117761826988,
+      "learning_rate": 0.00011836945831522284,
+      "loss": 1.3393,
+      "step": 8599
+    },
+    {
+      "epoch": 1.376,
+      "grad_norm": 2.2568210917292606,
+      "learning_rate": 0.00011835247814723129,
+      "loss": 1.4287,
+      "step": 8600
+    },
+    {
+      "epoch": 1.37616,
+      "grad_norm": 2.132525249107146,
+      "learning_rate": 0.00011833549743162853,
+      "loss": 1.3889,
+      "step": 8601
+    },
+    {
+      "epoch": 1.37632,
+      "grad_norm": 2.0512070757053684,
+      "learning_rate": 0.00011831851616892127,
+      "loss": 1.4218,
+      "step": 8602
+    },
+    {
+      "epoch": 1.37648,
+      "grad_norm": 1.8014013012885794,
+      "learning_rate": 0.00011830153435961614,
+      "loss": 1.3785,
+      "step": 8603
+    },
+    {
+      "epoch": 1.37664,
+      "grad_norm": 1.913817965128138,
+      "learning_rate": 0.00011828455200421993,
+      "loss": 1.2592,
+      "step": 8604
+    },
+    {
+      "epoch": 1.3768,
+      "grad_norm": 2.0157578509963106,
+      "learning_rate": 0.00011826756910323929,
+      "loss": 1.3951,
+      "step": 8605
+    },
+    {
+      "epoch": 1.37696,
+      "grad_norm": 2.2183238024295204,
+      "learning_rate": 0.00011825058565718104,
+      "loss": 1.498,
+      "step": 8606
+    },
+    {
+      "epoch": 1.3771200000000001,
+      "grad_norm": 2.2632167613570693,
+      "learning_rate": 0.00011823360166655188,
+      "loss": 1.4366,
+      "step": 8607
+    },
+    {
+      "epoch": 1.37728,
+      "grad_norm": 2.185015234197331,
+      "learning_rate": 0.00011821661713185866,
+      "loss": 1.3854,
+      "step": 8608
+    },
+    {
+      "epoch": 1.37744,
+      "grad_norm": 2.28243094464451,
+      "learning_rate": 0.00011819963205360808,
+      "loss": 1.3465,
+      "step": 8609
+    },
+    {
+      "epoch": 1.3776,
+      "grad_norm": 2.0904740323766373,
+      "learning_rate": 0.00011818264643230703,
+      "loss": 1.384,
+      "step": 8610
+    },
+    {
+      "epoch": 1.3777599999999999,
+      "grad_norm": 2.500145797208074,
+      "learning_rate": 0.00011816566026846229,
+      "loss": 1.3536,
+      "step": 8611
+    },
+    {
+      "epoch": 1.37792,
+      "grad_norm": 2.093043911233006,
+      "learning_rate": 0.00011814867356258075,
+      "loss": 1.3599,
+      "step": 8612
+    },
+    {
+      "epoch": 1.37808,
+      "grad_norm": 2.161507530037319,
+      "learning_rate": 0.00011813168631516918,
+      "loss": 1.3341,
+      "step": 8613
+    },
+    {
+      "epoch": 1.37824,
+      "grad_norm": 1.9621891081457925,
+      "learning_rate": 0.00011811469852673456,
+      "loss": 1.2873,
+      "step": 8614
+    },
+    {
+      "epoch": 1.3784,
+      "grad_norm": 2.502744446165013,
+      "learning_rate": 0.00011809771019778371,
+      "loss": 1.3529,
+      "step": 8615
+    },
+    {
+      "epoch": 1.37856,
+      "grad_norm": 2.002294447429704,
+      "learning_rate": 0.00011808072132882353,
+      "loss": 1.4089,
+      "step": 8616
+    },
+    {
+      "epoch": 1.37872,
+      "grad_norm": 2.8915064888815274,
+      "learning_rate": 0.00011806373192036101,
+      "loss": 1.4025,
+      "step": 8617
+    },
+    {
+      "epoch": 1.37888,
+      "grad_norm": 2.1088278935273927,
+      "learning_rate": 0.00011804674197290304,
+      "loss": 1.3599,
+      "step": 8618
+    },
+    {
+      "epoch": 1.37904,
+      "grad_norm": 2.496111945545731,
+      "learning_rate": 0.00011802975148695661,
+      "loss": 1.4338,
+      "step": 8619
+    },
+    {
+      "epoch": 1.3792,
+      "grad_norm": 2.553745573535591,
+      "learning_rate": 0.00011801276046302864,
+      "loss": 1.3605,
+      "step": 8620
+    },
+    {
+      "epoch": 1.37936,
+      "grad_norm": 2.401820408542824,
+      "learning_rate": 0.00011799576890162616,
+      "loss": 1.3549,
+      "step": 8621
+    },
+    {
+      "epoch": 1.37952,
+      "grad_norm": 2.5355920746256913,
+      "learning_rate": 0.00011797877680325613,
+      "loss": 1.4224,
+      "step": 8622
+    },
+    {
+      "epoch": 1.37968,
+      "grad_norm": 2.384167681840405,
+      "learning_rate": 0.00011796178416842565,
+      "loss": 1.4319,
+      "step": 8623
+    },
+    {
+      "epoch": 1.37984,
+      "grad_norm": 1.891894955273328,
+      "learning_rate": 0.00011794479099764166,
+      "loss": 1.3814,
+      "step": 8624
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 2.387625657166996,
+      "learning_rate": 0.00011792779729141127,
+      "loss": 1.3754,
+      "step": 8625
+    },
+    {
+      "epoch": 1.38016,
+      "grad_norm": 2.113012092114152,
+      "learning_rate": 0.00011791080305024154,
+      "loss": 1.4246,
+      "step": 8626
+    },
+    {
+      "epoch": 1.38032,
+      "grad_norm": 2.2081290566311416,
+      "learning_rate": 0.00011789380827463954,
+      "loss": 1.4068,
+      "step": 8627
+    },
+    {
+      "epoch": 1.38048,
+      "grad_norm": 2.317230003344263,
+      "learning_rate": 0.00011787681296511236,
+      "loss": 1.3914,
+      "step": 8628
+    },
+    {
+      "epoch": 1.38064,
+      "grad_norm": 2.0964381240957937,
+      "learning_rate": 0.00011785981712216715,
+      "loss": 1.3961,
+      "step": 8629
+    },
+    {
+      "epoch": 1.3808,
+      "grad_norm": 2.3116307004441348,
+      "learning_rate": 0.00011784282074631102,
+      "loss": 1.3362,
+      "step": 8630
+    },
+    {
+      "epoch": 1.38096,
+      "grad_norm": 2.0936465744072246,
+      "learning_rate": 0.0001178258238380511,
+      "loss": 1.3421,
+      "step": 8631
+    },
+    {
+      "epoch": 1.3811200000000001,
+      "grad_norm": 2.2660865392840353,
+      "learning_rate": 0.0001178088263978946,
+      "loss": 1.3929,
+      "step": 8632
+    },
+    {
+      "epoch": 1.38128,
+      "grad_norm": 1.923418045905484,
+      "learning_rate": 0.00011779182842634866,
+      "loss": 1.3018,
+      "step": 8633
+    },
+    {
+      "epoch": 1.38144,
+      "grad_norm": 2.2997542558114428,
+      "learning_rate": 0.00011777482992392048,
+      "loss": 1.3796,
+      "step": 8634
+    },
+    {
+      "epoch": 1.3816,
+      "grad_norm": 2.0092572917059335,
+      "learning_rate": 0.00011775783089111727,
+      "loss": 1.353,
+      "step": 8635
+    },
+    {
+      "epoch": 1.3817599999999999,
+      "grad_norm": 2.6754991835175623,
+      "learning_rate": 0.00011774083132844629,
+      "loss": 1.3863,
+      "step": 8636
+    },
+    {
+      "epoch": 1.38192,
+      "grad_norm": 2.4966155499004046,
+      "learning_rate": 0.00011772383123641473,
+      "loss": 1.3607,
+      "step": 8637
+    },
+    {
+      "epoch": 1.38208,
+      "grad_norm": 2.085353033531172,
+      "learning_rate": 0.00011770683061552985,
+      "loss": 1.3495,
+      "step": 8638
+    },
+    {
+      "epoch": 1.38224,
+      "grad_norm": 2.403368496305255,
+      "learning_rate": 0.00011768982946629898,
+      "loss": 1.3415,
+      "step": 8639
+    },
+    {
+      "epoch": 1.3824,
+      "grad_norm": 1.8852788251262067,
+      "learning_rate": 0.00011767282778922937,
+      "loss": 1.3526,
+      "step": 8640
+    },
+    {
+      "epoch": 1.38256,
+      "grad_norm": 2.113713654007087,
+      "learning_rate": 0.00011765582558482832,
+      "loss": 1.4019,
+      "step": 8641
+    },
+    {
+      "epoch": 1.38272,
+      "grad_norm": 1.962957399540839,
+      "learning_rate": 0.00011763882285360317,
+      "loss": 1.4063,
+      "step": 8642
+    },
+    {
+      "epoch": 1.38288,
+      "grad_norm": 2.016848550660758,
+      "learning_rate": 0.00011762181959606127,
+      "loss": 1.3264,
+      "step": 8643
+    },
+    {
+      "epoch": 1.38304,
+      "grad_norm": 2.0876196053302163,
+      "learning_rate": 0.00011760481581270993,
+      "loss": 1.3622,
+      "step": 8644
+    },
+    {
+      "epoch": 1.3832,
+      "grad_norm": 2.1394834496716375,
+      "learning_rate": 0.00011758781150405654,
+      "loss": 1.3625,
+      "step": 8645
+    },
+    {
+      "epoch": 1.38336,
+      "grad_norm": 2.207949440802264,
+      "learning_rate": 0.00011757080667060844,
+      "loss": 1.4299,
+      "step": 8646
+    },
+    {
+      "epoch": 1.38352,
+      "grad_norm": 1.9194801810680713,
+      "learning_rate": 0.00011755380131287314,
+      "loss": 1.3448,
+      "step": 8647
+    },
+    {
+      "epoch": 1.38368,
+      "grad_norm": 2.4786098124991316,
+      "learning_rate": 0.00011753679543135795,
+      "loss": 1.3992,
+      "step": 8648
+    },
+    {
+      "epoch": 1.38384,
+      "grad_norm": 2.1484841849933516,
+      "learning_rate": 0.00011751978902657037,
+      "loss": 1.3768,
+      "step": 8649
+    },
+    {
+      "epoch": 1.384,
+      "grad_norm": 2.2444789766437525,
+      "learning_rate": 0.00011750278209901779,
+      "loss": 1.3257,
+      "step": 8650
+    },
+    {
+      "epoch": 1.38416,
+      "grad_norm": 2.2294726262819733,
+      "learning_rate": 0.00011748577464920769,
+      "loss": 1.4087,
+      "step": 8651
+    },
+    {
+      "epoch": 1.38432,
+      "grad_norm": 2.4391573360723924,
+      "learning_rate": 0.00011746876667764756,
+      "loss": 1.3701,
+      "step": 8652
+    },
+    {
+      "epoch": 1.38448,
+      "grad_norm": 2.0955053299564637,
+      "learning_rate": 0.0001174517581848449,
+      "loss": 1.3434,
+      "step": 8653
+    },
+    {
+      "epoch": 1.38464,
+      "grad_norm": 1.9224087420016567,
+      "learning_rate": 0.00011743474917130719,
+      "loss": 1.3486,
+      "step": 8654
+    },
+    {
+      "epoch": 1.3848,
+      "grad_norm": 2.501454554126537,
+      "learning_rate": 0.00011741773963754197,
+      "loss": 1.3855,
+      "step": 8655
+    },
+    {
+      "epoch": 1.38496,
+      "grad_norm": 2.4163625029378033,
+      "learning_rate": 0.00011740072958405679,
+      "loss": 1.387,
+      "step": 8656
+    },
+    {
+      "epoch": 1.3851200000000001,
+      "grad_norm": 2.4068505099470165,
+      "learning_rate": 0.00011738371901135921,
+      "loss": 1.398,
+      "step": 8657
+    },
+    {
+      "epoch": 1.38528,
+      "grad_norm": 2.444501118809635,
+      "learning_rate": 0.00011736670791995676,
+      "loss": 1.3296,
+      "step": 8658
+    },
+    {
+      "epoch": 1.38544,
+      "grad_norm": 2.6574583073231866,
+      "learning_rate": 0.00011734969631035706,
+      "loss": 1.3972,
+      "step": 8659
+    },
+    {
+      "epoch": 1.3856,
+      "grad_norm": 2.155305927255454,
+      "learning_rate": 0.00011733268418306771,
+      "loss": 1.3269,
+      "step": 8660
+    },
+    {
+      "epoch": 1.3857599999999999,
+      "grad_norm": 2.782598369004534,
+      "learning_rate": 0.00011731567153859631,
+      "loss": 1.3834,
+      "step": 8661
+    },
+    {
+      "epoch": 1.38592,
+      "grad_norm": 2.575225872289105,
+      "learning_rate": 0.00011729865837745049,
+      "loss": 1.3612,
+      "step": 8662
+    },
+    {
+      "epoch": 1.38608,
+      "grad_norm": 2.12839268987896,
+      "learning_rate": 0.00011728164470013791,
+      "loss": 1.4146,
+      "step": 8663
+    },
+    {
+      "epoch": 1.38624,
+      "grad_norm": 3.337842022150363,
+      "learning_rate": 0.00011726463050716625,
+      "loss": 1.3645,
+      "step": 8664
+    },
+    {
+      "epoch": 1.3864,
+      "grad_norm": 2.2934124086494503,
+      "learning_rate": 0.0001172476157990432,
+      "loss": 1.3868,
+      "step": 8665
+    },
+    {
+      "epoch": 1.38656,
+      "grad_norm": 2.250059734405158,
+      "learning_rate": 0.0001172306005762764,
+      "loss": 1.5076,
+      "step": 8666
+    },
+    {
+      "epoch": 1.38672,
+      "grad_norm": 3.2372230021155914,
+      "learning_rate": 0.00011721358483937358,
+      "loss": 1.3824,
+      "step": 8667
+    },
+    {
+      "epoch": 1.3868800000000001,
+      "grad_norm": 2.2126056063125503,
+      "learning_rate": 0.00011719656858884247,
+      "loss": 1.4098,
+      "step": 8668
+    },
+    {
+      "epoch": 1.38704,
+      "grad_norm": 2.5603294749921908,
+      "learning_rate": 0.00011717955182519083,
+      "loss": 1.3327,
+      "step": 8669
+    },
+    {
+      "epoch": 1.3872,
+      "grad_norm": 2.9750797200401826,
+      "learning_rate": 0.00011716253454892636,
+      "loss": 1.4731,
+      "step": 8670
+    },
+    {
+      "epoch": 1.38736,
+      "grad_norm": 2.0719761591013546,
+      "learning_rate": 0.00011714551676055689,
+      "loss": 1.3383,
+      "step": 8671
+    },
+    {
+      "epoch": 1.3875199999999999,
+      "grad_norm": 2.9473686103100207,
+      "learning_rate": 0.0001171284984605902,
+      "loss": 1.377,
+      "step": 8672
+    },
+    {
+      "epoch": 1.38768,
+      "grad_norm": 2.990542270786769,
+      "learning_rate": 0.00011711147964953406,
+      "loss": 1.3863,
+      "step": 8673
+    },
+    {
+      "epoch": 1.38784,
+      "grad_norm": 2.044500395635488,
+      "learning_rate": 0.0001170944603278963,
+      "loss": 1.3283,
+      "step": 8674
+    },
+    {
+      "epoch": 1.388,
+      "grad_norm": 2.7574018482633353,
+      "learning_rate": 0.00011707744049618474,
+      "loss": 1.3189,
+      "step": 8675
+    },
+    {
+      "epoch": 1.38816,
+      "grad_norm": 2.388833708678039,
+      "learning_rate": 0.00011706042015490725,
+      "loss": 1.3846,
+      "step": 8676
+    },
+    {
+      "epoch": 1.38832,
+      "grad_norm": 2.0288880938143614,
+      "learning_rate": 0.0001170433993045717,
+      "loss": 1.386,
+      "step": 8677
+    },
+    {
+      "epoch": 1.38848,
+      "grad_norm": 2.1400156993367263,
+      "learning_rate": 0.00011702637794568594,
+      "loss": 1.3438,
+      "step": 8678
+    },
+    {
+      "epoch": 1.38864,
+      "grad_norm": 2.033865393967413,
+      "learning_rate": 0.00011700935607875787,
+      "loss": 1.4174,
+      "step": 8679
+    },
+    {
+      "epoch": 1.3888,
+      "grad_norm": 2.3554879776892346,
+      "learning_rate": 0.00011699233370429539,
+      "loss": 1.4202,
+      "step": 8680
+    },
+    {
+      "epoch": 1.38896,
+      "grad_norm": 2.182563065778188,
+      "learning_rate": 0.00011697531082280644,
+      "loss": 1.3538,
+      "step": 8681
+    },
+    {
+      "epoch": 1.3891200000000001,
+      "grad_norm": 2.293781879077588,
+      "learning_rate": 0.00011695828743479896,
+      "loss": 1.3292,
+      "step": 8682
+    },
+    {
+      "epoch": 1.38928,
+      "grad_norm": 2.8442483142237873,
+      "learning_rate": 0.00011694126354078088,
+      "loss": 1.4516,
+      "step": 8683
+    },
+    {
+      "epoch": 1.38944,
+      "grad_norm": 2.0714974526951933,
+      "learning_rate": 0.0001169242391412602,
+      "loss": 1.3227,
+      "step": 8684
+    },
+    {
+      "epoch": 1.3896,
+      "grad_norm": 2.785146781570821,
+      "learning_rate": 0.00011690721423674486,
+      "loss": 1.3668,
+      "step": 8685
+    },
+    {
+      "epoch": 1.3897599999999999,
+      "grad_norm": 2.436171087833042,
+      "learning_rate": 0.00011689018882774293,
+      "loss": 1.3441,
+      "step": 8686
+    },
+    {
+      "epoch": 1.38992,
+      "grad_norm": 2.558375945368871,
+      "learning_rate": 0.00011687316291476233,
+      "loss": 1.3681,
+      "step": 8687
+    },
+    {
+      "epoch": 1.39008,
+      "grad_norm": 3.550285016476457,
+      "learning_rate": 0.00011685613649831113,
+      "loss": 1.3901,
+      "step": 8688
+    },
+    {
+      "epoch": 1.39024,
+      "grad_norm": 2.112565050376393,
+      "learning_rate": 0.00011683910957889739,
+      "loss": 1.3567,
+      "step": 8689
+    },
+    {
+      "epoch": 1.3904,
+      "grad_norm": 2.2993170113465533,
+      "learning_rate": 0.00011682208215702917,
+      "loss": 1.3458,
+      "step": 8690
+    },
+    {
+      "epoch": 1.39056,
+      "grad_norm": 3.0658302802353345,
+      "learning_rate": 0.00011680505423321452,
+      "loss": 1.3718,
+      "step": 8691
+    },
+    {
+      "epoch": 1.39072,
+      "grad_norm": 1.913927118687387,
+      "learning_rate": 0.00011678802580796158,
+      "loss": 1.4008,
+      "step": 8692
+    },
+    {
+      "epoch": 1.3908800000000001,
+      "grad_norm": 2.198133744116908,
+      "learning_rate": 0.00011677099688177837,
+      "loss": 1.3456,
+      "step": 8693
+    },
+    {
+      "epoch": 1.39104,
+      "grad_norm": 2.324543103808429,
+      "learning_rate": 0.00011675396745517306,
+      "loss": 1.3808,
+      "step": 8694
+    },
+    {
+      "epoch": 1.3912,
+      "grad_norm": 2.43796882388488,
+      "learning_rate": 0.00011673693752865378,
+      "loss": 1.3672,
+      "step": 8695
+    },
+    {
+      "epoch": 1.39136,
+      "grad_norm": 2.4232488679762074,
+      "learning_rate": 0.00011671990710272864,
+      "loss": 1.3613,
+      "step": 8696
+    },
+    {
+      "epoch": 1.3915199999999999,
+      "grad_norm": 2.2967963291505975,
+      "learning_rate": 0.00011670287617790587,
+      "loss": 1.3216,
+      "step": 8697
+    },
+    {
+      "epoch": 1.39168,
+      "grad_norm": 2.367732961364043,
+      "learning_rate": 0.00011668584475469359,
+      "loss": 1.3885,
+      "step": 8698
+    },
+    {
+      "epoch": 1.39184,
+      "grad_norm": 2.4673527937607904,
+      "learning_rate": 0.00011666881283360004,
+      "loss": 1.383,
+      "step": 8699
+    },
+    {
+      "epoch": 1.392,
+      "grad_norm": 2.59743084596784,
+      "learning_rate": 0.0001166517804151334,
+      "loss": 1.4318,
+      "step": 8700
+    },
+    {
+      "epoch": 1.39216,
+      "grad_norm": 3.025741931531763,
+      "learning_rate": 0.0001166347474998019,
+      "loss": 1.3023,
+      "step": 8701
+    },
+    {
+      "epoch": 1.39232,
+      "grad_norm": 2.0487595853092206,
+      "learning_rate": 0.00011661771408811377,
+      "loss": 1.3532,
+      "step": 8702
+    },
+    {
+      "epoch": 1.39248,
+      "grad_norm": 3.433060087960804,
+      "learning_rate": 0.00011660068018057726,
+      "loss": 1.3795,
+      "step": 8703
+    },
+    {
+      "epoch": 1.39264,
+      "grad_norm": 2.3188817798274957,
+      "learning_rate": 0.00011658364577770062,
+      "loss": 1.4559,
+      "step": 8704
+    },
+    {
+      "epoch": 1.3928,
+      "grad_norm": 2.109133287081341,
+      "learning_rate": 0.00011656661087999219,
+      "loss": 1.293,
+      "step": 8705
+    },
+    {
+      "epoch": 1.39296,
+      "grad_norm": 2.2527170380360886,
+      "learning_rate": 0.00011654957548796024,
+      "loss": 1.3983,
+      "step": 8706
+    },
+    {
+      "epoch": 1.3931200000000001,
+      "grad_norm": 2.315501637911583,
+      "learning_rate": 0.00011653253960211306,
+      "loss": 1.3739,
+      "step": 8707
+    },
+    {
+      "epoch": 1.39328,
+      "grad_norm": 2.240479778357195,
+      "learning_rate": 0.00011651550322295897,
+      "loss": 1.3376,
+      "step": 8708
+    },
+    {
+      "epoch": 1.39344,
+      "grad_norm": 2.4217092626768806,
+      "learning_rate": 0.00011649846635100634,
+      "loss": 1.3149,
+      "step": 8709
+    },
+    {
+      "epoch": 1.3936,
+      "grad_norm": 2.1276848331648286,
+      "learning_rate": 0.00011648142898676354,
+      "loss": 1.3811,
+      "step": 8710
+    },
+    {
+      "epoch": 1.3937599999999999,
+      "grad_norm": 2.3151855559385766,
+      "learning_rate": 0.00011646439113073888,
+      "loss": 1.3448,
+      "step": 8711
+    },
+    {
+      "epoch": 1.39392,
+      "grad_norm": 2.0459531561864757,
+      "learning_rate": 0.00011644735278344078,
+      "loss": 1.3589,
+      "step": 8712
+    },
+    {
+      "epoch": 1.39408,
+      "grad_norm": 1.9087052804195508,
+      "learning_rate": 0.00011643031394537765,
+      "loss": 1.3328,
+      "step": 8713
+    },
+    {
+      "epoch": 1.39424,
+      "grad_norm": 3.012254228067962,
+      "learning_rate": 0.0001164132746170579,
+      "loss": 1.4347,
+      "step": 8714
+    },
+    {
+      "epoch": 1.3944,
+      "grad_norm": 2.3119403918491495,
+      "learning_rate": 0.00011639623479898996,
+      "loss": 1.3609,
+      "step": 8715
+    },
+    {
+      "epoch": 1.39456,
+      "grad_norm": 2.2762392753494503,
+      "learning_rate": 0.00011637919449168225,
+      "loss": 1.3704,
+      "step": 8716
+    },
+    {
+      "epoch": 1.39472,
+      "grad_norm": 2.2410270630683766,
+      "learning_rate": 0.00011636215369564324,
+      "loss": 1.3381,
+      "step": 8717
+    },
+    {
+      "epoch": 1.3948800000000001,
+      "grad_norm": 2.1307540223605796,
+      "learning_rate": 0.00011634511241138142,
+      "loss": 1.3937,
+      "step": 8718
+    },
+    {
+      "epoch": 1.39504,
+      "grad_norm": 2.143836790272914,
+      "learning_rate": 0.00011632807063940526,
+      "loss": 1.3863,
+      "step": 8719
+    },
+    {
+      "epoch": 1.3952,
+      "grad_norm": 2.0568724122652773,
+      "learning_rate": 0.00011631102838022325,
+      "loss": 1.3647,
+      "step": 8720
+    },
+    {
+      "epoch": 1.39536,
+      "grad_norm": 2.116366334223589,
+      "learning_rate": 0.00011629398563434392,
+      "loss": 1.362,
+      "step": 8721
+    },
+    {
+      "epoch": 1.3955199999999999,
+      "grad_norm": 2.0758082234328437,
+      "learning_rate": 0.00011627694240227583,
+      "loss": 1.4474,
+      "step": 8722
+    },
+    {
+      "epoch": 1.39568,
+      "grad_norm": 2.1048668189464665,
+      "learning_rate": 0.00011625989868452747,
+      "loss": 1.3622,
+      "step": 8723
+    },
+    {
+      "epoch": 1.39584,
+      "grad_norm": 1.921010656609407,
+      "learning_rate": 0.00011624285448160744,
+      "loss": 1.4024,
+      "step": 8724
+    },
+    {
+      "epoch": 1.396,
+      "grad_norm": 2.534618671635416,
+      "learning_rate": 0.00011622580979402431,
+      "loss": 1.3941,
+      "step": 8725
+    },
+    {
+      "epoch": 1.39616,
+      "grad_norm": 1.9494532141881267,
+      "learning_rate": 0.00011620876462228665,
+      "loss": 1.3844,
+      "step": 8726
+    },
+    {
+      "epoch": 1.39632,
+      "grad_norm": 1.8647176460106165,
+      "learning_rate": 0.00011619171896690303,
+      "loss": 1.3664,
+      "step": 8727
+    },
+    {
+      "epoch": 1.39648,
+      "grad_norm": 2.1804932539428292,
+      "learning_rate": 0.00011617467282838215,
+      "loss": 1.3937,
+      "step": 8728
+    },
+    {
+      "epoch": 1.39664,
+      "grad_norm": 1.9613804360212204,
+      "learning_rate": 0.00011615762620723261,
+      "loss": 1.4236,
+      "step": 8729
+    },
+    {
+      "epoch": 1.3968,
+      "grad_norm": 2.209887775103717,
+      "learning_rate": 0.00011614057910396303,
+      "loss": 1.288,
+      "step": 8730
+    },
+    {
+      "epoch": 1.39696,
+      "grad_norm": 2.6038822492104954,
+      "learning_rate": 0.00011612353151908211,
+      "loss": 1.4476,
+      "step": 8731
+    },
+    {
+      "epoch": 1.39712,
+      "grad_norm": 2.5149807286267563,
+      "learning_rate": 0.0001161064834530985,
+      "loss": 1.3955,
+      "step": 8732
+    },
+    {
+      "epoch": 1.39728,
+      "grad_norm": 3.232766116693168,
+      "learning_rate": 0.00011608943490652085,
+      "loss": 1.3464,
+      "step": 8733
+    },
+    {
+      "epoch": 1.39744,
+      "grad_norm": 1.9469760918482897,
+      "learning_rate": 0.00011607238587985794,
+      "loss": 1.3857,
+      "step": 8734
+    },
+    {
+      "epoch": 1.3976,
+      "grad_norm": 2.2633608710493744,
+      "learning_rate": 0.00011605533637361845,
+      "loss": 1.3386,
+      "step": 8735
+    },
+    {
+      "epoch": 1.39776,
+      "grad_norm": 2.411768357752614,
+      "learning_rate": 0.00011603828638831114,
+      "loss": 1.4052,
+      "step": 8736
+    },
+    {
+      "epoch": 1.39792,
+      "grad_norm": 2.1313128324126596,
+      "learning_rate": 0.0001160212359244447,
+      "loss": 1.3549,
+      "step": 8737
+    },
+    {
+      "epoch": 1.39808,
+      "grad_norm": 2.025796136247863,
+      "learning_rate": 0.00011600418498252793,
+      "loss": 1.3769,
+      "step": 8738
+    },
+    {
+      "epoch": 1.39824,
+      "grad_norm": 2.406287794732665,
+      "learning_rate": 0.00011598713356306963,
+      "loss": 1.3221,
+      "step": 8739
+    },
+    {
+      "epoch": 1.3984,
+      "grad_norm": 2.10890538505082,
+      "learning_rate": 0.00011597008166657855,
+      "loss": 1.3389,
+      "step": 8740
+    },
+    {
+      "epoch": 1.39856,
+      "grad_norm": 2.1298951730217537,
+      "learning_rate": 0.00011595302929356351,
+      "loss": 1.3467,
+      "step": 8741
+    },
+    {
+      "epoch": 1.39872,
+      "grad_norm": 1.9453906389380462,
+      "learning_rate": 0.00011593597644453331,
+      "loss": 1.353,
+      "step": 8742
+    },
+    {
+      "epoch": 1.3988800000000001,
+      "grad_norm": 2.1290807587266016,
+      "learning_rate": 0.0001159189231199968,
+      "loss": 1.3091,
+      "step": 8743
+    },
+    {
+      "epoch": 1.39904,
+      "grad_norm": 2.1828872525649827,
+      "learning_rate": 0.00011590186932046281,
+      "loss": 1.352,
+      "step": 8744
+    },
+    {
+      "epoch": 1.3992,
+      "grad_norm": 1.732158691800258,
+      "learning_rate": 0.0001158848150464402,
+      "loss": 1.3452,
+      "step": 8745
+    },
+    {
+      "epoch": 1.39936,
+      "grad_norm": 2.2250923492721193,
+      "learning_rate": 0.00011586776029843786,
+      "loss": 1.4064,
+      "step": 8746
+    },
+    {
+      "epoch": 1.3995199999999999,
+      "grad_norm": 1.8662606409205862,
+      "learning_rate": 0.00011585070507696469,
+      "loss": 1.358,
+      "step": 8747
+    },
+    {
+      "epoch": 1.39968,
+      "grad_norm": 2.3818528582430174,
+      "learning_rate": 0.00011583364938252959,
+      "loss": 1.4597,
+      "step": 8748
+    },
+    {
+      "epoch": 1.39984,
+      "grad_norm": 1.8959463344703236,
+      "learning_rate": 0.00011581659321564145,
+      "loss": 1.3528,
+      "step": 8749
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.795211579064814,
+      "learning_rate": 0.0001157995365768092,
+      "loss": 1.3986,
+      "step": 8750
+    },
+    {
+      "epoch": 1.40016,
+      "grad_norm": 2.186072372164095,
+      "learning_rate": 0.00011578247946654181,
+      "loss": 1.3832,
+      "step": 8751
+    },
+    {
+      "epoch": 1.40032,
+      "grad_norm": 1.921512154239876,
+      "learning_rate": 0.00011576542188534824,
+      "loss": 1.3736,
+      "step": 8752
+    },
+    {
+      "epoch": 1.40048,
+      "grad_norm": 2.1876979770955263,
+      "learning_rate": 0.00011574836383373743,
+      "loss": 1.442,
+      "step": 8753
+    },
+    {
+      "epoch": 1.40064,
+      "grad_norm": 2.323056961776821,
+      "learning_rate": 0.00011573130531221843,
+      "loss": 1.4444,
+      "step": 8754
+    },
+    {
+      "epoch": 1.4008,
+      "grad_norm": 2.199738426558667,
+      "learning_rate": 0.00011571424632130017,
+      "loss": 1.3672,
+      "step": 8755
+    },
+    {
+      "epoch": 1.40096,
+      "grad_norm": 1.9548753998816488,
+      "learning_rate": 0.00011569718686149171,
+      "loss": 1.3075,
+      "step": 8756
+    },
+    {
+      "epoch": 1.40112,
+      "grad_norm": 2.144090941529513,
+      "learning_rate": 0.00011568012693330208,
+      "loss": 1.3377,
+      "step": 8757
+    },
+    {
+      "epoch": 1.40128,
+      "grad_norm": 2.013292182930599,
+      "learning_rate": 0.00011566306653724028,
+      "loss": 1.4233,
+      "step": 8758
+    },
+    {
+      "epoch": 1.40144,
+      "grad_norm": 1.9683662450659136,
+      "learning_rate": 0.00011564600567381543,
+      "loss": 1.3888,
+      "step": 8759
+    },
+    {
+      "epoch": 1.4016,
+      "grad_norm": 2.0687583849630604,
+      "learning_rate": 0.00011562894434353655,
+      "loss": 1.4328,
+      "step": 8760
+    },
+    {
+      "epoch": 1.40176,
+      "grad_norm": 1.9165083518629693,
+      "learning_rate": 0.00011561188254691274,
+      "loss": 1.3663,
+      "step": 8761
+    },
+    {
+      "epoch": 1.40192,
+      "grad_norm": 2.385220768899916,
+      "learning_rate": 0.00011559482028445312,
+      "loss": 1.3326,
+      "step": 8762
+    },
+    {
+      "epoch": 1.40208,
+      "grad_norm": 1.9446107395452419,
+      "learning_rate": 0.00011557775755666678,
+      "loss": 1.3033,
+      "step": 8763
+    },
+    {
+      "epoch": 1.40224,
+      "grad_norm": 1.8496439772496611,
+      "learning_rate": 0.00011556069436406285,
+      "loss": 1.3993,
+      "step": 8764
+    },
+    {
+      "epoch": 1.4024,
+      "grad_norm": 2.2928175523630365,
+      "learning_rate": 0.00011554363070715049,
+      "loss": 1.3473,
+      "step": 8765
+    },
+    {
+      "epoch": 1.40256,
+      "grad_norm": 2.1021280580199275,
+      "learning_rate": 0.00011552656658643881,
+      "loss": 1.3547,
+      "step": 8766
+    },
+    {
+      "epoch": 1.40272,
+      "grad_norm": 2.192587679683647,
+      "learning_rate": 0.00011550950200243704,
+      "loss": 1.3588,
+      "step": 8767
+    },
+    {
+      "epoch": 1.4028800000000001,
+      "grad_norm": 2.1291343727100895,
+      "learning_rate": 0.00011549243695565433,
+      "loss": 1.3798,
+      "step": 8768
+    },
+    {
+      "epoch": 1.40304,
+      "grad_norm": 2.433915704563202,
+      "learning_rate": 0.00011547537144659984,
+      "loss": 1.3338,
+      "step": 8769
+    },
+    {
+      "epoch": 1.4032,
+      "grad_norm": 2.060606113933719,
+      "learning_rate": 0.00011545830547578283,
+      "loss": 1.3644,
+      "step": 8770
+    },
+    {
+      "epoch": 1.40336,
+      "grad_norm": 1.9901381687234907,
+      "learning_rate": 0.00011544123904371251,
+      "loss": 1.3369,
+      "step": 8771
+    },
+    {
+      "epoch": 1.4035199999999999,
+      "grad_norm": 2.6816304031963534,
+      "learning_rate": 0.00011542417215089814,
+      "loss": 1.3088,
+      "step": 8772
+    },
+    {
+      "epoch": 1.40368,
+      "grad_norm": 2.042769491784963,
+      "learning_rate": 0.00011540710479784893,
+      "loss": 1.3223,
+      "step": 8773
+    },
+    {
+      "epoch": 1.40384,
+      "grad_norm": 3.0918553296601075,
+      "learning_rate": 0.00011539003698507417,
+      "loss": 1.4071,
+      "step": 8774
+    },
+    {
+      "epoch": 1.404,
+      "grad_norm": 2.7377149623374164,
+      "learning_rate": 0.00011537296871308312,
+      "loss": 1.3238,
+      "step": 8775
+    },
+    {
+      "epoch": 1.40416,
+      "grad_norm": 2.317592946884226,
+      "learning_rate": 0.0001153558999823851,
+      "loss": 1.3885,
+      "step": 8776
+    },
+    {
+      "epoch": 1.40432,
+      "grad_norm": 2.7717348779055784,
+      "learning_rate": 0.00011533883079348937,
+      "loss": 1.3329,
+      "step": 8777
+    },
+    {
+      "epoch": 1.40448,
+      "grad_norm": 2.675048961075395,
+      "learning_rate": 0.00011532176114690531,
+      "loss": 1.4559,
+      "step": 8778
+    },
+    {
+      "epoch": 1.40464,
+      "grad_norm": 2.334751148867688,
+      "learning_rate": 0.00011530469104314222,
+      "loss": 1.4273,
+      "step": 8779
+    },
+    {
+      "epoch": 1.4048,
+      "grad_norm": 2.5797353577012268,
+      "learning_rate": 0.00011528762048270945,
+      "loss": 1.4383,
+      "step": 8780
+    },
+    {
+      "epoch": 1.40496,
+      "grad_norm": 2.5174615144589483,
+      "learning_rate": 0.00011527054946611635,
+      "loss": 1.3647,
+      "step": 8781
+    },
+    {
+      "epoch": 1.40512,
+      "grad_norm": 2.4164917646244883,
+      "learning_rate": 0.00011525347799387231,
+      "loss": 1.3032,
+      "step": 8782
+    },
+    {
+      "epoch": 1.4052799999999999,
+      "grad_norm": 3.103501567462139,
+      "learning_rate": 0.00011523640606648673,
+      "loss": 1.3874,
+      "step": 8783
+    },
+    {
+      "epoch": 1.40544,
+      "grad_norm": 1.956169517648007,
+      "learning_rate": 0.000115219333684469,
+      "loss": 1.356,
+      "step": 8784
+    },
+    {
+      "epoch": 1.4056,
+      "grad_norm": 1.9494910782751,
+      "learning_rate": 0.00011520226084832848,
+      "loss": 1.3706,
+      "step": 8785
+    },
+    {
+      "epoch": 1.40576,
+      "grad_norm": 2.8286527519667564,
+      "learning_rate": 0.00011518518755857469,
+      "loss": 1.4128,
+      "step": 8786
+    },
+    {
+      "epoch": 1.40592,
+      "grad_norm": 2.366586476578305,
+      "learning_rate": 0.00011516811381571702,
+      "loss": 1.418,
+      "step": 8787
+    },
+    {
+      "epoch": 1.40608,
+      "grad_norm": 2.211954047394811,
+      "learning_rate": 0.00011515103962026495,
+      "loss": 1.3462,
+      "step": 8788
+    },
+    {
+      "epoch": 1.40624,
+      "grad_norm": 2.5412640973769856,
+      "learning_rate": 0.0001151339649727279,
+      "loss": 1.3827,
+      "step": 8789
+    },
+    {
+      "epoch": 1.4064,
+      "grad_norm": 2.556881235602499,
+      "learning_rate": 0.00011511688987361541,
+      "loss": 1.3105,
+      "step": 8790
+    },
+    {
+      "epoch": 1.40656,
+      "grad_norm": 2.43521406756828,
+      "learning_rate": 0.00011509981432343697,
+      "loss": 1.4472,
+      "step": 8791
+    },
+    {
+      "epoch": 1.40672,
+      "grad_norm": 2.5675750897366343,
+      "learning_rate": 0.00011508273832270202,
+      "loss": 1.3395,
+      "step": 8792
+    },
+    {
+      "epoch": 1.4068800000000001,
+      "grad_norm": 2.2570579667243593,
+      "learning_rate": 0.00011506566187192018,
+      "loss": 1.346,
+      "step": 8793
+    },
+    {
+      "epoch": 1.40704,
+      "grad_norm": 2.368589688757452,
+      "learning_rate": 0.00011504858497160092,
+      "loss": 1.4131,
+      "step": 8794
+    },
+    {
+      "epoch": 1.4072,
+      "grad_norm": 2.543653236485865,
+      "learning_rate": 0.00011503150762225383,
+      "loss": 1.3564,
+      "step": 8795
+    },
+    {
+      "epoch": 1.40736,
+      "grad_norm": 2.2492673821212668,
+      "learning_rate": 0.00011501442982438846,
+      "loss": 1.3156,
+      "step": 8796
+    },
+    {
+      "epoch": 1.4075199999999999,
+      "grad_norm": 2.0273287665255815,
+      "learning_rate": 0.00011499735157851437,
+      "loss": 1.3659,
+      "step": 8797
+    },
+    {
+      "epoch": 1.40768,
+      "grad_norm": 2.1367227850644674,
+      "learning_rate": 0.00011498027288514117,
+      "loss": 1.3628,
+      "step": 8798
+    },
+    {
+      "epoch": 1.40784,
+      "grad_norm": 2.3800428714656214,
+      "learning_rate": 0.00011496319374477846,
+      "loss": 1.3952,
+      "step": 8799
+    },
+    {
+      "epoch": 1.408,
+      "grad_norm": 2.185178463634114,
+      "learning_rate": 0.00011494611415793586,
+      "loss": 1.3117,
+      "step": 8800
+    },
+    {
+      "epoch": 1.40816,
+      "grad_norm": 2.5697813921104524,
+      "learning_rate": 0.00011492903412512297,
+      "loss": 1.4239,
+      "step": 8801
+    },
+    {
+      "epoch": 1.40832,
+      "grad_norm": 2.1533154878973164,
+      "learning_rate": 0.00011491195364684943,
+      "loss": 1.3649,
+      "step": 8802
+    },
+    {
+      "epoch": 1.40848,
+      "grad_norm": 2.41535467671848,
+      "learning_rate": 0.00011489487272362498,
+      "loss": 1.376,
+      "step": 8803
+    },
+    {
+      "epoch": 1.4086400000000001,
+      "grad_norm": 2.3486512437204294,
+      "learning_rate": 0.0001148777913559592,
+      "loss": 1.3427,
+      "step": 8804
+    },
+    {
+      "epoch": 1.4088,
+      "grad_norm": 2.5657505858372205,
+      "learning_rate": 0.0001148607095443618,
+      "loss": 1.396,
+      "step": 8805
+    },
+    {
+      "epoch": 1.40896,
+      "grad_norm": 2.0491483192858144,
+      "learning_rate": 0.00011484362728934252,
+      "loss": 1.3734,
+      "step": 8806
+    },
+    {
+      "epoch": 1.40912,
+      "grad_norm": 2.019378634697237,
+      "learning_rate": 0.000114826544591411,
+      "loss": 1.3482,
+      "step": 8807
+    },
+    {
+      "epoch": 1.4092799999999999,
+      "grad_norm": 3.4853499618418646,
+      "learning_rate": 0.00011480946145107701,
+      "loss": 1.3881,
+      "step": 8808
+    },
+    {
+      "epoch": 1.40944,
+      "grad_norm": 1.9392718927216315,
+      "learning_rate": 0.00011479237786885026,
+      "loss": 1.3788,
+      "step": 8809
+    },
+    {
+      "epoch": 1.4096,
+      "grad_norm": 2.325318613417011,
+      "learning_rate": 0.00011477529384524048,
+      "loss": 1.3655,
+      "step": 8810
+    },
+    {
+      "epoch": 1.40976,
+      "grad_norm": 2.4415752943225906,
+      "learning_rate": 0.0001147582093807575,
+      "loss": 1.3903,
+      "step": 8811
+    },
+    {
+      "epoch": 1.40992,
+      "grad_norm": 2.0646644267708494,
+      "learning_rate": 0.00011474112447591106,
+      "loss": 1.3373,
+      "step": 8812
+    },
+    {
+      "epoch": 1.41008,
+      "grad_norm": 2.255769450441019,
+      "learning_rate": 0.00011472403913121092,
+      "loss": 1.3237,
+      "step": 8813
+    },
+    {
+      "epoch": 1.41024,
+      "grad_norm": 2.038509694280356,
+      "learning_rate": 0.00011470695334716693,
+      "loss": 1.3628,
+      "step": 8814
+    },
+    {
+      "epoch": 1.4104,
+      "grad_norm": 2.001755006294026,
+      "learning_rate": 0.00011468986712428889,
+      "loss": 1.3944,
+      "step": 8815
+    },
+    {
+      "epoch": 1.41056,
+      "grad_norm": 2.1271902717236526,
+      "learning_rate": 0.0001146727804630866,
+      "loss": 1.3107,
+      "step": 8816
+    },
+    {
+      "epoch": 1.41072,
+      "grad_norm": 2.123507579342678,
+      "learning_rate": 0.00011465569336406991,
+      "loss": 1.4168,
+      "step": 8817
+    },
+    {
+      "epoch": 1.4108800000000001,
+      "grad_norm": 2.0267736239325216,
+      "learning_rate": 0.00011463860582774871,
+      "loss": 1.2889,
+      "step": 8818
+    },
+    {
+      "epoch": 1.41104,
+      "grad_norm": 2.2666170916331967,
+      "learning_rate": 0.00011462151785463282,
+      "loss": 1.398,
+      "step": 8819
+    },
+    {
+      "epoch": 1.4112,
+      "grad_norm": 2.164155399006329,
+      "learning_rate": 0.00011460442944523216,
+      "loss": 1.3972,
+      "step": 8820
+    },
+    {
+      "epoch": 1.41136,
+      "grad_norm": 2.1566327193467694,
+      "learning_rate": 0.0001145873406000566,
+      "loss": 1.4363,
+      "step": 8821
+    },
+    {
+      "epoch": 1.4115199999999999,
+      "grad_norm": 2.7175178418673305,
+      "learning_rate": 0.00011457025131961607,
+      "loss": 1.3766,
+      "step": 8822
+    },
+    {
+      "epoch": 1.41168,
+      "grad_norm": 2.161714546413883,
+      "learning_rate": 0.00011455316160442046,
+      "loss": 1.3141,
+      "step": 8823
+    },
+    {
+      "epoch": 1.41184,
+      "grad_norm": 2.0082921549844013,
+      "learning_rate": 0.00011453607145497971,
+      "loss": 1.292,
+      "step": 8824
+    },
+    {
+      "epoch": 1.412,
+      "grad_norm": 1.9858308054918468,
+      "learning_rate": 0.00011451898087180377,
+      "loss": 1.4029,
+      "step": 8825
+    },
+    {
+      "epoch": 1.41216,
+      "grad_norm": 2.0853629803849056,
+      "learning_rate": 0.00011450188985540257,
+      "loss": 1.3948,
+      "step": 8826
+    },
+    {
+      "epoch": 1.41232,
+      "grad_norm": 1.8261563861095724,
+      "learning_rate": 0.00011448479840628614,
+      "loss": 1.2535,
+      "step": 8827
+    },
+    {
+      "epoch": 1.41248,
+      "grad_norm": 2.444581481741656,
+      "learning_rate": 0.00011446770652496443,
+      "loss": 1.4675,
+      "step": 8828
+    },
+    {
+      "epoch": 1.4126400000000001,
+      "grad_norm": 2.2770382631696204,
+      "learning_rate": 0.00011445061421194745,
+      "loss": 1.3378,
+      "step": 8829
+    },
+    {
+      "epoch": 1.4128,
+      "grad_norm": 1.6691810191467134,
+      "learning_rate": 0.00011443352146774517,
+      "loss": 1.3043,
+      "step": 8830
+    },
+    {
+      "epoch": 1.41296,
+      "grad_norm": 2.159507307853664,
+      "learning_rate": 0.00011441642829286767,
+      "loss": 1.37,
+      "step": 8831
+    },
+    {
+      "epoch": 1.41312,
+      "grad_norm": 2.3917949787728245,
+      "learning_rate": 0.00011439933468782495,
+      "loss": 1.4513,
+      "step": 8832
+    },
+    {
+      "epoch": 1.4132799999999999,
+      "grad_norm": 2.183717570413989,
+      "learning_rate": 0.00011438224065312705,
+      "loss": 1.366,
+      "step": 8833
+    },
+    {
+      "epoch": 1.41344,
+      "grad_norm": 3.2272412054461377,
+      "learning_rate": 0.00011436514618928405,
+      "loss": 1.4013,
+      "step": 8834
+    },
+    {
+      "epoch": 1.4136,
+      "grad_norm": 2.02668882159251,
+      "learning_rate": 0.00011434805129680603,
+      "loss": 1.3896,
+      "step": 8835
+    },
+    {
+      "epoch": 1.41376,
+      "grad_norm": 3.479984989383443,
+      "learning_rate": 0.00011433095597620305,
+      "loss": 1.3735,
+      "step": 8836
+    },
+    {
+      "epoch": 1.41392,
+      "grad_norm": 2.1788056467478087,
+      "learning_rate": 0.00011431386022798525,
+      "loss": 1.3186,
+      "step": 8837
+    },
+    {
+      "epoch": 1.41408,
+      "grad_norm": 2.040183967387108,
+      "learning_rate": 0.0001142967640526627,
+      "loss": 1.3486,
+      "step": 8838
+    },
+    {
+      "epoch": 1.41424,
+      "grad_norm": 2.6379631370084073,
+      "learning_rate": 0.00011427966745074556,
+      "loss": 1.4623,
+      "step": 8839
+    },
+    {
+      "epoch": 1.4144,
+      "grad_norm": 3.291042825829814,
+      "learning_rate": 0.00011426257042274394,
+      "loss": 1.4041,
+      "step": 8840
+    },
+    {
+      "epoch": 1.41456,
+      "grad_norm": 2.552280139497603,
+      "learning_rate": 0.00011424547296916801,
+      "loss": 1.3567,
+      "step": 8841
+    },
+    {
+      "epoch": 1.41472,
+      "grad_norm": 2.9805514186112068,
+      "learning_rate": 0.00011422837509052788,
+      "loss": 1.3495,
+      "step": 8842
+    },
+    {
+      "epoch": 1.4148800000000001,
+      "grad_norm": 2.1635810118125005,
+      "learning_rate": 0.00011421127678733381,
+      "loss": 1.4,
+      "step": 8843
+    },
+    {
+      "epoch": 1.41504,
+      "grad_norm": 2.191049695677754,
+      "learning_rate": 0.00011419417806009596,
+      "loss": 1.3414,
+      "step": 8844
+    },
+    {
+      "epoch": 1.4152,
+      "grad_norm": 2.55890377485513,
+      "learning_rate": 0.0001141770789093245,
+      "loss": 1.3855,
+      "step": 8845
+    },
+    {
+      "epoch": 1.41536,
+      "grad_norm": 1.9708594825889207,
+      "learning_rate": 0.00011415997933552967,
+      "loss": 1.3539,
+      "step": 8846
+    },
+    {
+      "epoch": 1.41552,
+      "grad_norm": 2.3814973539204387,
+      "learning_rate": 0.0001141428793392217,
+      "loss": 1.3413,
+      "step": 8847
+    },
+    {
+      "epoch": 1.41568,
+      "grad_norm": 2.280796831829884,
+      "learning_rate": 0.00011412577892091079,
+      "loss": 1.4137,
+      "step": 8848
+    },
+    {
+      "epoch": 1.41584,
+      "grad_norm": 2.6273538038509368,
+      "learning_rate": 0.00011410867808110725,
+      "loss": 1.2983,
+      "step": 8849
+    },
+    {
+      "epoch": 1.416,
+      "grad_norm": 2.038611847021983,
+      "learning_rate": 0.00011409157682032127,
+      "loss": 1.4019,
+      "step": 8850
+    },
+    {
+      "epoch": 1.41616,
+      "grad_norm": 2.2205145973090503,
+      "learning_rate": 0.00011407447513906322,
+      "loss": 1.353,
+      "step": 8851
+    },
+    {
+      "epoch": 1.41632,
+      "grad_norm": 2.2486564050336115,
+      "learning_rate": 0.0001140573730378433,
+      "loss": 1.4563,
+      "step": 8852
+    },
+    {
+      "epoch": 1.41648,
+      "grad_norm": 2.217331065159636,
+      "learning_rate": 0.00011404027051717188,
+      "loss": 1.4168,
+      "step": 8853
+    },
+    {
+      "epoch": 1.4166400000000001,
+      "grad_norm": 2.1896065367303534,
+      "learning_rate": 0.00011402316757755924,
+      "loss": 1.3452,
+      "step": 8854
+    },
+    {
+      "epoch": 1.4168,
+      "grad_norm": 1.8143881491758376,
+      "learning_rate": 0.00011400606421951572,
+      "loss": 1.3635,
+      "step": 8855
+    },
+    {
+      "epoch": 1.41696,
+      "grad_norm": 2.1829485906872983,
+      "learning_rate": 0.00011398896044355166,
+      "loss": 1.3544,
+      "step": 8856
+    },
+    {
+      "epoch": 1.41712,
+      "grad_norm": 1.706905522329279,
+      "learning_rate": 0.0001139718562501774,
+      "loss": 1.3208,
+      "step": 8857
+    },
+    {
+      "epoch": 1.4172799999999999,
+      "grad_norm": 2.286786697171408,
+      "learning_rate": 0.00011395475163990327,
+      "loss": 1.3246,
+      "step": 8858
+    },
+    {
+      "epoch": 1.41744,
+      "grad_norm": 2.446187371138573,
+      "learning_rate": 0.00011393764661323974,
+      "loss": 1.3736,
+      "step": 8859
+    },
+    {
+      "epoch": 1.4176,
+      "grad_norm": 1.6454167414403322,
+      "learning_rate": 0.0001139205411706971,
+      "loss": 1.2752,
+      "step": 8860
+    },
+    {
+      "epoch": 1.41776,
+      "grad_norm": 2.8055828473197346,
+      "learning_rate": 0.00011390343531278583,
+      "loss": 1.4813,
+      "step": 8861
+    },
+    {
+      "epoch": 1.41792,
+      "grad_norm": 2.287677893750505,
+      "learning_rate": 0.00011388632904001628,
+      "loss": 1.4364,
+      "step": 8862
+    },
+    {
+      "epoch": 1.41808,
+      "grad_norm": 2.5069957054802585,
+      "learning_rate": 0.00011386922235289894,
+      "loss": 1.3365,
+      "step": 8863
+    },
+    {
+      "epoch": 1.41824,
+      "grad_norm": 2.957390007235285,
+      "learning_rate": 0.00011385211525194419,
+      "loss": 1.4369,
+      "step": 8864
+    },
+    {
+      "epoch": 1.4184,
+      "grad_norm": 2.065669961589702,
+      "learning_rate": 0.00011383500773766252,
+      "loss": 1.4472,
+      "step": 8865
+    },
+    {
+      "epoch": 1.41856,
+      "grad_norm": 2.836948608823741,
+      "learning_rate": 0.00011381789981056437,
+      "loss": 1.4078,
+      "step": 8866
+    },
+    {
+      "epoch": 1.41872,
+      "grad_norm": 2.5393129821020017,
+      "learning_rate": 0.00011380079147116023,
+      "loss": 1.3116,
+      "step": 8867
+    },
+    {
+      "epoch": 1.41888,
+      "grad_norm": 2.127218046577203,
+      "learning_rate": 0.00011378368271996057,
+      "loss": 1.3832,
+      "step": 8868
+    },
+    {
+      "epoch": 1.41904,
+      "grad_norm": 2.9173378235468803,
+      "learning_rate": 0.00011376657355747591,
+      "loss": 1.3597,
+      "step": 8869
+    },
+    {
+      "epoch": 1.4192,
+      "grad_norm": 2.606467034383205,
+      "learning_rate": 0.00011374946398421675,
+      "loss": 1.3124,
+      "step": 8870
+    },
+    {
+      "epoch": 1.41936,
+      "grad_norm": 2.235586570850705,
+      "learning_rate": 0.00011373235400069361,
+      "loss": 1.3052,
+      "step": 8871
+    },
+    {
+      "epoch": 1.41952,
+      "grad_norm": 3.127742190045755,
+      "learning_rate": 0.00011371524360741708,
+      "loss": 1.3947,
+      "step": 8872
+    },
+    {
+      "epoch": 1.41968,
+      "grad_norm": 2.5741301965231904,
+      "learning_rate": 0.00011369813280489763,
+      "loss": 1.3275,
+      "step": 8873
+    },
+    {
+      "epoch": 1.41984,
+      "grad_norm": 2.1291178842959266,
+      "learning_rate": 0.00011368102159364587,
+      "loss": 1.3561,
+      "step": 8874
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 4.466439880723986,
+      "learning_rate": 0.00011366390997417236,
+      "loss": 1.3674,
+      "step": 8875
+    },
+    {
+      "epoch": 1.42016,
+      "grad_norm": 2.5263535687350847,
+      "learning_rate": 0.00011364679794698768,
+      "loss": 1.4133,
+      "step": 8876
+    },
+    {
+      "epoch": 1.42032,
+      "grad_norm": 2.808279302881082,
+      "learning_rate": 0.00011362968551260242,
+      "loss": 1.392,
+      "step": 8877
+    },
+    {
+      "epoch": 1.42048,
+      "grad_norm": 3.9363918979504975,
+      "learning_rate": 0.00011361257267152723,
+      "loss": 1.4351,
+      "step": 8878
+    },
+    {
+      "epoch": 1.4206400000000001,
+      "grad_norm": 2.5388381427438333,
+      "learning_rate": 0.0001135954594242727,
+      "loss": 1.4072,
+      "step": 8879
+    },
+    {
+      "epoch": 1.4208,
+      "grad_norm": 2.542488212923476,
+      "learning_rate": 0.00011357834577134948,
+      "loss": 1.327,
+      "step": 8880
+    },
+    {
+      "epoch": 1.42096,
+      "grad_norm": 3.7491342720882104,
+      "learning_rate": 0.00011356123171326821,
+      "loss": 1.4007,
+      "step": 8881
+    },
+    {
+      "epoch": 1.42112,
+      "grad_norm": 2.103844798180934,
+      "learning_rate": 0.00011354411725053953,
+      "loss": 1.3173,
+      "step": 8882
+    },
+    {
+      "epoch": 1.4212799999999999,
+      "grad_norm": 2.672718607216557,
+      "learning_rate": 0.00011352700238367411,
+      "loss": 1.3888,
+      "step": 8883
+    },
+    {
+      "epoch": 1.42144,
+      "grad_norm": 2.4902433230552536,
+      "learning_rate": 0.00011350988711318269,
+      "loss": 1.3578,
+      "step": 8884
+    },
+    {
+      "epoch": 1.4216,
+      "grad_norm": 2.3466687339241528,
+      "learning_rate": 0.0001134927714395759,
+      "loss": 1.3816,
+      "step": 8885
+    },
+    {
+      "epoch": 1.42176,
+      "grad_norm": 2.1203536819299282,
+      "learning_rate": 0.00011347565536336448,
+      "loss": 1.318,
+      "step": 8886
+    },
+    {
+      "epoch": 1.42192,
+      "grad_norm": 2.255850858899345,
+      "learning_rate": 0.00011345853888505914,
+      "loss": 1.2787,
+      "step": 8887
+    },
+    {
+      "epoch": 1.42208,
+      "grad_norm": 2.3765264586128994,
+      "learning_rate": 0.00011344142200517061,
+      "loss": 1.3038,
+      "step": 8888
+    },
+    {
+      "epoch": 1.42224,
+      "grad_norm": 2.1541286890458835,
+      "learning_rate": 0.00011342430472420964,
+      "loss": 1.3948,
+      "step": 8889
+    },
+    {
+      "epoch": 1.4224,
+      "grad_norm": 2.2169153455054187,
+      "learning_rate": 0.00011340718704268698,
+      "loss": 1.4093,
+      "step": 8890
+    },
+    {
+      "epoch": 1.42256,
+      "grad_norm": 2.1805831406530793,
+      "learning_rate": 0.00011339006896111337,
+      "loss": 1.3371,
+      "step": 8891
+    },
+    {
+      "epoch": 1.42272,
+      "grad_norm": 2.040098905805526,
+      "learning_rate": 0.00011337295047999962,
+      "loss": 1.3923,
+      "step": 8892
+    },
+    {
+      "epoch": 1.42288,
+      "grad_norm": 2.191592870949772,
+      "learning_rate": 0.00011335583159985654,
+      "loss": 1.3948,
+      "step": 8893
+    },
+    {
+      "epoch": 1.42304,
+      "grad_norm": 2.0477386935676862,
+      "learning_rate": 0.00011333871232119489,
+      "loss": 1.379,
+      "step": 8894
+    },
+    {
+      "epoch": 1.4232,
+      "grad_norm": 1.9427083804795615,
+      "learning_rate": 0.0001133215926445255,
+      "loss": 1.3127,
+      "step": 8895
+    },
+    {
+      "epoch": 1.42336,
+      "grad_norm": 2.0114781026497925,
+      "learning_rate": 0.00011330447257035921,
+      "loss": 1.4124,
+      "step": 8896
+    },
+    {
+      "epoch": 1.42352,
+      "grad_norm": 2.5476593908946135,
+      "learning_rate": 0.00011328735209920682,
+      "loss": 1.3422,
+      "step": 8897
+    },
+    {
+      "epoch": 1.42368,
+      "grad_norm": 2.3530893809529863,
+      "learning_rate": 0.00011327023123157923,
+      "loss": 1.3595,
+      "step": 8898
+    },
+    {
+      "epoch": 1.42384,
+      "grad_norm": 2.5165364754216224,
+      "learning_rate": 0.00011325310996798725,
+      "loss": 1.3601,
+      "step": 8899
+    },
+    {
+      "epoch": 1.424,
+      "grad_norm": 2.1471393823764755,
+      "learning_rate": 0.0001132359883089418,
+      "loss": 1.4067,
+      "step": 8900
+    },
+    {
+      "epoch": 1.42416,
+      "grad_norm": 2.023432935859794,
+      "learning_rate": 0.00011321886625495375,
+      "loss": 1.4275,
+      "step": 8901
+    },
+    {
+      "epoch": 1.42432,
+      "grad_norm": 2.5690837666673234,
+      "learning_rate": 0.00011320174380653399,
+      "loss": 1.4065,
+      "step": 8902
+    },
+    {
+      "epoch": 1.42448,
+      "grad_norm": 2.32001299223391,
+      "learning_rate": 0.00011318462096419343,
+      "loss": 1.3562,
+      "step": 8903
+    },
+    {
+      "epoch": 1.4246400000000001,
+      "grad_norm": 2.326294103674376,
+      "learning_rate": 0.00011316749772844297,
+      "loss": 1.3271,
+      "step": 8904
+    },
+    {
+      "epoch": 1.4248,
+      "grad_norm": 2.2259988905923276,
+      "learning_rate": 0.00011315037409979358,
+      "loss": 1.3489,
+      "step": 8905
+    },
+    {
+      "epoch": 1.42496,
+      "grad_norm": 2.42459203655878,
+      "learning_rate": 0.00011313325007875621,
+      "loss": 1.3788,
+      "step": 8906
+    },
+    {
+      "epoch": 1.42512,
+      "grad_norm": 2.1577823874410726,
+      "learning_rate": 0.00011311612566584175,
+      "loss": 1.312,
+      "step": 8907
+    },
+    {
+      "epoch": 1.4252799999999999,
+      "grad_norm": 2.173229028022804,
+      "learning_rate": 0.00011309900086156124,
+      "loss": 1.332,
+      "step": 8908
+    },
+    {
+      "epoch": 1.42544,
+      "grad_norm": 2.6753564277757,
+      "learning_rate": 0.00011308187566642562,
+      "loss": 1.3303,
+      "step": 8909
+    },
+    {
+      "epoch": 1.4256,
+      "grad_norm": 2.3777647258686523,
+      "learning_rate": 0.00011306475008094591,
+      "loss": 1.4001,
+      "step": 8910
+    },
+    {
+      "epoch": 1.42576,
+      "grad_norm": 2.382489050376952,
+      "learning_rate": 0.0001130476241056331,
+      "loss": 1.3173,
+      "step": 8911
+    },
+    {
+      "epoch": 1.42592,
+      "grad_norm": 2.4521847340981653,
+      "learning_rate": 0.00011303049774099819,
+      "loss": 1.3228,
+      "step": 8912
+    },
+    {
+      "epoch": 1.42608,
+      "grad_norm": 2.272680841488391,
+      "learning_rate": 0.0001130133709875522,
+      "loss": 1.3293,
+      "step": 8913
+    },
+    {
+      "epoch": 1.42624,
+      "grad_norm": 1.971208855687214,
+      "learning_rate": 0.0001129962438458062,
+      "loss": 1.3315,
+      "step": 8914
+    },
+    {
+      "epoch": 1.4264000000000001,
+      "grad_norm": 2.1298886431981963,
+      "learning_rate": 0.00011297911631627122,
+      "loss": 1.3603,
+      "step": 8915
+    },
+    {
+      "epoch": 1.42656,
+      "grad_norm": 3.0970294474366713,
+      "learning_rate": 0.00011296198839945833,
+      "loss": 1.3493,
+      "step": 8916
+    },
+    {
+      "epoch": 1.42672,
+      "grad_norm": 2.1592084523589983,
+      "learning_rate": 0.00011294486009587857,
+      "loss": 1.3462,
+      "step": 8917
+    },
+    {
+      "epoch": 1.42688,
+      "grad_norm": 2.369481639390076,
+      "learning_rate": 0.00011292773140604304,
+      "loss": 1.3309,
+      "step": 8918
+    },
+    {
+      "epoch": 1.4270399999999999,
+      "grad_norm": 1.9294397356686406,
+      "learning_rate": 0.00011291060233046287,
+      "loss": 1.4028,
+      "step": 8919
+    },
+    {
+      "epoch": 1.4272,
+      "grad_norm": 1.87507831371444,
+      "learning_rate": 0.00011289347286964914,
+      "loss": 1.3453,
+      "step": 8920
+    },
+    {
+      "epoch": 1.42736,
+      "grad_norm": 2.9375790813993206,
+      "learning_rate": 0.00011287634302411296,
+      "loss": 1.3579,
+      "step": 8921
+    },
+    {
+      "epoch": 1.42752,
+      "grad_norm": 2.418931864245853,
+      "learning_rate": 0.00011285921279436546,
+      "loss": 1.3444,
+      "step": 8922
+    },
+    {
+      "epoch": 1.42768,
+      "grad_norm": 2.5502489150250445,
+      "learning_rate": 0.00011284208218091781,
+      "loss": 1.399,
+      "step": 8923
+    },
+    {
+      "epoch": 1.42784,
+      "grad_norm": 2.2856550079933564,
+      "learning_rate": 0.00011282495118428111,
+      "loss": 1.3737,
+      "step": 8924
+    },
+    {
+      "epoch": 1.428,
+      "grad_norm": 2.061580465629217,
+      "learning_rate": 0.00011280781980496655,
+      "loss": 1.3332,
+      "step": 8925
+    },
+    {
+      "epoch": 1.42816,
+      "grad_norm": 2.3386585269566846,
+      "learning_rate": 0.00011279068804348534,
+      "loss": 1.433,
+      "step": 8926
+    },
+    {
+      "epoch": 1.42832,
+      "grad_norm": 2.277421911178962,
+      "learning_rate": 0.00011277355590034863,
+      "loss": 1.4445,
+      "step": 8927
+    },
+    {
+      "epoch": 1.42848,
+      "grad_norm": 2.3870793390988148,
+      "learning_rate": 0.00011275642337606762,
+      "loss": 1.3467,
+      "step": 8928
+    },
+    {
+      "epoch": 1.4286400000000001,
+      "grad_norm": 2.4756749137852876,
+      "learning_rate": 0.00011273929047115352,
+      "loss": 1.3416,
+      "step": 8929
+    },
+    {
+      "epoch": 1.4288,
+      "grad_norm": 2.016325514034362,
+      "learning_rate": 0.00011272215718611758,
+      "loss": 1.3591,
+      "step": 8930
+    },
+    {
+      "epoch": 1.42896,
+      "grad_norm": 2.2062384297804813,
+      "learning_rate": 0.00011270502352147098,
+      "loss": 1.3709,
+      "step": 8931
+    },
+    {
+      "epoch": 1.42912,
+      "grad_norm": 2.0080519781087527,
+      "learning_rate": 0.000112687889477725,
+      "loss": 1.3681,
+      "step": 8932
+    },
+    {
+      "epoch": 1.4292799999999999,
+      "grad_norm": 2.1049968049357424,
+      "learning_rate": 0.00011267075505539087,
+      "loss": 1.3845,
+      "step": 8933
+    },
+    {
+      "epoch": 1.42944,
+      "grad_norm": 2.0804292806583566,
+      "learning_rate": 0.0001126536202549799,
+      "loss": 1.2871,
+      "step": 8934
+    },
+    {
+      "epoch": 1.4296,
+      "grad_norm": 2.4041589740169997,
+      "learning_rate": 0.00011263648507700333,
+      "loss": 1.3519,
+      "step": 8935
+    },
+    {
+      "epoch": 1.42976,
+      "grad_norm": 2.37614288906804,
+      "learning_rate": 0.00011261934952197247,
+      "loss": 1.3161,
+      "step": 8936
+    },
+    {
+      "epoch": 1.42992,
+      "grad_norm": 1.93137303075798,
+      "learning_rate": 0.00011260221359039857,
+      "loss": 1.3847,
+      "step": 8937
+    },
+    {
+      "epoch": 1.43008,
+      "grad_norm": 2.55022919198689,
+      "learning_rate": 0.00011258507728279302,
+      "loss": 1.3709,
+      "step": 8938
+    },
+    {
+      "epoch": 1.43024,
+      "grad_norm": 1.8823155365287865,
+      "learning_rate": 0.00011256794059966709,
+      "loss": 1.3382,
+      "step": 8939
+    },
+    {
+      "epoch": 1.4304000000000001,
+      "grad_norm": 2.0829218987365703,
+      "learning_rate": 0.00011255080354153211,
+      "loss": 1.3083,
+      "step": 8940
+    },
+    {
+      "epoch": 1.43056,
+      "grad_norm": 2.0304630071715026,
+      "learning_rate": 0.00011253366610889944,
+      "loss": 1.3416,
+      "step": 8941
+    },
+    {
+      "epoch": 1.43072,
+      "grad_norm": 2.4683641745435687,
+      "learning_rate": 0.00011251652830228045,
+      "loss": 1.3486,
+      "step": 8942
+    },
+    {
+      "epoch": 1.43088,
+      "grad_norm": 2.0033697055628834,
+      "learning_rate": 0.0001124993901221865,
+      "loss": 1.3224,
+      "step": 8943
+    },
+    {
+      "epoch": 1.4310399999999999,
+      "grad_norm": 2.0022271020008393,
+      "learning_rate": 0.00011248225156912897,
+      "loss": 1.3757,
+      "step": 8944
+    },
+    {
+      "epoch": 1.4312,
+      "grad_norm": 2.361166366726185,
+      "learning_rate": 0.00011246511264361923,
+      "loss": 1.3382,
+      "step": 8945
+    },
+    {
+      "epoch": 1.43136,
+      "grad_norm": 2.2529408675609526,
+      "learning_rate": 0.0001124479733461687,
+      "loss": 1.3575,
+      "step": 8946
+    },
+    {
+      "epoch": 1.43152,
+      "grad_norm": 1.9997313702081285,
+      "learning_rate": 0.00011243083367728878,
+      "loss": 1.3622,
+      "step": 8947
+    },
+    {
+      "epoch": 1.43168,
+      "grad_norm": 2.1981190210469856,
+      "learning_rate": 0.00011241369363749088,
+      "loss": 1.4002,
+      "step": 8948
+    },
+    {
+      "epoch": 1.43184,
+      "grad_norm": 2.368339903886725,
+      "learning_rate": 0.00011239655322728644,
+      "loss": 1.4156,
+      "step": 8949
+    },
+    {
+      "epoch": 1.432,
+      "grad_norm": 2.2207902917992794,
+      "learning_rate": 0.00011237941244718695,
+      "loss": 1.5494,
+      "step": 8950
+    },
+    {
+      "epoch": 1.43216,
+      "grad_norm": 1.9305004623977131,
+      "learning_rate": 0.00011236227129770383,
+      "loss": 1.3597,
+      "step": 8951
+    },
+    {
+      "epoch": 1.43232,
+      "grad_norm": 2.212539660697231,
+      "learning_rate": 0.00011234512977934852,
+      "loss": 1.358,
+      "step": 8952
+    },
+    {
+      "epoch": 1.43248,
+      "grad_norm": 2.1415641971488166,
+      "learning_rate": 0.00011232798789263255,
+      "loss": 1.3665,
+      "step": 8953
+    },
+    {
+      "epoch": 1.4326400000000001,
+      "grad_norm": 2.2172712125060965,
+      "learning_rate": 0.00011231084563806737,
+      "loss": 1.3603,
+      "step": 8954
+    },
+    {
+      "epoch": 1.4328,
+      "grad_norm": 2.5247847756864523,
+      "learning_rate": 0.0001122937030161645,
+      "loss": 1.4689,
+      "step": 8955
+    },
+    {
+      "epoch": 1.43296,
+      "grad_norm": 1.9041968572390429,
+      "learning_rate": 0.00011227656002743547,
+      "loss": 1.3438,
+      "step": 8956
+    },
+    {
+      "epoch": 1.43312,
+      "grad_norm": 1.995478161280018,
+      "learning_rate": 0.00011225941667239172,
+      "loss": 1.4156,
+      "step": 8957
+    },
+    {
+      "epoch": 1.4332799999999999,
+      "grad_norm": 2.060968240653892,
+      "learning_rate": 0.0001122422729515449,
+      "loss": 1.3138,
+      "step": 8958
+    },
+    {
+      "epoch": 1.43344,
+      "grad_norm": 2.219543010575779,
+      "learning_rate": 0.00011222512886540645,
+      "loss": 1.3864,
+      "step": 8959
+    },
+    {
+      "epoch": 1.4336,
+      "grad_norm": 2.0966882102181335,
+      "learning_rate": 0.00011220798441448801,
+      "loss": 1.3715,
+      "step": 8960
+    },
+    {
+      "epoch": 1.43376,
+      "grad_norm": 2.728017340355712,
+      "learning_rate": 0.00011219083959930108,
+      "loss": 1.2958,
+      "step": 8961
+    },
+    {
+      "epoch": 1.43392,
+      "grad_norm": 2.0923439844423117,
+      "learning_rate": 0.00011217369442035727,
+      "loss": 1.3715,
+      "step": 8962
+    },
+    {
+      "epoch": 1.43408,
+      "grad_norm": 2.2425551773845385,
+      "learning_rate": 0.00011215654887816815,
+      "loss": 1.4231,
+      "step": 8963
+    },
+    {
+      "epoch": 1.43424,
+      "grad_norm": 2.214999621180169,
+      "learning_rate": 0.00011213940297324534,
+      "loss": 1.3928,
+      "step": 8964
+    },
+    {
+      "epoch": 1.4344000000000001,
+      "grad_norm": 2.1436995368210705,
+      "learning_rate": 0.0001121222567061004,
+      "loss": 1.4521,
+      "step": 8965
+    },
+    {
+      "epoch": 1.43456,
+      "grad_norm": 2.003460476122135,
+      "learning_rate": 0.000112105110077245,
+      "loss": 1.3776,
+      "step": 8966
+    },
+    {
+      "epoch": 1.43472,
+      "grad_norm": 2.423839006911758,
+      "learning_rate": 0.00011208796308719074,
+      "loss": 1.3664,
+      "step": 8967
+    },
+    {
+      "epoch": 1.43488,
+      "grad_norm": 2.0040192824641947,
+      "learning_rate": 0.00011207081573644931,
+      "loss": 1.2943,
+      "step": 8968
+    },
+    {
+      "epoch": 1.4350399999999999,
+      "grad_norm": 2.123825805248611,
+      "learning_rate": 0.0001120536680255323,
+      "loss": 1.3202,
+      "step": 8969
+    },
+    {
+      "epoch": 1.4352,
+      "grad_norm": 2.1810886567583907,
+      "learning_rate": 0.00011203651995495142,
+      "loss": 1.3202,
+      "step": 8970
+    },
+    {
+      "epoch": 1.43536,
+      "grad_norm": 1.8352492260732427,
+      "learning_rate": 0.00011201937152521833,
+      "loss": 1.3891,
+      "step": 8971
+    },
+    {
+      "epoch": 1.43552,
+      "grad_norm": 2.244632450940742,
+      "learning_rate": 0.00011200222273684469,
+      "loss": 1.3773,
+      "step": 8972
+    },
+    {
+      "epoch": 1.43568,
+      "grad_norm": 2.2880357060754988,
+      "learning_rate": 0.0001119850735903422,
+      "loss": 1.326,
+      "step": 8973
+    },
+    {
+      "epoch": 1.43584,
+      "grad_norm": 2.040104548414478,
+      "learning_rate": 0.00011196792408622259,
+      "loss": 1.3428,
+      "step": 8974
+    },
+    {
+      "epoch": 1.436,
+      "grad_norm": 2.2033691490389966,
+      "learning_rate": 0.00011195077422499757,
+      "loss": 1.4178,
+      "step": 8975
+    },
+    {
+      "epoch": 1.43616,
+      "grad_norm": 2.3002015077858835,
+      "learning_rate": 0.00011193362400717886,
+      "loss": 1.304,
+      "step": 8976
+    },
+    {
+      "epoch": 1.43632,
+      "grad_norm": 1.9583730772232883,
+      "learning_rate": 0.00011191647343327821,
+      "loss": 1.3733,
+      "step": 8977
+    },
+    {
+      "epoch": 1.43648,
+      "grad_norm": 2.3698046451031343,
+      "learning_rate": 0.00011189932250380734,
+      "loss": 1.366,
+      "step": 8978
+    },
+    {
+      "epoch": 1.43664,
+      "grad_norm": 2.166155246268235,
+      "learning_rate": 0.00011188217121927804,
+      "loss": 1.3325,
+      "step": 8979
+    },
+    {
+      "epoch": 1.4368,
+      "grad_norm": 2.1508841192347936,
+      "learning_rate": 0.00011186501958020206,
+      "loss": 1.2952,
+      "step": 8980
+    },
+    {
+      "epoch": 1.43696,
+      "grad_norm": 1.919376795088663,
+      "learning_rate": 0.00011184786758709119,
+      "loss": 1.3702,
+      "step": 8981
+    },
+    {
+      "epoch": 1.43712,
+      "grad_norm": 2.421052830630891,
+      "learning_rate": 0.00011183071524045718,
+      "loss": 1.4087,
+      "step": 8982
+    },
+    {
+      "epoch": 1.43728,
+      "grad_norm": 2.3226035520127404,
+      "learning_rate": 0.0001118135625408119,
+      "loss": 1.3007,
+      "step": 8983
+    },
+    {
+      "epoch": 1.43744,
+      "grad_norm": 1.9093503318514495,
+      "learning_rate": 0.00011179640948866712,
+      "loss": 1.3547,
+      "step": 8984
+    },
+    {
+      "epoch": 1.4376,
+      "grad_norm": 3.1164990291454795,
+      "learning_rate": 0.00011177925608453468,
+      "loss": 1.3405,
+      "step": 8985
+    },
+    {
+      "epoch": 1.43776,
+      "grad_norm": 2.131331001909804,
+      "learning_rate": 0.00011176210232892638,
+      "loss": 1.3684,
+      "step": 8986
+    },
+    {
+      "epoch": 1.43792,
+      "grad_norm": 2.5178940945028856,
+      "learning_rate": 0.0001117449482223541,
+      "loss": 1.4506,
+      "step": 8987
+    },
+    {
+      "epoch": 1.43808,
+      "grad_norm": 3.101634191454257,
+      "learning_rate": 0.00011172779376532969,
+      "loss": 1.3308,
+      "step": 8988
+    },
+    {
+      "epoch": 1.43824,
+      "grad_norm": 1.9705742159777182,
+      "learning_rate": 0.000111710638958365,
+      "loss": 1.4459,
+      "step": 8989
+    },
+    {
+      "epoch": 1.4384000000000001,
+      "grad_norm": 2.5234185985674413,
+      "learning_rate": 0.00011169348380197186,
+      "loss": 1.3982,
+      "step": 8990
+    },
+    {
+      "epoch": 1.43856,
+      "grad_norm": 2.601471499030682,
+      "learning_rate": 0.00011167632829666225,
+      "loss": 1.3915,
+      "step": 8991
+    },
+    {
+      "epoch": 1.43872,
+      "grad_norm": 2.1220713980558297,
+      "learning_rate": 0.000111659172442948,
+      "loss": 1.3505,
+      "step": 8992
+    },
+    {
+      "epoch": 1.43888,
+      "grad_norm": 2.4344856405376722,
+      "learning_rate": 0.00011164201624134108,
+      "loss": 1.3297,
+      "step": 8993
+    },
+    {
+      "epoch": 1.4390399999999999,
+      "grad_norm": 2.4714286787240436,
+      "learning_rate": 0.0001116248596923533,
+      "loss": 1.3567,
+      "step": 8994
+    },
+    {
+      "epoch": 1.4392,
+      "grad_norm": 2.53104936160666,
+      "learning_rate": 0.00011160770279649669,
+      "loss": 1.3315,
+      "step": 8995
+    },
+    {
+      "epoch": 1.43936,
+      "grad_norm": 3.085549993749397,
+      "learning_rate": 0.00011159054555428313,
+      "loss": 1.3317,
+      "step": 8996
+    },
+    {
+      "epoch": 1.43952,
+      "grad_norm": 2.073841731058718,
+      "learning_rate": 0.00011157338796622456,
+      "loss": 1.3799,
+      "step": 8997
+    },
+    {
+      "epoch": 1.43968,
+      "grad_norm": 2.4053167510648477,
+      "learning_rate": 0.00011155623003283295,
+      "loss": 1.3054,
+      "step": 8998
+    },
+    {
+      "epoch": 1.43984,
+      "grad_norm": 2.9242263208167576,
+      "learning_rate": 0.0001115390717546203,
+      "loss": 1.3643,
+      "step": 8999
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 2.0350918432448672,
+      "learning_rate": 0.00011152191313209854,
+      "loss": 1.4345,
+      "step": 9000
+    },
+    {
+      "epoch": 1.44016,
+      "grad_norm": 3.2552160682566456,
+      "learning_rate": 0.00011150475416577972,
+      "loss": 1.3777,
+      "step": 9001
+    },
+    {
+      "epoch": 1.44032,
+      "grad_norm": 2.0358245840794558,
+      "learning_rate": 0.00011148759485617576,
+      "loss": 1.3627,
+      "step": 9002
+    },
+    {
+      "epoch": 1.44048,
+      "grad_norm": 2.263738683981031,
+      "learning_rate": 0.00011147043520379874,
+      "loss": 1.3401,
+      "step": 9003
+    },
+    {
+      "epoch": 1.44064,
+      "grad_norm": 2.8556644753239984,
+      "learning_rate": 0.00011145327520916064,
+      "loss": 1.3968,
+      "step": 9004
+    },
+    {
+      "epoch": 1.4408,
+      "grad_norm": 2.1330103096367954,
+      "learning_rate": 0.0001114361148727735,
+      "loss": 1.3401,
+      "step": 9005
+    },
+    {
+      "epoch": 1.44096,
+      "grad_norm": 2.99481465865117,
+      "learning_rate": 0.00011141895419514932,
+      "loss": 1.3508,
+      "step": 9006
+    },
+    {
+      "epoch": 1.44112,
+      "grad_norm": 2.2539094073265833,
+      "learning_rate": 0.00011140179317680021,
+      "loss": 1.3065,
+      "step": 9007
+    },
+    {
+      "epoch": 1.44128,
+      "grad_norm": 2.576406413289231,
+      "learning_rate": 0.00011138463181823822,
+      "loss": 1.4176,
+      "step": 9008
+    },
+    {
+      "epoch": 1.44144,
+      "grad_norm": 3.1282602965087443,
+      "learning_rate": 0.0001113674701199754,
+      "loss": 1.3965,
+      "step": 9009
+    },
+    {
+      "epoch": 1.4416,
+      "grad_norm": 2.1296366059782397,
+      "learning_rate": 0.00011135030808252383,
+      "loss": 1.3308,
+      "step": 9010
+    },
+    {
+      "epoch": 1.44176,
+      "grad_norm": 2.21261853133655,
+      "learning_rate": 0.00011133314570639561,
+      "loss": 1.3089,
+      "step": 9011
+    },
+    {
+      "epoch": 1.44192,
+      "grad_norm": 3.5131257269303946,
+      "learning_rate": 0.00011131598299210284,
+      "loss": 1.4102,
+      "step": 9012
+    },
+    {
+      "epoch": 1.44208,
+      "grad_norm": 2.0423263372567595,
+      "learning_rate": 0.00011129881994015763,
+      "loss": 1.3655,
+      "step": 9013
+    },
+    {
+      "epoch": 1.44224,
+      "grad_norm": 2.284246133200354,
+      "learning_rate": 0.00011128165655107206,
+      "loss": 1.4184,
+      "step": 9014
+    },
+    {
+      "epoch": 1.4424000000000001,
+      "grad_norm": 2.3934372345959165,
+      "learning_rate": 0.00011126449282535833,
+      "loss": 1.3469,
+      "step": 9015
+    },
+    {
+      "epoch": 1.44256,
+      "grad_norm": 1.9780143525027563,
+      "learning_rate": 0.00011124732876352854,
+      "loss": 1.2982,
+      "step": 9016
+    },
+    {
+      "epoch": 1.44272,
+      "grad_norm": 2.0573746829288377,
+      "learning_rate": 0.00011123016436609486,
+      "loss": 1.3878,
+      "step": 9017
+    },
+    {
+      "epoch": 1.44288,
+      "grad_norm": 2.2472482135038168,
+      "learning_rate": 0.00011121299963356943,
+      "loss": 1.3478,
+      "step": 9018
+    },
+    {
+      "epoch": 1.4430399999999999,
+      "grad_norm": 2.3210891050792135,
+      "learning_rate": 0.00011119583456646443,
+      "loss": 1.3611,
+      "step": 9019
+    },
+    {
+      "epoch": 1.4432,
+      "grad_norm": 2.2098045726417803,
+      "learning_rate": 0.00011117866916529205,
+      "loss": 1.4196,
+      "step": 9020
+    },
+    {
+      "epoch": 1.44336,
+      "grad_norm": 1.828869825813337,
+      "learning_rate": 0.00011116150343056446,
+      "loss": 1.4228,
+      "step": 9021
+    },
+    {
+      "epoch": 1.44352,
+      "grad_norm": 2.248756661309979,
+      "learning_rate": 0.00011114433736279386,
+      "loss": 1.3761,
+      "step": 9022
+    },
+    {
+      "epoch": 1.44368,
+      "grad_norm": 1.9592544718089582,
+      "learning_rate": 0.0001111271709624925,
+      "loss": 1.3862,
+      "step": 9023
+    },
+    {
+      "epoch": 1.44384,
+      "grad_norm": 1.9754788504485308,
+      "learning_rate": 0.00011111000423017255,
+      "loss": 1.3003,
+      "step": 9024
+    },
+    {
+      "epoch": 1.444,
+      "grad_norm": 2.030103961907083,
+      "learning_rate": 0.00011109283716634627,
+      "loss": 1.3353,
+      "step": 9025
+    },
+    {
+      "epoch": 1.44416,
+      "grad_norm": 2.4202273547554496,
+      "learning_rate": 0.00011107566977152588,
+      "loss": 1.3679,
+      "step": 9026
+    },
+    {
+      "epoch": 1.44432,
+      "grad_norm": 2.1516638769481755,
+      "learning_rate": 0.00011105850204622365,
+      "loss": 1.3696,
+      "step": 9027
+    },
+    {
+      "epoch": 1.44448,
+      "grad_norm": 1.832033012847825,
+      "learning_rate": 0.00011104133399095185,
+      "loss": 1.4348,
+      "step": 9028
+    },
+    {
+      "epoch": 1.44464,
+      "grad_norm": 1.9682790209293741,
+      "learning_rate": 0.00011102416560622273,
+      "loss": 1.312,
+      "step": 9029
+    },
+    {
+      "epoch": 1.4447999999999999,
+      "grad_norm": 2.047683772871493,
+      "learning_rate": 0.00011100699689254856,
+      "loss": 1.407,
+      "step": 9030
+    },
+    {
+      "epoch": 1.44496,
+      "grad_norm": 2.435261756186513,
+      "learning_rate": 0.00011098982785044166,
+      "loss": 1.3287,
+      "step": 9031
+    },
+    {
+      "epoch": 1.44512,
+      "grad_norm": 2.423183622759019,
+      "learning_rate": 0.00011097265848041427,
+      "loss": 1.362,
+      "step": 9032
+    },
+    {
+      "epoch": 1.44528,
+      "grad_norm": 2.262894039305916,
+      "learning_rate": 0.00011095548878297877,
+      "loss": 1.3645,
+      "step": 9033
+    },
+    {
+      "epoch": 1.44544,
+      "grad_norm": 2.044020102992937,
+      "learning_rate": 0.00011093831875864746,
+      "loss": 1.3503,
+      "step": 9034
+    },
+    {
+      "epoch": 1.4456,
+      "grad_norm": 2.3098063323133506,
+      "learning_rate": 0.00011092114840793263,
+      "loss": 1.3206,
+      "step": 9035
+    },
+    {
+      "epoch": 1.44576,
+      "grad_norm": 1.9595583183869099,
+      "learning_rate": 0.00011090397773134667,
+      "loss": 1.3919,
+      "step": 9036
+    },
+    {
+      "epoch": 1.44592,
+      "grad_norm": 2.0211339992521595,
+      "learning_rate": 0.0001108868067294019,
+      "loss": 1.3408,
+      "step": 9037
+    },
+    {
+      "epoch": 1.44608,
+      "grad_norm": 1.9139374873013217,
+      "learning_rate": 0.00011086963540261067,
+      "loss": 1.3392,
+      "step": 9038
+    },
+    {
+      "epoch": 1.44624,
+      "grad_norm": 2.015995049740044,
+      "learning_rate": 0.00011085246375148536,
+      "loss": 1.3784,
+      "step": 9039
+    },
+    {
+      "epoch": 1.4464000000000001,
+      "grad_norm": 2.026305391765431,
+      "learning_rate": 0.00011083529177653833,
+      "loss": 1.3216,
+      "step": 9040
+    },
+    {
+      "epoch": 1.44656,
+      "grad_norm": 2.092651090550723,
+      "learning_rate": 0.000110818119478282,
+      "loss": 1.3424,
+      "step": 9041
+    },
+    {
+      "epoch": 1.44672,
+      "grad_norm": 2.202373887110044,
+      "learning_rate": 0.00011080094685722875,
+      "loss": 1.4167,
+      "step": 9042
+    },
+    {
+      "epoch": 1.44688,
+      "grad_norm": 2.2342700488522564,
+      "learning_rate": 0.000110783773913891,
+      "loss": 1.3726,
+      "step": 9043
+    },
+    {
+      "epoch": 1.4470399999999999,
+      "grad_norm": 1.8193262964391594,
+      "learning_rate": 0.00011076660064878112,
+      "loss": 1.2941,
+      "step": 9044
+    },
+    {
+      "epoch": 1.4472,
+      "grad_norm": 2.4280712888747256,
+      "learning_rate": 0.00011074942706241158,
+      "loss": 1.3457,
+      "step": 9045
+    },
+    {
+      "epoch": 1.44736,
+      "grad_norm": 2.1441754796805257,
+      "learning_rate": 0.0001107322531552948,
+      "loss": 1.3054,
+      "step": 9046
+    },
+    {
+      "epoch": 1.44752,
+      "grad_norm": 2.2151875366809772,
+      "learning_rate": 0.0001107150789279432,
+      "loss": 1.4205,
+      "step": 9047
+    },
+    {
+      "epoch": 1.44768,
+      "grad_norm": 2.2987914491833004,
+      "learning_rate": 0.0001106979043808693,
+      "loss": 1.3714,
+      "step": 9048
+    },
+    {
+      "epoch": 1.44784,
+      "grad_norm": 2.4204065383403504,
+      "learning_rate": 0.00011068072951458548,
+      "loss": 1.3531,
+      "step": 9049
+    },
+    {
+      "epoch": 1.448,
+      "grad_norm": 2.1506771377496694,
+      "learning_rate": 0.0001106635543296043,
+      "loss": 1.3885,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4481600000000001,
+      "grad_norm": 2.4937210597983825,
+      "learning_rate": 0.00011064637882643815,
+      "loss": 1.3618,
+      "step": 9051
+    },
+    {
+      "epoch": 1.44832,
+      "grad_norm": 2.2724581386029876,
+      "learning_rate": 0.0001106292030055996,
+      "loss": 1.3664,
+      "step": 9052
+    },
+    {
+      "epoch": 1.44848,
+      "grad_norm": 2.1976554872431695,
+      "learning_rate": 0.00011061202686760112,
+      "loss": 1.2603,
+      "step": 9053
+    },
+    {
+      "epoch": 1.44864,
+      "grad_norm": 2.6537282272012384,
+      "learning_rate": 0.0001105948504129552,
+      "loss": 1.2889,
+      "step": 9054
+    },
+    {
+      "epoch": 1.4487999999999999,
+      "grad_norm": 2.478589833626931,
+      "learning_rate": 0.0001105776736421744,
+      "loss": 1.3842,
+      "step": 9055
+    },
+    {
+      "epoch": 1.44896,
+      "grad_norm": 2.0905983934515593,
+      "learning_rate": 0.00011056049655577123,
+      "loss": 1.3304,
+      "step": 9056
+    },
+    {
+      "epoch": 1.44912,
+      "grad_norm": 4.345087414103366,
+      "learning_rate": 0.00011054331915425821,
+      "loss": 1.3957,
+      "step": 9057
+    },
+    {
+      "epoch": 1.44928,
+      "grad_norm": 2.019927586911027,
+      "learning_rate": 0.00011052614143814794,
+      "loss": 1.3845,
+      "step": 9058
+    },
+    {
+      "epoch": 1.44944,
+      "grad_norm": 2.810575367366931,
+      "learning_rate": 0.00011050896340795296,
+      "loss": 1.3977,
+      "step": 9059
+    },
+    {
+      "epoch": 1.4496,
+      "grad_norm": 2.0334818960867307,
+      "learning_rate": 0.0001104917850641858,
+      "loss": 1.32,
+      "step": 9060
+    },
+    {
+      "epoch": 1.44976,
+      "grad_norm": 2.1662694639232067,
+      "learning_rate": 0.00011047460640735905,
+      "loss": 1.3214,
+      "step": 9061
+    },
+    {
+      "epoch": 1.44992,
+      "grad_norm": 1.9664586911959085,
+      "learning_rate": 0.00011045742743798533,
+      "loss": 1.345,
+      "step": 9062
+    },
+    {
+      "epoch": 1.45008,
+      "grad_norm": 3.105765811058314,
+      "learning_rate": 0.00011044024815657719,
+      "loss": 1.311,
+      "step": 9063
+    },
+    {
+      "epoch": 1.45024,
+      "grad_norm": 2.21650767190607,
+      "learning_rate": 0.00011042306856364728,
+      "loss": 1.3688,
+      "step": 9064
+    },
+    {
+      "epoch": 1.4504000000000001,
+      "grad_norm": 2.3750351764058544,
+      "learning_rate": 0.00011040588865970818,
+      "loss": 1.375,
+      "step": 9065
+    },
+    {
+      "epoch": 1.45056,
+      "grad_norm": 2.2533878038062256,
+      "learning_rate": 0.00011038870844527254,
+      "loss": 1.3147,
+      "step": 9066
+    },
+    {
+      "epoch": 1.45072,
+      "grad_norm": 2.2434639353507455,
+      "learning_rate": 0.00011037152792085297,
+      "loss": 1.3092,
+      "step": 9067
+    },
+    {
+      "epoch": 1.45088,
+      "grad_norm": 1.95831787732361,
+      "learning_rate": 0.00011035434708696212,
+      "loss": 1.3307,
+      "step": 9068
+    },
+    {
+      "epoch": 1.4510399999999999,
+      "grad_norm": 3.0341869018595995,
+      "learning_rate": 0.00011033716594411263,
+      "loss": 1.3998,
+      "step": 9069
+    },
+    {
+      "epoch": 1.4512,
+      "grad_norm": 2.2435247627885215,
+      "learning_rate": 0.0001103199844928172,
+      "loss": 1.3314,
+      "step": 9070
+    },
+    {
+      "epoch": 1.45136,
+      "grad_norm": 2.7202676142761932,
+      "learning_rate": 0.00011030280273358841,
+      "loss": 1.341,
+      "step": 9071
+    },
+    {
+      "epoch": 1.45152,
+      "grad_norm": 3.2037854386947613,
+      "learning_rate": 0.00011028562066693906,
+      "loss": 1.3658,
+      "step": 9072
+    },
+    {
+      "epoch": 1.45168,
+      "grad_norm": 2.242013756346642,
+      "learning_rate": 0.00011026843829338175,
+      "loss": 1.3377,
+      "step": 9073
+    },
+    {
+      "epoch": 1.45184,
+      "grad_norm": 2.051040275761456,
+      "learning_rate": 0.00011025125561342921,
+      "loss": 1.3275,
+      "step": 9074
+    },
+    {
+      "epoch": 1.452,
+      "grad_norm": 3.0134884044161088,
+      "learning_rate": 0.00011023407262759415,
+      "loss": 1.3411,
+      "step": 9075
+    },
+    {
+      "epoch": 1.4521600000000001,
+      "grad_norm": 2.0608989251767085,
+      "learning_rate": 0.00011021688933638927,
+      "loss": 1.3772,
+      "step": 9076
+    },
+    {
+      "epoch": 1.45232,
+      "grad_norm": 2.559589046369143,
+      "learning_rate": 0.0001101997057403273,
+      "loss": 1.3485,
+      "step": 9077
+    },
+    {
+      "epoch": 1.45248,
+      "grad_norm": 2.2065905907183447,
+      "learning_rate": 0.00011018252183992098,
+      "loss": 1.3197,
+      "step": 9078
+    },
+    {
+      "epoch": 1.45264,
+      "grad_norm": 2.0712587421737116,
+      "learning_rate": 0.00011016533763568302,
+      "loss": 1.3796,
+      "step": 9079
+    },
+    {
+      "epoch": 1.4527999999999999,
+      "grad_norm": 2.1999004664177373,
+      "learning_rate": 0.0001101481531281262,
+      "loss": 1.3072,
+      "step": 9080
+    },
+    {
+      "epoch": 1.45296,
+      "grad_norm": 2.015346395970763,
+      "learning_rate": 0.00011013096831776331,
+      "loss": 1.2758,
+      "step": 9081
+    },
+    {
+      "epoch": 1.45312,
+      "grad_norm": 2.5886050312451845,
+      "learning_rate": 0.00011011378320510706,
+      "loss": 1.349,
+      "step": 9082
+    },
+    {
+      "epoch": 1.45328,
+      "grad_norm": 2.091576374457919,
+      "learning_rate": 0.00011009659779067027,
+      "loss": 1.3679,
+      "step": 9083
+    },
+    {
+      "epoch": 1.45344,
+      "grad_norm": 2.2303587012115473,
+      "learning_rate": 0.00011007941207496568,
+      "loss": 1.3234,
+      "step": 9084
+    },
+    {
+      "epoch": 1.4536,
+      "grad_norm": 2.2903722029992206,
+      "learning_rate": 0.00011006222605850616,
+      "loss": 1.4036,
+      "step": 9085
+    },
+    {
+      "epoch": 1.45376,
+      "grad_norm": 2.2559534961889187,
+      "learning_rate": 0.00011004503974180447,
+      "loss": 1.422,
+      "step": 9086
+    },
+    {
+      "epoch": 1.45392,
+      "grad_norm": 2.1170158486553383,
+      "learning_rate": 0.00011002785312537343,
+      "loss": 1.4422,
+      "step": 9087
+    },
+    {
+      "epoch": 1.45408,
+      "grad_norm": 2.1633807323587195,
+      "learning_rate": 0.00011001066620972587,
+      "loss": 1.3013,
+      "step": 9088
+    },
+    {
+      "epoch": 1.45424,
+      "grad_norm": 2.1373031229436807,
+      "learning_rate": 0.00010999347899537458,
+      "loss": 1.3162,
+      "step": 9089
+    },
+    {
+      "epoch": 1.4544000000000001,
+      "grad_norm": 2.023824649370338,
+      "learning_rate": 0.00010997629148283248,
+      "loss": 1.3398,
+      "step": 9090
+    },
+    {
+      "epoch": 1.45456,
+      "grad_norm": 1.7317399290361415,
+      "learning_rate": 0.00010995910367261237,
+      "loss": 1.306,
+      "step": 9091
+    },
+    {
+      "epoch": 1.45472,
+      "grad_norm": 2.368658096801512,
+      "learning_rate": 0.00010994191556522713,
+      "loss": 1.435,
+      "step": 9092
+    },
+    {
+      "epoch": 1.45488,
+      "grad_norm": 2.1584537461321287,
+      "learning_rate": 0.0001099247271611896,
+      "loss": 1.3152,
+      "step": 9093
+    },
+    {
+      "epoch": 1.45504,
+      "grad_norm": 2.116769571301139,
+      "learning_rate": 0.00010990753846101267,
+      "loss": 1.3604,
+      "step": 9094
+    },
+    {
+      "epoch": 1.4552,
+      "grad_norm": 1.9073869877048757,
+      "learning_rate": 0.00010989034946520925,
+      "loss": 1.3561,
+      "step": 9095
+    },
+    {
+      "epoch": 1.45536,
+      "grad_norm": 2.2985299646172805,
+      "learning_rate": 0.00010987316017429219,
+      "loss": 1.3237,
+      "step": 9096
+    },
+    {
+      "epoch": 1.45552,
+      "grad_norm": 2.1962663468642356,
+      "learning_rate": 0.00010985597058877442,
+      "loss": 1.2997,
+      "step": 9097
+    },
+    {
+      "epoch": 1.45568,
+      "grad_norm": 2.150361525826362,
+      "learning_rate": 0.00010983878070916884,
+      "loss": 1.3303,
+      "step": 9098
+    },
+    {
+      "epoch": 1.45584,
+      "grad_norm": 2.3382140874253388,
+      "learning_rate": 0.00010982159053598841,
+      "loss": 1.3362,
+      "step": 9099
+    },
+    {
+      "epoch": 1.456,
+      "grad_norm": 1.964036363121786,
+      "learning_rate": 0.00010980440006974602,
+      "loss": 1.3245,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4561600000000001,
+      "grad_norm": 1.9608958649121646,
+      "learning_rate": 0.00010978720931095461,
+      "loss": 1.3058,
+      "step": 9101
+    },
+    {
+      "epoch": 1.45632,
+      "grad_norm": 2.418768849598738,
+      "learning_rate": 0.00010977001826012714,
+      "loss": 1.3981,
+      "step": 9102
+    },
+    {
+      "epoch": 1.45648,
+      "grad_norm": 2.1234638285899377,
+      "learning_rate": 0.00010975282691777658,
+      "loss": 1.322,
+      "step": 9103
+    },
+    {
+      "epoch": 1.45664,
+      "grad_norm": 2.462576549353311,
+      "learning_rate": 0.00010973563528441586,
+      "loss": 1.3163,
+      "step": 9104
+    },
+    {
+      "epoch": 1.4567999999999999,
+      "grad_norm": 2.3071821736375315,
+      "learning_rate": 0.00010971844336055794,
+      "loss": 1.3378,
+      "step": 9105
+    },
+    {
+      "epoch": 1.45696,
+      "grad_norm": 2.4067005431723785,
+      "learning_rate": 0.00010970125114671585,
+      "loss": 1.3209,
+      "step": 9106
+    },
+    {
+      "epoch": 1.45712,
+      "grad_norm": 2.0962569795804584,
+      "learning_rate": 0.00010968405864340258,
+      "loss": 1.3283,
+      "step": 9107
+    },
+    {
+      "epoch": 1.45728,
+      "grad_norm": 1.9532881818790564,
+      "learning_rate": 0.00010966686585113111,
+      "loss": 1.3188,
+      "step": 9108
+    },
+    {
+      "epoch": 1.45744,
+      "grad_norm": 2.1510884298748,
+      "learning_rate": 0.00010964967277041444,
+      "loss": 1.3575,
+      "step": 9109
+    },
+    {
+      "epoch": 1.4576,
+      "grad_norm": 2.4208052071398125,
+      "learning_rate": 0.0001096324794017656,
+      "loss": 1.3282,
+      "step": 9110
+    },
+    {
+      "epoch": 1.45776,
+      "grad_norm": 2.1167663832230614,
+      "learning_rate": 0.00010961528574569761,
+      "loss": 1.3763,
+      "step": 9111
+    },
+    {
+      "epoch": 1.45792,
+      "grad_norm": 2.000539364127927,
+      "learning_rate": 0.00010959809180272351,
+      "loss": 1.3068,
+      "step": 9112
+    },
+    {
+      "epoch": 1.45808,
+      "grad_norm": 2.0520858006590603,
+      "learning_rate": 0.0001095808975733563,
+      "loss": 1.2966,
+      "step": 9113
+    },
+    {
+      "epoch": 1.45824,
+      "grad_norm": 1.9830185304869115,
+      "learning_rate": 0.00010956370305810911,
+      "loss": 1.3456,
+      "step": 9114
+    },
+    {
+      "epoch": 1.4584,
+      "grad_norm": 2.026560209679603,
+      "learning_rate": 0.00010954650825749493,
+      "loss": 1.3109,
+      "step": 9115
+    },
+    {
+      "epoch": 1.45856,
+      "grad_norm": 1.8101224451044444,
+      "learning_rate": 0.00010952931317202689,
+      "loss": 1.3284,
+      "step": 9116
+    },
+    {
+      "epoch": 1.45872,
+      "grad_norm": 2.01288488632475,
+      "learning_rate": 0.00010951211780221799,
+      "loss": 1.3441,
+      "step": 9117
+    },
+    {
+      "epoch": 1.45888,
+      "grad_norm": 2.3537794513042893,
+      "learning_rate": 0.00010949492214858139,
+      "loss": 1.3789,
+      "step": 9118
+    },
+    {
+      "epoch": 1.45904,
+      "grad_norm": 2.0311153966537185,
+      "learning_rate": 0.00010947772621163013,
+      "loss": 1.2964,
+      "step": 9119
+    },
+    {
+      "epoch": 1.4592,
+      "grad_norm": 2.211652984142663,
+      "learning_rate": 0.00010946052999187733,
+      "loss": 1.3741,
+      "step": 9120
+    },
+    {
+      "epoch": 1.45936,
+      "grad_norm": 2.241768888375975,
+      "learning_rate": 0.00010944333348983609,
+      "loss": 1.3285,
+      "step": 9121
+    },
+    {
+      "epoch": 1.45952,
+      "grad_norm": 2.3335962983061984,
+      "learning_rate": 0.00010942613670601956,
+      "loss": 1.3111,
+      "step": 9122
+    },
+    {
+      "epoch": 1.45968,
+      "grad_norm": 2.3467879503225904,
+      "learning_rate": 0.00010940893964094083,
+      "loss": 1.3594,
+      "step": 9123
+    },
+    {
+      "epoch": 1.45984,
+      "grad_norm": 2.4083935025053953,
+      "learning_rate": 0.00010939174229511307,
+      "loss": 1.3852,
+      "step": 9124
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 2.1827214685752727,
+      "learning_rate": 0.00010937454466904938,
+      "loss": 1.3124,
+      "step": 9125
+    },
+    {
+      "epoch": 1.4601600000000001,
+      "grad_norm": 2.4185858011598174,
+      "learning_rate": 0.00010935734676326295,
+      "loss": 1.324,
+      "step": 9126
+    },
+    {
+      "epoch": 1.46032,
+      "grad_norm": 2.833647344105225,
+      "learning_rate": 0.00010934014857826694,
+      "loss": 1.3224,
+      "step": 9127
+    },
+    {
+      "epoch": 1.46048,
+      "grad_norm": 2.560461681962002,
+      "learning_rate": 0.0001093229501145745,
+      "loss": 1.3766,
+      "step": 9128
+    },
+    {
+      "epoch": 1.46064,
+      "grad_norm": 2.1700719898593195,
+      "learning_rate": 0.00010930575137269879,
+      "loss": 1.337,
+      "step": 9129
+    },
+    {
+      "epoch": 1.4607999999999999,
+      "grad_norm": 2.2437748764240575,
+      "learning_rate": 0.00010928855235315303,
+      "loss": 1.3842,
+      "step": 9130
+    },
+    {
+      "epoch": 1.46096,
+      "grad_norm": 2.1792357076788442,
+      "learning_rate": 0.00010927135305645043,
+      "loss": 1.3926,
+      "step": 9131
+    },
+    {
+      "epoch": 1.46112,
+      "grad_norm": 2.463598762555451,
+      "learning_rate": 0.00010925415348310412,
+      "loss": 1.3738,
+      "step": 9132
+    },
+    {
+      "epoch": 1.46128,
+      "grad_norm": 2.3745170768681363,
+      "learning_rate": 0.00010923695363362738,
+      "loss": 1.3474,
+      "step": 9133
+    },
+    {
+      "epoch": 1.46144,
+      "grad_norm": 2.2527087573221785,
+      "learning_rate": 0.00010921975350853341,
+      "loss": 1.3397,
+      "step": 9134
+    },
+    {
+      "epoch": 1.4616,
+      "grad_norm": 2.957926605871652,
+      "learning_rate": 0.00010920255310833542,
+      "loss": 1.3567,
+      "step": 9135
+    },
+    {
+      "epoch": 1.46176,
+      "grad_norm": 2.0750977002495543,
+      "learning_rate": 0.00010918535243354667,
+      "loss": 1.3318,
+      "step": 9136
+    },
+    {
+      "epoch": 1.46192,
+      "grad_norm": 2.3676639438842333,
+      "learning_rate": 0.00010916815148468036,
+      "loss": 1.3888,
+      "step": 9137
+    },
+    {
+      "epoch": 1.46208,
+      "grad_norm": 2.9720118868195433,
+      "learning_rate": 0.00010915095026224977,
+      "loss": 1.4583,
+      "step": 9138
+    },
+    {
+      "epoch": 1.46224,
+      "grad_norm": 2.3918159264132917,
+      "learning_rate": 0.00010913374876676818,
+      "loss": 1.4396,
+      "step": 9139
+    },
+    {
+      "epoch": 1.4624,
+      "grad_norm": 1.8750921668339167,
+      "learning_rate": 0.00010911654699874884,
+      "loss": 1.3355,
+      "step": 9140
+    },
+    {
+      "epoch": 1.46256,
+      "grad_norm": 2.9996274153052136,
+      "learning_rate": 0.00010909934495870501,
+      "loss": 1.3121,
+      "step": 9141
+    },
+    {
+      "epoch": 1.46272,
+      "grad_norm": 2.8879876604695633,
+      "learning_rate": 0.00010908214264715,
+      "loss": 1.378,
+      "step": 9142
+    },
+    {
+      "epoch": 1.46288,
+      "grad_norm": 2.162754964506431,
+      "learning_rate": 0.00010906494006459705,
+      "loss": 1.3842,
+      "step": 9143
+    },
+    {
+      "epoch": 1.46304,
+      "grad_norm": 2.5277084809272834,
+      "learning_rate": 0.00010904773721155951,
+      "loss": 1.3583,
+      "step": 9144
+    },
+    {
+      "epoch": 1.4632,
+      "grad_norm": 2.3438378435799434,
+      "learning_rate": 0.00010903053408855071,
+      "loss": 1.3517,
+      "step": 9145
+    },
+    {
+      "epoch": 1.46336,
+      "grad_norm": 2.180612595884427,
+      "learning_rate": 0.00010901333069608389,
+      "loss": 1.3714,
+      "step": 9146
+    },
+    {
+      "epoch": 1.46352,
+      "grad_norm": 2.9603155143135274,
+      "learning_rate": 0.00010899612703467244,
+      "loss": 1.371,
+      "step": 9147
+    },
+    {
+      "epoch": 1.46368,
+      "grad_norm": 2.430052913276496,
+      "learning_rate": 0.00010897892310482965,
+      "loss": 1.3378,
+      "step": 9148
+    },
+    {
+      "epoch": 1.46384,
+      "grad_norm": 2.0424262644174718,
+      "learning_rate": 0.0001089617189070689,
+      "loss": 1.2861,
+      "step": 9149
+    },
+    {
+      "epoch": 1.464,
+      "grad_norm": 2.673481301741236,
+      "learning_rate": 0.0001089445144419035,
+      "loss": 1.2848,
+      "step": 9150
+    },
+    {
+      "epoch": 1.4641600000000001,
+      "grad_norm": 2.89978890084509,
+      "learning_rate": 0.00010892730970984683,
+      "loss": 1.2812,
+      "step": 9151
+    },
+    {
+      "epoch": 1.46432,
+      "grad_norm": 2.278742563401343,
+      "learning_rate": 0.00010891010471141226,
+      "loss": 1.3463,
+      "step": 9152
+    },
+    {
+      "epoch": 1.46448,
+      "grad_norm": 2.9603909333275302,
+      "learning_rate": 0.00010889289944711314,
+      "loss": 1.4025,
+      "step": 9153
+    },
+    {
+      "epoch": 1.46464,
+      "grad_norm": 2.5814865593328182,
+      "learning_rate": 0.00010887569391746281,
+      "loss": 1.4141,
+      "step": 9154
+    },
+    {
+      "epoch": 1.4647999999999999,
+      "grad_norm": 2.428615099177324,
+      "learning_rate": 0.00010885848812297477,
+      "loss": 1.4082,
+      "step": 9155
+    },
+    {
+      "epoch": 1.46496,
+      "grad_norm": 3.667107290500879,
+      "learning_rate": 0.00010884128206416233,
+      "loss": 1.4427,
+      "step": 9156
+    },
+    {
+      "epoch": 1.46512,
+      "grad_norm": 2.3460083699305145,
+      "learning_rate": 0.00010882407574153892,
+      "loss": 1.3527,
+      "step": 9157
+    },
+    {
+      "epoch": 1.46528,
+      "grad_norm": 2.143267243999209,
+      "learning_rate": 0.00010880686915561795,
+      "loss": 1.4253,
+      "step": 9158
+    },
+    {
+      "epoch": 1.46544,
+      "grad_norm": 2.7894893653480746,
+      "learning_rate": 0.00010878966230691284,
+      "loss": 1.3419,
+      "step": 9159
+    },
+    {
+      "epoch": 1.4656,
+      "grad_norm": 2.2464390672083017,
+      "learning_rate": 0.00010877245519593702,
+      "loss": 1.3359,
+      "step": 9160
+    },
+    {
+      "epoch": 1.46576,
+      "grad_norm": 2.2151551891656998,
+      "learning_rate": 0.0001087552478232039,
+      "loss": 1.3877,
+      "step": 9161
+    },
+    {
+      "epoch": 1.4659200000000001,
+      "grad_norm": 2.614786404867499,
+      "learning_rate": 0.00010873804018922693,
+      "loss": 1.3173,
+      "step": 9162
+    },
+    {
+      "epoch": 1.46608,
+      "grad_norm": 2.1630839832000537,
+      "learning_rate": 0.00010872083229451959,
+      "loss": 1.3228,
+      "step": 9163
+    },
+    {
+      "epoch": 1.46624,
+      "grad_norm": 2.1272174446739105,
+      "learning_rate": 0.00010870362413959535,
+      "loss": 1.2905,
+      "step": 9164
+    },
+    {
+      "epoch": 1.4664,
+      "grad_norm": 2.17562035785245,
+      "learning_rate": 0.00010868641572496763,
+      "loss": 1.3423,
+      "step": 9165
+    },
+    {
+      "epoch": 1.4665599999999999,
+      "grad_norm": 2.273337735957848,
+      "learning_rate": 0.0001086692070511499,
+      "loss": 1.4121,
+      "step": 9166
+    },
+    {
+      "epoch": 1.46672,
+      "grad_norm": 1.9958482109802784,
+      "learning_rate": 0.00010865199811865569,
+      "loss": 1.375,
+      "step": 9167
+    },
+    {
+      "epoch": 1.46688,
+      "grad_norm": 1.9379219483672612,
+      "learning_rate": 0.00010863478892799845,
+      "loss": 1.3349,
+      "step": 9168
+    },
+    {
+      "epoch": 1.46704,
+      "grad_norm": 1.8359160174409044,
+      "learning_rate": 0.00010861757947969171,
+      "loss": 1.3004,
+      "step": 9169
+    },
+    {
+      "epoch": 1.4672,
+      "grad_norm": 2.5983513498251343,
+      "learning_rate": 0.00010860036977424892,
+      "loss": 1.3668,
+      "step": 9170
+    },
+    {
+      "epoch": 1.46736,
+      "grad_norm": 2.078557900432423,
+      "learning_rate": 0.00010858315981218366,
+      "loss": 1.3498,
+      "step": 9171
+    },
+    {
+      "epoch": 1.46752,
+      "grad_norm": 2.296906558516605,
+      "learning_rate": 0.00010856594959400941,
+      "loss": 1.4073,
+      "step": 9172
+    },
+    {
+      "epoch": 1.46768,
+      "grad_norm": 2.2842283820517784,
+      "learning_rate": 0.00010854873912023971,
+      "loss": 1.3178,
+      "step": 9173
+    },
+    {
+      "epoch": 1.46784,
+      "grad_norm": 2.242711115586151,
+      "learning_rate": 0.00010853152839138807,
+      "loss": 1.3348,
+      "step": 9174
+    },
+    {
+      "epoch": 1.468,
+      "grad_norm": 2.975963526612128,
+      "learning_rate": 0.00010851431740796807,
+      "loss": 1.3789,
+      "step": 9175
+    },
+    {
+      "epoch": 1.4681600000000001,
+      "grad_norm": 1.9055023487299303,
+      "learning_rate": 0.00010849710617049324,
+      "loss": 1.363,
+      "step": 9176
+    },
+    {
+      "epoch": 1.46832,
+      "grad_norm": 2.205619506777857,
+      "learning_rate": 0.00010847989467947717,
+      "loss": 1.4416,
+      "step": 9177
+    },
+    {
+      "epoch": 1.46848,
+      "grad_norm": 2.450202281244698,
+      "learning_rate": 0.00010846268293543336,
+      "loss": 1.3635,
+      "step": 9178
+    },
+    {
+      "epoch": 1.46864,
+      "grad_norm": 2.359303272565032,
+      "learning_rate": 0.00010844547093887543,
+      "loss": 1.3214,
+      "step": 9179
+    },
+    {
+      "epoch": 1.4687999999999999,
+      "grad_norm": 1.8896908629630214,
+      "learning_rate": 0.00010842825869031697,
+      "loss": 1.3155,
+      "step": 9180
+    },
+    {
+      "epoch": 1.46896,
+      "grad_norm": 3.225576818283234,
+      "learning_rate": 0.00010841104619027155,
+      "loss": 1.4017,
+      "step": 9181
+    },
+    {
+      "epoch": 1.46912,
+      "grad_norm": 2.11648150646972,
+      "learning_rate": 0.00010839383343925276,
+      "loss": 1.2706,
+      "step": 9182
+    },
+    {
+      "epoch": 1.46928,
+      "grad_norm": 2.1126637571986917,
+      "learning_rate": 0.00010837662043777422,
+      "loss": 1.415,
+      "step": 9183
+    },
+    {
+      "epoch": 1.46944,
+      "grad_norm": 2.4249829013713966,
+      "learning_rate": 0.00010835940718634954,
+      "loss": 1.3405,
+      "step": 9184
+    },
+    {
+      "epoch": 1.4696,
+      "grad_norm": 2.168320841526641,
+      "learning_rate": 0.00010834219368549233,
+      "loss": 1.428,
+      "step": 9185
+    },
+    {
+      "epoch": 1.46976,
+      "grad_norm": 2.007899677510426,
+      "learning_rate": 0.00010832497993571618,
+      "loss": 1.3503,
+      "step": 9186
+    },
+    {
+      "epoch": 1.4699200000000001,
+      "grad_norm": 3.0990567142800973,
+      "learning_rate": 0.00010830776593753478,
+      "loss": 1.3099,
+      "step": 9187
+    },
+    {
+      "epoch": 1.47008,
+      "grad_norm": 2.035002365703692,
+      "learning_rate": 0.00010829055169146174,
+      "loss": 1.3892,
+      "step": 9188
+    },
+    {
+      "epoch": 1.47024,
+      "grad_norm": 2.0718442166308413,
+      "learning_rate": 0.00010827333719801074,
+      "loss": 1.2849,
+      "step": 9189
+    },
+    {
+      "epoch": 1.4704,
+      "grad_norm": 1.8856685651049456,
+      "learning_rate": 0.00010825612245769541,
+      "loss": 1.2902,
+      "step": 9190
+    },
+    {
+      "epoch": 1.4705599999999999,
+      "grad_norm": 1.9541615804917478,
+      "learning_rate": 0.00010823890747102941,
+      "loss": 1.3141,
+      "step": 9191
+    },
+    {
+      "epoch": 1.47072,
+      "grad_norm": 2.1670527724241544,
+      "learning_rate": 0.00010822169223852644,
+      "loss": 1.3233,
+      "step": 9192
+    },
+    {
+      "epoch": 1.47088,
+      "grad_norm": 2.026721821910129,
+      "learning_rate": 0.00010820447676070012,
+      "loss": 1.3423,
+      "step": 9193
+    },
+    {
+      "epoch": 1.47104,
+      "grad_norm": 1.917902326999671,
+      "learning_rate": 0.00010818726103806416,
+      "loss": 1.3584,
+      "step": 9194
+    },
+    {
+      "epoch": 1.4712,
+      "grad_norm": 2.0076190204150537,
+      "learning_rate": 0.00010817004507113228,
+      "loss": 1.3306,
+      "step": 9195
+    },
+    {
+      "epoch": 1.47136,
+      "grad_norm": 2.3043185001494737,
+      "learning_rate": 0.00010815282886041816,
+      "loss": 1.3641,
+      "step": 9196
+    },
+    {
+      "epoch": 1.47152,
+      "grad_norm": 2.165184835363473,
+      "learning_rate": 0.0001081356124064355,
+      "loss": 1.3802,
+      "step": 9197
+    },
+    {
+      "epoch": 1.47168,
+      "grad_norm": 1.9251685927329896,
+      "learning_rate": 0.00010811839570969804,
+      "loss": 1.3251,
+      "step": 9198
+    },
+    {
+      "epoch": 1.47184,
+      "grad_norm": 1.8966068480782707,
+      "learning_rate": 0.00010810117877071947,
+      "loss": 1.3332,
+      "step": 9199
+    },
+    {
+      "epoch": 1.472,
+      "grad_norm": 2.095212540191039,
+      "learning_rate": 0.00010808396159001355,
+      "loss": 1.342,
+      "step": 9200
+    },
+    {
+      "epoch": 1.4721600000000001,
+      "grad_norm": 2.009207930965022,
+      "learning_rate": 0.00010806674416809395,
+      "loss": 1.3267,
+      "step": 9201
+    },
+    {
+      "epoch": 1.47232,
+      "grad_norm": 2.029177814004488,
+      "learning_rate": 0.00010804952650547448,
+      "loss": 1.3284,
+      "step": 9202
+    },
+    {
+      "epoch": 1.47248,
+      "grad_norm": 1.8804230081681927,
+      "learning_rate": 0.00010803230860266886,
+      "loss": 1.3447,
+      "step": 9203
+    },
+    {
+      "epoch": 1.47264,
+      "grad_norm": 1.9515483962946498,
+      "learning_rate": 0.00010801509046019087,
+      "loss": 1.3155,
+      "step": 9204
+    },
+    {
+      "epoch": 1.4727999999999999,
+      "grad_norm": 2.180940508469754,
+      "learning_rate": 0.00010799787207855426,
+      "loss": 1.2843,
+      "step": 9205
+    },
+    {
+      "epoch": 1.47296,
+      "grad_norm": 1.8055090842724362,
+      "learning_rate": 0.0001079806534582728,
+      "loss": 1.4012,
+      "step": 9206
+    },
+    {
+      "epoch": 1.47312,
+      "grad_norm": 2.4331067358896337,
+      "learning_rate": 0.00010796343459986027,
+      "loss": 1.4312,
+      "step": 9207
+    },
+    {
+      "epoch": 1.47328,
+      "grad_norm": 2.110689658079541,
+      "learning_rate": 0.00010794621550383046,
+      "loss": 1.3662,
+      "step": 9208
+    },
+    {
+      "epoch": 1.47344,
+      "grad_norm": 1.7818770974857316,
+      "learning_rate": 0.00010792899617069716,
+      "loss": 1.3427,
+      "step": 9209
+    },
+    {
+      "epoch": 1.4736,
+      "grad_norm": 2.676858488907884,
+      "learning_rate": 0.00010791177660097415,
+      "loss": 1.3311,
+      "step": 9210
+    },
+    {
+      "epoch": 1.47376,
+      "grad_norm": 2.019678291935645,
+      "learning_rate": 0.00010789455679517527,
+      "loss": 1.2891,
+      "step": 9211
+    },
+    {
+      "epoch": 1.4739200000000001,
+      "grad_norm": 2.282979512827345,
+      "learning_rate": 0.00010787733675381431,
+      "loss": 1.3468,
+      "step": 9212
+    },
+    {
+      "epoch": 1.47408,
+      "grad_norm": 2.2436747650828193,
+      "learning_rate": 0.0001078601164774051,
+      "loss": 1.3756,
+      "step": 9213
+    },
+    {
+      "epoch": 1.47424,
+      "grad_norm": 2.201325391563732,
+      "learning_rate": 0.00010784289596646147,
+      "loss": 1.3034,
+      "step": 9214
+    },
+    {
+      "epoch": 1.4744,
+      "grad_norm": 2.0167947561816213,
+      "learning_rate": 0.00010782567522149726,
+      "loss": 1.3159,
+      "step": 9215
+    },
+    {
+      "epoch": 1.4745599999999999,
+      "grad_norm": 2.342023003483174,
+      "learning_rate": 0.0001078084542430263,
+      "loss": 1.3412,
+      "step": 9216
+    },
+    {
+      "epoch": 1.47472,
+      "grad_norm": 2.2474983833692086,
+      "learning_rate": 0.00010779123303156244,
+      "loss": 1.3359,
+      "step": 9217
+    },
+    {
+      "epoch": 1.47488,
+      "grad_norm": 3.026120596202229,
+      "learning_rate": 0.00010777401158761955,
+      "loss": 1.3879,
+      "step": 9218
+    },
+    {
+      "epoch": 1.47504,
+      "grad_norm": 2.25713809362002,
+      "learning_rate": 0.00010775678991171146,
+      "loss": 1.4416,
+      "step": 9219
+    },
+    {
+      "epoch": 1.4752,
+      "grad_norm": 2.2007663724912576,
+      "learning_rate": 0.00010773956800435208,
+      "loss": 1.3532,
+      "step": 9220
+    },
+    {
+      "epoch": 1.47536,
+      "grad_norm": 2.069605965701038,
+      "learning_rate": 0.00010772234586605527,
+      "loss": 1.2887,
+      "step": 9221
+    },
+    {
+      "epoch": 1.47552,
+      "grad_norm": 2.1026507718698833,
+      "learning_rate": 0.0001077051234973349,
+      "loss": 1.3827,
+      "step": 9222
+    },
+    {
+      "epoch": 1.47568,
+      "grad_norm": 2.2215599471812646,
+      "learning_rate": 0.00010768790089870488,
+      "loss": 1.3242,
+      "step": 9223
+    },
+    {
+      "epoch": 1.47584,
+      "grad_norm": 2.0172443745738304,
+      "learning_rate": 0.0001076706780706791,
+      "loss": 1.367,
+      "step": 9224
+    },
+    {
+      "epoch": 1.476,
+      "grad_norm": 2.0548149412175944,
+      "learning_rate": 0.00010765345501377146,
+      "loss": 1.2875,
+      "step": 9225
+    },
+    {
+      "epoch": 1.4761600000000001,
+      "grad_norm": 2.21089998650866,
+      "learning_rate": 0.00010763623172849589,
+      "loss": 1.4814,
+      "step": 9226
+    },
+    {
+      "epoch": 1.47632,
+      "grad_norm": 1.9847564645136249,
+      "learning_rate": 0.00010761900821536625,
+      "loss": 1.36,
+      "step": 9227
+    },
+    {
+      "epoch": 1.47648,
+      "grad_norm": 2.73663379723976,
+      "learning_rate": 0.00010760178447489654,
+      "loss": 1.3835,
+      "step": 9228
+    },
+    {
+      "epoch": 1.47664,
+      "grad_norm": 1.915039522619839,
+      "learning_rate": 0.00010758456050760068,
+      "loss": 1.3693,
+      "step": 9229
+    },
+    {
+      "epoch": 1.4768,
+      "grad_norm": 2.306387747290432,
+      "learning_rate": 0.00010756733631399256,
+      "loss": 1.3435,
+      "step": 9230
+    },
+    {
+      "epoch": 1.47696,
+      "grad_norm": 2.020676234875487,
+      "learning_rate": 0.00010755011189458615,
+      "loss": 1.3635,
+      "step": 9231
+    },
+    {
+      "epoch": 1.47712,
+      "grad_norm": 2.101379029462507,
+      "learning_rate": 0.00010753288724989541,
+      "loss": 1.3494,
+      "step": 9232
+    },
+    {
+      "epoch": 1.47728,
+      "grad_norm": 1.8565102314835513,
+      "learning_rate": 0.00010751566238043429,
+      "loss": 1.3762,
+      "step": 9233
+    },
+    {
+      "epoch": 1.47744,
+      "grad_norm": 1.8858604745939886,
+      "learning_rate": 0.00010749843728671676,
+      "loss": 1.3921,
+      "step": 9234
+    },
+    {
+      "epoch": 1.4776,
+      "grad_norm": 2.0866166025202584,
+      "learning_rate": 0.00010748121196925676,
+      "loss": 1.33,
+      "step": 9235
+    },
+    {
+      "epoch": 1.47776,
+      "grad_norm": 2.1788241467286893,
+      "learning_rate": 0.00010746398642856832,
+      "loss": 1.2747,
+      "step": 9236
+    },
+    {
+      "epoch": 1.4779200000000001,
+      "grad_norm": 1.9301817489353124,
+      "learning_rate": 0.00010744676066516539,
+      "loss": 1.2828,
+      "step": 9237
+    },
+    {
+      "epoch": 1.47808,
+      "grad_norm": 1.9465518485392976,
+      "learning_rate": 0.00010742953467956198,
+      "loss": 1.3533,
+      "step": 9238
+    },
+    {
+      "epoch": 1.47824,
+      "grad_norm": 2.3514840657369267,
+      "learning_rate": 0.00010741230847227208,
+      "loss": 1.3862,
+      "step": 9239
+    },
+    {
+      "epoch": 1.4784,
+      "grad_norm": 1.7429737507215008,
+      "learning_rate": 0.0001073950820438097,
+      "loss": 1.2866,
+      "step": 9240
+    },
+    {
+      "epoch": 1.4785599999999999,
+      "grad_norm": 1.9592540531558915,
+      "learning_rate": 0.00010737785539468885,
+      "loss": 1.3093,
+      "step": 9241
+    },
+    {
+      "epoch": 1.47872,
+      "grad_norm": 1.9956823033165543,
+      "learning_rate": 0.00010736062852542355,
+      "loss": 1.3329,
+      "step": 9242
+    },
+    {
+      "epoch": 1.47888,
+      "grad_norm": 2.100387388771799,
+      "learning_rate": 0.0001073434014365278,
+      "loss": 1.3979,
+      "step": 9243
+    },
+    {
+      "epoch": 1.47904,
+      "grad_norm": 1.9850629578221184,
+      "learning_rate": 0.00010732617412851567,
+      "loss": 1.3612,
+      "step": 9244
+    },
+    {
+      "epoch": 1.4792,
+      "grad_norm": 2.2087177631580794,
+      "learning_rate": 0.00010730894660190116,
+      "loss": 1.3518,
+      "step": 9245
+    },
+    {
+      "epoch": 1.47936,
+      "grad_norm": 2.176019365725934,
+      "learning_rate": 0.00010729171885719835,
+      "loss": 1.3178,
+      "step": 9246
+    },
+    {
+      "epoch": 1.47952,
+      "grad_norm": 2.217992248492873,
+      "learning_rate": 0.00010727449089492127,
+      "loss": 1.3416,
+      "step": 9247
+    },
+    {
+      "epoch": 1.47968,
+      "grad_norm": 2.005207395806561,
+      "learning_rate": 0.00010725726271558398,
+      "loss": 1.3715,
+      "step": 9248
+    },
+    {
+      "epoch": 1.47984,
+      "grad_norm": 2.2422912555659766,
+      "learning_rate": 0.00010724003431970056,
+      "loss": 1.3682,
+      "step": 9249
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 2.0980627023117306,
+      "learning_rate": 0.00010722280570778505,
+      "loss": 1.4211,
+      "step": 9250
+    },
+    {
+      "epoch": 1.48016,
+      "grad_norm": 2.2019457807618474,
+      "learning_rate": 0.00010720557688035153,
+      "loss": 1.351,
+      "step": 9251
+    },
+    {
+      "epoch": 1.48032,
+      "grad_norm": 2.0000974981985915,
+      "learning_rate": 0.00010718834783791412,
+      "loss": 1.3853,
+      "step": 9252
+    },
+    {
+      "epoch": 1.48048,
+      "grad_norm": 2.018641587492983,
+      "learning_rate": 0.00010717111858098688,
+      "loss": 1.4036,
+      "step": 9253
+    },
+    {
+      "epoch": 1.48064,
+      "grad_norm": 2.596499674235602,
+      "learning_rate": 0.0001071538891100839,
+      "loss": 1.3653,
+      "step": 9254
+    },
+    {
+      "epoch": 1.4808,
+      "grad_norm": 1.8173380254472353,
+      "learning_rate": 0.00010713665942571931,
+      "loss": 1.291,
+      "step": 9255
+    },
+    {
+      "epoch": 1.48096,
+      "grad_norm": 2.218343203500608,
+      "learning_rate": 0.00010711942952840719,
+      "loss": 1.3083,
+      "step": 9256
+    },
+    {
+      "epoch": 1.48112,
+      "grad_norm": 2.128067644974136,
+      "learning_rate": 0.00010710219941866167,
+      "loss": 1.3279,
+      "step": 9257
+    },
+    {
+      "epoch": 1.48128,
+      "grad_norm": 2.683844269642813,
+      "learning_rate": 0.00010708496909699688,
+      "loss": 1.3697,
+      "step": 9258
+    },
+    {
+      "epoch": 1.48144,
+      "grad_norm": 1.9399638594170356,
+      "learning_rate": 0.00010706773856392689,
+      "loss": 1.3564,
+      "step": 9259
+    },
+    {
+      "epoch": 1.4816,
+      "grad_norm": 2.159143530294562,
+      "learning_rate": 0.00010705050781996593,
+      "loss": 1.3937,
+      "step": 9260
+    },
+    {
+      "epoch": 1.48176,
+      "grad_norm": 2.3421505734206907,
+      "learning_rate": 0.00010703327686562806,
+      "loss": 1.3148,
+      "step": 9261
+    },
+    {
+      "epoch": 1.4819200000000001,
+      "grad_norm": 2.0881358799590775,
+      "learning_rate": 0.00010701604570142744,
+      "loss": 1.3438,
+      "step": 9262
+    },
+    {
+      "epoch": 1.48208,
+      "grad_norm": 2.250991639399151,
+      "learning_rate": 0.00010699881432787828,
+      "loss": 1.3761,
+      "step": 9263
+    },
+    {
+      "epoch": 1.48224,
+      "grad_norm": 2.340815944286105,
+      "learning_rate": 0.0001069815827454947,
+      "loss": 1.317,
+      "step": 9264
+    },
+    {
+      "epoch": 1.4824,
+      "grad_norm": 2.2857554234663016,
+      "learning_rate": 0.00010696435095479085,
+      "loss": 1.3267,
+      "step": 9265
+    },
+    {
+      "epoch": 1.4825599999999999,
+      "grad_norm": 1.9292820850997343,
+      "learning_rate": 0.00010694711895628093,
+      "loss": 1.3632,
+      "step": 9266
+    },
+    {
+      "epoch": 1.48272,
+      "grad_norm": 2.16873051083669,
+      "learning_rate": 0.00010692988675047911,
+      "loss": 1.3215,
+      "step": 9267
+    },
+    {
+      "epoch": 1.48288,
+      "grad_norm": 2.290342290904153,
+      "learning_rate": 0.00010691265433789955,
+      "loss": 1.3727,
+      "step": 9268
+    },
+    {
+      "epoch": 1.48304,
+      "grad_norm": 2.1134654256564755,
+      "learning_rate": 0.00010689542171905642,
+      "loss": 1.3983,
+      "step": 9269
+    },
+    {
+      "epoch": 1.4832,
+      "grad_norm": 2.037604681237324,
+      "learning_rate": 0.000106878188894464,
+      "loss": 1.3156,
+      "step": 9270
+    },
+    {
+      "epoch": 1.48336,
+      "grad_norm": 2.190126096364874,
+      "learning_rate": 0.00010686095586463644,
+      "loss": 1.4021,
+      "step": 9271
+    },
+    {
+      "epoch": 1.48352,
+      "grad_norm": 2.018595255218939,
+      "learning_rate": 0.00010684372263008795,
+      "loss": 1.3209,
+      "step": 9272
+    },
+    {
+      "epoch": 1.48368,
+      "grad_norm": 2.2606641948029655,
+      "learning_rate": 0.00010682648919133276,
+      "loss": 1.2674,
+      "step": 9273
+    },
+    {
+      "epoch": 1.48384,
+      "grad_norm": 2.392202213215808,
+      "learning_rate": 0.00010680925554888507,
+      "loss": 1.3482,
+      "step": 9274
+    },
+    {
+      "epoch": 1.484,
+      "grad_norm": 2.0264627825496175,
+      "learning_rate": 0.00010679202170325912,
+      "loss": 1.4021,
+      "step": 9275
+    },
+    {
+      "epoch": 1.48416,
+      "grad_norm": 1.6767496238710167,
+      "learning_rate": 0.00010677478765496914,
+      "loss": 1.313,
+      "step": 9276
+    },
+    {
+      "epoch": 1.4843199999999999,
+      "grad_norm": 1.8994328044445379,
+      "learning_rate": 0.00010675755340452934,
+      "loss": 1.2514,
+      "step": 9277
+    },
+    {
+      "epoch": 1.48448,
+      "grad_norm": 1.6674515532650644,
+      "learning_rate": 0.00010674031895245402,
+      "loss": 1.3661,
+      "step": 9278
+    },
+    {
+      "epoch": 1.48464,
+      "grad_norm": 1.9049275782334314,
+      "learning_rate": 0.00010672308429925741,
+      "loss": 1.3845,
+      "step": 9279
+    },
+    {
+      "epoch": 1.4848,
+      "grad_norm": 2.320046197279997,
+      "learning_rate": 0.00010670584944545374,
+      "loss": 1.3073,
+      "step": 9280
+    },
+    {
+      "epoch": 1.48496,
+      "grad_norm": 2.025820087822876,
+      "learning_rate": 0.00010668861439155733,
+      "loss": 1.3344,
+      "step": 9281
+    },
+    {
+      "epoch": 1.48512,
+      "grad_norm": 1.904430665786156,
+      "learning_rate": 0.00010667137913808239,
+      "loss": 1.3033,
+      "step": 9282
+    },
+    {
+      "epoch": 1.48528,
+      "grad_norm": 2.016704680986312,
+      "learning_rate": 0.00010665414368554321,
+      "loss": 1.3426,
+      "step": 9283
+    },
+    {
+      "epoch": 1.48544,
+      "grad_norm": 1.9760224621971005,
+      "learning_rate": 0.00010663690803445409,
+      "loss": 1.2793,
+      "step": 9284
+    },
+    {
+      "epoch": 1.4856,
+      "grad_norm": 1.9797640085666752,
+      "learning_rate": 0.00010661967218532929,
+      "loss": 1.3167,
+      "step": 9285
+    },
+    {
+      "epoch": 1.48576,
+      "grad_norm": 2.206608772252896,
+      "learning_rate": 0.00010660243613868313,
+      "loss": 1.3609,
+      "step": 9286
+    },
+    {
+      "epoch": 1.4859200000000001,
+      "grad_norm": 2.265967112339692,
+      "learning_rate": 0.00010658519989502989,
+      "loss": 1.351,
+      "step": 9287
+    },
+    {
+      "epoch": 1.48608,
+      "grad_norm": 1.9263050043368652,
+      "learning_rate": 0.00010656796345488389,
+      "loss": 1.2677,
+      "step": 9288
+    },
+    {
+      "epoch": 1.48624,
+      "grad_norm": 2.1910581094399504,
+      "learning_rate": 0.00010655072681875944,
+      "loss": 1.3507,
+      "step": 9289
+    },
+    {
+      "epoch": 1.4864,
+      "grad_norm": 2.1340338535049272,
+      "learning_rate": 0.00010653348998717084,
+      "loss": 1.333,
+      "step": 9290
+    },
+    {
+      "epoch": 1.4865599999999999,
+      "grad_norm": 2.2922867425427804,
+      "learning_rate": 0.00010651625296063242,
+      "loss": 1.334,
+      "step": 9291
+    },
+    {
+      "epoch": 1.48672,
+      "grad_norm": 2.35747631994888,
+      "learning_rate": 0.00010649901573965848,
+      "loss": 1.3281,
+      "step": 9292
+    },
+    {
+      "epoch": 1.48688,
+      "grad_norm": 2.2858914117262965,
+      "learning_rate": 0.00010648177832476338,
+      "loss": 1.3426,
+      "step": 9293
+    },
+    {
+      "epoch": 1.48704,
+      "grad_norm": 2.035917866040706,
+      "learning_rate": 0.0001064645407164615,
+      "loss": 1.4111,
+      "step": 9294
+    },
+    {
+      "epoch": 1.4872,
+      "grad_norm": 2.134319380790155,
+      "learning_rate": 0.00010644730291526712,
+      "loss": 1.3485,
+      "step": 9295
+    },
+    {
+      "epoch": 1.48736,
+      "grad_norm": 2.1378005715070048,
+      "learning_rate": 0.00010643006492169463,
+      "loss": 1.3642,
+      "step": 9296
+    },
+    {
+      "epoch": 1.48752,
+      "grad_norm": 2.3184195287825062,
+      "learning_rate": 0.00010641282673625836,
+      "loss": 1.4076,
+      "step": 9297
+    },
+    {
+      "epoch": 1.4876800000000001,
+      "grad_norm": 1.9972495739053528,
+      "learning_rate": 0.0001063955883594727,
+      "loss": 1.3447,
+      "step": 9298
+    },
+    {
+      "epoch": 1.48784,
+      "grad_norm": 2.1812432668257364,
+      "learning_rate": 0.00010637834979185197,
+      "loss": 1.416,
+      "step": 9299
+    },
+    {
+      "epoch": 1.488,
+      "grad_norm": 2.2112635111596477,
+      "learning_rate": 0.0001063611110339106,
+      "loss": 1.3536,
+      "step": 9300
+    },
+    {
+      "epoch": 1.48816,
+      "grad_norm": 2.0161979270476205,
+      "learning_rate": 0.00010634387208616295,
+      "loss": 1.334,
+      "step": 9301
+    },
+    {
+      "epoch": 1.4883199999999999,
+      "grad_norm": 2.0852501185956736,
+      "learning_rate": 0.0001063266329491234,
+      "loss": 1.3589,
+      "step": 9302
+    },
+    {
+      "epoch": 1.48848,
+      "grad_norm": 2.1940157030469454,
+      "learning_rate": 0.00010630939362330631,
+      "loss": 1.3846,
+      "step": 9303
+    },
+    {
+      "epoch": 1.48864,
+      "grad_norm": 2.2043662805406936,
+      "learning_rate": 0.00010629215410922614,
+      "loss": 1.4016,
+      "step": 9304
+    },
+    {
+      "epoch": 1.4888,
+      "grad_norm": 2.3669974545368495,
+      "learning_rate": 0.00010627491440739725,
+      "loss": 1.3404,
+      "step": 9305
+    },
+    {
+      "epoch": 1.48896,
+      "grad_norm": 2.2409676736390876,
+      "learning_rate": 0.00010625767451833409,
+      "loss": 1.4156,
+      "step": 9306
+    },
+    {
+      "epoch": 1.48912,
+      "grad_norm": 2.197581509931036,
+      "learning_rate": 0.00010624043444255101,
+      "loss": 1.3926,
+      "step": 9307
+    },
+    {
+      "epoch": 1.48928,
+      "grad_norm": 1.9764400811157075,
+      "learning_rate": 0.00010622319418056243,
+      "loss": 1.3169,
+      "step": 9308
+    },
+    {
+      "epoch": 1.48944,
+      "grad_norm": 2.8367784011981767,
+      "learning_rate": 0.00010620595373288285,
+      "loss": 1.3889,
+      "step": 9309
+    },
+    {
+      "epoch": 1.4896,
+      "grad_norm": 2.415028103680356,
+      "learning_rate": 0.00010618871310002663,
+      "loss": 1.3347,
+      "step": 9310
+    },
+    {
+      "epoch": 1.48976,
+      "grad_norm": 2.2180771531154058,
+      "learning_rate": 0.00010617147228250825,
+      "loss": 1.359,
+      "step": 9311
+    },
+    {
+      "epoch": 1.4899200000000001,
+      "grad_norm": 2.283905251178199,
+      "learning_rate": 0.00010615423128084212,
+      "loss": 1.3009,
+      "step": 9312
+    },
+    {
+      "epoch": 1.49008,
+      "grad_norm": 2.191047172560278,
+      "learning_rate": 0.0001061369900955427,
+      "loss": 1.3385,
+      "step": 9313
+    },
+    {
+      "epoch": 1.49024,
+      "grad_norm": 2.0890228500277828,
+      "learning_rate": 0.00010611974872712443,
+      "loss": 1.3466,
+      "step": 9314
+    },
+    {
+      "epoch": 1.4904,
+      "grad_norm": 2.2535256505050385,
+      "learning_rate": 0.00010610250717610177,
+      "loss": 1.3428,
+      "step": 9315
+    },
+    {
+      "epoch": 1.4905599999999999,
+      "grad_norm": 2.233908620541982,
+      "learning_rate": 0.00010608526544298917,
+      "loss": 1.3843,
+      "step": 9316
+    },
+    {
+      "epoch": 1.49072,
+      "grad_norm": 2.0586456397557846,
+      "learning_rate": 0.00010606802352830116,
+      "loss": 1.3941,
+      "step": 9317
+    },
+    {
+      "epoch": 1.49088,
+      "grad_norm": 2.2282009047749667,
+      "learning_rate": 0.00010605078143255213,
+      "loss": 1.361,
+      "step": 9318
+    },
+    {
+      "epoch": 1.49104,
+      "grad_norm": 2.3312335962736404,
+      "learning_rate": 0.00010603353915625662,
+      "loss": 1.3407,
+      "step": 9319
+    },
+    {
+      "epoch": 1.4912,
+      "grad_norm": 2.3035295497044106,
+      "learning_rate": 0.0001060162966999291,
+      "loss": 1.3835,
+      "step": 9320
+    },
+    {
+      "epoch": 1.49136,
+      "grad_norm": 2.705571288232817,
+      "learning_rate": 0.00010599905406408406,
+      "loss": 1.3345,
+      "step": 9321
+    },
+    {
+      "epoch": 1.49152,
+      "grad_norm": 2.2694331689570264,
+      "learning_rate": 0.00010598181124923596,
+      "loss": 1.271,
+      "step": 9322
+    },
+    {
+      "epoch": 1.4916800000000001,
+      "grad_norm": 2.208946260524557,
+      "learning_rate": 0.00010596456825589936,
+      "loss": 1.3524,
+      "step": 9323
+    },
+    {
+      "epoch": 1.49184,
+      "grad_norm": 1.8669288081229605,
+      "learning_rate": 0.00010594732508458872,
+      "loss": 1.3299,
+      "step": 9324
+    },
+    {
+      "epoch": 1.492,
+      "grad_norm": 2.5255684411765054,
+      "learning_rate": 0.00010593008173581858,
+      "loss": 1.2943,
+      "step": 9325
+    },
+    {
+      "epoch": 1.49216,
+      "grad_norm": 2.056581476360406,
+      "learning_rate": 0.0001059128382101034,
+      "loss": 1.3321,
+      "step": 9326
+    },
+    {
+      "epoch": 1.4923199999999999,
+      "grad_norm": 2.555474743813545,
+      "learning_rate": 0.00010589559450795776,
+      "loss": 1.3589,
+      "step": 9327
+    },
+    {
+      "epoch": 1.49248,
+      "grad_norm": 2.191322713480925,
+      "learning_rate": 0.00010587835062989619,
+      "loss": 1.3209,
+      "step": 9328
+    },
+    {
+      "epoch": 1.49264,
+      "grad_norm": 1.9691004120036426,
+      "learning_rate": 0.0001058611065764332,
+      "loss": 1.2756,
+      "step": 9329
+    },
+    {
+      "epoch": 1.4928,
+      "grad_norm": 2.192134935492358,
+      "learning_rate": 0.00010584386234808332,
+      "loss": 1.2777,
+      "step": 9330
+    },
+    {
+      "epoch": 1.49296,
+      "grad_norm": 2.0625544378518126,
+      "learning_rate": 0.0001058266179453611,
+      "loss": 1.2974,
+      "step": 9331
+    },
+    {
+      "epoch": 1.49312,
+      "grad_norm": 2.2634737668164466,
+      "learning_rate": 0.00010580937336878111,
+      "loss": 1.3331,
+      "step": 9332
+    },
+    {
+      "epoch": 1.49328,
+      "grad_norm": 1.9750654320952694,
+      "learning_rate": 0.00010579212861885788,
+      "loss": 1.3746,
+      "step": 9333
+    },
+    {
+      "epoch": 1.49344,
+      "grad_norm": 1.9928017991459175,
+      "learning_rate": 0.00010577488369610595,
+      "loss": 1.2914,
+      "step": 9334
+    },
+    {
+      "epoch": 1.4936,
+      "grad_norm": 2.049056245150514,
+      "learning_rate": 0.00010575763860103991,
+      "loss": 1.2821,
+      "step": 9335
+    },
+    {
+      "epoch": 1.49376,
+      "grad_norm": 2.0902085394551775,
+      "learning_rate": 0.00010574039333417434,
+      "loss": 1.3476,
+      "step": 9336
+    },
+    {
+      "epoch": 1.4939200000000001,
+      "grad_norm": 2.235845217592357,
+      "learning_rate": 0.0001057231478960238,
+      "loss": 1.3038,
+      "step": 9337
+    },
+    {
+      "epoch": 1.49408,
+      "grad_norm": 1.9832686315409587,
+      "learning_rate": 0.00010570590228710287,
+      "loss": 1.335,
+      "step": 9338
+    },
+    {
+      "epoch": 1.49424,
+      "grad_norm": 2.1888332998851405,
+      "learning_rate": 0.0001056886565079261,
+      "loss": 1.3411,
+      "step": 9339
+    },
+    {
+      "epoch": 1.4944,
+      "grad_norm": 2.1953502256774438,
+      "learning_rate": 0.00010567141055900813,
+      "loss": 1.2642,
+      "step": 9340
+    },
+    {
+      "epoch": 1.4945599999999999,
+      "grad_norm": 1.9870557056010179,
+      "learning_rate": 0.00010565416444086353,
+      "loss": 1.3119,
+      "step": 9341
+    },
+    {
+      "epoch": 1.49472,
+      "grad_norm": 1.987435265607743,
+      "learning_rate": 0.00010563691815400689,
+      "loss": 1.3675,
+      "step": 9342
+    },
+    {
+      "epoch": 1.49488,
+      "grad_norm": 2.5413378978387486,
+      "learning_rate": 0.00010561967169895284,
+      "loss": 1.3678,
+      "step": 9343
+    },
+    {
+      "epoch": 1.49504,
+      "grad_norm": 2.083670401485469,
+      "learning_rate": 0.00010560242507621599,
+      "loss": 1.3646,
+      "step": 9344
+    },
+    {
+      "epoch": 1.4952,
+      "grad_norm": 2.1957173756831065,
+      "learning_rate": 0.00010558517828631091,
+      "loss": 1.3052,
+      "step": 9345
+    },
+    {
+      "epoch": 1.49536,
+      "grad_norm": 1.9815596630982046,
+      "learning_rate": 0.00010556793132975227,
+      "loss": 1.3082,
+      "step": 9346
+    },
+    {
+      "epoch": 1.49552,
+      "grad_norm": 1.8299691732446728,
+      "learning_rate": 0.00010555068420705465,
+      "loss": 1.2764,
+      "step": 9347
+    },
+    {
+      "epoch": 1.4956800000000001,
+      "grad_norm": 1.9005272372588236,
+      "learning_rate": 0.00010553343691873271,
+      "loss": 1.2992,
+      "step": 9348
+    },
+    {
+      "epoch": 1.49584,
+      "grad_norm": 2.0071835996748844,
+      "learning_rate": 0.00010551618946530111,
+      "loss": 1.38,
+      "step": 9349
+    },
+    {
+      "epoch": 1.496,
+      "grad_norm": 2.3105699915482285,
+      "learning_rate": 0.0001054989418472744,
+      "loss": 1.3298,
+      "step": 9350
+    },
+    {
+      "epoch": 1.49616,
+      "grad_norm": 1.987869512744095,
+      "learning_rate": 0.00010548169406516731,
+      "loss": 1.3376,
+      "step": 9351
+    },
+    {
+      "epoch": 1.4963199999999999,
+      "grad_norm": 2.3871831632864136,
+      "learning_rate": 0.00010546444611949445,
+      "loss": 1.3933,
+      "step": 9352
+    },
+    {
+      "epoch": 1.49648,
+      "grad_norm": 1.8689965439340572,
+      "learning_rate": 0.00010544719801077048,
+      "loss": 1.284,
+      "step": 9353
+    },
+    {
+      "epoch": 1.49664,
+      "grad_norm": 1.9857056392841932,
+      "learning_rate": 0.00010542994973951006,
+      "loss": 1.2999,
+      "step": 9354
+    },
+    {
+      "epoch": 1.4968,
+      "grad_norm": 2.0224188317909206,
+      "learning_rate": 0.00010541270130622784,
+      "loss": 1.3667,
+      "step": 9355
+    },
+    {
+      "epoch": 1.49696,
+      "grad_norm": 2.1514117845898446,
+      "learning_rate": 0.00010539545271143852,
+      "loss": 1.3469,
+      "step": 9356
+    },
+    {
+      "epoch": 1.49712,
+      "grad_norm": 3.0291872583072177,
+      "learning_rate": 0.00010537820395565675,
+      "loss": 1.4209,
+      "step": 9357
+    },
+    {
+      "epoch": 1.49728,
+      "grad_norm": 2.2045002676805776,
+      "learning_rate": 0.00010536095503939718,
+      "loss": 1.3453,
+      "step": 9358
+    },
+    {
+      "epoch": 1.49744,
+      "grad_norm": 2.344215561558867,
+      "learning_rate": 0.00010534370596317455,
+      "loss": 1.2991,
+      "step": 9359
+    },
+    {
+      "epoch": 1.4976,
+      "grad_norm": 2.7754381272617716,
+      "learning_rate": 0.00010532645672750352,
+      "loss": 1.347,
+      "step": 9360
+    },
+    {
+      "epoch": 1.49776,
+      "grad_norm": 1.9459429321481785,
+      "learning_rate": 0.00010530920733289878,
+      "loss": 1.3628,
+      "step": 9361
+    },
+    {
+      "epoch": 1.49792,
+      "grad_norm": 2.373603990008819,
+      "learning_rate": 0.00010529195777987504,
+      "loss": 1.3289,
+      "step": 9362
+    },
+    {
+      "epoch": 1.49808,
+      "grad_norm": 2.017313241570652,
+      "learning_rate": 0.00010527470806894698,
+      "loss": 1.3432,
+      "step": 9363
+    },
+    {
+      "epoch": 1.49824,
+      "grad_norm": 2.5773380765114595,
+      "learning_rate": 0.00010525745820062933,
+      "loss": 1.3554,
+      "step": 9364
+    },
+    {
+      "epoch": 1.4984,
+      "grad_norm": 2.326871677365062,
+      "learning_rate": 0.00010524020817543676,
+      "loss": 1.3803,
+      "step": 9365
+    },
+    {
+      "epoch": 1.49856,
+      "grad_norm": 2.3417537602743064,
+      "learning_rate": 0.00010522295799388401,
+      "loss": 1.3549,
+      "step": 9366
+    },
+    {
+      "epoch": 1.49872,
+      "grad_norm": 2.439855899216894,
+      "learning_rate": 0.00010520570765648583,
+      "loss": 1.3011,
+      "step": 9367
+    },
+    {
+      "epoch": 1.49888,
+      "grad_norm": 2.0077433557980084,
+      "learning_rate": 0.00010518845716375693,
+      "loss": 1.2936,
+      "step": 9368
+    },
+    {
+      "epoch": 1.49904,
+      "grad_norm": 2.0661047028461836,
+      "learning_rate": 0.00010517120651621202,
+      "loss": 1.267,
+      "step": 9369
+    },
+    {
+      "epoch": 1.4992,
+      "grad_norm": 2.320912421638497,
+      "learning_rate": 0.00010515395571436584,
+      "loss": 1.3381,
+      "step": 9370
+    },
+    {
+      "epoch": 1.49936,
+      "grad_norm": 2.224405805881181,
+      "learning_rate": 0.00010513670475873312,
+      "loss": 1.3635,
+      "step": 9371
+    },
+    {
+      "epoch": 1.49952,
+      "grad_norm": 2.244780606173291,
+      "learning_rate": 0.00010511945364982862,
+      "loss": 1.2948,
+      "step": 9372
+    },
+    {
+      "epoch": 1.4996800000000001,
+      "grad_norm": 2.0605244305298447,
+      "learning_rate": 0.00010510220238816711,
+      "loss": 1.3423,
+      "step": 9373
+    },
+    {
+      "epoch": 1.49984,
+      "grad_norm": 1.9371593057221235,
+      "learning_rate": 0.00010508495097426326,
+      "loss": 1.3192,
+      "step": 9374
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 2.132476979368442,
+      "learning_rate": 0.0001050676994086319,
+      "loss": 1.2872,
+      "step": 9375
+    },
+    {
+      "epoch": 1.5001600000000002,
+      "grad_norm": 1.8193957811628299,
+      "learning_rate": 0.0001050504476917878,
+      "loss": 1.3556,
+      "step": 9376
+    },
+    {
+      "epoch": 1.5003199999999999,
+      "grad_norm": 2.1324276155951534,
+      "learning_rate": 0.00010503319582424566,
+      "loss": 1.3447,
+      "step": 9377
+    },
+    {
+      "epoch": 1.50048,
+      "grad_norm": 2.2476582091361768,
+      "learning_rate": 0.00010501594380652033,
+      "loss": 1.4121,
+      "step": 9378
+    },
+    {
+      "epoch": 1.50064,
+      "grad_norm": 2.7459289648676943,
+      "learning_rate": 0.00010499869163912652,
+      "loss": 1.4695,
+      "step": 9379
+    },
+    {
+      "epoch": 1.5008,
+      "grad_norm": 2.3408026195017806,
+      "learning_rate": 0.00010498143932257906,
+      "loss": 1.3614,
+      "step": 9380
+    },
+    {
+      "epoch": 1.50096,
+      "grad_norm": 2.0834342789749782,
+      "learning_rate": 0.00010496418685739271,
+      "loss": 1.3764,
+      "step": 9381
+    },
+    {
+      "epoch": 1.50112,
+      "grad_norm": 2.236765077734712,
+      "learning_rate": 0.00010494693424408224,
+      "loss": 1.3551,
+      "step": 9382
+    },
+    {
+      "epoch": 1.50128,
+      "grad_norm": 2.4643858806779244,
+      "learning_rate": 0.00010492968148316245,
+      "loss": 1.2962,
+      "step": 9383
+    },
+    {
+      "epoch": 1.50144,
+      "grad_norm": 2.152474623690615,
+      "learning_rate": 0.00010491242857514816,
+      "loss": 1.3364,
+      "step": 9384
+    },
+    {
+      "epoch": 1.5016,
+      "grad_norm": 1.9534519639742154,
+      "learning_rate": 0.00010489517552055416,
+      "loss": 1.2876,
+      "step": 9385
+    },
+    {
+      "epoch": 1.50176,
+      "grad_norm": 1.952117235277611,
+      "learning_rate": 0.00010487792231989528,
+      "loss": 1.2695,
+      "step": 9386
+    },
+    {
+      "epoch": 1.5019200000000001,
+      "grad_norm": 1.9017292599523128,
+      "learning_rate": 0.0001048606689736863,
+      "loss": 1.3295,
+      "step": 9387
+    },
+    {
+      "epoch": 1.5020799999999999,
+      "grad_norm": 2.2702170403345763,
+      "learning_rate": 0.00010484341548244202,
+      "loss": 1.2644,
+      "step": 9388
+    },
+    {
+      "epoch": 1.50224,
+      "grad_norm": 1.9713492751030486,
+      "learning_rate": 0.00010482616184667728,
+      "loss": 1.2937,
+      "step": 9389
+    },
+    {
+      "epoch": 1.5024,
+      "grad_norm": 1.8507043535460532,
+      "learning_rate": 0.00010480890806690691,
+      "loss": 1.3369,
+      "step": 9390
+    },
+    {
+      "epoch": 1.50256,
+      "grad_norm": 2.1589711991430436,
+      "learning_rate": 0.00010479165414364571,
+      "loss": 1.4387,
+      "step": 9391
+    },
+    {
+      "epoch": 1.50272,
+      "grad_norm": 2.0531911489221533,
+      "learning_rate": 0.00010477440007740855,
+      "loss": 1.3249,
+      "step": 9392
+    },
+    {
+      "epoch": 1.50288,
+      "grad_norm": 2.054776124586282,
+      "learning_rate": 0.00010475714586871026,
+      "loss": 1.4076,
+      "step": 9393
+    },
+    {
+      "epoch": 1.50304,
+      "grad_norm": 2.022502272064121,
+      "learning_rate": 0.00010473989151806567,
+      "loss": 1.3427,
+      "step": 9394
+    },
+    {
+      "epoch": 1.5032,
+      "grad_norm": 2.5389143017700477,
+      "learning_rate": 0.00010472263702598961,
+      "loss": 1.3925,
+      "step": 9395
+    },
+    {
+      "epoch": 1.50336,
+      "grad_norm": 1.900597523259774,
+      "learning_rate": 0.00010470538239299694,
+      "loss": 1.3342,
+      "step": 9396
+    },
+    {
+      "epoch": 1.50352,
+      "grad_norm": 2.0810144883100237,
+      "learning_rate": 0.00010468812761960253,
+      "loss": 1.3763,
+      "step": 9397
+    },
+    {
+      "epoch": 1.5036800000000001,
+      "grad_norm": 2.287329763486525,
+      "learning_rate": 0.00010467087270632122,
+      "loss": 1.3524,
+      "step": 9398
+    },
+    {
+      "epoch": 1.5038399999999998,
+      "grad_norm": 2.012834350418238,
+      "learning_rate": 0.00010465361765366785,
+      "loss": 1.3483,
+      "step": 9399
+    },
+    {
+      "epoch": 1.504,
+      "grad_norm": 2.496991093348757,
+      "learning_rate": 0.00010463636246215734,
+      "loss": 1.3725,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5041600000000002,
+      "grad_norm": 2.0402362029982033,
+      "learning_rate": 0.00010461910713230452,
+      "loss": 1.3404,
+      "step": 9401
+    },
+    {
+      "epoch": 1.5043199999999999,
+      "grad_norm": 2.090396874243138,
+      "learning_rate": 0.00010460185166462426,
+      "loss": 1.3221,
+      "step": 9402
+    },
+    {
+      "epoch": 1.50448,
+      "grad_norm": 2.455201749027555,
+      "learning_rate": 0.00010458459605963149,
+      "loss": 1.2586,
+      "step": 9403
+    },
+    {
+      "epoch": 1.50464,
+      "grad_norm": 2.1089204442208427,
+      "learning_rate": 0.00010456734031784101,
+      "loss": 1.3644,
+      "step": 9404
+    },
+    {
+      "epoch": 1.5048,
+      "grad_norm": 2.208454734137223,
+      "learning_rate": 0.0001045500844397678,
+      "loss": 1.3746,
+      "step": 9405
+    },
+    {
+      "epoch": 1.50496,
+      "grad_norm": 2.163161353478156,
+      "learning_rate": 0.00010453282842592666,
+      "loss": 1.2781,
+      "step": 9406
+    },
+    {
+      "epoch": 1.50512,
+      "grad_norm": 2.0815768264086345,
+      "learning_rate": 0.0001045155722768325,
+      "loss": 1.3461,
+      "step": 9407
+    },
+    {
+      "epoch": 1.50528,
+      "grad_norm": 2.217055136990908,
+      "learning_rate": 0.00010449831599300031,
+      "loss": 1.4433,
+      "step": 9408
+    },
+    {
+      "epoch": 1.5054400000000001,
+      "grad_norm": 2.374003051011858,
+      "learning_rate": 0.00010448105957494488,
+      "loss": 1.3384,
+      "step": 9409
+    },
+    {
+      "epoch": 1.5056,
+      "grad_norm": 2.2423629864639802,
+      "learning_rate": 0.00010446380302318117,
+      "loss": 1.3375,
+      "step": 9410
+    },
+    {
+      "epoch": 1.50576,
+      "grad_norm": 2.3045798143943865,
+      "learning_rate": 0.0001044465463382241,
+      "loss": 1.3581,
+      "step": 9411
+    },
+    {
+      "epoch": 1.5059200000000001,
+      "grad_norm": 1.9151589761025618,
+      "learning_rate": 0.00010442928952058857,
+      "loss": 1.2757,
+      "step": 9412
+    },
+    {
+      "epoch": 1.5060799999999999,
+      "grad_norm": 2.821135587888029,
+      "learning_rate": 0.00010441203257078947,
+      "loss": 1.3352,
+      "step": 9413
+    },
+    {
+      "epoch": 1.50624,
+      "grad_norm": 2.4022403043842693,
+      "learning_rate": 0.00010439477548934176,
+      "loss": 1.3485,
+      "step": 9414
+    },
+    {
+      "epoch": 1.5064,
+      "grad_norm": 2.4863827499147635,
+      "learning_rate": 0.00010437751827676033,
+      "loss": 1.3489,
+      "step": 9415
+    },
+    {
+      "epoch": 1.50656,
+      "grad_norm": 2.2844769660574515,
+      "learning_rate": 0.00010436026093356016,
+      "loss": 1.3408,
+      "step": 9416
+    },
+    {
+      "epoch": 1.50672,
+      "grad_norm": 2.1852345477617576,
+      "learning_rate": 0.00010434300346025616,
+      "loss": 1.3148,
+      "step": 9417
+    },
+    {
+      "epoch": 1.50688,
+      "grad_norm": 2.055478722499212,
+      "learning_rate": 0.00010432574585736326,
+      "loss": 1.2933,
+      "step": 9418
+    },
+    {
+      "epoch": 1.50704,
+      "grad_norm": 2.389484937705755,
+      "learning_rate": 0.0001043084881253964,
+      "loss": 1.33,
+      "step": 9419
+    },
+    {
+      "epoch": 1.5072,
+      "grad_norm": 2.128543694669618,
+      "learning_rate": 0.00010429123026487053,
+      "loss": 1.3951,
+      "step": 9420
+    },
+    {
+      "epoch": 1.50736,
+      "grad_norm": 2.1135065941956883,
+      "learning_rate": 0.0001042739722763006,
+      "loss": 1.4158,
+      "step": 9421
+    },
+    {
+      "epoch": 1.50752,
+      "grad_norm": 2.4172726972740035,
+      "learning_rate": 0.0001042567141602016,
+      "loss": 1.2915,
+      "step": 9422
+    },
+    {
+      "epoch": 1.5076800000000001,
+      "grad_norm": 2.1542712560634385,
+      "learning_rate": 0.00010423945591708839,
+      "loss": 1.2888,
+      "step": 9423
+    },
+    {
+      "epoch": 1.5078399999999998,
+      "grad_norm": 1.8538613952778746,
+      "learning_rate": 0.00010422219754747604,
+      "loss": 1.3772,
+      "step": 9424
+    },
+    {
+      "epoch": 1.508,
+      "grad_norm": 2.16935231101122,
+      "learning_rate": 0.00010420493905187945,
+      "loss": 1.3102,
+      "step": 9425
+    },
+    {
+      "epoch": 1.50816,
+      "grad_norm": 2.2515869198403857,
+      "learning_rate": 0.00010418768043081362,
+      "loss": 1.35,
+      "step": 9426
+    },
+    {
+      "epoch": 1.5083199999999999,
+      "grad_norm": 2.1205533628081135,
+      "learning_rate": 0.0001041704216847935,
+      "loss": 1.3373,
+      "step": 9427
+    },
+    {
+      "epoch": 1.50848,
+      "grad_norm": 2.1575322978724945,
+      "learning_rate": 0.00010415316281433407,
+      "loss": 1.276,
+      "step": 9428
+    },
+    {
+      "epoch": 1.50864,
+      "grad_norm": 2.2328746779358557,
+      "learning_rate": 0.00010413590381995031,
+      "loss": 1.393,
+      "step": 9429
+    },
+    {
+      "epoch": 1.5088,
+      "grad_norm": 2.065885691589535,
+      "learning_rate": 0.00010411864470215721,
+      "loss": 1.2869,
+      "step": 9430
+    },
+    {
+      "epoch": 1.50896,
+      "grad_norm": 2.0420693692389538,
+      "learning_rate": 0.00010410138546146975,
+      "loss": 1.3021,
+      "step": 9431
+    },
+    {
+      "epoch": 1.50912,
+      "grad_norm": 1.8870857628104643,
+      "learning_rate": 0.00010408412609840295,
+      "loss": 1.3302,
+      "step": 9432
+    },
+    {
+      "epoch": 1.50928,
+      "grad_norm": 1.9676471251692806,
+      "learning_rate": 0.00010406686661347174,
+      "loss": 1.3679,
+      "step": 9433
+    },
+    {
+      "epoch": 1.5094400000000001,
+      "grad_norm": 2.005664243762482,
+      "learning_rate": 0.0001040496070071912,
+      "loss": 1.3564,
+      "step": 9434
+    },
+    {
+      "epoch": 1.5096,
+      "grad_norm": 1.9507397767415968,
+      "learning_rate": 0.00010403234728007622,
+      "loss": 1.3296,
+      "step": 9435
+    },
+    {
+      "epoch": 1.50976,
+      "grad_norm": 2.0459562208692015,
+      "learning_rate": 0.00010401508743264192,
+      "loss": 1.3621,
+      "step": 9436
+    },
+    {
+      "epoch": 1.5099200000000002,
+      "grad_norm": 1.9923395891562181,
+      "learning_rate": 0.00010399782746540327,
+      "loss": 1.3798,
+      "step": 9437
+    },
+    {
+      "epoch": 1.5100799999999999,
+      "grad_norm": 2.1816049330251905,
+      "learning_rate": 0.00010398056737887528,
+      "loss": 1.3056,
+      "step": 9438
+    },
+    {
+      "epoch": 1.51024,
+      "grad_norm": 2.080628213753893,
+      "learning_rate": 0.00010396330717357292,
+      "loss": 1.3115,
+      "step": 9439
+    },
+    {
+      "epoch": 1.5104,
+      "grad_norm": 2.1311238597969937,
+      "learning_rate": 0.00010394604685001128,
+      "loss": 1.3133,
+      "step": 9440
+    },
+    {
+      "epoch": 1.51056,
+      "grad_norm": 2.245311796967875,
+      "learning_rate": 0.00010392878640870531,
+      "loss": 1.3836,
+      "step": 9441
+    },
+    {
+      "epoch": 1.51072,
+      "grad_norm": 1.9425197894900352,
+      "learning_rate": 0.00010391152585017012,
+      "loss": 1.3224,
+      "step": 9442
+    },
+    {
+      "epoch": 1.51088,
+      "grad_norm": 2.231669570671843,
+      "learning_rate": 0.0001038942651749207,
+      "loss": 1.3492,
+      "step": 9443
+    },
+    {
+      "epoch": 1.51104,
+      "grad_norm": 2.2116042431740106,
+      "learning_rate": 0.00010387700438347208,
+      "loss": 1.3125,
+      "step": 9444
+    },
+    {
+      "epoch": 1.5112,
+      "grad_norm": 2.116012531532905,
+      "learning_rate": 0.00010385974347633929,
+      "loss": 1.3307,
+      "step": 9445
+    },
+    {
+      "epoch": 1.51136,
+      "grad_norm": 2.7639076521556305,
+      "learning_rate": 0.00010384248245403739,
+      "loss": 1.2908,
+      "step": 9446
+    },
+    {
+      "epoch": 1.51152,
+      "grad_norm": 1.9477768116842216,
+      "learning_rate": 0.00010382522131708139,
+      "loss": 1.3058,
+      "step": 9447
+    },
+    {
+      "epoch": 1.5116800000000001,
+      "grad_norm": 2.1320471330121107,
+      "learning_rate": 0.00010380796006598636,
+      "loss": 1.3501,
+      "step": 9448
+    },
+    {
+      "epoch": 1.5118399999999999,
+      "grad_norm": 2.217056523201507,
+      "learning_rate": 0.00010379069870126734,
+      "loss": 1.3741,
+      "step": 9449
+    },
+    {
+      "epoch": 1.512,
+      "grad_norm": 2.120738555597105,
+      "learning_rate": 0.00010377343722343944,
+      "loss": 1.3626,
+      "step": 9450
+    },
+    {
+      "epoch": 1.51216,
+      "grad_norm": 1.9414472951556685,
+      "learning_rate": 0.00010375617563301765,
+      "loss": 1.2563,
+      "step": 9451
+    },
+    {
+      "epoch": 1.5123199999999999,
+      "grad_norm": 2.1951763859755533,
+      "learning_rate": 0.00010373891393051704,
+      "loss": 1.3365,
+      "step": 9452
+    },
+    {
+      "epoch": 1.51248,
+      "grad_norm": 2.071804921510366,
+      "learning_rate": 0.00010372165211645271,
+      "loss": 1.3428,
+      "step": 9453
+    },
+    {
+      "epoch": 1.51264,
+      "grad_norm": 2.068790198944966,
+      "learning_rate": 0.00010370439019133968,
+      "loss": 1.3204,
+      "step": 9454
+    },
+    {
+      "epoch": 1.5128,
+      "grad_norm": 2.392143746291116,
+      "learning_rate": 0.00010368712815569306,
+      "loss": 1.3517,
+      "step": 9455
+    },
+    {
+      "epoch": 1.51296,
+      "grad_norm": 2.1460659989120825,
+      "learning_rate": 0.0001036698660100279,
+      "loss": 1.3372,
+      "step": 9456
+    },
+    {
+      "epoch": 1.51312,
+      "grad_norm": 2.416927896388729,
+      "learning_rate": 0.00010365260375485928,
+      "loss": 1.3297,
+      "step": 9457
+    },
+    {
+      "epoch": 1.51328,
+      "grad_norm": 1.9782256236339868,
+      "learning_rate": 0.0001036353413907023,
+      "loss": 1.2964,
+      "step": 9458
+    },
+    {
+      "epoch": 1.5134400000000001,
+      "grad_norm": 1.9935366425838186,
+      "learning_rate": 0.00010361807891807202,
+      "loss": 1.347,
+      "step": 9459
+    },
+    {
+      "epoch": 1.5135999999999998,
+      "grad_norm": 2.5405445300433627,
+      "learning_rate": 0.00010360081633748356,
+      "loss": 1.421,
+      "step": 9460
+    },
+    {
+      "epoch": 1.51376,
+      "grad_norm": 2.015897947610362,
+      "learning_rate": 0.00010358355364945196,
+      "loss": 1.3284,
+      "step": 9461
+    },
+    {
+      "epoch": 1.5139200000000002,
+      "grad_norm": 1.9863266743706955,
+      "learning_rate": 0.00010356629085449236,
+      "loss": 1.2895,
+      "step": 9462
+    },
+    {
+      "epoch": 1.5140799999999999,
+      "grad_norm": 2.6547865281540886,
+      "learning_rate": 0.00010354902795311982,
+      "loss": 1.448,
+      "step": 9463
+    },
+    {
+      "epoch": 1.51424,
+      "grad_norm": 2.1806360384007473,
+      "learning_rate": 0.00010353176494584947,
+      "loss": 1.3116,
+      "step": 9464
+    },
+    {
+      "epoch": 1.5144,
+      "grad_norm": 2.0338734697971677,
+      "learning_rate": 0.00010351450183319638,
+      "loss": 1.4184,
+      "step": 9465
+    },
+    {
+      "epoch": 1.51456,
+      "grad_norm": 2.282279622782178,
+      "learning_rate": 0.00010349723861567571,
+      "loss": 1.3812,
+      "step": 9466
+    },
+    {
+      "epoch": 1.51472,
+      "grad_norm": 1.957561831305409,
+      "learning_rate": 0.00010347997529380251,
+      "loss": 1.3602,
+      "step": 9467
+    },
+    {
+      "epoch": 1.51488,
+      "grad_norm": 2.2527615225563484,
+      "learning_rate": 0.00010346271186809193,
+      "loss": 1.4063,
+      "step": 9468
+    },
+    {
+      "epoch": 1.51504,
+      "grad_norm": 2.0242070337978415,
+      "learning_rate": 0.00010344544833905907,
+      "loss": 1.3969,
+      "step": 9469
+    },
+    {
+      "epoch": 1.5152,
+      "grad_norm": 2.084495112475654,
+      "learning_rate": 0.00010342818470721907,
+      "loss": 1.3142,
+      "step": 9470
+    },
+    {
+      "epoch": 1.51536,
+      "grad_norm": 2.156991903613979,
+      "learning_rate": 0.000103410920973087,
+      "loss": 1.2909,
+      "step": 9471
+    },
+    {
+      "epoch": 1.51552,
+      "grad_norm": 1.9943956419727813,
+      "learning_rate": 0.00010339365713717801,
+      "loss": 1.3992,
+      "step": 9472
+    },
+    {
+      "epoch": 1.5156800000000001,
+      "grad_norm": 1.991434039181839,
+      "learning_rate": 0.00010337639320000727,
+      "loss": 1.3348,
+      "step": 9473
+    },
+    {
+      "epoch": 1.5158399999999999,
+      "grad_norm": 1.9783799858405273,
+      "learning_rate": 0.00010335912916208988,
+      "loss": 1.3334,
+      "step": 9474
+    },
+    {
+      "epoch": 1.516,
+      "grad_norm": 1.9382142477562498,
+      "learning_rate": 0.00010334186502394094,
+      "loss": 1.2394,
+      "step": 9475
+    },
+    {
+      "epoch": 1.51616,
+      "grad_norm": 2.098727345570236,
+      "learning_rate": 0.00010332460078607565,
+      "loss": 1.3209,
+      "step": 9476
+    },
+    {
+      "epoch": 1.51632,
+      "grad_norm": 1.8515705547202832,
+      "learning_rate": 0.0001033073364490091,
+      "loss": 1.3339,
+      "step": 9477
+    },
+    {
+      "epoch": 1.51648,
+      "grad_norm": 2.2481592311357725,
+      "learning_rate": 0.00010329007201325646,
+      "loss": 1.346,
+      "step": 9478
+    },
+    {
+      "epoch": 1.51664,
+      "grad_norm": 2.7239941238057535,
+      "learning_rate": 0.00010327280747933285,
+      "loss": 1.3609,
+      "step": 9479
+    },
+    {
+      "epoch": 1.5168,
+      "grad_norm": 2.077147096479128,
+      "learning_rate": 0.00010325554284775343,
+      "loss": 1.35,
+      "step": 9480
+    },
+    {
+      "epoch": 1.51696,
+      "grad_norm": 2.049505432761305,
+      "learning_rate": 0.00010323827811903337,
+      "loss": 1.321,
+      "step": 9481
+    },
+    {
+      "epoch": 1.51712,
+      "grad_norm": 2.921572313200667,
+      "learning_rate": 0.00010322101329368781,
+      "loss": 1.3048,
+      "step": 9482
+    },
+    {
+      "epoch": 1.51728,
+      "grad_norm": 2.0211790558076115,
+      "learning_rate": 0.0001032037483722319,
+      "loss": 1.3519,
+      "step": 9483
+    },
+    {
+      "epoch": 1.5174400000000001,
+      "grad_norm": 1.8077341771181583,
+      "learning_rate": 0.00010318648335518082,
+      "loss": 1.3206,
+      "step": 9484
+    },
+    {
+      "epoch": 1.5175999999999998,
+      "grad_norm": 2.197353762951714,
+      "learning_rate": 0.0001031692182430497,
+      "loss": 1.3129,
+      "step": 9485
+    },
+    {
+      "epoch": 1.51776,
+      "grad_norm": 2.3167182117154437,
+      "learning_rate": 0.00010315195303635375,
+      "loss": 1.3191,
+      "step": 9486
+    },
+    {
+      "epoch": 1.5179200000000002,
+      "grad_norm": 2.1610144124844055,
+      "learning_rate": 0.0001031346877356081,
+      "loss": 1.3403,
+      "step": 9487
+    },
+    {
+      "epoch": 1.5180799999999999,
+      "grad_norm": 2.9069827824362737,
+      "learning_rate": 0.00010311742234132791,
+      "loss": 1.3512,
+      "step": 9488
+    },
+    {
+      "epoch": 1.51824,
+      "grad_norm": 1.970690938692273,
+      "learning_rate": 0.00010310015685402841,
+      "loss": 1.2792,
+      "step": 9489
+    },
+    {
+      "epoch": 1.5184,
+      "grad_norm": 2.199073714663901,
+      "learning_rate": 0.00010308289127422475,
+      "loss": 1.2766,
+      "step": 9490
+    },
+    {
+      "epoch": 1.51856,
+      "grad_norm": 1.995366811321799,
+      "learning_rate": 0.0001030656256024321,
+      "loss": 1.2353,
+      "step": 9491
+    },
+    {
+      "epoch": 1.51872,
+      "grad_norm": 2.331584936308505,
+      "learning_rate": 0.00010304835983916566,
+      "loss": 1.3333,
+      "step": 9492
+    },
+    {
+      "epoch": 1.51888,
+      "grad_norm": 2.063381775256606,
+      "learning_rate": 0.00010303109398494057,
+      "loss": 1.3081,
+      "step": 9493
+    },
+    {
+      "epoch": 1.51904,
+      "grad_norm": 2.246798385239283,
+      "learning_rate": 0.00010301382804027206,
+      "loss": 1.3892,
+      "step": 9494
+    },
+    {
+      "epoch": 1.5192,
+      "grad_norm": 2.0256998568742537,
+      "learning_rate": 0.00010299656200567533,
+      "loss": 1.3328,
+      "step": 9495
+    },
+    {
+      "epoch": 1.51936,
+      "grad_norm": 2.5740981347391583,
+      "learning_rate": 0.00010297929588166557,
+      "loss": 1.4081,
+      "step": 9496
+    },
+    {
+      "epoch": 1.51952,
+      "grad_norm": 2.693682039205087,
+      "learning_rate": 0.00010296202966875794,
+      "loss": 1.3288,
+      "step": 9497
+    },
+    {
+      "epoch": 1.5196800000000001,
+      "grad_norm": 1.9277304173335796,
+      "learning_rate": 0.00010294476336746766,
+      "loss": 1.3535,
+      "step": 9498
+    },
+    {
+      "epoch": 1.5198399999999999,
+      "grad_norm": 2.6298412104528794,
+      "learning_rate": 0.00010292749697830992,
+      "loss": 1.4136,
+      "step": 9499
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 2.570471318202771,
+      "learning_rate": 0.00010291023050179995,
+      "loss": 1.3829,
+      "step": 9500
+    },
+    {
+      "epoch": 1.52016,
+      "grad_norm": 2.1898447201522977,
+      "learning_rate": 0.00010289296393845296,
+      "loss": 1.3943,
+      "step": 9501
+    },
+    {
+      "epoch": 1.52032,
+      "grad_norm": 2.1166298208113012,
+      "learning_rate": 0.00010287569728878415,
+      "loss": 1.305,
+      "step": 9502
+    },
+    {
+      "epoch": 1.52048,
+      "grad_norm": 2.57929239800215,
+      "learning_rate": 0.00010285843055330869,
+      "loss": 1.3621,
+      "step": 9503
+    },
+    {
+      "epoch": 1.52064,
+      "grad_norm": 1.9045863016886593,
+      "learning_rate": 0.00010284116373254185,
+      "loss": 1.383,
+      "step": 9504
+    },
+    {
+      "epoch": 1.5208,
+      "grad_norm": 2.3396097214058407,
+      "learning_rate": 0.0001028238968269988,
+      "loss": 1.3016,
+      "step": 9505
+    },
+    {
+      "epoch": 1.52096,
+      "grad_norm": 2.320008733541773,
+      "learning_rate": 0.00010280662983719477,
+      "loss": 1.3262,
+      "step": 9506
+    },
+    {
+      "epoch": 1.52112,
+      "grad_norm": 2.048370583027092,
+      "learning_rate": 0.00010278936276364503,
+      "loss": 1.349,
+      "step": 9507
+    },
+    {
+      "epoch": 1.52128,
+      "grad_norm": 2.6622008619222646,
+      "learning_rate": 0.00010277209560686478,
+      "loss": 1.3309,
+      "step": 9508
+    },
+    {
+      "epoch": 1.5214400000000001,
+      "grad_norm": 1.8973263007549606,
+      "learning_rate": 0.0001027548283673692,
+      "loss": 1.3412,
+      "step": 9509
+    },
+    {
+      "epoch": 1.5215999999999998,
+      "grad_norm": 2.178367598045612,
+      "learning_rate": 0.00010273756104567359,
+      "loss": 1.3229,
+      "step": 9510
+    },
+    {
+      "epoch": 1.52176,
+      "grad_norm": 2.159878733925123,
+      "learning_rate": 0.00010272029364229314,
+      "loss": 1.367,
+      "step": 9511
+    },
+    {
+      "epoch": 1.5219200000000002,
+      "grad_norm": 2.083132474420144,
+      "learning_rate": 0.00010270302615774308,
+      "loss": 1.2648,
+      "step": 9512
+    },
+    {
+      "epoch": 1.5220799999999999,
+      "grad_norm": 2.1587842588655755,
+      "learning_rate": 0.00010268575859253866,
+      "loss": 1.3485,
+      "step": 9513
+    },
+    {
+      "epoch": 1.52224,
+      "grad_norm": 2.0929324826406126,
+      "learning_rate": 0.0001026684909471951,
+      "loss": 1.352,
+      "step": 9514
+    },
+    {
+      "epoch": 1.5224,
+      "grad_norm": 2.0360686732328293,
+      "learning_rate": 0.00010265122322222769,
+      "loss": 1.3376,
+      "step": 9515
+    },
+    {
+      "epoch": 1.52256,
+      "grad_norm": 2.0731746974491605,
+      "learning_rate": 0.00010263395541815162,
+      "loss": 1.3219,
+      "step": 9516
+    },
+    {
+      "epoch": 1.52272,
+      "grad_norm": 1.7861612315013158,
+      "learning_rate": 0.00010261668753548217,
+      "loss": 1.2839,
+      "step": 9517
+    },
+    {
+      "epoch": 1.52288,
+      "grad_norm": 2.475945439795313,
+      "learning_rate": 0.00010259941957473457,
+      "loss": 1.3588,
+      "step": 9518
+    },
+    {
+      "epoch": 1.52304,
+      "grad_norm": 2.739887040641089,
+      "learning_rate": 0.0001025821515364241,
+      "loss": 1.3508,
+      "step": 9519
+    },
+    {
+      "epoch": 1.5232,
+      "grad_norm": 1.9878411899479143,
+      "learning_rate": 0.00010256488342106598,
+      "loss": 1.4153,
+      "step": 9520
+    },
+    {
+      "epoch": 1.52336,
+      "grad_norm": 1.965270363228157,
+      "learning_rate": 0.00010254761522917546,
+      "loss": 1.3528,
+      "step": 9521
+    },
+    {
+      "epoch": 1.52352,
+      "grad_norm": 2.3067117171047733,
+      "learning_rate": 0.0001025303469612678,
+      "loss": 1.3707,
+      "step": 9522
+    },
+    {
+      "epoch": 1.5236800000000001,
+      "grad_norm": 2.247942466502029,
+      "learning_rate": 0.0001025130786178583,
+      "loss": 1.2922,
+      "step": 9523
+    },
+    {
+      "epoch": 1.5238399999999999,
+      "grad_norm": 2.254994898997879,
+      "learning_rate": 0.0001024958101994622,
+      "loss": 1.2924,
+      "step": 9524
+    },
+    {
+      "epoch": 1.524,
+      "grad_norm": 1.9403514334527299,
+      "learning_rate": 0.00010247854170659476,
+      "loss": 1.3172,
+      "step": 9525
+    },
+    {
+      "epoch": 1.52416,
+      "grad_norm": 1.990263356414204,
+      "learning_rate": 0.00010246127313977125,
+      "loss": 1.347,
+      "step": 9526
+    },
+    {
+      "epoch": 1.52432,
+      "grad_norm": 2.3204136750669,
+      "learning_rate": 0.00010244400449950692,
+      "loss": 1.3626,
+      "step": 9527
+    },
+    {
+      "epoch": 1.52448,
+      "grad_norm": 2.0685145413173074,
+      "learning_rate": 0.00010242673578631707,
+      "loss": 1.3039,
+      "step": 9528
+    },
+    {
+      "epoch": 1.52464,
+      "grad_norm": 1.9802409830434673,
+      "learning_rate": 0.00010240946700071693,
+      "loss": 1.267,
+      "step": 9529
+    },
+    {
+      "epoch": 1.5248,
+      "grad_norm": 2.3819181099808486,
+      "learning_rate": 0.00010239219814322183,
+      "loss": 1.3593,
+      "step": 9530
+    },
+    {
+      "epoch": 1.52496,
+      "grad_norm": 2.114186801203799,
+      "learning_rate": 0.00010237492921434703,
+      "loss": 1.3204,
+      "step": 9531
+    },
+    {
+      "epoch": 1.52512,
+      "grad_norm": 1.8612014093974067,
+      "learning_rate": 0.00010235766021460779,
+      "loss": 1.2979,
+      "step": 9532
+    },
+    {
+      "epoch": 1.52528,
+      "grad_norm": 2.326778850950231,
+      "learning_rate": 0.00010234039114451942,
+      "loss": 1.4867,
+      "step": 9533
+    },
+    {
+      "epoch": 1.5254400000000001,
+      "grad_norm": 2.436838453149964,
+      "learning_rate": 0.00010232312200459717,
+      "loss": 1.3484,
+      "step": 9534
+    },
+    {
+      "epoch": 1.5255999999999998,
+      "grad_norm": 2.1004550847900085,
+      "learning_rate": 0.00010230585279535636,
+      "loss": 1.3785,
+      "step": 9535
+    },
+    {
+      "epoch": 1.52576,
+      "grad_norm": 2.1832419092568176,
+      "learning_rate": 0.00010228858351731225,
+      "loss": 1.3532,
+      "step": 9536
+    },
+    {
+      "epoch": 1.5259200000000002,
+      "grad_norm": 2.0809941977659494,
+      "learning_rate": 0.00010227131417098018,
+      "loss": 1.3019,
+      "step": 9537
+    },
+    {
+      "epoch": 1.5260799999999999,
+      "grad_norm": 2.2628460764872775,
+      "learning_rate": 0.00010225404475687534,
+      "loss": 1.3171,
+      "step": 9538
+    },
+    {
+      "epoch": 1.52624,
+      "grad_norm": 2.1569026143465497,
+      "learning_rate": 0.00010223677527551311,
+      "loss": 1.3086,
+      "step": 9539
+    },
+    {
+      "epoch": 1.5264,
+      "grad_norm": 2.14625065263954,
+      "learning_rate": 0.0001022195057274088,
+      "loss": 1.3688,
+      "step": 9540
+    },
+    {
+      "epoch": 1.52656,
+      "grad_norm": 2.0451071160597025,
+      "learning_rate": 0.00010220223611307764,
+      "loss": 1.3247,
+      "step": 9541
+    },
+    {
+      "epoch": 1.52672,
+      "grad_norm": 1.880628185785983,
+      "learning_rate": 0.000102184966433035,
+      "loss": 1.2889,
+      "step": 9542
+    },
+    {
+      "epoch": 1.52688,
+      "grad_norm": 1.9651300711405124,
+      "learning_rate": 0.00010216769668779612,
+      "loss": 1.321,
+      "step": 9543
+    },
+    {
+      "epoch": 1.52704,
+      "grad_norm": 2.110917987329504,
+      "learning_rate": 0.0001021504268778763,
+      "loss": 1.393,
+      "step": 9544
+    },
+    {
+      "epoch": 1.5272000000000001,
+      "grad_norm": 2.047503925618877,
+      "learning_rate": 0.00010213315700379092,
+      "loss": 1.3497,
+      "step": 9545
+    },
+    {
+      "epoch": 1.52736,
+      "grad_norm": 1.9700666453120783,
+      "learning_rate": 0.00010211588706605522,
+      "loss": 1.3383,
+      "step": 9546
+    },
+    {
+      "epoch": 1.52752,
+      "grad_norm": 1.714503651919512,
+      "learning_rate": 0.00010209861706518454,
+      "loss": 1.3092,
+      "step": 9547
+    },
+    {
+      "epoch": 1.5276800000000001,
+      "grad_norm": 2.0417731964021377,
+      "learning_rate": 0.00010208134700169418,
+      "loss": 1.2858,
+      "step": 9548
+    },
+    {
+      "epoch": 1.5278399999999999,
+      "grad_norm": 2.326340608678765,
+      "learning_rate": 0.00010206407687609947,
+      "loss": 1.3374,
+      "step": 9549
+    },
+    {
+      "epoch": 1.528,
+      "grad_norm": 2.1006844868477303,
+      "learning_rate": 0.0001020468066889157,
+      "loss": 1.3305,
+      "step": 9550
+    },
+    {
+      "epoch": 1.52816,
+      "grad_norm": 3.002829514844704,
+      "learning_rate": 0.00010202953644065822,
+      "loss": 1.3139,
+      "step": 9551
+    },
+    {
+      "epoch": 1.52832,
+      "grad_norm": 2.074303203824907,
+      "learning_rate": 0.0001020122661318423,
+      "loss": 1.3149,
+      "step": 9552
+    },
+    {
+      "epoch": 1.52848,
+      "grad_norm": 2.2229367386542243,
+      "learning_rate": 0.0001019949957629833,
+      "loss": 1.4306,
+      "step": 9553
+    },
+    {
+      "epoch": 1.52864,
+      "grad_norm": 2.612277133742859,
+      "learning_rate": 0.00010197772533459655,
+      "loss": 1.322,
+      "step": 9554
+    },
+    {
+      "epoch": 1.5288,
+      "grad_norm": 1.9858527834629331,
+      "learning_rate": 0.00010196045484719733,
+      "loss": 1.3256,
+      "step": 9555
+    },
+    {
+      "epoch": 1.52896,
+      "grad_norm": 2.2683776274048446,
+      "learning_rate": 0.00010194318430130101,
+      "loss": 1.3387,
+      "step": 9556
+    },
+    {
+      "epoch": 1.52912,
+      "grad_norm": 2.7509883714518253,
+      "learning_rate": 0.00010192591369742292,
+      "loss": 1.3332,
+      "step": 9557
+    },
+    {
+      "epoch": 1.52928,
+      "grad_norm": 2.315822011848309,
+      "learning_rate": 0.00010190864303607837,
+      "loss": 1.286,
+      "step": 9558
+    },
+    {
+      "epoch": 1.5294400000000001,
+      "grad_norm": 2.3726793921897484,
+      "learning_rate": 0.00010189137231778267,
+      "loss": 1.3599,
+      "step": 9559
+    },
+    {
+      "epoch": 1.5295999999999998,
+      "grad_norm": 3.727461744222217,
+      "learning_rate": 0.0001018741015430512,
+      "loss": 1.2934,
+      "step": 9560
+    },
+    {
+      "epoch": 1.52976,
+      "grad_norm": 2.3334713622998384,
+      "learning_rate": 0.00010185683071239924,
+      "loss": 1.4207,
+      "step": 9561
+    },
+    {
+      "epoch": 1.52992,
+      "grad_norm": 2.59443177250899,
+      "learning_rate": 0.0001018395598263422,
+      "loss": 1.3514,
+      "step": 9562
+    },
+    {
+      "epoch": 1.5300799999999999,
+      "grad_norm": 3.163970636282255,
+      "learning_rate": 0.00010182228888539532,
+      "loss": 1.347,
+      "step": 9563
+    },
+    {
+      "epoch": 1.53024,
+      "grad_norm": 2.2779108527742187,
+      "learning_rate": 0.00010180501789007404,
+      "loss": 1.348,
+      "step": 9564
+    },
+    {
+      "epoch": 1.5304,
+      "grad_norm": 1.8907075207682102,
+      "learning_rate": 0.00010178774684089363,
+      "loss": 1.4085,
+      "step": 9565
+    },
+    {
+      "epoch": 1.53056,
+      "grad_norm": 2.2432271510793536,
+      "learning_rate": 0.00010177047573836948,
+      "loss": 1.3373,
+      "step": 9566
+    },
+    {
+      "epoch": 1.53072,
+      "grad_norm": 2.4817316094874236,
+      "learning_rate": 0.00010175320458301691,
+      "loss": 1.3105,
+      "step": 9567
+    },
+    {
+      "epoch": 1.53088,
+      "grad_norm": 2.2373846332600453,
+      "learning_rate": 0.00010173593337535125,
+      "loss": 1.2331,
+      "step": 9568
+    },
+    {
+      "epoch": 1.53104,
+      "grad_norm": 2.266465048311983,
+      "learning_rate": 0.00010171866211588789,
+      "loss": 1.2821,
+      "step": 9569
+    },
+    {
+      "epoch": 1.5312000000000001,
+      "grad_norm": 2.1985831165573053,
+      "learning_rate": 0.00010170139080514216,
+      "loss": 1.3979,
+      "step": 9570
+    },
+    {
+      "epoch": 1.53136,
+      "grad_norm": 1.9093225309985586,
+      "learning_rate": 0.00010168411944362937,
+      "loss": 1.3206,
+      "step": 9571
+    },
+    {
+      "epoch": 1.53152,
+      "grad_norm": 2.2022124107712657,
+      "learning_rate": 0.00010166684803186494,
+      "loss": 1.3469,
+      "step": 9572
+    },
+    {
+      "epoch": 1.5316800000000002,
+      "grad_norm": 2.5413803452168517,
+      "learning_rate": 0.00010164957657036421,
+      "loss": 1.3134,
+      "step": 9573
+    },
+    {
+      "epoch": 1.5318399999999999,
+      "grad_norm": 2.055948291412524,
+      "learning_rate": 0.0001016323050596425,
+      "loss": 1.3476,
+      "step": 9574
+    },
+    {
+      "epoch": 1.532,
+      "grad_norm": 2.157401835978069,
+      "learning_rate": 0.00010161503350021517,
+      "loss": 1.3855,
+      "step": 9575
+    },
+    {
+      "epoch": 1.53216,
+      "grad_norm": 3.1999732347126395,
+      "learning_rate": 0.00010159776189259762,
+      "loss": 1.3288,
+      "step": 9576
+    },
+    {
+      "epoch": 1.53232,
+      "grad_norm": 1.7386983784860521,
+      "learning_rate": 0.00010158049023730518,
+      "loss": 1.3482,
+      "step": 9577
+    },
+    {
+      "epoch": 1.53248,
+      "grad_norm": 2.3784410199974055,
+      "learning_rate": 0.0001015632185348532,
+      "loss": 1.335,
+      "step": 9578
+    },
+    {
+      "epoch": 1.53264,
+      "grad_norm": 2.5080122511974787,
+      "learning_rate": 0.00010154594678575704,
+      "loss": 1.3778,
+      "step": 9579
+    },
+    {
+      "epoch": 1.5328,
+      "grad_norm": 2.2616310135618036,
+      "learning_rate": 0.00010152867499053212,
+      "loss": 1.3209,
+      "step": 9580
+    },
+    {
+      "epoch": 1.53296,
+      "grad_norm": 2.3928960471561167,
+      "learning_rate": 0.00010151140314969373,
+      "loss": 1.2514,
+      "step": 9581
+    },
+    {
+      "epoch": 1.53312,
+      "grad_norm": 2.7350346234758534,
+      "learning_rate": 0.00010149413126375731,
+      "loss": 1.3374,
+      "step": 9582
+    },
+    {
+      "epoch": 1.53328,
+      "grad_norm": 1.8947255918681256,
+      "learning_rate": 0.00010147685933323816,
+      "loss": 1.3013,
+      "step": 9583
+    },
+    {
+      "epoch": 1.5334400000000001,
+      "grad_norm": 2.7244113870479127,
+      "learning_rate": 0.00010145958735865171,
+      "loss": 1.3532,
+      "step": 9584
+    },
+    {
+      "epoch": 1.5335999999999999,
+      "grad_norm": 2.462999274201729,
+      "learning_rate": 0.00010144231534051327,
+      "loss": 1.3954,
+      "step": 9585
+    },
+    {
+      "epoch": 1.53376,
+      "grad_norm": 2.152782732912722,
+      "learning_rate": 0.00010142504327933826,
+      "loss": 1.2759,
+      "step": 9586
+    },
+    {
+      "epoch": 1.53392,
+      "grad_norm": 2.463298655352083,
+      "learning_rate": 0.00010140777117564202,
+      "loss": 1.3137,
+      "step": 9587
+    },
+    {
+      "epoch": 1.5340799999999999,
+      "grad_norm": 2.6067348119487694,
+      "learning_rate": 0.00010139049902993993,
+      "loss": 1.28,
+      "step": 9588
+    },
+    {
+      "epoch": 1.53424,
+      "grad_norm": 2.04837386785889,
+      "learning_rate": 0.00010137322684274742,
+      "loss": 1.3048,
+      "step": 9589
+    },
+    {
+      "epoch": 1.5344,
+      "grad_norm": 2.822610546623702,
+      "learning_rate": 0.00010135595461457979,
+      "loss": 1.4078,
+      "step": 9590
+    },
+    {
+      "epoch": 1.53456,
+      "grad_norm": 2.3448484077366127,
+      "learning_rate": 0.00010133868234595246,
+      "loss": 1.3333,
+      "step": 9591
+    },
+    {
+      "epoch": 1.53472,
+      "grad_norm": 2.0679714065955506,
+      "learning_rate": 0.00010132141003738079,
+      "loss": 1.3343,
+      "step": 9592
+    },
+    {
+      "epoch": 1.53488,
+      "grad_norm": 2.737692550430764,
+      "learning_rate": 0.00010130413768938018,
+      "loss": 1.3449,
+      "step": 9593
+    },
+    {
+      "epoch": 1.53504,
+      "grad_norm": 2.2599362215944643,
+      "learning_rate": 0.00010128686530246601,
+      "loss": 1.3206,
+      "step": 9594
+    },
+    {
+      "epoch": 1.5352000000000001,
+      "grad_norm": 2.2904065953050474,
+      "learning_rate": 0.0001012695928771536,
+      "loss": 1.2918,
+      "step": 9595
+    },
+    {
+      "epoch": 1.5353599999999998,
+      "grad_norm": 2.7280545903618405,
+      "learning_rate": 0.00010125232041395844,
+      "loss": 1.3642,
+      "step": 9596
+    },
+    {
+      "epoch": 1.53552,
+      "grad_norm": 2.123179462687601,
+      "learning_rate": 0.00010123504791339587,
+      "loss": 1.2916,
+      "step": 9597
+    },
+    {
+      "epoch": 1.5356800000000002,
+      "grad_norm": 2.078967879052199,
+      "learning_rate": 0.00010121777537598123,
+      "loss": 1.298,
+      "step": 9598
+    },
+    {
+      "epoch": 1.5358399999999999,
+      "grad_norm": 2.802286253148417,
+      "learning_rate": 0.00010120050280223,
+      "loss": 1.3522,
+      "step": 9599
+    },
+    {
+      "epoch": 1.536,
+      "grad_norm": 2.324989983638433,
+      "learning_rate": 0.00010118323019265748,
+      "loss": 1.298,
+      "step": 9600
+    },
+    {
+      "epoch": 1.53616,
+      "grad_norm": 2.251680461158374,
+      "learning_rate": 0.00010116595754777909,
+      "loss": 1.3444,
+      "step": 9601
+    },
+    {
+      "epoch": 1.53632,
+      "grad_norm": 2.623176301085076,
+      "learning_rate": 0.00010114868486811025,
+      "loss": 1.2949,
+      "step": 9602
+    },
+    {
+      "epoch": 1.53648,
+      "grad_norm": 1.9712213521520163,
+      "learning_rate": 0.00010113141215416628,
+      "loss": 1.2726,
+      "step": 9603
+    },
+    {
+      "epoch": 1.53664,
+      "grad_norm": 1.973321766352145,
+      "learning_rate": 0.00010111413940646266,
+      "loss": 1.2628,
+      "step": 9604
+    },
+    {
+      "epoch": 1.5368,
+      "grad_norm": 2.1627104826653896,
+      "learning_rate": 0.00010109686662551474,
+      "loss": 1.3806,
+      "step": 9605
+    },
+    {
+      "epoch": 1.53696,
+      "grad_norm": 2.07517568099329,
+      "learning_rate": 0.00010107959381183792,
+      "loss": 1.3054,
+      "step": 9606
+    },
+    {
+      "epoch": 1.53712,
+      "grad_norm": 1.8967551609312983,
+      "learning_rate": 0.00010106232096594759,
+      "loss": 1.3136,
+      "step": 9607
+    },
+    {
+      "epoch": 1.53728,
+      "grad_norm": 2.2722798766126173,
+      "learning_rate": 0.00010104504808835915,
+      "loss": 1.436,
+      "step": 9608
+    },
+    {
+      "epoch": 1.5374400000000001,
+      "grad_norm": 2.01755493664041,
+      "learning_rate": 0.00010102777517958801,
+      "loss": 1.4093,
+      "step": 9609
+    },
+    {
+      "epoch": 1.5375999999999999,
+      "grad_norm": 2.3411170368038055,
+      "learning_rate": 0.00010101050224014954,
+      "loss": 1.3552,
+      "step": 9610
+    },
+    {
+      "epoch": 1.53776,
+      "grad_norm": 2.0258067972815472,
+      "learning_rate": 0.00010099322927055915,
+      "loss": 1.3415,
+      "step": 9611
+    },
+    {
+      "epoch": 1.53792,
+      "grad_norm": 1.8461444914070941,
+      "learning_rate": 0.00010097595627133225,
+      "loss": 1.3109,
+      "step": 9612
+    },
+    {
+      "epoch": 1.53808,
+      "grad_norm": 1.9326190010058557,
+      "learning_rate": 0.00010095868324298423,
+      "loss": 1.3176,
+      "step": 9613
+    },
+    {
+      "epoch": 1.53824,
+      "grad_norm": 2.124252326337016,
+      "learning_rate": 0.00010094141018603052,
+      "loss": 1.357,
+      "step": 9614
+    },
+    {
+      "epoch": 1.5384,
+      "grad_norm": 1.9438177232287015,
+      "learning_rate": 0.0001009241371009865,
+      "loss": 1.3506,
+      "step": 9615
+    },
+    {
+      "epoch": 1.53856,
+      "grad_norm": 1.8817442185017086,
+      "learning_rate": 0.00010090686398836756,
+      "loss": 1.28,
+      "step": 9616
+    },
+    {
+      "epoch": 1.53872,
+      "grad_norm": 1.9683306025261316,
+      "learning_rate": 0.00010088959084868912,
+      "loss": 1.2701,
+      "step": 9617
+    },
+    {
+      "epoch": 1.53888,
+      "grad_norm": 2.736733992288128,
+      "learning_rate": 0.00010087231768246662,
+      "loss": 1.3133,
+      "step": 9618
+    },
+    {
+      "epoch": 1.53904,
+      "grad_norm": 2.3014590861450444,
+      "learning_rate": 0.0001008550444902154,
+      "loss": 1.3873,
+      "step": 9619
+    },
+    {
+      "epoch": 1.5392000000000001,
+      "grad_norm": 2.5582609993559884,
+      "learning_rate": 0.00010083777127245089,
+      "loss": 1.4241,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5393599999999998,
+      "grad_norm": 3.319534263952595,
+      "learning_rate": 0.00010082049802968852,
+      "loss": 1.3329,
+      "step": 9621
+    },
+    {
+      "epoch": 1.53952,
+      "grad_norm": 1.9126020189032955,
+      "learning_rate": 0.00010080322476244369,
+      "loss": 1.3682,
+      "step": 9622
+    },
+    {
+      "epoch": 1.5396800000000002,
+      "grad_norm": 2.220075387552874,
+      "learning_rate": 0.00010078595147123178,
+      "loss": 1.3091,
+      "step": 9623
+    },
+    {
+      "epoch": 1.5398399999999999,
+      "grad_norm": 2.5875740163266863,
+      "learning_rate": 0.00010076867815656825,
+      "loss": 1.3412,
+      "step": 9624
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 2.192379726961838,
+      "learning_rate": 0.00010075140481896848,
+      "loss": 1.3407,
+      "step": 9625
+    },
+    {
+      "epoch": 1.54016,
+      "grad_norm": 2.2299899192049755,
+      "learning_rate": 0.00010073413145894788,
+      "loss": 1.3997,
+      "step": 9626
+    },
+    {
+      "epoch": 1.54032,
+      "grad_norm": 2.313653259672008,
+      "learning_rate": 0.00010071685807702187,
+      "loss": 1.3069,
+      "step": 9627
+    },
+    {
+      "epoch": 1.54048,
+      "grad_norm": 2.4287271974404785,
+      "learning_rate": 0.00010069958467370583,
+      "loss": 1.2872,
+      "step": 9628
+    },
+    {
+      "epoch": 1.54064,
+      "grad_norm": 2.390264317089359,
+      "learning_rate": 0.00010068231124951524,
+      "loss": 1.4064,
+      "step": 9629
+    },
+    {
+      "epoch": 1.5408,
+      "grad_norm": 2.649486859225252,
+      "learning_rate": 0.00010066503780496548,
+      "loss": 1.3175,
+      "step": 9630
+    },
+    {
+      "epoch": 1.54096,
+      "grad_norm": 2.032850336083663,
+      "learning_rate": 0.00010064776434057196,
+      "loss": 1.3027,
+      "step": 9631
+    },
+    {
+      "epoch": 1.54112,
+      "grad_norm": 2.494409802962139,
+      "learning_rate": 0.00010063049085685006,
+      "loss": 1.2717,
+      "step": 9632
+    },
+    {
+      "epoch": 1.54128,
+      "grad_norm": 1.9411718410551582,
+      "learning_rate": 0.00010061321735431526,
+      "loss": 1.3973,
+      "step": 9633
+    },
+    {
+      "epoch": 1.5414400000000001,
+      "grad_norm": 2.221319743698561,
+      "learning_rate": 0.00010059594383348294,
+      "loss": 1.3923,
+      "step": 9634
+    },
+    {
+      "epoch": 1.5415999999999999,
+      "grad_norm": 2.169373432661991,
+      "learning_rate": 0.00010057867029486853,
+      "loss": 1.3505,
+      "step": 9635
+    },
+    {
+      "epoch": 1.54176,
+      "grad_norm": 2.074663737478676,
+      "learning_rate": 0.00010056139673898744,
+      "loss": 1.3308,
+      "step": 9636
+    },
+    {
+      "epoch": 1.54192,
+      "grad_norm": 2.662440659239399,
+      "learning_rate": 0.00010054412316635506,
+      "loss": 1.3615,
+      "step": 9637
+    },
+    {
+      "epoch": 1.54208,
+      "grad_norm": 2.4088237636926735,
+      "learning_rate": 0.00010052684957748689,
+      "loss": 1.3202,
+      "step": 9638
+    },
+    {
+      "epoch": 1.54224,
+      "grad_norm": 1.8515337733563306,
+      "learning_rate": 0.00010050957597289828,
+      "loss": 1.2487,
+      "step": 9639
+    },
+    {
+      "epoch": 1.5424,
+      "grad_norm": 2.124490359665448,
+      "learning_rate": 0.00010049230235310466,
+      "loss": 1.2991,
+      "step": 9640
+    },
+    {
+      "epoch": 1.54256,
+      "grad_norm": 2.3777518040637666,
+      "learning_rate": 0.00010047502871862145,
+      "loss": 1.3885,
+      "step": 9641
+    },
+    {
+      "epoch": 1.54272,
+      "grad_norm": 2.0869423900915463,
+      "learning_rate": 0.00010045775506996406,
+      "loss": 1.2817,
+      "step": 9642
+    },
+    {
+      "epoch": 1.54288,
+      "grad_norm": 2.555532570485144,
+      "learning_rate": 0.00010044048140764795,
+      "loss": 1.2786,
+      "step": 9643
+    },
+    {
+      "epoch": 1.54304,
+      "grad_norm": 1.9262799548579015,
+      "learning_rate": 0.00010042320773218849,
+      "loss": 1.281,
+      "step": 9644
+    },
+    {
+      "epoch": 1.5432000000000001,
+      "grad_norm": 2.509246161036638,
+      "learning_rate": 0.00010040593404410115,
+      "loss": 1.2873,
+      "step": 9645
+    },
+    {
+      "epoch": 1.5433599999999998,
+      "grad_norm": 2.012790572248557,
+      "learning_rate": 0.00010038866034390133,
+      "loss": 1.3206,
+      "step": 9646
+    },
+    {
+      "epoch": 1.54352,
+      "grad_norm": 2.1487109165258227,
+      "learning_rate": 0.00010037138663210446,
+      "loss": 1.3382,
+      "step": 9647
+    },
+    {
+      "epoch": 1.5436800000000002,
+      "grad_norm": 2.0312947066165927,
+      "learning_rate": 0.00010035411290922594,
+      "loss": 1.3657,
+      "step": 9648
+    },
+    {
+      "epoch": 1.5438399999999999,
+      "grad_norm": 2.8232839485226533,
+      "learning_rate": 0.0001003368391757812,
+      "loss": 1.3032,
+      "step": 9649
+    },
+    {
+      "epoch": 1.544,
+      "grad_norm": 2.1179762909036075,
+      "learning_rate": 0.0001003195654322857,
+      "loss": 1.3354,
+      "step": 9650
+    },
+    {
+      "epoch": 1.54416,
+      "grad_norm": 2.2476627594654683,
+      "learning_rate": 0.00010030229167925482,
+      "loss": 1.3001,
+      "step": 9651
+    },
+    {
+      "epoch": 1.54432,
+      "grad_norm": 2.5856575005758544,
+      "learning_rate": 0.00010028501791720397,
+      "loss": 1.3565,
+      "step": 9652
+    },
+    {
+      "epoch": 1.54448,
+      "grad_norm": 2.393177643371751,
+      "learning_rate": 0.0001002677441466486,
+      "loss": 1.407,
+      "step": 9653
+    },
+    {
+      "epoch": 1.54464,
+      "grad_norm": 2.0917636527990418,
+      "learning_rate": 0.00010025047036810418,
+      "loss": 1.3557,
+      "step": 9654
+    },
+    {
+      "epoch": 1.5448,
+      "grad_norm": 2.1702654605816654,
+      "learning_rate": 0.00010023319658208606,
+      "loss": 1.3,
+      "step": 9655
+    },
+    {
+      "epoch": 1.54496,
+      "grad_norm": 1.9492974787251076,
+      "learning_rate": 0.0001002159227891097,
+      "loss": 1.2601,
+      "step": 9656
+    },
+    {
+      "epoch": 1.54512,
+      "grad_norm": 1.738046185918809,
+      "learning_rate": 0.00010019864898969051,
+      "loss": 1.3388,
+      "step": 9657
+    },
+    {
+      "epoch": 1.54528,
+      "grad_norm": 2.0530269424774183,
+      "learning_rate": 0.00010018137518434394,
+      "loss": 1.3127,
+      "step": 9658
+    },
+    {
+      "epoch": 1.5454400000000001,
+      "grad_norm": 2.433707516259465,
+      "learning_rate": 0.0001001641013735854,
+      "loss": 1.3015,
+      "step": 9659
+    },
+    {
+      "epoch": 1.5455999999999999,
+      "grad_norm": 2.286016119311383,
+      "learning_rate": 0.00010014682755793028,
+      "loss": 1.3228,
+      "step": 9660
+    },
+    {
+      "epoch": 1.54576,
+      "grad_norm": 2.313379424615351,
+      "learning_rate": 0.00010012955373789406,
+      "loss": 1.3302,
+      "step": 9661
+    },
+    {
+      "epoch": 1.54592,
+      "grad_norm": 2.3078029614569546,
+      "learning_rate": 0.00010011227991399215,
+      "loss": 1.3221,
+      "step": 9662
+    },
+    {
+      "epoch": 1.54608,
+      "grad_norm": 2.0465448275870606,
+      "learning_rate": 0.00010009500608673998,
+      "loss": 1.3776,
+      "step": 9663
+    },
+    {
+      "epoch": 1.54624,
+      "grad_norm": 1.8727418628318706,
+      "learning_rate": 0.00010007773225665296,
+      "loss": 1.3976,
+      "step": 9664
+    },
+    {
+      "epoch": 1.5464,
+      "grad_norm": 2.3591079253232037,
+      "learning_rate": 0.00010006045842424652,
+      "loss": 1.2821,
+      "step": 9665
+    },
+    {
+      "epoch": 1.54656,
+      "grad_norm": 1.8868186551441544,
+      "learning_rate": 0.00010004318459003608,
+      "loss": 1.2713,
+      "step": 9666
+    },
+    {
+      "epoch": 1.54672,
+      "grad_norm": 2.1028018962432706,
+      "learning_rate": 0.00010002591075453709,
+      "loss": 1.3361,
+      "step": 9667
+    },
+    {
+      "epoch": 1.54688,
+      "grad_norm": 2.3886448297145058,
+      "learning_rate": 0.00010000863691826494,
+      "loss": 1.3,
+      "step": 9668
+    },
+    {
+      "epoch": 1.54704,
+      "grad_norm": 2.2478738259058924,
+      "learning_rate": 9.999136308173509e-05,
+      "loss": 1.3469,
+      "step": 9669
+    },
+    {
+      "epoch": 1.5472000000000001,
+      "grad_norm": 1.7460573436084703,
+      "learning_rate": 9.997408924546295e-05,
+      "loss": 1.3257,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5473599999999998,
+      "grad_norm": 1.9598708242384753,
+      "learning_rate": 9.995681540996393e-05,
+      "loss": 1.3356,
+      "step": 9671
+    },
+    {
+      "epoch": 1.54752,
+      "grad_norm": 2.2844648417020705,
+      "learning_rate": 9.993954157575353e-05,
+      "loss": 1.2808,
+      "step": 9672
+    },
+    {
+      "epoch": 1.54768,
+      "grad_norm": 1.836328236086077,
+      "learning_rate": 9.992226774334706e-05,
+      "loss": 1.2962,
+      "step": 9673
+    },
+    {
+      "epoch": 1.5478399999999999,
+      "grad_norm": 2.2755017417359937,
+      "learning_rate": 9.990499391326007e-05,
+      "loss": 1.3861,
+      "step": 9674
+    },
+    {
+      "epoch": 1.548,
+      "grad_norm": 2.0404014711818133,
+      "learning_rate": 9.988772008600786e-05,
+      "loss": 1.3014,
+      "step": 9675
+    },
+    {
+      "epoch": 1.54816,
+      "grad_norm": 1.793316493614561,
+      "learning_rate": 9.987044626210595e-05,
+      "loss": 1.1949,
+      "step": 9676
+    },
+    {
+      "epoch": 1.54832,
+      "grad_norm": 2.1637108485400525,
+      "learning_rate": 9.985317244206975e-05,
+      "loss": 1.3055,
+      "step": 9677
+    },
+    {
+      "epoch": 1.54848,
+      "grad_norm": 1.9807707019374494,
+      "learning_rate": 9.983589862641465e-05,
+      "loss": 1.3048,
+      "step": 9678
+    },
+    {
+      "epoch": 1.54864,
+      "grad_norm": 2.1989783737025563,
+      "learning_rate": 9.981862481565608e-05,
+      "loss": 1.3038,
+      "step": 9679
+    },
+    {
+      "epoch": 1.5488,
+      "grad_norm": 1.9973974553671356,
+      "learning_rate": 9.980135101030952e-05,
+      "loss": 1.255,
+      "step": 9680
+    },
+    {
+      "epoch": 1.5489600000000001,
+      "grad_norm": 2.0091480410369384,
+      "learning_rate": 9.978407721089032e-05,
+      "loss": 1.4033,
+      "step": 9681
+    },
+    {
+      "epoch": 1.54912,
+      "grad_norm": 2.3024913855577505,
+      "learning_rate": 9.976680341791398e-05,
+      "loss": 1.3322,
+      "step": 9682
+    },
+    {
+      "epoch": 1.54928,
+      "grad_norm": 2.449624057411108,
+      "learning_rate": 9.974952963189585e-05,
+      "loss": 1.3863,
+      "step": 9683
+    },
+    {
+      "epoch": 1.5494400000000002,
+      "grad_norm": 1.9294600512279612,
+      "learning_rate": 9.97322558533514e-05,
+      "loss": 1.2631,
+      "step": 9684
+    },
+    {
+      "epoch": 1.5495999999999999,
+      "grad_norm": 2.010029686885603,
+      "learning_rate": 9.971498208279606e-05,
+      "loss": 1.3387,
+      "step": 9685
+    },
+    {
+      "epoch": 1.54976,
+      "grad_norm": 2.027672492453531,
+      "learning_rate": 9.969770832074523e-05,
+      "loss": 1.3546,
+      "step": 9686
+    },
+    {
+      "epoch": 1.54992,
+      "grad_norm": 1.989915616691982,
+      "learning_rate": 9.968043456771431e-05,
+      "loss": 1.2899,
+      "step": 9687
+    },
+    {
+      "epoch": 1.55008,
+      "grad_norm": 2.0051488120967114,
+      "learning_rate": 9.966316082421882e-05,
+      "loss": 1.2938,
+      "step": 9688
+    },
+    {
+      "epoch": 1.55024,
+      "grad_norm": 2.194168056247239,
+      "learning_rate": 9.964588709077409e-05,
+      "loss": 1.314,
+      "step": 9689
+    },
+    {
+      "epoch": 1.5504,
+      "grad_norm": 1.9886721822191278,
+      "learning_rate": 9.962861336789558e-05,
+      "loss": 1.2947,
+      "step": 9690
+    },
+    {
+      "epoch": 1.55056,
+      "grad_norm": 2.1005590702713,
+      "learning_rate": 9.961133965609869e-05,
+      "loss": 1.4009,
+      "step": 9691
+    },
+    {
+      "epoch": 1.55072,
+      "grad_norm": 2.0376248870206295,
+      "learning_rate": 9.959406595589886e-05,
+      "loss": 1.3217,
+      "step": 9692
+    },
+    {
+      "epoch": 1.55088,
+      "grad_norm": 2.482209071868788,
+      "learning_rate": 9.957679226781152e-05,
+      "loss": 1.3715,
+      "step": 9693
+    },
+    {
+      "epoch": 1.55104,
+      "grad_norm": 1.9188721780272062,
+      "learning_rate": 9.955951859235209e-05,
+      "loss": 1.3153,
+      "step": 9694
+    },
+    {
+      "epoch": 1.5512000000000001,
+      "grad_norm": 2.2326379086415695,
+      "learning_rate": 9.954224493003596e-05,
+      "loss": 1.2678,
+      "step": 9695
+    },
+    {
+      "epoch": 1.5513599999999999,
+      "grad_norm": 2.472505898821232,
+      "learning_rate": 9.95249712813786e-05,
+      "loss": 1.3364,
+      "step": 9696
+    },
+    {
+      "epoch": 1.55152,
+      "grad_norm": 1.927031872793261,
+      "learning_rate": 9.950769764689538e-05,
+      "loss": 1.2663,
+      "step": 9697
+    },
+    {
+      "epoch": 1.55168,
+      "grad_norm": 1.9640505097452314,
+      "learning_rate": 9.949042402710173e-05,
+      "loss": 1.3223,
+      "step": 9698
+    },
+    {
+      "epoch": 1.5518399999999999,
+      "grad_norm": 2.107513913692691,
+      "learning_rate": 9.947315042251314e-05,
+      "loss": 1.3591,
+      "step": 9699
+    },
+    {
+      "epoch": 1.552,
+      "grad_norm": 2.4589608879592477,
+      "learning_rate": 9.945587683364492e-05,
+      "loss": 1.3858,
+      "step": 9700
+    },
+    {
+      "epoch": 1.55216,
+      "grad_norm": 1.9338541222914991,
+      "learning_rate": 9.94386032610126e-05,
+      "loss": 1.3083,
+      "step": 9701
+    },
+    {
+      "epoch": 1.55232,
+      "grad_norm": 2.3928004164121957,
+      "learning_rate": 9.942132970513148e-05,
+      "loss": 1.3014,
+      "step": 9702
+    },
+    {
+      "epoch": 1.55248,
+      "grad_norm": 1.9679258694470843,
+      "learning_rate": 9.940405616651707e-05,
+      "loss": 1.2964,
+      "step": 9703
+    },
+    {
+      "epoch": 1.55264,
+      "grad_norm": 1.988064948758037,
+      "learning_rate": 9.938678264568474e-05,
+      "loss": 1.31,
+      "step": 9704
+    },
+    {
+      "epoch": 1.5528,
+      "grad_norm": 2.434500779682388,
+      "learning_rate": 9.936950914314995e-05,
+      "loss": 1.3624,
+      "step": 9705
+    },
+    {
+      "epoch": 1.5529600000000001,
+      "grad_norm": 1.8428271739560156,
+      "learning_rate": 9.935223565942805e-05,
+      "loss": 1.3571,
+      "step": 9706
+    },
+    {
+      "epoch": 1.5531199999999998,
+      "grad_norm": 2.215289107476072,
+      "learning_rate": 9.933496219503455e-05,
+      "loss": 1.3272,
+      "step": 9707
+    },
+    {
+      "epoch": 1.55328,
+      "grad_norm": 1.968144500832969,
+      "learning_rate": 9.931768875048476e-05,
+      "loss": 1.3361,
+      "step": 9708
+    },
+    {
+      "epoch": 1.5534400000000002,
+      "grad_norm": 2.101204749913191,
+      "learning_rate": 9.930041532629418e-05,
+      "loss": 1.3314,
+      "step": 9709
+    },
+    {
+      "epoch": 1.5535999999999999,
+      "grad_norm": 2.0258968161239275,
+      "learning_rate": 9.928314192297815e-05,
+      "loss": 1.2941,
+      "step": 9710
+    },
+    {
+      "epoch": 1.55376,
+      "grad_norm": 1.8149100741734345,
+      "learning_rate": 9.926586854105214e-05,
+      "loss": 1.3734,
+      "step": 9711
+    },
+    {
+      "epoch": 1.55392,
+      "grad_norm": 1.9183947919623523,
+      "learning_rate": 9.924859518103152e-05,
+      "loss": 1.3344,
+      "step": 9712
+    },
+    {
+      "epoch": 1.55408,
+      "grad_norm": 1.876058265878141,
+      "learning_rate": 9.923132184343177e-05,
+      "loss": 1.2703,
+      "step": 9713
+    },
+    {
+      "epoch": 1.55424,
+      "grad_norm": 2.598287348790994,
+      "learning_rate": 9.92140485287682e-05,
+      "loss": 1.3039,
+      "step": 9714
+    },
+    {
+      "epoch": 1.5544,
+      "grad_norm": 2.2662406576892264,
+      "learning_rate": 9.919677523755635e-05,
+      "loss": 1.3354,
+      "step": 9715
+    },
+    {
+      "epoch": 1.55456,
+      "grad_norm": 2.2406379736594726,
+      "learning_rate": 9.917950197031149e-05,
+      "loss": 1.2617,
+      "step": 9716
+    },
+    {
+      "epoch": 1.55472,
+      "grad_norm": 1.9087210761030327,
+      "learning_rate": 9.916222872754915e-05,
+      "loss": 1.2639,
+      "step": 9717
+    },
+    {
+      "epoch": 1.55488,
+      "grad_norm": 2.0076815980478226,
+      "learning_rate": 9.914495550978463e-05,
+      "loss": 1.3088,
+      "step": 9718
+    },
+    {
+      "epoch": 1.55504,
+      "grad_norm": 2.097730682490871,
+      "learning_rate": 9.912768231753342e-05,
+      "loss": 1.38,
+      "step": 9719
+    },
+    {
+      "epoch": 1.5552000000000001,
+      "grad_norm": 1.9918686350952142,
+      "learning_rate": 9.911040915131087e-05,
+      "loss": 1.3342,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5553599999999999,
+      "grad_norm": 1.9770679206392572,
+      "learning_rate": 9.909313601163247e-05,
+      "loss": 1.358,
+      "step": 9721
+    },
+    {
+      "epoch": 1.55552,
+      "grad_norm": 2.6537234589222423,
+      "learning_rate": 9.907586289901351e-05,
+      "loss": 1.334,
+      "step": 9722
+    },
+    {
+      "epoch": 1.55568,
+      "grad_norm": 2.3274503971447564,
+      "learning_rate": 9.905858981396951e-05,
+      "loss": 1.3109,
+      "step": 9723
+    },
+    {
+      "epoch": 1.55584,
+      "grad_norm": 2.8689643464252206,
+      "learning_rate": 9.904131675701577e-05,
+      "loss": 1.3026,
+      "step": 9724
+    },
+    {
+      "epoch": 1.556,
+      "grad_norm": 2.0035687437427407,
+      "learning_rate": 9.902404372866777e-05,
+      "loss": 1.3579,
+      "step": 9725
+    },
+    {
+      "epoch": 1.55616,
+      "grad_norm": 2.488775005042439,
+      "learning_rate": 9.900677072944087e-05,
+      "loss": 1.2609,
+      "step": 9726
+    },
+    {
+      "epoch": 1.55632,
+      "grad_norm": 2.4755141979559725,
+      "learning_rate": 9.89894977598505e-05,
+      "loss": 1.4212,
+      "step": 9727
+    },
+    {
+      "epoch": 1.55648,
+      "grad_norm": 2.196022405371127,
+      "learning_rate": 9.897222482041202e-05,
+      "loss": 1.3359,
+      "step": 9728
+    },
+    {
+      "epoch": 1.55664,
+      "grad_norm": 1.9146738680598803,
+      "learning_rate": 9.895495191164089e-05,
+      "loss": 1.2975,
+      "step": 9729
+    },
+    {
+      "epoch": 1.5568,
+      "grad_norm": 1.9880406597532754,
+      "learning_rate": 9.893767903405243e-05,
+      "loss": 1.2902,
+      "step": 9730
+    },
+    {
+      "epoch": 1.5569600000000001,
+      "grad_norm": 2.1328929744952965,
+      "learning_rate": 9.892040618816212e-05,
+      "loss": 1.3562,
+      "step": 9731
+    },
+    {
+      "epoch": 1.5571199999999998,
+      "grad_norm": 2.0136027459535812,
+      "learning_rate": 9.890313337448528e-05,
+      "loss": 1.3186,
+      "step": 9732
+    },
+    {
+      "epoch": 1.55728,
+      "grad_norm": 1.9574792196516666,
+      "learning_rate": 9.888586059353736e-05,
+      "loss": 1.3463,
+      "step": 9733
+    },
+    {
+      "epoch": 1.5574400000000002,
+      "grad_norm": 1.9847211871658292,
+      "learning_rate": 9.886858784583374e-05,
+      "loss": 1.3148,
+      "step": 9734
+    },
+    {
+      "epoch": 1.5575999999999999,
+      "grad_norm": 2.7754437622838104,
+      "learning_rate": 9.88513151318898e-05,
+      "loss": 1.3475,
+      "step": 9735
+    },
+    {
+      "epoch": 1.55776,
+      "grad_norm": 1.89369077639806,
+      "learning_rate": 9.883404245222093e-05,
+      "loss": 1.3323,
+      "step": 9736
+    },
+    {
+      "epoch": 1.55792,
+      "grad_norm": 2.3727784534686434,
+      "learning_rate": 9.881676980734258e-05,
+      "loss": 1.3545,
+      "step": 9737
+    },
+    {
+      "epoch": 1.55808,
+      "grad_norm": 2.0431403699570567,
+      "learning_rate": 9.879949719777004e-05,
+      "loss": 1.3668,
+      "step": 9738
+    },
+    {
+      "epoch": 1.55824,
+      "grad_norm": 2.071017444738438,
+      "learning_rate": 9.87822246240188e-05,
+      "loss": 1.2589,
+      "step": 9739
+    },
+    {
+      "epoch": 1.5584,
+      "grad_norm": 2.1742382647092495,
+      "learning_rate": 9.876495208660417e-05,
+      "loss": 1.3184,
+      "step": 9740
+    },
+    {
+      "epoch": 1.55856,
+      "grad_norm": 2.1226828307445613,
+      "learning_rate": 9.874767958604158e-05,
+      "loss": 1.3045,
+      "step": 9741
+    },
+    {
+      "epoch": 1.55872,
+      "grad_norm": 2.370398302521163,
+      "learning_rate": 9.873040712284642e-05,
+      "loss": 1.3343,
+      "step": 9742
+    },
+    {
+      "epoch": 1.55888,
+      "grad_norm": 1.878043867892099,
+      "learning_rate": 9.871313469753405e-05,
+      "loss": 1.241,
+      "step": 9743
+    },
+    {
+      "epoch": 1.55904,
+      "grad_norm": 2.1911852578571467,
+      "learning_rate": 9.869586231061984e-05,
+      "loss": 1.2501,
+      "step": 9744
+    },
+    {
+      "epoch": 1.5592000000000001,
+      "grad_norm": 2.089843003082643,
+      "learning_rate": 9.867858996261925e-05,
+      "loss": 1.3265,
+      "step": 9745
+    },
+    {
+      "epoch": 1.5593599999999999,
+      "grad_norm": 1.9950390891205465,
+      "learning_rate": 9.866131765404757e-05,
+      "loss": 1.3113,
+      "step": 9746
+    },
+    {
+      "epoch": 1.55952,
+      "grad_norm": 1.838856360915904,
+      "learning_rate": 9.864404538542026e-05,
+      "loss": 1.3242,
+      "step": 9747
+    },
+    {
+      "epoch": 1.55968,
+      "grad_norm": 2.1474940085260705,
+      "learning_rate": 9.862677315725262e-05,
+      "loss": 1.3303,
+      "step": 9748
+    },
+    {
+      "epoch": 1.55984,
+      "grad_norm": 2.2698450308235074,
+      "learning_rate": 9.860950097006008e-05,
+      "loss": 1.3502,
+      "step": 9749
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 2.078750511997093,
+      "learning_rate": 9.859222882435802e-05,
+      "loss": 1.3267,
+      "step": 9750
+    },
+    {
+      "epoch": 1.56016,
+      "grad_norm": 2.5317674836272026,
+      "learning_rate": 9.857495672066178e-05,
+      "loss": 1.3019,
+      "step": 9751
+    },
+    {
+      "epoch": 1.56032,
+      "grad_norm": 2.2786532483547908,
+      "learning_rate": 9.855768465948675e-05,
+      "loss": 1.3938,
+      "step": 9752
+    },
+    {
+      "epoch": 1.56048,
+      "grad_norm": 2.684432856745637,
+      "learning_rate": 9.854041264134834e-05,
+      "loss": 1.2766,
+      "step": 9753
+    },
+    {
+      "epoch": 1.56064,
+      "grad_norm": 2.138306536317449,
+      "learning_rate": 9.852314066676185e-05,
+      "loss": 1.3023,
+      "step": 9754
+    },
+    {
+      "epoch": 1.5608,
+      "grad_norm": 1.9755854696196882,
+      "learning_rate": 9.850586873624274e-05,
+      "loss": 1.3192,
+      "step": 9755
+    },
+    {
+      "epoch": 1.5609600000000001,
+      "grad_norm": 2.491152517323662,
+      "learning_rate": 9.848859685030628e-05,
+      "loss": 1.3364,
+      "step": 9756
+    },
+    {
+      "epoch": 1.5611199999999998,
+      "grad_norm": 2.034504441577449,
+      "learning_rate": 9.847132500946788e-05,
+      "loss": 1.3931,
+      "step": 9757
+    },
+    {
+      "epoch": 1.56128,
+      "grad_norm": 1.938319024160232,
+      "learning_rate": 9.845405321424298e-05,
+      "loss": 1.2956,
+      "step": 9758
+    },
+    {
+      "epoch": 1.5614400000000002,
+      "grad_norm": 2.2358033166475044,
+      "learning_rate": 9.843678146514682e-05,
+      "loss": 1.3482,
+      "step": 9759
+    },
+    {
+      "epoch": 1.5615999999999999,
+      "grad_norm": 2.0862172431834725,
+      "learning_rate": 9.841950976269485e-05,
+      "loss": 1.2904,
+      "step": 9760
+    },
+    {
+      "epoch": 1.56176,
+      "grad_norm": 2.172794669717783,
+      "learning_rate": 9.840223810740238e-05,
+      "loss": 1.4007,
+      "step": 9761
+    },
+    {
+      "epoch": 1.56192,
+      "grad_norm": 1.7981027675847272,
+      "learning_rate": 9.838496649978485e-05,
+      "loss": 1.3327,
+      "step": 9762
+    },
+    {
+      "epoch": 1.56208,
+      "grad_norm": 2.0514387329620987,
+      "learning_rate": 9.836769494035751e-05,
+      "loss": 1.3094,
+      "step": 9763
+    },
+    {
+      "epoch": 1.56224,
+      "grad_norm": 1.8884499086907183,
+      "learning_rate": 9.835042342963583e-05,
+      "loss": 1.2698,
+      "step": 9764
+    },
+    {
+      "epoch": 1.5624,
+      "grad_norm": 2.1273670713791213,
+      "learning_rate": 9.833315196813505e-05,
+      "loss": 1.3552,
+      "step": 9765
+    },
+    {
+      "epoch": 1.56256,
+      "grad_norm": 2.0094419649699335,
+      "learning_rate": 9.831588055637064e-05,
+      "loss": 1.3493,
+      "step": 9766
+    },
+    {
+      "epoch": 1.56272,
+      "grad_norm": 2.027415725982128,
+      "learning_rate": 9.829860919485786e-05,
+      "loss": 1.3529,
+      "step": 9767
+    },
+    {
+      "epoch": 1.56288,
+      "grad_norm": 1.9217516727230772,
+      "learning_rate": 9.828133788411212e-05,
+      "loss": 1.3474,
+      "step": 9768
+    },
+    {
+      "epoch": 1.56304,
+      "grad_norm": 1.9765455029144576,
+      "learning_rate": 9.826406662464874e-05,
+      "loss": 1.2643,
+      "step": 9769
+    },
+    {
+      "epoch": 1.5632000000000001,
+      "grad_norm": 1.9982679581651819,
+      "learning_rate": 9.824679541698311e-05,
+      "loss": 1.2649,
+      "step": 9770
+    },
+    {
+      "epoch": 1.5633599999999999,
+      "grad_norm": 2.049785406321914,
+      "learning_rate": 9.822952426163052e-05,
+      "loss": 1.3013,
+      "step": 9771
+    },
+    {
+      "epoch": 1.56352,
+      "grad_norm": 1.9875802732360732,
+      "learning_rate": 9.821225315910639e-05,
+      "loss": 1.3673,
+      "step": 9772
+    },
+    {
+      "epoch": 1.56368,
+      "grad_norm": 2.4335132573797384,
+      "learning_rate": 9.819498210992597e-05,
+      "loss": 1.3579,
+      "step": 9773
+    },
+    {
+      "epoch": 1.56384,
+      "grad_norm": 2.133928256928764,
+      "learning_rate": 9.817771111460469e-05,
+      "loss": 1.3732,
+      "step": 9774
+    },
+    {
+      "epoch": 1.564,
+      "grad_norm": 2.0734802162081345,
+      "learning_rate": 9.816044017365783e-05,
+      "loss": 1.374,
+      "step": 9775
+    },
+    {
+      "epoch": 1.56416,
+      "grad_norm": 2.2424710986500282,
+      "learning_rate": 9.814316928760077e-05,
+      "loss": 1.2952,
+      "step": 9776
+    },
+    {
+      "epoch": 1.56432,
+      "grad_norm": 2.0418288140256737,
+      "learning_rate": 9.81258984569488e-05,
+      "loss": 1.3037,
+      "step": 9777
+    },
+    {
+      "epoch": 1.56448,
+      "grad_norm": 2.6428093771477075,
+      "learning_rate": 9.810862768221735e-05,
+      "loss": 1.3175,
+      "step": 9778
+    },
+    {
+      "epoch": 1.56464,
+      "grad_norm": 2.1703690661487425,
+      "learning_rate": 9.809135696392164e-05,
+      "loss": 1.3716,
+      "step": 9779
+    },
+    {
+      "epoch": 1.5648,
+      "grad_norm": 2.7563782968372315,
+      "learning_rate": 9.80740863025771e-05,
+      "loss": 1.4132,
+      "step": 9780
+    },
+    {
+      "epoch": 1.5649600000000001,
+      "grad_norm": 2.0201464197903642,
+      "learning_rate": 9.805681569869897e-05,
+      "loss": 1.3867,
+      "step": 9781
+    },
+    {
+      "epoch": 1.5651199999999998,
+      "grad_norm": 2.4150363721188537,
+      "learning_rate": 9.803954515280269e-05,
+      "loss": 1.291,
+      "step": 9782
+    },
+    {
+      "epoch": 1.56528,
+      "grad_norm": 3.0303641737924716,
+      "learning_rate": 9.802227466540346e-05,
+      "loss": 1.314,
+      "step": 9783
+    },
+    {
+      "epoch": 1.5654400000000002,
+      "grad_norm": 2.422860349676599,
+      "learning_rate": 9.800500423701672e-05,
+      "loss": 1.374,
+      "step": 9784
+    },
+    {
+      "epoch": 1.5655999999999999,
+      "grad_norm": 2.7105003332292243,
+      "learning_rate": 9.79877338681577e-05,
+      "loss": 1.3158,
+      "step": 9785
+    },
+    {
+      "epoch": 1.56576,
+      "grad_norm": 2.193712609777907,
+      "learning_rate": 9.797046355934183e-05,
+      "loss": 1.3522,
+      "step": 9786
+    },
+    {
+      "epoch": 1.56592,
+      "grad_norm": 2.2959025681792964,
+      "learning_rate": 9.795319331108432e-05,
+      "loss": 1.3199,
+      "step": 9787
+    },
+    {
+      "epoch": 1.56608,
+      "grad_norm": 2.2936737834945773,
+      "learning_rate": 9.793592312390058e-05,
+      "loss": 1.3373,
+      "step": 9788
+    },
+    {
+      "epoch": 1.56624,
+      "grad_norm": 2.3355682856942637,
+      "learning_rate": 9.791865299830585e-05,
+      "loss": 1.3409,
+      "step": 9789
+    },
+    {
+      "epoch": 1.5664,
+      "grad_norm": 1.7772740036073542,
+      "learning_rate": 9.790138293481551e-05,
+      "loss": 1.3241,
+      "step": 9790
+    },
+    {
+      "epoch": 1.56656,
+      "grad_norm": 2.088234598865119,
+      "learning_rate": 9.788411293394482e-05,
+      "loss": 1.4407,
+      "step": 9791
+    },
+    {
+      "epoch": 1.5667200000000001,
+      "grad_norm": 2.0705322861265216,
+      "learning_rate": 9.786684299620914e-05,
+      "loss": 1.287,
+      "step": 9792
+    },
+    {
+      "epoch": 1.56688,
+      "grad_norm": 1.9215042263728335,
+      "learning_rate": 9.784957312212371e-05,
+      "loss": 1.3403,
+      "step": 9793
+    },
+    {
+      "epoch": 1.56704,
+      "grad_norm": 2.1702779851678726,
+      "learning_rate": 9.783230331220395e-05,
+      "loss": 1.3795,
+      "step": 9794
+    },
+    {
+      "epoch": 1.5672000000000001,
+      "grad_norm": 2.2431179339599745,
+      "learning_rate": 9.781503356696504e-05,
+      "loss": 1.3246,
+      "step": 9795
+    },
+    {
+      "epoch": 1.5673599999999999,
+      "grad_norm": 2.451410227206375,
+      "learning_rate": 9.77977638869224e-05,
+      "loss": 1.3236,
+      "step": 9796
+    },
+    {
+      "epoch": 1.56752,
+      "grad_norm": 2.1984445318983514,
+      "learning_rate": 9.778049427259123e-05,
+      "loss": 1.26,
+      "step": 9797
+    },
+    {
+      "epoch": 1.56768,
+      "grad_norm": 1.9130734263216866,
+      "learning_rate": 9.776322472448692e-05,
+      "loss": 1.2933,
+      "step": 9798
+    },
+    {
+      "epoch": 1.56784,
+      "grad_norm": 2.4494469018540936,
+      "learning_rate": 9.774595524312469e-05,
+      "loss": 1.321,
+      "step": 9799
+    },
+    {
+      "epoch": 1.568,
+      "grad_norm": 2.6975332739526636,
+      "learning_rate": 9.772868582901989e-05,
+      "loss": 1.3344,
+      "step": 9800
+    },
+    {
+      "epoch": 1.56816,
+      "grad_norm": 2.3464147430686815,
+      "learning_rate": 9.771141648268776e-05,
+      "loss": 1.3392,
+      "step": 9801
+    },
+    {
+      "epoch": 1.56832,
+      "grad_norm": 2.4227534816243494,
+      "learning_rate": 9.769414720464369e-05,
+      "loss": 1.3953,
+      "step": 9802
+    },
+    {
+      "epoch": 1.56848,
+      "grad_norm": 2.341492202630362,
+      "learning_rate": 9.767687799540285e-05,
+      "loss": 1.3227,
+      "step": 9803
+    },
+    {
+      "epoch": 1.56864,
+      "grad_norm": 1.794483944959297,
+      "learning_rate": 9.765960885548062e-05,
+      "loss": 1.3134,
+      "step": 9804
+    },
+    {
+      "epoch": 1.5688,
+      "grad_norm": 2.1103555588138816,
+      "learning_rate": 9.764233978539222e-05,
+      "loss": 1.2241,
+      "step": 9805
+    },
+    {
+      "epoch": 1.5689600000000001,
+      "grad_norm": 2.5856255574850673,
+      "learning_rate": 9.762507078565302e-05,
+      "loss": 1.3492,
+      "step": 9806
+    },
+    {
+      "epoch": 1.5691199999999998,
+      "grad_norm": 2.1235380212661314,
+      "learning_rate": 9.760780185677819e-05,
+      "loss": 1.3331,
+      "step": 9807
+    },
+    {
+      "epoch": 1.56928,
+      "grad_norm": 1.9813676641561706,
+      "learning_rate": 9.759053299928309e-05,
+      "loss": 1.3201,
+      "step": 9808
+    },
+    {
+      "epoch": 1.56944,
+      "grad_norm": 2.8885281365861126,
+      "learning_rate": 9.757326421368295e-05,
+      "loss": 1.3488,
+      "step": 9809
+    },
+    {
+      "epoch": 1.5695999999999999,
+      "grad_norm": 1.9169108005412814,
+      "learning_rate": 9.755599550049313e-05,
+      "loss": 1.241,
+      "step": 9810
+    },
+    {
+      "epoch": 1.56976,
+      "grad_norm": 2.5824726565779414,
+      "learning_rate": 9.753872686022878e-05,
+      "loss": 1.248,
+      "step": 9811
+    },
+    {
+      "epoch": 1.56992,
+      "grad_norm": 2.676185885551868,
+      "learning_rate": 9.752145829340529e-05,
+      "loss": 1.3891,
+      "step": 9812
+    },
+    {
+      "epoch": 1.57008,
+      "grad_norm": 2.090269255777329,
+      "learning_rate": 9.750418980053782e-05,
+      "loss": 1.3332,
+      "step": 9813
+    },
+    {
+      "epoch": 1.57024,
+      "grad_norm": 3.043826440102301,
+      "learning_rate": 9.748692138214173e-05,
+      "loss": 1.3142,
+      "step": 9814
+    },
+    {
+      "epoch": 1.5704,
+      "grad_norm": 2.088701426731462,
+      "learning_rate": 9.74696530387322e-05,
+      "loss": 1.3717,
+      "step": 9815
+    },
+    {
+      "epoch": 1.57056,
+      "grad_norm": 2.323431170038478,
+      "learning_rate": 9.745238477082455e-05,
+      "loss": 1.3174,
+      "step": 9816
+    },
+    {
+      "epoch": 1.5707200000000001,
+      "grad_norm": 2.328503013145964,
+      "learning_rate": 9.743511657893405e-05,
+      "loss": 1.3802,
+      "step": 9817
+    },
+    {
+      "epoch": 1.57088,
+      "grad_norm": 2.030051070374394,
+      "learning_rate": 9.74178484635759e-05,
+      "loss": 1.3346,
+      "step": 9818
+    },
+    {
+      "epoch": 1.57104,
+      "grad_norm": 2.3021199641170944,
+      "learning_rate": 9.740058042526544e-05,
+      "loss": 1.323,
+      "step": 9819
+    },
+    {
+      "epoch": 1.5712000000000002,
+      "grad_norm": 2.0607592383721824,
+      "learning_rate": 9.738331246451783e-05,
+      "loss": 1.3048,
+      "step": 9820
+    },
+    {
+      "epoch": 1.5713599999999999,
+      "grad_norm": 2.2225326873545703,
+      "learning_rate": 9.736604458184839e-05,
+      "loss": 1.3239,
+      "step": 9821
+    },
+    {
+      "epoch": 1.57152,
+      "grad_norm": 2.1656989260377904,
+      "learning_rate": 9.734877677777232e-05,
+      "loss": 1.3279,
+      "step": 9822
+    },
+    {
+      "epoch": 1.57168,
+      "grad_norm": 2.145990066579529,
+      "learning_rate": 9.733150905280492e-05,
+      "loss": 1.3056,
+      "step": 9823
+    },
+    {
+      "epoch": 1.57184,
+      "grad_norm": 2.7607509576338045,
+      "learning_rate": 9.731424140746136e-05,
+      "loss": 1.3159,
+      "step": 9824
+    },
+    {
+      "epoch": 1.572,
+      "grad_norm": 2.117118344017946,
+      "learning_rate": 9.729697384225694e-05,
+      "loss": 1.3283,
+      "step": 9825
+    },
+    {
+      "epoch": 1.57216,
+      "grad_norm": 2.8135687144370025,
+      "learning_rate": 9.727970635770686e-05,
+      "loss": 1.3103,
+      "step": 9826
+    },
+    {
+      "epoch": 1.57232,
+      "grad_norm": 2.352932351057987,
+      "learning_rate": 9.726243895432642e-05,
+      "loss": 1.256,
+      "step": 9827
+    },
+    {
+      "epoch": 1.57248,
+      "grad_norm": 1.9642484388124697,
+      "learning_rate": 9.724517163263078e-05,
+      "loss": 1.2852,
+      "step": 9828
+    },
+    {
+      "epoch": 1.57264,
+      "grad_norm": 2.6250828842663263,
+      "learning_rate": 9.722790439313525e-05,
+      "loss": 1.2839,
+      "step": 9829
+    },
+    {
+      "epoch": 1.5728,
+      "grad_norm": 1.9599052544997968,
+      "learning_rate": 9.721063723635496e-05,
+      "loss": 1.2903,
+      "step": 9830
+    },
+    {
+      "epoch": 1.5729600000000001,
+      "grad_norm": 1.9718503938648124,
+      "learning_rate": 9.719337016280524e-05,
+      "loss": 1.259,
+      "step": 9831
+    },
+    {
+      "epoch": 1.5731199999999999,
+      "grad_norm": 2.1743619997647414,
+      "learning_rate": 9.717610317300121e-05,
+      "loss": 1.2851,
+      "step": 9832
+    },
+    {
+      "epoch": 1.57328,
+      "grad_norm": 2.118866280400504,
+      "learning_rate": 9.715883626745819e-05,
+      "loss": 1.3877,
+      "step": 9833
+    },
+    {
+      "epoch": 1.57344,
+      "grad_norm": 2.2316335311718705,
+      "learning_rate": 9.714156944669131e-05,
+      "loss": 1.3599,
+      "step": 9834
+    },
+    {
+      "epoch": 1.5735999999999999,
+      "grad_norm": 1.648347249600059,
+      "learning_rate": 9.712430271121589e-05,
+      "loss": 1.2761,
+      "step": 9835
+    },
+    {
+      "epoch": 1.57376,
+      "grad_norm": 2.0264071759549696,
+      "learning_rate": 9.710703606154704e-05,
+      "loss": 1.3506,
+      "step": 9836
+    },
+    {
+      "epoch": 1.57392,
+      "grad_norm": 2.064376687775498,
+      "learning_rate": 9.708976949820005e-05,
+      "loss": 1.2595,
+      "step": 9837
+    },
+    {
+      "epoch": 1.57408,
+      "grad_norm": 2.2916225404075568,
+      "learning_rate": 9.707250302169008e-05,
+      "loss": 1.2961,
+      "step": 9838
+    },
+    {
+      "epoch": 1.57424,
+      "grad_norm": 1.9077091238389747,
+      "learning_rate": 9.705523663253237e-05,
+      "loss": 1.3315,
+      "step": 9839
+    },
+    {
+      "epoch": 1.5744,
+      "grad_norm": 2.6705942477323714,
+      "learning_rate": 9.703797033124207e-05,
+      "loss": 1.338,
+      "step": 9840
+    },
+    {
+      "epoch": 1.57456,
+      "grad_norm": 2.3159243563031717,
+      "learning_rate": 9.702070411833446e-05,
+      "loss": 1.3098,
+      "step": 9841
+    },
+    {
+      "epoch": 1.5747200000000001,
+      "grad_norm": 2.1135865942419496,
+      "learning_rate": 9.700343799432467e-05,
+      "loss": 1.3217,
+      "step": 9842
+    },
+    {
+      "epoch": 1.5748799999999998,
+      "grad_norm": 2.253548853003411,
+      "learning_rate": 9.698617195972796e-05,
+      "loss": 1.2665,
+      "step": 9843
+    },
+    {
+      "epoch": 1.57504,
+      "grad_norm": 2.330038354473096,
+      "learning_rate": 9.696890601505944e-05,
+      "loss": 1.3452,
+      "step": 9844
+    },
+    {
+      "epoch": 1.5752000000000002,
+      "grad_norm": 1.9968733726718562,
+      "learning_rate": 9.69516401608344e-05,
+      "loss": 1.3602,
+      "step": 9845
+    },
+    {
+      "epoch": 1.5753599999999999,
+      "grad_norm": 2.192275383454218,
+      "learning_rate": 9.693437439756792e-05,
+      "loss": 1.3064,
+      "step": 9846
+    },
+    {
+      "epoch": 1.57552,
+      "grad_norm": 2.4195115388537363,
+      "learning_rate": 9.69171087257753e-05,
+      "loss": 1.3245,
+      "step": 9847
+    },
+    {
+      "epoch": 1.57568,
+      "grad_norm": 1.920421686660415,
+      "learning_rate": 9.689984314597161e-05,
+      "loss": 1.2829,
+      "step": 9848
+    },
+    {
+      "epoch": 1.57584,
+      "grad_norm": 1.9572202488902726,
+      "learning_rate": 9.688257765867212e-05,
+      "loss": 1.2923,
+      "step": 9849
+    },
+    {
+      "epoch": 1.576,
+      "grad_norm": 3.2864009677977886,
+      "learning_rate": 9.686531226439193e-05,
+      "loss": 1.3656,
+      "step": 9850
+    },
+    {
+      "epoch": 1.57616,
+      "grad_norm": 2.0154434593524777,
+      "learning_rate": 9.68480469636463e-05,
+      "loss": 1.3134,
+      "step": 9851
+    },
+    {
+      "epoch": 1.57632,
+      "grad_norm": 2.297770083779517,
+      "learning_rate": 9.683078175695032e-05,
+      "loss": 1.2651,
+      "step": 9852
+    },
+    {
+      "epoch": 1.57648,
+      "grad_norm": 1.8776418355545532,
+      "learning_rate": 9.681351664481922e-05,
+      "loss": 1.251,
+      "step": 9853
+    },
+    {
+      "epoch": 1.57664,
+      "grad_norm": 2.008575492087276,
+      "learning_rate": 9.679625162776811e-05,
+      "loss": 1.3219,
+      "step": 9854
+    },
+    {
+      "epoch": 1.5768,
+      "grad_norm": 2.4352041311068873,
+      "learning_rate": 9.677898670631224e-05,
+      "loss": 1.3915,
+      "step": 9855
+    },
+    {
+      "epoch": 1.5769600000000001,
+      "grad_norm": 2.322405732883078,
+      "learning_rate": 9.676172188096666e-05,
+      "loss": 1.4297,
+      "step": 9856
+    },
+    {
+      "epoch": 1.5771199999999999,
+      "grad_norm": 2.297887789299443,
+      "learning_rate": 9.67444571522466e-05,
+      "loss": 1.2668,
+      "step": 9857
+    },
+    {
+      "epoch": 1.57728,
+      "grad_norm": 2.0398463154983797,
+      "learning_rate": 9.672719252066717e-05,
+      "loss": 1.3218,
+      "step": 9858
+    },
+    {
+      "epoch": 1.57744,
+      "grad_norm": 1.910808114069276,
+      "learning_rate": 9.670992798674359e-05,
+      "loss": 1.3382,
+      "step": 9859
+    },
+    {
+      "epoch": 1.5776,
+      "grad_norm": 2.275001216732449,
+      "learning_rate": 9.669266355099092e-05,
+      "loss": 1.3318,
+      "step": 9860
+    },
+    {
+      "epoch": 1.57776,
+      "grad_norm": 1.9592772701620442,
+      "learning_rate": 9.667539921392439e-05,
+      "loss": 1.3244,
+      "step": 9861
+    },
+    {
+      "epoch": 1.57792,
+      "grad_norm": 2.28472428857257,
+      "learning_rate": 9.665813497605907e-05,
+      "loss": 1.2833,
+      "step": 9862
+    },
+    {
+      "epoch": 1.57808,
+      "grad_norm": 2.068738742536845,
+      "learning_rate": 9.664087083791017e-05,
+      "loss": 1.3716,
+      "step": 9863
+    },
+    {
+      "epoch": 1.57824,
+      "grad_norm": 1.9358120075469134,
+      "learning_rate": 9.662360679999275e-05,
+      "loss": 1.2565,
+      "step": 9864
+    },
+    {
+      "epoch": 1.5784,
+      "grad_norm": 1.9525725760785,
+      "learning_rate": 9.6606342862822e-05,
+      "loss": 1.3871,
+      "step": 9865
+    },
+    {
+      "epoch": 1.57856,
+      "grad_norm": 2.4297364867202327,
+      "learning_rate": 9.658907902691302e-05,
+      "loss": 1.2835,
+      "step": 9866
+    },
+    {
+      "epoch": 1.5787200000000001,
+      "grad_norm": 1.90502743393676,
+      "learning_rate": 9.657181529278098e-05,
+      "loss": 1.3447,
+      "step": 9867
+    },
+    {
+      "epoch": 1.5788799999999998,
+      "grad_norm": 2.109573996290363,
+      "learning_rate": 9.655455166094095e-05,
+      "loss": 1.271,
+      "step": 9868
+    },
+    {
+      "epoch": 1.57904,
+      "grad_norm": 2.1660804961330538,
+      "learning_rate": 9.653728813190812e-05,
+      "loss": 1.2794,
+      "step": 9869
+    },
+    {
+      "epoch": 1.5792000000000002,
+      "grad_norm": 2.079571946451225,
+      "learning_rate": 9.652002470619751e-05,
+      "loss": 1.2624,
+      "step": 9870
+    },
+    {
+      "epoch": 1.5793599999999999,
+      "grad_norm": 2.5085936835113984,
+      "learning_rate": 9.650276138432434e-05,
+      "loss": 1.3401,
+      "step": 9871
+    },
+    {
+      "epoch": 1.57952,
+      "grad_norm": 2.411555065532334,
+      "learning_rate": 9.648549816680363e-05,
+      "loss": 1.3158,
+      "step": 9872
+    },
+    {
+      "epoch": 1.57968,
+      "grad_norm": 2.1669412134259463,
+      "learning_rate": 9.646823505415057e-05,
+      "loss": 1.3187,
+      "step": 9873
+    },
+    {
+      "epoch": 1.57984,
+      "grad_norm": 2.291617425392996,
+      "learning_rate": 9.645097204688019e-05,
+      "loss": 1.3154,
+      "step": 9874
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 2.0049616494010194,
+      "learning_rate": 9.643370914550765e-05,
+      "loss": 1.3614,
+      "step": 9875
+    },
+    {
+      "epoch": 1.58016,
+      "grad_norm": 2.1876832898418743,
+      "learning_rate": 9.641644635054807e-05,
+      "loss": 1.364,
+      "step": 9876
+    },
+    {
+      "epoch": 1.58032,
+      "grad_norm": 2.0655091033428374,
+      "learning_rate": 9.639918366251644e-05,
+      "loss": 1.3358,
+      "step": 9877
+    },
+    {
+      "epoch": 1.58048,
+      "grad_norm": 2.1151784344343354,
+      "learning_rate": 9.6381921081928e-05,
+      "loss": 1.32,
+      "step": 9878
+    },
+    {
+      "epoch": 1.58064,
+      "grad_norm": 2.002927318846052,
+      "learning_rate": 9.636465860929772e-05,
+      "loss": 1.3204,
+      "step": 9879
+    },
+    {
+      "epoch": 1.5808,
+      "grad_norm": 1.963419375845091,
+      "learning_rate": 9.634739624514075e-05,
+      "loss": 1.3663,
+      "step": 9880
+    },
+    {
+      "epoch": 1.5809600000000001,
+      "grad_norm": 1.808347379031711,
+      "learning_rate": 9.633013398997212e-05,
+      "loss": 1.331,
+      "step": 9881
+    },
+    {
+      "epoch": 1.5811199999999999,
+      "grad_norm": 1.862966445053054,
+      "learning_rate": 9.631287184430696e-05,
+      "loss": 1.2318,
+      "step": 9882
+    },
+    {
+      "epoch": 1.58128,
+      "grad_norm": 1.9901288441527216,
+      "learning_rate": 9.629560980866031e-05,
+      "loss": 1.3283,
+      "step": 9883
+    },
+    {
+      "epoch": 1.58144,
+      "grad_norm": 1.8229394622948087,
+      "learning_rate": 9.627834788354732e-05,
+      "loss": 1.3344,
+      "step": 9884
+    },
+    {
+      "epoch": 1.5816,
+      "grad_norm": 1.9858306825521237,
+      "learning_rate": 9.626108606948295e-05,
+      "loss": 1.3131,
+      "step": 9885
+    },
+    {
+      "epoch": 1.58176,
+      "grad_norm": 2.1887495899049427,
+      "learning_rate": 9.624382436698239e-05,
+      "loss": 1.2578,
+      "step": 9886
+    },
+    {
+      "epoch": 1.58192,
+      "grad_norm": 2.2087566432546817,
+      "learning_rate": 9.622656277656057e-05,
+      "loss": 1.3495,
+      "step": 9887
+    },
+    {
+      "epoch": 1.58208,
+      "grad_norm": 2.287277771149838,
+      "learning_rate": 9.620930129873267e-05,
+      "loss": 1.3346,
+      "step": 9888
+    },
+    {
+      "epoch": 1.58224,
+      "grad_norm": 2.2736197124437565,
+      "learning_rate": 9.619203993401366e-05,
+      "loss": 1.3201,
+      "step": 9889
+    },
+    {
+      "epoch": 1.5824,
+      "grad_norm": 1.9404638268173817,
+      "learning_rate": 9.617477868291863e-05,
+      "loss": 1.351,
+      "step": 9890
+    },
+    {
+      "epoch": 1.58256,
+      "grad_norm": 2.261064220760276,
+      "learning_rate": 9.615751754596262e-05,
+      "loss": 1.294,
+      "step": 9891
+    },
+    {
+      "epoch": 1.5827200000000001,
+      "grad_norm": 2.080286507511277,
+      "learning_rate": 9.614025652366075e-05,
+      "loss": 1.3353,
+      "step": 9892
+    },
+    {
+      "epoch": 1.5828799999999998,
+      "grad_norm": 2.2982499779177097,
+      "learning_rate": 9.612299561652793e-05,
+      "loss": 1.3751,
+      "step": 9893
+    },
+    {
+      "epoch": 1.58304,
+      "grad_norm": 2.0494904620880967,
+      "learning_rate": 9.610573482507933e-05,
+      "loss": 1.3362,
+      "step": 9894
+    },
+    {
+      "epoch": 1.5832000000000002,
+      "grad_norm": 2.4185631246130033,
+      "learning_rate": 9.608847414982988e-05,
+      "loss": 1.2973,
+      "step": 9895
+    },
+    {
+      "epoch": 1.5833599999999999,
+      "grad_norm": 1.8852418613030535,
+      "learning_rate": 9.60712135912947e-05,
+      "loss": 1.3794,
+      "step": 9896
+    },
+    {
+      "epoch": 1.58352,
+      "grad_norm": 2.1847888073250985,
+      "learning_rate": 9.605395314998874e-05,
+      "loss": 1.2825,
+      "step": 9897
+    },
+    {
+      "epoch": 1.58368,
+      "grad_norm": 2.1094688641340253,
+      "learning_rate": 9.60366928264271e-05,
+      "loss": 1.3445,
+      "step": 9898
+    },
+    {
+      "epoch": 1.58384,
+      "grad_norm": 1.9264770663743813,
+      "learning_rate": 9.601943262112473e-05,
+      "loss": 1.2673,
+      "step": 9899
+    },
+    {
+      "epoch": 1.584,
+      "grad_norm": 2.2841709407946866,
+      "learning_rate": 9.600217253459675e-05,
+      "loss": 1.3044,
+      "step": 9900
+    },
+    {
+      "epoch": 1.58416,
+      "grad_norm": 2.451989750119401,
+      "learning_rate": 9.598491256735807e-05,
+      "loss": 1.3357,
+      "step": 9901
+    },
+    {
+      "epoch": 1.58432,
+      "grad_norm": 2.222108186530279,
+      "learning_rate": 9.596765271992379e-05,
+      "loss": 1.4035,
+      "step": 9902
+    },
+    {
+      "epoch": 1.58448,
+      "grad_norm": 2.5176301769230514,
+      "learning_rate": 9.595039299280883e-05,
+      "loss": 1.2992,
+      "step": 9903
+    },
+    {
+      "epoch": 1.58464,
+      "grad_norm": 1.984669834743459,
+      "learning_rate": 9.59331333865283e-05,
+      "loss": 1.3093,
+      "step": 9904
+    },
+    {
+      "epoch": 1.5848,
+      "grad_norm": 2.5466959920590795,
+      "learning_rate": 9.591587390159709e-05,
+      "loss": 1.3342,
+      "step": 9905
+    },
+    {
+      "epoch": 1.5849600000000001,
+      "grad_norm": 2.2946383231080563,
+      "learning_rate": 9.589861453853028e-05,
+      "loss": 1.2674,
+      "step": 9906
+    },
+    {
+      "epoch": 1.5851199999999999,
+      "grad_norm": 2.1720117602668667,
+      "learning_rate": 9.58813552978428e-05,
+      "loss": 1.3405,
+      "step": 9907
+    },
+    {
+      "epoch": 1.58528,
+      "grad_norm": 1.8581884992889013,
+      "learning_rate": 9.586409618004973e-05,
+      "loss": 1.2567,
+      "step": 9908
+    },
+    {
+      "epoch": 1.58544,
+      "grad_norm": 2.7008693334650196,
+      "learning_rate": 9.584683718566596e-05,
+      "loss": 1.366,
+      "step": 9909
+    },
+    {
+      "epoch": 1.5856,
+      "grad_norm": 1.927469138405371,
+      "learning_rate": 9.582957831520655e-05,
+      "loss": 1.3248,
+      "step": 9910
+    },
+    {
+      "epoch": 1.58576,
+      "grad_norm": 2.4709151164101346,
+      "learning_rate": 9.58123195691864e-05,
+      "loss": 1.2858,
+      "step": 9911
+    },
+    {
+      "epoch": 1.58592,
+      "grad_norm": 2.874560628637777,
+      "learning_rate": 9.57950609481206e-05,
+      "loss": 1.38,
+      "step": 9912
+    },
+    {
+      "epoch": 1.58608,
+      "grad_norm": 1.9278766726519219,
+      "learning_rate": 9.577780245252399e-05,
+      "loss": 1.2677,
+      "step": 9913
+    },
+    {
+      "epoch": 1.58624,
+      "grad_norm": 1.8305543254186303,
+      "learning_rate": 9.576054408291162e-05,
+      "loss": 1.2883,
+      "step": 9914
+    },
+    {
+      "epoch": 1.5864,
+      "grad_norm": 2.0680057523450226,
+      "learning_rate": 9.574328583979844e-05,
+      "loss": 1.2856,
+      "step": 9915
+    },
+    {
+      "epoch": 1.58656,
+      "grad_norm": 1.9662279526142887,
+      "learning_rate": 9.572602772369943e-05,
+      "loss": 1.2566,
+      "step": 9916
+    },
+    {
+      "epoch": 1.5867200000000001,
+      "grad_norm": 1.9597878591704854,
+      "learning_rate": 9.570876973512949e-05,
+      "loss": 1.3095,
+      "step": 9917
+    },
+    {
+      "epoch": 1.5868799999999998,
+      "grad_norm": 2.3905445549225837,
+      "learning_rate": 9.569151187460365e-05,
+      "loss": 1.3012,
+      "step": 9918
+    },
+    {
+      "epoch": 1.58704,
+      "grad_norm": 2.0166701067776067,
+      "learning_rate": 9.567425414263678e-05,
+      "loss": 1.3161,
+      "step": 9919
+    },
+    {
+      "epoch": 1.5872000000000002,
+      "grad_norm": 2.3204800204267872,
+      "learning_rate": 9.56569965397439e-05,
+      "loss": 1.2471,
+      "step": 9920
+    },
+    {
+      "epoch": 1.5873599999999999,
+      "grad_norm": 1.9590920136114658,
+      "learning_rate": 9.563973906643985e-05,
+      "loss": 1.2724,
+      "step": 9921
+    },
+    {
+      "epoch": 1.58752,
+      "grad_norm": 2.1127804659534557,
+      "learning_rate": 9.562248172323968e-05,
+      "loss": 1.2859,
+      "step": 9922
+    },
+    {
+      "epoch": 1.58768,
+      "grad_norm": 2.0750796627117807,
+      "learning_rate": 9.560522451065826e-05,
+      "loss": 1.2866,
+      "step": 9923
+    },
+    {
+      "epoch": 1.58784,
+      "grad_norm": 2.0618922277421103,
+      "learning_rate": 9.558796742921057e-05,
+      "loss": 1.3781,
+      "step": 9924
+    },
+    {
+      "epoch": 1.588,
+      "grad_norm": 2.068991151791092,
+      "learning_rate": 9.557071047941146e-05,
+      "loss": 1.3675,
+      "step": 9925
+    },
+    {
+      "epoch": 1.58816,
+      "grad_norm": 1.9154629105846932,
+      "learning_rate": 9.555345366177594e-05,
+      "loss": 1.2693,
+      "step": 9926
+    },
+    {
+      "epoch": 1.58832,
+      "grad_norm": 2.2864620934047664,
+      "learning_rate": 9.553619697681884e-05,
+      "loss": 1.3252,
+      "step": 9927
+    },
+    {
+      "epoch": 1.5884800000000001,
+      "grad_norm": 1.9470081085684354,
+      "learning_rate": 9.551894042505516e-05,
+      "loss": 1.2913,
+      "step": 9928
+    },
+    {
+      "epoch": 1.58864,
+      "grad_norm": 2.1851625387146374,
+      "learning_rate": 9.550168400699973e-05,
+      "loss": 1.38,
+      "step": 9929
+    },
+    {
+      "epoch": 1.5888,
+      "grad_norm": 1.842796820187463,
+      "learning_rate": 9.548442772316751e-05,
+      "loss": 1.2976,
+      "step": 9930
+    },
+    {
+      "epoch": 1.5889600000000002,
+      "grad_norm": 2.5197077139984123,
+      "learning_rate": 9.546717157407336e-05,
+      "loss": 1.2822,
+      "step": 9931
+    },
+    {
+      "epoch": 1.5891199999999999,
+      "grad_norm": 1.9795852259586346,
+      "learning_rate": 9.544991556023224e-05,
+      "loss": 1.3511,
+      "step": 9932
+    },
+    {
+      "epoch": 1.58928,
+      "grad_norm": 2.0646669648736813,
+      "learning_rate": 9.5432659682159e-05,
+      "loss": 1.3298,
+      "step": 9933
+    },
+    {
+      "epoch": 1.58944,
+      "grad_norm": 2.2061295619851884,
+      "learning_rate": 9.541540394036852e-05,
+      "loss": 1.321,
+      "step": 9934
+    },
+    {
+      "epoch": 1.5896,
+      "grad_norm": 1.8436503547183467,
+      "learning_rate": 9.539814833537574e-05,
+      "loss": 1.3685,
+      "step": 9935
+    },
+    {
+      "epoch": 1.58976,
+      "grad_norm": 1.9302730993095003,
+      "learning_rate": 9.538089286769548e-05,
+      "loss": 1.2688,
+      "step": 9936
+    },
+    {
+      "epoch": 1.58992,
+      "grad_norm": 1.9302730993095003,
+      "learning_rate": 9.538089286769548e-05,
+      "loss": 1.2943,
+      "step": 9937
+    },
+    {
+      "epoch": 1.59008,
+      "grad_norm": 2.053611452486623,
+      "learning_rate": 9.536363753784268e-05,
+      "loss": 1.4138,
+      "step": 9938
+    },
+    {
+      "epoch": 1.59024,
+      "grad_norm": 2.053611452486623,
+      "learning_rate": 9.536363753784268e-05,
+      "loss": 1.3825,
+      "step": 9939
+    },
+    {
+      "epoch": 1.5904,
+      "grad_norm": 2.4367068786645705,
+      "learning_rate": 9.534638234633216e-05,
+      "loss": 1.286,
+      "step": 9940
+    },
+    {
+      "epoch": 1.59056,
+      "grad_norm": 2.297950512920456,
+      "learning_rate": 9.53291272936788e-05,
+      "loss": 1.3535,
+      "step": 9941
+    },
+    {
+      "epoch": 1.5907200000000001,
+      "grad_norm": 2.13403472628532,
+      "learning_rate": 9.531187238039748e-05,
+      "loss": 1.3301,
+      "step": 9942
+    },
+    {
+      "epoch": 1.5908799999999998,
+      "grad_norm": 2.0554676650162063,
+      "learning_rate": 9.529461760700308e-05,
+      "loss": 1.346,
+      "step": 9943
+    },
+    {
+      "epoch": 1.59104,
+      "grad_norm": 1.9757711080492215,
+      "learning_rate": 9.527736297401039e-05,
+      "loss": 1.3215,
+      "step": 9944
+    },
+    {
+      "epoch": 1.5912,
+      "grad_norm": 2.021934829602552,
+      "learning_rate": 9.526010848193437e-05,
+      "loss": 1.3348,
+      "step": 9945
+    },
+    {
+      "epoch": 1.5913599999999999,
+      "grad_norm": 2.4908143875229873,
+      "learning_rate": 9.524285413128974e-05,
+      "loss": 1.3507,
+      "step": 9946
+    },
+    {
+      "epoch": 1.59152,
+      "grad_norm": 1.8003380974523655,
+      "learning_rate": 9.522559992259147e-05,
+      "loss": 1.2888,
+      "step": 9947
+    },
+    {
+      "epoch": 1.59168,
+      "grad_norm": 2.0870758930582807,
+      "learning_rate": 9.520834585635428e-05,
+      "loss": 1.3955,
+      "step": 9948
+    },
+    {
+      "epoch": 1.59184,
+      "grad_norm": 1.7913585543378727,
+      "learning_rate": 9.519109193309311e-05,
+      "loss": 1.384,
+      "step": 9949
+    },
+    {
+      "epoch": 1.592,
+      "grad_norm": 1.8352756437991518,
+      "learning_rate": 9.517383815332271e-05,
+      "loss": 1.2806,
+      "step": 9950
+    },
+    {
+      "epoch": 1.59216,
+      "grad_norm": 2.0028139145189416,
+      "learning_rate": 9.515658451755801e-05,
+      "loss": 1.2979,
+      "step": 9951
+    },
+    {
+      "epoch": 1.59232,
+      "grad_norm": 2.8010366853421265,
+      "learning_rate": 9.513933102631371e-05,
+      "loss": 1.3081,
+      "step": 9952
+    },
+    {
+      "epoch": 1.5924800000000001,
+      "grad_norm": 2.154658545957246,
+      "learning_rate": 9.512207768010475e-05,
+      "loss": 1.385,
+      "step": 9953
+    },
+    {
+      "epoch": 1.5926399999999998,
+      "grad_norm": 2.184918457284557,
+      "learning_rate": 9.510482447944582e-05,
+      "loss": 1.3475,
+      "step": 9954
+    },
+    {
+      "epoch": 1.5928,
+      "grad_norm": 2.861423227929576,
+      "learning_rate": 9.508757142485185e-05,
+      "loss": 1.3457,
+      "step": 9955
+    },
+    {
+      "epoch": 1.5929600000000002,
+      "grad_norm": 2.050589966737101,
+      "learning_rate": 9.507031851683755e-05,
+      "loss": 1.2775,
+      "step": 9956
+    },
+    {
+      "epoch": 1.5931199999999999,
+      "grad_norm": 2.3081089412254743,
+      "learning_rate": 9.505306575591779e-05,
+      "loss": 1.216,
+      "step": 9957
+    },
+    {
+      "epoch": 1.59328,
+      "grad_norm": 2.2807875099461357,
+      "learning_rate": 9.50358131426073e-05,
+      "loss": 1.3122,
+      "step": 9958
+    },
+    {
+      "epoch": 1.59344,
+      "grad_norm": 2.152104630068182,
+      "learning_rate": 9.501856067742096e-05,
+      "loss": 1.2454,
+      "step": 9959
+    },
+    {
+      "epoch": 1.5936,
+      "grad_norm": 2.221193315641849,
+      "learning_rate": 9.500130836087346e-05,
+      "loss": 1.3145,
+      "step": 9960
+    },
+    {
+      "epoch": 1.59376,
+      "grad_norm": 2.433613988670585,
+      "learning_rate": 9.49840561934797e-05,
+      "loss": 1.3223,
+      "step": 9961
+    },
+    {
+      "epoch": 1.59392,
+      "grad_norm": 2.293724649285314,
+      "learning_rate": 9.496680417575433e-05,
+      "loss": 1.3443,
+      "step": 9962
+    },
+    {
+      "epoch": 1.59408,
+      "grad_norm": 2.0061565327663287,
+      "learning_rate": 9.494955230821224e-05,
+      "loss": 1.264,
+      "step": 9963
+    },
+    {
+      "epoch": 1.59424,
+      "grad_norm": 3.0543095280031416,
+      "learning_rate": 9.493230059136811e-05,
+      "loss": 1.2828,
+      "step": 9964
+    },
+    {
+      "epoch": 1.5944,
+      "grad_norm": 1.7300058726418002,
+      "learning_rate": 9.491504902573677e-05,
+      "loss": 1.2394,
+      "step": 9965
+    },
+    {
+      "epoch": 1.59456,
+      "grad_norm": 2.2389077154208517,
+      "learning_rate": 9.489779761183293e-05,
+      "loss": 1.2723,
+      "step": 9966
+    },
+    {
+      "epoch": 1.5947200000000001,
+      "grad_norm": 2.486471967706228,
+      "learning_rate": 9.488054635017142e-05,
+      "loss": 1.3444,
+      "step": 9967
+    },
+    {
+      "epoch": 1.5948799999999999,
+      "grad_norm": 2.1599129460631645,
+      "learning_rate": 9.48632952412669e-05,
+      "loss": 1.3586,
+      "step": 9968
+    },
+    {
+      "epoch": 1.59504,
+      "grad_norm": 2.019791450796009,
+      "learning_rate": 9.484604428563422e-05,
+      "loss": 1.346,
+      "step": 9969
+    },
+    {
+      "epoch": 1.5952,
+      "grad_norm": 2.030283021382468,
+      "learning_rate": 9.482879348378801e-05,
+      "loss": 1.2825,
+      "step": 9970
+    },
+    {
+      "epoch": 1.59536,
+      "grad_norm": 2.1572561670928563,
+      "learning_rate": 9.481154283624312e-05,
+      "loss": 1.2557,
+      "step": 9971
+    },
+    {
+      "epoch": 1.59552,
+      "grad_norm": 2.0356921862141153,
+      "learning_rate": 9.479429234351418e-05,
+      "loss": 1.3749,
+      "step": 9972
+    },
+    {
+      "epoch": 1.59568,
+      "grad_norm": 2.1545064772830966,
+      "learning_rate": 9.4777042006116e-05,
+      "loss": 1.3727,
+      "step": 9973
+    },
+    {
+      "epoch": 1.59584,
+      "grad_norm": 1.9167965281606378,
+      "learning_rate": 9.475979182456325e-05,
+      "loss": 1.305,
+      "step": 9974
+    },
+    {
+      "epoch": 1.596,
+      "grad_norm": 1.865959540699213,
+      "learning_rate": 9.474254179937072e-05,
+      "loss": 1.2793,
+      "step": 9975
+    },
+    {
+      "epoch": 1.59616,
+      "grad_norm": 2.2828296005949746,
+      "learning_rate": 9.472529193105303e-05,
+      "loss": 1.3334,
+      "step": 9976
+    },
+    {
+      "epoch": 1.59632,
+      "grad_norm": 2.0959447746700843,
+      "learning_rate": 9.470804222012501e-05,
+      "loss": 1.3429,
+      "step": 9977
+    },
+    {
+      "epoch": 1.5964800000000001,
+      "grad_norm": 2.318877123211011,
+      "learning_rate": 9.469079266710123e-05,
+      "loss": 1.2877,
+      "step": 9978
+    },
+    {
+      "epoch": 1.5966399999999998,
+      "grad_norm": 2.2903168586680285,
+      "learning_rate": 9.467354327249652e-05,
+      "loss": 1.3566,
+      "step": 9979
+    },
+    {
+      "epoch": 1.5968,
+      "grad_norm": 2.2292529022689065,
+      "learning_rate": 9.465629403682548e-05,
+      "loss": 1.3144,
+      "step": 9980
+    },
+    {
+      "epoch": 1.5969600000000002,
+      "grad_norm": 1.9810672872395436,
+      "learning_rate": 9.463904496060284e-05,
+      "loss": 1.2686,
+      "step": 9981
+    },
+    {
+      "epoch": 1.5971199999999999,
+      "grad_norm": 2.432693688021202,
+      "learning_rate": 9.462179604434328e-05,
+      "loss": 1.4186,
+      "step": 9982
+    },
+    {
+      "epoch": 1.59728,
+      "grad_norm": 2.838554070375258,
+      "learning_rate": 9.460454728856153e-05,
+      "loss": 1.3346,
+      "step": 9983
+    },
+    {
+      "epoch": 1.59744,
+      "grad_norm": 2.6779477350639778,
+      "learning_rate": 9.458729869377217e-05,
+      "loss": 1.333,
+      "step": 9984
+    },
+    {
+      "epoch": 1.5976,
+      "grad_norm": 2.1399200061638646,
+      "learning_rate": 9.457005026049e-05,
+      "loss": 1.312,
+      "step": 9985
+    },
+    {
+      "epoch": 1.59776,
+      "grad_norm": 2.2097440398361003,
+      "learning_rate": 9.455280198922956e-05,
+      "loss": 1.3068,
+      "step": 9986
+    },
+    {
+      "epoch": 1.59792,
+      "grad_norm": 2.1178976255804316,
+      "learning_rate": 9.45355538805056e-05,
+      "loss": 1.2882,
+      "step": 9987
+    },
+    {
+      "epoch": 1.59808,
+      "grad_norm": 2.072998420100706,
+      "learning_rate": 9.451830593483273e-05,
+      "loss": 1.2517,
+      "step": 9988
+    },
+    {
+      "epoch": 1.59824,
+      "grad_norm": 2.098122542430632,
+      "learning_rate": 9.450105815272563e-05,
+      "loss": 1.3129,
+      "step": 9989
+    },
+    {
+      "epoch": 1.5984,
+      "grad_norm": 2.203569456845842,
+      "learning_rate": 9.448381053469893e-05,
+      "loss": 1.3572,
+      "step": 9990
+    },
+    {
+      "epoch": 1.59856,
+      "grad_norm": 2.173772608957403,
+      "learning_rate": 9.446656308126732e-05,
+      "loss": 1.3172,
+      "step": 9991
+    },
+    {
+      "epoch": 1.5987200000000001,
+      "grad_norm": 2.030760142751847,
+      "learning_rate": 9.444931579294536e-05,
+      "loss": 1.3033,
+      "step": 9992
+    },
+    {
+      "epoch": 1.5988799999999999,
+      "grad_norm": 2.208115135903921,
+      "learning_rate": 9.443206867024774e-05,
+      "loss": 1.287,
+      "step": 9993
+    },
+    {
+      "epoch": 1.59904,
+      "grad_norm": 2.224684972964484,
+      "learning_rate": 9.441482171368911e-05,
+      "loss": 1.2221,
+      "step": 9994
+    },
+    {
+      "epoch": 1.5992,
+      "grad_norm": 2.177599496230135,
+      "learning_rate": 9.439757492378402e-05,
+      "loss": 1.3162,
+      "step": 9995
+    },
+    {
+      "epoch": 1.59936,
+      "grad_norm": 2.334960901016772,
+      "learning_rate": 9.438032830104718e-05,
+      "loss": 1.4246,
+      "step": 9996
+    },
+    {
+      "epoch": 1.59952,
+      "grad_norm": 2.0071129536491257,
+      "learning_rate": 9.436308184599311e-05,
+      "loss": 1.3313,
+      "step": 9997
+    },
+    {
+      "epoch": 1.59968,
+      "grad_norm": 2.014556521833955,
+      "learning_rate": 9.434583555913648e-05,
+      "loss": 1.3331,
+      "step": 9998
+    },
+    {
+      "epoch": 1.59984,
+      "grad_norm": 2.0342380961014594,
+      "learning_rate": 9.432858944099189e-05,
+      "loss": 1.3057,
+      "step": 9999
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 2.4378620123386545,
+      "learning_rate": 9.431134349207391e-05,
+      "loss": 1.3735,
+      "step": 10000
+    },
+    {
+      "epoch": 1.60016,
+      "grad_norm": 2.031563392219701,
+      "learning_rate": 9.429409771289714e-05,
+      "loss": 1.3046,
+      "step": 10001
+    },
+    {
+      "epoch": 1.60032,
+      "grad_norm": 2.5767275936800678,
+      "learning_rate": 9.427685210397622e-05,
+      "loss": 1.2956,
+      "step": 10002
+    },
+    {
+      "epoch": 1.6004800000000001,
+      "grad_norm": 2.0630023385801977,
+      "learning_rate": 9.425960666582566e-05,
+      "loss": 1.337,
+      "step": 10003
+    },
+    {
+      "epoch": 1.6006399999999998,
+      "grad_norm": 2.9601381591626397,
+      "learning_rate": 9.42423613989601e-05,
+      "loss": 1.3092,
+      "step": 10004
+    },
+    {
+      "epoch": 1.6008,
+      "grad_norm": 3.010555627243759,
+      "learning_rate": 9.422511630389406e-05,
+      "loss": 1.3546,
+      "step": 10005
+    },
+    {
+      "epoch": 1.6009600000000002,
+      "grad_norm": 2.4369907312802184,
+      "learning_rate": 9.420787138114214e-05,
+      "loss": 1.3225,
+      "step": 10006
+    },
+    {
+      "epoch": 1.6011199999999999,
+      "grad_norm": 2.45225987626601,
+      "learning_rate": 9.41906266312189e-05,
+      "loss": 1.3373,
+      "step": 10007
+    },
+    {
+      "epoch": 1.60128,
+      "grad_norm": 2.0670587718637927,
+      "learning_rate": 9.417338205463891e-05,
+      "loss": 1.3352,
+      "step": 10008
+    },
+    {
+      "epoch": 1.60144,
+      "grad_norm": 1.9888477496130244,
+      "learning_rate": 9.415613765191668e-05,
+      "loss": 1.302,
+      "step": 10009
+    },
+    {
+      "epoch": 1.6016,
+      "grad_norm": 1.9196902876497712,
+      "learning_rate": 9.413889342356682e-05,
+      "loss": 1.2307,
+      "step": 10010
+    },
+    {
+      "epoch": 1.60176,
+      "grad_norm": 2.4499507052362204,
+      "learning_rate": 9.412164937010381e-05,
+      "loss": 1.2776,
+      "step": 10011
+    },
+    {
+      "epoch": 1.60192,
+      "grad_norm": 2.302474821040069,
+      "learning_rate": 9.410440549204225e-05,
+      "loss": 1.3143,
+      "step": 10012
+    },
+    {
+      "epoch": 1.60208,
+      "grad_norm": 2.0367060898860894,
+      "learning_rate": 9.40871617898966e-05,
+      "loss": 1.3487,
+      "step": 10013
+    },
+    {
+      "epoch": 1.60224,
+      "grad_norm": 2.4136936152351787,
+      "learning_rate": 9.406991826418146e-05,
+      "loss": 1.3053,
+      "step": 10014
+    },
+    {
+      "epoch": 1.6024,
+      "grad_norm": 2.694313487983494,
+      "learning_rate": 9.405267491541129e-05,
+      "loss": 1.2445,
+      "step": 10015
+    },
+    {
+      "epoch": 1.60256,
+      "grad_norm": 2.130520435050084,
+      "learning_rate": 9.403543174410067e-05,
+      "loss": 1.3302,
+      "step": 10016
+    },
+    {
+      "epoch": 1.6027200000000001,
+      "grad_norm": 2.1189888965526382,
+      "learning_rate": 9.401818875076402e-05,
+      "loss": 1.3112,
+      "step": 10017
+    },
+    {
+      "epoch": 1.6028799999999999,
+      "grad_norm": 3.1130543285394308,
+      "learning_rate": 9.400094593591598e-05,
+      "loss": 1.2516,
+      "step": 10018
+    },
+    {
+      "epoch": 1.60304,
+      "grad_norm": 2.1819577560507035,
+      "learning_rate": 9.39837033000709e-05,
+      "loss": 1.2922,
+      "step": 10019
+    },
+    {
+      "epoch": 1.6032,
+      "grad_norm": 3.0365231073660834,
+      "learning_rate": 9.39664608437434e-05,
+      "loss": 1.2694,
+      "step": 10020
+    },
+    {
+      "epoch": 1.60336,
+      "grad_norm": 1.7985981545123768,
+      "learning_rate": 9.394921856744788e-05,
+      "loss": 1.261,
+      "step": 10021
+    },
+    {
+      "epoch": 1.60352,
+      "grad_norm": 1.9764514930605235,
+      "learning_rate": 9.393197647169887e-05,
+      "loss": 1.2643,
+      "step": 10022
+    },
+    {
+      "epoch": 1.60368,
+      "grad_norm": 2.355055614276912,
+      "learning_rate": 9.391473455701085e-05,
+      "loss": 1.4034,
+      "step": 10023
+    },
+    {
+      "epoch": 1.60384,
+      "grad_norm": 2.067795945951789,
+      "learning_rate": 9.389749282389827e-05,
+      "loss": 1.3144,
+      "step": 10024
+    },
+    {
+      "epoch": 1.604,
+      "grad_norm": 1.8195610172387124,
+      "learning_rate": 9.38802512728756e-05,
+      "loss": 1.2757,
+      "step": 10025
+    },
+    {
+      "epoch": 1.60416,
+      "grad_norm": 2.057048923467534,
+      "learning_rate": 9.386300990445736e-05,
+      "loss": 1.3651,
+      "step": 10026
+    },
+    {
+      "epoch": 1.60432,
+      "grad_norm": 2.5135958932127873,
+      "learning_rate": 9.384576871915792e-05,
+      "loss": 1.2925,
+      "step": 10027
+    },
+    {
+      "epoch": 1.6044800000000001,
+      "grad_norm": 2.1301906993353366,
+      "learning_rate": 9.38285277174918e-05,
+      "loss": 1.2842,
+      "step": 10028
+    },
+    {
+      "epoch": 1.6046399999999998,
+      "grad_norm": 2.0154509470378703,
+      "learning_rate": 9.381128689997338e-05,
+      "loss": 1.301,
+      "step": 10029
+    },
+    {
+      "epoch": 1.6048,
+      "grad_norm": 2.1373955566206595,
+      "learning_rate": 9.379404626711718e-05,
+      "loss": 1.2934,
+      "step": 10030
+    },
+    {
+      "epoch": 1.6049600000000002,
+      "grad_norm": 2.3273604242576105,
+      "learning_rate": 9.377680581943759e-05,
+      "loss": 1.352,
+      "step": 10031
+    },
+    {
+      "epoch": 1.6051199999999999,
+      "grad_norm": 2.29855492775319,
+      "learning_rate": 9.375956555744904e-05,
+      "loss": 1.3475,
+      "step": 10032
+    },
+    {
+      "epoch": 1.60528,
+      "grad_norm": 2.0742902136365564,
+      "learning_rate": 9.374232548166595e-05,
+      "loss": 1.2762,
+      "step": 10033
+    },
+    {
+      "epoch": 1.60544,
+      "grad_norm": 2.1435465805605345,
+      "learning_rate": 9.372508559260278e-05,
+      "loss": 1.2915,
+      "step": 10034
+    },
+    {
+      "epoch": 1.6056,
+      "grad_norm": 2.4467475970926906,
+      "learning_rate": 9.370784589077386e-05,
+      "loss": 1.4196,
+      "step": 10035
+    },
+    {
+      "epoch": 1.60576,
+      "grad_norm": 2.1179228133819477,
+      "learning_rate": 9.369060637669372e-05,
+      "loss": 1.3303,
+      "step": 10036
+    },
+    {
+      "epoch": 1.60592,
+      "grad_norm": 2.593885452439388,
+      "learning_rate": 9.367336705087664e-05,
+      "loss": 1.3552,
+      "step": 10037
+    },
+    {
+      "epoch": 1.60608,
+      "grad_norm": 1.899107111072277,
+      "learning_rate": 9.365612791383709e-05,
+      "loss": 1.3068,
+      "step": 10038
+    },
+    {
+      "epoch": 1.6062400000000001,
+      "grad_norm": 2.6971480254718703,
+      "learning_rate": 9.363888896608943e-05,
+      "loss": 1.299,
+      "step": 10039
+    },
+    {
+      "epoch": 1.6064,
+      "grad_norm": 2.3204364025439452,
+      "learning_rate": 9.362165020814806e-05,
+      "loss": 1.354,
+      "step": 10040
+    },
+    {
+      "epoch": 1.60656,
+      "grad_norm": 2.46837579711636,
+      "learning_rate": 9.360441164052734e-05,
+      "loss": 1.3122,
+      "step": 10041
+    },
+    {
+      "epoch": 1.6067200000000001,
+      "grad_norm": 2.345976907908757,
+      "learning_rate": 9.358717326374169e-05,
+      "loss": 1.3106,
+      "step": 10042
+    },
+    {
+      "epoch": 1.6068799999999999,
+      "grad_norm": 2.194915695022561,
+      "learning_rate": 9.35699350783054e-05,
+      "loss": 1.2748,
+      "step": 10043
+    },
+    {
+      "epoch": 1.60704,
+      "grad_norm": 2.08851923872709,
+      "learning_rate": 9.355269708473292e-05,
+      "loss": 1.2883,
+      "step": 10044
+    },
+    {
+      "epoch": 1.6072,
+      "grad_norm": 1.980077569875961,
+      "learning_rate": 9.353545928353853e-05,
+      "loss": 1.286,
+      "step": 10045
+    },
+    {
+      "epoch": 1.60736,
+      "grad_norm": 1.9249528795740007,
+      "learning_rate": 9.351822167523663e-05,
+      "loss": 1.2632,
+      "step": 10046
+    },
+    {
+      "epoch": 1.60752,
+      "grad_norm": 2.211426110049411,
+      "learning_rate": 9.350098426034153e-05,
+      "loss": 1.3642,
+      "step": 10047
+    },
+    {
+      "epoch": 1.60768,
+      "grad_norm": 1.948183973694207,
+      "learning_rate": 9.348374703936765e-05,
+      "loss": 1.3345,
+      "step": 10048
+    },
+    {
+      "epoch": 1.60784,
+      "grad_norm": 1.9245854280996118,
+      "learning_rate": 9.346651001282919e-05,
+      "loss": 1.2671,
+      "step": 10049
+    },
+    {
+      "epoch": 1.608,
+      "grad_norm": 1.9578404838602295,
+      "learning_rate": 9.344927318124061e-05,
+      "loss": 1.3281,
+      "step": 10050
+    },
+    {
+      "epoch": 1.60816,
+      "grad_norm": 1.715045168690731,
+      "learning_rate": 9.343203654511613e-05,
+      "loss": 1.3108,
+      "step": 10051
+    },
+    {
+      "epoch": 1.60832,
+      "grad_norm": 1.9446687761050574,
+      "learning_rate": 9.34148001049701e-05,
+      "loss": 1.2431,
+      "step": 10052
+    },
+    {
+      "epoch": 1.6084800000000001,
+      "grad_norm": 1.9963862886672936,
+      "learning_rate": 9.339756386131689e-05,
+      "loss": 1.2907,
+      "step": 10053
+    },
+    {
+      "epoch": 1.6086399999999998,
+      "grad_norm": 1.8207360605964265,
+      "learning_rate": 9.338032781467072e-05,
+      "loss": 1.3284,
+      "step": 10054
+    },
+    {
+      "epoch": 1.6088,
+      "grad_norm": 2.028673994121587,
+      "learning_rate": 9.336309196554593e-05,
+      "loss": 1.3047,
+      "step": 10055
+    },
+    {
+      "epoch": 1.60896,
+      "grad_norm": 2.0546853406761296,
+      "learning_rate": 9.33458563144568e-05,
+      "loss": 1.2621,
+      "step": 10056
+    },
+    {
+      "epoch": 1.6091199999999999,
+      "grad_norm": 2.197668881936546,
+      "learning_rate": 9.332862086191763e-05,
+      "loss": 1.3373,
+      "step": 10057
+    },
+    {
+      "epoch": 1.60928,
+      "grad_norm": 2.005957254306174,
+      "learning_rate": 9.331138560844268e-05,
+      "loss": 1.274,
+      "step": 10058
+    },
+    {
+      "epoch": 1.60944,
+      "grad_norm": 2.188420669811347,
+      "learning_rate": 9.329415055454627e-05,
+      "loss": 1.2904,
+      "step": 10059
+    },
+    {
+      "epoch": 1.6096,
+      "grad_norm": 2.1735569358960714,
+      "learning_rate": 9.32769157007426e-05,
+      "loss": 1.3493,
+      "step": 10060
+    },
+    {
+      "epoch": 1.60976,
+      "grad_norm": 1.9742401809165087,
+      "learning_rate": 9.3259681047546e-05,
+      "loss": 1.3587,
+      "step": 10061
+    },
+    {
+      "epoch": 1.60992,
+      "grad_norm": 2.0448910836035963,
+      "learning_rate": 9.324244659547066e-05,
+      "loss": 1.3763,
+      "step": 10062
+    },
+    {
+      "epoch": 1.61008,
+      "grad_norm": 2.1659712350138407,
+      "learning_rate": 9.322521234503088e-05,
+      "loss": 1.3912,
+      "step": 10063
+    },
+    {
+      "epoch": 1.6102400000000001,
+      "grad_norm": 1.9182547140890538,
+      "learning_rate": 9.32079782967409e-05,
+      "loss": 1.3104,
+      "step": 10064
+    },
+    {
+      "epoch": 1.6104,
+      "grad_norm": 2.023560856035014,
+      "learning_rate": 9.319074445111495e-05,
+      "loss": 1.3136,
+      "step": 10065
+    },
+    {
+      "epoch": 1.61056,
+      "grad_norm": 2.1633153591111607,
+      "learning_rate": 9.317351080866724e-05,
+      "loss": 1.3203,
+      "step": 10066
+    },
+    {
+      "epoch": 1.6107200000000002,
+      "grad_norm": 2.1075190611291865,
+      "learning_rate": 9.315627736991207e-05,
+      "loss": 1.3899,
+      "step": 10067
+    },
+    {
+      "epoch": 1.6108799999999999,
+      "grad_norm": 2.5648018371330865,
+      "learning_rate": 9.313904413536357e-05,
+      "loss": 1.3036,
+      "step": 10068
+    },
+    {
+      "epoch": 1.61104,
+      "grad_norm": 2.062075622682776,
+      "learning_rate": 9.312181110553601e-05,
+      "loss": 1.3387,
+      "step": 10069
+    },
+    {
+      "epoch": 1.6112,
+      "grad_norm": 2.2624753924693475,
+      "learning_rate": 9.310457828094358e-05,
+      "loss": 1.3614,
+      "step": 10070
+    },
+    {
+      "epoch": 1.61136,
+      "grad_norm": 2.011467592516461,
+      "learning_rate": 9.308734566210048e-05,
+      "loss": 1.2695,
+      "step": 10071
+    },
+    {
+      "epoch": 1.61152,
+      "grad_norm": 2.1911634424731523,
+      "learning_rate": 9.307011324952093e-05,
+      "loss": 1.2757,
+      "step": 10072
+    },
+    {
+      "epoch": 1.61168,
+      "grad_norm": 1.9564801841338875,
+      "learning_rate": 9.305288104371909e-05,
+      "loss": 1.3058,
+      "step": 10073
+    },
+    {
+      "epoch": 1.61184,
+      "grad_norm": 2.0463008280837,
+      "learning_rate": 9.303564904520915e-05,
+      "loss": 1.2541,
+      "step": 10074
+    },
+    {
+      "epoch": 1.612,
+      "grad_norm": 2.1356424658176434,
+      "learning_rate": 9.301841725450532e-05,
+      "loss": 1.3236,
+      "step": 10075
+    },
+    {
+      "epoch": 1.61216,
+      "grad_norm": 2.2152031715747493,
+      "learning_rate": 9.30011856721217e-05,
+      "loss": 1.2999,
+      "step": 10076
+    },
+    {
+      "epoch": 1.61232,
+      "grad_norm": 2.319612460278761,
+      "learning_rate": 9.298395429857256e-05,
+      "loss": 1.3068,
+      "step": 10077
+    },
+    {
+      "epoch": 1.6124800000000001,
+      "grad_norm": 2.0465768501492994,
+      "learning_rate": 9.296672313437195e-05,
+      "loss": 1.2723,
+      "step": 10078
+    },
+    {
+      "epoch": 1.6126399999999999,
+      "grad_norm": 2.400003220472012,
+      "learning_rate": 9.294949218003411e-05,
+      "loss": 1.3559,
+      "step": 10079
+    },
+    {
+      "epoch": 1.6128,
+      "grad_norm": 2.2906427321640224,
+      "learning_rate": 9.293226143607313e-05,
+      "loss": 1.3022,
+      "step": 10080
+    },
+    {
+      "epoch": 1.61296,
+      "grad_norm": 2.059209811360842,
+      "learning_rate": 9.291503090300318e-05,
+      "loss": 1.3021,
+      "step": 10081
+    },
+    {
+      "epoch": 1.6131199999999999,
+      "grad_norm": 1.972813943814558,
+      "learning_rate": 9.289780058133835e-05,
+      "loss": 1.2762,
+      "step": 10082
+    },
+    {
+      "epoch": 1.61328,
+      "grad_norm": 2.376204220996477,
+      "learning_rate": 9.288057047159285e-05,
+      "loss": 1.3077,
+      "step": 10083
+    },
+    {
+      "epoch": 1.61344,
+      "grad_norm": 1.9769554163394234,
+      "learning_rate": 9.286334057428073e-05,
+      "loss": 1.3045,
+      "step": 10084
+    },
+    {
+      "epoch": 1.6136,
+      "grad_norm": 1.8606828618605018,
+      "learning_rate": 9.284611088991614e-05,
+      "loss": 1.2776,
+      "step": 10085
+    },
+    {
+      "epoch": 1.61376,
+      "grad_norm": 2.2483673259191264,
+      "learning_rate": 9.282888141901315e-05,
+      "loss": 1.2744,
+      "step": 10086
+    },
+    {
+      "epoch": 1.61392,
+      "grad_norm": 2.2765865730852224,
+      "learning_rate": 9.281165216208592e-05,
+      "loss": 1.3846,
+      "step": 10087
+    },
+    {
+      "epoch": 1.61408,
+      "grad_norm": 2.2963361315768,
+      "learning_rate": 9.27944231196485e-05,
+      "loss": 1.2426,
+      "step": 10088
+    },
+    {
+      "epoch": 1.6142400000000001,
+      "grad_norm": 2.405819407866077,
+      "learning_rate": 9.277719429221499e-05,
+      "loss": 1.3337,
+      "step": 10089
+    },
+    {
+      "epoch": 1.6143999999999998,
+      "grad_norm": 2.3309288634434573,
+      "learning_rate": 9.275996568029946e-05,
+      "loss": 1.2926,
+      "step": 10090
+    },
+    {
+      "epoch": 1.61456,
+      "grad_norm": 2.0837409326815313,
+      "learning_rate": 9.274273728441606e-05,
+      "loss": 1.2234,
+      "step": 10091
+    },
+    {
+      "epoch": 1.6147200000000002,
+      "grad_norm": 2.7463679144898725,
+      "learning_rate": 9.272550910507875e-05,
+      "loss": 1.2707,
+      "step": 10092
+    },
+    {
+      "epoch": 1.6148799999999999,
+      "grad_norm": 3.1375590502437984,
+      "learning_rate": 9.27082811428017e-05,
+      "loss": 1.2619,
+      "step": 10093
+    },
+    {
+      "epoch": 1.61504,
+      "grad_norm": 2.098521439146592,
+      "learning_rate": 9.269105339809886e-05,
+      "loss": 1.2653,
+      "step": 10094
+    },
+    {
+      "epoch": 1.6152,
+      "grad_norm": 2.268808766634168,
+      "learning_rate": 9.267382587148437e-05,
+      "loss": 1.2424,
+      "step": 10095
+    },
+    {
+      "epoch": 1.61536,
+      "grad_norm": 2.725688701904059,
+      "learning_rate": 9.265659856347223e-05,
+      "loss": 1.3706,
+      "step": 10096
+    },
+    {
+      "epoch": 1.61552,
+      "grad_norm": 1.938203647911437,
+      "learning_rate": 9.26393714745765e-05,
+      "loss": 1.3395,
+      "step": 10097
+    },
+    {
+      "epoch": 1.61568,
+      "grad_norm": 3.210207359555544,
+      "learning_rate": 9.262214460531117e-05,
+      "loss": 1.3582,
+      "step": 10098
+    },
+    {
+      "epoch": 1.61584,
+      "grad_norm": 2.0409403606205574,
+      "learning_rate": 9.260491795619033e-05,
+      "loss": 1.3355,
+      "step": 10099
+    },
+    {
+      "epoch": 1.616,
+      "grad_norm": 1.7348986072111632,
+      "learning_rate": 9.258769152772793e-05,
+      "loss": 1.2784,
+      "step": 10100
+    },
+    {
+      "epoch": 1.61616,
+      "grad_norm": 1.8402527510790678,
+      "learning_rate": 9.257046532043805e-05,
+      "loss": 1.3074,
+      "step": 10101
+    },
+    {
+      "epoch": 1.61632,
+      "grad_norm": 2.0684661330564977,
+      "learning_rate": 9.255323933483463e-05,
+      "loss": 1.2785,
+      "step": 10102
+    },
+    {
+      "epoch": 1.6164800000000001,
+      "grad_norm": 1.7934168914621698,
+      "learning_rate": 9.253601357143171e-05,
+      "loss": 1.2987,
+      "step": 10103
+    },
+    {
+      "epoch": 1.6166399999999999,
+      "grad_norm": 2.1819094584817704,
+      "learning_rate": 9.251878803074326e-05,
+      "loss": 1.3592,
+      "step": 10104
+    },
+    {
+      "epoch": 1.6168,
+      "grad_norm": 2.23715152343947,
+      "learning_rate": 9.250156271328329e-05,
+      "loss": 1.366,
+      "step": 10105
+    },
+    {
+      "epoch": 1.61696,
+      "grad_norm": 2.073306145750615,
+      "learning_rate": 9.248433761956573e-05,
+      "loss": 1.2824,
+      "step": 10106
+    },
+    {
+      "epoch": 1.61712,
+      "grad_norm": 2.4706990129239648,
+      "learning_rate": 9.246711275010463e-05,
+      "loss": 1.2867,
+      "step": 10107
+    },
+    {
+      "epoch": 1.61728,
+      "grad_norm": 2.182881051700165,
+      "learning_rate": 9.244988810541387e-05,
+      "loss": 1.3063,
+      "step": 10108
+    },
+    {
+      "epoch": 1.61744,
+      "grad_norm": 2.12957824002543,
+      "learning_rate": 9.243266368600749e-05,
+      "loss": 1.3012,
+      "step": 10109
+    },
+    {
+      "epoch": 1.6176,
+      "grad_norm": 2.194466093446013,
+      "learning_rate": 9.241543949239934e-05,
+      "loss": 1.3382,
+      "step": 10110
+    },
+    {
+      "epoch": 1.61776,
+      "grad_norm": 2.070288508176117,
+      "learning_rate": 9.239821552510345e-05,
+      "loss": 1.3465,
+      "step": 10111
+    },
+    {
+      "epoch": 1.61792,
+      "grad_norm": 2.486570911425829,
+      "learning_rate": 9.238099178463376e-05,
+      "loss": 1.315,
+      "step": 10112
+    },
+    {
+      "epoch": 1.61808,
+      "grad_norm": 2.111797583605007,
+      "learning_rate": 9.236376827150412e-05,
+      "loss": 1.2398,
+      "step": 10113
+    },
+    {
+      "epoch": 1.6182400000000001,
+      "grad_norm": 1.941581604717175,
+      "learning_rate": 9.234654498622855e-05,
+      "loss": 1.2841,
+      "step": 10114
+    },
+    {
+      "epoch": 1.6183999999999998,
+      "grad_norm": 1.9225168164868534,
+      "learning_rate": 9.232932192932089e-05,
+      "loss": 1.3183,
+      "step": 10115
+    },
+    {
+      "epoch": 1.61856,
+      "grad_norm": 2.461955098147224,
+      "learning_rate": 9.231209910129513e-05,
+      "loss": 1.4342,
+      "step": 10116
+    },
+    {
+      "epoch": 1.6187200000000002,
+      "grad_norm": 1.8494496225312356,
+      "learning_rate": 9.229487650266509e-05,
+      "loss": 1.2921,
+      "step": 10117
+    },
+    {
+      "epoch": 1.6188799999999999,
+      "grad_norm": 2.6122522014510206,
+      "learning_rate": 9.227765413394476e-05,
+      "loss": 1.255,
+      "step": 10118
+    },
+    {
+      "epoch": 1.61904,
+      "grad_norm": 2.189560428038332,
+      "learning_rate": 9.226043199564793e-05,
+      "loss": 1.2746,
+      "step": 10119
+    },
+    {
+      "epoch": 1.6192,
+      "grad_norm": 2.028359297867644,
+      "learning_rate": 9.224321008828855e-05,
+      "loss": 1.3669,
+      "step": 10120
+    },
+    {
+      "epoch": 1.61936,
+      "grad_norm": 2.3949812350142374,
+      "learning_rate": 9.222598841238047e-05,
+      "loss": 1.2715,
+      "step": 10121
+    },
+    {
+      "epoch": 1.61952,
+      "grad_norm": 2.58731577780364,
+      "learning_rate": 9.220876696843758e-05,
+      "loss": 1.3495,
+      "step": 10122
+    },
+    {
+      "epoch": 1.61968,
+      "grad_norm": 1.8260621555776313,
+      "learning_rate": 9.21915457569737e-05,
+      "loss": 1.3203,
+      "step": 10123
+    },
+    {
+      "epoch": 1.61984,
+      "grad_norm": 3.349872748934355,
+      "learning_rate": 9.217432477850276e-05,
+      "loss": 1.2461,
+      "step": 10124
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 2.1410141137748644,
+      "learning_rate": 9.215710403353852e-05,
+      "loss": 1.2895,
+      "step": 10125
+    },
+    {
+      "epoch": 1.62016,
+      "grad_norm": 1.9664750788153922,
+      "learning_rate": 9.213988352259492e-05,
+      "loss": 1.3169,
+      "step": 10126
+    },
+    {
+      "epoch": 1.62032,
+      "grad_norm": 2.3186033052733768,
+      "learning_rate": 9.21226632461857e-05,
+      "loss": 1.3088,
+      "step": 10127
+    },
+    {
+      "epoch": 1.6204800000000001,
+      "grad_norm": 2.3014468115954596,
+      "learning_rate": 9.210544320482477e-05,
+      "loss": 1.3154,
+      "step": 10128
+    },
+    {
+      "epoch": 1.6206399999999999,
+      "grad_norm": 2.0944826081091104,
+      "learning_rate": 9.208822339902586e-05,
+      "loss": 1.2686,
+      "step": 10129
+    },
+    {
+      "epoch": 1.6208,
+      "grad_norm": 2.170548808349764,
+      "learning_rate": 9.207100382930288e-05,
+      "loss": 1.2984,
+      "step": 10130
+    },
+    {
+      "epoch": 1.62096,
+      "grad_norm": 1.98446469853381,
+      "learning_rate": 9.205378449616953e-05,
+      "loss": 1.35,
+      "step": 10131
+    },
+    {
+      "epoch": 1.62112,
+      "grad_norm": 1.8563903986121137,
+      "learning_rate": 9.203656540013975e-05,
+      "loss": 1.2876,
+      "step": 10132
+    },
+    {
+      "epoch": 1.62128,
+      "grad_norm": 1.9359526522983115,
+      "learning_rate": 9.20193465417272e-05,
+      "loss": 1.2486,
+      "step": 10133
+    },
+    {
+      "epoch": 1.62144,
+      "grad_norm": 2.4817703338323134,
+      "learning_rate": 9.200212792144577e-05,
+      "loss": 1.2849,
+      "step": 10134
+    },
+    {
+      "epoch": 1.6216,
+      "grad_norm": 2.0480936395769116,
+      "learning_rate": 9.198490953980914e-05,
+      "loss": 1.3057,
+      "step": 10135
+    },
+    {
+      "epoch": 1.62176,
+      "grad_norm": 2.1113149209308792,
+      "learning_rate": 9.196769139733117e-05,
+      "loss": 1.3749,
+      "step": 10136
+    },
+    {
+      "epoch": 1.62192,
+      "grad_norm": 2.399453299113869,
+      "learning_rate": 9.195047349452553e-05,
+      "loss": 1.3178,
+      "step": 10137
+    },
+    {
+      "epoch": 1.62208,
+      "grad_norm": 1.8963707585141742,
+      "learning_rate": 9.193325583190608e-05,
+      "loss": 1.3043,
+      "step": 10138
+    },
+    {
+      "epoch": 1.6222400000000001,
+      "grad_norm": 2.352106537746045,
+      "learning_rate": 9.191603840998649e-05,
+      "loss": 1.4027,
+      "step": 10139
+    },
+    {
+      "epoch": 1.6223999999999998,
+      "grad_norm": 2.08316845855348,
+      "learning_rate": 9.189882122928058e-05,
+      "loss": 1.4088,
+      "step": 10140
+    },
+    {
+      "epoch": 1.62256,
+      "grad_norm": 2.204119876419778,
+      "learning_rate": 9.188160429030198e-05,
+      "loss": 1.3023,
+      "step": 10141
+    },
+    {
+      "epoch": 1.6227200000000002,
+      "grad_norm": 2.1533410806854567,
+      "learning_rate": 9.186438759356453e-05,
+      "loss": 1.3774,
+      "step": 10142
+    },
+    {
+      "epoch": 1.6228799999999999,
+      "grad_norm": 2.285114131199537,
+      "learning_rate": 9.184717113958185e-05,
+      "loss": 1.3084,
+      "step": 10143
+    },
+    {
+      "epoch": 1.62304,
+      "grad_norm": 2.1115219902418314,
+      "learning_rate": 9.182995492886774e-05,
+      "loss": 1.3846,
+      "step": 10144
+    },
+    {
+      "epoch": 1.6232,
+      "grad_norm": 2.0948536901811403,
+      "learning_rate": 9.181273896193587e-05,
+      "loss": 1.323,
+      "step": 10145
+    },
+    {
+      "epoch": 1.62336,
+      "grad_norm": 1.9236026820319738,
+      "learning_rate": 9.179552323929994e-05,
+      "loss": 1.2595,
+      "step": 10146
+    },
+    {
+      "epoch": 1.62352,
+      "grad_norm": 2.2200225972208254,
+      "learning_rate": 9.17783077614736e-05,
+      "loss": 1.2767,
+      "step": 10147
+    },
+    {
+      "epoch": 1.62368,
+      "grad_norm": 2.0273683269190266,
+      "learning_rate": 9.176109252897063e-05,
+      "loss": 1.2717,
+      "step": 10148
+    },
+    {
+      "epoch": 1.62384,
+      "grad_norm": 2.971285085065248,
+      "learning_rate": 9.174387754230462e-05,
+      "loss": 1.2468,
+      "step": 10149
+    },
+    {
+      "epoch": 1.624,
+      "grad_norm": 1.744337919348975,
+      "learning_rate": 9.17266628019893e-05,
+      "loss": 1.2448,
+      "step": 10150
+    },
+    {
+      "epoch": 1.62416,
+      "grad_norm": 2.2957745021923643,
+      "learning_rate": 9.170944830853827e-05,
+      "loss": 1.3142,
+      "step": 10151
+    },
+    {
+      "epoch": 1.62432,
+      "grad_norm": 2.2312514336968157,
+      "learning_rate": 9.169223406246525e-05,
+      "loss": 1.298,
+      "step": 10152
+    },
+    {
+      "epoch": 1.6244800000000001,
+      "grad_norm": 2.078429104539196,
+      "learning_rate": 9.167502006428386e-05,
+      "loss": 1.2385,
+      "step": 10153
+    },
+    {
+      "epoch": 1.6246399999999999,
+      "grad_norm": 2.305500503396384,
+      "learning_rate": 9.165780631450774e-05,
+      "loss": 1.3523,
+      "step": 10154
+    },
+    {
+      "epoch": 1.6248,
+      "grad_norm": 2.064434105166101,
+      "learning_rate": 9.16405928136505e-05,
+      "loss": 1.2369,
+      "step": 10155
+    },
+    {
+      "epoch": 1.62496,
+      "grad_norm": 2.0308849344085793,
+      "learning_rate": 9.162337956222582e-05,
+      "loss": 1.2873,
+      "step": 10156
+    },
+    {
+      "epoch": 1.62512,
+      "grad_norm": 2.0111830606913768,
+      "learning_rate": 9.160616656074725e-05,
+      "loss": 1.3304,
+      "step": 10157
+    },
+    {
+      "epoch": 1.62528,
+      "grad_norm": 2.5360182518279784,
+      "learning_rate": 9.15889538097285e-05,
+      "loss": 1.3234,
+      "step": 10158
+    },
+    {
+      "epoch": 1.62544,
+      "grad_norm": 2.1378614284180535,
+      "learning_rate": 9.157174130968304e-05,
+      "loss": 1.3307,
+      "step": 10159
+    },
+    {
+      "epoch": 1.6256,
+      "grad_norm": 1.9822112141585106,
+      "learning_rate": 9.155452906112458e-05,
+      "loss": 1.2045,
+      "step": 10160
+    },
+    {
+      "epoch": 1.62576,
+      "grad_norm": 2.502928919965081,
+      "learning_rate": 9.153731706456666e-05,
+      "loss": 1.3005,
+      "step": 10161
+    },
+    {
+      "epoch": 1.62592,
+      "grad_norm": 1.9108211601329848,
+      "learning_rate": 9.152010532052287e-05,
+      "loss": 1.3434,
+      "step": 10162
+    },
+    {
+      "epoch": 1.62608,
+      "grad_norm": 2.3836477550625648,
+      "learning_rate": 9.150289382950677e-05,
+      "loss": 1.2909,
+      "step": 10163
+    },
+    {
+      "epoch": 1.6262400000000001,
+      "grad_norm": 1.9749150353551448,
+      "learning_rate": 9.148568259203197e-05,
+      "loss": 1.2655,
+      "step": 10164
+    },
+    {
+      "epoch": 1.6263999999999998,
+      "grad_norm": 2.1052405954210447,
+      "learning_rate": 9.146847160861194e-05,
+      "loss": 1.323,
+      "step": 10165
+    },
+    {
+      "epoch": 1.62656,
+      "grad_norm": 2.620048330612001,
+      "learning_rate": 9.145126087976034e-05,
+      "loss": 1.3348,
+      "step": 10166
+    },
+    {
+      "epoch": 1.6267200000000002,
+      "grad_norm": 2.1186136801898123,
+      "learning_rate": 9.143405040599063e-05,
+      "loss": 1.2695,
+      "step": 10167
+    },
+    {
+      "epoch": 1.6268799999999999,
+      "grad_norm": 2.1235849023151117,
+      "learning_rate": 9.141684018781634e-05,
+      "loss": 1.3265,
+      "step": 10168
+    },
+    {
+      "epoch": 1.62704,
+      "grad_norm": 2.027619246078777,
+      "learning_rate": 9.13996302257511e-05,
+      "loss": 1.2775,
+      "step": 10169
+    },
+    {
+      "epoch": 1.6272,
+      "grad_norm": 2.1495324447004345,
+      "learning_rate": 9.138242052030831e-05,
+      "loss": 1.2681,
+      "step": 10170
+    },
+    {
+      "epoch": 1.62736,
+      "grad_norm": 2.150953347002324,
+      "learning_rate": 9.136521107200156e-05,
+      "loss": 1.3342,
+      "step": 10171
+    },
+    {
+      "epoch": 1.62752,
+      "grad_norm": 2.284658405464728,
+      "learning_rate": 9.13480018813443e-05,
+      "loss": 1.2557,
+      "step": 10172
+    },
+    {
+      "epoch": 1.62768,
+      "grad_norm": 2.036880583431236,
+      "learning_rate": 9.13307929488501e-05,
+      "loss": 1.3689,
+      "step": 10173
+    },
+    {
+      "epoch": 1.62784,
+      "grad_norm": 2.3039838449521093,
+      "learning_rate": 9.131358427503238e-05,
+      "loss": 1.3163,
+      "step": 10174
+    },
+    {
+      "epoch": 1.6280000000000001,
+      "grad_norm": 1.994563508603159,
+      "learning_rate": 9.129637586040468e-05,
+      "loss": 1.2832,
+      "step": 10175
+    },
+    {
+      "epoch": 1.62816,
+      "grad_norm": 2.588065510229754,
+      "learning_rate": 9.12791677054804e-05,
+      "loss": 1.3556,
+      "step": 10176
+    },
+    {
+      "epoch": 1.62832,
+      "grad_norm": 2.1102017723045754,
+      "learning_rate": 9.126195981077309e-05,
+      "loss": 1.2881,
+      "step": 10177
+    },
+    {
+      "epoch": 1.6284800000000001,
+      "grad_norm": 2.3885974084967043,
+      "learning_rate": 9.124475217679612e-05,
+      "loss": 1.2614,
+      "step": 10178
+    },
+    {
+      "epoch": 1.6286399999999999,
+      "grad_norm": 1.9010260828927197,
+      "learning_rate": 9.122754480406301e-05,
+      "loss": 1.3355,
+      "step": 10179
+    },
+    {
+      "epoch": 1.6288,
+      "grad_norm": 2.2492516076982496,
+      "learning_rate": 9.121033769308716e-05,
+      "loss": 1.352,
+      "step": 10180
+    },
+    {
+      "epoch": 1.62896,
+      "grad_norm": 2.117531627093576,
+      "learning_rate": 9.119313084438207e-05,
+      "loss": 1.3323,
+      "step": 10181
+    },
+    {
+      "epoch": 1.62912,
+      "grad_norm": 2.1525766439894176,
+      "learning_rate": 9.117592425846107e-05,
+      "loss": 1.4227,
+      "step": 10182
+    },
+    {
+      "epoch": 1.62928,
+      "grad_norm": 2.300756537695544,
+      "learning_rate": 9.115871793583769e-05,
+      "loss": 1.3371,
+      "step": 10183
+    },
+    {
+      "epoch": 1.62944,
+      "grad_norm": 1.8421950862878238,
+      "learning_rate": 9.114151187702524e-05,
+      "loss": 1.283,
+      "step": 10184
+    },
+    {
+      "epoch": 1.6296,
+      "grad_norm": 2.9398170713313716,
+      "learning_rate": 9.112430608253721e-05,
+      "loss": 1.3076,
+      "step": 10185
+    },
+    {
+      "epoch": 1.62976,
+      "grad_norm": 2.0582091536429545,
+      "learning_rate": 9.110710055288689e-05,
+      "loss": 1.3336,
+      "step": 10186
+    },
+    {
+      "epoch": 1.62992,
+      "grad_norm": 2.257058986118604,
+      "learning_rate": 9.108989528858777e-05,
+      "loss": 1.3889,
+      "step": 10187
+    },
+    {
+      "epoch": 1.63008,
+      "grad_norm": 2.3023167785905887,
+      "learning_rate": 9.107269029015317e-05,
+      "loss": 1.2725,
+      "step": 10188
+    },
+    {
+      "epoch": 1.6302400000000001,
+      "grad_norm": 2.0708777125029365,
+      "learning_rate": 9.105548555809653e-05,
+      "loss": 1.3141,
+      "step": 10189
+    },
+    {
+      "epoch": 1.6303999999999998,
+      "grad_norm": 2.166948077498621,
+      "learning_rate": 9.103828109293111e-05,
+      "loss": 1.3012,
+      "step": 10190
+    },
+    {
+      "epoch": 1.63056,
+      "grad_norm": 2.783275755041253,
+      "learning_rate": 9.102107689517037e-05,
+      "loss": 1.3264,
+      "step": 10191
+    },
+    {
+      "epoch": 1.63072,
+      "grad_norm": 2.45331353560529,
+      "learning_rate": 9.100387296532757e-05,
+      "loss": 1.355,
+      "step": 10192
+    },
+    {
+      "epoch": 1.6308799999999999,
+      "grad_norm": 2.137426845525394,
+      "learning_rate": 9.098666930391615e-05,
+      "loss": 1.3233,
+      "step": 10193
+    },
+    {
+      "epoch": 1.63104,
+      "grad_norm": 2.757028008925231,
+      "learning_rate": 9.096946591144933e-05,
+      "loss": 1.3835,
+      "step": 10194
+    },
+    {
+      "epoch": 1.6312,
+      "grad_norm": 3.1515564514529215,
+      "learning_rate": 9.09522627884405e-05,
+      "loss": 1.3167,
+      "step": 10195
+    },
+    {
+      "epoch": 1.63136,
+      "grad_norm": 2.238239201141696,
+      "learning_rate": 9.093505993540295e-05,
+      "loss": 1.2812,
+      "step": 10196
+    },
+    {
+      "epoch": 1.63152,
+      "grad_norm": 1.9459701206835593,
+      "learning_rate": 9.091785735285006e-05,
+      "loss": 1.3055,
+      "step": 10197
+    },
+    {
+      "epoch": 1.63168,
+      "grad_norm": 2.102314888304819,
+      "learning_rate": 9.090065504129501e-05,
+      "loss": 1.3101,
+      "step": 10198
+    },
+    {
+      "epoch": 1.63184,
+      "grad_norm": 3.002475602002844,
+      "learning_rate": 9.08834530012512e-05,
+      "loss": 1.2545,
+      "step": 10199
+    },
+    {
+      "epoch": 1.6320000000000001,
+      "grad_norm": 2.1065215328678795,
+      "learning_rate": 9.086625123323183e-05,
+      "loss": 1.2656,
+      "step": 10200
+    },
+    {
+      "epoch": 1.63216,
+      "grad_norm": 2.1740961317158565,
+      "learning_rate": 9.084904973775025e-05,
+      "loss": 1.2895,
+      "step": 10201
+    },
+    {
+      "epoch": 1.63232,
+      "grad_norm": 3.747927895453893,
+      "learning_rate": 9.083184851531966e-05,
+      "loss": 1.3814,
+      "step": 10202
+    },
+    {
+      "epoch": 1.6324800000000002,
+      "grad_norm": 1.9542610778348048,
+      "learning_rate": 9.081464756645337e-05,
+      "loss": 1.3316,
+      "step": 10203
+    },
+    {
+      "epoch": 1.6326399999999999,
+      "grad_norm": 2.9924905775764206,
+      "learning_rate": 9.079744689166459e-05,
+      "loss": 1.3502,
+      "step": 10204
+    },
+    {
+      "epoch": 1.6328,
+      "grad_norm": 3.118837358274488,
+      "learning_rate": 9.078024649146662e-05,
+      "loss": 1.3836,
+      "step": 10205
+    },
+    {
+      "epoch": 1.63296,
+      "grad_norm": 2.0077250859576923,
+      "learning_rate": 9.076304636637263e-05,
+      "loss": 1.2441,
+      "step": 10206
+    },
+    {
+      "epoch": 1.63312,
+      "grad_norm": 2.3314415531420942,
+      "learning_rate": 9.074584651689591e-05,
+      "loss": 1.2561,
+      "step": 10207
+    },
+    {
+      "epoch": 1.63328,
+      "grad_norm": 2.7704057201216488,
+      "learning_rate": 9.072864694354961e-05,
+      "loss": 1.2964,
+      "step": 10208
+    },
+    {
+      "epoch": 1.63344,
+      "grad_norm": 2.1216079224031654,
+      "learning_rate": 9.071144764684702e-05,
+      "loss": 1.3702,
+      "step": 10209
+    },
+    {
+      "epoch": 1.6336,
+      "grad_norm": 3.012081402184221,
+      "learning_rate": 9.069424862730124e-05,
+      "loss": 1.3251,
+      "step": 10210
+    },
+    {
+      "epoch": 1.63376,
+      "grad_norm": 2.686949490076241,
+      "learning_rate": 9.067704988542556e-05,
+      "loss": 1.2948,
+      "step": 10211
+    },
+    {
+      "epoch": 1.63392,
+      "grad_norm": 2.1384895542226734,
+      "learning_rate": 9.065985142173308e-05,
+      "loss": 1.3493,
+      "step": 10212
+    },
+    {
+      "epoch": 1.63408,
+      "grad_norm": 3.126421340937311,
+      "learning_rate": 9.064265323673709e-05,
+      "loss": 1.3647,
+      "step": 10213
+    },
+    {
+      "epoch": 1.6342400000000001,
+      "grad_norm": 2.450657369430489,
+      "learning_rate": 9.062545533095063e-05,
+      "loss": 1.2814,
+      "step": 10214
+    },
+    {
+      "epoch": 1.6343999999999999,
+      "grad_norm": 2.083330175819898,
+      "learning_rate": 9.0608257704887e-05,
+      "loss": 1.3045,
+      "step": 10215
+    },
+    {
+      "epoch": 1.63456,
+      "grad_norm": 2.204249891061981,
+      "learning_rate": 9.05910603590592e-05,
+      "loss": 1.2766,
+      "step": 10216
+    },
+    {
+      "epoch": 1.63472,
+      "grad_norm": 2.37318665017752,
+      "learning_rate": 9.057386329398049e-05,
+      "loss": 1.3188,
+      "step": 10217
+    },
+    {
+      "epoch": 1.6348799999999999,
+      "grad_norm": 1.9105806228518112,
+      "learning_rate": 9.055666651016395e-05,
+      "loss": 1.2606,
+      "step": 10218
+    },
+    {
+      "epoch": 1.63504,
+      "grad_norm": 2.308385651313117,
+      "learning_rate": 9.053947000812272e-05,
+      "loss": 1.2469,
+      "step": 10219
+    },
+    {
+      "epoch": 1.6352,
+      "grad_norm": 2.0176396827701115,
+      "learning_rate": 9.052227378836989e-05,
+      "loss": 1.2762,
+      "step": 10220
+    },
+    {
+      "epoch": 1.63536,
+      "grad_norm": 2.139681423567448,
+      "learning_rate": 9.050507785141866e-05,
+      "loss": 1.3439,
+      "step": 10221
+    },
+    {
+      "epoch": 1.63552,
+      "grad_norm": 2.1641419346554907,
+      "learning_rate": 9.048788219778202e-05,
+      "loss": 1.3378,
+      "step": 10222
+    },
+    {
+      "epoch": 1.63568,
+      "grad_norm": 1.9524832423988352,
+      "learning_rate": 9.047068682797318e-05,
+      "loss": 1.323,
+      "step": 10223
+    },
+    {
+      "epoch": 1.63584,
+      "grad_norm": 2.1468778469368073,
+      "learning_rate": 9.045349174250508e-05,
+      "loss": 1.2818,
+      "step": 10224
+    },
+    {
+      "epoch": 1.6360000000000001,
+      "grad_norm": 2.124443920364309,
+      "learning_rate": 9.043629694189094e-05,
+      "loss": 1.2618,
+      "step": 10225
+    },
+    {
+      "epoch": 1.6361599999999998,
+      "grad_norm": 2.1513839259234584,
+      "learning_rate": 9.041910242664372e-05,
+      "loss": 1.3611,
+      "step": 10226
+    },
+    {
+      "epoch": 1.63632,
+      "grad_norm": 1.85889016489573,
+      "learning_rate": 9.040190819727652e-05,
+      "loss": 1.2842,
+      "step": 10227
+    },
+    {
+      "epoch": 1.6364800000000002,
+      "grad_norm": 2.2119728769302434,
+      "learning_rate": 9.03847142543024e-05,
+      "loss": 1.2963,
+      "step": 10228
+    },
+    {
+      "epoch": 1.6366399999999999,
+      "grad_norm": 1.9593698818760918,
+      "learning_rate": 9.03675205982344e-05,
+      "loss": 1.3584,
+      "step": 10229
+    },
+    {
+      "epoch": 1.6368,
+      "grad_norm": 1.9275749901880803,
+      "learning_rate": 9.035032722958558e-05,
+      "loss": 1.2755,
+      "step": 10230
+    },
+    {
+      "epoch": 1.63696,
+      "grad_norm": 1.8128197765465541,
+      "learning_rate": 9.03331341488689e-05,
+      "loss": 1.288,
+      "step": 10231
+    },
+    {
+      "epoch": 1.63712,
+      "grad_norm": 2.6189628065186086,
+      "learning_rate": 9.031594135659743e-05,
+      "loss": 1.3411,
+      "step": 10232
+    },
+    {
+      "epoch": 1.63728,
+      "grad_norm": 2.2440426622390786,
+      "learning_rate": 9.029874885328413e-05,
+      "loss": 1.3331,
+      "step": 10233
+    },
+    {
+      "epoch": 1.63744,
+      "grad_norm": 1.984314256129323,
+      "learning_rate": 9.028155663944207e-05,
+      "loss": 1.3118,
+      "step": 10234
+    },
+    {
+      "epoch": 1.6376,
+      "grad_norm": 2.132578921851436,
+      "learning_rate": 9.026436471558416e-05,
+      "loss": 1.349,
+      "step": 10235
+    },
+    {
+      "epoch": 1.63776,
+      "grad_norm": 2.0305476675335976,
+      "learning_rate": 9.024717308222344e-05,
+      "loss": 1.3122,
+      "step": 10236
+    },
+    {
+      "epoch": 1.63792,
+      "grad_norm": 2.913892762331326,
+      "learning_rate": 9.022998173987285e-05,
+      "loss": 1.2797,
+      "step": 10237
+    },
+    {
+      "epoch": 1.63808,
+      "grad_norm": 2.0790608595062823,
+      "learning_rate": 9.02127906890454e-05,
+      "loss": 1.3278,
+      "step": 10238
+    },
+    {
+      "epoch": 1.6382400000000001,
+      "grad_norm": 2.204987709941652,
+      "learning_rate": 9.019559993025398e-05,
+      "loss": 1.3181,
+      "step": 10239
+    },
+    {
+      "epoch": 1.6383999999999999,
+      "grad_norm": 1.8862523548221457,
+      "learning_rate": 9.017840946401161e-05,
+      "loss": 1.2374,
+      "step": 10240
+    },
+    {
+      "epoch": 1.63856,
+      "grad_norm": 2.6555282831348874,
+      "learning_rate": 9.016121929083115e-05,
+      "loss": 1.3107,
+      "step": 10241
+    },
+    {
+      "epoch": 1.63872,
+      "grad_norm": 1.9933535799355362,
+      "learning_rate": 9.014402941122561e-05,
+      "loss": 1.3141,
+      "step": 10242
+    },
+    {
+      "epoch": 1.63888,
+      "grad_norm": 2.607733618079008,
+      "learning_rate": 9.012683982570782e-05,
+      "loss": 1.2752,
+      "step": 10243
+    },
+    {
+      "epoch": 1.63904,
+      "grad_norm": 2.80267107684195,
+      "learning_rate": 9.010965053479078e-05,
+      "loss": 1.2655,
+      "step": 10244
+    },
+    {
+      "epoch": 1.6392,
+      "grad_norm": 2.0646314154114975,
+      "learning_rate": 9.009246153898733e-05,
+      "loss": 1.2895,
+      "step": 10245
+    },
+    {
+      "epoch": 1.63936,
+      "grad_norm": 2.8392096233446744,
+      "learning_rate": 9.007527283881042e-05,
+      "loss": 1.3609,
+      "step": 10246
+    },
+    {
+      "epoch": 1.63952,
+      "grad_norm": 2.5930659601347967,
+      "learning_rate": 9.005808443477288e-05,
+      "loss": 1.3522,
+      "step": 10247
+    },
+    {
+      "epoch": 1.63968,
+      "grad_norm": 1.8972952370522527,
+      "learning_rate": 9.004089632738765e-05,
+      "loss": 1.2906,
+      "step": 10248
+    },
+    {
+      "epoch": 1.63984,
+      "grad_norm": 2.4594685734408612,
+      "learning_rate": 9.002370851716752e-05,
+      "loss": 1.3234,
+      "step": 10249
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 2.7479424509313324,
+      "learning_rate": 9.000652100462543e-05,
+      "loss": 1.3485,
+      "step": 10250
+    },
+    {
+      "epoch": 1.6401599999999998,
+      "grad_norm": 2.2116800852232656,
+      "learning_rate": 8.998933379027416e-05,
+      "loss": 1.4052,
+      "step": 10251
+    },
+    {
+      "epoch": 1.64032,
+      "grad_norm": 3.232655323647722,
+      "learning_rate": 8.997214687462658e-05,
+      "loss": 1.2954,
+      "step": 10252
+    },
+    {
+      "epoch": 1.6404800000000002,
+      "grad_norm": 1.9763499606435204,
+      "learning_rate": 8.995496025819554e-05,
+      "loss": 1.2905,
+      "step": 10253
+    },
+    {
+      "epoch": 1.6406399999999999,
+      "grad_norm": 2.4520072832887494,
+      "learning_rate": 8.993777394149386e-05,
+      "loss": 1.247,
+      "step": 10254
+    },
+    {
+      "epoch": 1.6408,
+      "grad_norm": 2.4028158130391053,
+      "learning_rate": 8.99205879250343e-05,
+      "loss": 1.3463,
+      "step": 10255
+    },
+    {
+      "epoch": 1.64096,
+      "grad_norm": 1.8970017007036621,
+      "learning_rate": 8.990340220932978e-05,
+      "loss": 1.3062,
+      "step": 10256
+    },
+    {
+      "epoch": 1.64112,
+      "grad_norm": 2.329963085870251,
+      "learning_rate": 8.988621679489296e-05,
+      "loss": 1.3534,
+      "step": 10257
+    },
+    {
+      "epoch": 1.64128,
+      "grad_norm": 2.464950144120055,
+      "learning_rate": 8.986903168223675e-05,
+      "loss": 1.3564,
+      "step": 10258
+    },
+    {
+      "epoch": 1.64144,
+      "grad_norm": 2.011968554823058,
+      "learning_rate": 8.98518468718738e-05,
+      "loss": 1.3371,
+      "step": 10259
+    },
+    {
+      "epoch": 1.6416,
+      "grad_norm": 2.1429135892406665,
+      "learning_rate": 8.983466236431701e-05,
+      "loss": 1.2997,
+      "step": 10260
+    },
+    {
+      "epoch": 1.64176,
+      "grad_norm": 2.043477308386063,
+      "learning_rate": 8.981747816007905e-05,
+      "loss": 1.287,
+      "step": 10261
+    },
+    {
+      "epoch": 1.64192,
+      "grad_norm": 2.0949504056133863,
+      "learning_rate": 8.980029425967274e-05,
+      "loss": 1.3642,
+      "step": 10262
+    },
+    {
+      "epoch": 1.64208,
+      "grad_norm": 1.9901843031270867,
+      "learning_rate": 8.978311066361075e-05,
+      "loss": 1.2973,
+      "step": 10263
+    },
+    {
+      "epoch": 1.6422400000000001,
+      "grad_norm": 1.919578338710487,
+      "learning_rate": 8.976592737240588e-05,
+      "loss": 1.3582,
+      "step": 10264
+    },
+    {
+      "epoch": 1.6423999999999999,
+      "grad_norm": 2.2453184025908026,
+      "learning_rate": 8.97487443865708e-05,
+      "loss": 1.3228,
+      "step": 10265
+    },
+    {
+      "epoch": 1.64256,
+      "grad_norm": 2.374581368593367,
+      "learning_rate": 8.973156170661828e-05,
+      "loss": 1.2825,
+      "step": 10266
+    },
+    {
+      "epoch": 1.64272,
+      "grad_norm": 2.2831590504196964,
+      "learning_rate": 8.971437933306096e-05,
+      "loss": 1.3367,
+      "step": 10267
+    },
+    {
+      "epoch": 1.64288,
+      "grad_norm": 2.0543263891651957,
+      "learning_rate": 8.96971972664116e-05,
+      "loss": 1.3287,
+      "step": 10268
+    },
+    {
+      "epoch": 1.64304,
+      "grad_norm": 1.9649014675559215,
+      "learning_rate": 8.968001550718283e-05,
+      "loss": 1.2562,
+      "step": 10269
+    },
+    {
+      "epoch": 1.6432,
+      "grad_norm": 2.4421359856424005,
+      "learning_rate": 8.966283405588741e-05,
+      "loss": 1.2768,
+      "step": 10270
+    },
+    {
+      "epoch": 1.64336,
+      "grad_norm": 2.059638407419992,
+      "learning_rate": 8.96456529130379e-05,
+      "loss": 1.3831,
+      "step": 10271
+    },
+    {
+      "epoch": 1.64352,
+      "grad_norm": 2.344320176601425,
+      "learning_rate": 8.962847207914707e-05,
+      "loss": 1.2635,
+      "step": 10272
+    },
+    {
+      "epoch": 1.64368,
+      "grad_norm": 2.491693121279653,
+      "learning_rate": 8.961129155472748e-05,
+      "loss": 1.3084,
+      "step": 10273
+    },
+    {
+      "epoch": 1.64384,
+      "grad_norm": 2.367076549371859,
+      "learning_rate": 8.959411134029187e-05,
+      "loss": 1.2396,
+      "step": 10274
+    },
+    {
+      "epoch": 1.6440000000000001,
+      "grad_norm": 2.193870317492558,
+      "learning_rate": 8.957693143635275e-05,
+      "loss": 1.3308,
+      "step": 10275
+    },
+    {
+      "epoch": 1.6441599999999998,
+      "grad_norm": 2.2162205775435693,
+      "learning_rate": 8.955975184342283e-05,
+      "loss": 1.2784,
+      "step": 10276
+    },
+    {
+      "epoch": 1.64432,
+      "grad_norm": 2.4124864495268343,
+      "learning_rate": 8.954257256201468e-05,
+      "loss": 1.3014,
+      "step": 10277
+    },
+    {
+      "epoch": 1.6444800000000002,
+      "grad_norm": 2.071701044417173,
+      "learning_rate": 8.952539359264098e-05,
+      "loss": 1.2959,
+      "step": 10278
+    },
+    {
+      "epoch": 1.6446399999999999,
+      "grad_norm": 3.1510348754879742,
+      "learning_rate": 8.950821493581422e-05,
+      "loss": 1.3663,
+      "step": 10279
+    },
+    {
+      "epoch": 1.6448,
+      "grad_norm": 2.0672286139256277,
+      "learning_rate": 8.94910365920471e-05,
+      "loss": 1.3574,
+      "step": 10280
+    },
+    {
+      "epoch": 1.64496,
+      "grad_norm": 2.314026802621631,
+      "learning_rate": 8.947385856185208e-05,
+      "loss": 1.284,
+      "step": 10281
+    },
+    {
+      "epoch": 1.64512,
+      "grad_norm": 2.996817995505046,
+      "learning_rate": 8.945668084574181e-05,
+      "loss": 1.2975,
+      "step": 10282
+    },
+    {
+      "epoch": 1.64528,
+      "grad_norm": 2.2396609202247335,
+      "learning_rate": 8.94395034442288e-05,
+      "loss": 1.268,
+      "step": 10283
+    },
+    {
+      "epoch": 1.64544,
+      "grad_norm": 2.020352154721596,
+      "learning_rate": 8.942232635782562e-05,
+      "loss": 1.3278,
+      "step": 10284
+    },
+    {
+      "epoch": 1.6456,
+      "grad_norm": 2.020352154721596,
+      "learning_rate": 8.942232635782562e-05,
+      "loss": 1.3166,
+      "step": 10285
+    },
+    {
+      "epoch": 1.6457600000000001,
+      "grad_norm": 2.9954741471333226,
+      "learning_rate": 8.94051495870448e-05,
+      "loss": 1.3038,
+      "step": 10286
+    },
+    {
+      "epoch": 1.64592,
+      "grad_norm": 2.7913985003945134,
+      "learning_rate": 8.938797313239888e-05,
+      "loss": 1.2731,
+      "step": 10287
+    },
+    {
+      "epoch": 1.64608,
+      "grad_norm": 2.337449756466775,
+      "learning_rate": 8.937079699440041e-05,
+      "loss": 1.3345,
+      "step": 10288
+    },
+    {
+      "epoch": 1.6462400000000001,
+      "grad_norm": 2.881933316594347,
+      "learning_rate": 8.935362117356184e-05,
+      "loss": 1.3108,
+      "step": 10289
+    },
+    {
+      "epoch": 1.6463999999999999,
+      "grad_norm": 2.0024281838439086,
+      "learning_rate": 8.933644567039574e-05,
+      "loss": 1.2999,
+      "step": 10290
+    },
+    {
+      "epoch": 1.64656,
+      "grad_norm": 2.007959768232671,
+      "learning_rate": 8.931927048541452e-05,
+      "loss": 1.2855,
+      "step": 10291
+    },
+    {
+      "epoch": 1.64672,
+      "grad_norm": 2.2080183386920833,
+      "learning_rate": 8.930209561913074e-05,
+      "loss": 1.3033,
+      "step": 10292
+    },
+    {
+      "epoch": 1.64688,
+      "grad_norm": 2.008367348912445,
+      "learning_rate": 8.928492107205681e-05,
+      "loss": 1.2774,
+      "step": 10293
+    },
+    {
+      "epoch": 1.64704,
+      "grad_norm": 2.0336014448589266,
+      "learning_rate": 8.926774684470523e-05,
+      "loss": 1.3061,
+      "step": 10294
+    },
+    {
+      "epoch": 1.6472,
+      "grad_norm": 1.9334271530186948,
+      "learning_rate": 8.925057293758842e-05,
+      "loss": 1.2594,
+      "step": 10295
+    },
+    {
+      "epoch": 1.64736,
+      "grad_norm": 2.0509977489887024,
+      "learning_rate": 8.92333993512189e-05,
+      "loss": 1.3362,
+      "step": 10296
+    },
+    {
+      "epoch": 1.64752,
+      "grad_norm": 2.0475747231681423,
+      "learning_rate": 8.921622608610902e-05,
+      "loss": 1.2544,
+      "step": 10297
+    },
+    {
+      "epoch": 1.64768,
+      "grad_norm": 2.131639641206925,
+      "learning_rate": 8.919905314277127e-05,
+      "loss": 1.2522,
+      "step": 10298
+    },
+    {
+      "epoch": 1.64784,
+      "grad_norm": 2.1001766135001403,
+      "learning_rate": 8.9181880521718e-05,
+      "loss": 1.3689,
+      "step": 10299
+    },
+    {
+      "epoch": 1.6480000000000001,
+      "grad_norm": 2.02383495448114,
+      "learning_rate": 8.916470822346169e-05,
+      "loss": 1.2865,
+      "step": 10300
+    },
+    {
+      "epoch": 1.6481599999999998,
+      "grad_norm": 1.9232425596848766,
+      "learning_rate": 8.914753624851466e-05,
+      "loss": 1.2976,
+      "step": 10301
+    },
+    {
+      "epoch": 1.64832,
+      "grad_norm": 2.408860697319923,
+      "learning_rate": 8.913036459738934e-05,
+      "loss": 1.3894,
+      "step": 10302
+    },
+    {
+      "epoch": 1.64848,
+      "grad_norm": 1.8669983088342939,
+      "learning_rate": 8.91131932705981e-05,
+      "loss": 1.2496,
+      "step": 10303
+    },
+    {
+      "epoch": 1.6486399999999999,
+      "grad_norm": 2.260345737083035,
+      "learning_rate": 8.909602226865334e-05,
+      "loss": 1.2608,
+      "step": 10304
+    },
+    {
+      "epoch": 1.6488,
+      "grad_norm": 2.3054692125461864,
+      "learning_rate": 8.907885159206736e-05,
+      "loss": 1.33,
+      "step": 10305
+    },
+    {
+      "epoch": 1.64896,
+      "grad_norm": 2.104026329393729,
+      "learning_rate": 8.906168124135257e-05,
+      "loss": 1.2363,
+      "step": 10306
+    },
+    {
+      "epoch": 1.64912,
+      "grad_norm": 2.122659579694764,
+      "learning_rate": 8.904451121702123e-05,
+      "loss": 1.2906,
+      "step": 10307
+    },
+    {
+      "epoch": 1.64928,
+      "grad_norm": 2.4815462200211535,
+      "learning_rate": 8.902734151958575e-05,
+      "loss": 1.3527,
+      "step": 10308
+    },
+    {
+      "epoch": 1.64944,
+      "grad_norm": 2.0582061915523333,
+      "learning_rate": 8.901017214955836e-05,
+      "loss": 1.3408,
+      "step": 10309
+    },
+    {
+      "epoch": 1.6496,
+      "grad_norm": 2.1874279512068364,
+      "learning_rate": 8.899300310745145e-05,
+      "loss": 1.2546,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6497600000000001,
+      "grad_norm": 2.403546537797011,
+      "learning_rate": 8.897583439377726e-05,
+      "loss": 1.281,
+      "step": 10311
+    },
+    {
+      "epoch": 1.64992,
+      "grad_norm": 1.849527992610508,
+      "learning_rate": 8.895866600904816e-05,
+      "loss": 1.2621,
+      "step": 10312
+    },
+    {
+      "epoch": 1.65008,
+      "grad_norm": 2.0166123526653394,
+      "learning_rate": 8.894149795377634e-05,
+      "loss": 1.2654,
+      "step": 10313
+    },
+    {
+      "epoch": 1.6502400000000002,
+      "grad_norm": 2.289893863574801,
+      "learning_rate": 8.892433022847414e-05,
+      "loss": 1.3125,
+      "step": 10314
+    },
+    {
+      "epoch": 1.6503999999999999,
+      "grad_norm": 1.8939165925146948,
+      "learning_rate": 8.890716283365374e-05,
+      "loss": 1.3258,
+      "step": 10315
+    },
+    {
+      "epoch": 1.65056,
+      "grad_norm": 2.093083367916446,
+      "learning_rate": 8.88899957698275e-05,
+      "loss": 1.2988,
+      "step": 10316
+    },
+    {
+      "epoch": 1.65072,
+      "grad_norm": 2.0615023820728844,
+      "learning_rate": 8.887282903750753e-05,
+      "loss": 1.2789,
+      "step": 10317
+    },
+    {
+      "epoch": 1.65088,
+      "grad_norm": 2.2100355473884523,
+      "learning_rate": 8.885566263720616e-05,
+      "loss": 1.3095,
+      "step": 10318
+    },
+    {
+      "epoch": 1.65104,
+      "grad_norm": 1.7810646413962257,
+      "learning_rate": 8.883849656943556e-05,
+      "loss": 1.2039,
+      "step": 10319
+    },
+    {
+      "epoch": 1.6512,
+      "grad_norm": 1.942594183990999,
+      "learning_rate": 8.8821330834708e-05,
+      "loss": 1.2765,
+      "step": 10320
+    },
+    {
+      "epoch": 1.65136,
+      "grad_norm": 2.0007110020184484,
+      "learning_rate": 8.880416543353559e-05,
+      "loss": 1.307,
+      "step": 10321
+    },
+    {
+      "epoch": 1.65152,
+      "grad_norm": 1.9742038851358983,
+      "learning_rate": 8.878700036643062e-05,
+      "loss": 1.2008,
+      "step": 10322
+    },
+    {
+      "epoch": 1.65168,
+      "grad_norm": 1.808674869297754,
+      "learning_rate": 8.876983563390516e-05,
+      "loss": 1.2448,
+      "step": 10323
+    },
+    {
+      "epoch": 1.65184,
+      "grad_norm": 1.92229562713748,
+      "learning_rate": 8.87526712364715e-05,
+      "loss": 1.2886,
+      "step": 10324
+    },
+    {
+      "epoch": 1.6520000000000001,
+      "grad_norm": 1.9725603361376784,
+      "learning_rate": 8.873550717464169e-05,
+      "loss": 1.2621,
+      "step": 10325
+    },
+    {
+      "epoch": 1.6521599999999999,
+      "grad_norm": 1.6132433409644475,
+      "learning_rate": 8.871834344892796e-05,
+      "loss": 1.278,
+      "step": 10326
+    },
+    {
+      "epoch": 1.65232,
+      "grad_norm": 2.0921639004332997,
+      "learning_rate": 8.87011800598424e-05,
+      "loss": 1.2657,
+      "step": 10327
+    },
+    {
+      "epoch": 1.65248,
+      "grad_norm": 2.3210651559116773,
+      "learning_rate": 8.86840170078972e-05,
+      "loss": 1.3661,
+      "step": 10328
+    },
+    {
+      "epoch": 1.6526399999999999,
+      "grad_norm": 2.0362744924854135,
+      "learning_rate": 8.866685429360441e-05,
+      "loss": 1.3135,
+      "step": 10329
+    },
+    {
+      "epoch": 1.6528,
+      "grad_norm": 2.019857356984012,
+      "learning_rate": 8.86496919174762e-05,
+      "loss": 1.3119,
+      "step": 10330
+    },
+    {
+      "epoch": 1.65296,
+      "grad_norm": 2.302369651381357,
+      "learning_rate": 8.863252988002461e-05,
+      "loss": 1.2515,
+      "step": 10331
+    },
+    {
+      "epoch": 1.65312,
+      "grad_norm": 1.8402812656897807,
+      "learning_rate": 8.861536818176182e-05,
+      "loss": 1.2271,
+      "step": 10332
+    },
+    {
+      "epoch": 1.65328,
+      "grad_norm": 1.854124193719967,
+      "learning_rate": 8.859820682319981e-05,
+      "loss": 1.3629,
+      "step": 10333
+    },
+    {
+      "epoch": 1.65344,
+      "grad_norm": 2.4116333894243347,
+      "learning_rate": 8.858104580485069e-05,
+      "loss": 1.3056,
+      "step": 10334
+    },
+    {
+      "epoch": 1.6536,
+      "grad_norm": 1.8887335422870455,
+      "learning_rate": 8.856388512722653e-05,
+      "loss": 1.241,
+      "step": 10335
+    },
+    {
+      "epoch": 1.6537600000000001,
+      "grad_norm": 2.187122472733626,
+      "learning_rate": 8.85467247908394e-05,
+      "loss": 1.3714,
+      "step": 10336
+    },
+    {
+      "epoch": 1.6539199999999998,
+      "grad_norm": 2.3681853778607382,
+      "learning_rate": 8.852956479620128e-05,
+      "loss": 1.309,
+      "step": 10337
+    },
+    {
+      "epoch": 1.65408,
+      "grad_norm": 2.04410943315252,
+      "learning_rate": 8.851240514382426e-05,
+      "loss": 1.3847,
+      "step": 10338
+    },
+    {
+      "epoch": 1.6542400000000002,
+      "grad_norm": 2.298323042029432,
+      "learning_rate": 8.84952458342203e-05,
+      "loss": 1.3304,
+      "step": 10339
+    },
+    {
+      "epoch": 1.6543999999999999,
+      "grad_norm": 2.129111867263525,
+      "learning_rate": 8.847808686790149e-05,
+      "loss": 1.421,
+      "step": 10340
+    },
+    {
+      "epoch": 1.65456,
+      "grad_norm": 2.0366827549005015,
+      "learning_rate": 8.846092824537972e-05,
+      "loss": 1.2824,
+      "step": 10341
+    },
+    {
+      "epoch": 1.65472,
+      "grad_norm": 1.7591371263497542,
+      "learning_rate": 8.844376996716709e-05,
+      "loss": 1.3342,
+      "step": 10342
+    },
+    {
+      "epoch": 1.65488,
+      "grad_norm": 2.1119740893915235,
+      "learning_rate": 8.842661203377546e-05,
+      "loss": 1.2952,
+      "step": 10343
+    },
+    {
+      "epoch": 1.65504,
+      "grad_norm": 2.1241218385360896,
+      "learning_rate": 8.840945444571694e-05,
+      "loss": 1.3249,
+      "step": 10344
+    },
+    {
+      "epoch": 1.6552,
+      "grad_norm": 2.159646438812963,
+      "learning_rate": 8.839229720350333e-05,
+      "loss": 1.2701,
+      "step": 10345
+    },
+    {
+      "epoch": 1.65536,
+      "grad_norm": 2.374123993187256,
+      "learning_rate": 8.837514030764669e-05,
+      "loss": 1.329,
+      "step": 10346
+    },
+    {
+      "epoch": 1.65552,
+      "grad_norm": 2.169211159833914,
+      "learning_rate": 8.835798375865896e-05,
+      "loss": 1.2969,
+      "step": 10347
+    },
+    {
+      "epoch": 1.65568,
+      "grad_norm": 2.4824415948850915,
+      "learning_rate": 8.834082755705199e-05,
+      "loss": 1.2771,
+      "step": 10348
+    },
+    {
+      "epoch": 1.65584,
+      "grad_norm": 1.9676241743689051,
+      "learning_rate": 8.832367170333777e-05,
+      "loss": 1.2812,
+      "step": 10349
+    },
+    {
+      "epoch": 1.6560000000000001,
+      "grad_norm": 2.0346669267054875,
+      "learning_rate": 8.830651619802813e-05,
+      "loss": 1.2936,
+      "step": 10350
+    },
+    {
+      "epoch": 1.6561599999999999,
+      "grad_norm": 1.936620033961032,
+      "learning_rate": 8.828936104163503e-05,
+      "loss": 1.3295,
+      "step": 10351
+    },
+    {
+      "epoch": 1.65632,
+      "grad_norm": 1.9165741768409728,
+      "learning_rate": 8.827220623467034e-05,
+      "loss": 1.2024,
+      "step": 10352
+    },
+    {
+      "epoch": 1.65648,
+      "grad_norm": 2.085375682989734,
+      "learning_rate": 8.825505177764592e-05,
+      "loss": 1.3056,
+      "step": 10353
+    },
+    {
+      "epoch": 1.65664,
+      "grad_norm": 2.0342998037949775,
+      "learning_rate": 8.823789767107361e-05,
+      "loss": 1.2712,
+      "step": 10354
+    },
+    {
+      "epoch": 1.6568,
+      "grad_norm": 1.8313297882846182,
+      "learning_rate": 8.822074391546535e-05,
+      "loss": 1.2859,
+      "step": 10355
+    },
+    {
+      "epoch": 1.65696,
+      "grad_norm": 1.996572671491305,
+      "learning_rate": 8.820359051133289e-05,
+      "loss": 1.3547,
+      "step": 10356
+    },
+    {
+      "epoch": 1.65712,
+      "grad_norm": 2.2787500290776177,
+      "learning_rate": 8.818643745918814e-05,
+      "loss": 1.3108,
+      "step": 10357
+    },
+    {
+      "epoch": 1.65728,
+      "grad_norm": 1.901012054971844,
+      "learning_rate": 8.816928475954283e-05,
+      "loss": 1.2345,
+      "step": 10358
+    },
+    {
+      "epoch": 1.65744,
+      "grad_norm": 2.007363624645287,
+      "learning_rate": 8.815213241290884e-05,
+      "loss": 1.2885,
+      "step": 10359
+    },
+    {
+      "epoch": 1.6576,
+      "grad_norm": 2.687482828067898,
+      "learning_rate": 8.813498041979797e-05,
+      "loss": 1.3057,
+      "step": 10360
+    },
+    {
+      "epoch": 1.6577600000000001,
+      "grad_norm": 2.038562195850941,
+      "learning_rate": 8.811782878072199e-05,
+      "loss": 1.3151,
+      "step": 10361
+    },
+    {
+      "epoch": 1.6579199999999998,
+      "grad_norm": 2.1084346486671937,
+      "learning_rate": 8.810067749619266e-05,
+      "loss": 1.279,
+      "step": 10362
+    },
+    {
+      "epoch": 1.65808,
+      "grad_norm": 2.29730341533835,
+      "learning_rate": 8.808352656672182e-05,
+      "loss": 1.3673,
+      "step": 10363
+    },
+    {
+      "epoch": 1.6582400000000002,
+      "grad_norm": 2.1340252426885673,
+      "learning_rate": 8.806637599282114e-05,
+      "loss": 1.2662,
+      "step": 10364
+    },
+    {
+      "epoch": 1.6583999999999999,
+      "grad_norm": 2.233092273350457,
+      "learning_rate": 8.804922577500245e-05,
+      "loss": 1.2781,
+      "step": 10365
+    },
+    {
+      "epoch": 1.65856,
+      "grad_norm": 2.0780127818991088,
+      "learning_rate": 8.803207591377741e-05,
+      "loss": 1.2561,
+      "step": 10366
+    },
+    {
+      "epoch": 1.65872,
+      "grad_norm": 1.8703265610567763,
+      "learning_rate": 8.801492640965781e-05,
+      "loss": 1.3099,
+      "step": 10367
+    },
+    {
+      "epoch": 1.65888,
+      "grad_norm": 2.1361889416813624,
+      "learning_rate": 8.799777726315532e-05,
+      "loss": 1.27,
+      "step": 10368
+    },
+    {
+      "epoch": 1.65904,
+      "grad_norm": 2.3467192270787685,
+      "learning_rate": 8.79806284747817e-05,
+      "loss": 1.3166,
+      "step": 10369
+    },
+    {
+      "epoch": 1.6592,
+      "grad_norm": 1.8814872906734443,
+      "learning_rate": 8.796348004504858e-05,
+      "loss": 1.3799,
+      "step": 10370
+    },
+    {
+      "epoch": 1.65936,
+      "grad_norm": 2.240516852586248,
+      "learning_rate": 8.79463319744677e-05,
+      "loss": 1.2776,
+      "step": 10371
+    },
+    {
+      "epoch": 1.65952,
+      "grad_norm": 2.792778216086862,
+      "learning_rate": 8.79291842635507e-05,
+      "loss": 1.3536,
+      "step": 10372
+    },
+    {
+      "epoch": 1.65968,
+      "grad_norm": 2.0868797706717817,
+      "learning_rate": 8.791203691280928e-05,
+      "loss": 1.2767,
+      "step": 10373
+    },
+    {
+      "epoch": 1.65984,
+      "grad_norm": 1.943526510513798,
+      "learning_rate": 8.789488992275502e-05,
+      "loss": 1.2774,
+      "step": 10374
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 2.0277883378046786,
+      "learning_rate": 8.787774329389963e-05,
+      "loss": 1.2575,
+      "step": 10375
+    },
+    {
+      "epoch": 1.6601599999999999,
+      "grad_norm": 2.0541561761727474,
+      "learning_rate": 8.786059702675468e-05,
+      "loss": 1.294,
+      "step": 10376
+    },
+    {
+      "epoch": 1.66032,
+      "grad_norm": 2.0592087399780827,
+      "learning_rate": 8.784345112183189e-05,
+      "loss": 1.3215,
+      "step": 10377
+    },
+    {
+      "epoch": 1.66048,
+      "grad_norm": 2.9493625035741515,
+      "learning_rate": 8.782630557964275e-05,
+      "loss": 1.2562,
+      "step": 10378
+    },
+    {
+      "epoch": 1.66064,
+      "grad_norm": 2.049808291455188,
+      "learning_rate": 8.780916040069895e-05,
+      "loss": 1.232,
+      "step": 10379
+    },
+    {
+      "epoch": 1.6608,
+      "grad_norm": 2.3250765112122247,
+      "learning_rate": 8.779201558551201e-05,
+      "loss": 1.3568,
+      "step": 10380
+    },
+    {
+      "epoch": 1.66096,
+      "grad_norm": 1.915660138857673,
+      "learning_rate": 8.777487113459357e-05,
+      "loss": 1.2704,
+      "step": 10381
+    },
+    {
+      "epoch": 1.66112,
+      "grad_norm": 2.033558831331463,
+      "learning_rate": 8.775772704845513e-05,
+      "loss": 1.3408,
+      "step": 10382
+    },
+    {
+      "epoch": 1.66128,
+      "grad_norm": 2.136159405452753,
+      "learning_rate": 8.774058332760829e-05,
+      "loss": 1.3015,
+      "step": 10383
+    },
+    {
+      "epoch": 1.66144,
+      "grad_norm": 2.0418296603095563,
+      "learning_rate": 8.772343997256455e-05,
+      "loss": 1.2545,
+      "step": 10384
+    },
+    {
+      "epoch": 1.6616,
+      "grad_norm": 1.601516319399952,
+      "learning_rate": 8.770629698383552e-05,
+      "loss": 1.2434,
+      "step": 10385
+    },
+    {
+      "epoch": 1.6617600000000001,
+      "grad_norm": 2.272741209649695,
+      "learning_rate": 8.768915436193264e-05,
+      "loss": 1.3034,
+      "step": 10386
+    },
+    {
+      "epoch": 1.6619199999999998,
+      "grad_norm": 2.1480069694274975,
+      "learning_rate": 8.76720121073675e-05,
+      "loss": 1.3011,
+      "step": 10387
+    },
+    {
+      "epoch": 1.66208,
+      "grad_norm": 1.9554401716152476,
+      "learning_rate": 8.765487022065149e-05,
+      "loss": 1.2704,
+      "step": 10388
+    },
+    {
+      "epoch": 1.6622400000000002,
+      "grad_norm": 2.0974902195057163,
+      "learning_rate": 8.763772870229623e-05,
+      "loss": 1.3259,
+      "step": 10389
+    },
+    {
+      "epoch": 1.6623999999999999,
+      "grad_norm": 2.2271806526125864,
+      "learning_rate": 8.762058755281309e-05,
+      "loss": 1.3321,
+      "step": 10390
+    },
+    {
+      "epoch": 1.66256,
+      "grad_norm": 2.149718789942447,
+      "learning_rate": 8.760344677271359e-05,
+      "loss": 1.3397,
+      "step": 10391
+    },
+    {
+      "epoch": 1.66272,
+      "grad_norm": 1.8525868804867858,
+      "learning_rate": 8.758630636250914e-05,
+      "loss": 1.2205,
+      "step": 10392
+    },
+    {
+      "epoch": 1.66288,
+      "grad_norm": 1.7868825135957216,
+      "learning_rate": 8.756916632271129e-05,
+      "loss": 1.2246,
+      "step": 10393
+    },
+    {
+      "epoch": 1.66304,
+      "grad_norm": 2.3102616289632705,
+      "learning_rate": 8.755202665383133e-05,
+      "loss": 1.3055,
+      "step": 10394
+    },
+    {
+      "epoch": 1.6632,
+      "grad_norm": 2.018074163636113,
+      "learning_rate": 8.753488735638082e-05,
+      "loss": 1.2413,
+      "step": 10395
+    },
+    {
+      "epoch": 1.66336,
+      "grad_norm": 1.9806526601450667,
+      "learning_rate": 8.751774843087107e-05,
+      "loss": 1.2528,
+      "step": 10396
+    },
+    {
+      "epoch": 1.66352,
+      "grad_norm": 1.782559278169615,
+      "learning_rate": 8.750060987781355e-05,
+      "loss": 1.3047,
+      "step": 10397
+    },
+    {
+      "epoch": 1.66368,
+      "grad_norm": 1.985960454973673,
+      "learning_rate": 8.748347169771957e-05,
+      "loss": 1.2772,
+      "step": 10398
+    },
+    {
+      "epoch": 1.66384,
+      "grad_norm": 1.9360247461843645,
+      "learning_rate": 8.746633389110056e-05,
+      "loss": 1.2494,
+      "step": 10399
+    },
+    {
+      "epoch": 1.6640000000000001,
+      "grad_norm": 1.8269222777626362,
+      "learning_rate": 8.74491964584679e-05,
+      "loss": 1.2435,
+      "step": 10400
+    },
+    {
+      "epoch": 1.6641599999999999,
+      "grad_norm": 2.0008244752508726,
+      "learning_rate": 8.743205940033296e-05,
+      "loss": 1.3081,
+      "step": 10401
+    },
+    {
+      "epoch": 1.66432,
+      "grad_norm": 1.8978780750944566,
+      "learning_rate": 8.7414922717207e-05,
+      "loss": 1.2637,
+      "step": 10402
+    },
+    {
+      "epoch": 1.66448,
+      "grad_norm": 1.8968942002939,
+      "learning_rate": 8.739778640960142e-05,
+      "loss": 1.2499,
+      "step": 10403
+    },
+    {
+      "epoch": 1.66464,
+      "grad_norm": 2.0256448637200415,
+      "learning_rate": 8.738065047802757e-05,
+      "loss": 1.1938,
+      "step": 10404
+    },
+    {
+      "epoch": 1.6648,
+      "grad_norm": 2.076765565951669,
+      "learning_rate": 8.736351492299668e-05,
+      "loss": 1.2553,
+      "step": 10405
+    },
+    {
+      "epoch": 1.66496,
+      "grad_norm": 2.406206655832804,
+      "learning_rate": 8.734637974502014e-05,
+      "loss": 1.3158,
+      "step": 10406
+    },
+    {
+      "epoch": 1.66512,
+      "grad_norm": 1.9386402286616846,
+      "learning_rate": 8.732924494460913e-05,
+      "loss": 1.3135,
+      "step": 10407
+    },
+    {
+      "epoch": 1.66528,
+      "grad_norm": 2.359284063615304,
+      "learning_rate": 8.731211052227501e-05,
+      "loss": 1.302,
+      "step": 10408
+    },
+    {
+      "epoch": 1.66544,
+      "grad_norm": 1.9852148163430754,
+      "learning_rate": 8.729497647852904e-05,
+      "loss": 1.279,
+      "step": 10409
+    },
+    {
+      "epoch": 1.6656,
+      "grad_norm": 1.9583524264412315,
+      "learning_rate": 8.727784281388244e-05,
+      "loss": 1.2929,
+      "step": 10410
+    },
+    {
+      "epoch": 1.6657600000000001,
+      "grad_norm": 2.0032189013149813,
+      "learning_rate": 8.726070952884647e-05,
+      "loss": 1.3024,
+      "step": 10411
+    },
+    {
+      "epoch": 1.6659199999999998,
+      "grad_norm": 1.9303737193022108,
+      "learning_rate": 8.72435766239324e-05,
+      "loss": 1.2987,
+      "step": 10412
+    },
+    {
+      "epoch": 1.66608,
+      "grad_norm": 1.9112999241303024,
+      "learning_rate": 8.722644409965138e-05,
+      "loss": 1.2448,
+      "step": 10413
+    },
+    {
+      "epoch": 1.6662400000000002,
+      "grad_norm": 1.8931526352112742,
+      "learning_rate": 8.720931195651468e-05,
+      "loss": 1.2502,
+      "step": 10414
+    },
+    {
+      "epoch": 1.6663999999999999,
+      "grad_norm": 2.113472492228133,
+      "learning_rate": 8.719218019503343e-05,
+      "loss": 1.329,
+      "step": 10415
+    },
+    {
+      "epoch": 1.66656,
+      "grad_norm": 2.2624203686126427,
+      "learning_rate": 8.71750488157189e-05,
+      "loss": 1.2952,
+      "step": 10416
+    },
+    {
+      "epoch": 1.66672,
+      "grad_norm": 2.1515297882239484,
+      "learning_rate": 8.715791781908223e-05,
+      "loss": 1.2568,
+      "step": 10417
+    },
+    {
+      "epoch": 1.66688,
+      "grad_norm": 2.057814516385789,
+      "learning_rate": 8.714078720563455e-05,
+      "loss": 1.3041,
+      "step": 10418
+    },
+    {
+      "epoch": 1.66704,
+      "grad_norm": 2.229177840897874,
+      "learning_rate": 8.712365697588704e-05,
+      "loss": 1.314,
+      "step": 10419
+    },
+    {
+      "epoch": 1.6672,
+      "grad_norm": 1.9375598974310184,
+      "learning_rate": 8.710652713035087e-05,
+      "loss": 1.2825,
+      "step": 10420
+    },
+    {
+      "epoch": 1.66736,
+      "grad_norm": 2.1768551830797938,
+      "learning_rate": 8.708939766953712e-05,
+      "loss": 1.3727,
+      "step": 10421
+    },
+    {
+      "epoch": 1.6675200000000001,
+      "grad_norm": 2.153978594453145,
+      "learning_rate": 8.707226859395697e-05,
+      "loss": 1.306,
+      "step": 10422
+    },
+    {
+      "epoch": 1.66768,
+      "grad_norm": 2.2415683729209483,
+      "learning_rate": 8.705513990412144e-05,
+      "loss": 1.2708,
+      "step": 10423
+    },
+    {
+      "epoch": 1.66784,
+      "grad_norm": 1.9260401712811022,
+      "learning_rate": 8.70380116005417e-05,
+      "loss": 1.2833,
+      "step": 10424
+    },
+    {
+      "epoch": 1.6680000000000001,
+      "grad_norm": 2.074910862620396,
+      "learning_rate": 8.702088368372879e-05,
+      "loss": 1.2511,
+      "step": 10425
+    },
+    {
+      "epoch": 1.6681599999999999,
+      "grad_norm": 1.9601929993167628,
+      "learning_rate": 8.700375615419382e-05,
+      "loss": 1.311,
+      "step": 10426
+    },
+    {
+      "epoch": 1.66832,
+      "grad_norm": 1.8185550015860756,
+      "learning_rate": 8.698662901244778e-05,
+      "loss": 1.3064,
+      "step": 10427
+    },
+    {
+      "epoch": 1.66848,
+      "grad_norm": 1.9036227893770463,
+      "learning_rate": 8.696950225900185e-05,
+      "loss": 1.2482,
+      "step": 10428
+    },
+    {
+      "epoch": 1.66864,
+      "grad_norm": 2.0409545342942668,
+      "learning_rate": 8.69523758943669e-05,
+      "loss": 1.325,
+      "step": 10429
+    },
+    {
+      "epoch": 1.6688,
+      "grad_norm": 2.2021331395116905,
+      "learning_rate": 8.69352499190541e-05,
+      "loss": 1.2975,
+      "step": 10430
+    },
+    {
+      "epoch": 1.66896,
+      "grad_norm": 1.9021793534201703,
+      "learning_rate": 8.691812433357438e-05,
+      "loss": 1.255,
+      "step": 10431
+    },
+    {
+      "epoch": 1.66912,
+      "grad_norm": 1.8125239452371165,
+      "learning_rate": 8.690099913843877e-05,
+      "loss": 1.2153,
+      "step": 10432
+    },
+    {
+      "epoch": 1.66928,
+      "grad_norm": 1.7992349155142149,
+      "learning_rate": 8.688387433415827e-05,
+      "loss": 1.2962,
+      "step": 10433
+    },
+    {
+      "epoch": 1.66944,
+      "grad_norm": 1.6640461199928365,
+      "learning_rate": 8.686674992124384e-05,
+      "loss": 1.3557,
+      "step": 10434
+    },
+    {
+      "epoch": 1.6696,
+      "grad_norm": 2.0509678694773235,
+      "learning_rate": 8.684962590020643e-05,
+      "loss": 1.3369,
+      "step": 10435
+    },
+    {
+      "epoch": 1.6697600000000001,
+      "grad_norm": 2.4732618333746,
+      "learning_rate": 8.683250227155705e-05,
+      "loss": 1.3415,
+      "step": 10436
+    },
+    {
+      "epoch": 1.6699199999999998,
+      "grad_norm": 2.390196480316856,
+      "learning_rate": 8.681537903580661e-05,
+      "loss": 1.2899,
+      "step": 10437
+    },
+    {
+      "epoch": 1.67008,
+      "grad_norm": 1.8282707908412459,
+      "learning_rate": 8.679825619346606e-05,
+      "loss": 1.3628,
+      "step": 10438
+    },
+    {
+      "epoch": 1.67024,
+      "grad_norm": 1.9900531606163583,
+      "learning_rate": 8.678113374504628e-05,
+      "loss": 1.2352,
+      "step": 10439
+    },
+    {
+      "epoch": 1.6703999999999999,
+      "grad_norm": 2.088403373332559,
+      "learning_rate": 8.676401169105822e-05,
+      "loss": 1.2902,
+      "step": 10440
+    },
+    {
+      "epoch": 1.67056,
+      "grad_norm": 2.175352207467906,
+      "learning_rate": 8.674689003201277e-05,
+      "loss": 1.3274,
+      "step": 10441
+    },
+    {
+      "epoch": 1.67072,
+      "grad_norm": 1.9728227818447936,
+      "learning_rate": 8.672976876842082e-05,
+      "loss": 1.3159,
+      "step": 10442
+    },
+    {
+      "epoch": 1.67088,
+      "grad_norm": 2.0523437587135205,
+      "learning_rate": 8.671264790079319e-05,
+      "loss": 1.2975,
+      "step": 10443
+    },
+    {
+      "epoch": 1.67104,
+      "grad_norm": 2.1251326795224768,
+      "learning_rate": 8.669552742964084e-05,
+      "loss": 1.2712,
+      "step": 10444
+    },
+    {
+      "epoch": 1.6712,
+      "grad_norm": 2.0816408730258846,
+      "learning_rate": 8.667840735547452e-05,
+      "loss": 1.269,
+      "step": 10445
+    },
+    {
+      "epoch": 1.67136,
+      "grad_norm": 2.326873369944956,
+      "learning_rate": 8.666128767880516e-05,
+      "loss": 1.3326,
+      "step": 10446
+    },
+    {
+      "epoch": 1.6715200000000001,
+      "grad_norm": 2.0651106667438555,
+      "learning_rate": 8.664416840014349e-05,
+      "loss": 1.2785,
+      "step": 10447
+    },
+    {
+      "epoch": 1.67168,
+      "grad_norm": 2.3471732552246944,
+      "learning_rate": 8.66270495200004e-05,
+      "loss": 1.2827,
+      "step": 10448
+    },
+    {
+      "epoch": 1.67184,
+      "grad_norm": 2.028489693247594,
+      "learning_rate": 8.660993103888666e-05,
+      "loss": 1.2778,
+      "step": 10449
+    },
+    {
+      "epoch": 1.6720000000000002,
+      "grad_norm": 2.207429815227544,
+      "learning_rate": 8.659281295731307e-05,
+      "loss": 1.2883,
+      "step": 10450
+    },
+    {
+      "epoch": 1.6721599999999999,
+      "grad_norm": 2.1053660087833914,
+      "learning_rate": 8.657569527579038e-05,
+      "loss": 1.2748,
+      "step": 10451
+    },
+    {
+      "epoch": 1.67232,
+      "grad_norm": 1.996214493679529,
+      "learning_rate": 8.655857799482943e-05,
+      "loss": 1.3265,
+      "step": 10452
+    },
+    {
+      "epoch": 1.67248,
+      "grad_norm": 2.530786307593522,
+      "learning_rate": 8.654146111494088e-05,
+      "loss": 1.2924,
+      "step": 10453
+    },
+    {
+      "epoch": 1.67264,
+      "grad_norm": 1.9561299452558774,
+      "learning_rate": 8.652434463663557e-05,
+      "loss": 1.2645,
+      "step": 10454
+    },
+    {
+      "epoch": 1.6728,
+      "grad_norm": 2.070257861163132,
+      "learning_rate": 8.650722856042413e-05,
+      "loss": 1.2656,
+      "step": 10455
+    },
+    {
+      "epoch": 1.67296,
+      "grad_norm": 2.1985735120595287,
+      "learning_rate": 8.649011288681735e-05,
+      "loss": 1.2921,
+      "step": 10456
+    },
+    {
+      "epoch": 1.67312,
+      "grad_norm": 2.0416194633592397,
+      "learning_rate": 8.64729976163259e-05,
+      "loss": 1.3047,
+      "step": 10457
+    },
+    {
+      "epoch": 1.67328,
+      "grad_norm": 2.155849822955977,
+      "learning_rate": 8.645588274946051e-05,
+      "loss": 1.3187,
+      "step": 10458
+    },
+    {
+      "epoch": 1.67344,
+      "grad_norm": 1.8292109951999407,
+      "learning_rate": 8.643876828673182e-05,
+      "loss": 1.2721,
+      "step": 10459
+    },
+    {
+      "epoch": 1.6736,
+      "grad_norm": 2.1451879689706663,
+      "learning_rate": 8.642165422865056e-05,
+      "loss": 1.251,
+      "step": 10460
+    },
+    {
+      "epoch": 1.6737600000000001,
+      "grad_norm": 1.8635540798852646,
+      "learning_rate": 8.640454057572731e-05,
+      "loss": 1.2962,
+      "step": 10461
+    },
+    {
+      "epoch": 1.6739199999999999,
+      "grad_norm": 2.063702833274743,
+      "learning_rate": 8.638742732847278e-05,
+      "loss": 1.2801,
+      "step": 10462
+    },
+    {
+      "epoch": 1.67408,
+      "grad_norm": 2.216283515101553,
+      "learning_rate": 8.637031448739759e-05,
+      "loss": 1.3418,
+      "step": 10463
+    },
+    {
+      "epoch": 1.67424,
+      "grad_norm": 2.111604701122306,
+      "learning_rate": 8.635320205301233e-05,
+      "loss": 1.3181,
+      "step": 10464
+    },
+    {
+      "epoch": 1.6743999999999999,
+      "grad_norm": 2.0222753595010596,
+      "learning_rate": 8.633609002582765e-05,
+      "loss": 1.2566,
+      "step": 10465
+    },
+    {
+      "epoch": 1.67456,
+      "grad_norm": 2.0204133637943897,
+      "learning_rate": 8.631897840635414e-05,
+      "loss": 1.2328,
+      "step": 10466
+    },
+    {
+      "epoch": 1.67472,
+      "grad_norm": 1.9698637973954074,
+      "learning_rate": 8.630186719510238e-05,
+      "loss": 1.3134,
+      "step": 10467
+    },
+    {
+      "epoch": 1.67488,
+      "grad_norm": 2.0130490794484515,
+      "learning_rate": 8.628475639258293e-05,
+      "loss": 1.3557,
+      "step": 10468
+    },
+    {
+      "epoch": 1.67504,
+      "grad_norm": 1.869783272513802,
+      "learning_rate": 8.62676459993064e-05,
+      "loss": 1.2862,
+      "step": 10469
+    },
+    {
+      "epoch": 1.6752,
+      "grad_norm": 2.3061258732737753,
+      "learning_rate": 8.625053601578325e-05,
+      "loss": 1.3132,
+      "step": 10470
+    },
+    {
+      "epoch": 1.67536,
+      "grad_norm": 2.438330158379585,
+      "learning_rate": 8.623342644252413e-05,
+      "loss": 1.2894,
+      "step": 10471
+    },
+    {
+      "epoch": 1.6755200000000001,
+      "grad_norm": 1.9498684230065544,
+      "learning_rate": 8.621631728003944e-05,
+      "loss": 1.2768,
+      "step": 10472
+    },
+    {
+      "epoch": 1.6756799999999998,
+      "grad_norm": 2.370257554222099,
+      "learning_rate": 8.619920852883979e-05,
+      "loss": 1.2565,
+      "step": 10473
+    },
+    {
+      "epoch": 1.67584,
+      "grad_norm": 1.9662204098755138,
+      "learning_rate": 8.618210018943566e-05,
+      "loss": 1.2639,
+      "step": 10474
+    },
+    {
+      "epoch": 1.6760000000000002,
+      "grad_norm": 1.9529174760450192,
+      "learning_rate": 8.61649922623375e-05,
+      "loss": 1.2313,
+      "step": 10475
+    },
+    {
+      "epoch": 1.6761599999999999,
+      "grad_norm": 2.1570007318786577,
+      "learning_rate": 8.614788474805581e-05,
+      "loss": 1.2985,
+      "step": 10476
+    },
+    {
+      "epoch": 1.67632,
+      "grad_norm": 1.8514134372756408,
+      "learning_rate": 8.61307776471011e-05,
+      "loss": 1.2506,
+      "step": 10477
+    },
+    {
+      "epoch": 1.67648,
+      "grad_norm": 1.944134089454685,
+      "learning_rate": 8.611367095998372e-05,
+      "loss": 1.257,
+      "step": 10478
+    },
+    {
+      "epoch": 1.67664,
+      "grad_norm": 2.3429422477354436,
+      "learning_rate": 8.609656468721421e-05,
+      "loss": 1.2856,
+      "step": 10479
+    },
+    {
+      "epoch": 1.6768,
+      "grad_norm": 2.14043291216433,
+      "learning_rate": 8.607945882930291e-05,
+      "loss": 1.2956,
+      "step": 10480
+    },
+    {
+      "epoch": 1.67696,
+      "grad_norm": 1.8460167369091658,
+      "learning_rate": 8.606235338676029e-05,
+      "loss": 1.298,
+      "step": 10481
+    },
+    {
+      "epoch": 1.67712,
+      "grad_norm": 2.2216786759565483,
+      "learning_rate": 8.604524836009674e-05,
+      "loss": 1.275,
+      "step": 10482
+    },
+    {
+      "epoch": 1.67728,
+      "grad_norm": 2.0592044022534064,
+      "learning_rate": 8.602814374982264e-05,
+      "loss": 1.2928,
+      "step": 10483
+    },
+    {
+      "epoch": 1.67744,
+      "grad_norm": 2.1627130874352742,
+      "learning_rate": 8.601103955644834e-05,
+      "loss": 1.2922,
+      "step": 10484
+    },
+    {
+      "epoch": 1.6776,
+      "grad_norm": 1.8230715252531033,
+      "learning_rate": 8.59939357804843e-05,
+      "loss": 1.2385,
+      "step": 10485
+    },
+    {
+      "epoch": 1.6777600000000001,
+      "grad_norm": 2.1738703561005925,
+      "learning_rate": 8.597683242244076e-05,
+      "loss": 1.302,
+      "step": 10486
+    },
+    {
+      "epoch": 1.6779199999999999,
+      "grad_norm": 2.577136871270649,
+      "learning_rate": 8.595972948282814e-05,
+      "loss": 1.2292,
+      "step": 10487
+    },
+    {
+      "epoch": 1.67808,
+      "grad_norm": 1.7306761245862279,
+      "learning_rate": 8.594262696215671e-05,
+      "loss": 1.324,
+      "step": 10488
+    },
+    {
+      "epoch": 1.67824,
+      "grad_norm": 2.1632755603836866,
+      "learning_rate": 8.592552486093682e-05,
+      "loss": 1.3122,
+      "step": 10489
+    },
+    {
+      "epoch": 1.6784,
+      "grad_norm": 2.371054807029723,
+      "learning_rate": 8.590842317967875e-05,
+      "loss": 1.2703,
+      "step": 10490
+    },
+    {
+      "epoch": 1.67856,
+      "grad_norm": 2.0367257603892592,
+      "learning_rate": 8.58913219188928e-05,
+      "loss": 1.2975,
+      "step": 10491
+    },
+    {
+      "epoch": 1.67872,
+      "grad_norm": 2.870281161957289,
+      "learning_rate": 8.587422107908922e-05,
+      "loss": 1.3001,
+      "step": 10492
+    },
+    {
+      "epoch": 1.67888,
+      "grad_norm": 2.1489737989995175,
+      "learning_rate": 8.585712066077835e-05,
+      "loss": 1.2647,
+      "step": 10493
+    },
+    {
+      "epoch": 1.67904,
+      "grad_norm": 1.8274739357780962,
+      "learning_rate": 8.584002066447036e-05,
+      "loss": 1.2457,
+      "step": 10494
+    },
+    {
+      "epoch": 1.6792,
+      "grad_norm": 2.240390940355793,
+      "learning_rate": 8.582292109067554e-05,
+      "loss": 1.2621,
+      "step": 10495
+    },
+    {
+      "epoch": 1.67936,
+      "grad_norm": 1.8878394508937382,
+      "learning_rate": 8.580582193990407e-05,
+      "loss": 1.329,
+      "step": 10496
+    },
+    {
+      "epoch": 1.6795200000000001,
+      "grad_norm": 2.2015761110127516,
+      "learning_rate": 8.57887232126662e-05,
+      "loss": 1.2819,
+      "step": 10497
+    },
+    {
+      "epoch": 1.6796799999999998,
+      "grad_norm": 2.146648156959622,
+      "learning_rate": 8.577162490947214e-05,
+      "loss": 1.3803,
+      "step": 10498
+    },
+    {
+      "epoch": 1.67984,
+      "grad_norm": 2.0025521115758367,
+      "learning_rate": 8.575452703083204e-05,
+      "loss": 1.3507,
+      "step": 10499
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 1.8615161057662324,
+      "learning_rate": 8.573742957725607e-05,
+      "loss": 1.3155,
+      "step": 10500
+    },
+    {
+      "epoch": 1.6801599999999999,
+      "grad_norm": 2.1971407926120223,
+      "learning_rate": 8.572033254925449e-05,
+      "loss": 1.2376,
+      "step": 10501
+    },
+    {
+      "epoch": 1.68032,
+      "grad_norm": 1.9632951361010802,
+      "learning_rate": 8.570323594733731e-05,
+      "loss": 1.3111,
+      "step": 10502
+    },
+    {
+      "epoch": 1.68048,
+      "grad_norm": 2.4701866513579662,
+      "learning_rate": 8.56861397720148e-05,
+      "loss": 1.3056,
+      "step": 10503
+    },
+    {
+      "epoch": 1.68064,
+      "grad_norm": 2.0686081434811756,
+      "learning_rate": 8.566904402379696e-05,
+      "loss": 1.2999,
+      "step": 10504
+    },
+    {
+      "epoch": 1.6808,
+      "grad_norm": 1.840442217685499,
+      "learning_rate": 8.565194870319399e-05,
+      "loss": 1.2792,
+      "step": 10505
+    },
+    {
+      "epoch": 1.68096,
+      "grad_norm": 2.239079549928991,
+      "learning_rate": 8.563485381071598e-05,
+      "loss": 1.2773,
+      "step": 10506
+    },
+    {
+      "epoch": 1.68112,
+      "grad_norm": 1.8213806369058858,
+      "learning_rate": 8.561775934687298e-05,
+      "loss": 1.2159,
+      "step": 10507
+    },
+    {
+      "epoch": 1.68128,
+      "grad_norm": 2.0134752954288158,
+      "learning_rate": 8.560066531217508e-05,
+      "loss": 1.2963,
+      "step": 10508
+    },
+    {
+      "epoch": 1.68144,
+      "grad_norm": 2.419253725796109,
+      "learning_rate": 8.558357170713237e-05,
+      "loss": 1.2535,
+      "step": 10509
+    },
+    {
+      "epoch": 1.6816,
+      "grad_norm": 2.0433130590989173,
+      "learning_rate": 8.556647853225484e-05,
+      "loss": 1.3035,
+      "step": 10510
+    },
+    {
+      "epoch": 1.6817600000000001,
+      "grad_norm": 2.0689680844572256,
+      "learning_rate": 8.554938578805259e-05,
+      "loss": 1.2604,
+      "step": 10511
+    },
+    {
+      "epoch": 1.6819199999999999,
+      "grad_norm": 1.8074406562789695,
+      "learning_rate": 8.553229347503558e-05,
+      "loss": 1.282,
+      "step": 10512
+    },
+    {
+      "epoch": 1.68208,
+      "grad_norm": 2.05795462743957,
+      "learning_rate": 8.551520159371387e-05,
+      "loss": 1.3506,
+      "step": 10513
+    },
+    {
+      "epoch": 1.68224,
+      "grad_norm": 2.173755626455444,
+      "learning_rate": 8.549811014459745e-05,
+      "loss": 1.3021,
+      "step": 10514
+    },
+    {
+      "epoch": 1.6824,
+      "grad_norm": 1.6956745281975267,
+      "learning_rate": 8.548101912819628e-05,
+      "loss": 1.2771,
+      "step": 10515
+    },
+    {
+      "epoch": 1.68256,
+      "grad_norm": 1.806319773308401,
+      "learning_rate": 8.546392854502031e-05,
+      "loss": 1.3049,
+      "step": 10516
+    },
+    {
+      "epoch": 1.68272,
+      "grad_norm": 2.031977573268048,
+      "learning_rate": 8.544683839557959e-05,
+      "loss": 1.2962,
+      "step": 10517
+    },
+    {
+      "epoch": 1.68288,
+      "grad_norm": 1.9861597393420394,
+      "learning_rate": 8.542974868038396e-05,
+      "loss": 1.2839,
+      "step": 10518
+    },
+    {
+      "epoch": 1.68304,
+      "grad_norm": 2.2937124405511007,
+      "learning_rate": 8.541265939994344e-05,
+      "loss": 1.2251,
+      "step": 10519
+    },
+    {
+      "epoch": 1.6832,
+      "grad_norm": 2.2144315671567476,
+      "learning_rate": 8.539557055476785e-05,
+      "loss": 1.4146,
+      "step": 10520
+    },
+    {
+      "epoch": 1.68336,
+      "grad_norm": 2.263590434677726,
+      "learning_rate": 8.537848214536718e-05,
+      "loss": 1.3345,
+      "step": 10521
+    },
+    {
+      "epoch": 1.6835200000000001,
+      "grad_norm": 2.2222815358780577,
+      "learning_rate": 8.536139417225133e-05,
+      "loss": 1.3276,
+      "step": 10522
+    },
+    {
+      "epoch": 1.6836799999999998,
+      "grad_norm": 1.8672931569205935,
+      "learning_rate": 8.53443066359301e-05,
+      "loss": 1.2834,
+      "step": 10523
+    },
+    {
+      "epoch": 1.68384,
+      "grad_norm": 2.0682248219814867,
+      "learning_rate": 8.532721953691343e-05,
+      "loss": 1.228,
+      "step": 10524
+    },
+    {
+      "epoch": 1.6840000000000002,
+      "grad_norm": 1.9143351652144773,
+      "learning_rate": 8.531013287571112e-05,
+      "loss": 1.2969,
+      "step": 10525
+    },
+    {
+      "epoch": 1.6841599999999999,
+      "grad_norm": 1.7841138176788742,
+      "learning_rate": 8.529304665283308e-05,
+      "loss": 1.2777,
+      "step": 10526
+    },
+    {
+      "epoch": 1.68432,
+      "grad_norm": 2.009443583721984,
+      "learning_rate": 8.527596086878906e-05,
+      "loss": 1.2268,
+      "step": 10527
+    },
+    {
+      "epoch": 1.68448,
+      "grad_norm": 2.0134965258496376,
+      "learning_rate": 8.525887552408897e-05,
+      "loss": 1.3357,
+      "step": 10528
+    },
+    {
+      "epoch": 1.68464,
+      "grad_norm": 2.06750168546443,
+      "learning_rate": 8.524179061924248e-05,
+      "loss": 1.318,
+      "step": 10529
+    },
+    {
+      "epoch": 1.6848,
+      "grad_norm": 1.9604707283895233,
+      "learning_rate": 8.522470615475953e-05,
+      "loss": 1.2632,
+      "step": 10530
+    },
+    {
+      "epoch": 1.68496,
+      "grad_norm": 2.11188556293993,
+      "learning_rate": 8.520762213114976e-05,
+      "loss": 1.2816,
+      "step": 10531
+    },
+    {
+      "epoch": 1.68512,
+      "grad_norm": 2.0645262174120314,
+      "learning_rate": 8.519053854892302e-05,
+      "loss": 1.3286,
+      "step": 10532
+    },
+    {
+      "epoch": 1.6852800000000001,
+      "grad_norm": 2.0608979096910476,
+      "learning_rate": 8.517345540858899e-05,
+      "loss": 1.2402,
+      "step": 10533
+    },
+    {
+      "epoch": 1.68544,
+      "grad_norm": 2.1812744372986046,
+      "learning_rate": 8.515637271065751e-05,
+      "loss": 1.2558,
+      "step": 10534
+    },
+    {
+      "epoch": 1.6856,
+      "grad_norm": 2.091416445399735,
+      "learning_rate": 8.513929045563818e-05,
+      "loss": 1.3123,
+      "step": 10535
+    },
+    {
+      "epoch": 1.6857600000000001,
+      "grad_norm": 2.290336828543859,
+      "learning_rate": 8.512220864404081e-05,
+      "loss": 1.2808,
+      "step": 10536
+    },
+    {
+      "epoch": 1.6859199999999999,
+      "grad_norm": 2.1496798506377335,
+      "learning_rate": 8.510512727637503e-05,
+      "loss": 1.3457,
+      "step": 10537
+    },
+    {
+      "epoch": 1.68608,
+      "grad_norm": 1.9829181799715119,
+      "learning_rate": 8.508804635315058e-05,
+      "loss": 1.3098,
+      "step": 10538
+    },
+    {
+      "epoch": 1.68624,
+      "grad_norm": 1.6174601388615912,
+      "learning_rate": 8.507096587487705e-05,
+      "loss": 1.2437,
+      "step": 10539
+    },
+    {
+      "epoch": 1.6864,
+      "grad_norm": 1.8944228433241865,
+      "learning_rate": 8.505388584206417e-05,
+      "loss": 1.2689,
+      "step": 10540
+    },
+    {
+      "epoch": 1.68656,
+      "grad_norm": 1.8606503298978987,
+      "learning_rate": 8.503680625522154e-05,
+      "loss": 1.3068,
+      "step": 10541
+    },
+    {
+      "epoch": 1.68672,
+      "grad_norm": 1.9508483676805792,
+      "learning_rate": 8.501972711485885e-05,
+      "loss": 1.2441,
+      "step": 10542
+    },
+    {
+      "epoch": 1.68688,
+      "grad_norm": 2.076222249481502,
+      "learning_rate": 8.500264842148562e-05,
+      "loss": 1.3605,
+      "step": 10543
+    },
+    {
+      "epoch": 1.68704,
+      "grad_norm": 2.345969552383731,
+      "learning_rate": 8.498557017561157e-05,
+      "loss": 1.2896,
+      "step": 10544
+    },
+    {
+      "epoch": 1.6872,
+      "grad_norm": 2.1805216856984058,
+      "learning_rate": 8.496849237774616e-05,
+      "loss": 1.2618,
+      "step": 10545
+    },
+    {
+      "epoch": 1.68736,
+      "grad_norm": 2.0961990429822763,
+      "learning_rate": 8.49514150283991e-05,
+      "loss": 1.2549,
+      "step": 10546
+    },
+    {
+      "epoch": 1.6875200000000001,
+      "grad_norm": 2.5776712279937226,
+      "learning_rate": 8.493433812807984e-05,
+      "loss": 1.3073,
+      "step": 10547
+    },
+    {
+      "epoch": 1.6876799999999998,
+      "grad_norm": 2.0386617819031794,
+      "learning_rate": 8.4917261677298e-05,
+      "loss": 1.2336,
+      "step": 10548
+    },
+    {
+      "epoch": 1.68784,
+      "grad_norm": 2.2809655468721513,
+      "learning_rate": 8.490018567656306e-05,
+      "loss": 1.2911,
+      "step": 10549
+    },
+    {
+      "epoch": 1.688,
+      "grad_norm": 2.0785752929669408,
+      "learning_rate": 8.488311012638462e-05,
+      "loss": 1.314,
+      "step": 10550
+    },
+    {
+      "epoch": 1.6881599999999999,
+      "grad_norm": 2.228013739519785,
+      "learning_rate": 8.486603502727212e-05,
+      "loss": 1.2664,
+      "step": 10551
+    },
+    {
+      "epoch": 1.68832,
+      "grad_norm": 1.8973400524617112,
+      "learning_rate": 8.484896037973511e-05,
+      "loss": 1.2619,
+      "step": 10552
+    },
+    {
+      "epoch": 1.68848,
+      "grad_norm": 1.8914813191007152,
+      "learning_rate": 8.483188618428301e-05,
+      "loss": 1.2581,
+      "step": 10553
+    },
+    {
+      "epoch": 1.68864,
+      "grad_norm": 2.5565300644969438,
+      "learning_rate": 8.481481244142537e-05,
+      "loss": 1.2747,
+      "step": 10554
+    },
+    {
+      "epoch": 1.6888,
+      "grad_norm": 2.413123303702621,
+      "learning_rate": 8.479773915167155e-05,
+      "loss": 1.2528,
+      "step": 10555
+    },
+    {
+      "epoch": 1.68896,
+      "grad_norm": 1.8332462636389049,
+      "learning_rate": 8.478066631553106e-05,
+      "loss": 1.3357,
+      "step": 10556
+    },
+    {
+      "epoch": 1.68912,
+      "grad_norm": 1.8567674669866814,
+      "learning_rate": 8.476359393351328e-05,
+      "loss": 1.2898,
+      "step": 10557
+    },
+    {
+      "epoch": 1.6892800000000001,
+      "grad_norm": 1.9981411468149441,
+      "learning_rate": 8.474652200612771e-05,
+      "loss": 1.2653,
+      "step": 10558
+    },
+    {
+      "epoch": 1.68944,
+      "grad_norm": 2.1102932585914407,
+      "learning_rate": 8.472945053388366e-05,
+      "loss": 1.2976,
+      "step": 10559
+    },
+    {
+      "epoch": 1.6896,
+      "grad_norm": 1.999305779875528,
+      "learning_rate": 8.47123795172906e-05,
+      "loss": 1.2635,
+      "step": 10560
+    },
+    {
+      "epoch": 1.6897600000000002,
+      "grad_norm": 2.3007190578223504,
+      "learning_rate": 8.469530895685779e-05,
+      "loss": 1.272,
+      "step": 10561
+    },
+    {
+      "epoch": 1.6899199999999999,
+      "grad_norm": 1.9156069486494944,
+      "learning_rate": 8.467823885309471e-05,
+      "loss": 1.2991,
+      "step": 10562
+    },
+    {
+      "epoch": 1.69008,
+      "grad_norm": 1.982223606280667,
+      "learning_rate": 8.466116920651065e-05,
+      "loss": 1.2749,
+      "step": 10563
+    },
+    {
+      "epoch": 1.69024,
+      "grad_norm": 2.1086468875426245,
+      "learning_rate": 8.464410001761493e-05,
+      "loss": 1.4219,
+      "step": 10564
+    },
+    {
+      "epoch": 1.6904,
+      "grad_norm": 1.9496648038351,
+      "learning_rate": 8.462703128691689e-05,
+      "loss": 1.2779,
+      "step": 10565
+    },
+    {
+      "epoch": 1.69056,
+      "grad_norm": 2.205865330007897,
+      "learning_rate": 8.460996301492588e-05,
+      "loss": 1.3167,
+      "step": 10566
+    },
+    {
+      "epoch": 1.69072,
+      "grad_norm": 1.9737402908897728,
+      "learning_rate": 8.459289520215109e-05,
+      "loss": 1.2028,
+      "step": 10567
+    },
+    {
+      "epoch": 1.69088,
+      "grad_norm": 1.8115579396012202,
+      "learning_rate": 8.45758278491019e-05,
+      "loss": 1.2891,
+      "step": 10568
+    },
+    {
+      "epoch": 1.69104,
+      "grad_norm": 2.0417502813031656,
+      "learning_rate": 8.45587609562875e-05,
+      "loss": 1.3296,
+      "step": 10569
+    },
+    {
+      "epoch": 1.6912,
+      "grad_norm": 1.906358225504338,
+      "learning_rate": 8.454169452421719e-05,
+      "loss": 1.2874,
+      "step": 10570
+    },
+    {
+      "epoch": 1.69136,
+      "grad_norm": 1.8680422037074411,
+      "learning_rate": 8.45246285534002e-05,
+      "loss": 1.2651,
+      "step": 10571
+    },
+    {
+      "epoch": 1.6915200000000001,
+      "grad_norm": 2.058531160799662,
+      "learning_rate": 8.450756304434572e-05,
+      "loss": 1.2454,
+      "step": 10572
+    },
+    {
+      "epoch": 1.6916799999999999,
+      "grad_norm": 2.038796730653941,
+      "learning_rate": 8.449049799756297e-05,
+      "loss": 1.2851,
+      "step": 10573
+    },
+    {
+      "epoch": 1.69184,
+      "grad_norm": 2.2224733801670773,
+      "learning_rate": 8.447343341356121e-05,
+      "loss": 1.3766,
+      "step": 10574
+    },
+    {
+      "epoch": 1.692,
+      "grad_norm": 2.2700309633510543,
+      "learning_rate": 8.445636929284953e-05,
+      "loss": 1.291,
+      "step": 10575
+    },
+    {
+      "epoch": 1.6921599999999999,
+      "grad_norm": 2.129428568025176,
+      "learning_rate": 8.44393056359372e-05,
+      "loss": 1.2984,
+      "step": 10576
+    },
+    {
+      "epoch": 1.69232,
+      "grad_norm": 2.5725742474034896,
+      "learning_rate": 8.442224244333324e-05,
+      "loss": 1.2911,
+      "step": 10577
+    },
+    {
+      "epoch": 1.69248,
+      "grad_norm": 2.1030998376496006,
+      "learning_rate": 8.440517971554689e-05,
+      "loss": 1.2821,
+      "step": 10578
+    },
+    {
+      "epoch": 1.69264,
+      "grad_norm": 2.2133836374901823,
+      "learning_rate": 8.438811745308727e-05,
+      "loss": 1.2703,
+      "step": 10579
+    },
+    {
+      "epoch": 1.6928,
+      "grad_norm": 2.2392888457954823,
+      "learning_rate": 8.437105565646346e-05,
+      "loss": 1.289,
+      "step": 10580
+    },
+    {
+      "epoch": 1.69296,
+      "grad_norm": 2.337426868448619,
+      "learning_rate": 8.435399432618458e-05,
+      "loss": 1.2711,
+      "step": 10581
+    },
+    {
+      "epoch": 1.69312,
+      "grad_norm": 2.192982388942916,
+      "learning_rate": 8.43369334627597e-05,
+      "loss": 1.2964,
+      "step": 10582
+    },
+    {
+      "epoch": 1.6932800000000001,
+      "grad_norm": 3.129368400832922,
+      "learning_rate": 8.431987306669795e-05,
+      "loss": 1.2685,
+      "step": 10583
+    },
+    {
+      "epoch": 1.6934399999999998,
+      "grad_norm": 2.402440594722727,
+      "learning_rate": 8.430281313850829e-05,
+      "loss": 1.3373,
+      "step": 10584
+    },
+    {
+      "epoch": 1.6936,
+      "grad_norm": 1.849186360242952,
+      "learning_rate": 8.428575367869984e-05,
+      "loss": 1.3026,
+      "step": 10585
+    },
+    {
+      "epoch": 1.6937600000000002,
+      "grad_norm": 2.644446875671077,
+      "learning_rate": 8.426869468778158e-05,
+      "loss": 1.2973,
+      "step": 10586
+    },
+    {
+      "epoch": 1.6939199999999999,
+      "grad_norm": 2.5827794390611905,
+      "learning_rate": 8.425163616626258e-05,
+      "loss": 1.3693,
+      "step": 10587
+    },
+    {
+      "epoch": 1.69408,
+      "grad_norm": 2.0440000377022147,
+      "learning_rate": 8.423457811465178e-05,
+      "loss": 1.3319,
+      "step": 10588
+    },
+    {
+      "epoch": 1.69424,
+      "grad_norm": 2.6360319734271753,
+      "learning_rate": 8.42175205334582e-05,
+      "loss": 1.2961,
+      "step": 10589
+    },
+    {
+      "epoch": 1.6944,
+      "grad_norm": 2.3940308163374517,
+      "learning_rate": 8.420046342319079e-05,
+      "loss": 1.2584,
+      "step": 10590
+    },
+    {
+      "epoch": 1.69456,
+      "grad_norm": 2.0162238594424458,
+      "learning_rate": 8.418340678435859e-05,
+      "loss": 1.337,
+      "step": 10591
+    },
+    {
+      "epoch": 1.69472,
+      "grad_norm": 2.380086308297931,
+      "learning_rate": 8.416635061747042e-05,
+      "loss": 1.2923,
+      "step": 10592
+    },
+    {
+      "epoch": 1.69488,
+      "grad_norm": 2.2832765405024507,
+      "learning_rate": 8.414929492303533e-05,
+      "loss": 1.287,
+      "step": 10593
+    },
+    {
+      "epoch": 1.69504,
+      "grad_norm": 2.0345038692867323,
+      "learning_rate": 8.413223970156213e-05,
+      "loss": 1.3413,
+      "step": 10594
+    },
+    {
+      "epoch": 1.6952,
+      "grad_norm": 2.3952233099134586,
+      "learning_rate": 8.411518495355984e-05,
+      "loss": 1.2668,
+      "step": 10595
+    },
+    {
+      "epoch": 1.69536,
+      "grad_norm": 2.118510008571881,
+      "learning_rate": 8.409813067953721e-05,
+      "loss": 1.3601,
+      "step": 10596
+    },
+    {
+      "epoch": 1.6955200000000001,
+      "grad_norm": 2.507193860664787,
+      "learning_rate": 8.408107688000324e-05,
+      "loss": 1.3008,
+      "step": 10597
+    },
+    {
+      "epoch": 1.6956799999999999,
+      "grad_norm": 2.856270339367003,
+      "learning_rate": 8.40640235554667e-05,
+      "loss": 1.2927,
+      "step": 10598
+    },
+    {
+      "epoch": 1.69584,
+      "grad_norm": 2.094083305523969,
+      "learning_rate": 8.404697070643652e-05,
+      "loss": 1.3624,
+      "step": 10599
+    },
+    {
+      "epoch": 1.696,
+      "grad_norm": 2.2697777010555185,
+      "learning_rate": 8.402991833342145e-05,
+      "loss": 1.2262,
+      "step": 10600
+    },
+    {
+      "epoch": 1.69616,
+      "grad_norm": 2.5914095524824576,
+      "learning_rate": 8.401286643693039e-05,
+      "loss": 1.3186,
+      "step": 10601
+    },
+    {
+      "epoch": 1.69632,
+      "grad_norm": 2.0514542349192038,
+      "learning_rate": 8.399581501747206e-05,
+      "loss": 1.3219,
+      "step": 10602
+    },
+    {
+      "epoch": 1.69648,
+      "grad_norm": 1.7803142718370568,
+      "learning_rate": 8.397876407555532e-05,
+      "loss": 1.2954,
+      "step": 10603
+    },
+    {
+      "epoch": 1.69664,
+      "grad_norm": 2.238172779777928,
+      "learning_rate": 8.396171361168888e-05,
+      "loss": 1.2557,
+      "step": 10604
+    },
+    {
+      "epoch": 1.6968,
+      "grad_norm": 1.7670645482446128,
+      "learning_rate": 8.394466362638156e-05,
+      "loss": 1.2586,
+      "step": 10605
+    },
+    {
+      "epoch": 1.69696,
+      "grad_norm": 2.2226548658676313,
+      "learning_rate": 8.392761412014205e-05,
+      "loss": 1.244,
+      "step": 10606
+    },
+    {
+      "epoch": 1.69712,
+      "grad_norm": 1.9096252053142873,
+      "learning_rate": 8.391056509347916e-05,
+      "loss": 1.2712,
+      "step": 10607
+    },
+    {
+      "epoch": 1.6972800000000001,
+      "grad_norm": 2.0931027145334116,
+      "learning_rate": 8.389351654690154e-05,
+      "loss": 1.2329,
+      "step": 10608
+    },
+    {
+      "epoch": 1.6974399999999998,
+      "grad_norm": 2.300070352614812,
+      "learning_rate": 8.387646848091794e-05,
+      "loss": 1.3822,
+      "step": 10609
+    },
+    {
+      "epoch": 1.6976,
+      "grad_norm": 2.355592602712172,
+      "learning_rate": 8.385942089603699e-05,
+      "loss": 1.3487,
+      "step": 10610
+    },
+    {
+      "epoch": 1.6977600000000002,
+      "grad_norm": 2.218002531683446,
+      "learning_rate": 8.384237379276744e-05,
+      "loss": 1.3095,
+      "step": 10611
+    },
+    {
+      "epoch": 1.6979199999999999,
+      "grad_norm": 2.551330780681239,
+      "learning_rate": 8.382532717161786e-05,
+      "loss": 1.3091,
+      "step": 10612
+    },
+    {
+      "epoch": 1.69808,
+      "grad_norm": 2.2289009472089125,
+      "learning_rate": 8.380828103309699e-05,
+      "loss": 1.3174,
+      "step": 10613
+    },
+    {
+      "epoch": 1.69824,
+      "grad_norm": 2.069931750462454,
+      "learning_rate": 8.379123537771338e-05,
+      "loss": 1.2917,
+      "step": 10614
+    },
+    {
+      "epoch": 1.6984,
+      "grad_norm": 2.0589551623675186,
+      "learning_rate": 8.377419020597574e-05,
+      "loss": 1.2491,
+      "step": 10615
+    },
+    {
+      "epoch": 1.69856,
+      "grad_norm": 2.594151587812958,
+      "learning_rate": 8.375714551839257e-05,
+      "loss": 1.3337,
+      "step": 10616
+    },
+    {
+      "epoch": 1.69872,
+      "grad_norm": 2.5253375428872626,
+      "learning_rate": 8.374010131547258e-05,
+      "loss": 1.324,
+      "step": 10617
+    },
+    {
+      "epoch": 1.69888,
+      "grad_norm": 2.1154406122962977,
+      "learning_rate": 8.372305759772421e-05,
+      "loss": 1.2847,
+      "step": 10618
+    },
+    {
+      "epoch": 1.69904,
+      "grad_norm": 1.954361857085983,
+      "learning_rate": 8.370601436565611e-05,
+      "loss": 1.3349,
+      "step": 10619
+    },
+    {
+      "epoch": 1.6992,
+      "grad_norm": 1.9683771825132466,
+      "learning_rate": 8.368897161977678e-05,
+      "loss": 1.2708,
+      "step": 10620
+    },
+    {
+      "epoch": 1.69936,
+      "grad_norm": 2.058901990687157,
+      "learning_rate": 8.367192936059478e-05,
+      "loss": 1.2634,
+      "step": 10621
+    },
+    {
+      "epoch": 1.6995200000000001,
+      "grad_norm": 1.934429281032226,
+      "learning_rate": 8.36548875886186e-05,
+      "loss": 1.3583,
+      "step": 10622
+    },
+    {
+      "epoch": 1.6996799999999999,
+      "grad_norm": 2.2194140495999064,
+      "learning_rate": 8.36378463043568e-05,
+      "loss": 1.3243,
+      "step": 10623
+    },
+    {
+      "epoch": 1.69984,
+      "grad_norm": 2.0959637270718674,
+      "learning_rate": 8.362080550831776e-05,
+      "loss": 1.299,
+      "step": 10624
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 1.95635571407355,
+      "learning_rate": 8.360376520101009e-05,
+      "loss": 1.2799,
+      "step": 10625
+    },
+    {
+      "epoch": 1.70016,
+      "grad_norm": 2.00336223158127,
+      "learning_rate": 8.358672538294212e-05,
+      "loss": 1.2348,
+      "step": 10626
+    },
+    {
+      "epoch": 1.70032,
+      "grad_norm": 1.9401122904685515,
+      "learning_rate": 8.356968605462239e-05,
+      "loss": 1.3329,
+      "step": 10627
+    },
+    {
+      "epoch": 1.70048,
+      "grad_norm": 2.137329906049735,
+      "learning_rate": 8.355264721655924e-05,
+      "loss": 1.2733,
+      "step": 10628
+    },
+    {
+      "epoch": 1.70064,
+      "grad_norm": 1.8080148673475591,
+      "learning_rate": 8.353560886926115e-05,
+      "loss": 1.2138,
+      "step": 10629
+    },
+    {
+      "epoch": 1.7008,
+      "grad_norm": 2.955661614091977,
+      "learning_rate": 8.351857101323649e-05,
+      "loss": 1.2705,
+      "step": 10630
+    },
+    {
+      "epoch": 1.70096,
+      "grad_norm": 1.879958729008374,
+      "learning_rate": 8.350153364899369e-05,
+      "loss": 1.2994,
+      "step": 10631
+    },
+    {
+      "epoch": 1.70112,
+      "grad_norm": 2.2600190330497507,
+      "learning_rate": 8.348449677704104e-05,
+      "loss": 1.2642,
+      "step": 10632
+    },
+    {
+      "epoch": 1.7012800000000001,
+      "grad_norm": 2.2807539283930343,
+      "learning_rate": 8.3467460397887e-05,
+      "loss": 1.3179,
+      "step": 10633
+    },
+    {
+      "epoch": 1.7014399999999998,
+      "grad_norm": 2.021561249139995,
+      "learning_rate": 8.345042451203979e-05,
+      "loss": 1.3152,
+      "step": 10634
+    },
+    {
+      "epoch": 1.7016,
+      "grad_norm": 1.9915894086824564,
+      "learning_rate": 8.343338912000785e-05,
+      "loss": 1.3057,
+      "step": 10635
+    },
+    {
+      "epoch": 1.7017600000000002,
+      "grad_norm": 1.7551014039457358,
+      "learning_rate": 8.341635422229939e-05,
+      "loss": 1.2955,
+      "step": 10636
+    },
+    {
+      "epoch": 1.7019199999999999,
+      "grad_norm": 1.9489364584599826,
+      "learning_rate": 8.339931981942275e-05,
+      "loss": 1.3325,
+      "step": 10637
+    },
+    {
+      "epoch": 1.70208,
+      "grad_norm": 2.26269526191194,
+      "learning_rate": 8.338228591188624e-05,
+      "loss": 1.2797,
+      "step": 10638
+    },
+    {
+      "epoch": 1.70224,
+      "grad_norm": 2.146315948282092,
+      "learning_rate": 8.33652525001981e-05,
+      "loss": 1.4223,
+      "step": 10639
+    },
+    {
+      "epoch": 1.7024,
+      "grad_norm": 1.8935153898790282,
+      "learning_rate": 8.334821958486662e-05,
+      "loss": 1.2171,
+      "step": 10640
+    },
+    {
+      "epoch": 1.70256,
+      "grad_norm": 2.0702099015852924,
+      "learning_rate": 8.333118716639995e-05,
+      "loss": 1.3076,
+      "step": 10641
+    },
+    {
+      "epoch": 1.70272,
+      "grad_norm": 2.001374028153196,
+      "learning_rate": 8.331415524530642e-05,
+      "loss": 1.3246,
+      "step": 10642
+    },
+    {
+      "epoch": 1.70288,
+      "grad_norm": 2.337204839443255,
+      "learning_rate": 8.329712382209414e-05,
+      "loss": 1.1855,
+      "step": 10643
+    },
+    {
+      "epoch": 1.70304,
+      "grad_norm": 1.9938459781260185,
+      "learning_rate": 8.328009289727138e-05,
+      "loss": 1.287,
+      "step": 10644
+    },
+    {
+      "epoch": 1.7032,
+      "grad_norm": 1.8857504696181986,
+      "learning_rate": 8.326306247134624e-05,
+      "loss": 1.2653,
+      "step": 10645
+    },
+    {
+      "epoch": 1.70336,
+      "grad_norm": 1.813167298301399,
+      "learning_rate": 8.324603254482696e-05,
+      "loss": 1.2964,
+      "step": 10646
+    },
+    {
+      "epoch": 1.7035200000000001,
+      "grad_norm": 1.9061816845245152,
+      "learning_rate": 8.322900311822164e-05,
+      "loss": 1.2449,
+      "step": 10647
+    },
+    {
+      "epoch": 1.7036799999999999,
+      "grad_norm": 2.252836742848457,
+      "learning_rate": 8.321197419203845e-05,
+      "loss": 1.3109,
+      "step": 10648
+    },
+    {
+      "epoch": 1.70384,
+      "grad_norm": 1.7279548194513137,
+      "learning_rate": 8.319494576678546e-05,
+      "loss": 1.231,
+      "step": 10649
+    },
+    {
+      "epoch": 1.704,
+      "grad_norm": 2.131426046973759,
+      "learning_rate": 8.317791784297084e-05,
+      "loss": 1.2896,
+      "step": 10650
+    },
+    {
+      "epoch": 1.70416,
+      "grad_norm": 2.0235218370168506,
+      "learning_rate": 8.31608904211026e-05,
+      "loss": 1.242,
+      "step": 10651
+    },
+    {
+      "epoch": 1.70432,
+      "grad_norm": 1.9331627595573941,
+      "learning_rate": 8.31438635016889e-05,
+      "loss": 1.2498,
+      "step": 10652
+    },
+    {
+      "epoch": 1.70448,
+      "grad_norm": 2.0021169454686722,
+      "learning_rate": 8.312683708523769e-05,
+      "loss": 1.2587,
+      "step": 10653
+    },
+    {
+      "epoch": 1.70464,
+      "grad_norm": 2.1064215314043424,
+      "learning_rate": 8.310981117225711e-05,
+      "loss": 1.2923,
+      "step": 10654
+    },
+    {
+      "epoch": 1.7048,
+      "grad_norm": 2.0409071292927146,
+      "learning_rate": 8.309278576325513e-05,
+      "loss": 1.2936,
+      "step": 10655
+    },
+    {
+      "epoch": 1.70496,
+      "grad_norm": 2.065577270359159,
+      "learning_rate": 8.307576085873982e-05,
+      "loss": 1.2118,
+      "step": 10656
+    },
+    {
+      "epoch": 1.70512,
+      "grad_norm": 2.2031319759120858,
+      "learning_rate": 8.30587364592191e-05,
+      "loss": 1.2784,
+      "step": 10657
+    },
+    {
+      "epoch": 1.7052800000000001,
+      "grad_norm": 1.9739533969128775,
+      "learning_rate": 8.304171256520106e-05,
+      "loss": 1.2868,
+      "step": 10658
+    },
+    {
+      "epoch": 1.7054399999999998,
+      "grad_norm": 1.853283491456626,
+      "learning_rate": 8.302468917719355e-05,
+      "loss": 1.2498,
+      "step": 10659
+    },
+    {
+      "epoch": 1.7056,
+      "grad_norm": 2.0160133327470935,
+      "learning_rate": 8.300766629570465e-05,
+      "loss": 1.2676,
+      "step": 10660
+    },
+    {
+      "epoch": 1.7057600000000002,
+      "grad_norm": 2.310827684558638,
+      "learning_rate": 8.299064392124215e-05,
+      "loss": 1.2862,
+      "step": 10661
+    },
+    {
+      "epoch": 1.7059199999999999,
+      "grad_norm": 2.351624440093474,
+      "learning_rate": 8.297362205431409e-05,
+      "loss": 1.3039,
+      "step": 10662
+    },
+    {
+      "epoch": 1.70608,
+      "grad_norm": 1.9728964741179926,
+      "learning_rate": 8.29566006954283e-05,
+      "loss": 1.2963,
+      "step": 10663
+    },
+    {
+      "epoch": 1.70624,
+      "grad_norm": 2.212530837470629,
+      "learning_rate": 8.293957984509276e-05,
+      "loss": 1.3064,
+      "step": 10664
+    },
+    {
+      "epoch": 1.7064,
+      "grad_norm": 2.1284220318849543,
+      "learning_rate": 8.292255950381526e-05,
+      "loss": 1.2067,
+      "step": 10665
+    },
+    {
+      "epoch": 1.70656,
+      "grad_norm": 1.9833849002440955,
+      "learning_rate": 8.290553967210374e-05,
+      "loss": 1.3265,
+      "step": 10666
+    },
+    {
+      "epoch": 1.70672,
+      "grad_norm": 1.9793873663252741,
+      "learning_rate": 8.288852035046596e-05,
+      "loss": 1.2394,
+      "step": 10667
+    },
+    {
+      "epoch": 1.70688,
+      "grad_norm": 1.7565168471447827,
+      "learning_rate": 8.287150153940984e-05,
+      "loss": 1.2292,
+      "step": 10668
+    },
+    {
+      "epoch": 1.7070400000000001,
+      "grad_norm": 2.041747572603573,
+      "learning_rate": 8.285448323944312e-05,
+      "loss": 1.298,
+      "step": 10669
+    },
+    {
+      "epoch": 1.7072,
+      "grad_norm": 1.9870480209476151,
+      "learning_rate": 8.283746545107366e-05,
+      "loss": 1.2831,
+      "step": 10670
+    },
+    {
+      "epoch": 1.70736,
+      "grad_norm": 2.0797040758182823,
+      "learning_rate": 8.282044817480919e-05,
+      "loss": 1.3494,
+      "step": 10671
+    },
+    {
+      "epoch": 1.7075200000000001,
+      "grad_norm": 2.3932859315927466,
+      "learning_rate": 8.280343141115756e-05,
+      "loss": 1.2685,
+      "step": 10672
+    },
+    {
+      "epoch": 1.7076799999999999,
+      "grad_norm": 1.9205867223694668,
+      "learning_rate": 8.278641516062643e-05,
+      "loss": 1.268,
+      "step": 10673
+    },
+    {
+      "epoch": 1.70784,
+      "grad_norm": 2.135821139996562,
+      "learning_rate": 8.276939942372366e-05,
+      "loss": 1.2439,
+      "step": 10674
+    },
+    {
+      "epoch": 1.708,
+      "grad_norm": 1.8725331720021627,
+      "learning_rate": 8.275238420095684e-05,
+      "loss": 1.3075,
+      "step": 10675
+    },
+    {
+      "epoch": 1.70816,
+      "grad_norm": 2.1704691294710754,
+      "learning_rate": 8.273536949283378e-05,
+      "loss": 1.2938,
+      "step": 10676
+    },
+    {
+      "epoch": 1.70832,
+      "grad_norm": 2.0362413071653283,
+      "learning_rate": 8.27183552998621e-05,
+      "loss": 1.2338,
+      "step": 10677
+    },
+    {
+      "epoch": 1.70848,
+      "grad_norm": 1.721446555329932,
+      "learning_rate": 8.270134162254955e-05,
+      "loss": 1.258,
+      "step": 10678
+    },
+    {
+      "epoch": 1.70864,
+      "grad_norm": 1.9986220523409353,
+      "learning_rate": 8.268432846140372e-05,
+      "loss": 1.2152,
+      "step": 10679
+    },
+    {
+      "epoch": 1.7088,
+      "grad_norm": 2.1509749675645193,
+      "learning_rate": 8.266731581693234e-05,
+      "loss": 1.3014,
+      "step": 10680
+    },
+    {
+      "epoch": 1.70896,
+      "grad_norm": 1.8907381760733262,
+      "learning_rate": 8.265030368964296e-05,
+      "loss": 1.3093,
+      "step": 10681
+    },
+    {
+      "epoch": 1.70912,
+      "grad_norm": 2.341384529109516,
+      "learning_rate": 8.263329208004329e-05,
+      "loss": 1.2936,
+      "step": 10682
+    },
+    {
+      "epoch": 1.7092800000000001,
+      "grad_norm": 1.870970994904789,
+      "learning_rate": 8.261628098864083e-05,
+      "loss": 1.2802,
+      "step": 10683
+    },
+    {
+      "epoch": 1.7094399999999998,
+      "grad_norm": 1.8900597243254713,
+      "learning_rate": 8.259927041594325e-05,
+      "loss": 1.3319,
+      "step": 10684
+    },
+    {
+      "epoch": 1.7096,
+      "grad_norm": 2.1338009499108535,
+      "learning_rate": 8.258226036245804e-05,
+      "loss": 1.2959,
+      "step": 10685
+    },
+    {
+      "epoch": 1.70976,
+      "grad_norm": 2.1888747634231227,
+      "learning_rate": 8.256525082869282e-05,
+      "loss": 1.2788,
+      "step": 10686
+    },
+    {
+      "epoch": 1.7099199999999999,
+      "grad_norm": 2.0927156535245275,
+      "learning_rate": 8.254824181515511e-05,
+      "loss": 1.2411,
+      "step": 10687
+    },
+    {
+      "epoch": 1.71008,
+      "grad_norm": 2.073876562700155,
+      "learning_rate": 8.253123332235246e-05,
+      "loss": 1.3396,
+      "step": 10688
+    },
+    {
+      "epoch": 1.71024,
+      "grad_norm": 1.7230049457311802,
+      "learning_rate": 8.251422535079232e-05,
+      "loss": 1.3077,
+      "step": 10689
+    },
+    {
+      "epoch": 1.7104,
+      "grad_norm": 2.1257421862804153,
+      "learning_rate": 8.249721790098226e-05,
+      "loss": 1.2956,
+      "step": 10690
+    },
+    {
+      "epoch": 1.71056,
+      "grad_norm": 2.0796151524083246,
+      "learning_rate": 8.248021097342967e-05,
+      "loss": 1.2601,
+      "step": 10691
+    },
+    {
+      "epoch": 1.71072,
+      "grad_norm": 2.1409412164966493,
+      "learning_rate": 8.246320456864208e-05,
+      "loss": 1.2662,
+      "step": 10692
+    },
+    {
+      "epoch": 1.71088,
+      "grad_norm": 1.9358351388454982,
+      "learning_rate": 8.244619868712688e-05,
+      "loss": 1.281,
+      "step": 10693
+    },
+    {
+      "epoch": 1.7110400000000001,
+      "grad_norm": 1.7945877927421485,
+      "learning_rate": 8.242919332939156e-05,
+      "loss": 1.2478,
+      "step": 10694
+    },
+    {
+      "epoch": 1.7112,
+      "grad_norm": 2.333920171759413,
+      "learning_rate": 8.241218849594349e-05,
+      "loss": 1.231,
+      "step": 10695
+    },
+    {
+      "epoch": 1.71136,
+      "grad_norm": 1.9432809957526258,
+      "learning_rate": 8.239518418729008e-05,
+      "loss": 1.3108,
+      "step": 10696
+    },
+    {
+      "epoch": 1.7115200000000002,
+      "grad_norm": 2.130770714954646,
+      "learning_rate": 8.237818040393877e-05,
+      "loss": 1.2737,
+      "step": 10697
+    },
+    {
+      "epoch": 1.7116799999999999,
+      "grad_norm": 1.8466238626747973,
+      "learning_rate": 8.236117714639681e-05,
+      "loss": 1.2712,
+      "step": 10698
+    },
+    {
+      "epoch": 1.71184,
+      "grad_norm": 1.9480832431583737,
+      "learning_rate": 8.234417441517169e-05,
+      "loss": 1.3074,
+      "step": 10699
+    },
+    {
+      "epoch": 1.712,
+      "grad_norm": 2.3363942128099633,
+      "learning_rate": 8.232717221077063e-05,
+      "loss": 1.2528,
+      "step": 10700
+    },
+    {
+      "epoch": 1.71216,
+      "grad_norm": 2.01112026770422,
+      "learning_rate": 8.231017053370104e-05,
+      "loss": 1.2625,
+      "step": 10701
+    },
+    {
+      "epoch": 1.71232,
+      "grad_norm": 2.283304910479188,
+      "learning_rate": 8.229316938447015e-05,
+      "loss": 1.297,
+      "step": 10702
+    },
+    {
+      "epoch": 1.71248,
+      "grad_norm": 2.0802725170532086,
+      "learning_rate": 8.22761687635853e-05,
+      "loss": 1.238,
+      "step": 10703
+    },
+    {
+      "epoch": 1.71264,
+      "grad_norm": 2.3960349662065363,
+      "learning_rate": 8.225916867155372e-05,
+      "loss": 1.3404,
+      "step": 10704
+    },
+    {
+      "epoch": 1.7128,
+      "grad_norm": 2.0383911539588278,
+      "learning_rate": 8.224216910888274e-05,
+      "loss": 1.2417,
+      "step": 10705
+    },
+    {
+      "epoch": 1.71296,
+      "grad_norm": 1.9274331152997666,
+      "learning_rate": 8.222517007607951e-05,
+      "loss": 1.2863,
+      "step": 10706
+    },
+    {
+      "epoch": 1.71312,
+      "grad_norm": 2.155653880213729,
+      "learning_rate": 8.220817157365137e-05,
+      "loss": 1.2749,
+      "step": 10707
+    },
+    {
+      "epoch": 1.7132800000000001,
+      "grad_norm": 2.000942309033394,
+      "learning_rate": 8.21911736021054e-05,
+      "loss": 1.2364,
+      "step": 10708
+    },
+    {
+      "epoch": 1.7134399999999999,
+      "grad_norm": 2.035140339129986,
+      "learning_rate": 8.217417616194891e-05,
+      "loss": 1.4047,
+      "step": 10709
+    },
+    {
+      "epoch": 1.7136,
+      "grad_norm": 1.8857504762700603,
+      "learning_rate": 8.215717925368899e-05,
+      "loss": 1.2525,
+      "step": 10710
+    },
+    {
+      "epoch": 1.71376,
+      "grad_norm": 2.356935356179782,
+      "learning_rate": 8.214018287783285e-05,
+      "loss": 1.227,
+      "step": 10711
+    },
+    {
+      "epoch": 1.7139199999999999,
+      "grad_norm": 2.5652897164788717,
+      "learning_rate": 8.212318703488762e-05,
+      "loss": 1.288,
+      "step": 10712
+    },
+    {
+      "epoch": 1.71408,
+      "grad_norm": 1.9898318278822007,
+      "learning_rate": 8.210619172536048e-05,
+      "loss": 1.2581,
+      "step": 10713
+    },
+    {
+      "epoch": 1.71424,
+      "grad_norm": 2.1908342518624018,
+      "learning_rate": 8.208919694975846e-05,
+      "loss": 1.2918,
+      "step": 10714
+    },
+    {
+      "epoch": 1.7144,
+      "grad_norm": 2.2341529307793078,
+      "learning_rate": 8.207220270858873e-05,
+      "loss": 1.3638,
+      "step": 10715
+    },
+    {
+      "epoch": 1.71456,
+      "grad_norm": 2.2448591452526476,
+      "learning_rate": 8.205520900235833e-05,
+      "loss": 1.2626,
+      "step": 10716
+    },
+    {
+      "epoch": 1.71472,
+      "grad_norm": 2.2053112031982147,
+      "learning_rate": 8.203821583157438e-05,
+      "loss": 1.2748,
+      "step": 10717
+    },
+    {
+      "epoch": 1.71488,
+      "grad_norm": 1.9090302207425178,
+      "learning_rate": 8.202122319674386e-05,
+      "loss": 1.2548,
+      "step": 10718
+    },
+    {
+      "epoch": 1.7150400000000001,
+      "grad_norm": 2.6664776501819354,
+      "learning_rate": 8.200423109837385e-05,
+      "loss": 1.3014,
+      "step": 10719
+    },
+    {
+      "epoch": 1.7151999999999998,
+      "grad_norm": 1.978069104304482,
+      "learning_rate": 8.198723953697135e-05,
+      "loss": 1.2484,
+      "step": 10720
+    },
+    {
+      "epoch": 1.71536,
+      "grad_norm": 1.9529476327230917,
+      "learning_rate": 8.197024851304341e-05,
+      "loss": 1.3019,
+      "step": 10721
+    },
+    {
+      "epoch": 1.7155200000000002,
+      "grad_norm": 1.8542973844479358,
+      "learning_rate": 8.195325802709694e-05,
+      "loss": 1.2584,
+      "step": 10722
+    },
+    {
+      "epoch": 1.7156799999999999,
+      "grad_norm": 2.2191291823444246,
+      "learning_rate": 8.193626807963901e-05,
+      "loss": 1.3215,
+      "step": 10723
+    },
+    {
+      "epoch": 1.71584,
+      "grad_norm": 2.0102627385147604,
+      "learning_rate": 8.191927867117645e-05,
+      "loss": 1.275,
+      "step": 10724
+    },
+    {
+      "epoch": 1.716,
+      "grad_norm": 1.877539187037927,
+      "learning_rate": 8.190228980221634e-05,
+      "loss": 1.28,
+      "step": 10725
+    },
+    {
+      "epoch": 1.71616,
+      "grad_norm": 2.397380652510999,
+      "learning_rate": 8.188530147326548e-05,
+      "loss": 1.3411,
+      "step": 10726
+    },
+    {
+      "epoch": 1.71632,
+      "grad_norm": 2.224015729726687,
+      "learning_rate": 8.186831368483083e-05,
+      "loss": 1.2777,
+      "step": 10727
+    },
+    {
+      "epoch": 1.71648,
+      "grad_norm": 2.327328512710468,
+      "learning_rate": 8.185132643741929e-05,
+      "loss": 1.2868,
+      "step": 10728
+    },
+    {
+      "epoch": 1.71664,
+      "grad_norm": 2.0587721598804802,
+      "learning_rate": 8.183433973153773e-05,
+      "loss": 1.2252,
+      "step": 10729
+    },
+    {
+      "epoch": 1.7168,
+      "grad_norm": 2.142519559321697,
+      "learning_rate": 8.181735356769299e-05,
+      "loss": 1.3098,
+      "step": 10730
+    },
+    {
+      "epoch": 1.71696,
+      "grad_norm": 1.9816231717396744,
+      "learning_rate": 8.180036794639195e-05,
+      "loss": 1.2185,
+      "step": 10731
+    },
+    {
+      "epoch": 1.71712,
+      "grad_norm": 2.1371710215433413,
+      "learning_rate": 8.178338286814138e-05,
+      "loss": 1.2914,
+      "step": 10732
+    },
+    {
+      "epoch": 1.7172800000000001,
+      "grad_norm": 2.249379970294028,
+      "learning_rate": 8.176639833344815e-05,
+      "loss": 1.3366,
+      "step": 10733
+    },
+    {
+      "epoch": 1.7174399999999999,
+      "grad_norm": 1.7207197910139325,
+      "learning_rate": 8.1749414342819e-05,
+      "loss": 1.303,
+      "step": 10734
+    },
+    {
+      "epoch": 1.7176,
+      "grad_norm": 1.9710467126648563,
+      "learning_rate": 8.173243089676073e-05,
+      "loss": 1.3161,
+      "step": 10735
+    },
+    {
+      "epoch": 1.71776,
+      "grad_norm": 2.491039826443675,
+      "learning_rate": 8.171544799578011e-05,
+      "loss": 1.2983,
+      "step": 10736
+    },
+    {
+      "epoch": 1.71792,
+      "grad_norm": 1.8800990864583655,
+      "learning_rate": 8.169846564038389e-05,
+      "loss": 1.251,
+      "step": 10737
+    },
+    {
+      "epoch": 1.71808,
+      "grad_norm": 2.2412487084048833,
+      "learning_rate": 8.168148383107876e-05,
+      "loss": 1.3178,
+      "step": 10738
+    },
+    {
+      "epoch": 1.71824,
+      "grad_norm": 2.101556384920154,
+      "learning_rate": 8.16645025683715e-05,
+      "loss": 1.3987,
+      "step": 10739
+    },
+    {
+      "epoch": 1.7184,
+      "grad_norm": 2.3304842481971417,
+      "learning_rate": 8.164752185276873e-05,
+      "loss": 1.3367,
+      "step": 10740
+    },
+    {
+      "epoch": 1.71856,
+      "grad_norm": 2.3575487316411676,
+      "learning_rate": 8.163054168477721e-05,
+      "loss": 1.2645,
+      "step": 10741
+    },
+    {
+      "epoch": 1.71872,
+      "grad_norm": 2.408353326576776,
+      "learning_rate": 8.161356206490351e-05,
+      "loss": 1.2188,
+      "step": 10742
+    },
+    {
+      "epoch": 1.71888,
+      "grad_norm": 2.2484066940737426,
+      "learning_rate": 8.159658299365435e-05,
+      "loss": 1.3635,
+      "step": 10743
+    },
+    {
+      "epoch": 1.7190400000000001,
+      "grad_norm": 2.5365270847210586,
+      "learning_rate": 8.157960447153631e-05,
+      "loss": 1.2922,
+      "step": 10744
+    },
+    {
+      "epoch": 1.7191999999999998,
+      "grad_norm": 2.3247183592910825,
+      "learning_rate": 8.156262649905607e-05,
+      "loss": 1.3357,
+      "step": 10745
+    },
+    {
+      "epoch": 1.71936,
+      "grad_norm": 2.0495202647940065,
+      "learning_rate": 8.154564907672015e-05,
+      "loss": 1.3422,
+      "step": 10746
+    },
+    {
+      "epoch": 1.7195200000000002,
+      "grad_norm": 2.554047945982881,
+      "learning_rate": 8.152867220503522e-05,
+      "loss": 1.319,
+      "step": 10747
+    },
+    {
+      "epoch": 1.7196799999999999,
+      "grad_norm": 1.8821725773741311,
+      "learning_rate": 8.151169588450776e-05,
+      "loss": 1.2905,
+      "step": 10748
+    },
+    {
+      "epoch": 1.71984,
+      "grad_norm": 1.942533331216054,
+      "learning_rate": 8.149472011564439e-05,
+      "loss": 1.2872,
+      "step": 10749
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 1.9246898441775704,
+      "learning_rate": 8.147774489895154e-05,
+      "loss": 1.2712,
+      "step": 10750
+    },
+    {
+      "epoch": 1.72016,
+      "grad_norm": 2.2117702664493013,
+      "learning_rate": 8.146077023493585e-05,
+      "loss": 1.3119,
+      "step": 10751
+    },
+    {
+      "epoch": 1.72032,
+      "grad_norm": 1.9007032700491067,
+      "learning_rate": 8.144379612410372e-05,
+      "loss": 1.3158,
+      "step": 10752
+    },
+    {
+      "epoch": 1.72048,
+      "grad_norm": 2.082105754584741,
+      "learning_rate": 8.142682256696172e-05,
+      "loss": 1.2379,
+      "step": 10753
+    },
+    {
+      "epoch": 1.72064,
+      "grad_norm": 2.0078797094175544,
+      "learning_rate": 8.140984956401623e-05,
+      "loss": 1.2611,
+      "step": 10754
+    },
+    {
+      "epoch": 1.7208,
+      "grad_norm": 2.338563315407231,
+      "learning_rate": 8.139287711577375e-05,
+      "loss": 1.3672,
+      "step": 10755
+    },
+    {
+      "epoch": 1.72096,
+      "grad_norm": 2.7842813458920177,
+      "learning_rate": 8.137590522274072e-05,
+      "loss": 1.2918,
+      "step": 10756
+    },
+    {
+      "epoch": 1.72112,
+      "grad_norm": 1.822529305240263,
+      "learning_rate": 8.135893388542353e-05,
+      "loss": 1.2771,
+      "step": 10757
+    },
+    {
+      "epoch": 1.7212800000000001,
+      "grad_norm": 1.8231631451431007,
+      "learning_rate": 8.134196310432861e-05,
+      "loss": 1.2714,
+      "step": 10758
+    },
+    {
+      "epoch": 1.7214399999999999,
+      "grad_norm": 2.0877290341164083,
+      "learning_rate": 8.132499287996228e-05,
+      "loss": 1.2574,
+      "step": 10759
+    },
+    {
+      "epoch": 1.7216,
+      "grad_norm": 2.043587474742896,
+      "learning_rate": 8.130802321283099e-05,
+      "loss": 1.2676,
+      "step": 10760
+    },
+    {
+      "epoch": 1.72176,
+      "grad_norm": 1.8883752531315148,
+      "learning_rate": 8.129105410344104e-05,
+      "loss": 1.2325,
+      "step": 10761
+    },
+    {
+      "epoch": 1.72192,
+      "grad_norm": 2.04381063724899,
+      "learning_rate": 8.127408555229876e-05,
+      "loss": 1.2857,
+      "step": 10762
+    },
+    {
+      "epoch": 1.72208,
+      "grad_norm": 2.161007459124944,
+      "learning_rate": 8.125711755991048e-05,
+      "loss": 1.3443,
+      "step": 10763
+    },
+    {
+      "epoch": 1.72224,
+      "grad_norm": 2.147840197021629,
+      "learning_rate": 8.124015012678253e-05,
+      "loss": 1.2393,
+      "step": 10764
+    },
+    {
+      "epoch": 1.7224,
+      "grad_norm": 2.0897807754140048,
+      "learning_rate": 8.122318325342114e-05,
+      "loss": 1.2497,
+      "step": 10765
+    },
+    {
+      "epoch": 1.72256,
+      "grad_norm": 2.582758504560119,
+      "learning_rate": 8.120621694033261e-05,
+      "loss": 1.3425,
+      "step": 10766
+    },
+    {
+      "epoch": 1.72272,
+      "grad_norm": 2.0223947714881114,
+      "learning_rate": 8.118925118802317e-05,
+      "loss": 1.2987,
+      "step": 10767
+    },
+    {
+      "epoch": 1.72288,
+      "grad_norm": 2.3627551588526323,
+      "learning_rate": 8.117228599699906e-05,
+      "loss": 1.2708,
+      "step": 10768
+    },
+    {
+      "epoch": 1.7230400000000001,
+      "grad_norm": 2.193368541158803,
+      "learning_rate": 8.115532136776652e-05,
+      "loss": 1.2317,
+      "step": 10769
+    },
+    {
+      "epoch": 1.7231999999999998,
+      "grad_norm": 2.0742656915011057,
+      "learning_rate": 8.113835730083174e-05,
+      "loss": 1.2321,
+      "step": 10770
+    },
+    {
+      "epoch": 1.72336,
+      "grad_norm": 1.9745730860278705,
+      "learning_rate": 8.112139379670085e-05,
+      "loss": 1.2842,
+      "step": 10771
+    },
+    {
+      "epoch": 1.7235200000000002,
+      "grad_norm": 2.3747314707167053,
+      "learning_rate": 8.110443085588011e-05,
+      "loss": 1.2786,
+      "step": 10772
+    },
+    {
+      "epoch": 1.7236799999999999,
+      "grad_norm": 2.039390529522154,
+      "learning_rate": 8.108746847887558e-05,
+      "loss": 1.3729,
+      "step": 10773
+    },
+    {
+      "epoch": 1.72384,
+      "grad_norm": 2.4812826005946196,
+      "learning_rate": 8.107050666619347e-05,
+      "loss": 1.265,
+      "step": 10774
+    },
+    {
+      "epoch": 1.724,
+      "grad_norm": 1.9262188664241218,
+      "learning_rate": 8.105354541833983e-05,
+      "loss": 1.2543,
+      "step": 10775
+    },
+    {
+      "epoch": 1.72416,
+      "grad_norm": 2.1409058283323423,
+      "learning_rate": 8.103658473582079e-05,
+      "loss": 1.383,
+      "step": 10776
+    },
+    {
+      "epoch": 1.72432,
+      "grad_norm": 2.000552563470156,
+      "learning_rate": 8.101962461914244e-05,
+      "loss": 1.2793,
+      "step": 10777
+    },
+    {
+      "epoch": 1.72448,
+      "grad_norm": 1.9882039744210815,
+      "learning_rate": 8.100266506881082e-05,
+      "loss": 1.3737,
+      "step": 10778
+    },
+    {
+      "epoch": 1.72464,
+      "grad_norm": 1.9733368679598347,
+      "learning_rate": 8.098570608533198e-05,
+      "loss": 1.2242,
+      "step": 10779
+    },
+    {
+      "epoch": 1.7248,
+      "grad_norm": 2.0061050901350788,
+      "learning_rate": 8.096874766921201e-05,
+      "loss": 1.1893,
+      "step": 10780
+    },
+    {
+      "epoch": 1.72496,
+      "grad_norm": 1.9629796012938132,
+      "learning_rate": 8.095178982095684e-05,
+      "loss": 1.2931,
+      "step": 10781
+    },
+    {
+      "epoch": 1.72512,
+      "grad_norm": 2.1740902725051785,
+      "learning_rate": 8.093483254107253e-05,
+      "loss": 1.3361,
+      "step": 10782
+    },
+    {
+      "epoch": 1.7252800000000001,
+      "grad_norm": 1.9805172712162358,
+      "learning_rate": 8.0917875830065e-05,
+      "loss": 1.3111,
+      "step": 10783
+    },
+    {
+      "epoch": 1.7254399999999999,
+      "grad_norm": 2.0891250375244583,
+      "learning_rate": 8.090091968844027e-05,
+      "loss": 1.2383,
+      "step": 10784
+    },
+    {
+      "epoch": 1.7256,
+      "grad_norm": 2.3692714069779606,
+      "learning_rate": 8.088396411670426e-05,
+      "loss": 1.3579,
+      "step": 10785
+    },
+    {
+      "epoch": 1.72576,
+      "grad_norm": 2.196848989232728,
+      "learning_rate": 8.086700911536293e-05,
+      "loss": 1.3054,
+      "step": 10786
+    },
+    {
+      "epoch": 1.72592,
+      "grad_norm": 2.667455563078945,
+      "learning_rate": 8.085005468492215e-05,
+      "loss": 1.3383,
+      "step": 10787
+    },
+    {
+      "epoch": 1.72608,
+      "grad_norm": 1.9849424021190103,
+      "learning_rate": 8.083310082588786e-05,
+      "loss": 1.2743,
+      "step": 10788
+    },
+    {
+      "epoch": 1.72624,
+      "grad_norm": 1.9673225641629233,
+      "learning_rate": 8.081614753876587e-05,
+      "loss": 1.3504,
+      "step": 10789
+    },
+    {
+      "epoch": 1.7264,
+      "grad_norm": 2.4478984383121456,
+      "learning_rate": 8.079919482406211e-05,
+      "loss": 1.2978,
+      "step": 10790
+    },
+    {
+      "epoch": 1.72656,
+      "grad_norm": 1.9014454605559261,
+      "learning_rate": 8.078224268228238e-05,
+      "loss": 1.2319,
+      "step": 10791
+    },
+    {
+      "epoch": 1.72672,
+      "grad_norm": 2.396331164590954,
+      "learning_rate": 8.076529111393254e-05,
+      "loss": 1.2847,
+      "step": 10792
+    },
+    {
+      "epoch": 1.72688,
+      "grad_norm": 2.101717970894753,
+      "learning_rate": 8.074834011951834e-05,
+      "loss": 1.3776,
+      "step": 10793
+    },
+    {
+      "epoch": 1.7270400000000001,
+      "grad_norm": 1.7250379218297132,
+      "learning_rate": 8.07313896995457e-05,
+      "loss": 1.2974,
+      "step": 10794
+    },
+    {
+      "epoch": 1.7271999999999998,
+      "grad_norm": 2.1307985351761065,
+      "learning_rate": 8.071443985452025e-05,
+      "loss": 1.1912,
+      "step": 10795
+    },
+    {
+      "epoch": 1.72736,
+      "grad_norm": 2.2085282312422376,
+      "learning_rate": 8.069749058494787e-05,
+      "loss": 1.2173,
+      "step": 10796
+    },
+    {
+      "epoch": 1.72752,
+      "grad_norm": 1.9412310661513508,
+      "learning_rate": 8.068054189133418e-05,
+      "loss": 1.3003,
+      "step": 10797
+    },
+    {
+      "epoch": 1.7276799999999999,
+      "grad_norm": 1.8877335564477287,
+      "learning_rate": 8.066359377418502e-05,
+      "loss": 1.2888,
+      "step": 10798
+    },
+    {
+      "epoch": 1.72784,
+      "grad_norm": 1.9061389975298229,
+      "learning_rate": 8.0646646234006e-05,
+      "loss": 1.2699,
+      "step": 10799
+    },
+    {
+      "epoch": 1.728,
+      "grad_norm": 2.2143387120466134,
+      "learning_rate": 8.06296992713029e-05,
+      "loss": 1.3538,
+      "step": 10800
+    },
+    {
+      "epoch": 1.72816,
+      "grad_norm": 2.047663118662529,
+      "learning_rate": 8.06127528865813e-05,
+      "loss": 1.235,
+      "step": 10801
+    },
+    {
+      "epoch": 1.72832,
+      "grad_norm": 1.9299015358221208,
+      "learning_rate": 8.059580708034695e-05,
+      "loss": 1.2374,
+      "step": 10802
+    },
+    {
+      "epoch": 1.72848,
+      "grad_norm": 1.8276101128136952,
+      "learning_rate": 8.057886185310542e-05,
+      "loss": 1.261,
+      "step": 10803
+    },
+    {
+      "epoch": 1.72864,
+      "grad_norm": 2.2527490627290456,
+      "learning_rate": 8.056191720536236e-05,
+      "loss": 1.3162,
+      "step": 10804
+    },
+    {
+      "epoch": 1.7288000000000001,
+      "grad_norm": 2.1291817747098825,
+      "learning_rate": 8.054497313762335e-05,
+      "loss": 1.2401,
+      "step": 10805
+    },
+    {
+      "epoch": 1.72896,
+      "grad_norm": 2.1871297289634515,
+      "learning_rate": 8.052802965039404e-05,
+      "loss": 1.333,
+      "step": 10806
+    },
+    {
+      "epoch": 1.72912,
+      "grad_norm": 1.9516418515285383,
+      "learning_rate": 8.051108674417989e-05,
+      "loss": 1.2958,
+      "step": 10807
+    },
+    {
+      "epoch": 1.7292800000000002,
+      "grad_norm": 2.368188135956289,
+      "learning_rate": 8.049414441948653e-05,
+      "loss": 1.3613,
+      "step": 10808
+    },
+    {
+      "epoch": 1.7294399999999999,
+      "grad_norm": 2.2674598388203893,
+      "learning_rate": 8.047720267681947e-05,
+      "loss": 1.2773,
+      "step": 10809
+    },
+    {
+      "epoch": 1.7296,
+      "grad_norm": 1.871792293647255,
+      "learning_rate": 8.046026151668428e-05,
+      "loss": 1.2642,
+      "step": 10810
+    },
+    {
+      "epoch": 1.72976,
+      "grad_norm": 2.0010426012185776,
+      "learning_rate": 8.044332093958638e-05,
+      "loss": 1.3314,
+      "step": 10811
+    },
+    {
+      "epoch": 1.72992,
+      "grad_norm": 1.9851657535129934,
+      "learning_rate": 8.042638094603125e-05,
+      "loss": 1.322,
+      "step": 10812
+    },
+    {
+      "epoch": 1.73008,
+      "grad_norm": 1.9931525634429716,
+      "learning_rate": 8.040944153652446e-05,
+      "loss": 1.2186,
+      "step": 10813
+    },
+    {
+      "epoch": 1.73024,
+      "grad_norm": 2.5933942078268664,
+      "learning_rate": 8.039250271157134e-05,
+      "loss": 1.2866,
+      "step": 10814
+    },
+    {
+      "epoch": 1.7304,
+      "grad_norm": 2.249795433134664,
+      "learning_rate": 8.037556447167742e-05,
+      "loss": 1.2537,
+      "step": 10815
+    },
+    {
+      "epoch": 1.73056,
+      "grad_norm": 1.8671542617172059,
+      "learning_rate": 8.0358626817348e-05,
+      "loss": 1.2029,
+      "step": 10816
+    },
+    {
+      "epoch": 1.73072,
+      "grad_norm": 2.160934571327781,
+      "learning_rate": 8.034168974908858e-05,
+      "loss": 1.268,
+      "step": 10817
+    },
+    {
+      "epoch": 1.73088,
+      "grad_norm": 2.2700201087957286,
+      "learning_rate": 8.032475326740448e-05,
+      "loss": 1.4278,
+      "step": 10818
+    },
+    {
+      "epoch": 1.7310400000000001,
+      "grad_norm": 1.9988325047242899,
+      "learning_rate": 8.030781737280106e-05,
+      "loss": 1.2814,
+      "step": 10819
+    },
+    {
+      "epoch": 1.7311999999999999,
+      "grad_norm": 2.1668971310220475,
+      "learning_rate": 8.029088206578367e-05,
+      "loss": 1.2958,
+      "step": 10820
+    },
+    {
+      "epoch": 1.73136,
+      "grad_norm": 2.920596319968182,
+      "learning_rate": 8.027394734685768e-05,
+      "loss": 1.2788,
+      "step": 10821
+    },
+    {
+      "epoch": 1.73152,
+      "grad_norm": 2.271812433885345,
+      "learning_rate": 8.025701321652831e-05,
+      "loss": 1.3433,
+      "step": 10822
+    },
+    {
+      "epoch": 1.7316799999999999,
+      "grad_norm": 2.2272960567280697,
+      "learning_rate": 8.024007967530093e-05,
+      "loss": 1.2701,
+      "step": 10823
+    },
+    {
+      "epoch": 1.73184,
+      "grad_norm": 1.8567127048257908,
+      "learning_rate": 8.022314672368077e-05,
+      "loss": 1.3291,
+      "step": 10824
+    },
+    {
+      "epoch": 1.732,
+      "grad_norm": 1.853081111479709,
+      "learning_rate": 8.020621436217308e-05,
+      "loss": 1.2292,
+      "step": 10825
+    },
+    {
+      "epoch": 1.73216,
+      "grad_norm": 2.1980177551024926,
+      "learning_rate": 8.018928259128314e-05,
+      "loss": 1.3594,
+      "step": 10826
+    },
+    {
+      "epoch": 1.73232,
+      "grad_norm": 2.1978724923645174,
+      "learning_rate": 8.017235141151612e-05,
+      "loss": 1.3034,
+      "step": 10827
+    },
+    {
+      "epoch": 1.73248,
+      "grad_norm": 1.9596553615450016,
+      "learning_rate": 8.015542082337722e-05,
+      "loss": 1.2857,
+      "step": 10828
+    },
+    {
+      "epoch": 1.73264,
+      "grad_norm": 1.950827139757765,
+      "learning_rate": 8.013849082737169e-05,
+      "loss": 1.2969,
+      "step": 10829
+    },
+    {
+      "epoch": 1.7328000000000001,
+      "grad_norm": 1.914892401555294,
+      "learning_rate": 8.012156142400462e-05,
+      "loss": 1.329,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7329599999999998,
+      "grad_norm": 2.0113485470702828,
+      "learning_rate": 8.010463261378123e-05,
+      "loss": 1.2539,
+      "step": 10831
+    },
+    {
+      "epoch": 1.73312,
+      "grad_norm": 1.8878361073347592,
+      "learning_rate": 8.008770439720656e-05,
+      "loss": 1.228,
+      "step": 10832
+    },
+    {
+      "epoch": 1.7332800000000002,
+      "grad_norm": 2.002425037971521,
+      "learning_rate": 8.007077677478581e-05,
+      "loss": 1.281,
+      "step": 10833
+    },
+    {
+      "epoch": 1.7334399999999999,
+      "grad_norm": 2.3593541171020322,
+      "learning_rate": 8.005384974702403e-05,
+      "loss": 1.2558,
+      "step": 10834
+    },
+    {
+      "epoch": 1.7336,
+      "grad_norm": 1.8589797362893477,
+      "learning_rate": 8.003692331442632e-05,
+      "loss": 1.2632,
+      "step": 10835
+    },
+    {
+      "epoch": 1.73376,
+      "grad_norm": 2.3096655673521855,
+      "learning_rate": 8.00199974774977e-05,
+      "loss": 1.2151,
+      "step": 10836
+    },
+    {
+      "epoch": 1.73392,
+      "grad_norm": 2.2194697378182386,
+      "learning_rate": 8.000307223674328e-05,
+      "loss": 1.2542,
+      "step": 10837
+    },
+    {
+      "epoch": 1.73408,
+      "grad_norm": 2.1863309804932514,
+      "learning_rate": 7.998614759266802e-05,
+      "loss": 1.3042,
+      "step": 10838
+    },
+    {
+      "epoch": 1.73424,
+      "grad_norm": 1.742948994689287,
+      "learning_rate": 7.996922354577699e-05,
+      "loss": 1.2787,
+      "step": 10839
+    },
+    {
+      "epoch": 1.7344,
+      "grad_norm": 2.2830695644969214,
+      "learning_rate": 7.995230009657508e-05,
+      "loss": 1.2874,
+      "step": 10840
+    },
+    {
+      "epoch": 1.73456,
+      "grad_norm": 2.1947288145362527,
+      "learning_rate": 7.993537724556737e-05,
+      "loss": 1.313,
+      "step": 10841
+    },
+    {
+      "epoch": 1.73472,
+      "grad_norm": 2.2635007801511304,
+      "learning_rate": 7.991845499325874e-05,
+      "loss": 1.2816,
+      "step": 10842
+    },
+    {
+      "epoch": 1.73488,
+      "grad_norm": 1.9773467429121512,
+      "learning_rate": 7.990153334015417e-05,
+      "loss": 1.2584,
+      "step": 10843
+    },
+    {
+      "epoch": 1.7350400000000001,
+      "grad_norm": 1.8566156677705687,
+      "learning_rate": 7.988461228675854e-05,
+      "loss": 1.2846,
+      "step": 10844
+    },
+    {
+      "epoch": 1.7351999999999999,
+      "grad_norm": 2.31395020392754,
+      "learning_rate": 7.986769183357678e-05,
+      "loss": 1.2476,
+      "step": 10845
+    },
+    {
+      "epoch": 1.73536,
+      "grad_norm": 2.0621110622115837,
+      "learning_rate": 7.985077198111375e-05,
+      "loss": 1.2664,
+      "step": 10846
+    },
+    {
+      "epoch": 1.73552,
+      "grad_norm": 1.9662131809762222,
+      "learning_rate": 7.983385272987436e-05,
+      "loss": 1.3012,
+      "step": 10847
+    },
+    {
+      "epoch": 1.73568,
+      "grad_norm": 1.9883155471345797,
+      "learning_rate": 7.981693408036337e-05,
+      "loss": 1.2768,
+      "step": 10848
+    },
+    {
+      "epoch": 1.73584,
+      "grad_norm": 1.8653452347532833,
+      "learning_rate": 7.980001603308568e-05,
+      "loss": 1.3588,
+      "step": 10849
+    },
+    {
+      "epoch": 1.736,
+      "grad_norm": 2.108798088410331,
+      "learning_rate": 7.978309858854606e-05,
+      "loss": 1.2984,
+      "step": 10850
+    },
+    {
+      "epoch": 1.73616,
+      "grad_norm": 1.8153261485541363,
+      "learning_rate": 7.976618174724935e-05,
+      "loss": 1.2219,
+      "step": 10851
+    },
+    {
+      "epoch": 1.73632,
+      "grad_norm": 2.27677049714505,
+      "learning_rate": 7.974926550970026e-05,
+      "loss": 1.2948,
+      "step": 10852
+    },
+    {
+      "epoch": 1.73648,
+      "grad_norm": 2.2230839561604574,
+      "learning_rate": 7.973234987640362e-05,
+      "loss": 1.2838,
+      "step": 10853
+    },
+    {
+      "epoch": 1.73664,
+      "grad_norm": 2.524872862816278,
+      "learning_rate": 7.971543484786408e-05,
+      "loss": 1.2836,
+      "step": 10854
+    },
+    {
+      "epoch": 1.7368000000000001,
+      "grad_norm": 1.666550612929415,
+      "learning_rate": 7.969852042458645e-05,
+      "loss": 1.209,
+      "step": 10855
+    },
+    {
+      "epoch": 1.7369599999999998,
+      "grad_norm": 1.8740997411240727,
+      "learning_rate": 7.968160660707536e-05,
+      "loss": 1.2268,
+      "step": 10856
+    },
+    {
+      "epoch": 1.73712,
+      "grad_norm": 1.955076154755863,
+      "learning_rate": 7.966469339583552e-05,
+      "loss": 1.2329,
+      "step": 10857
+    },
+    {
+      "epoch": 1.7372800000000002,
+      "grad_norm": 2.1472608061081555,
+      "learning_rate": 7.964778079137161e-05,
+      "loss": 1.3022,
+      "step": 10858
+    },
+    {
+      "epoch": 1.7374399999999999,
+      "grad_norm": 2.108267420446109,
+      "learning_rate": 7.963086879418827e-05,
+      "loss": 1.3043,
+      "step": 10859
+    },
+    {
+      "epoch": 1.7376,
+      "grad_norm": 2.1562902216651993,
+      "learning_rate": 7.96139574047901e-05,
+      "loss": 1.2256,
+      "step": 10860
+    },
+    {
+      "epoch": 1.73776,
+      "grad_norm": 2.2010923823484063,
+      "learning_rate": 7.959704662368177e-05,
+      "loss": 1.269,
+      "step": 10861
+    },
+    {
+      "epoch": 1.73792,
+      "grad_norm": 1.6672980645548299,
+      "learning_rate": 7.958013645136779e-05,
+      "loss": 1.2426,
+      "step": 10862
+    },
+    {
+      "epoch": 1.73808,
+      "grad_norm": 1.9145550797190325,
+      "learning_rate": 7.956322688835284e-05,
+      "loss": 1.2749,
+      "step": 10863
+    },
+    {
+      "epoch": 1.73824,
+      "grad_norm": 1.8649787048550748,
+      "learning_rate": 7.954631793514138e-05,
+      "loss": 1.2635,
+      "step": 10864
+    },
+    {
+      "epoch": 1.7384,
+      "grad_norm": 1.9877023318476033,
+      "learning_rate": 7.9529409592238e-05,
+      "loss": 1.3494,
+      "step": 10865
+    },
+    {
+      "epoch": 1.73856,
+      "grad_norm": 2.059741878355681,
+      "learning_rate": 7.951250186014721e-05,
+      "loss": 1.2981,
+      "step": 10866
+    },
+    {
+      "epoch": 1.73872,
+      "grad_norm": 1.890454841347806,
+      "learning_rate": 7.949559473937351e-05,
+      "loss": 1.2768,
+      "step": 10867
+    },
+    {
+      "epoch": 1.73888,
+      "grad_norm": 1.9192392428373934,
+      "learning_rate": 7.947868823042137e-05,
+      "loss": 1.2677,
+      "step": 10868
+    },
+    {
+      "epoch": 1.7390400000000001,
+      "grad_norm": 1.956005176615943,
+      "learning_rate": 7.946178233379531e-05,
+      "loss": 1.3495,
+      "step": 10869
+    },
+    {
+      "epoch": 1.7391999999999999,
+      "grad_norm": 1.9911978027883162,
+      "learning_rate": 7.94448770499997e-05,
+      "loss": 1.2699,
+      "step": 10870
+    },
+    {
+      "epoch": 1.73936,
+      "grad_norm": 2.092391478080497,
+      "learning_rate": 7.9427972379539e-05,
+      "loss": 1.3262,
+      "step": 10871
+    },
+    {
+      "epoch": 1.73952,
+      "grad_norm": 1.9330997101408194,
+      "learning_rate": 7.941106832291767e-05,
+      "loss": 1.2796,
+      "step": 10872
+    },
+    {
+      "epoch": 1.73968,
+      "grad_norm": 2.0045793011037643,
+      "learning_rate": 7.939416488064002e-05,
+      "loss": 1.2606,
+      "step": 10873
+    },
+    {
+      "epoch": 1.73984,
+      "grad_norm": 1.9610556524004892,
+      "learning_rate": 7.937726205321051e-05,
+      "loss": 1.2564,
+      "step": 10874
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 1.8647837000466239,
+      "learning_rate": 7.93603598411334e-05,
+      "loss": 1.2849,
+      "step": 10875
+    },
+    {
+      "epoch": 1.74016,
+      "grad_norm": 2.1186990203900566,
+      "learning_rate": 7.93434582449131e-05,
+      "loss": 1.3164,
+      "step": 10876
+    },
+    {
+      "epoch": 1.74032,
+      "grad_norm": 2.1481949213832774,
+      "learning_rate": 7.932655726505389e-05,
+      "loss": 1.3109,
+      "step": 10877
+    },
+    {
+      "epoch": 1.74048,
+      "grad_norm": 1.7453132929570419,
+      "learning_rate": 7.930965690206014e-05,
+      "loss": 1.3318,
+      "step": 10878
+    },
+    {
+      "epoch": 1.74064,
+      "grad_norm": 1.7696310435848952,
+      "learning_rate": 7.929275715643602e-05,
+      "loss": 1.2915,
+      "step": 10879
+    },
+    {
+      "epoch": 1.7408000000000001,
+      "grad_norm": 1.9012827289191965,
+      "learning_rate": 7.927585802868589e-05,
+      "loss": 1.2676,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7409599999999998,
+      "grad_norm": 1.9361180349901592,
+      "learning_rate": 7.925895951931393e-05,
+      "loss": 1.2616,
+      "step": 10881
+    },
+    {
+      "epoch": 1.74112,
+      "grad_norm": 1.9894037244543976,
+      "learning_rate": 7.924206162882441e-05,
+      "loss": 1.298,
+      "step": 10882
+    },
+    {
+      "epoch": 1.7412800000000002,
+      "grad_norm": 2.3113113111472514,
+      "learning_rate": 7.922516435772153e-05,
+      "loss": 1.2564,
+      "step": 10883
+    },
+    {
+      "epoch": 1.7414399999999999,
+      "grad_norm": 1.9172235674969598,
+      "learning_rate": 7.920826770650947e-05,
+      "loss": 1.2665,
+      "step": 10884
+    },
+    {
+      "epoch": 1.7416,
+      "grad_norm": 1.8524358833660037,
+      "learning_rate": 7.919137167569237e-05,
+      "loss": 1.2843,
+      "step": 10885
+    },
+    {
+      "epoch": 1.74176,
+      "grad_norm": 2.2000614388222863,
+      "learning_rate": 7.917447626577448e-05,
+      "loss": 1.2806,
+      "step": 10886
+    },
+    {
+      "epoch": 1.74192,
+      "grad_norm": 2.2388085555593062,
+      "learning_rate": 7.915758147725982e-05,
+      "loss": 1.3061,
+      "step": 10887
+    },
+    {
+      "epoch": 1.74208,
+      "grad_norm": 1.7742755333762354,
+      "learning_rate": 7.91406873106526e-05,
+      "loss": 1.3261,
+      "step": 10888
+    },
+    {
+      "epoch": 1.74224,
+      "grad_norm": 2.0261852792804182,
+      "learning_rate": 7.912379376645683e-05,
+      "loss": 1.2445,
+      "step": 10889
+    },
+    {
+      "epoch": 1.7424,
+      "grad_norm": 1.928032876709938,
+      "learning_rate": 7.910690084517666e-05,
+      "loss": 1.2674,
+      "step": 10890
+    },
+    {
+      "epoch": 1.74256,
+      "grad_norm": 2.4851290119549456,
+      "learning_rate": 7.909000854731612e-05,
+      "loss": 1.2247,
+      "step": 10891
+    },
+    {
+      "epoch": 1.74272,
+      "grad_norm": 1.6822427917665015,
+      "learning_rate": 7.907311687337925e-05,
+      "loss": 1.2737,
+      "step": 10892
+    },
+    {
+      "epoch": 1.74288,
+      "grad_norm": 2.402274979332955,
+      "learning_rate": 7.905622582387008e-05,
+      "loss": 1.3012,
+      "step": 10893
+    },
+    {
+      "epoch": 1.7430400000000001,
+      "grad_norm": 2.296040045212306,
+      "learning_rate": 7.903933539929265e-05,
+      "loss": 1.2612,
+      "step": 10894
+    },
+    {
+      "epoch": 1.7431999999999999,
+      "grad_norm": 2.4971222059277274,
+      "learning_rate": 7.902244560015086e-05,
+      "loss": 1.2341,
+      "step": 10895
+    },
+    {
+      "epoch": 1.74336,
+      "grad_norm": 2.5193289397857663,
+      "learning_rate": 7.900555642694877e-05,
+      "loss": 1.2343,
+      "step": 10896
+    },
+    {
+      "epoch": 1.74352,
+      "grad_norm": 2.028968952332189,
+      "learning_rate": 7.898866788019024e-05,
+      "loss": 1.2915,
+      "step": 10897
+    },
+    {
+      "epoch": 1.74368,
+      "grad_norm": 2.064357515941974,
+      "learning_rate": 7.897177996037925e-05,
+      "loss": 1.3193,
+      "step": 10898
+    },
+    {
+      "epoch": 1.74384,
+      "grad_norm": 2.3173028715216835,
+      "learning_rate": 7.895489266801973e-05,
+      "loss": 1.2939,
+      "step": 10899
+    },
+    {
+      "epoch": 1.744,
+      "grad_norm": 1.9082880530866375,
+      "learning_rate": 7.893800600361554e-05,
+      "loss": 1.2366,
+      "step": 10900
+    },
+    {
+      "epoch": 1.74416,
+      "grad_norm": 2.0354051671127094,
+      "learning_rate": 7.892111996767053e-05,
+      "loss": 1.2798,
+      "step": 10901
+    },
+    {
+      "epoch": 1.74432,
+      "grad_norm": 2.197856474021588,
+      "learning_rate": 7.890423456068862e-05,
+      "loss": 1.3238,
+      "step": 10902
+    },
+    {
+      "epoch": 1.74448,
+      "grad_norm": 2.0161884781949233,
+      "learning_rate": 7.88873497831736e-05,
+      "loss": 1.3437,
+      "step": 10903
+    },
+    {
+      "epoch": 1.74464,
+      "grad_norm": 1.8470076464976763,
+      "learning_rate": 7.887046563562932e-05,
+      "loss": 1.2835,
+      "step": 10904
+    },
+    {
+      "epoch": 1.7448000000000001,
+      "grad_norm": 1.977413003805217,
+      "learning_rate": 7.88535821185595e-05,
+      "loss": 1.2774,
+      "step": 10905
+    },
+    {
+      "epoch": 1.7449599999999998,
+      "grad_norm": 1.8397818476170233,
+      "learning_rate": 7.883669923246803e-05,
+      "loss": 1.2635,
+      "step": 10906
+    },
+    {
+      "epoch": 1.74512,
+      "grad_norm": 2.0740948289573677,
+      "learning_rate": 7.881981697785859e-05,
+      "loss": 1.275,
+      "step": 10907
+    },
+    {
+      "epoch": 1.7452800000000002,
+      "grad_norm": 2.0856377918122284,
+      "learning_rate": 7.880293535523497e-05,
+      "loss": 1.2381,
+      "step": 10908
+    },
+    {
+      "epoch": 1.7454399999999999,
+      "grad_norm": 2.0078091996799894,
+      "learning_rate": 7.878605436510083e-05,
+      "loss": 1.2983,
+      "step": 10909
+    },
+    {
+      "epoch": 1.7456,
+      "grad_norm": 1.9337373200297125,
+      "learning_rate": 7.876917400795997e-05,
+      "loss": 1.2747,
+      "step": 10910
+    },
+    {
+      "epoch": 1.74576,
+      "grad_norm": 2.0404273565268682,
+      "learning_rate": 7.875229428431597e-05,
+      "loss": 1.2248,
+      "step": 10911
+    },
+    {
+      "epoch": 1.74592,
+      "grad_norm": 1.941211040063585,
+      "learning_rate": 7.87354151946726e-05,
+      "loss": 1.3488,
+      "step": 10912
+    },
+    {
+      "epoch": 1.74608,
+      "grad_norm": 2.4791977492479607,
+      "learning_rate": 7.87185367395334e-05,
+      "loss": 1.2355,
+      "step": 10913
+    },
+    {
+      "epoch": 1.74624,
+      "grad_norm": 1.9028629210720145,
+      "learning_rate": 7.870165891940209e-05,
+      "loss": 1.2478,
+      "step": 10914
+    },
+    {
+      "epoch": 1.7464,
+      "grad_norm": 1.9030261246949949,
+      "learning_rate": 7.868478173478223e-05,
+      "loss": 1.2883,
+      "step": 10915
+    },
+    {
+      "epoch": 1.7465600000000001,
+      "grad_norm": 2.4888705244944145,
+      "learning_rate": 7.866790518617742e-05,
+      "loss": 1.2583,
+      "step": 10916
+    },
+    {
+      "epoch": 1.74672,
+      "grad_norm": 1.9146669732928674,
+      "learning_rate": 7.865102927409122e-05,
+      "loss": 1.2325,
+      "step": 10917
+    },
+    {
+      "epoch": 1.74688,
+      "grad_norm": 2.303011029947197,
+      "learning_rate": 7.863415399902725e-05,
+      "loss": 1.2711,
+      "step": 10918
+    },
+    {
+      "epoch": 1.7470400000000001,
+      "grad_norm": 1.8556722736653133,
+      "learning_rate": 7.861727936148895e-05,
+      "loss": 1.2352,
+      "step": 10919
+    },
+    {
+      "epoch": 1.7471999999999999,
+      "grad_norm": 1.8391588031458928,
+      "learning_rate": 7.86004053619799e-05,
+      "loss": 1.4038,
+      "step": 10920
+    },
+    {
+      "epoch": 1.74736,
+      "grad_norm": 2.6273346851175057,
+      "learning_rate": 7.858353200100354e-05,
+      "loss": 1.2634,
+      "step": 10921
+    },
+    {
+      "epoch": 1.74752,
+      "grad_norm": 2.202501794027214,
+      "learning_rate": 7.856665927906341e-05,
+      "loss": 1.2937,
+      "step": 10922
+    },
+    {
+      "epoch": 1.74768,
+      "grad_norm": 1.9948542580763164,
+      "learning_rate": 7.854978719666291e-05,
+      "loss": 1.2875,
+      "step": 10923
+    },
+    {
+      "epoch": 1.74784,
+      "grad_norm": 2.7690420057002134,
+      "learning_rate": 7.853291575430553e-05,
+      "loss": 1.3011,
+      "step": 10924
+    },
+    {
+      "epoch": 1.748,
+      "grad_norm": 2.3006572658991122,
+      "learning_rate": 7.851604495249463e-05,
+      "loss": 1.2634,
+      "step": 10925
+    },
+    {
+      "epoch": 1.74816,
+      "grad_norm": 2.0531442483863667,
+      "learning_rate": 7.849917479173368e-05,
+      "loss": 1.2227,
+      "step": 10926
+    },
+    {
+      "epoch": 1.74832,
+      "grad_norm": 2.2209190529562686,
+      "learning_rate": 7.848230527252598e-05,
+      "loss": 1.3657,
+      "step": 10927
+    },
+    {
+      "epoch": 1.74848,
+      "grad_norm": 2.6851749972012113,
+      "learning_rate": 7.846543639537498e-05,
+      "loss": 1.2338,
+      "step": 10928
+    },
+    {
+      "epoch": 1.74864,
+      "grad_norm": 2.0949993834806797,
+      "learning_rate": 7.844856816078394e-05,
+      "loss": 1.3414,
+      "step": 10929
+    },
+    {
+      "epoch": 1.7488000000000001,
+      "grad_norm": 1.6324892803242794,
+      "learning_rate": 7.843170056925623e-05,
+      "loss": 1.2623,
+      "step": 10930
+    },
+    {
+      "epoch": 1.7489599999999998,
+      "grad_norm": 2.0683944964432563,
+      "learning_rate": 7.841483362129517e-05,
+      "loss": 1.2909,
+      "step": 10931
+    },
+    {
+      "epoch": 1.74912,
+      "grad_norm": 2.406307585930877,
+      "learning_rate": 7.839796731740401e-05,
+      "loss": 1.2388,
+      "step": 10932
+    },
+    {
+      "epoch": 1.74928,
+      "grad_norm": 2.089631073433548,
+      "learning_rate": 7.838110165808604e-05,
+      "loss": 1.2645,
+      "step": 10933
+    },
+    {
+      "epoch": 1.7494399999999999,
+      "grad_norm": 2.7279039612548126,
+      "learning_rate": 7.836423664384445e-05,
+      "loss": 1.2698,
+      "step": 10934
+    },
+    {
+      "epoch": 1.7496,
+      "grad_norm": 2.5452297187284043,
+      "learning_rate": 7.834737227518258e-05,
+      "loss": 1.3556,
+      "step": 10935
+    },
+    {
+      "epoch": 1.74976,
+      "grad_norm": 2.163736396627657,
+      "learning_rate": 7.833050855260352e-05,
+      "loss": 1.2862,
+      "step": 10936
+    },
+    {
+      "epoch": 1.74992,
+      "grad_norm": 2.30128363412988,
+      "learning_rate": 7.831364547661054e-05,
+      "loss": 1.2024,
+      "step": 10937
+    },
+    {
+      "epoch": 1.75008,
+      "grad_norm": 2.7030742348632133,
+      "learning_rate": 7.829678304770674e-05,
+      "loss": 1.2977,
+      "step": 10938
+    },
+    {
+      "epoch": 1.75024,
+      "grad_norm": 2.4787655178041086,
+      "learning_rate": 7.827992126639535e-05,
+      "loss": 1.2691,
+      "step": 10939
+    },
+    {
+      "epoch": 1.7504,
+      "grad_norm": 2.4000997296445834,
+      "learning_rate": 7.826306013317942e-05,
+      "loss": 1.2426,
+      "step": 10940
+    },
+    {
+      "epoch": 1.7505600000000001,
+      "grad_norm": 2.1474888313710134,
+      "learning_rate": 7.824619964856214e-05,
+      "loss": 1.2711,
+      "step": 10941
+    },
+    {
+      "epoch": 1.75072,
+      "grad_norm": 2.5060123103347904,
+      "learning_rate": 7.822933981304652e-05,
+      "loss": 1.2776,
+      "step": 10942
+    },
+    {
+      "epoch": 1.75088,
+      "grad_norm": 1.930308801160628,
+      "learning_rate": 7.821248062713574e-05,
+      "loss": 1.2416,
+      "step": 10943
+    },
+    {
+      "epoch": 1.7510400000000002,
+      "grad_norm": 2.4920571878478195,
+      "learning_rate": 7.819562209133272e-05,
+      "loss": 1.3211,
+      "step": 10944
+    },
+    {
+      "epoch": 1.7511999999999999,
+      "grad_norm": 2.119632075825193,
+      "learning_rate": 7.817876420614063e-05,
+      "loss": 1.2563,
+      "step": 10945
+    },
+    {
+      "epoch": 1.75136,
+      "grad_norm": 2.1440410622896344,
+      "learning_rate": 7.816190697206235e-05,
+      "loss": 1.2807,
+      "step": 10946
+    },
+    {
+      "epoch": 1.75152,
+      "grad_norm": 2.0386591319927727,
+      "learning_rate": 7.814505038960101e-05,
+      "loss": 1.2209,
+      "step": 10947
+    },
+    {
+      "epoch": 1.75168,
+      "grad_norm": 2.429438990807633,
+      "learning_rate": 7.812819445925948e-05,
+      "loss": 1.2784,
+      "step": 10948
+    },
+    {
+      "epoch": 1.75184,
+      "grad_norm": 1.821464052307826,
+      "learning_rate": 7.811133918154078e-05,
+      "loss": 1.2353,
+      "step": 10949
+    },
+    {
+      "epoch": 1.752,
+      "grad_norm": 1.9744212856855996,
+      "learning_rate": 7.809448455694779e-05,
+      "loss": 1.2007,
+      "step": 10950
+    },
+    {
+      "epoch": 1.75216,
+      "grad_norm": 2.131452959243337,
+      "learning_rate": 7.807763058598351e-05,
+      "loss": 1.3209,
+      "step": 10951
+    },
+    {
+      "epoch": 1.75232,
+      "grad_norm": 2.068797602844522,
+      "learning_rate": 7.806077726915074e-05,
+      "loss": 1.276,
+      "step": 10952
+    },
+    {
+      "epoch": 1.75248,
+      "grad_norm": 2.2690489191891956,
+      "learning_rate": 7.804392460695246e-05,
+      "loss": 1.3271,
+      "step": 10953
+    },
+    {
+      "epoch": 1.75264,
+      "grad_norm": 1.9890983789055163,
+      "learning_rate": 7.802707259989142e-05,
+      "loss": 1.2657,
+      "step": 10954
+    },
+    {
+      "epoch": 1.7528000000000001,
+      "grad_norm": 2.1258362549840952,
+      "learning_rate": 7.801022124847057e-05,
+      "loss": 1.2976,
+      "step": 10955
+    },
+    {
+      "epoch": 1.7529599999999999,
+      "grad_norm": 1.9598592302187596,
+      "learning_rate": 7.799337055319262e-05,
+      "loss": 1.3488,
+      "step": 10956
+    },
+    {
+      "epoch": 1.75312,
+      "grad_norm": 1.9264585291301595,
+      "learning_rate": 7.797652051456046e-05,
+      "loss": 1.2065,
+      "step": 10957
+    },
+    {
+      "epoch": 1.75328,
+      "grad_norm": 1.8184822352529217,
+      "learning_rate": 7.795967113307682e-05,
+      "loss": 1.2789,
+      "step": 10958
+    },
+    {
+      "epoch": 1.7534399999999999,
+      "grad_norm": 2.0230064662446736,
+      "learning_rate": 7.794282240924451e-05,
+      "loss": 1.2341,
+      "step": 10959
+    },
+    {
+      "epoch": 1.7536,
+      "grad_norm": 1.6378713410219103,
+      "learning_rate": 7.79259743435662e-05,
+      "loss": 1.2279,
+      "step": 10960
+    },
+    {
+      "epoch": 1.75376,
+      "grad_norm": 2.036725605940992,
+      "learning_rate": 7.790912693654468e-05,
+      "loss": 1.2399,
+      "step": 10961
+    },
+    {
+      "epoch": 1.75392,
+      "grad_norm": 2.010331306226739,
+      "learning_rate": 7.78922801886826e-05,
+      "loss": 1.3782,
+      "step": 10962
+    },
+    {
+      "epoch": 1.75408,
+      "grad_norm": 2.180729278975877,
+      "learning_rate": 7.787543410048269e-05,
+      "loss": 1.2367,
+      "step": 10963
+    },
+    {
+      "epoch": 1.75424,
+      "grad_norm": 2.1921938564185717,
+      "learning_rate": 7.785858867244755e-05,
+      "loss": 1.3993,
+      "step": 10964
+    },
+    {
+      "epoch": 1.7544,
+      "grad_norm": 2.5425712943040275,
+      "learning_rate": 7.784174390507989e-05,
+      "loss": 1.2938,
+      "step": 10965
+    },
+    {
+      "epoch": 1.7545600000000001,
+      "grad_norm": 2.465152799683136,
+      "learning_rate": 7.782489979888227e-05,
+      "loss": 1.1895,
+      "step": 10966
+    },
+    {
+      "epoch": 1.7547199999999998,
+      "grad_norm": 1.9833613055423924,
+      "learning_rate": 7.780805635435738e-05,
+      "loss": 1.2659,
+      "step": 10967
+    },
+    {
+      "epoch": 1.75488,
+      "grad_norm": 2.4417368156813697,
+      "learning_rate": 7.779121357200771e-05,
+      "loss": 1.3254,
+      "step": 10968
+    },
+    {
+      "epoch": 1.7550400000000002,
+      "grad_norm": 1.831107945295511,
+      "learning_rate": 7.777437145233591e-05,
+      "loss": 1.2286,
+      "step": 10969
+    },
+    {
+      "epoch": 1.7551999999999999,
+      "grad_norm": 1.576730103185691,
+      "learning_rate": 7.775752999584442e-05,
+      "loss": 1.3175,
+      "step": 10970
+    },
+    {
+      "epoch": 1.75536,
+      "grad_norm": 2.045303412267316,
+      "learning_rate": 7.774068920303588e-05,
+      "loss": 1.3177,
+      "step": 10971
+    },
+    {
+      "epoch": 1.75552,
+      "grad_norm": 2.1956423793409656,
+      "learning_rate": 7.77238490744127e-05,
+      "loss": 1.2612,
+      "step": 10972
+    },
+    {
+      "epoch": 1.75568,
+      "grad_norm": 1.788410183069408,
+      "learning_rate": 7.77070096104774e-05,
+      "loss": 1.2495,
+      "step": 10973
+    },
+    {
+      "epoch": 1.75584,
+      "grad_norm": 2.023629067818476,
+      "learning_rate": 7.769017081173246e-05,
+      "loss": 1.2387,
+      "step": 10974
+    },
+    {
+      "epoch": 1.756,
+      "grad_norm": 1.7919170745561164,
+      "learning_rate": 7.767333267868034e-05,
+      "loss": 1.2736,
+      "step": 10975
+    },
+    {
+      "epoch": 1.75616,
+      "grad_norm": 2.1786593695479084,
+      "learning_rate": 7.765649521182339e-05,
+      "loss": 1.2786,
+      "step": 10976
+    },
+    {
+      "epoch": 1.75632,
+      "grad_norm": 1.856006443242633,
+      "learning_rate": 7.763965841166412e-05,
+      "loss": 1.2466,
+      "step": 10977
+    },
+    {
+      "epoch": 1.75648,
+      "grad_norm": 1.8050311129004124,
+      "learning_rate": 7.762282227870483e-05,
+      "loss": 1.3201,
+      "step": 10978
+    },
+    {
+      "epoch": 1.75664,
+      "grad_norm": 1.874482348226392,
+      "learning_rate": 7.76059868134479e-05,
+      "loss": 1.2606,
+      "step": 10979
+    },
+    {
+      "epoch": 1.7568000000000001,
+      "grad_norm": 1.8074064322529781,
+      "learning_rate": 7.758915201639573e-05,
+      "loss": 1.2731,
+      "step": 10980
+    },
+    {
+      "epoch": 1.7569599999999999,
+      "grad_norm": 1.8391553229874122,
+      "learning_rate": 7.75723178880506e-05,
+      "loss": 1.234,
+      "step": 10981
+    },
+    {
+      "epoch": 1.75712,
+      "grad_norm": 2.383782090431216,
+      "learning_rate": 7.755548442891479e-05,
+      "loss": 1.3848,
+      "step": 10982
+    },
+    {
+      "epoch": 1.75728,
+      "grad_norm": 2.000778709397533,
+      "learning_rate": 7.753865163949066e-05,
+      "loss": 1.2334,
+      "step": 10983
+    },
+    {
+      "epoch": 1.75744,
+      "grad_norm": 2.2811344467078563,
+      "learning_rate": 7.75218195202804e-05,
+      "loss": 1.2175,
+      "step": 10984
+    },
+    {
+      "epoch": 1.7576,
+      "grad_norm": 2.5875063781970833,
+      "learning_rate": 7.750498807178633e-05,
+      "loss": 1.3128,
+      "step": 10985
+    },
+    {
+      "epoch": 1.75776,
+      "grad_norm": 2.0267983362825035,
+      "learning_rate": 7.74881572945106e-05,
+      "loss": 1.201,
+      "step": 10986
+    },
+    {
+      "epoch": 1.75792,
+      "grad_norm": 2.5339235019990225,
+      "learning_rate": 7.747132718895547e-05,
+      "loss": 1.2691,
+      "step": 10987
+    },
+    {
+      "epoch": 1.75808,
+      "grad_norm": 2.2412938998773355,
+      "learning_rate": 7.745449775562312e-05,
+      "loss": 1.305,
+      "step": 10988
+    },
+    {
+      "epoch": 1.75824,
+      "grad_norm": 1.9408272858719156,
+      "learning_rate": 7.743766899501567e-05,
+      "loss": 1.247,
+      "step": 10989
+    },
+    {
+      "epoch": 1.7584,
+      "grad_norm": 2.0178718754502123,
+      "learning_rate": 7.742084090763533e-05,
+      "loss": 1.2661,
+      "step": 10990
+    },
+    {
+      "epoch": 1.7585600000000001,
+      "grad_norm": 2.053126188756517,
+      "learning_rate": 7.740401349398416e-05,
+      "loss": 1.2178,
+      "step": 10991
+    },
+    {
+      "epoch": 1.7587199999999998,
+      "grad_norm": 1.905372170066427,
+      "learning_rate": 7.738718675456436e-05,
+      "loss": 1.2972,
+      "step": 10992
+    },
+    {
+      "epoch": 1.75888,
+      "grad_norm": 2.5845929449592635,
+      "learning_rate": 7.737036068987792e-05,
+      "loss": 1.3523,
+      "step": 10993
+    },
+    {
+      "epoch": 1.7590400000000002,
+      "grad_norm": 2.1569798835994205,
+      "learning_rate": 7.735353530042697e-05,
+      "loss": 1.3101,
+      "step": 10994
+    },
+    {
+      "epoch": 1.7591999999999999,
+      "grad_norm": 1.7971287469028348,
+      "learning_rate": 7.733671058671349e-05,
+      "loss": 1.2071,
+      "step": 10995
+    },
+    {
+      "epoch": 1.75936,
+      "grad_norm": 2.057642129219373,
+      "learning_rate": 7.731988654923959e-05,
+      "loss": 1.2976,
+      "step": 10996
+    },
+    {
+      "epoch": 1.75952,
+      "grad_norm": 2.15297246666982,
+      "learning_rate": 7.730306318850718e-05,
+      "loss": 1.3522,
+      "step": 10997
+    },
+    {
+      "epoch": 1.75968,
+      "grad_norm": 1.9829274609395866,
+      "learning_rate": 7.72862405050183e-05,
+      "loss": 1.2426,
+      "step": 10998
+    },
+    {
+      "epoch": 1.75984,
+      "grad_norm": 2.7070332077280415,
+      "learning_rate": 7.726941849927489e-05,
+      "loss": 1.2874,
+      "step": 10999
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 2.0262525520875445,
+      "learning_rate": 7.725259717177897e-05,
+      "loss": 1.3124,
+      "step": 11000
+    },
+    {
+      "epoch": 1.76016,
+      "grad_norm": 2.550549437579223,
+      "learning_rate": 7.723577652303235e-05,
+      "loss": 1.2165,
+      "step": 11001
+    },
+    {
+      "epoch": 1.76032,
+      "grad_norm": 2.340137552364698,
+      "learning_rate": 7.721895655353703e-05,
+      "loss": 1.2626,
+      "step": 11002
+    },
+    {
+      "epoch": 1.76048,
+      "grad_norm": 1.9230704455041538,
+      "learning_rate": 7.720213726379481e-05,
+      "loss": 1.2172,
+      "step": 11003
+    },
+    {
+      "epoch": 1.76064,
+      "grad_norm": 2.6484078895568435,
+      "learning_rate": 7.718531865430763e-05,
+      "loss": 1.3403,
+      "step": 11004
+    },
+    {
+      "epoch": 1.7608000000000001,
+      "grad_norm": 2.7311891749526787,
+      "learning_rate": 7.716850072557727e-05,
+      "loss": 1.2431,
+      "step": 11005
+    },
+    {
+      "epoch": 1.7609599999999999,
+      "grad_norm": 2.0087547593645922,
+      "learning_rate": 7.715168347810558e-05,
+      "loss": 1.2685,
+      "step": 11006
+    },
+    {
+      "epoch": 1.76112,
+      "grad_norm": 1.9494952623442066,
+      "learning_rate": 7.713486691239435e-05,
+      "loss": 1.2843,
+      "step": 11007
+    },
+    {
+      "epoch": 1.76128,
+      "grad_norm": 2.3196718556179086,
+      "learning_rate": 7.711805102894541e-05,
+      "loss": 1.1815,
+      "step": 11008
+    },
+    {
+      "epoch": 1.76144,
+      "grad_norm": 3.0023761352459863,
+      "learning_rate": 7.710123582826045e-05,
+      "loss": 1.2917,
+      "step": 11009
+    },
+    {
+      "epoch": 1.7616,
+      "grad_norm": 1.9435415524083175,
+      "learning_rate": 7.70844213108413e-05,
+      "loss": 1.2655,
+      "step": 11010
+    },
+    {
+      "epoch": 1.76176,
+      "grad_norm": 2.6849688763131514,
+      "learning_rate": 7.706760747718956e-05,
+      "loss": 1.3859,
+      "step": 11011
+    },
+    {
+      "epoch": 1.76192,
+      "grad_norm": 2.7500362889669328,
+      "learning_rate": 7.705079432780705e-05,
+      "loss": 1.2779,
+      "step": 11012
+    },
+    {
+      "epoch": 1.76208,
+      "grad_norm": 2.1021793548722516,
+      "learning_rate": 7.703398186319536e-05,
+      "loss": 1.2508,
+      "step": 11013
+    },
+    {
+      "epoch": 1.76224,
+      "grad_norm": 2.232653673877257,
+      "learning_rate": 7.701717008385619e-05,
+      "loss": 1.3315,
+      "step": 11014
+    },
+    {
+      "epoch": 1.7624,
+      "grad_norm": 2.934936250361298,
+      "learning_rate": 7.700035899029116e-05,
+      "loss": 1.2735,
+      "step": 11015
+    },
+    {
+      "epoch": 1.7625600000000001,
+      "grad_norm": 2.629546528950467,
+      "learning_rate": 7.698354858300196e-05,
+      "loss": 1.2136,
+      "step": 11016
+    },
+    {
+      "epoch": 1.7627199999999998,
+      "grad_norm": 1.7620455264266224,
+      "learning_rate": 7.696673886249008e-05,
+      "loss": 1.218,
+      "step": 11017
+    },
+    {
+      "epoch": 1.76288,
+      "grad_norm": 2.721270222384936,
+      "learning_rate": 7.69499298292572e-05,
+      "loss": 1.2426,
+      "step": 11018
+    },
+    {
+      "epoch": 1.7630400000000002,
+      "grad_norm": 2.2086141022063086,
+      "learning_rate": 7.693312148380478e-05,
+      "loss": 1.2446,
+      "step": 11019
+    },
+    {
+      "epoch": 1.7631999999999999,
+      "grad_norm": 2.1324040378936977,
+      "learning_rate": 7.691631382663443e-05,
+      "loss": 1.2778,
+      "step": 11020
+    },
+    {
+      "epoch": 1.76336,
+      "grad_norm": 2.0703804197211273,
+      "learning_rate": 7.689950685824762e-05,
+      "loss": 1.2327,
+      "step": 11021
+    },
+    {
+      "epoch": 1.76352,
+      "grad_norm": 2.680276515398174,
+      "learning_rate": 7.688270057914587e-05,
+      "loss": 1.3083,
+      "step": 11022
+    },
+    {
+      "epoch": 1.76368,
+      "grad_norm": 2.521221144482408,
+      "learning_rate": 7.686589498983064e-05,
+      "loss": 1.2661,
+      "step": 11023
+    },
+    {
+      "epoch": 1.76384,
+      "grad_norm": 2.2447111889734175,
+      "learning_rate": 7.684909009080342e-05,
+      "loss": 1.2631,
+      "step": 11024
+    },
+    {
+      "epoch": 1.764,
+      "grad_norm": 2.8639524252755106,
+      "learning_rate": 7.683228588256558e-05,
+      "loss": 1.3112,
+      "step": 11025
+    },
+    {
+      "epoch": 1.76416,
+      "grad_norm": 2.466065457475747,
+      "learning_rate": 7.681548236561861e-05,
+      "loss": 1.3259,
+      "step": 11026
+    },
+    {
+      "epoch": 1.76432,
+      "grad_norm": 1.8577872108742153,
+      "learning_rate": 7.679867954046384e-05,
+      "loss": 1.2619,
+      "step": 11027
+    },
+    {
+      "epoch": 1.76448,
+      "grad_norm": 2.792536147010362,
+      "learning_rate": 7.678187740760269e-05,
+      "loss": 1.2253,
+      "step": 11028
+    },
+    {
+      "epoch": 1.76464,
+      "grad_norm": 2.7318084463882624,
+      "learning_rate": 7.676507596753645e-05,
+      "loss": 1.3457,
+      "step": 11029
+    },
+    {
+      "epoch": 1.7648000000000001,
+      "grad_norm": 2.3344072754222265,
+      "learning_rate": 7.67482752207665e-05,
+      "loss": 1.3428,
+      "step": 11030
+    },
+    {
+      "epoch": 1.7649599999999999,
+      "grad_norm": 2.0808460032102425,
+      "learning_rate": 7.673147516779412e-05,
+      "loss": 1.2355,
+      "step": 11031
+    },
+    {
+      "epoch": 1.76512,
+      "grad_norm": 1.9862560290165876,
+      "learning_rate": 7.671467580912066e-05,
+      "loss": 1.3045,
+      "step": 11032
+    },
+    {
+      "epoch": 1.76528,
+      "grad_norm": 2.030255677904479,
+      "learning_rate": 7.669787714524729e-05,
+      "loss": 1.24,
+      "step": 11033
+    },
+    {
+      "epoch": 1.76544,
+      "grad_norm": 1.926746527911234,
+      "learning_rate": 7.668107917667537e-05,
+      "loss": 1.2524,
+      "step": 11034
+    },
+    {
+      "epoch": 1.7656,
+      "grad_norm": 2.0321669861613127,
+      "learning_rate": 7.6664281903906e-05,
+      "loss": 1.257,
+      "step": 11035
+    },
+    {
+      "epoch": 1.76576,
+      "grad_norm": 1.924307541138958,
+      "learning_rate": 7.664748532744052e-05,
+      "loss": 1.2529,
+      "step": 11036
+    },
+    {
+      "epoch": 1.76592,
+      "grad_norm": 1.9500879629233203,
+      "learning_rate": 7.663068944778e-05,
+      "loss": 1.2697,
+      "step": 11037
+    },
+    {
+      "epoch": 1.76608,
+      "grad_norm": 1.9868622032271168,
+      "learning_rate": 7.661389426542569e-05,
+      "loss": 1.2456,
+      "step": 11038
+    },
+    {
+      "epoch": 1.76624,
+      "grad_norm": 2.1445720870848466,
+      "learning_rate": 7.659709978087865e-05,
+      "loss": 1.3069,
+      "step": 11039
+    },
+    {
+      "epoch": 1.7664,
+      "grad_norm": 1.931897819668494,
+      "learning_rate": 7.658030599464011e-05,
+      "loss": 1.312,
+      "step": 11040
+    },
+    {
+      "epoch": 1.7665600000000001,
+      "grad_norm": 1.9975491976843351,
+      "learning_rate": 7.656351290721107e-05,
+      "loss": 1.3204,
+      "step": 11041
+    },
+    {
+      "epoch": 1.7667199999999998,
+      "grad_norm": 2.08523651622216,
+      "learning_rate": 7.65467205190927e-05,
+      "loss": 1.2715,
+      "step": 11042
+    },
+    {
+      "epoch": 1.76688,
+      "grad_norm": 2.1282091253035977,
+      "learning_rate": 7.652992883078595e-05,
+      "loss": 1.2485,
+      "step": 11043
+    },
+    {
+      "epoch": 1.7670400000000002,
+      "grad_norm": 1.9079871030231599,
+      "learning_rate": 7.6513137842792e-05,
+      "loss": 1.2888,
+      "step": 11044
+    },
+    {
+      "epoch": 1.7671999999999999,
+      "grad_norm": 1.7989495814618053,
+      "learning_rate": 7.649634755561173e-05,
+      "loss": 1.2087,
+      "step": 11045
+    },
+    {
+      "epoch": 1.76736,
+      "grad_norm": 2.0513246294423353,
+      "learning_rate": 7.64795579697462e-05,
+      "loss": 1.2581,
+      "step": 11046
+    },
+    {
+      "epoch": 1.76752,
+      "grad_norm": 1.8758256179114365,
+      "learning_rate": 7.64627690856964e-05,
+      "loss": 1.2962,
+      "step": 11047
+    },
+    {
+      "epoch": 1.76768,
+      "grad_norm": 1.863164700038692,
+      "learning_rate": 7.644598090396328e-05,
+      "loss": 1.2741,
+      "step": 11048
+    },
+    {
+      "epoch": 1.76784,
+      "grad_norm": 1.5687136937987407,
+      "learning_rate": 7.642919342504778e-05,
+      "loss": 1.2817,
+      "step": 11049
+    },
+    {
+      "epoch": 1.768,
+      "grad_norm": 2.1064951067701867,
+      "learning_rate": 7.641240664945078e-05,
+      "loss": 1.3231,
+      "step": 11050
+    },
+    {
+      "epoch": 1.76816,
+      "grad_norm": 1.8358710528553013,
+      "learning_rate": 7.639562057767324e-05,
+      "loss": 1.2629,
+      "step": 11051
+    },
+    {
+      "epoch": 1.7683200000000001,
+      "grad_norm": 2.5238297882716307,
+      "learning_rate": 7.637883521021592e-05,
+      "loss": 1.2522,
+      "step": 11052
+    },
+    {
+      "epoch": 1.76848,
+      "grad_norm": 2.15815141302912,
+      "learning_rate": 7.63620505475798e-05,
+      "loss": 1.2469,
+      "step": 11053
+    },
+    {
+      "epoch": 1.76864,
+      "grad_norm": 2.1539827801592524,
+      "learning_rate": 7.63452665902656e-05,
+      "loss": 1.2881,
+      "step": 11054
+    },
+    {
+      "epoch": 1.7688000000000001,
+      "grad_norm": 2.564106227381595,
+      "learning_rate": 7.63284833387742e-05,
+      "loss": 1.2903,
+      "step": 11055
+    },
+    {
+      "epoch": 1.7689599999999999,
+      "grad_norm": 2.2964093895555955,
+      "learning_rate": 7.631170079360634e-05,
+      "loss": 1.2672,
+      "step": 11056
+    },
+    {
+      "epoch": 1.76912,
+      "grad_norm": 1.8567379924676208,
+      "learning_rate": 7.629491895526286e-05,
+      "loss": 1.2326,
+      "step": 11057
+    },
+    {
+      "epoch": 1.76928,
+      "grad_norm": 2.427540662665745,
+      "learning_rate": 7.62781378242444e-05,
+      "loss": 1.2927,
+      "step": 11058
+    },
+    {
+      "epoch": 1.76944,
+      "grad_norm": 3.0903812942799007,
+      "learning_rate": 7.62613574010518e-05,
+      "loss": 1.2916,
+      "step": 11059
+    },
+    {
+      "epoch": 1.7696,
+      "grad_norm": 1.9681044025937813,
+      "learning_rate": 7.624457768618564e-05,
+      "loss": 1.2447,
+      "step": 11060
+    },
+    {
+      "epoch": 1.76976,
+      "grad_norm": 2.2856668346619102,
+      "learning_rate": 7.622779868014673e-05,
+      "loss": 1.3322,
+      "step": 11061
+    },
+    {
+      "epoch": 1.76992,
+      "grad_norm": 2.4257053327227487,
+      "learning_rate": 7.621102038343561e-05,
+      "loss": 1.2559,
+      "step": 11062
+    },
+    {
+      "epoch": 1.77008,
+      "grad_norm": 1.9245970003237185,
+      "learning_rate": 7.619424279655301e-05,
+      "loss": 1.2701,
+      "step": 11063
+    },
+    {
+      "epoch": 1.77024,
+      "grad_norm": 2.0712940345358937,
+      "learning_rate": 7.617746591999951e-05,
+      "loss": 1.338,
+      "step": 11064
+    },
+    {
+      "epoch": 1.7704,
+      "grad_norm": 1.970669852102364,
+      "learning_rate": 7.616068975427573e-05,
+      "loss": 1.3227,
+      "step": 11065
+    },
+    {
+      "epoch": 1.7705600000000001,
+      "grad_norm": 1.6239234609193454,
+      "learning_rate": 7.614391429988219e-05,
+      "loss": 1.3108,
+      "step": 11066
+    },
+    {
+      "epoch": 1.7707199999999998,
+      "grad_norm": 1.9953136743243447,
+      "learning_rate": 7.612713955731954e-05,
+      "loss": 1.2587,
+      "step": 11067
+    },
+    {
+      "epoch": 1.77088,
+      "grad_norm": 1.8643585962968459,
+      "learning_rate": 7.61103655270882e-05,
+      "loss": 1.3121,
+      "step": 11068
+    },
+    {
+      "epoch": 1.77104,
+      "grad_norm": 2.3027639870063177,
+      "learning_rate": 7.609359220968879e-05,
+      "loss": 1.2534,
+      "step": 11069
+    },
+    {
+      "epoch": 1.7711999999999999,
+      "grad_norm": 1.8962828554224265,
+      "learning_rate": 7.607681960562172e-05,
+      "loss": 1.248,
+      "step": 11070
+    },
+    {
+      "epoch": 1.77136,
+      "grad_norm": 2.28911570641883,
+      "learning_rate": 7.606004771538751e-05,
+      "loss": 1.2647,
+      "step": 11071
+    },
+    {
+      "epoch": 1.77152,
+      "grad_norm": 1.8876721562695387,
+      "learning_rate": 7.604327653948657e-05,
+      "loss": 1.2897,
+      "step": 11072
+    },
+    {
+      "epoch": 1.77168,
+      "grad_norm": 1.9844154754566288,
+      "learning_rate": 7.602650607841939e-05,
+      "loss": 1.262,
+      "step": 11073
+    },
+    {
+      "epoch": 1.77184,
+      "grad_norm": 1.988653653064187,
+      "learning_rate": 7.60097363326863e-05,
+      "loss": 1.2538,
+      "step": 11074
+    },
+    {
+      "epoch": 1.772,
+      "grad_norm": 1.8264390918703945,
+      "learning_rate": 7.599296730278776e-05,
+      "loss": 1.1979,
+      "step": 11075
+    },
+    {
+      "epoch": 1.77216,
+      "grad_norm": 1.9551235086540812,
+      "learning_rate": 7.597619898922404e-05,
+      "loss": 1.2772,
+      "step": 11076
+    },
+    {
+      "epoch": 1.7723200000000001,
+      "grad_norm": 1.8281530942997124,
+      "learning_rate": 7.59594313924956e-05,
+      "loss": 1.2406,
+      "step": 11077
+    },
+    {
+      "epoch": 1.77248,
+      "grad_norm": 2.0088988834321038,
+      "learning_rate": 7.594266451310264e-05,
+      "loss": 1.3192,
+      "step": 11078
+    },
+    {
+      "epoch": 1.77264,
+      "grad_norm": 1.8743134494908333,
+      "learning_rate": 7.592589835154553e-05,
+      "loss": 1.2906,
+      "step": 11079
+    },
+    {
+      "epoch": 1.7728000000000002,
+      "grad_norm": 1.8890518280416875,
+      "learning_rate": 7.590913290832452e-05,
+      "loss": 1.2086,
+      "step": 11080
+    },
+    {
+      "epoch": 1.7729599999999999,
+      "grad_norm": 2.205622128763326,
+      "learning_rate": 7.589236818393992e-05,
+      "loss": 1.2648,
+      "step": 11081
+    },
+    {
+      "epoch": 1.77312,
+      "grad_norm": 2.5062432691805783,
+      "learning_rate": 7.58756041788919e-05,
+      "loss": 1.337,
+      "step": 11082
+    },
+    {
+      "epoch": 1.77328,
+      "grad_norm": 1.9328982837657176,
+      "learning_rate": 7.585884089368071e-05,
+      "loss": 1.2575,
+      "step": 11083
+    },
+    {
+      "epoch": 1.77344,
+      "grad_norm": 1.928780339273491,
+      "learning_rate": 7.58420783288065e-05,
+      "loss": 1.2305,
+      "step": 11084
+    },
+    {
+      "epoch": 1.7736,
+      "grad_norm": 1.9882733372115011,
+      "learning_rate": 7.582531648476953e-05,
+      "loss": 1.2767,
+      "step": 11085
+    },
+    {
+      "epoch": 1.77376,
+      "grad_norm": 2.1331230827474386,
+      "learning_rate": 7.580855536206984e-05,
+      "loss": 1.2392,
+      "step": 11086
+    },
+    {
+      "epoch": 1.77392,
+      "grad_norm": 2.111784786183339,
+      "learning_rate": 7.579179496120761e-05,
+      "loss": 1.2853,
+      "step": 11087
+    },
+    {
+      "epoch": 1.77408,
+      "grad_norm": 1.856000774431983,
+      "learning_rate": 7.577503528268294e-05,
+      "loss": 1.2545,
+      "step": 11088
+    },
+    {
+      "epoch": 1.77424,
+      "grad_norm": 2.036982343483296,
+      "learning_rate": 7.575827632699595e-05,
+      "loss": 1.2288,
+      "step": 11089
+    },
+    {
+      "epoch": 1.7744,
+      "grad_norm": 1.9873198675331845,
+      "learning_rate": 7.574151809464663e-05,
+      "loss": 1.2826,
+      "step": 11090
+    },
+    {
+      "epoch": 1.7745600000000001,
+      "grad_norm": 2.226180752136115,
+      "learning_rate": 7.572476058613511e-05,
+      "loss": 1.2906,
+      "step": 11091
+    },
+    {
+      "epoch": 1.7747199999999999,
+      "grad_norm": 2.0084413613984116,
+      "learning_rate": 7.570800380196132e-05,
+      "loss": 1.2463,
+      "step": 11092
+    },
+    {
+      "epoch": 1.77488,
+      "grad_norm": 1.83336359251165,
+      "learning_rate": 7.569124774262534e-05,
+      "loss": 1.2652,
+      "step": 11093
+    },
+    {
+      "epoch": 1.77504,
+      "grad_norm": 1.8369563591848634,
+      "learning_rate": 7.567449240862706e-05,
+      "loss": 1.1877,
+      "step": 11094
+    },
+    {
+      "epoch": 1.7752,
+      "grad_norm": 1.9223835546554768,
+      "learning_rate": 7.56577378004665e-05,
+      "loss": 1.2227,
+      "step": 11095
+    },
+    {
+      "epoch": 1.77536,
+      "grad_norm": 2.0626875966010916,
+      "learning_rate": 7.564098391864356e-05,
+      "loss": 1.2965,
+      "step": 11096
+    },
+    {
+      "epoch": 1.77552,
+      "grad_norm": 2.049219628403665,
+      "learning_rate": 7.562423076365822e-05,
+      "loss": 1.2693,
+      "step": 11097
+    },
+    {
+      "epoch": 1.77568,
+      "grad_norm": 1.8744721945151002,
+      "learning_rate": 7.560747833601024e-05,
+      "loss": 1.3267,
+      "step": 11098
+    },
+    {
+      "epoch": 1.77584,
+      "grad_norm": 2.488714223658638,
+      "learning_rate": 7.559072663619962e-05,
+      "loss": 1.2999,
+      "step": 11099
+    },
+    {
+      "epoch": 1.776,
+      "grad_norm": 2.1809938396064186,
+      "learning_rate": 7.557397566472611e-05,
+      "loss": 1.282,
+      "step": 11100
+    },
+    {
+      "epoch": 1.77616,
+      "grad_norm": 2.2664622889960575,
+      "learning_rate": 7.55572254220896e-05,
+      "loss": 1.2894,
+      "step": 11101
+    },
+    {
+      "epoch": 1.7763200000000001,
+      "grad_norm": 1.738213441579651,
+      "learning_rate": 7.554047590878986e-05,
+      "loss": 1.3503,
+      "step": 11102
+    },
+    {
+      "epoch": 1.7764799999999998,
+      "grad_norm": 2.5611740771756533,
+      "learning_rate": 7.552372712532667e-05,
+      "loss": 1.2318,
+      "step": 11103
+    },
+    {
+      "epoch": 1.77664,
+      "grad_norm": 2.0526206253159884,
+      "learning_rate": 7.550697907219978e-05,
+      "loss": 1.2975,
+      "step": 11104
+    },
+    {
+      "epoch": 1.7768000000000002,
+      "grad_norm": 3.0626978466184953,
+      "learning_rate": 7.549023174990895e-05,
+      "loss": 1.2815,
+      "step": 11105
+    },
+    {
+      "epoch": 1.7769599999999999,
+      "grad_norm": 2.0771941582710376,
+      "learning_rate": 7.547348515895393e-05,
+      "loss": 1.215,
+      "step": 11106
+    },
+    {
+      "epoch": 1.77712,
+      "grad_norm": 2.237382144359357,
+      "learning_rate": 7.545673929983433e-05,
+      "loss": 1.1974,
+      "step": 11107
+    },
+    {
+      "epoch": 1.77728,
+      "grad_norm": 2.028915001223952,
+      "learning_rate": 7.54399941730499e-05,
+      "loss": 1.3038,
+      "step": 11108
+    },
+    {
+      "epoch": 1.77744,
+      "grad_norm": 1.9180501862201929,
+      "learning_rate": 7.542324977910022e-05,
+      "loss": 1.2338,
+      "step": 11109
+    },
+    {
+      "epoch": 1.7776,
+      "grad_norm": 2.013940102918092,
+      "learning_rate": 7.540650611848502e-05,
+      "loss": 1.3423,
+      "step": 11110
+    },
+    {
+      "epoch": 1.77776,
+      "grad_norm": 2.162640659153854,
+      "learning_rate": 7.538976319170377e-05,
+      "loss": 1.2513,
+      "step": 11111
+    },
+    {
+      "epoch": 1.77792,
+      "grad_norm": 2.165094847507836,
+      "learning_rate": 7.537302099925615e-05,
+      "loss": 1.3118,
+      "step": 11112
+    },
+    {
+      "epoch": 1.77808,
+      "grad_norm": 1.8999337078274396,
+      "learning_rate": 7.535627954164168e-05,
+      "loss": 1.2888,
+      "step": 11113
+    },
+    {
+      "epoch": 1.77824,
+      "grad_norm": 2.0505399859659508,
+      "learning_rate": 7.533953881935996e-05,
+      "loss": 1.2431,
+      "step": 11114
+    },
+    {
+      "epoch": 1.7784,
+      "grad_norm": 1.9850541922049831,
+      "learning_rate": 7.532279883291042e-05,
+      "loss": 1.2481,
+      "step": 11115
+    },
+    {
+      "epoch": 1.7785600000000001,
+      "grad_norm": 1.903142326191867,
+      "learning_rate": 7.530605958279267e-05,
+      "loss": 1.338,
+      "step": 11116
+    },
+    {
+      "epoch": 1.7787199999999999,
+      "grad_norm": 2.3868273548435956,
+      "learning_rate": 7.528932106950605e-05,
+      "loss": 1.2505,
+      "step": 11117
+    },
+    {
+      "epoch": 1.77888,
+      "grad_norm": 1.9739836464629104,
+      "learning_rate": 7.527258329355013e-05,
+      "loss": 1.3044,
+      "step": 11118
+    },
+    {
+      "epoch": 1.77904,
+      "grad_norm": 1.8466169423171224,
+      "learning_rate": 7.525584625542427e-05,
+      "loss": 1.2273,
+      "step": 11119
+    },
+    {
+      "epoch": 1.7792,
+      "grad_norm": 2.1208913646489584,
+      "learning_rate": 7.523910995562791e-05,
+      "loss": 1.3662,
+      "step": 11120
+    },
+    {
+      "epoch": 1.77936,
+      "grad_norm": 2.0620426280220543,
+      "learning_rate": 7.52223743946604e-05,
+      "loss": 1.3017,
+      "step": 11121
+    },
+    {
+      "epoch": 1.77952,
+      "grad_norm": 2.1992251753819074,
+      "learning_rate": 7.520563957302118e-05,
+      "loss": 1.2987,
+      "step": 11122
+    },
+    {
+      "epoch": 1.77968,
+      "grad_norm": 2.23448237668541,
+      "learning_rate": 7.518890549120952e-05,
+      "loss": 1.319,
+      "step": 11123
+    },
+    {
+      "epoch": 1.77984,
+      "grad_norm": 1.895155996630388,
+      "learning_rate": 7.517217214972478e-05,
+      "loss": 1.242,
+      "step": 11124
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 2.0627091177205883,
+      "learning_rate": 7.515543954906622e-05,
+      "loss": 1.2822,
+      "step": 11125
+    },
+    {
+      "epoch": 1.78016,
+      "grad_norm": 2.160076365021221,
+      "learning_rate": 7.513870768973319e-05,
+      "loss": 1.2343,
+      "step": 11126
+    },
+    {
+      "epoch": 1.7803200000000001,
+      "grad_norm": 2.10298496460739,
+      "learning_rate": 7.512197657222483e-05,
+      "loss": 1.268,
+      "step": 11127
+    },
+    {
+      "epoch": 1.7804799999999998,
+      "grad_norm": 1.9934451233819743,
+      "learning_rate": 7.510524619704048e-05,
+      "loss": 1.2941,
+      "step": 11128
+    },
+    {
+      "epoch": 1.78064,
+      "grad_norm": 1.9311655364470581,
+      "learning_rate": 7.508851656467928e-05,
+      "loss": 1.1958,
+      "step": 11129
+    },
+    {
+      "epoch": 1.7808000000000002,
+      "grad_norm": 2.162661399174278,
+      "learning_rate": 7.507178767564049e-05,
+      "loss": 1.2384,
+      "step": 11130
+    },
+    {
+      "epoch": 1.7809599999999999,
+      "grad_norm": 1.8359906511494788,
+      "learning_rate": 7.505505953042319e-05,
+      "loss": 1.2225,
+      "step": 11131
+    },
+    {
+      "epoch": 1.78112,
+      "grad_norm": 1.9969615033115908,
+      "learning_rate": 7.50383321295266e-05,
+      "loss": 1.2416,
+      "step": 11132
+    },
+    {
+      "epoch": 1.78128,
+      "grad_norm": 2.118990306615421,
+      "learning_rate": 7.502160547344978e-05,
+      "loss": 1.2828,
+      "step": 11133
+    },
+    {
+      "epoch": 1.78144,
+      "grad_norm": 1.9741244766092318,
+      "learning_rate": 7.500487956269189e-05,
+      "loss": 1.2817,
+      "step": 11134
+    },
+    {
+      "epoch": 1.7816,
+      "grad_norm": 2.0283449670639024,
+      "learning_rate": 7.498815439775192e-05,
+      "loss": 1.3006,
+      "step": 11135
+    },
+    {
+      "epoch": 1.78176,
+      "grad_norm": 2.4281007258889984,
+      "learning_rate": 7.4971429979129e-05,
+      "loss": 1.2687,
+      "step": 11136
+    },
+    {
+      "epoch": 1.78192,
+      "grad_norm": 1.778188816502412,
+      "learning_rate": 7.495470630732213e-05,
+      "loss": 1.2181,
+      "step": 11137
+    },
+    {
+      "epoch": 1.78208,
+      "grad_norm": 2.1216595687884516,
+      "learning_rate": 7.493798338283036e-05,
+      "loss": 1.3075,
+      "step": 11138
+    },
+    {
+      "epoch": 1.78224,
+      "grad_norm": 1.9133922864822184,
+      "learning_rate": 7.492126120615261e-05,
+      "loss": 1.2342,
+      "step": 11139
+    },
+    {
+      "epoch": 1.7824,
+      "grad_norm": 2.0860875758808,
+      "learning_rate": 7.490453977778792e-05,
+      "loss": 1.3059,
+      "step": 11140
+    },
+    {
+      "epoch": 1.7825600000000001,
+      "grad_norm": 2.1036065762073717,
+      "learning_rate": 7.488781909823515e-05,
+      "loss": 1.2834,
+      "step": 11141
+    },
+    {
+      "epoch": 1.7827199999999999,
+      "grad_norm": 2.03767868776038,
+      "learning_rate": 7.487109916799332e-05,
+      "loss": 1.2849,
+      "step": 11142
+    },
+    {
+      "epoch": 1.78288,
+      "grad_norm": 1.9896850993163069,
+      "learning_rate": 7.485437998756122e-05,
+      "loss": 1.2618,
+      "step": 11143
+    },
+    {
+      "epoch": 1.78304,
+      "grad_norm": 1.893631869284916,
+      "learning_rate": 7.483766155743781e-05,
+      "loss": 1.2854,
+      "step": 11144
+    },
+    {
+      "epoch": 1.7832,
+      "grad_norm": 2.324641892545001,
+      "learning_rate": 7.482094387812188e-05,
+      "loss": 1.2657,
+      "step": 11145
+    },
+    {
+      "epoch": 1.78336,
+      "grad_norm": 1.883750531395554,
+      "learning_rate": 7.480422695011236e-05,
+      "loss": 1.285,
+      "step": 11146
+    },
+    {
+      "epoch": 1.78352,
+      "grad_norm": 2.1462584206710793,
+      "learning_rate": 7.478751077390794e-05,
+      "loss": 1.2458,
+      "step": 11147
+    },
+    {
+      "epoch": 1.78368,
+      "grad_norm": 2.5498337127879145,
+      "learning_rate": 7.477079535000748e-05,
+      "loss": 1.2907,
+      "step": 11148
+    },
+    {
+      "epoch": 1.78384,
+      "grad_norm": 1.7561951542856824,
+      "learning_rate": 7.475408067890972e-05,
+      "loss": 1.2555,
+      "step": 11149
+    },
+    {
+      "epoch": 1.784,
+      "grad_norm": 2.105195434665773,
+      "learning_rate": 7.473736676111342e-05,
+      "loss": 1.2363,
+      "step": 11150
+    },
+    {
+      "epoch": 1.78416,
+      "grad_norm": 2.4581869413981843,
+      "learning_rate": 7.472065359711725e-05,
+      "loss": 1.2935,
+      "step": 11151
+    },
+    {
+      "epoch": 1.7843200000000001,
+      "grad_norm": 2.383617477926634,
+      "learning_rate": 7.470394118741997e-05,
+      "loss": 1.2262,
+      "step": 11152
+    },
+    {
+      "epoch": 1.7844799999999998,
+      "grad_norm": 2.3633269857900174,
+      "learning_rate": 7.468722953252019e-05,
+      "loss": 1.2182,
+      "step": 11153
+    },
+    {
+      "epoch": 1.78464,
+      "grad_norm": 2.604629827637804,
+      "learning_rate": 7.467051863291666e-05,
+      "loss": 1.3435,
+      "step": 11154
+    },
+    {
+      "epoch": 1.7848000000000002,
+      "grad_norm": 1.7971235969594053,
+      "learning_rate": 7.465380848910787e-05,
+      "loss": 1.2629,
+      "step": 11155
+    },
+    {
+      "epoch": 1.7849599999999999,
+      "grad_norm": 2.4426130425706694,
+      "learning_rate": 7.463709910159256e-05,
+      "loss": 1.281,
+      "step": 11156
+    },
+    {
+      "epoch": 1.78512,
+      "grad_norm": 2.4082971355457357,
+      "learning_rate": 7.462039047086922e-05,
+      "loss": 1.2603,
+      "step": 11157
+    },
+    {
+      "epoch": 1.78528,
+      "grad_norm": 2.0617254397090563,
+      "learning_rate": 7.460368259743649e-05,
+      "loss": 1.2875,
+      "step": 11158
+    },
+    {
+      "epoch": 1.78544,
+      "grad_norm": 1.730231430094228,
+      "learning_rate": 7.458697548179282e-05,
+      "loss": 1.2613,
+      "step": 11159
+    },
+    {
+      "epoch": 1.7856,
+      "grad_norm": 2.469711677921504,
+      "learning_rate": 7.457026912443678e-05,
+      "loss": 1.2607,
+      "step": 11160
+    },
+    {
+      "epoch": 1.78576,
+      "grad_norm": 2.9363989389247362,
+      "learning_rate": 7.455356352586683e-05,
+      "loss": 1.2931,
+      "step": 11161
+    },
+    {
+      "epoch": 1.78592,
+      "grad_norm": 2.0381684732392604,
+      "learning_rate": 7.453685868658152e-05,
+      "loss": 1.296,
+      "step": 11162
+    },
+    {
+      "epoch": 1.7860800000000001,
+      "grad_norm": 2.4132056729880786,
+      "learning_rate": 7.45201546070792e-05,
+      "loss": 1.2745,
+      "step": 11163
+    },
+    {
+      "epoch": 1.78624,
+      "grad_norm": 1.9956584689717691,
+      "learning_rate": 7.450345128785833e-05,
+      "loss": 1.2505,
+      "step": 11164
+    },
+    {
+      "epoch": 1.7864,
+      "grad_norm": 2.060761002649513,
+      "learning_rate": 7.448674872941738e-05,
+      "loss": 1.1988,
+      "step": 11165
+    },
+    {
+      "epoch": 1.7865600000000001,
+      "grad_norm": 1.9442787023826615,
+      "learning_rate": 7.447004693225462e-05,
+      "loss": 1.2887,
+      "step": 11166
+    },
+    {
+      "epoch": 1.7867199999999999,
+      "grad_norm": 1.9751890352769295,
+      "learning_rate": 7.44533458968685e-05,
+      "loss": 1.2167,
+      "step": 11167
+    },
+    {
+      "epoch": 1.78688,
+      "grad_norm": 2.3299073106376937,
+      "learning_rate": 7.443664562375727e-05,
+      "loss": 1.2089,
+      "step": 11168
+    },
+    {
+      "epoch": 1.78704,
+      "grad_norm": 1.9625633841985377,
+      "learning_rate": 7.441994611341932e-05,
+      "loss": 1.2377,
+      "step": 11169
+    },
+    {
+      "epoch": 1.7872,
+      "grad_norm": 1.9686843410036128,
+      "learning_rate": 7.440324736635288e-05,
+      "loss": 1.2633,
+      "step": 11170
+    },
+    {
+      "epoch": 1.78736,
+      "grad_norm": 2.2536446634844594,
+      "learning_rate": 7.438654938305625e-05,
+      "loss": 1.2613,
+      "step": 11171
+    },
+    {
+      "epoch": 1.78752,
+      "grad_norm": 2.2283718141374496,
+      "learning_rate": 7.436985216402765e-05,
+      "loss": 1.3078,
+      "step": 11172
+    },
+    {
+      "epoch": 1.78768,
+      "grad_norm": 2.066638157392055,
+      "learning_rate": 7.435315570976534e-05,
+      "loss": 1.256,
+      "step": 11173
+    },
+    {
+      "epoch": 1.78784,
+      "grad_norm": 2.0324087434136673,
+      "learning_rate": 7.433646002076747e-05,
+      "loss": 1.2488,
+      "step": 11174
+    },
+    {
+      "epoch": 1.788,
+      "grad_norm": 2.4033437292004685,
+      "learning_rate": 7.431976509753226e-05,
+      "loss": 1.2375,
+      "step": 11175
+    },
+    {
+      "epoch": 1.78816,
+      "grad_norm": 2.2526968771423324,
+      "learning_rate": 7.430307094055781e-05,
+      "loss": 1.3179,
+      "step": 11176
+    },
+    {
+      "epoch": 1.7883200000000001,
+      "grad_norm": 1.8682004462167368,
+      "learning_rate": 7.42863775503423e-05,
+      "loss": 1.2676,
+      "step": 11177
+    },
+    {
+      "epoch": 1.7884799999999998,
+      "grad_norm": 1.9408624245835169,
+      "learning_rate": 7.426968492738382e-05,
+      "loss": 1.2786,
+      "step": 11178
+    },
+    {
+      "epoch": 1.78864,
+      "grad_norm": 2.2637468885682215,
+      "learning_rate": 7.425299307218043e-05,
+      "loss": 1.3399,
+      "step": 11179
+    },
+    {
+      "epoch": 1.7888,
+      "grad_norm": 1.9993642409838037,
+      "learning_rate": 7.42363019852302e-05,
+      "loss": 1.3208,
+      "step": 11180
+    },
+    {
+      "epoch": 1.7889599999999999,
+      "grad_norm": 2.952553559222856,
+      "learning_rate": 7.421961166703122e-05,
+      "loss": 1.2603,
+      "step": 11181
+    },
+    {
+      "epoch": 1.78912,
+      "grad_norm": 2.144266322393598,
+      "learning_rate": 7.420292211808141e-05,
+      "loss": 1.2578,
+      "step": 11182
+    },
+    {
+      "epoch": 1.78928,
+      "grad_norm": 2.1979325052712877,
+      "learning_rate": 7.418623333887887e-05,
+      "loss": 1.2359,
+      "step": 11183
+    },
+    {
+      "epoch": 1.78944,
+      "grad_norm": 2.4769279079916293,
+      "learning_rate": 7.416954532992147e-05,
+      "loss": 1.2911,
+      "step": 11184
+    },
+    {
+      "epoch": 1.7896,
+      "grad_norm": 2.2464651667233455,
+      "learning_rate": 7.41528580917072e-05,
+      "loss": 1.2688,
+      "step": 11185
+    },
+    {
+      "epoch": 1.78976,
+      "grad_norm": 1.940360442630734,
+      "learning_rate": 7.4136171624734e-05,
+      "loss": 1.2795,
+      "step": 11186
+    },
+    {
+      "epoch": 1.78992,
+      "grad_norm": 2.439799924315086,
+      "learning_rate": 7.411948592949975e-05,
+      "loss": 1.2591,
+      "step": 11187
+    },
+    {
+      "epoch": 1.7900800000000001,
+      "grad_norm": 2.111854618759018,
+      "learning_rate": 7.41028010065023e-05,
+      "loss": 1.2393,
+      "step": 11188
+    },
+    {
+      "epoch": 1.79024,
+      "grad_norm": 1.7164928517627513,
+      "learning_rate": 7.408611685623959e-05,
+      "loss": 1.2458,
+      "step": 11189
+    },
+    {
+      "epoch": 1.7904,
+      "grad_norm": 2.3733559531539927,
+      "learning_rate": 7.406943347920931e-05,
+      "loss": 1.2855,
+      "step": 11190
+    },
+    {
+      "epoch": 1.7905600000000002,
+      "grad_norm": 2.383897214017547,
+      "learning_rate": 7.405275087590944e-05,
+      "loss": 1.1936,
+      "step": 11191
+    },
+    {
+      "epoch": 1.7907199999999999,
+      "grad_norm": 1.6379671390486157,
+      "learning_rate": 7.403606904683759e-05,
+      "loss": 1.2526,
+      "step": 11192
+    },
+    {
+      "epoch": 1.79088,
+      "grad_norm": 2.1791275132761054,
+      "learning_rate": 7.401938799249166e-05,
+      "loss": 1.2507,
+      "step": 11193
+    },
+    {
+      "epoch": 1.79104,
+      "grad_norm": 2.617425938643105,
+      "learning_rate": 7.400270771336933e-05,
+      "loss": 1.2601,
+      "step": 11194
+    },
+    {
+      "epoch": 1.7912,
+      "grad_norm": 2.151499723469867,
+      "learning_rate": 7.398602820996832e-05,
+      "loss": 1.2785,
+      "step": 11195
+    },
+    {
+      "epoch": 1.79136,
+      "grad_norm": 1.6954491049516167,
+      "learning_rate": 7.39693494827863e-05,
+      "loss": 1.232,
+      "step": 11196
+    },
+    {
+      "epoch": 1.79152,
+      "grad_norm": 2.3275299073015874,
+      "learning_rate": 7.395267153232099e-05,
+      "loss": 1.3218,
+      "step": 11197
+    },
+    {
+      "epoch": 1.79168,
+      "grad_norm": 2.307537436860288,
+      "learning_rate": 7.393599435906997e-05,
+      "loss": 1.2412,
+      "step": 11198
+    },
+    {
+      "epoch": 1.79184,
+      "grad_norm": 1.9586647614796315,
+      "learning_rate": 7.391931796353093e-05,
+      "loss": 1.2545,
+      "step": 11199
+    },
+    {
+      "epoch": 1.792,
+      "grad_norm": 2.363234965579525,
+      "learning_rate": 7.390264234620141e-05,
+      "loss": 1.2882,
+      "step": 11200
+    },
+    {
+      "epoch": 1.79216,
+      "grad_norm": 2.159604323433302,
+      "learning_rate": 7.388596750757903e-05,
+      "loss": 1.2875,
+      "step": 11201
+    },
+    {
+      "epoch": 1.7923200000000001,
+      "grad_norm": 2.2642874968678877,
+      "learning_rate": 7.386929344816133e-05,
+      "loss": 1.2342,
+      "step": 11202
+    },
+    {
+      "epoch": 1.7924799999999999,
+      "grad_norm": 2.0374084787728832,
+      "learning_rate": 7.385262016844585e-05,
+      "loss": 1.2597,
+      "step": 11203
+    },
+    {
+      "epoch": 1.79264,
+      "grad_norm": 2.626014038359289,
+      "learning_rate": 7.383594766893005e-05,
+      "loss": 1.2542,
+      "step": 11204
+    },
+    {
+      "epoch": 1.7928,
+      "grad_norm": 1.8926631903882798,
+      "learning_rate": 7.38192759501115e-05,
+      "loss": 1.2917,
+      "step": 11205
+    },
+    {
+      "epoch": 1.7929599999999999,
+      "grad_norm": 2.055329156528195,
+      "learning_rate": 7.380260501248755e-05,
+      "loss": 1.2741,
+      "step": 11206
+    },
+    {
+      "epoch": 1.79312,
+      "grad_norm": 2.054780758061585,
+      "learning_rate": 7.378593485655573e-05,
+      "loss": 1.3155,
+      "step": 11207
+    },
+    {
+      "epoch": 1.79328,
+      "grad_norm": 2.2328592216557928,
+      "learning_rate": 7.37692654828134e-05,
+      "loss": 1.2898,
+      "step": 11208
+    },
+    {
+      "epoch": 1.79344,
+      "grad_norm": 1.7716480001665924,
+      "learning_rate": 7.375259689175797e-05,
+      "loss": 1.2863,
+      "step": 11209
+    },
+    {
+      "epoch": 1.7936,
+      "grad_norm": 2.1795269970081206,
+      "learning_rate": 7.373592908388678e-05,
+      "loss": 1.2428,
+      "step": 11210
+    },
+    {
+      "epoch": 1.79376,
+      "grad_norm": 2.034050879036329,
+      "learning_rate": 7.371926205969724e-05,
+      "loss": 1.258,
+      "step": 11211
+    },
+    {
+      "epoch": 1.79392,
+      "grad_norm": 1.8518642172352546,
+      "learning_rate": 7.370259581968656e-05,
+      "loss": 1.2904,
+      "step": 11212
+    },
+    {
+      "epoch": 1.7940800000000001,
+      "grad_norm": 2.091093828342361,
+      "learning_rate": 7.368593036435218e-05,
+      "loss": 1.204,
+      "step": 11213
+    },
+    {
+      "epoch": 1.7942399999999998,
+      "grad_norm": 1.8890999261982482,
+      "learning_rate": 7.366926569419123e-05,
+      "loss": 1.1882,
+      "step": 11214
+    },
+    {
+      "epoch": 1.7944,
+      "grad_norm": 1.7117136033777254,
+      "learning_rate": 7.365260180970108e-05,
+      "loss": 1.2227,
+      "step": 11215
+    },
+    {
+      "epoch": 1.7945600000000002,
+      "grad_norm": 2.331660285015401,
+      "learning_rate": 7.363593871137885e-05,
+      "loss": 1.2439,
+      "step": 11216
+    },
+    {
+      "epoch": 1.7947199999999999,
+      "grad_norm": 1.90035385165663,
+      "learning_rate": 7.361927639972181e-05,
+      "loss": 1.2298,
+      "step": 11217
+    },
+    {
+      "epoch": 1.79488,
+      "grad_norm": 1.862875629758189,
+      "learning_rate": 7.36026148752271e-05,
+      "loss": 1.2682,
+      "step": 11218
+    },
+    {
+      "epoch": 1.79504,
+      "grad_norm": 2.1257013108499985,
+      "learning_rate": 7.358595413839196e-05,
+      "loss": 1.2688,
+      "step": 11219
+    },
+    {
+      "epoch": 1.7952,
+      "grad_norm": 2.105368646602599,
+      "learning_rate": 7.356929418971339e-05,
+      "loss": 1.2956,
+      "step": 11220
+    },
+    {
+      "epoch": 1.79536,
+      "grad_norm": 1.905164045920572,
+      "learning_rate": 7.355263502968863e-05,
+      "loss": 1.2956,
+      "step": 11221
+    },
+    {
+      "epoch": 1.79552,
+      "grad_norm": 1.7163447241198682,
+      "learning_rate": 7.353597665881465e-05,
+      "loss": 1.2103,
+      "step": 11222
+    },
+    {
+      "epoch": 1.79568,
+      "grad_norm": 2.133277072139544,
+      "learning_rate": 7.351931907758855e-05,
+      "loss": 1.3098,
+      "step": 11223
+    },
+    {
+      "epoch": 1.79584,
+      "grad_norm": 1.9939635561704812,
+      "learning_rate": 7.350266228650746e-05,
+      "loss": 1.2528,
+      "step": 11224
+    },
+    {
+      "epoch": 1.796,
+      "grad_norm": 1.8883887929837457,
+      "learning_rate": 7.348600628606824e-05,
+      "loss": 1.2604,
+      "step": 11225
+    },
+    {
+      "epoch": 1.79616,
+      "grad_norm": 2.078083869531086,
+      "learning_rate": 7.3469351076768e-05,
+      "loss": 1.2292,
+      "step": 11226
+    },
+    {
+      "epoch": 1.7963200000000001,
+      "grad_norm": 2.0837508247421463,
+      "learning_rate": 7.345269665910366e-05,
+      "loss": 1.3434,
+      "step": 11227
+    },
+    {
+      "epoch": 1.7964799999999999,
+      "grad_norm": 2.3204889151085597,
+      "learning_rate": 7.343604303357217e-05,
+      "loss": 1.2661,
+      "step": 11228
+    },
+    {
+      "epoch": 1.79664,
+      "grad_norm": 2.1253390282521676,
+      "learning_rate": 7.341939020067041e-05,
+      "loss": 1.2804,
+      "step": 11229
+    },
+    {
+      "epoch": 1.7968,
+      "grad_norm": 1.9104592686635447,
+      "learning_rate": 7.340273816089538e-05,
+      "loss": 1.2439,
+      "step": 11230
+    },
+    {
+      "epoch": 1.79696,
+      "grad_norm": 2.3340218183624413,
+      "learning_rate": 7.338608691474383e-05,
+      "loss": 1.2127,
+      "step": 11231
+    },
+    {
+      "epoch": 1.79712,
+      "grad_norm": 1.8771804799238436,
+      "learning_rate": 7.33694364627127e-05,
+      "loss": 1.3243,
+      "step": 11232
+    },
+    {
+      "epoch": 1.79728,
+      "grad_norm": 2.165292614993149,
+      "learning_rate": 7.335278680529875e-05,
+      "loss": 1.3028,
+      "step": 11233
+    },
+    {
+      "epoch": 1.79744,
+      "grad_norm": 2.4234348460649895,
+      "learning_rate": 7.333613794299885e-05,
+      "loss": 1.3286,
+      "step": 11234
+    },
+    {
+      "epoch": 1.7976,
+      "grad_norm": 1.9749343719183918,
+      "learning_rate": 7.331948987630972e-05,
+      "loss": 1.253,
+      "step": 11235
+    },
+    {
+      "epoch": 1.79776,
+      "grad_norm": 1.9695833320075034,
+      "learning_rate": 7.330284260572815e-05,
+      "loss": 1.2567,
+      "step": 11236
+    },
+    {
+      "epoch": 1.79792,
+      "grad_norm": 2.1866644512864326,
+      "learning_rate": 7.328619613175082e-05,
+      "loss": 1.2601,
+      "step": 11237
+    },
+    {
+      "epoch": 1.7980800000000001,
+      "grad_norm": 2.429730885062175,
+      "learning_rate": 7.326955045487452e-05,
+      "loss": 1.3195,
+      "step": 11238
+    },
+    {
+      "epoch": 1.7982399999999998,
+      "grad_norm": 1.883273091644581,
+      "learning_rate": 7.325290557559583e-05,
+      "loss": 1.2979,
+      "step": 11239
+    },
+    {
+      "epoch": 1.7984,
+      "grad_norm": 2.037237494112446,
+      "learning_rate": 7.323626149441152e-05,
+      "loss": 1.2384,
+      "step": 11240
+    },
+    {
+      "epoch": 1.7985600000000002,
+      "grad_norm": 2.068077408953883,
+      "learning_rate": 7.321961821181811e-05,
+      "loss": 1.2831,
+      "step": 11241
+    },
+    {
+      "epoch": 1.7987199999999999,
+      "grad_norm": 2.6601121591007724,
+      "learning_rate": 7.320297572831232e-05,
+      "loss": 1.2677,
+      "step": 11242
+    },
+    {
+      "epoch": 1.79888,
+      "grad_norm": 1.9260660647103527,
+      "learning_rate": 7.318633404439066e-05,
+      "loss": 1.1991,
+      "step": 11243
+    },
+    {
+      "epoch": 1.79904,
+      "grad_norm": 1.9995144993811613,
+      "learning_rate": 7.316969316054973e-05,
+      "loss": 1.293,
+      "step": 11244
+    },
+    {
+      "epoch": 1.7992,
+      "grad_norm": 2.056844780063517,
+      "learning_rate": 7.315305307728604e-05,
+      "loss": 1.2373,
+      "step": 11245
+    },
+    {
+      "epoch": 1.79936,
+      "grad_norm": 1.8265048204826992,
+      "learning_rate": 7.313641379509616e-05,
+      "loss": 1.3099,
+      "step": 11246
+    },
+    {
+      "epoch": 1.79952,
+      "grad_norm": 2.0308503685631445,
+      "learning_rate": 7.311977531447651e-05,
+      "loss": 1.2278,
+      "step": 11247
+    },
+    {
+      "epoch": 1.79968,
+      "grad_norm": 2.0801045354755385,
+      "learning_rate": 7.310313763592364e-05,
+      "loss": 1.2267,
+      "step": 11248
+    },
+    {
+      "epoch": 1.79984,
+      "grad_norm": 2.03789482273933,
+      "learning_rate": 7.308650075993391e-05,
+      "loss": 1.2419,
+      "step": 11249
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 2.0540843791798133,
+      "learning_rate": 7.306986468700379e-05,
+      "loss": 1.1942,
+      "step": 11250
+    },
+    {
+      "epoch": 1.80016,
+      "grad_norm": 2.316473775355317,
+      "learning_rate": 7.305322941762968e-05,
+      "loss": 1.2482,
+      "step": 11251
+    },
+    {
+      "epoch": 1.8003200000000001,
+      "grad_norm": 2.005948101515914,
+      "learning_rate": 7.303659495230793e-05,
+      "loss": 1.2539,
+      "step": 11252
+    },
+    {
+      "epoch": 1.8004799999999999,
+      "grad_norm": 2.2580279943311474,
+      "learning_rate": 7.301996129153487e-05,
+      "loss": 1.1819,
+      "step": 11253
+    },
+    {
+      "epoch": 1.80064,
+      "grad_norm": 2.6334594494165913,
+      "learning_rate": 7.30033284358069e-05,
+      "loss": 1.2279,
+      "step": 11254
+    },
+    {
+      "epoch": 1.8008,
+      "grad_norm": 2.7371399743373126,
+      "learning_rate": 7.298669638562024e-05,
+      "loss": 1.2487,
+      "step": 11255
+    },
+    {
+      "epoch": 1.80096,
+      "grad_norm": 1.9211887089984094,
+      "learning_rate": 7.297006514147124e-05,
+      "loss": 1.291,
+      "step": 11256
+    },
+    {
+      "epoch": 1.80112,
+      "grad_norm": 2.177289873676676,
+      "learning_rate": 7.295343470385605e-05,
+      "loss": 1.1975,
+      "step": 11257
+    },
+    {
+      "epoch": 1.80128,
+      "grad_norm": 2.678888251907692,
+      "learning_rate": 7.293680507327098e-05,
+      "loss": 1.2855,
+      "step": 11258
+    },
+    {
+      "epoch": 1.80144,
+      "grad_norm": 2.2780000867541443,
+      "learning_rate": 7.292017625021223e-05,
+      "loss": 1.2609,
+      "step": 11259
+    },
+    {
+      "epoch": 1.8016,
+      "grad_norm": 2.0365721336833573,
+      "learning_rate": 7.290354823517594e-05,
+      "loss": 1.2473,
+      "step": 11260
+    },
+    {
+      "epoch": 1.80176,
+      "grad_norm": 2.833111922385791,
+      "learning_rate": 7.288692102865828e-05,
+      "loss": 1.2239,
+      "step": 11261
+    },
+    {
+      "epoch": 1.80192,
+      "grad_norm": 2.0056026508699256,
+      "learning_rate": 7.287029463115542e-05,
+      "loss": 1.321,
+      "step": 11262
+    },
+    {
+      "epoch": 1.8020800000000001,
+      "grad_norm": 2.105611118403836,
+      "learning_rate": 7.28536690431634e-05,
+      "loss": 1.2151,
+      "step": 11263
+    },
+    {
+      "epoch": 1.8022399999999998,
+      "grad_norm": 2.1004615776640128,
+      "learning_rate": 7.283704426517839e-05,
+      "loss": 1.1702,
+      "step": 11264
+    },
+    {
+      "epoch": 1.8024,
+      "grad_norm": 1.934864286072801,
+      "learning_rate": 7.282042029769634e-05,
+      "loss": 1.293,
+      "step": 11265
+    },
+    {
+      "epoch": 1.8025600000000002,
+      "grad_norm": 2.3697355920603402,
+      "learning_rate": 7.280379714121338e-05,
+      "loss": 1.2899,
+      "step": 11266
+    },
+    {
+      "epoch": 1.8027199999999999,
+      "grad_norm": 1.9238534516032608,
+      "learning_rate": 7.278717479622547e-05,
+      "loss": 1.2671,
+      "step": 11267
+    },
+    {
+      "epoch": 1.80288,
+      "grad_norm": 1.9014099179402129,
+      "learning_rate": 7.277055326322862e-05,
+      "loss": 1.3117,
+      "step": 11268
+    },
+    {
+      "epoch": 1.80304,
+      "grad_norm": 2.1900079539676343,
+      "learning_rate": 7.275393254271876e-05,
+      "loss": 1.2771,
+      "step": 11269
+    },
+    {
+      "epoch": 1.8032,
+      "grad_norm": 1.704065678375464,
+      "learning_rate": 7.27373126351919e-05,
+      "loss": 1.278,
+      "step": 11270
+    },
+    {
+      "epoch": 1.80336,
+      "grad_norm": 2.075120979439852,
+      "learning_rate": 7.272069354114385e-05,
+      "loss": 1.2692,
+      "step": 11271
+    },
+    {
+      "epoch": 1.80352,
+      "grad_norm": 1.8457083158003764,
+      "learning_rate": 7.27040752610706e-05,
+      "loss": 1.2226,
+      "step": 11272
+    },
+    {
+      "epoch": 1.80368,
+      "grad_norm": 1.784869678336958,
+      "learning_rate": 7.268745779546793e-05,
+      "loss": 1.2447,
+      "step": 11273
+    },
+    {
+      "epoch": 1.80384,
+      "grad_norm": 2.0317347342190426,
+      "learning_rate": 7.267084114483173e-05,
+      "loss": 1.2923,
+      "step": 11274
+    },
+    {
+      "epoch": 1.804,
+      "grad_norm": 2.097395979259557,
+      "learning_rate": 7.265422530965783e-05,
+      "loss": 1.2603,
+      "step": 11275
+    },
+    {
+      "epoch": 1.80416,
+      "grad_norm": 1.9803445160740862,
+      "learning_rate": 7.263761029044199e-05,
+      "loss": 1.2216,
+      "step": 11276
+    },
+    {
+      "epoch": 1.8043200000000001,
+      "grad_norm": 2.2375753305343773,
+      "learning_rate": 7.262099608767996e-05,
+      "loss": 1.2522,
+      "step": 11277
+    },
+    {
+      "epoch": 1.8044799999999999,
+      "grad_norm": 2.339360586632346,
+      "learning_rate": 7.260438270186756e-05,
+      "loss": 1.3166,
+      "step": 11278
+    },
+    {
+      "epoch": 1.80464,
+      "grad_norm": 2.2164684923000455,
+      "learning_rate": 7.258777013350042e-05,
+      "loss": 1.2901,
+      "step": 11279
+    },
+    {
+      "epoch": 1.8048,
+      "grad_norm": 1.8948331165868486,
+      "learning_rate": 7.257115838307428e-05,
+      "loss": 1.2608,
+      "step": 11280
+    },
+    {
+      "epoch": 1.80496,
+      "grad_norm": 2.124797006974263,
+      "learning_rate": 7.255454745108483e-05,
+      "loss": 1.2764,
+      "step": 11281
+    },
+    {
+      "epoch": 1.80512,
+      "grad_norm": 2.041322866107434,
+      "learning_rate": 7.253793733802767e-05,
+      "loss": 1.3241,
+      "step": 11282
+    },
+    {
+      "epoch": 1.80528,
+      "grad_norm": 1.9534578555761815,
+      "learning_rate": 7.252132804439847e-05,
+      "loss": 1.3345,
+      "step": 11283
+    },
+    {
+      "epoch": 1.80544,
+      "grad_norm": 2.484772089576805,
+      "learning_rate": 7.250471957069277e-05,
+      "loss": 1.2578,
+      "step": 11284
+    },
+    {
+      "epoch": 1.8056,
+      "grad_norm": 2.6414940670158176,
+      "learning_rate": 7.248811191740618e-05,
+      "loss": 1.2636,
+      "step": 11285
+    },
+    {
+      "epoch": 1.80576,
+      "grad_norm": 2.052558185466259,
+      "learning_rate": 7.247150508503424e-05,
+      "loss": 1.2236,
+      "step": 11286
+    },
+    {
+      "epoch": 1.80592,
+      "grad_norm": 2.4295671187537113,
+      "learning_rate": 7.24548990740725e-05,
+      "loss": 1.2755,
+      "step": 11287
+    },
+    {
+      "epoch": 1.8060800000000001,
+      "grad_norm": 1.9560337458107644,
+      "learning_rate": 7.243829388501639e-05,
+      "loss": 1.2698,
+      "step": 11288
+    },
+    {
+      "epoch": 1.8062399999999998,
+      "grad_norm": 2.102670025346469,
+      "learning_rate": 7.242168951836148e-05,
+      "loss": 1.264,
+      "step": 11289
+    },
+    {
+      "epoch": 1.8064,
+      "grad_norm": 1.853596227946366,
+      "learning_rate": 7.240508597460314e-05,
+      "loss": 1.2487,
+      "step": 11290
+    },
+    {
+      "epoch": 1.8065600000000002,
+      "grad_norm": 2.3382075238585567,
+      "learning_rate": 7.238848325423684e-05,
+      "loss": 1.2454,
+      "step": 11291
+    },
+    {
+      "epoch": 1.8067199999999999,
+      "grad_norm": 1.8319699311244273,
+      "learning_rate": 7.237188135775794e-05,
+      "loss": 1.2537,
+      "step": 11292
+    },
+    {
+      "epoch": 1.80688,
+      "grad_norm": 1.798054698020417,
+      "learning_rate": 7.235528028566185e-05,
+      "loss": 1.25,
+      "step": 11293
+    },
+    {
+      "epoch": 1.80704,
+      "grad_norm": 2.0829863913942934,
+      "learning_rate": 7.23386800384439e-05,
+      "loss": 1.2984,
+      "step": 11294
+    },
+    {
+      "epoch": 1.8072,
+      "grad_norm": 2.035106843599318,
+      "learning_rate": 7.232208061659949e-05,
+      "loss": 1.4086,
+      "step": 11295
+    },
+    {
+      "epoch": 1.80736,
+      "grad_norm": 2.109676016643844,
+      "learning_rate": 7.23054820206238e-05,
+      "loss": 1.288,
+      "step": 11296
+    },
+    {
+      "epoch": 1.80752,
+      "grad_norm": 1.9158923045921743,
+      "learning_rate": 7.228888425101221e-05,
+      "loss": 1.2039,
+      "step": 11297
+    },
+    {
+      "epoch": 1.80768,
+      "grad_norm": 1.8588473576851103,
+      "learning_rate": 7.227228730825991e-05,
+      "loss": 1.2457,
+      "step": 11298
+    },
+    {
+      "epoch": 1.8078400000000001,
+      "grad_norm": 2.6366521088194337,
+      "learning_rate": 7.225569119286217e-05,
+      "loss": 1.2751,
+      "step": 11299
+    },
+    {
+      "epoch": 1.808,
+      "grad_norm": 2.092238990037614,
+      "learning_rate": 7.223909590531416e-05,
+      "loss": 1.2348,
+      "step": 11300
+    },
+    {
+      "epoch": 1.80816,
+      "grad_norm": 2.0496634267484617,
+      "learning_rate": 7.222250144611109e-05,
+      "loss": 1.2078,
+      "step": 11301
+    },
+    {
+      "epoch": 1.8083200000000001,
+      "grad_norm": 2.132248589876724,
+      "learning_rate": 7.220590781574809e-05,
+      "loss": 1.2526,
+      "step": 11302
+    },
+    {
+      "epoch": 1.8084799999999999,
+      "grad_norm": 2.457293231861916,
+      "learning_rate": 7.218931501472033e-05,
+      "loss": 1.2621,
+      "step": 11303
+    },
+    {
+      "epoch": 1.80864,
+      "grad_norm": 2.337162313214365,
+      "learning_rate": 7.217272304352285e-05,
+      "loss": 1.255,
+      "step": 11304
+    },
+    {
+      "epoch": 1.8088,
+      "grad_norm": 1.9868573972992718,
+      "learning_rate": 7.215613190265081e-05,
+      "loss": 1.2706,
+      "step": 11305
+    },
+    {
+      "epoch": 1.80896,
+      "grad_norm": 1.6410950574853325,
+      "learning_rate": 7.213954159259916e-05,
+      "loss": 1.2253,
+      "step": 11306
+    },
+    {
+      "epoch": 1.80912,
+      "grad_norm": 1.8472602427781284,
+      "learning_rate": 7.212295211386302e-05,
+      "loss": 1.2998,
+      "step": 11307
+    },
+    {
+      "epoch": 1.80928,
+      "grad_norm": 2.0778817970957055,
+      "learning_rate": 7.210636346693738e-05,
+      "loss": 1.2684,
+      "step": 11308
+    },
+    {
+      "epoch": 1.80944,
+      "grad_norm": 2.2980723439884203,
+      "learning_rate": 7.208977565231719e-05,
+      "loss": 1.2418,
+      "step": 11309
+    },
+    {
+      "epoch": 1.8096,
+      "grad_norm": 2.353016035562769,
+      "learning_rate": 7.207318867049741e-05,
+      "loss": 1.2048,
+      "step": 11310
+    },
+    {
+      "epoch": 1.80976,
+      "grad_norm": 2.2167169466318755,
+      "learning_rate": 7.205660252197303e-05,
+      "loss": 1.2355,
+      "step": 11311
+    },
+    {
+      "epoch": 1.80992,
+      "grad_norm": 1.9988953036405601,
+      "learning_rate": 7.204001720723888e-05,
+      "loss": 1.314,
+      "step": 11312
+    },
+    {
+      "epoch": 1.8100800000000001,
+      "grad_norm": 1.9639226705712323,
+      "learning_rate": 7.202343272678992e-05,
+      "loss": 1.2966,
+      "step": 11313
+    },
+    {
+      "epoch": 1.8102399999999998,
+      "grad_norm": 1.8140574207498428,
+      "learning_rate": 7.200684908112091e-05,
+      "loss": 1.2404,
+      "step": 11314
+    },
+    {
+      "epoch": 1.8104,
+      "grad_norm": 2.3358522957119616,
+      "learning_rate": 7.199026627072676e-05,
+      "loss": 1.2603,
+      "step": 11315
+    },
+    {
+      "epoch": 1.81056,
+      "grad_norm": 1.8976837542345693,
+      "learning_rate": 7.197368429610224e-05,
+      "loss": 1.2812,
+      "step": 11316
+    },
+    {
+      "epoch": 1.8107199999999999,
+      "grad_norm": 1.8324195975265178,
+      "learning_rate": 7.195710315774214e-05,
+      "loss": 1.3268,
+      "step": 11317
+    },
+    {
+      "epoch": 1.81088,
+      "grad_norm": 1.7630164815515044,
+      "learning_rate": 7.194052285614121e-05,
+      "loss": 1.2402,
+      "step": 11318
+    },
+    {
+      "epoch": 1.81104,
+      "grad_norm": 1.9815595158246202,
+      "learning_rate": 7.192394339179422e-05,
+      "loss": 1.289,
+      "step": 11319
+    },
+    {
+      "epoch": 1.8112,
+      "grad_norm": 2.0881688121087083,
+      "learning_rate": 7.190736476519582e-05,
+      "loss": 1.3312,
+      "step": 11320
+    },
+    {
+      "epoch": 1.81136,
+      "grad_norm": 2.072417855977184,
+      "learning_rate": 7.189078697684075e-05,
+      "loss": 1.2366,
+      "step": 11321
+    },
+    {
+      "epoch": 1.81152,
+      "grad_norm": 1.7106673717226135,
+      "learning_rate": 7.187421002722361e-05,
+      "loss": 1.1915,
+      "step": 11322
+    },
+    {
+      "epoch": 1.81168,
+      "grad_norm": 1.8962520881493277,
+      "learning_rate": 7.185763391683907e-05,
+      "loss": 1.2924,
+      "step": 11323
+    },
+    {
+      "epoch": 1.8118400000000001,
+      "grad_norm": 1.8915833685024086,
+      "learning_rate": 7.184105864618172e-05,
+      "loss": 1.2348,
+      "step": 11324
+    },
+    {
+      "epoch": 1.812,
+      "grad_norm": 1.9526544190657071,
+      "learning_rate": 7.182448421574616e-05,
+      "loss": 1.3451,
+      "step": 11325
+    },
+    {
+      "epoch": 1.81216,
+      "grad_norm": 1.8206372772653265,
+      "learning_rate": 7.180791062602691e-05,
+      "loss": 1.269,
+      "step": 11326
+    },
+    {
+      "epoch": 1.8123200000000002,
+      "grad_norm": 2.4916465877191443,
+      "learning_rate": 7.179133787751857e-05,
+      "loss": 1.2842,
+      "step": 11327
+    },
+    {
+      "epoch": 1.8124799999999999,
+      "grad_norm": 1.9582521926730878,
+      "learning_rate": 7.177476597071555e-05,
+      "loss": 1.2908,
+      "step": 11328
+    },
+    {
+      "epoch": 1.81264,
+      "grad_norm": 2.3445438330713113,
+      "learning_rate": 7.175819490611244e-05,
+      "loss": 1.2917,
+      "step": 11329
+    },
+    {
+      "epoch": 1.8128,
+      "grad_norm": 2.1472731843446833,
+      "learning_rate": 7.17416246842036e-05,
+      "loss": 1.3184,
+      "step": 11330
+    },
+    {
+      "epoch": 1.81296,
+      "grad_norm": 2.211950093339304,
+      "learning_rate": 7.172505530548351e-05,
+      "loss": 1.2364,
+      "step": 11331
+    },
+    {
+      "epoch": 1.81312,
+      "grad_norm": 2.277184330736975,
+      "learning_rate": 7.170848677044658e-05,
+      "loss": 1.2826,
+      "step": 11332
+    },
+    {
+      "epoch": 1.81328,
+      "grad_norm": 2.1541801332354997,
+      "learning_rate": 7.169191907958718e-05,
+      "loss": 1.2163,
+      "step": 11333
+    },
+    {
+      "epoch": 1.81344,
+      "grad_norm": 2.1192299418178533,
+      "learning_rate": 7.167535223339964e-05,
+      "loss": 1.3004,
+      "step": 11334
+    },
+    {
+      "epoch": 1.8136,
+      "grad_norm": 2.0939752130174467,
+      "learning_rate": 7.165878623237837e-05,
+      "loss": 1.2776,
+      "step": 11335
+    },
+    {
+      "epoch": 1.81376,
+      "grad_norm": 2.0471302352194045,
+      "learning_rate": 7.164222107701756e-05,
+      "loss": 1.3124,
+      "step": 11336
+    },
+    {
+      "epoch": 1.81392,
+      "grad_norm": 2.120660416078606,
+      "learning_rate": 7.162565676781161e-05,
+      "loss": 1.3133,
+      "step": 11337
+    },
+    {
+      "epoch": 1.8140800000000001,
+      "grad_norm": 2.1159585963685923,
+      "learning_rate": 7.160909330525466e-05,
+      "loss": 1.3242,
+      "step": 11338
+    },
+    {
+      "epoch": 1.8142399999999999,
+      "grad_norm": 1.839541497571212,
+      "learning_rate": 7.159253068984099e-05,
+      "loss": 1.2166,
+      "step": 11339
+    },
+    {
+      "epoch": 1.8144,
+      "grad_norm": 2.2419169030128647,
+      "learning_rate": 7.157596892206489e-05,
+      "loss": 1.2797,
+      "step": 11340
+    },
+    {
+      "epoch": 1.81456,
+      "grad_norm": 2.052101852651955,
+      "learning_rate": 7.155940800242038e-05,
+      "loss": 1.3234,
+      "step": 11341
+    },
+    {
+      "epoch": 1.8147199999999999,
+      "grad_norm": 1.8852139486078832,
+      "learning_rate": 7.154284793140173e-05,
+      "loss": 1.2498,
+      "step": 11342
+    },
+    {
+      "epoch": 1.81488,
+      "grad_norm": 1.8352734587931925,
+      "learning_rate": 7.152628870950299e-05,
+      "loss": 1.2041,
+      "step": 11343
+    },
+    {
+      "epoch": 1.81504,
+      "grad_norm": 1.9353404834322336,
+      "learning_rate": 7.150973033721838e-05,
+      "loss": 1.2396,
+      "step": 11344
+    },
+    {
+      "epoch": 1.8152,
+      "grad_norm": 2.2257598619744794,
+      "learning_rate": 7.149317281504185e-05,
+      "loss": 1.2528,
+      "step": 11345
+    },
+    {
+      "epoch": 1.81536,
+      "grad_norm": 2.0623738314098765,
+      "learning_rate": 7.147661614346752e-05,
+      "loss": 1.2989,
+      "step": 11346
+    },
+    {
+      "epoch": 1.81552,
+      "grad_norm": 2.1366635320393947,
+      "learning_rate": 7.146006032298937e-05,
+      "loss": 1.2206,
+      "step": 11347
+    },
+    {
+      "epoch": 1.81568,
+      "grad_norm": 1.9600309368862596,
+      "learning_rate": 7.144350535410149e-05,
+      "loss": 1.2543,
+      "step": 11348
+    },
+    {
+      "epoch": 1.8158400000000001,
+      "grad_norm": 2.377845086576981,
+      "learning_rate": 7.142695123729776e-05,
+      "loss": 1.3127,
+      "step": 11349
+    },
+    {
+      "epoch": 1.8159999999999998,
+      "grad_norm": 2.0683515127862653,
+      "learning_rate": 7.141039797307218e-05,
+      "loss": 1.3077,
+      "step": 11350
+    },
+    {
+      "epoch": 1.81616,
+      "grad_norm": 3.121355097349925,
+      "learning_rate": 7.139384556191863e-05,
+      "loss": 1.325,
+      "step": 11351
+    },
+    {
+      "epoch": 1.8163200000000002,
+      "grad_norm": 2.275624317202857,
+      "learning_rate": 7.137729400433111e-05,
+      "loss": 1.3199,
+      "step": 11352
+    },
+    {
+      "epoch": 1.8164799999999999,
+      "grad_norm": 2.041072896567108,
+      "learning_rate": 7.136074330080338e-05,
+      "loss": 1.2723,
+      "step": 11353
+    },
+    {
+      "epoch": 1.81664,
+      "grad_norm": 2.540272781748478,
+      "learning_rate": 7.134419345182938e-05,
+      "loss": 1.3463,
+      "step": 11354
+    },
+    {
+      "epoch": 1.8168,
+      "grad_norm": 2.2765078971623782,
+      "learning_rate": 7.132764445790286e-05,
+      "loss": 1.2326,
+      "step": 11355
+    },
+    {
+      "epoch": 1.81696,
+      "grad_norm": 1.9692816336340515,
+      "learning_rate": 7.131109631951766e-05,
+      "loss": 1.238,
+      "step": 11356
+    },
+    {
+      "epoch": 1.81712,
+      "grad_norm": 2.1064381480482,
+      "learning_rate": 7.129454903716752e-05,
+      "loss": 1.2816,
+      "step": 11357
+    },
+    {
+      "epoch": 1.81728,
+      "grad_norm": 2.06111645370169,
+      "learning_rate": 7.127800261134624e-05,
+      "loss": 1.2509,
+      "step": 11358
+    },
+    {
+      "epoch": 1.81744,
+      "grad_norm": 2.141067763437914,
+      "learning_rate": 7.126145704254748e-05,
+      "loss": 1.3079,
+      "step": 11359
+    },
+    {
+      "epoch": 1.8176,
+      "grad_norm": 1.930337818963564,
+      "learning_rate": 7.1244912331265e-05,
+      "loss": 1.3591,
+      "step": 11360
+    },
+    {
+      "epoch": 1.81776,
+      "grad_norm": 2.1116827947460077,
+      "learning_rate": 7.122836847799238e-05,
+      "loss": 1.2184,
+      "step": 11361
+    },
+    {
+      "epoch": 1.81792,
+      "grad_norm": 1.8556501771197644,
+      "learning_rate": 7.121182548322341e-05,
+      "loss": 1.2188,
+      "step": 11362
+    },
+    {
+      "epoch": 1.8180800000000001,
+      "grad_norm": 2.1172807441691694,
+      "learning_rate": 7.119528334745154e-05,
+      "loss": 1.2841,
+      "step": 11363
+    },
+    {
+      "epoch": 1.8182399999999999,
+      "grad_norm": 2.068572260184804,
+      "learning_rate": 7.117874207117048e-05,
+      "loss": 1.2539,
+      "step": 11364
+    },
+    {
+      "epoch": 1.8184,
+      "grad_norm": 2.1689825913156406,
+      "learning_rate": 7.116220165487373e-05,
+      "loss": 1.3199,
+      "step": 11365
+    },
+    {
+      "epoch": 1.81856,
+      "grad_norm": 2.013157189340842,
+      "learning_rate": 7.114566209905489e-05,
+      "loss": 1.2538,
+      "step": 11366
+    },
+    {
+      "epoch": 1.81872,
+      "grad_norm": 1.803149476223646,
+      "learning_rate": 7.112912340420741e-05,
+      "loss": 1.2044,
+      "step": 11367
+    },
+    {
+      "epoch": 1.81888,
+      "grad_norm": 2.228956083774956,
+      "learning_rate": 7.111258557082488e-05,
+      "loss": 1.2249,
+      "step": 11368
+    },
+    {
+      "epoch": 1.81904,
+      "grad_norm": 2.0869265088756763,
+      "learning_rate": 7.109604859940064e-05,
+      "loss": 1.1894,
+      "step": 11369
+    },
+    {
+      "epoch": 1.8192,
+      "grad_norm": 1.983110510465884,
+      "learning_rate": 7.107951249042824e-05,
+      "loss": 1.3181,
+      "step": 11370
+    },
+    {
+      "epoch": 1.81936,
+      "grad_norm": 2.2318728813034805,
+      "learning_rate": 7.106297724440101e-05,
+      "loss": 1.2049,
+      "step": 11371
+    },
+    {
+      "epoch": 1.81952,
+      "grad_norm": 1.90236807727521,
+      "learning_rate": 7.104644286181242e-05,
+      "loss": 1.3069,
+      "step": 11372
+    },
+    {
+      "epoch": 1.81968,
+      "grad_norm": 1.6807191893638613,
+      "learning_rate": 7.102990934315573e-05,
+      "loss": 1.2102,
+      "step": 11373
+    },
+    {
+      "epoch": 1.8198400000000001,
+      "grad_norm": 1.885650963973001,
+      "learning_rate": 7.101337668892436e-05,
+      "loss": 1.3049,
+      "step": 11374
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 2.0574469407228717,
+      "learning_rate": 7.099684489961157e-05,
+      "loss": 1.2497,
+      "step": 11375
+    },
+    {
+      "epoch": 1.82016,
+      "grad_norm": 2.31877612637252,
+      "learning_rate": 7.09803139757107e-05,
+      "loss": 1.2769,
+      "step": 11376
+    },
+    {
+      "epoch": 1.8203200000000002,
+      "grad_norm": 2.160078023516488,
+      "learning_rate": 7.096378391771494e-05,
+      "loss": 1.2447,
+      "step": 11377
+    },
+    {
+      "epoch": 1.8204799999999999,
+      "grad_norm": 1.8345866378029303,
+      "learning_rate": 7.094725472611761e-05,
+      "loss": 1.2409,
+      "step": 11378
+    },
+    {
+      "epoch": 1.82064,
+      "grad_norm": 2.1603057610092535,
+      "learning_rate": 7.093072640141181e-05,
+      "loss": 1.2276,
+      "step": 11379
+    },
+    {
+      "epoch": 1.8208,
+      "grad_norm": 2.7225603993726883,
+      "learning_rate": 7.091419894409082e-05,
+      "loss": 1.3277,
+      "step": 11380
+    },
+    {
+      "epoch": 1.82096,
+      "grad_norm": 2.0007351422451394,
+      "learning_rate": 7.08976723546477e-05,
+      "loss": 1.2741,
+      "step": 11381
+    },
+    {
+      "epoch": 1.82112,
+      "grad_norm": 2.0712203842068155,
+      "learning_rate": 7.088114663357567e-05,
+      "loss": 1.2449,
+      "step": 11382
+    },
+    {
+      "epoch": 1.82128,
+      "grad_norm": 2.7400065179246815,
+      "learning_rate": 7.086462178136777e-05,
+      "loss": 1.2304,
+      "step": 11383
+    },
+    {
+      "epoch": 1.82144,
+      "grad_norm": 2.280947268168966,
+      "learning_rate": 7.084809779851714e-05,
+      "loss": 1.2582,
+      "step": 11384
+    },
+    {
+      "epoch": 1.8216,
+      "grad_norm": 1.899290199189756,
+      "learning_rate": 7.083157468551673e-05,
+      "loss": 1.2797,
+      "step": 11385
+    },
+    {
+      "epoch": 1.82176,
+      "grad_norm": 1.9317322875894452,
+      "learning_rate": 7.081505244285968e-05,
+      "loss": 1.2656,
+      "step": 11386
+    },
+    {
+      "epoch": 1.82192,
+      "grad_norm": 2.2338944886090695,
+      "learning_rate": 7.07985310710389e-05,
+      "loss": 1.2717,
+      "step": 11387
+    },
+    {
+      "epoch": 1.8220800000000001,
+      "grad_norm": 1.9540722672714415,
+      "learning_rate": 7.078201057054746e-05,
+      "loss": 1.2657,
+      "step": 11388
+    },
+    {
+      "epoch": 1.8222399999999999,
+      "grad_norm": 1.7747397171563424,
+      "learning_rate": 7.076549094187819e-05,
+      "loss": 1.2344,
+      "step": 11389
+    },
+    {
+      "epoch": 1.8224,
+      "grad_norm": 2.828112645263847,
+      "learning_rate": 7.07489721855241e-05,
+      "loss": 1.3207,
+      "step": 11390
+    },
+    {
+      "epoch": 1.82256,
+      "grad_norm": 1.9519299608336258,
+      "learning_rate": 7.073245430197804e-05,
+      "loss": 1.2767,
+      "step": 11391
+    },
+    {
+      "epoch": 1.82272,
+      "grad_norm": 1.9723488536009461,
+      "learning_rate": 7.071593729173293e-05,
+      "loss": 1.2189,
+      "step": 11392
+    },
+    {
+      "epoch": 1.82288,
+      "grad_norm": 2.0525036904250813,
+      "learning_rate": 7.069942115528154e-05,
+      "loss": 1.2416,
+      "step": 11393
+    },
+    {
+      "epoch": 1.82304,
+      "grad_norm": 1.8920637324746468,
+      "learning_rate": 7.068290589311677e-05,
+      "loss": 1.3147,
+      "step": 11394
+    },
+    {
+      "epoch": 1.8232,
+      "grad_norm": 2.119871028138239,
+      "learning_rate": 7.066639150573132e-05,
+      "loss": 1.2566,
+      "step": 11395
+    },
+    {
+      "epoch": 1.82336,
+      "grad_norm": 1.9708753243815578,
+      "learning_rate": 7.064987799361806e-05,
+      "loss": 1.2154,
+      "step": 11396
+    },
+    {
+      "epoch": 1.82352,
+      "grad_norm": 1.9318074980658202,
+      "learning_rate": 7.063336535726961e-05,
+      "loss": 1.234,
+      "step": 11397
+    },
+    {
+      "epoch": 1.82368,
+      "grad_norm": 1.8444613892364645,
+      "learning_rate": 7.061685359717876e-05,
+      "loss": 1.2113,
+      "step": 11398
+    },
+    {
+      "epoch": 1.8238400000000001,
+      "grad_norm": 1.9359828137048345,
+      "learning_rate": 7.06003427138382e-05,
+      "loss": 1.2499,
+      "step": 11399
+    },
+    {
+      "epoch": 1.8239999999999998,
+      "grad_norm": 2.2405240904563453,
+      "learning_rate": 7.058383270774056e-05,
+      "loss": 1.3106,
+      "step": 11400
+    },
+    {
+      "epoch": 1.82416,
+      "grad_norm": 2.109224589125321,
+      "learning_rate": 7.056732357937852e-05,
+      "loss": 1.2613,
+      "step": 11401
+    },
+    {
+      "epoch": 1.8243200000000002,
+      "grad_norm": 1.9348648827129977,
+      "learning_rate": 7.055081532924462e-05,
+      "loss": 1.2529,
+      "step": 11402
+    },
+    {
+      "epoch": 1.8244799999999999,
+      "grad_norm": 2.337343051816114,
+      "learning_rate": 7.053430795783153e-05,
+      "loss": 1.2443,
+      "step": 11403
+    },
+    {
+      "epoch": 1.82464,
+      "grad_norm": 2.337129631672779,
+      "learning_rate": 7.05178014656317e-05,
+      "loss": 1.249,
+      "step": 11404
+    },
+    {
+      "epoch": 1.8248,
+      "grad_norm": 1.9081712333277938,
+      "learning_rate": 7.050129585313776e-05,
+      "loss": 1.2457,
+      "step": 11405
+    },
+    {
+      "epoch": 1.82496,
+      "grad_norm": 1.9392306672949389,
+      "learning_rate": 7.048479112084212e-05,
+      "loss": 1.319,
+      "step": 11406
+    },
+    {
+      "epoch": 1.82512,
+      "grad_norm": 2.2534420733593645,
+      "learning_rate": 7.046828726923733e-05,
+      "loss": 1.2109,
+      "step": 11407
+    },
+    {
+      "epoch": 1.82528,
+      "grad_norm": 1.8339157426966022,
+      "learning_rate": 7.045178429881579e-05,
+      "loss": 1.2013,
+      "step": 11408
+    },
+    {
+      "epoch": 1.82544,
+      "grad_norm": 2.5182227723024084,
+      "learning_rate": 7.043528221007e-05,
+      "loss": 1.279,
+      "step": 11409
+    },
+    {
+      "epoch": 1.8256000000000001,
+      "grad_norm": 2.287499925267215,
+      "learning_rate": 7.041878100349227e-05,
+      "loss": 1.3043,
+      "step": 11410
+    },
+    {
+      "epoch": 1.82576,
+      "grad_norm": 1.9882114572800573,
+      "learning_rate": 7.040228067957504e-05,
+      "loss": 1.2556,
+      "step": 11411
+    },
+    {
+      "epoch": 1.82592,
+      "grad_norm": 2.064666351007924,
+      "learning_rate": 7.038578123881059e-05,
+      "loss": 1.2464,
+      "step": 11412
+    },
+    {
+      "epoch": 1.8260800000000001,
+      "grad_norm": 2.093169216123032,
+      "learning_rate": 7.036928268169132e-05,
+      "loss": 1.2639,
+      "step": 11413
+    },
+    {
+      "epoch": 1.8262399999999999,
+      "grad_norm": 2.540400031522787,
+      "learning_rate": 7.035278500870944e-05,
+      "loss": 1.2921,
+      "step": 11414
+    },
+    {
+      "epoch": 1.8264,
+      "grad_norm": 1.9537810760612682,
+      "learning_rate": 7.033628822035725e-05,
+      "loss": 1.2636,
+      "step": 11415
+    },
+    {
+      "epoch": 1.82656,
+      "grad_norm": 1.9456487178764388,
+      "learning_rate": 7.0319792317127e-05,
+      "loss": 1.2958,
+      "step": 11416
+    },
+    {
+      "epoch": 1.82672,
+      "grad_norm": 2.6830588730136484,
+      "learning_rate": 7.030329729951093e-05,
+      "loss": 1.2088,
+      "step": 11417
+    },
+    {
+      "epoch": 1.82688,
+      "grad_norm": 2.3839765810737026,
+      "learning_rate": 7.028680316800115e-05,
+      "loss": 1.2389,
+      "step": 11418
+    },
+    {
+      "epoch": 1.82704,
+      "grad_norm": 1.8299515797413575,
+      "learning_rate": 7.027030992308992e-05,
+      "loss": 1.2272,
+      "step": 11419
+    },
+    {
+      "epoch": 1.8272,
+      "grad_norm": 2.6820009268429983,
+      "learning_rate": 7.025381756526928e-05,
+      "loss": 1.2602,
+      "step": 11420
+    },
+    {
+      "epoch": 1.82736,
+      "grad_norm": 2.0841351771995327,
+      "learning_rate": 7.023732609503142e-05,
+      "loss": 1.2752,
+      "step": 11421
+    },
+    {
+      "epoch": 1.82752,
+      "grad_norm": 2.2081544789478538,
+      "learning_rate": 7.022083551286834e-05,
+      "loss": 1.3287,
+      "step": 11422
+    },
+    {
+      "epoch": 1.82768,
+      "grad_norm": 1.7694049864597907,
+      "learning_rate": 7.020434581927214e-05,
+      "loss": 1.2264,
+      "step": 11423
+    },
+    {
+      "epoch": 1.8278400000000001,
+      "grad_norm": 2.402123102225793,
+      "learning_rate": 7.018785701473484e-05,
+      "loss": 1.2319,
+      "step": 11424
+    },
+    {
+      "epoch": 1.8279999999999998,
+      "grad_norm": 1.9712404793819507,
+      "learning_rate": 7.017136909974849e-05,
+      "loss": 1.2223,
+      "step": 11425
+    },
+    {
+      "epoch": 1.82816,
+      "grad_norm": 1.783059234157386,
+      "learning_rate": 7.015488207480497e-05,
+      "loss": 1.3236,
+      "step": 11426
+    },
+    {
+      "epoch": 1.82832,
+      "grad_norm": 3.1203369984037543,
+      "learning_rate": 7.013839594039631e-05,
+      "loss": 1.301,
+      "step": 11427
+    },
+    {
+      "epoch": 1.8284799999999999,
+      "grad_norm": 2.1645073985365917,
+      "learning_rate": 7.012191069701437e-05,
+      "loss": 1.2538,
+      "step": 11428
+    },
+    {
+      "epoch": 1.82864,
+      "grad_norm": 1.95662234899832,
+      "learning_rate": 7.010542634515114e-05,
+      "loss": 1.3289,
+      "step": 11429
+    },
+    {
+      "epoch": 1.8288,
+      "grad_norm": 2.2000153915259464,
+      "learning_rate": 7.008894288529836e-05,
+      "loss": 1.2007,
+      "step": 11430
+    },
+    {
+      "epoch": 1.82896,
+      "grad_norm": 2.0626818931069204,
+      "learning_rate": 7.007246031794796e-05,
+      "loss": 1.2453,
+      "step": 11431
+    },
+    {
+      "epoch": 1.82912,
+      "grad_norm": 1.8908183901790407,
+      "learning_rate": 7.005597864359172e-05,
+      "loss": 1.2233,
+      "step": 11432
+    },
+    {
+      "epoch": 1.82928,
+      "grad_norm": 2.076405266619584,
+      "learning_rate": 7.00394978627215e-05,
+      "loss": 1.2923,
+      "step": 11433
+    },
+    {
+      "epoch": 1.82944,
+      "grad_norm": 2.568871215112906,
+      "learning_rate": 7.002301797582896e-05,
+      "loss": 1.3236,
+      "step": 11434
+    },
+    {
+      "epoch": 1.8296000000000001,
+      "grad_norm": 1.9333771869430598,
+      "learning_rate": 7.000653898340591e-05,
+      "loss": 1.2148,
+      "step": 11435
+    },
+    {
+      "epoch": 1.82976,
+      "grad_norm": 2.0123024759463375,
+      "learning_rate": 6.9990060885944e-05,
+      "loss": 1.2335,
+      "step": 11436
+    },
+    {
+      "epoch": 1.82992,
+      "grad_norm": 2.445064829825236,
+      "learning_rate": 6.997358368393498e-05,
+      "loss": 1.2713,
+      "step": 11437
+    },
+    {
+      "epoch": 1.8300800000000002,
+      "grad_norm": 1.9288310103787547,
+      "learning_rate": 6.995710737787044e-05,
+      "loss": 1.2346,
+      "step": 11438
+    },
+    {
+      "epoch": 1.8302399999999999,
+      "grad_norm": 2.164443262637675,
+      "learning_rate": 6.994063196824204e-05,
+      "loss": 1.2137,
+      "step": 11439
+    },
+    {
+      "epoch": 1.8304,
+      "grad_norm": 2.0792183522394936,
+      "learning_rate": 6.992415745554137e-05,
+      "loss": 1.2529,
+      "step": 11440
+    },
+    {
+      "epoch": 1.83056,
+      "grad_norm": 1.9905124716182558,
+      "learning_rate": 6.990768384026006e-05,
+      "loss": 1.1697,
+      "step": 11441
+    },
+    {
+      "epoch": 1.83072,
+      "grad_norm": 1.836302811918715,
+      "learning_rate": 6.989121112288958e-05,
+      "loss": 1.2051,
+      "step": 11442
+    },
+    {
+      "epoch": 1.83088,
+      "grad_norm": 2.411176494396652,
+      "learning_rate": 6.987473930392152e-05,
+      "loss": 1.2557,
+      "step": 11443
+    },
+    {
+      "epoch": 1.83104,
+      "grad_norm": 2.3579347177903127,
+      "learning_rate": 6.98582683838473e-05,
+      "loss": 1.2608,
+      "step": 11444
+    },
+    {
+      "epoch": 1.8312,
+      "grad_norm": 2.477422556033281,
+      "learning_rate": 6.984179836315847e-05,
+      "loss": 1.3018,
+      "step": 11445
+    },
+    {
+      "epoch": 1.83136,
+      "grad_norm": 2.160557290958769,
+      "learning_rate": 6.982532924234639e-05,
+      "loss": 1.2318,
+      "step": 11446
+    },
+    {
+      "epoch": 1.83152,
+      "grad_norm": 2.177509822255071,
+      "learning_rate": 6.980886102190253e-05,
+      "loss": 1.1831,
+      "step": 11447
+    },
+    {
+      "epoch": 1.83168,
+      "grad_norm": 1.8587400360773965,
+      "learning_rate": 6.979239370231828e-05,
+      "loss": 1.2355,
+      "step": 11448
+    },
+    {
+      "epoch": 1.8318400000000001,
+      "grad_norm": 1.979533063028851,
+      "learning_rate": 6.9775927284085e-05,
+      "loss": 1.2919,
+      "step": 11449
+    },
+    {
+      "epoch": 1.8319999999999999,
+      "grad_norm": 2.054693923651083,
+      "learning_rate": 6.975946176769395e-05,
+      "loss": 1.2382,
+      "step": 11450
+    },
+    {
+      "epoch": 1.83216,
+      "grad_norm": 1.8914925648738208,
+      "learning_rate": 6.974299715363657e-05,
+      "loss": 1.2265,
+      "step": 11451
+    },
+    {
+      "epoch": 1.83232,
+      "grad_norm": 1.9228661981393027,
+      "learning_rate": 6.972653344240401e-05,
+      "loss": 1.3099,
+      "step": 11452
+    },
+    {
+      "epoch": 1.8324799999999999,
+      "grad_norm": 2.4360121103318417,
+      "learning_rate": 6.97100706344876e-05,
+      "loss": 1.2168,
+      "step": 11453
+    },
+    {
+      "epoch": 1.83264,
+      "grad_norm": 1.999069702678809,
+      "learning_rate": 6.969360873037853e-05,
+      "loss": 1.3079,
+      "step": 11454
+    },
+    {
+      "epoch": 1.8328,
+      "grad_norm": 1.9269458329753475,
+      "learning_rate": 6.967714773056804e-05,
+      "loss": 1.2781,
+      "step": 11455
+    },
+    {
+      "epoch": 1.83296,
+      "grad_norm": 2.8180927549122603,
+      "learning_rate": 6.966068763554722e-05,
+      "loss": 1.2946,
+      "step": 11456
+    },
+    {
+      "epoch": 1.83312,
+      "grad_norm": 2.235865875787127,
+      "learning_rate": 6.964422844580731e-05,
+      "loss": 1.2404,
+      "step": 11457
+    },
+    {
+      "epoch": 1.83328,
+      "grad_norm": 2.0355778027064515,
+      "learning_rate": 6.962777016183942e-05,
+      "loss": 1.2892,
+      "step": 11458
+    },
+    {
+      "epoch": 1.83344,
+      "grad_norm": 2.1518810508113884,
+      "learning_rate": 6.961131278413458e-05,
+      "loss": 1.2715,
+      "step": 11459
+    },
+    {
+      "epoch": 1.8336000000000001,
+      "grad_norm": 2.2994262719720253,
+      "learning_rate": 6.959485631318389e-05,
+      "loss": 1.2698,
+      "step": 11460
+    },
+    {
+      "epoch": 1.8337599999999998,
+      "grad_norm": 2.1323839729408527,
+      "learning_rate": 6.957840074947838e-05,
+      "loss": 1.2601,
+      "step": 11461
+    },
+    {
+      "epoch": 1.83392,
+      "grad_norm": 1.5597726230011832,
+      "learning_rate": 6.956194609350908e-05,
+      "loss": 1.2612,
+      "step": 11462
+    },
+    {
+      "epoch": 1.8340800000000002,
+      "grad_norm": 2.020923822675642,
+      "learning_rate": 6.954549234576692e-05,
+      "loss": 1.2148,
+      "step": 11463
+    },
+    {
+      "epoch": 1.8342399999999999,
+      "grad_norm": 2.265370074904109,
+      "learning_rate": 6.952903950674292e-05,
+      "loss": 1.2905,
+      "step": 11464
+    },
+    {
+      "epoch": 1.8344,
+      "grad_norm": 2.0544998849792253,
+      "learning_rate": 6.951258757692796e-05,
+      "loss": 1.2231,
+      "step": 11465
+    },
+    {
+      "epoch": 1.83456,
+      "grad_norm": 2.1518326974709248,
+      "learning_rate": 6.949613655681299e-05,
+      "loss": 1.3214,
+      "step": 11466
+    },
+    {
+      "epoch": 1.83472,
+      "grad_norm": 2.1289714911508377,
+      "learning_rate": 6.947968644688884e-05,
+      "loss": 1.2218,
+      "step": 11467
+    },
+    {
+      "epoch": 1.83488,
+      "grad_norm": 1.8775023408179594,
+      "learning_rate": 6.94632372476464e-05,
+      "loss": 1.2551,
+      "step": 11468
+    },
+    {
+      "epoch": 1.83504,
+      "grad_norm": 2.147279806280661,
+      "learning_rate": 6.944678895957642e-05,
+      "loss": 1.3012,
+      "step": 11469
+    },
+    {
+      "epoch": 1.8352,
+      "grad_norm": 2.10718377887536,
+      "learning_rate": 6.943034158316978e-05,
+      "loss": 1.2197,
+      "step": 11470
+    },
+    {
+      "epoch": 1.83536,
+      "grad_norm": 2.227957458213076,
+      "learning_rate": 6.941389511891716e-05,
+      "loss": 1.2957,
+      "step": 11471
+    },
+    {
+      "epoch": 1.83552,
+      "grad_norm": 1.8990484593251842,
+      "learning_rate": 6.939744956730936e-05,
+      "loss": 1.2107,
+      "step": 11472
+    },
+    {
+      "epoch": 1.83568,
+      "grad_norm": 1.8927322683135597,
+      "learning_rate": 6.938100492883706e-05,
+      "loss": 1.2331,
+      "step": 11473
+    },
+    {
+      "epoch": 1.8358400000000001,
+      "grad_norm": 1.8444022692351032,
+      "learning_rate": 6.936456120399099e-05,
+      "loss": 1.2441,
+      "step": 11474
+    },
+    {
+      "epoch": 1.8359999999999999,
+      "grad_norm": 2.0019539055081332,
+      "learning_rate": 6.934811839326173e-05,
+      "loss": 1.2645,
+      "step": 11475
+    },
+    {
+      "epoch": 1.83616,
+      "grad_norm": 1.9531596355645455,
+      "learning_rate": 6.933167649714e-05,
+      "loss": 1.2567,
+      "step": 11476
+    },
+    {
+      "epoch": 1.83632,
+      "grad_norm": 1.8542371701843818,
+      "learning_rate": 6.93152355161163e-05,
+      "loss": 1.2433,
+      "step": 11477
+    },
+    {
+      "epoch": 1.83648,
+      "grad_norm": 2.1691921909521086,
+      "learning_rate": 6.929879545068132e-05,
+      "loss": 1.2552,
+      "step": 11478
+    },
+    {
+      "epoch": 1.83664,
+      "grad_norm": 2.6762265617741683,
+      "learning_rate": 6.928235630132548e-05,
+      "loss": 1.2404,
+      "step": 11479
+    },
+    {
+      "epoch": 1.8368,
+      "grad_norm": 1.903317417597453,
+      "learning_rate": 6.926591806853941e-05,
+      "loss": 1.2121,
+      "step": 11480
+    },
+    {
+      "epoch": 1.83696,
+      "grad_norm": 2.431121713323863,
+      "learning_rate": 6.924948075281355e-05,
+      "loss": 1.2546,
+      "step": 11481
+    },
+    {
+      "epoch": 1.83712,
+      "grad_norm": 1.8536858208400115,
+      "learning_rate": 6.92330443546384e-05,
+      "loss": 1.2295,
+      "step": 11482
+    },
+    {
+      "epoch": 1.83728,
+      "grad_norm": 1.9213537106562741,
+      "learning_rate": 6.921660887450435e-05,
+      "loss": 1.2436,
+      "step": 11483
+    },
+    {
+      "epoch": 1.83744,
+      "grad_norm": 2.046875677675693,
+      "learning_rate": 6.920017431290184e-05,
+      "loss": 1.2378,
+      "step": 11484
+    },
+    {
+      "epoch": 1.8376000000000001,
+      "grad_norm": 1.951301840844347,
+      "learning_rate": 6.918374067032124e-05,
+      "loss": 1.3189,
+      "step": 11485
+    },
+    {
+      "epoch": 1.8377599999999998,
+      "grad_norm": 1.9284371648988259,
+      "learning_rate": 6.916730794725294e-05,
+      "loss": 1.2113,
+      "step": 11486
+    },
+    {
+      "epoch": 1.83792,
+      "grad_norm": 2.7679295806428392,
+      "learning_rate": 6.915087614418722e-05,
+      "loss": 1.2871,
+      "step": 11487
+    },
+    {
+      "epoch": 1.8380800000000002,
+      "grad_norm": 2.0959369755816155,
+      "learning_rate": 6.913444526161441e-05,
+      "loss": 1.2422,
+      "step": 11488
+    },
+    {
+      "epoch": 1.8382399999999999,
+      "grad_norm": 1.9909018624685715,
+      "learning_rate": 6.911801530002475e-05,
+      "loss": 1.2711,
+      "step": 11489
+    },
+    {
+      "epoch": 1.8384,
+      "grad_norm": 1.6889544512889283,
+      "learning_rate": 6.910158625990857e-05,
+      "loss": 1.2304,
+      "step": 11490
+    },
+    {
+      "epoch": 1.83856,
+      "grad_norm": 2.138406804263857,
+      "learning_rate": 6.908515814175598e-05,
+      "loss": 1.3354,
+      "step": 11491
+    },
+    {
+      "epoch": 1.83872,
+      "grad_norm": 1.804521570594894,
+      "learning_rate": 6.906873094605727e-05,
+      "loss": 1.2254,
+      "step": 11492
+    },
+    {
+      "epoch": 1.83888,
+      "grad_norm": 2.264629880806814,
+      "learning_rate": 6.905230467330251e-05,
+      "loss": 1.2533,
+      "step": 11493
+    },
+    {
+      "epoch": 1.83904,
+      "grad_norm": 2.0515340097046426,
+      "learning_rate": 6.903587932398193e-05,
+      "loss": 1.2385,
+      "step": 11494
+    },
+    {
+      "epoch": 1.8392,
+      "grad_norm": 2.719171181796359,
+      "learning_rate": 6.901945489858555e-05,
+      "loss": 1.3212,
+      "step": 11495
+    },
+    {
+      "epoch": 1.83936,
+      "grad_norm": 1.9636607412746157,
+      "learning_rate": 6.900303139760351e-05,
+      "loss": 1.2612,
+      "step": 11496
+    },
+    {
+      "epoch": 1.83952,
+      "grad_norm": 1.9793123887454485,
+      "learning_rate": 6.898660882152583e-05,
+      "loss": 1.305,
+      "step": 11497
+    },
+    {
+      "epoch": 1.83968,
+      "grad_norm": 2.105939652976314,
+      "learning_rate": 6.89701871708426e-05,
+      "loss": 1.2167,
+      "step": 11498
+    },
+    {
+      "epoch": 1.8398400000000001,
+      "grad_norm": 1.983565815349978,
+      "learning_rate": 6.89537664460437e-05,
+      "loss": 1.1937,
+      "step": 11499
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 2.0124330405636552,
+      "learning_rate": 6.893734664761924e-05,
+      "loss": 1.248,
+      "step": 11500
+    },
+    {
+      "epoch": 1.84016,
+      "grad_norm": 1.925190058066443,
+      "learning_rate": 6.892092777605905e-05,
+      "loss": 1.2518,
+      "step": 11501
+    },
+    {
+      "epoch": 1.84032,
+      "grad_norm": 1.8863912263886828,
+      "learning_rate": 6.890450983185312e-05,
+      "loss": 1.2551,
+      "step": 11502
+    },
+    {
+      "epoch": 1.84048,
+      "grad_norm": 2.002236237620234,
+      "learning_rate": 6.888809281549126e-05,
+      "loss": 1.2332,
+      "step": 11503
+    },
+    {
+      "epoch": 1.84064,
+      "grad_norm": 1.9174934333007698,
+      "learning_rate": 6.88716767274634e-05,
+      "loss": 1.2451,
+      "step": 11504
+    },
+    {
+      "epoch": 1.8408,
+      "grad_norm": 1.8764872913794737,
+      "learning_rate": 6.885526156825934e-05,
+      "loss": 1.2038,
+      "step": 11505
+    },
+    {
+      "epoch": 1.84096,
+      "grad_norm": 1.8424729846546677,
+      "learning_rate": 6.883884733836891e-05,
+      "loss": 1.3349,
+      "step": 11506
+    },
+    {
+      "epoch": 1.84112,
+      "grad_norm": 1.7440803082448793,
+      "learning_rate": 6.882243403828184e-05,
+      "loss": 1.2342,
+      "step": 11507
+    },
+    {
+      "epoch": 1.84128,
+      "grad_norm": 2.121872083297867,
+      "learning_rate": 6.880602166848796e-05,
+      "loss": 1.2902,
+      "step": 11508
+    },
+    {
+      "epoch": 1.84144,
+      "grad_norm": 1.8276160978152207,
+      "learning_rate": 6.878961022947688e-05,
+      "loss": 1.23,
+      "step": 11509
+    },
+    {
+      "epoch": 1.8416000000000001,
+      "grad_norm": 2.0799490640818195,
+      "learning_rate": 6.877319972173839e-05,
+      "loss": 1.2458,
+      "step": 11510
+    },
+    {
+      "epoch": 1.8417599999999998,
+      "grad_norm": 1.6632236286882158,
+      "learning_rate": 6.87567901457621e-05,
+      "loss": 1.2349,
+      "step": 11511
+    },
+    {
+      "epoch": 1.84192,
+      "grad_norm": 1.91809276095803,
+      "learning_rate": 6.874038150203768e-05,
+      "loss": 1.2093,
+      "step": 11512
+    },
+    {
+      "epoch": 1.8420800000000002,
+      "grad_norm": 2.22785866199359,
+      "learning_rate": 6.872397379105468e-05,
+      "loss": 1.2657,
+      "step": 11513
+    },
+    {
+      "epoch": 1.8422399999999999,
+      "grad_norm": 1.9062226599442422,
+      "learning_rate": 6.870756701330277e-05,
+      "loss": 1.2628,
+      "step": 11514
+    },
+    {
+      "epoch": 1.8424,
+      "grad_norm": 1.9062226599442422,
+      "learning_rate": 6.870756701330277e-05,
+      "loss": 1.227,
+      "step": 11515
+    },
+    {
+      "epoch": 1.84256,
+      "grad_norm": 2.4220682408596814,
+      "learning_rate": 6.869116116927146e-05,
+      "loss": 1.3369,
+      "step": 11516
+    },
+    {
+      "epoch": 1.84272,
+      "grad_norm": 2.217933380416251,
+      "learning_rate": 6.867475625945026e-05,
+      "loss": 1.3959,
+      "step": 11517
+    },
+    {
+      "epoch": 1.84288,
+      "grad_norm": 2.026405105411997,
+      "learning_rate": 6.865835228432874e-05,
+      "loss": 1.2449,
+      "step": 11518
+    },
+    {
+      "epoch": 1.84304,
+      "grad_norm": 2.511079464493939,
+      "learning_rate": 6.864194924439626e-05,
+      "loss": 1.272,
+      "step": 11519
+    },
+    {
+      "epoch": 1.8432,
+      "grad_norm": 2.3292242112220736,
+      "learning_rate": 6.862554714014236e-05,
+      "loss": 1.2901,
+      "step": 11520
+    },
+    {
+      "epoch": 1.84336,
+      "grad_norm": 1.7085024271065334,
+      "learning_rate": 6.860914597205638e-05,
+      "loss": 1.1656,
+      "step": 11521
+    },
+    {
+      "epoch": 1.84352,
+      "grad_norm": 2.471859038275067,
+      "learning_rate": 6.859274574062776e-05,
+      "loss": 1.2151,
+      "step": 11522
+    },
+    {
+      "epoch": 1.84368,
+      "grad_norm": 2.2168177784998133,
+      "learning_rate": 6.857634644634583e-05,
+      "loss": 1.3284,
+      "step": 11523
+    },
+    {
+      "epoch": 1.8438400000000001,
+      "grad_norm": 1.9716058480443015,
+      "learning_rate": 6.855994808969996e-05,
+      "loss": 1.2138,
+      "step": 11524
+    },
+    {
+      "epoch": 1.8439999999999999,
+      "grad_norm": 2.1215748132729964,
+      "learning_rate": 6.85435506711794e-05,
+      "loss": 1.2622,
+      "step": 11525
+    },
+    {
+      "epoch": 1.84416,
+      "grad_norm": 1.890591609358627,
+      "learning_rate": 6.85271541912735e-05,
+      "loss": 1.2335,
+      "step": 11526
+    },
+    {
+      "epoch": 1.84432,
+      "grad_norm": 2.310752376435117,
+      "learning_rate": 6.851075865047139e-05,
+      "loss": 1.2319,
+      "step": 11527
+    },
+    {
+      "epoch": 1.84448,
+      "grad_norm": 1.9988320535552249,
+      "learning_rate": 6.849436404926241e-05,
+      "loss": 1.2522,
+      "step": 11528
+    },
+    {
+      "epoch": 1.84464,
+      "grad_norm": 1.7944517342792314,
+      "learning_rate": 6.847797038813567e-05,
+      "loss": 1.1982,
+      "step": 11529
+    },
+    {
+      "epoch": 1.8448,
+      "grad_norm": 2.5525104550523494,
+      "learning_rate": 6.846157766758036e-05,
+      "loss": 1.2576,
+      "step": 11530
+    },
+    {
+      "epoch": 1.84496,
+      "grad_norm": 1.9908792025495188,
+      "learning_rate": 6.844518588808562e-05,
+      "loss": 1.2369,
+      "step": 11531
+    },
+    {
+      "epoch": 1.84512,
+      "grad_norm": 1.8691710566993218,
+      "learning_rate": 6.842879505014057e-05,
+      "loss": 1.2536,
+      "step": 11532
+    },
+    {
+      "epoch": 1.84528,
+      "grad_norm": 2.299771896713701,
+      "learning_rate": 6.841240515423424e-05,
+      "loss": 1.2455,
+      "step": 11533
+    },
+    {
+      "epoch": 1.84544,
+      "grad_norm": 2.089147074236884,
+      "learning_rate": 6.839601620085576e-05,
+      "loss": 1.264,
+      "step": 11534
+    },
+    {
+      "epoch": 1.8456000000000001,
+      "grad_norm": 2.1846173132010605,
+      "learning_rate": 6.837962819049407e-05,
+      "loss": 1.2457,
+      "step": 11535
+    },
+    {
+      "epoch": 1.8457599999999998,
+      "grad_norm": 2.068890249913892,
+      "learning_rate": 6.836324112363821e-05,
+      "loss": 1.2154,
+      "step": 11536
+    },
+    {
+      "epoch": 1.84592,
+      "grad_norm": 2.068890249913892,
+      "learning_rate": 6.836324112363821e-05,
+      "loss": 1.2211,
+      "step": 11537
+    },
+    {
+      "epoch": 1.8460800000000002,
+      "grad_norm": 4.109345830958428,
+      "learning_rate": 6.834685500077711e-05,
+      "loss": 1.251,
+      "step": 11538
+    },
+    {
+      "epoch": 1.8462399999999999,
+      "grad_norm": 1.8977799639599584,
+      "learning_rate": 6.833046982239977e-05,
+      "loss": 1.3024,
+      "step": 11539
+    },
+    {
+      "epoch": 1.8464,
+      "grad_norm": 1.8905111657252227,
+      "learning_rate": 6.831408558899501e-05,
+      "loss": 1.2673,
+      "step": 11540
+    },
+    {
+      "epoch": 1.84656,
+      "grad_norm": 2.5744373027466856,
+      "learning_rate": 6.829770230105184e-05,
+      "loss": 1.2498,
+      "step": 11541
+    },
+    {
+      "epoch": 1.84672,
+      "grad_norm": 2.479539802925768,
+      "learning_rate": 6.828131995905898e-05,
+      "loss": 1.2606,
+      "step": 11542
+    },
+    {
+      "epoch": 1.84688,
+      "grad_norm": 1.8493500973928658,
+      "learning_rate": 6.826493856350535e-05,
+      "loss": 1.1624,
+      "step": 11543
+    },
+    {
+      "epoch": 1.84704,
+      "grad_norm": 2.7588361733346964,
+      "learning_rate": 6.824855811487968e-05,
+      "loss": 1.2293,
+      "step": 11544
+    },
+    {
+      "epoch": 1.8472,
+      "grad_norm": 2.429640313272014,
+      "learning_rate": 6.82321786136708e-05,
+      "loss": 1.3602,
+      "step": 11545
+    },
+    {
+      "epoch": 1.8473600000000001,
+      "grad_norm": 2.0056881072134796,
+      "learning_rate": 6.821580006036739e-05,
+      "loss": 1.2534,
+      "step": 11546
+    },
+    {
+      "epoch": 1.84752,
+      "grad_norm": 1.922670830161631,
+      "learning_rate": 6.81994224554582e-05,
+      "loss": 1.2403,
+      "step": 11547
+    },
+    {
+      "epoch": 1.84768,
+      "grad_norm": 2.64940948070479,
+      "learning_rate": 6.818304579943188e-05,
+      "loss": 1.3354,
+      "step": 11548
+    },
+    {
+      "epoch": 1.8478400000000001,
+      "grad_norm": 2.099151856217977,
+      "learning_rate": 6.816667009277718e-05,
+      "loss": 1.2936,
+      "step": 11549
+    },
+    {
+      "epoch": 1.8479999999999999,
+      "grad_norm": 2.0853088971872116,
+      "learning_rate": 6.81502953359826e-05,
+      "loss": 1.219,
+      "step": 11550
+    },
+    {
+      "epoch": 1.84816,
+      "grad_norm": 2.6468855641958506,
+      "learning_rate": 6.813392152953684e-05,
+      "loss": 1.2345,
+      "step": 11551
+    },
+    {
+      "epoch": 1.84832,
+      "grad_norm": 2.8279746491219377,
+      "learning_rate": 6.811754867392839e-05,
+      "loss": 1.2385,
+      "step": 11552
+    },
+    {
+      "epoch": 1.84848,
+      "grad_norm": 1.9103194407902198,
+      "learning_rate": 6.810117676964587e-05,
+      "loss": 1.2992,
+      "step": 11553
+    },
+    {
+      "epoch": 1.84864,
+      "grad_norm": 1.9740867577962473,
+      "learning_rate": 6.80848058171777e-05,
+      "loss": 1.2229,
+      "step": 11554
+    },
+    {
+      "epoch": 1.8488,
+      "grad_norm": 2.0449082902977547,
+      "learning_rate": 6.806843581701246e-05,
+      "loss": 1.2876,
+      "step": 11555
+    },
+    {
+      "epoch": 1.84896,
+      "grad_norm": 1.930734580395984,
+      "learning_rate": 6.805206676963853e-05,
+      "loss": 1.2477,
+      "step": 11556
+    },
+    {
+      "epoch": 1.84912,
+      "grad_norm": 1.8513982392181003,
+      "learning_rate": 6.803569867554442e-05,
+      "loss": 1.2197,
+      "step": 11557
+    },
+    {
+      "epoch": 1.84928,
+      "grad_norm": 2.225930362135353,
+      "learning_rate": 6.801933153521844e-05,
+      "loss": 1.317,
+      "step": 11558
+    },
+    {
+      "epoch": 1.84944,
+      "grad_norm": 2.2782572573862017,
+      "learning_rate": 6.800296534914905e-05,
+      "loss": 1.2066,
+      "step": 11559
+    },
+    {
+      "epoch": 1.8496000000000001,
+      "grad_norm": 2.188953388569527,
+      "learning_rate": 6.79866001178245e-05,
+      "loss": 1.261,
+      "step": 11560
+    },
+    {
+      "epoch": 1.8497599999999998,
+      "grad_norm": 2.038396125233829,
+      "learning_rate": 6.797023584173322e-05,
+      "loss": 1.2529,
+      "step": 11561
+    },
+    {
+      "epoch": 1.84992,
+      "grad_norm": 2.074860428253641,
+      "learning_rate": 6.795387252136337e-05,
+      "loss": 1.2086,
+      "step": 11562
+    },
+    {
+      "epoch": 1.85008,
+      "grad_norm": 2.165129005929371,
+      "learning_rate": 6.793751015720329e-05,
+      "loss": 1.2836,
+      "step": 11563
+    },
+    {
+      "epoch": 1.8502399999999999,
+      "grad_norm": 1.947364986972419,
+      "learning_rate": 6.792114874974116e-05,
+      "loss": 1.2478,
+      "step": 11564
+    },
+    {
+      "epoch": 1.8504,
+      "grad_norm": 2.106449401493485,
+      "learning_rate": 6.790478829946526e-05,
+      "loss": 1.2389,
+      "step": 11565
+    },
+    {
+      "epoch": 1.85056,
+      "grad_norm": 2.4730221190264747,
+      "learning_rate": 6.788842880686367e-05,
+      "loss": 1.2288,
+      "step": 11566
+    },
+    {
+      "epoch": 1.85072,
+      "grad_norm": 1.9645555542118331,
+      "learning_rate": 6.787207027242462e-05,
+      "loss": 1.4122,
+      "step": 11567
+    },
+    {
+      "epoch": 1.85088,
+      "grad_norm": 2.138275002984744,
+      "learning_rate": 6.78557126966361e-05,
+      "loss": 1.2647,
+      "step": 11568
+    },
+    {
+      "epoch": 1.85104,
+      "grad_norm": 2.055894698427754,
+      "learning_rate": 6.783935607998635e-05,
+      "loss": 1.3107,
+      "step": 11569
+    },
+    {
+      "epoch": 1.8512,
+      "grad_norm": 2.195611864048128,
+      "learning_rate": 6.78230004229633e-05,
+      "loss": 1.226,
+      "step": 11570
+    },
+    {
+      "epoch": 1.8513600000000001,
+      "grad_norm": 1.8025436704507927,
+      "learning_rate": 6.780664572605506e-05,
+      "loss": 1.2192,
+      "step": 11571
+    },
+    {
+      "epoch": 1.85152,
+      "grad_norm": 1.9831224630136701,
+      "learning_rate": 6.779029198974957e-05,
+      "loss": 1.2205,
+      "step": 11572
+    },
+    {
+      "epoch": 1.85168,
+      "grad_norm": 1.8712129518450344,
+      "learning_rate": 6.777393921453487e-05,
+      "loss": 1.3274,
+      "step": 11573
+    },
+    {
+      "epoch": 1.8518400000000002,
+      "grad_norm": 1.8009957385689306,
+      "learning_rate": 6.775758740089883e-05,
+      "loss": 1.2664,
+      "step": 11574
+    },
+    {
+      "epoch": 1.8519999999999999,
+      "grad_norm": 1.7952209672104003,
+      "learning_rate": 6.77412365493294e-05,
+      "loss": 1.1827,
+      "step": 11575
+    },
+    {
+      "epoch": 1.85216,
+      "grad_norm": 1.9292031384943957,
+      "learning_rate": 6.772488666031451e-05,
+      "loss": 1.249,
+      "step": 11576
+    },
+    {
+      "epoch": 1.85232,
+      "grad_norm": 1.816286141612241,
+      "learning_rate": 6.770853773434194e-05,
+      "loss": 1.2807,
+      "step": 11577
+    },
+    {
+      "epoch": 1.85248,
+      "grad_norm": 2.1627353986819395,
+      "learning_rate": 6.769218977189957e-05,
+      "loss": 1.2193,
+      "step": 11578
+    },
+    {
+      "epoch": 1.85264,
+      "grad_norm": 2.0551490959116,
+      "learning_rate": 6.767584277347516e-05,
+      "loss": 1.2442,
+      "step": 11579
+    },
+    {
+      "epoch": 1.8528,
+      "grad_norm": 1.910319409748786,
+      "learning_rate": 6.76594967395565e-05,
+      "loss": 1.2555,
+      "step": 11580
+    },
+    {
+      "epoch": 1.85296,
+      "grad_norm": 1.9845327573459985,
+      "learning_rate": 6.764315167063134e-05,
+      "loss": 1.2872,
+      "step": 11581
+    },
+    {
+      "epoch": 1.85312,
+      "grad_norm": 2.1139638529631823,
+      "learning_rate": 6.762680756718739e-05,
+      "loss": 1.2869,
+      "step": 11582
+    },
+    {
+      "epoch": 1.85328,
+      "grad_norm": 1.9163334495980395,
+      "learning_rate": 6.761046442971231e-05,
+      "loss": 1.2522,
+      "step": 11583
+    },
+    {
+      "epoch": 1.85344,
+      "grad_norm": 3.0711855478327754,
+      "learning_rate": 6.75941222586938e-05,
+      "loss": 1.2155,
+      "step": 11584
+    },
+    {
+      "epoch": 1.8536000000000001,
+      "grad_norm": 1.8335009596810237,
+      "learning_rate": 6.757778105461944e-05,
+      "loss": 1.2142,
+      "step": 11585
+    },
+    {
+      "epoch": 1.8537599999999999,
+      "grad_norm": 1.9517158835305763,
+      "learning_rate": 6.756144081797689e-05,
+      "loss": 1.182,
+      "step": 11586
+    },
+    {
+      "epoch": 1.85392,
+      "grad_norm": 2.0934071954243563,
+      "learning_rate": 6.754510154925362e-05,
+      "loss": 1.2759,
+      "step": 11587
+    },
+    {
+      "epoch": 1.85408,
+      "grad_norm": 2.434695367758684,
+      "learning_rate": 6.752876324893726e-05,
+      "loss": 1.2064,
+      "step": 11588
+    },
+    {
+      "epoch": 1.8542399999999999,
+      "grad_norm": 2.0197898717567697,
+      "learning_rate": 6.751242591751527e-05,
+      "loss": 1.2407,
+      "step": 11589
+    },
+    {
+      "epoch": 1.8544,
+      "grad_norm": 2.1778383333764704,
+      "learning_rate": 6.749608955547517e-05,
+      "loss": 1.2881,
+      "step": 11590
+    },
+    {
+      "epoch": 1.85456,
+      "grad_norm": 2.0289013364575017,
+      "learning_rate": 6.747975416330439e-05,
+      "loss": 1.205,
+      "step": 11591
+    },
+    {
+      "epoch": 1.85472,
+      "grad_norm": 1.9417181295263526,
+      "learning_rate": 6.746341974149037e-05,
+      "loss": 1.1923,
+      "step": 11592
+    },
+    {
+      "epoch": 1.85488,
+      "grad_norm": 1.7320542033175086,
+      "learning_rate": 6.744708629052047e-05,
+      "loss": 1.2463,
+      "step": 11593
+    },
+    {
+      "epoch": 1.85504,
+      "grad_norm": 2.3440472304973476,
+      "learning_rate": 6.74307538108821e-05,
+      "loss": 1.1909,
+      "step": 11594
+    },
+    {
+      "epoch": 1.8552,
+      "grad_norm": 2.0865690220848077,
+      "learning_rate": 6.741442230306256e-05,
+      "loss": 1.2434,
+      "step": 11595
+    },
+    {
+      "epoch": 1.8553600000000001,
+      "grad_norm": 2.0077211944035054,
+      "learning_rate": 6.73980917675492e-05,
+      "loss": 1.2824,
+      "step": 11596
+    },
+    {
+      "epoch": 1.8555199999999998,
+      "grad_norm": 2.323821372860007,
+      "learning_rate": 6.738176220482925e-05,
+      "loss": 1.2048,
+      "step": 11597
+    },
+    {
+      "epoch": 1.85568,
+      "grad_norm": 2.7389093826300095,
+      "learning_rate": 6.736543361539002e-05,
+      "loss": 1.2171,
+      "step": 11598
+    },
+    {
+      "epoch": 1.8558400000000002,
+      "grad_norm": 2.0894865431261187,
+      "learning_rate": 6.734910599971865e-05,
+      "loss": 1.2442,
+      "step": 11599
+    },
+    {
+      "epoch": 1.8559999999999999,
+      "grad_norm": 2.124241154668811,
+      "learning_rate": 6.733277935830243e-05,
+      "loss": 1.2304,
+      "step": 11600
+    },
+    {
+      "epoch": 1.85616,
+      "grad_norm": 2.7820554914289732,
+      "learning_rate": 6.731645369162843e-05,
+      "loss": 1.2412,
+      "step": 11601
+    },
+    {
+      "epoch": 1.85632,
+      "grad_norm": 2.2212072592862646,
+      "learning_rate": 6.730012900018388e-05,
+      "loss": 1.3106,
+      "step": 11602
+    },
+    {
+      "epoch": 1.85648,
+      "grad_norm": 2.126493487818017,
+      "learning_rate": 6.728380528445578e-05,
+      "loss": 1.2327,
+      "step": 11603
+    },
+    {
+      "epoch": 1.85664,
+      "grad_norm": 2.621233880316085,
+      "learning_rate": 6.726748254493126e-05,
+      "loss": 1.2089,
+      "step": 11604
+    },
+    {
+      "epoch": 1.8568,
+      "grad_norm": 2.281209771000826,
+      "learning_rate": 6.725116078209739e-05,
+      "loss": 1.3289,
+      "step": 11605
+    },
+    {
+      "epoch": 1.85696,
+      "grad_norm": 2.167678543618077,
+      "learning_rate": 6.723483999644115e-05,
+      "loss": 1.2842,
+      "step": 11606
+    },
+    {
+      "epoch": 1.85712,
+      "grad_norm": 2.0746625397703147,
+      "learning_rate": 6.721852018844953e-05,
+      "loss": 1.2779,
+      "step": 11607
+    },
+    {
+      "epoch": 1.85728,
+      "grad_norm": 3.0083095184114,
+      "learning_rate": 6.720220135860953e-05,
+      "loss": 1.2335,
+      "step": 11608
+    },
+    {
+      "epoch": 1.85744,
+      "grad_norm": 1.8673672486614985,
+      "learning_rate": 6.718588350740802e-05,
+      "loss": 1.3219,
+      "step": 11609
+    },
+    {
+      "epoch": 1.8576000000000001,
+      "grad_norm": 1.8385023152464788,
+      "learning_rate": 6.716956663533196e-05,
+      "loss": 1.2463,
+      "step": 11610
+    },
+    {
+      "epoch": 1.8577599999999999,
+      "grad_norm": 1.8901811102545445,
+      "learning_rate": 6.715325074286816e-05,
+      "loss": 1.2568,
+      "step": 11611
+    },
+    {
+      "epoch": 1.85792,
+      "grad_norm": 2.2723050983933746,
+      "learning_rate": 6.71369358305035e-05,
+      "loss": 1.1957,
+      "step": 11612
+    },
+    {
+      "epoch": 1.85808,
+      "grad_norm": 2.360961253539026,
+      "learning_rate": 6.712062189872481e-05,
+      "loss": 1.2571,
+      "step": 11613
+    },
+    {
+      "epoch": 1.85824,
+      "grad_norm": 1.9598257245401554,
+      "learning_rate": 6.710430894801885e-05,
+      "loss": 1.2299,
+      "step": 11614
+    },
+    {
+      "epoch": 1.8584,
+      "grad_norm": 2.1700613274687917,
+      "learning_rate": 6.708799697887236e-05,
+      "loss": 1.2774,
+      "step": 11615
+    },
+    {
+      "epoch": 1.85856,
+      "grad_norm": 2.0142254676922566,
+      "learning_rate": 6.707168599177212e-05,
+      "loss": 1.2154,
+      "step": 11616
+    },
+    {
+      "epoch": 1.85872,
+      "grad_norm": 2.017893749734541,
+      "learning_rate": 6.705537598720472e-05,
+      "loss": 1.1999,
+      "step": 11617
+    },
+    {
+      "epoch": 1.85888,
+      "grad_norm": 2.042089704278091,
+      "learning_rate": 6.703906696565696e-05,
+      "loss": 1.2599,
+      "step": 11618
+    },
+    {
+      "epoch": 1.85904,
+      "grad_norm": 1.9393348011586464,
+      "learning_rate": 6.702275892761538e-05,
+      "loss": 1.1898,
+      "step": 11619
+    },
+    {
+      "epoch": 1.8592,
+      "grad_norm": 2.0756192114625764,
+      "learning_rate": 6.700645187356662e-05,
+      "loss": 1.2314,
+      "step": 11620
+    },
+    {
+      "epoch": 1.8593600000000001,
+      "grad_norm": 2.127527364096815,
+      "learning_rate": 6.699014580399725e-05,
+      "loss": 1.2464,
+      "step": 11621
+    },
+    {
+      "epoch": 1.8595199999999998,
+      "grad_norm": 2.430983179164445,
+      "learning_rate": 6.697384071939388e-05,
+      "loss": 1.2074,
+      "step": 11622
+    },
+    {
+      "epoch": 1.85968,
+      "grad_norm": 2.201497585026808,
+      "learning_rate": 6.695753662024291e-05,
+      "loss": 1.2321,
+      "step": 11623
+    },
+    {
+      "epoch": 1.8598400000000002,
+      "grad_norm": 2.0400900200614167,
+      "learning_rate": 6.694123350703097e-05,
+      "loss": 1.2654,
+      "step": 11624
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 2.2393595002250875,
+      "learning_rate": 6.69249313802444e-05,
+      "loss": 1.2044,
+      "step": 11625
+    },
+    {
+      "epoch": 1.86016,
+      "grad_norm": 1.788091241866051,
+      "learning_rate": 6.69086302403697e-05,
+      "loss": 1.1851,
+      "step": 11626
+    },
+    {
+      "epoch": 1.86032,
+      "grad_norm": 2.2270784486629838,
+      "learning_rate": 6.689233008789324e-05,
+      "loss": 1.216,
+      "step": 11627
+    },
+    {
+      "epoch": 1.86048,
+      "grad_norm": 1.8584862049974202,
+      "learning_rate": 6.687603092330142e-05,
+      "loss": 1.2564,
+      "step": 11628
+    },
+    {
+      "epoch": 1.86064,
+      "grad_norm": 2.1910552631231073,
+      "learning_rate": 6.685973274708055e-05,
+      "loss": 1.2294,
+      "step": 11629
+    },
+    {
+      "epoch": 1.8608,
+      "grad_norm": 1.9894720886771136,
+      "learning_rate": 6.684343555971702e-05,
+      "loss": 1.2013,
+      "step": 11630
+    },
+    {
+      "epoch": 1.86096,
+      "grad_norm": 2.008884998673261,
+      "learning_rate": 6.6827139361697e-05,
+      "loss": 1.3156,
+      "step": 11631
+    },
+    {
+      "epoch": 1.86112,
+      "grad_norm": 2.1420771481472207,
+      "learning_rate": 6.681084415350684e-05,
+      "loss": 1.2398,
+      "step": 11632
+    },
+    {
+      "epoch": 1.86128,
+      "grad_norm": 2.656410525545274,
+      "learning_rate": 6.67945499356327e-05,
+      "loss": 1.2616,
+      "step": 11633
+    },
+    {
+      "epoch": 1.86144,
+      "grad_norm": 1.8032331078328407,
+      "learning_rate": 6.67782567085608e-05,
+      "loss": 1.2352,
+      "step": 11634
+    },
+    {
+      "epoch": 1.8616000000000001,
+      "grad_norm": 2.10626117834696,
+      "learning_rate": 6.676196447277735e-05,
+      "loss": 1.2594,
+      "step": 11635
+    },
+    {
+      "epoch": 1.8617599999999999,
+      "grad_norm": 2.1354647564367433,
+      "learning_rate": 6.674567322876841e-05,
+      "loss": 1.3105,
+      "step": 11636
+    },
+    {
+      "epoch": 1.86192,
+      "grad_norm": 2.5816457337479197,
+      "learning_rate": 6.672938297702014e-05,
+      "loss": 1.299,
+      "step": 11637
+    },
+    {
+      "epoch": 1.86208,
+      "grad_norm": 2.0136134508474512,
+      "learning_rate": 6.67130937180186e-05,
+      "loss": 1.2365,
+      "step": 11638
+    },
+    {
+      "epoch": 1.86224,
+      "grad_norm": 2.172384148476005,
+      "learning_rate": 6.669680545224986e-05,
+      "loss": 1.2413,
+      "step": 11639
+    },
+    {
+      "epoch": 1.8624,
+      "grad_norm": 2.6438717911183915,
+      "learning_rate": 6.668051818019987e-05,
+      "loss": 1.2474,
+      "step": 11640
+    },
+    {
+      "epoch": 1.86256,
+      "grad_norm": 2.2684817405695035,
+      "learning_rate": 6.666423190235471e-05,
+      "loss": 1.2489,
+      "step": 11641
+    },
+    {
+      "epoch": 1.86272,
+      "grad_norm": 2.0079978127689033,
+      "learning_rate": 6.664794661920026e-05,
+      "loss": 1.219,
+      "step": 11642
+    },
+    {
+      "epoch": 1.86288,
+      "grad_norm": 3.3969600215497047,
+      "learning_rate": 6.663166233122253e-05,
+      "loss": 1.2968,
+      "step": 11643
+    },
+    {
+      "epoch": 1.86304,
+      "grad_norm": 2.0208206009910805,
+      "learning_rate": 6.661537903890733e-05,
+      "loss": 1.2202,
+      "step": 11644
+    },
+    {
+      "epoch": 1.8632,
+      "grad_norm": 2.1917407468409733,
+      "learning_rate": 6.659909674274059e-05,
+      "loss": 1.2777,
+      "step": 11645
+    },
+    {
+      "epoch": 1.8633600000000001,
+      "grad_norm": 2.094335542348349,
+      "learning_rate": 6.658281544320814e-05,
+      "loss": 1.2117,
+      "step": 11646
+    },
+    {
+      "epoch": 1.8635199999999998,
+      "grad_norm": 1.9263050656627956,
+      "learning_rate": 6.656653514079578e-05,
+      "loss": 1.2647,
+      "step": 11647
+    },
+    {
+      "epoch": 1.86368,
+      "grad_norm": 2.0492318356824835,
+      "learning_rate": 6.655025583598927e-05,
+      "loss": 1.2293,
+      "step": 11648
+    },
+    {
+      "epoch": 1.8638400000000002,
+      "grad_norm": 2.0266081341794817,
+      "learning_rate": 6.653397752927441e-05,
+      "loss": 1.2508,
+      "step": 11649
+    },
+    {
+      "epoch": 1.8639999999999999,
+      "grad_norm": 1.953547859878233,
+      "learning_rate": 6.651770022113688e-05,
+      "loss": 1.1946,
+      "step": 11650
+    },
+    {
+      "epoch": 1.86416,
+      "grad_norm": 2.0775826676164835,
+      "learning_rate": 6.650142391206243e-05,
+      "loss": 1.2359,
+      "step": 11651
+    },
+    {
+      "epoch": 1.86432,
+      "grad_norm": 1.9350302618065136,
+      "learning_rate": 6.648514860253662e-05,
+      "loss": 1.2645,
+      "step": 11652
+    },
+    {
+      "epoch": 1.86448,
+      "grad_norm": 2.098260729458225,
+      "learning_rate": 6.646887429304516e-05,
+      "loss": 1.2455,
+      "step": 11653
+    },
+    {
+      "epoch": 1.86464,
+      "grad_norm": 2.0004128957825835,
+      "learning_rate": 6.645260098407364e-05,
+      "loss": 1.2915,
+      "step": 11654
+    },
+    {
+      "epoch": 1.8648,
+      "grad_norm": 2.271158154726798,
+      "learning_rate": 6.643632867610762e-05,
+      "loss": 1.3395,
+      "step": 11655
+    },
+    {
+      "epoch": 1.86496,
+      "grad_norm": 1.7694347724251254,
+      "learning_rate": 6.642005736963262e-05,
+      "loss": 1.245,
+      "step": 11656
+    },
+    {
+      "epoch": 1.8651200000000001,
+      "grad_norm": 2.5468465119686843,
+      "learning_rate": 6.640378706513421e-05,
+      "loss": 1.2265,
+      "step": 11657
+    },
+    {
+      "epoch": 1.86528,
+      "grad_norm": 2.1995495979739994,
+      "learning_rate": 6.638751776309782e-05,
+      "loss": 1.2243,
+      "step": 11658
+    },
+    {
+      "epoch": 1.86544,
+      "grad_norm": 1.930626267313218,
+      "learning_rate": 6.637124946400894e-05,
+      "loss": 1.2572,
+      "step": 11659
+    },
+    {
+      "epoch": 1.8656000000000001,
+      "grad_norm": 2.4763459706805055,
+      "learning_rate": 6.635498216835296e-05,
+      "loss": 1.2272,
+      "step": 11660
+    },
+    {
+      "epoch": 1.8657599999999999,
+      "grad_norm": 2.102439187105528,
+      "learning_rate": 6.633871587661528e-05,
+      "loss": 1.2919,
+      "step": 11661
+    },
+    {
+      "epoch": 1.86592,
+      "grad_norm": 2.1334927917210593,
+      "learning_rate": 6.632245058928128e-05,
+      "loss": 1.2383,
+      "step": 11662
+    },
+    {
+      "epoch": 1.86608,
+      "grad_norm": 2.4749535821390456,
+      "learning_rate": 6.630618630683629e-05,
+      "loss": 1.2484,
+      "step": 11663
+    },
+    {
+      "epoch": 1.86624,
+      "grad_norm": 2.0100351562831347,
+      "learning_rate": 6.628992302976558e-05,
+      "loss": 1.2296,
+      "step": 11664
+    },
+    {
+      "epoch": 1.8664,
+      "grad_norm": 2.005048794592142,
+      "learning_rate": 6.627366075855447e-05,
+      "loss": 1.1978,
+      "step": 11665
+    },
+    {
+      "epoch": 1.86656,
+      "grad_norm": 2.2157008576315955,
+      "learning_rate": 6.625739949368814e-05,
+      "loss": 1.3205,
+      "step": 11666
+    },
+    {
+      "epoch": 1.86672,
+      "grad_norm": 2.0803752222403733,
+      "learning_rate": 6.624113923565189e-05,
+      "loss": 1.2738,
+      "step": 11667
+    },
+    {
+      "epoch": 1.86688,
+      "grad_norm": 1.9413192953247453,
+      "learning_rate": 6.622487998493083e-05,
+      "loss": 1.236,
+      "step": 11668
+    },
+    {
+      "epoch": 1.86704,
+      "grad_norm": 2.2256144790495225,
+      "learning_rate": 6.620862174201012e-05,
+      "loss": 1.2299,
+      "step": 11669
+    },
+    {
+      "epoch": 1.8672,
+      "grad_norm": 2.014231161997279,
+      "learning_rate": 6.619236450737493e-05,
+      "loss": 1.2992,
+      "step": 11670
+    },
+    {
+      "epoch": 1.8673600000000001,
+      "grad_norm": 2.2775357586649454,
+      "learning_rate": 6.61761082815103e-05,
+      "loss": 1.2381,
+      "step": 11671
+    },
+    {
+      "epoch": 1.8675199999999998,
+      "grad_norm": 1.8933516024396821,
+      "learning_rate": 6.615985306490131e-05,
+      "loss": 1.2663,
+      "step": 11672
+    },
+    {
+      "epoch": 1.86768,
+      "grad_norm": 1.9716825243572622,
+      "learning_rate": 6.614359885803302e-05,
+      "loss": 1.2444,
+      "step": 11673
+    },
+    {
+      "epoch": 1.86784,
+      "grad_norm": 2.034997152130842,
+      "learning_rate": 6.612734566139037e-05,
+      "loss": 1.2325,
+      "step": 11674
+    },
+    {
+      "epoch": 1.8679999999999999,
+      "grad_norm": 2.0836932202498235,
+      "learning_rate": 6.61110934754584e-05,
+      "loss": 1.243,
+      "step": 11675
+    },
+    {
+      "epoch": 1.86816,
+      "grad_norm": 2.088592883035201,
+      "learning_rate": 6.609484230072199e-05,
+      "loss": 1.2407,
+      "step": 11676
+    },
+    {
+      "epoch": 1.86832,
+      "grad_norm": 1.8535747394054394,
+      "learning_rate": 6.607859213766609e-05,
+      "loss": 1.2208,
+      "step": 11677
+    },
+    {
+      "epoch": 1.86848,
+      "grad_norm": 1.8296905367987768,
+      "learning_rate": 6.606234298677558e-05,
+      "loss": 1.1942,
+      "step": 11678
+    },
+    {
+      "epoch": 1.86864,
+      "grad_norm": 2.209025358913212,
+      "learning_rate": 6.604609484853531e-05,
+      "loss": 1.2581,
+      "step": 11679
+    },
+    {
+      "epoch": 1.8688,
+      "grad_norm": 2.1087651841995174,
+      "learning_rate": 6.602984772343006e-05,
+      "loss": 1.2467,
+      "step": 11680
+    },
+    {
+      "epoch": 1.86896,
+      "grad_norm": 1.8306341192287563,
+      "learning_rate": 6.60136016119447e-05,
+      "loss": 1.1585,
+      "step": 11681
+    },
+    {
+      "epoch": 1.8691200000000001,
+      "grad_norm": 2.030938583195338,
+      "learning_rate": 6.599735651456391e-05,
+      "loss": 1.2147,
+      "step": 11682
+    },
+    {
+      "epoch": 1.86928,
+      "grad_norm": 1.9796088992238197,
+      "learning_rate": 6.59811124317725e-05,
+      "loss": 1.2148,
+      "step": 11683
+    },
+    {
+      "epoch": 1.86944,
+      "grad_norm": 2.1461361900777,
+      "learning_rate": 6.596486936405506e-05,
+      "loss": 1.2784,
+      "step": 11684
+    },
+    {
+      "epoch": 1.8696000000000002,
+      "grad_norm": 1.9591727633877276,
+      "learning_rate": 6.594862731189636e-05,
+      "loss": 1.1741,
+      "step": 11685
+    },
+    {
+      "epoch": 1.8697599999999999,
+      "grad_norm": 1.959441307936662,
+      "learning_rate": 6.5932386275781e-05,
+      "loss": 1.1775,
+      "step": 11686
+    },
+    {
+      "epoch": 1.86992,
+      "grad_norm": 2.078557583271092,
+      "learning_rate": 6.59161462561936e-05,
+      "loss": 1.1898,
+      "step": 11687
+    },
+    {
+      "epoch": 1.87008,
+      "grad_norm": 2.5254785795141794,
+      "learning_rate": 6.58999072536187e-05,
+      "loss": 1.2307,
+      "step": 11688
+    },
+    {
+      "epoch": 1.87024,
+      "grad_norm": 1.8437803695555277,
+      "learning_rate": 6.588366926854092e-05,
+      "loss": 1.237,
+      "step": 11689
+    },
+    {
+      "epoch": 1.8704,
+      "grad_norm": 2.3208384507411948,
+      "learning_rate": 6.586743230144468e-05,
+      "loss": 1.196,
+      "step": 11690
+    },
+    {
+      "epoch": 1.87056,
+      "grad_norm": 2.4344414801965866,
+      "learning_rate": 6.585119635281459e-05,
+      "loss": 1.2629,
+      "step": 11691
+    },
+    {
+      "epoch": 1.87072,
+      "grad_norm": 2.117403059410469,
+      "learning_rate": 6.583496142313499e-05,
+      "loss": 1.2261,
+      "step": 11692
+    },
+    {
+      "epoch": 1.87088,
+      "grad_norm": 1.894100937063961,
+      "learning_rate": 6.581872751289036e-05,
+      "loss": 1.2141,
+      "step": 11693
+    },
+    {
+      "epoch": 1.87104,
+      "grad_norm": 2.3687555552436885,
+      "learning_rate": 6.580249462256513e-05,
+      "loss": 1.2975,
+      "step": 11694
+    },
+    {
+      "epoch": 1.8712,
+      "grad_norm": 2.052017258299675,
+      "learning_rate": 6.578626275264359e-05,
+      "loss": 1.2543,
+      "step": 11695
+    },
+    {
+      "epoch": 1.8713600000000001,
+      "grad_norm": 1.9154230874511793,
+      "learning_rate": 6.577003190361014e-05,
+      "loss": 1.163,
+      "step": 11696
+    },
+    {
+      "epoch": 1.8715199999999999,
+      "grad_norm": 2.133656730670283,
+      "learning_rate": 6.575380207594904e-05,
+      "loss": 1.213,
+      "step": 11697
+    },
+    {
+      "epoch": 1.87168,
+      "grad_norm": 2.4782919615010517,
+      "learning_rate": 6.573757327014461e-05,
+      "loss": 1.2227,
+      "step": 11698
+    },
+    {
+      "epoch": 1.87184,
+      "grad_norm": 2.0691649464248973,
+      "learning_rate": 6.572134548668104e-05,
+      "loss": 1.2606,
+      "step": 11699
+    },
+    {
+      "epoch": 1.8719999999999999,
+      "grad_norm": 1.8483730215389569,
+      "learning_rate": 6.570511872604261e-05,
+      "loss": 1.2386,
+      "step": 11700
+    },
+    {
+      "epoch": 1.87216,
+      "grad_norm": 2.1524486357985593,
+      "learning_rate": 6.568889298871342e-05,
+      "loss": 1.2382,
+      "step": 11701
+    },
+    {
+      "epoch": 1.87232,
+      "grad_norm": 2.0843436710471956,
+      "learning_rate": 6.567266827517771e-05,
+      "loss": 1.217,
+      "step": 11702
+    },
+    {
+      "epoch": 1.87248,
+      "grad_norm": 1.8617140798020155,
+      "learning_rate": 6.565644458591953e-05,
+      "loss": 1.2622,
+      "step": 11703
+    },
+    {
+      "epoch": 1.87264,
+      "grad_norm": 2.228871868721989,
+      "learning_rate": 6.564022192142301e-05,
+      "loss": 1.2515,
+      "step": 11704
+    },
+    {
+      "epoch": 1.8728,
+      "grad_norm": 2.636476737225187,
+      "learning_rate": 6.562400028217217e-05,
+      "loss": 1.2807,
+      "step": 11705
+    },
+    {
+      "epoch": 1.87296,
+      "grad_norm": 1.7113554541977969,
+      "learning_rate": 6.560777966865113e-05,
+      "loss": 1.1751,
+      "step": 11706
+    },
+    {
+      "epoch": 1.8731200000000001,
+      "grad_norm": 2.8951607936466353,
+      "learning_rate": 6.559156008134377e-05,
+      "loss": 1.2922,
+      "step": 11707
+    },
+    {
+      "epoch": 1.8732799999999998,
+      "grad_norm": 2.1013333913190206,
+      "learning_rate": 6.557534152073417e-05,
+      "loss": 1.2314,
+      "step": 11708
+    },
+    {
+      "epoch": 1.87344,
+      "grad_norm": 2.097257470715156,
+      "learning_rate": 6.555912398730617e-05,
+      "loss": 1.3331,
+      "step": 11709
+    },
+    {
+      "epoch": 1.8736000000000002,
+      "grad_norm": 1.8233286111331009,
+      "learning_rate": 6.554290748154377e-05,
+      "loss": 1.2683,
+      "step": 11710
+    },
+    {
+      "epoch": 1.8737599999999999,
+      "grad_norm": 2.0096087034340098,
+      "learning_rate": 6.552669200393077e-05,
+      "loss": 1.226,
+      "step": 11711
+    },
+    {
+      "epoch": 1.87392,
+      "grad_norm": 1.8256853233120953,
+      "learning_rate": 6.551047755495106e-05,
+      "loss": 1.2287,
+      "step": 11712
+    },
+    {
+      "epoch": 1.87408,
+      "grad_norm": 2.3025800876507767,
+      "learning_rate": 6.549426413508841e-05,
+      "loss": 1.2784,
+      "step": 11713
+    },
+    {
+      "epoch": 1.87424,
+      "grad_norm": 1.8627474238701163,
+      "learning_rate": 6.54780517448267e-05,
+      "loss": 1.2694,
+      "step": 11714
+    },
+    {
+      "epoch": 1.8744,
+      "grad_norm": 1.9295105030690871,
+      "learning_rate": 6.546184038464958e-05,
+      "loss": 1.2652,
+      "step": 11715
+    },
+    {
+      "epoch": 1.87456,
+      "grad_norm": 2.330312541792262,
+      "learning_rate": 6.544563005504084e-05,
+      "loss": 1.2387,
+      "step": 11716
+    },
+    {
+      "epoch": 1.87472,
+      "grad_norm": 2.219534760528536,
+      "learning_rate": 6.542942075648411e-05,
+      "loss": 1.2285,
+      "step": 11717
+    },
+    {
+      "epoch": 1.87488,
+      "grad_norm": 1.8075921387723413,
+      "learning_rate": 6.541321248946314e-05,
+      "loss": 1.2348,
+      "step": 11718
+    },
+    {
+      "epoch": 1.87504,
+      "grad_norm": 1.7960013731004139,
+      "learning_rate": 6.539700525446147e-05,
+      "loss": 1.2238,
+      "step": 11719
+    },
+    {
+      "epoch": 1.8752,
+      "grad_norm": 2.2002762175318913,
+      "learning_rate": 6.538079905196275e-05,
+      "loss": 1.2214,
+      "step": 11720
+    },
+    {
+      "epoch": 1.8753600000000001,
+      "grad_norm": 1.9196210586479792,
+      "learning_rate": 6.536459388245053e-05,
+      "loss": 1.179,
+      "step": 11721
+    },
+    {
+      "epoch": 1.8755199999999999,
+      "grad_norm": 2.497624198753668,
+      "learning_rate": 6.53483897464084e-05,
+      "loss": 1.1618,
+      "step": 11722
+    },
+    {
+      "epoch": 1.87568,
+      "grad_norm": 2.222534216457938,
+      "learning_rate": 6.53321866443198e-05,
+      "loss": 1.2482,
+      "step": 11723
+    },
+    {
+      "epoch": 1.87584,
+      "grad_norm": 1.9317105317266878,
+      "learning_rate": 6.531598457666825e-05,
+      "loss": 1.2552,
+      "step": 11724
+    },
+    {
+      "epoch": 1.876,
+      "grad_norm": 2.069860051616139,
+      "learning_rate": 6.529978354393716e-05,
+      "loss": 1.2567,
+      "step": 11725
+    },
+    {
+      "epoch": 1.87616,
+      "grad_norm": 2.5768642544345894,
+      "learning_rate": 6.528358354660996e-05,
+      "loss": 1.2541,
+      "step": 11726
+    },
+    {
+      "epoch": 1.87632,
+      "grad_norm": 2.072204509001515,
+      "learning_rate": 6.526738458517005e-05,
+      "loss": 1.2101,
+      "step": 11727
+    },
+    {
+      "epoch": 1.87648,
+      "grad_norm": 2.0761147853292905,
+      "learning_rate": 6.525118666010079e-05,
+      "loss": 1.2163,
+      "step": 11728
+    },
+    {
+      "epoch": 1.87664,
+      "grad_norm": 2.160729105204736,
+      "learning_rate": 6.523498977188544e-05,
+      "loss": 1.2719,
+      "step": 11729
+    },
+    {
+      "epoch": 1.8768,
+      "grad_norm": 2.0019823945699944,
+      "learning_rate": 6.52187939210074e-05,
+      "loss": 1.235,
+      "step": 11730
+    },
+    {
+      "epoch": 1.87696,
+      "grad_norm": 2.0012048025757223,
+      "learning_rate": 6.520259910794982e-05,
+      "loss": 1.2306,
+      "step": 11731
+    },
+    {
+      "epoch": 1.8771200000000001,
+      "grad_norm": 1.910745439873446,
+      "learning_rate": 6.518640533319603e-05,
+      "loss": 1.2511,
+      "step": 11732
+    },
+    {
+      "epoch": 1.8772799999999998,
+      "grad_norm": 2.3195110324971946,
+      "learning_rate": 6.51702125972291e-05,
+      "loss": 1.2516,
+      "step": 11733
+    },
+    {
+      "epoch": 1.87744,
+      "grad_norm": 1.9935615378815295,
+      "learning_rate": 6.515402090053233e-05,
+      "loss": 1.2011,
+      "step": 11734
+    },
+    {
+      "epoch": 1.8776000000000002,
+      "grad_norm": 1.8880045833218901,
+      "learning_rate": 6.513783024358879e-05,
+      "loss": 1.1789,
+      "step": 11735
+    },
+    {
+      "epoch": 1.8777599999999999,
+      "grad_norm": 2.0444035006637415,
+      "learning_rate": 6.512164062688158e-05,
+      "loss": 1.2251,
+      "step": 11736
+    },
+    {
+      "epoch": 1.87792,
+      "grad_norm": 1.937457114253839,
+      "learning_rate": 6.51054520508938e-05,
+      "loss": 1.1996,
+      "step": 11737
+    },
+    {
+      "epoch": 1.87808,
+      "grad_norm": 1.8650857913668615,
+      "learning_rate": 6.50892645161085e-05,
+      "loss": 1.2648,
+      "step": 11738
+    },
+    {
+      "epoch": 1.87824,
+      "grad_norm": 2.2698542512072035,
+      "learning_rate": 6.507307802300863e-05,
+      "loss": 1.2771,
+      "step": 11739
+    },
+    {
+      "epoch": 1.8784,
+      "grad_norm": 1.7677215033115534,
+      "learning_rate": 6.505689257207729e-05,
+      "loss": 1.2318,
+      "step": 11740
+    },
+    {
+      "epoch": 1.87856,
+      "grad_norm": 2.3026965670817483,
+      "learning_rate": 6.50407081637973e-05,
+      "loss": 1.2464,
+      "step": 11741
+    },
+    {
+      "epoch": 1.87872,
+      "grad_norm": 1.9483037018759666,
+      "learning_rate": 6.502452479865166e-05,
+      "loss": 1.2521,
+      "step": 11742
+    },
+    {
+      "epoch": 1.87888,
+      "grad_norm": 1.996219219753525,
+      "learning_rate": 6.500834247712322e-05,
+      "loss": 1.2839,
+      "step": 11743
+    },
+    {
+      "epoch": 1.87904,
+      "grad_norm": 1.9181400976801746,
+      "learning_rate": 6.499216119969487e-05,
+      "loss": 1.2619,
+      "step": 11744
+    },
+    {
+      "epoch": 1.8792,
+      "grad_norm": 2.1002163665900366,
+      "learning_rate": 6.49759809668494e-05,
+      "loss": 1.2487,
+      "step": 11745
+    },
+    {
+      "epoch": 1.8793600000000001,
+      "grad_norm": 2.263491147968287,
+      "learning_rate": 6.495980177906965e-05,
+      "loss": 1.2133,
+      "step": 11746
+    },
+    {
+      "epoch": 1.8795199999999999,
+      "grad_norm": 2.0776181209949978,
+      "learning_rate": 6.494362363683833e-05,
+      "loss": 1.2667,
+      "step": 11747
+    },
+    {
+      "epoch": 1.87968,
+      "grad_norm": 2.06697365538723,
+      "learning_rate": 6.492744654063823e-05,
+      "loss": 1.1972,
+      "step": 11748
+    },
+    {
+      "epoch": 1.87984,
+      "grad_norm": 1.9990344044479342,
+      "learning_rate": 6.491127049095197e-05,
+      "loss": 1.2894,
+      "step": 11749
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 2.2627111744314408,
+      "learning_rate": 6.489509548826227e-05,
+      "loss": 1.3436,
+      "step": 11750
+    },
+    {
+      "epoch": 1.88016,
+      "grad_norm": 1.9970353286099323,
+      "learning_rate": 6.487892153305182e-05,
+      "loss": 1.2582,
+      "step": 11751
+    },
+    {
+      "epoch": 1.88032,
+      "grad_norm": 1.8445867726524636,
+      "learning_rate": 6.486274862580313e-05,
+      "loss": 1.1939,
+      "step": 11752
+    },
+    {
+      "epoch": 1.88048,
+      "grad_norm": 1.9569735902896286,
+      "learning_rate": 6.484657676699883e-05,
+      "loss": 1.2085,
+      "step": 11753
+    },
+    {
+      "epoch": 1.88064,
+      "grad_norm": 1.899102265206817,
+      "learning_rate": 6.483040595712146e-05,
+      "loss": 1.3289,
+      "step": 11754
+    },
+    {
+      "epoch": 1.8808,
+      "grad_norm": 1.7880105831245439,
+      "learning_rate": 6.481423619665354e-05,
+      "loss": 1.2794,
+      "step": 11755
+    },
+    {
+      "epoch": 1.88096,
+      "grad_norm": 1.932397981343941,
+      "learning_rate": 6.479806748607753e-05,
+      "loss": 1.1923,
+      "step": 11756
+    },
+    {
+      "epoch": 1.8811200000000001,
+      "grad_norm": 1.9988640137087006,
+      "learning_rate": 6.478189982587591e-05,
+      "loss": 1.2623,
+      "step": 11757
+    },
+    {
+      "epoch": 1.8812799999999998,
+      "grad_norm": 2.2224871150709564,
+      "learning_rate": 6.476573321653105e-05,
+      "loss": 1.3236,
+      "step": 11758
+    },
+    {
+      "epoch": 1.88144,
+      "grad_norm": 2.0160897079082445,
+      "learning_rate": 6.47495676585254e-05,
+      "loss": 1.2224,
+      "step": 11759
+    },
+    {
+      "epoch": 1.8816000000000002,
+      "grad_norm": 2.209100408816434,
+      "learning_rate": 6.473340315234127e-05,
+      "loss": 1.2562,
+      "step": 11760
+    },
+    {
+      "epoch": 1.8817599999999999,
+      "grad_norm": 2.0092688791701034,
+      "learning_rate": 6.4717239698461e-05,
+      "loss": 1.175,
+      "step": 11761
+    },
+    {
+      "epoch": 1.88192,
+      "grad_norm": 2.174324324665122,
+      "learning_rate": 6.470107729736686e-05,
+      "loss": 1.2955,
+      "step": 11762
+    },
+    {
+      "epoch": 1.88208,
+      "grad_norm": 1.5525457282881692,
+      "learning_rate": 6.46849159495412e-05,
+      "loss": 1.2704,
+      "step": 11763
+    },
+    {
+      "epoch": 1.88224,
+      "grad_norm": 2.051779832368887,
+      "learning_rate": 6.466875565546612e-05,
+      "loss": 1.2617,
+      "step": 11764
+    },
+    {
+      "epoch": 1.8824,
+      "grad_norm": 1.6581110553036094,
+      "learning_rate": 6.465259641562395e-05,
+      "loss": 1.2151,
+      "step": 11765
+    },
+    {
+      "epoch": 1.88256,
+      "grad_norm": 2.11453752315871,
+      "learning_rate": 6.463643823049674e-05,
+      "loss": 1.2659,
+      "step": 11766
+    },
+    {
+      "epoch": 1.88272,
+      "grad_norm": 2.374689688595992,
+      "learning_rate": 6.462028110056673e-05,
+      "loss": 1.2608,
+      "step": 11767
+    },
+    {
+      "epoch": 1.88288,
+      "grad_norm": 1.838226361685178,
+      "learning_rate": 6.460412502631594e-05,
+      "loss": 1.2586,
+      "step": 11768
+    },
+    {
+      "epoch": 1.88304,
+      "grad_norm": 2.1312569691377017,
+      "learning_rate": 6.458797000822652e-05,
+      "loss": 1.2198,
+      "step": 11769
+    },
+    {
+      "epoch": 1.8832,
+      "grad_norm": 2.1162414702885046,
+      "learning_rate": 6.457181604678042e-05,
+      "loss": 1.1876,
+      "step": 11770
+    },
+    {
+      "epoch": 1.8833600000000001,
+      "grad_norm": 1.7151169357436702,
+      "learning_rate": 6.455566314245978e-05,
+      "loss": 1.1927,
+      "step": 11771
+    },
+    {
+      "epoch": 1.8835199999999999,
+      "grad_norm": 1.7798249026361914,
+      "learning_rate": 6.453951129574644e-05,
+      "loss": 1.2392,
+      "step": 11772
+    },
+    {
+      "epoch": 1.88368,
+      "grad_norm": 2.198822433749176,
+      "learning_rate": 6.452336050712245e-05,
+      "loss": 1.1933,
+      "step": 11773
+    },
+    {
+      "epoch": 1.88384,
+      "grad_norm": 2.0214543465040262,
+      "learning_rate": 6.450721077706965e-05,
+      "loss": 1.2434,
+      "step": 11774
+    },
+    {
+      "epoch": 1.884,
+      "grad_norm": 1.9204627723515368,
+      "learning_rate": 6.449106210607e-05,
+      "loss": 1.2729,
+      "step": 11775
+    },
+    {
+      "epoch": 1.88416,
+      "grad_norm": 1.9531875721065133,
+      "learning_rate": 6.447491449460529e-05,
+      "loss": 1.2575,
+      "step": 11776
+    },
+    {
+      "epoch": 1.88432,
+      "grad_norm": 2.018916372429281,
+      "learning_rate": 6.445876794315739e-05,
+      "loss": 1.2126,
+      "step": 11777
+    },
+    {
+      "epoch": 1.88448,
+      "grad_norm": 1.8540554866417214,
+      "learning_rate": 6.444262245220803e-05,
+      "loss": 1.3141,
+      "step": 11778
+    },
+    {
+      "epoch": 1.88464,
+      "grad_norm": 2.152939448678783,
+      "learning_rate": 6.442647802223904e-05,
+      "loss": 1.2375,
+      "step": 11779
+    },
+    {
+      "epoch": 1.8848,
+      "grad_norm": 1.8457122067544673,
+      "learning_rate": 6.441033465373207e-05,
+      "loss": 1.19,
+      "step": 11780
+    },
+    {
+      "epoch": 1.88496,
+      "grad_norm": 1.8594201113721645,
+      "learning_rate": 6.439419234716891e-05,
+      "loss": 1.2685,
+      "step": 11781
+    },
+    {
+      "epoch": 1.8851200000000001,
+      "grad_norm": 1.888881253929464,
+      "learning_rate": 6.437805110303111e-05,
+      "loss": 1.2424,
+      "step": 11782
+    },
+    {
+      "epoch": 1.8852799999999998,
+      "grad_norm": 2.2544608153397174,
+      "learning_rate": 6.436191092180042e-05,
+      "loss": 1.2582,
+      "step": 11783
+    },
+    {
+      "epoch": 1.88544,
+      "grad_norm": 2.3012514753476125,
+      "learning_rate": 6.43457718039583e-05,
+      "loss": 1.2514,
+      "step": 11784
+    },
+    {
+      "epoch": 1.8856000000000002,
+      "grad_norm": 2.1209223702688496,
+      "learning_rate": 6.432963374998645e-05,
+      "loss": 1.2368,
+      "step": 11785
+    },
+    {
+      "epoch": 1.8857599999999999,
+      "grad_norm": 1.9963881361032236,
+      "learning_rate": 6.431349676036633e-05,
+      "loss": 1.2101,
+      "step": 11786
+    },
+    {
+      "epoch": 1.88592,
+      "grad_norm": 1.8918474943094934,
+      "learning_rate": 6.429736083557949e-05,
+      "loss": 1.2666,
+      "step": 11787
+    },
+    {
+      "epoch": 1.88608,
+      "grad_norm": 2.0434385868120857,
+      "learning_rate": 6.428122597610735e-05,
+      "loss": 1.2313,
+      "step": 11788
+    },
+    {
+      "epoch": 1.88624,
+      "grad_norm": 1.7961895725116652,
+      "learning_rate": 6.426509218243142e-05,
+      "loss": 1.1948,
+      "step": 11789
+    },
+    {
+      "epoch": 1.8864,
+      "grad_norm": 2.092369400823018,
+      "learning_rate": 6.424895945503301e-05,
+      "loss": 1.2495,
+      "step": 11790
+    },
+    {
+      "epoch": 1.88656,
+      "grad_norm": 1.999335473526272,
+      "learning_rate": 6.423282779439361e-05,
+      "loss": 1.2428,
+      "step": 11791
+    },
+    {
+      "epoch": 1.88672,
+      "grad_norm": 1.8853334622994105,
+      "learning_rate": 6.421669720099447e-05,
+      "loss": 1.2877,
+      "step": 11792
+    },
+    {
+      "epoch": 1.8868800000000001,
+      "grad_norm": 1.915327000778218,
+      "learning_rate": 6.420056767531697e-05,
+      "loss": 1.2012,
+      "step": 11793
+    },
+    {
+      "epoch": 1.88704,
+      "grad_norm": 2.2232441809919763,
+      "learning_rate": 6.418443921784233e-05,
+      "loss": 1.2876,
+      "step": 11794
+    },
+    {
+      "epoch": 1.8872,
+      "grad_norm": 2.224016812017464,
+      "learning_rate": 6.416831182905188e-05,
+      "loss": 1.3012,
+      "step": 11795
+    },
+    {
+      "epoch": 1.8873600000000001,
+      "grad_norm": 1.8017209216123558,
+      "learning_rate": 6.415218550942675e-05,
+      "loss": 1.3113,
+      "step": 11796
+    },
+    {
+      "epoch": 1.8875199999999999,
+      "grad_norm": 1.9448207862946065,
+      "learning_rate": 6.413606025944823e-05,
+      "loss": 1.1702,
+      "step": 11797
+    },
+    {
+      "epoch": 1.88768,
+      "grad_norm": 1.6669492396833028,
+      "learning_rate": 6.411993607959734e-05,
+      "loss": 1.2561,
+      "step": 11798
+    },
+    {
+      "epoch": 1.88784,
+      "grad_norm": 2.0704287240307555,
+      "learning_rate": 6.410381297035533e-05,
+      "loss": 1.2834,
+      "step": 11799
+    },
+    {
+      "epoch": 1.888,
+      "grad_norm": 2.00957609143022,
+      "learning_rate": 6.40876909322032e-05,
+      "loss": 1.2517,
+      "step": 11800
+    },
+    {
+      "epoch": 1.88816,
+      "grad_norm": 1.8075582351299022,
+      "learning_rate": 6.407156996562205e-05,
+      "loss": 1.2539,
+      "step": 11801
+    },
+    {
+      "epoch": 1.88832,
+      "grad_norm": 1.919018959495834,
+      "learning_rate": 6.40554500710929e-05,
+      "loss": 1.2291,
+      "step": 11802
+    },
+    {
+      "epoch": 1.88848,
+      "grad_norm": 1.8361304414867565,
+      "learning_rate": 6.403933124909677e-05,
+      "loss": 1.239,
+      "step": 11803
+    },
+    {
+      "epoch": 1.88864,
+      "grad_norm": 1.8799229839061162,
+      "learning_rate": 6.402321350011455e-05,
+      "loss": 1.2719,
+      "step": 11804
+    },
+    {
+      "epoch": 1.8888,
+      "grad_norm": 1.9336013151843252,
+      "learning_rate": 6.400709682462727e-05,
+      "loss": 1.2261,
+      "step": 11805
+    },
+    {
+      "epoch": 1.88896,
+      "grad_norm": 2.021692305209107,
+      "learning_rate": 6.399098122311571e-05,
+      "loss": 1.2905,
+      "step": 11806
+    },
+    {
+      "epoch": 1.8891200000000001,
+      "grad_norm": 1.9500668588274142,
+      "learning_rate": 6.397486669606086e-05,
+      "loss": 1.204,
+      "step": 11807
+    },
+    {
+      "epoch": 1.8892799999999998,
+      "grad_norm": 1.8504033505013455,
+      "learning_rate": 6.395875324394344e-05,
+      "loss": 1.3071,
+      "step": 11808
+    },
+    {
+      "epoch": 1.88944,
+      "grad_norm": 1.9521559178313161,
+      "learning_rate": 6.394264086724433e-05,
+      "loss": 1.2434,
+      "step": 11809
+    },
+    {
+      "epoch": 1.8896,
+      "grad_norm": 2.164570677802104,
+      "learning_rate": 6.392652956644427e-05,
+      "loss": 1.3005,
+      "step": 11810
+    },
+    {
+      "epoch": 1.8897599999999999,
+      "grad_norm": 2.038068707357993,
+      "learning_rate": 6.3910419342024e-05,
+      "loss": 1.3382,
+      "step": 11811
+    },
+    {
+      "epoch": 1.88992,
+      "grad_norm": 1.8237247611534855,
+      "learning_rate": 6.389431019446425e-05,
+      "loss": 1.2999,
+      "step": 11812
+    },
+    {
+      "epoch": 1.89008,
+      "grad_norm": 1.9766074715757134,
+      "learning_rate": 6.387820212424565e-05,
+      "loss": 1.2426,
+      "step": 11813
+    },
+    {
+      "epoch": 1.89024,
+      "grad_norm": 2.197986094810292,
+      "learning_rate": 6.386209513184888e-05,
+      "loss": 1.2703,
+      "step": 11814
+    },
+    {
+      "epoch": 1.8904,
+      "grad_norm": 1.7781820009585412,
+      "learning_rate": 6.384598921775453e-05,
+      "loss": 1.1996,
+      "step": 11815
+    },
+    {
+      "epoch": 1.89056,
+      "grad_norm": 1.926099048128838,
+      "learning_rate": 6.382988438244319e-05,
+      "loss": 1.2603,
+      "step": 11816
+    },
+    {
+      "epoch": 1.89072,
+      "grad_norm": 2.0923726379248437,
+      "learning_rate": 6.381378062639536e-05,
+      "loss": 1.2349,
+      "step": 11817
+    },
+    {
+      "epoch": 1.8908800000000001,
+      "grad_norm": 1.9388451880714561,
+      "learning_rate": 6.379767795009161e-05,
+      "loss": 1.2303,
+      "step": 11818
+    },
+    {
+      "epoch": 1.89104,
+      "grad_norm": 1.8086734458857747,
+      "learning_rate": 6.378157635401239e-05,
+      "loss": 1.1993,
+      "step": 11819
+    },
+    {
+      "epoch": 1.8912,
+      "grad_norm": 1.848254994520406,
+      "learning_rate": 6.376547583863818e-05,
+      "loss": 1.2609,
+      "step": 11820
+    },
+    {
+      "epoch": 1.8913600000000002,
+      "grad_norm": 2.175316292394081,
+      "learning_rate": 6.374937640444933e-05,
+      "loss": 1.2405,
+      "step": 11821
+    },
+    {
+      "epoch": 1.8915199999999999,
+      "grad_norm": 1.6752570217168625,
+      "learning_rate": 6.373327805192631e-05,
+      "loss": 1.2561,
+      "step": 11822
+    },
+    {
+      "epoch": 1.89168,
+      "grad_norm": 1.8235233113547067,
+      "learning_rate": 6.371718078154938e-05,
+      "loss": 1.1974,
+      "step": 11823
+    },
+    {
+      "epoch": 1.89184,
+      "grad_norm": 2.248997828793139,
+      "learning_rate": 6.370108459379897e-05,
+      "loss": 1.2629,
+      "step": 11824
+    },
+    {
+      "epoch": 1.892,
+      "grad_norm": 2.1200726357515647,
+      "learning_rate": 6.368498948915524e-05,
+      "loss": 1.2685,
+      "step": 11825
+    },
+    {
+      "epoch": 1.89216,
+      "grad_norm": 1.9673150516947806,
+      "learning_rate": 6.366889546809853e-05,
+      "loss": 1.3197,
+      "step": 11826
+    },
+    {
+      "epoch": 1.89232,
+      "grad_norm": 2.2918953998494356,
+      "learning_rate": 6.365280253110904e-05,
+      "loss": 1.2927,
+      "step": 11827
+    },
+    {
+      "epoch": 1.89248,
+      "grad_norm": 1.7852164078679365,
+      "learning_rate": 6.363671067866697e-05,
+      "loss": 1.2435,
+      "step": 11828
+    },
+    {
+      "epoch": 1.89264,
+      "grad_norm": 1.7453031015615317,
+      "learning_rate": 6.362061991125243e-05,
+      "loss": 1.2323,
+      "step": 11829
+    },
+    {
+      "epoch": 1.8928,
+      "grad_norm": 1.9159386822662101,
+      "learning_rate": 6.360453022934564e-05,
+      "loss": 1.247,
+      "step": 11830
+    },
+    {
+      "epoch": 1.89296,
+      "grad_norm": 2.362156769495419,
+      "learning_rate": 6.358844163342658e-05,
+      "loss": 1.3213,
+      "step": 11831
+    },
+    {
+      "epoch": 1.8931200000000001,
+      "grad_norm": 2.000171740608968,
+      "learning_rate": 6.35723541239754e-05,
+      "loss": 1.2486,
+      "step": 11832
+    },
+    {
+      "epoch": 1.8932799999999999,
+      "grad_norm": 1.7852597108873618,
+      "learning_rate": 6.355626770147207e-05,
+      "loss": 1.2418,
+      "step": 11833
+    },
+    {
+      "epoch": 1.89344,
+      "grad_norm": 1.9072303145293472,
+      "learning_rate": 6.354018236639662e-05,
+      "loss": 1.2507,
+      "step": 11834
+    },
+    {
+      "epoch": 1.8936,
+      "grad_norm": 1.8840765055910325,
+      "learning_rate": 6.352409811922897e-05,
+      "loss": 1.2693,
+      "step": 11835
+    },
+    {
+      "epoch": 1.8937599999999999,
+      "grad_norm": 2.0110033659793873,
+      "learning_rate": 6.350801496044914e-05,
+      "loss": 1.2919,
+      "step": 11836
+    },
+    {
+      "epoch": 1.89392,
+      "grad_norm": 1.9472993527314977,
+      "learning_rate": 6.349193289053692e-05,
+      "loss": 1.2015,
+      "step": 11837
+    },
+    {
+      "epoch": 1.89408,
+      "grad_norm": 1.9680705134804353,
+      "learning_rate": 6.347585190997226e-05,
+      "loss": 1.3388,
+      "step": 11838
+    },
+    {
+      "epoch": 1.89424,
+      "grad_norm": 2.4675475244588627,
+      "learning_rate": 6.345977201923492e-05,
+      "loss": 1.2414,
+      "step": 11839
+    },
+    {
+      "epoch": 1.8944,
+      "grad_norm": 1.8190525853396324,
+      "learning_rate": 6.344369321880478e-05,
+      "loss": 1.2444,
+      "step": 11840
+    },
+    {
+      "epoch": 1.89456,
+      "grad_norm": 1.9679171286325126,
+      "learning_rate": 6.342761550916152e-05,
+      "loss": 1.2639,
+      "step": 11841
+    },
+    {
+      "epoch": 1.89472,
+      "grad_norm": 2.0908179712416053,
+      "learning_rate": 6.341153889078495e-05,
+      "loss": 1.2286,
+      "step": 11842
+    },
+    {
+      "epoch": 1.8948800000000001,
+      "grad_norm": 1.9885271928359307,
+      "learning_rate": 6.339546336415472e-05,
+      "loss": 1.2214,
+      "step": 11843
+    },
+    {
+      "epoch": 1.8950399999999998,
+      "grad_norm": 2.265370154737069,
+      "learning_rate": 6.337938892975057e-05,
+      "loss": 1.2901,
+      "step": 11844
+    },
+    {
+      "epoch": 1.8952,
+      "grad_norm": 1.7436264847065346,
+      "learning_rate": 6.336331558805204e-05,
+      "loss": 1.296,
+      "step": 11845
+    },
+    {
+      "epoch": 1.8953600000000002,
+      "grad_norm": 1.7974272287998945,
+      "learning_rate": 6.334724333953886e-05,
+      "loss": 1.2292,
+      "step": 11846
+    },
+    {
+      "epoch": 1.8955199999999999,
+      "grad_norm": 2.033948010839846,
+      "learning_rate": 6.333117218469045e-05,
+      "loss": 1.2584,
+      "step": 11847
+    },
+    {
+      "epoch": 1.89568,
+      "grad_norm": 2.27377040105681,
+      "learning_rate": 6.331510212398649e-05,
+      "loss": 1.2434,
+      "step": 11848
+    },
+    {
+      "epoch": 1.89584,
+      "grad_norm": 1.8040838082390203,
+      "learning_rate": 6.329903315790638e-05,
+      "loss": 1.2175,
+      "step": 11849
+    },
+    {
+      "epoch": 1.896,
+      "grad_norm": 1.8911867711409773,
+      "learning_rate": 6.328296528692967e-05,
+      "loss": 1.252,
+      "step": 11850
+    },
+    {
+      "epoch": 1.89616,
+      "grad_norm": 2.2038870710212772,
+      "learning_rate": 6.326689851153575e-05,
+      "loss": 1.2322,
+      "step": 11851
+    },
+    {
+      "epoch": 1.89632,
+      "grad_norm": 1.8701508242928657,
+      "learning_rate": 6.325083283220409e-05,
+      "loss": 1.2687,
+      "step": 11852
+    },
+    {
+      "epoch": 1.89648,
+      "grad_norm": 1.8905882557654676,
+      "learning_rate": 6.3234768249414e-05,
+      "loss": 1.1867,
+      "step": 11853
+    },
+    {
+      "epoch": 1.89664,
+      "grad_norm": 2.365346793007005,
+      "learning_rate": 6.321870476364487e-05,
+      "loss": 1.2424,
+      "step": 11854
+    },
+    {
+      "epoch": 1.8968,
+      "grad_norm": 1.7835464544320292,
+      "learning_rate": 6.320264237537596e-05,
+      "loss": 1.2665,
+      "step": 11855
+    },
+    {
+      "epoch": 1.89696,
+      "grad_norm": 2.3103559082363496,
+      "learning_rate": 6.318658108508664e-05,
+      "loss": 1.2085,
+      "step": 11856
+    },
+    {
+      "epoch": 1.8971200000000001,
+      "grad_norm": 2.4325599648605256,
+      "learning_rate": 6.317052089325605e-05,
+      "loss": 1.2489,
+      "step": 11857
+    },
+    {
+      "epoch": 1.8972799999999999,
+      "grad_norm": 2.0788516179074876,
+      "learning_rate": 6.315446180036346e-05,
+      "loss": 1.2641,
+      "step": 11858
+    },
+    {
+      "epoch": 1.89744,
+      "grad_norm": 2.136620540632228,
+      "learning_rate": 6.313840380688802e-05,
+      "loss": 1.2618,
+      "step": 11859
+    },
+    {
+      "epoch": 1.8976,
+      "grad_norm": 2.2785960896432482,
+      "learning_rate": 6.312234691330896e-05,
+      "loss": 1.2964,
+      "step": 11860
+    },
+    {
+      "epoch": 1.89776,
+      "grad_norm": 1.7788454062480985,
+      "learning_rate": 6.310629112010527e-05,
+      "loss": 1.2468,
+      "step": 11861
+    },
+    {
+      "epoch": 1.89792,
+      "grad_norm": 2.450657605975211,
+      "learning_rate": 6.309023642775614e-05,
+      "loss": 1.316,
+      "step": 11862
+    },
+    {
+      "epoch": 1.89808,
+      "grad_norm": 2.21543100363564,
+      "learning_rate": 6.307418283674054e-05,
+      "loss": 1.2558,
+      "step": 11863
+    },
+    {
+      "epoch": 1.89824,
+      "grad_norm": 2.049694371987737,
+      "learning_rate": 6.305813034753754e-05,
+      "loss": 1.2313,
+      "step": 11864
+    },
+    {
+      "epoch": 1.8984,
+      "grad_norm": 2.1517812251030644,
+      "learning_rate": 6.304207896062607e-05,
+      "loss": 1.2363,
+      "step": 11865
+    },
+    {
+      "epoch": 1.89856,
+      "grad_norm": 2.1552400424899663,
+      "learning_rate": 6.302602867648513e-05,
+      "loss": 1.2553,
+      "step": 11866
+    },
+    {
+      "epoch": 1.89872,
+      "grad_norm": 2.2659265447259958,
+      "learning_rate": 6.30099794955936e-05,
+      "loss": 1.3017,
+      "step": 11867
+    },
+    {
+      "epoch": 1.8988800000000001,
+      "grad_norm": 1.9998471746913025,
+      "learning_rate": 6.299393141843038e-05,
+      "loss": 1.2726,
+      "step": 11868
+    },
+    {
+      "epoch": 1.8990399999999998,
+      "grad_norm": 2.247326793706704,
+      "learning_rate": 6.297788444547435e-05,
+      "loss": 1.2237,
+      "step": 11869
+    },
+    {
+      "epoch": 1.8992,
+      "grad_norm": 1.9519490112603937,
+      "learning_rate": 6.296183857720428e-05,
+      "loss": 1.2648,
+      "step": 11870
+    },
+    {
+      "epoch": 1.8993600000000002,
+      "grad_norm": 2.245074566754017,
+      "learning_rate": 6.294579381409901e-05,
+      "loss": 1.1946,
+      "step": 11871
+    },
+    {
+      "epoch": 1.8995199999999999,
+      "grad_norm": 2.3366178180847332,
+      "learning_rate": 6.292975015663723e-05,
+      "loss": 1.2501,
+      "step": 11872
+    },
+    {
+      "epoch": 1.89968,
+      "grad_norm": 1.690916637877415,
+      "learning_rate": 6.29137076052977e-05,
+      "loss": 1.247,
+      "step": 11873
+    },
+    {
+      "epoch": 1.89984,
+      "grad_norm": 1.760781626411232,
+      "learning_rate": 6.289766616055908e-05,
+      "loss": 1.1969,
+      "step": 11874
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 2.08370617718986,
+      "learning_rate": 6.288162582290005e-05,
+      "loss": 1.1989,
+      "step": 11875
+    },
+    {
+      "epoch": 1.90016,
+      "grad_norm": 1.7555952497795588,
+      "learning_rate": 6.28655865927992e-05,
+      "loss": 1.3054,
+      "step": 11876
+    },
+    {
+      "epoch": 1.90032,
+      "grad_norm": 2.3405510713895206,
+      "learning_rate": 6.284954847073516e-05,
+      "loss": 1.3199,
+      "step": 11877
+    },
+    {
+      "epoch": 1.90048,
+      "grad_norm": 2.13779450396127,
+      "learning_rate": 6.283351145718643e-05,
+      "loss": 1.2535,
+      "step": 11878
+    },
+    {
+      "epoch": 1.90064,
+      "grad_norm": 1.5996183739887988,
+      "learning_rate": 6.281747555263161e-05,
+      "loss": 1.2153,
+      "step": 11879
+    },
+    {
+      "epoch": 1.9008,
+      "grad_norm": 1.6770729285633255,
+      "learning_rate": 6.280144075754911e-05,
+      "loss": 1.2476,
+      "step": 11880
+    },
+    {
+      "epoch": 1.90096,
+      "grad_norm": 1.8088079615248622,
+      "learning_rate": 6.278540707241743e-05,
+      "loss": 1.2342,
+      "step": 11881
+    },
+    {
+      "epoch": 1.9011200000000001,
+      "grad_norm": 2.0367593783046716,
+      "learning_rate": 6.276937449771494e-05,
+      "loss": 1.267,
+      "step": 11882
+    },
+    {
+      "epoch": 1.9012799999999999,
+      "grad_norm": 2.144278239204203,
+      "learning_rate": 6.275334303392009e-05,
+      "loss": 1.2144,
+      "step": 11883
+    },
+    {
+      "epoch": 1.90144,
+      "grad_norm": 1.9381400104145934,
+      "learning_rate": 6.273731268151117e-05,
+      "loss": 1.1849,
+      "step": 11884
+    },
+    {
+      "epoch": 1.9016,
+      "grad_norm": 1.8961974516168252,
+      "learning_rate": 6.27212834409666e-05,
+      "loss": 1.3005,
+      "step": 11885
+    },
+    {
+      "epoch": 1.90176,
+      "grad_norm": 2.080843321548718,
+      "learning_rate": 6.270525531276457e-05,
+      "loss": 1.2455,
+      "step": 11886
+    },
+    {
+      "epoch": 1.90192,
+      "grad_norm": 2.5657739261413433,
+      "learning_rate": 6.268922829738342e-05,
+      "loss": 1.2471,
+      "step": 11887
+    },
+    {
+      "epoch": 1.90208,
+      "grad_norm": 2.04363262351125,
+      "learning_rate": 6.267320239530128e-05,
+      "loss": 1.2684,
+      "step": 11888
+    },
+    {
+      "epoch": 1.90224,
+      "grad_norm": 1.9723143134128764,
+      "learning_rate": 6.265717760699645e-05,
+      "loss": 1.2464,
+      "step": 11889
+    },
+    {
+      "epoch": 1.9024,
+      "grad_norm": 2.839778890817673,
+      "learning_rate": 6.264115393294695e-05,
+      "loss": 1.2339,
+      "step": 11890
+    },
+    {
+      "epoch": 1.90256,
+      "grad_norm": 2.1852675712938066,
+      "learning_rate": 6.262513137363104e-05,
+      "loss": 1.3008,
+      "step": 11891
+    },
+    {
+      "epoch": 1.90272,
+      "grad_norm": 1.8337984521666133,
+      "learning_rate": 6.260910992952671e-05,
+      "loss": 1.2269,
+      "step": 11892
+    },
+    {
+      "epoch": 1.9028800000000001,
+      "grad_norm": 2.258882562155936,
+      "learning_rate": 6.25930896011121e-05,
+      "loss": 1.1677,
+      "step": 11893
+    },
+    {
+      "epoch": 1.9030399999999998,
+      "grad_norm": 1.8748067922376181,
+      "learning_rate": 6.257707038886513e-05,
+      "loss": 1.2563,
+      "step": 11894
+    },
+    {
+      "epoch": 1.9032,
+      "grad_norm": 1.7471481735718604,
+      "learning_rate": 6.25610522932639e-05,
+      "loss": 1.2449,
+      "step": 11895
+    },
+    {
+      "epoch": 1.9033600000000002,
+      "grad_norm": 1.9362756307913171,
+      "learning_rate": 6.254503531478628e-05,
+      "loss": 1.2904,
+      "step": 11896
+    },
+    {
+      "epoch": 1.9035199999999999,
+      "grad_norm": 2.0253508325022676,
+      "learning_rate": 6.252901945391028e-05,
+      "loss": 1.1964,
+      "step": 11897
+    },
+    {
+      "epoch": 1.90368,
+      "grad_norm": 2.0831056265252856,
+      "learning_rate": 6.251300471111368e-05,
+      "loss": 1.2849,
+      "step": 11898
+    },
+    {
+      "epoch": 1.90384,
+      "grad_norm": 2.223647565490654,
+      "learning_rate": 6.249699108687441e-05,
+      "loss": 1.2547,
+      "step": 11899
+    },
+    {
+      "epoch": 1.904,
+      "grad_norm": 1.7790844384386146,
+      "learning_rate": 6.248097858167028e-05,
+      "loss": 1.2253,
+      "step": 11900
+    },
+    {
+      "epoch": 1.90416,
+      "grad_norm": 2.2860963627597393,
+      "learning_rate": 6.24649671959791e-05,
+      "loss": 1.4087,
+      "step": 11901
+    },
+    {
+      "epoch": 1.90432,
+      "grad_norm": 2.01622156646983,
+      "learning_rate": 6.244895693027857e-05,
+      "loss": 1.2693,
+      "step": 11902
+    },
+    {
+      "epoch": 1.90448,
+      "grad_norm": 2.0980092711965432,
+      "learning_rate": 6.243294778504649e-05,
+      "loss": 1.1886,
+      "step": 11903
+    },
+    {
+      "epoch": 1.90464,
+      "grad_norm": 2.102251080142354,
+      "learning_rate": 6.241693976076047e-05,
+      "loss": 1.2143,
+      "step": 11904
+    },
+    {
+      "epoch": 1.9048,
+      "grad_norm": 2.042616399569263,
+      "learning_rate": 6.240093285789823e-05,
+      "loss": 1.2142,
+      "step": 11905
+    },
+    {
+      "epoch": 1.90496,
+      "grad_norm": 1.8179853417654992,
+      "learning_rate": 6.238492707693734e-05,
+      "loss": 1.2453,
+      "step": 11906
+    },
+    {
+      "epoch": 1.9051200000000001,
+      "grad_norm": 1.9948177897809622,
+      "learning_rate": 6.236892241835544e-05,
+      "loss": 1.2079,
+      "step": 11907
+    },
+    {
+      "epoch": 1.9052799999999999,
+      "grad_norm": 1.823555987905523,
+      "learning_rate": 6.235291888263002e-05,
+      "loss": 1.2046,
+      "step": 11908
+    },
+    {
+      "epoch": 1.90544,
+      "grad_norm": 2.570954616926009,
+      "learning_rate": 6.23369164702387e-05,
+      "loss": 1.2483,
+      "step": 11909
+    },
+    {
+      "epoch": 1.9056,
+      "grad_norm": 1.9897501211110997,
+      "learning_rate": 6.232091518165886e-05,
+      "loss": 1.2736,
+      "step": 11910
+    },
+    {
+      "epoch": 1.90576,
+      "grad_norm": 1.900653737068585,
+      "learning_rate": 6.230491501736806e-05,
+      "loss": 1.3017,
+      "step": 11911
+    },
+    {
+      "epoch": 1.90592,
+      "grad_norm": 2.320920305029226,
+      "learning_rate": 6.22889159778436e-05,
+      "loss": 1.2281,
+      "step": 11912
+    },
+    {
+      "epoch": 1.90608,
+      "grad_norm": 2.045912799018047,
+      "learning_rate": 6.227291806356301e-05,
+      "loss": 1.2305,
+      "step": 11913
+    },
+    {
+      "epoch": 1.90624,
+      "grad_norm": 2.011745204703852,
+      "learning_rate": 6.225692127500352e-05,
+      "loss": 1.2461,
+      "step": 11914
+    },
+    {
+      "epoch": 1.9064,
+      "grad_norm": 2.01170428379943,
+      "learning_rate": 6.224092561264252e-05,
+      "loss": 1.2756,
+      "step": 11915
+    },
+    {
+      "epoch": 1.90656,
+      "grad_norm": 2.0147425835958694,
+      "learning_rate": 6.222493107695727e-05,
+      "loss": 1.283,
+      "step": 11916
+    },
+    {
+      "epoch": 1.90672,
+      "grad_norm": 2.169267343733785,
+      "learning_rate": 6.220893766842505e-05,
+      "loss": 1.2115,
+      "step": 11917
+    },
+    {
+      "epoch": 1.9068800000000001,
+      "grad_norm": 1.7871429871599298,
+      "learning_rate": 6.219294538752303e-05,
+      "loss": 1.2612,
+      "step": 11918
+    },
+    {
+      "epoch": 1.9070399999999998,
+      "grad_norm": 1.9704591174681574,
+      "learning_rate": 6.217695423472847e-05,
+      "loss": 1.278,
+      "step": 11919
+    },
+    {
+      "epoch": 1.9072,
+      "grad_norm": 3.004748592414135,
+      "learning_rate": 6.216096421051845e-05,
+      "loss": 1.1918,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9073600000000002,
+      "grad_norm": 2.4596129227765107,
+      "learning_rate": 6.214497531537015e-05,
+      "loss": 1.201,
+      "step": 11921
+    },
+    {
+      "epoch": 1.9075199999999999,
+      "grad_norm": 1.7194624124717088,
+      "learning_rate": 6.21289875497606e-05,
+      "loss": 1.2223,
+      "step": 11922
+    },
+    {
+      "epoch": 1.90768,
+      "grad_norm": 2.2275332994864185,
+      "learning_rate": 6.211300091416687e-05,
+      "loss": 1.1812,
+      "step": 11923
+    },
+    {
+      "epoch": 1.90784,
+      "grad_norm": 2.169644972995022,
+      "learning_rate": 6.209701540906599e-05,
+      "loss": 1.2401,
+      "step": 11924
+    },
+    {
+      "epoch": 1.908,
+      "grad_norm": 1.7029010116075194,
+      "learning_rate": 6.208103103493496e-05,
+      "loss": 1.2826,
+      "step": 11925
+    },
+    {
+      "epoch": 1.90816,
+      "grad_norm": 1.9548372966247107,
+      "learning_rate": 6.20650477922507e-05,
+      "loss": 1.2715,
+      "step": 11926
+    },
+    {
+      "epoch": 1.90832,
+      "grad_norm": 2.3720203614290667,
+      "learning_rate": 6.204906568149012e-05,
+      "loss": 1.2169,
+      "step": 11927
+    },
+    {
+      "epoch": 1.90848,
+      "grad_norm": 2.1692812338735634,
+      "learning_rate": 6.203308470313016e-05,
+      "loss": 1.2726,
+      "step": 11928
+    },
+    {
+      "epoch": 1.9086400000000001,
+      "grad_norm": 1.857945024972868,
+      "learning_rate": 6.201710485764759e-05,
+      "loss": 1.2567,
+      "step": 11929
+    },
+    {
+      "epoch": 1.9088,
+      "grad_norm": 1.9348922345512318,
+      "learning_rate": 6.20011261455193e-05,
+      "loss": 1.2154,
+      "step": 11930
+    },
+    {
+      "epoch": 1.90896,
+      "grad_norm": 2.616074146142158,
+      "learning_rate": 6.1985148567222e-05,
+      "loss": 1.2946,
+      "step": 11931
+    },
+    {
+      "epoch": 1.9091200000000002,
+      "grad_norm": 2.0369164115674763,
+      "learning_rate": 6.19691721232325e-05,
+      "loss": 1.2307,
+      "step": 11932
+    },
+    {
+      "epoch": 1.9092799999999999,
+      "grad_norm": 2.0757454215782953,
+      "learning_rate": 6.195319681402747e-05,
+      "loss": 1.2266,
+      "step": 11933
+    },
+    {
+      "epoch": 1.90944,
+      "grad_norm": 2.4457307137786177,
+      "learning_rate": 6.193722264008362e-05,
+      "loss": 1.2913,
+      "step": 11934
+    },
+    {
+      "epoch": 1.9096,
+      "grad_norm": 2.479334752232671,
+      "learning_rate": 6.192124960187756e-05,
+      "loss": 1.2944,
+      "step": 11935
+    },
+    {
+      "epoch": 1.90976,
+      "grad_norm": 1.5739437514465155,
+      "learning_rate": 6.190527769988597e-05,
+      "loss": 1.2067,
+      "step": 11936
+    },
+    {
+      "epoch": 1.90992,
+      "grad_norm": 1.914839942615626,
+      "learning_rate": 6.188930693458534e-05,
+      "loss": 1.2294,
+      "step": 11937
+    },
+    {
+      "epoch": 1.91008,
+      "grad_norm": 2.239407854530631,
+      "learning_rate": 6.187333730645228e-05,
+      "loss": 1.2127,
+      "step": 11938
+    },
+    {
+      "epoch": 1.91024,
+      "grad_norm": 2.008253940481331,
+      "learning_rate": 6.185736881596327e-05,
+      "loss": 1.1544,
+      "step": 11939
+    },
+    {
+      "epoch": 1.9104,
+      "grad_norm": 1.903729489756749,
+      "learning_rate": 6.184140146359478e-05,
+      "loss": 1.2387,
+      "step": 11940
+    },
+    {
+      "epoch": 1.91056,
+      "grad_norm": 2.008099813272581,
+      "learning_rate": 6.182543524982327e-05,
+      "loss": 1.2082,
+      "step": 11941
+    },
+    {
+      "epoch": 1.91072,
+      "grad_norm": 1.8455086338479112,
+      "learning_rate": 6.18094701751252e-05,
+      "loss": 1.2291,
+      "step": 11942
+    },
+    {
+      "epoch": 1.9108800000000001,
+      "grad_norm": 2.556174898708449,
+      "learning_rate": 6.179350623997681e-05,
+      "loss": 1.2261,
+      "step": 11943
+    },
+    {
+      "epoch": 1.9110399999999998,
+      "grad_norm": 1.9632684304909505,
+      "learning_rate": 6.17775434448546e-05,
+      "loss": 1.244,
+      "step": 11944
+    },
+    {
+      "epoch": 1.9112,
+      "grad_norm": 2.3293499429661373,
+      "learning_rate": 6.176158179023474e-05,
+      "loss": 1.217,
+      "step": 11945
+    },
+    {
+      "epoch": 1.91136,
+      "grad_norm": 2.4199778919297343,
+      "learning_rate": 6.17456212765936e-05,
+      "loss": 1.198,
+      "step": 11946
+    },
+    {
+      "epoch": 1.9115199999999999,
+      "grad_norm": 1.9355217335404098,
+      "learning_rate": 6.172966190440735e-05,
+      "loss": 1.2228,
+      "step": 11947
+    },
+    {
+      "epoch": 1.91168,
+      "grad_norm": 2.081253238721236,
+      "learning_rate": 6.171370367415223e-05,
+      "loss": 1.2726,
+      "step": 11948
+    },
+    {
+      "epoch": 1.91184,
+      "grad_norm": 2.3757043988227253,
+      "learning_rate": 6.169774658630438e-05,
+      "loss": 1.2818,
+      "step": 11949
+    },
+    {
+      "epoch": 1.912,
+      "grad_norm": 2.0153673342667418,
+      "learning_rate": 6.168179064134e-05,
+      "loss": 1.2815,
+      "step": 11950
+    },
+    {
+      "epoch": 1.91216,
+      "grad_norm": 1.7827911318346847,
+      "learning_rate": 6.166583583973513e-05,
+      "loss": 1.1971,
+      "step": 11951
+    },
+    {
+      "epoch": 1.91232,
+      "grad_norm": 1.888158134495538,
+      "learning_rate": 6.164988218196588e-05,
+      "loss": 1.2468,
+      "step": 11952
+    },
+    {
+      "epoch": 1.91248,
+      "grad_norm": 2.526649965323868,
+      "learning_rate": 6.163392966850823e-05,
+      "loss": 1.3042,
+      "step": 11953
+    },
+    {
+      "epoch": 1.9126400000000001,
+      "grad_norm": 1.9364259146015983,
+      "learning_rate": 6.161797829983826e-05,
+      "loss": 1.2052,
+      "step": 11954
+    },
+    {
+      "epoch": 1.9127999999999998,
+      "grad_norm": 2.8222207101915777,
+      "learning_rate": 6.160202807643184e-05,
+      "loss": 1.2589,
+      "step": 11955
+    },
+    {
+      "epoch": 1.91296,
+      "grad_norm": 1.8958024057320946,
+      "learning_rate": 6.158607899876497e-05,
+      "loss": 1.2184,
+      "step": 11956
+    },
+    {
+      "epoch": 1.9131200000000002,
+      "grad_norm": 2.074953528161197,
+      "learning_rate": 6.157013106731352e-05,
+      "loss": 1.2004,
+      "step": 11957
+    },
+    {
+      "epoch": 1.9132799999999999,
+      "grad_norm": 2.341224274599962,
+      "learning_rate": 6.15541842825534e-05,
+      "loss": 1.2068,
+      "step": 11958
+    },
+    {
+      "epoch": 1.91344,
+      "grad_norm": 2.543173866488912,
+      "learning_rate": 6.153823864496034e-05,
+      "loss": 1.2424,
+      "step": 11959
+    },
+    {
+      "epoch": 1.9136,
+      "grad_norm": 1.8212715012091054,
+      "learning_rate": 6.152229415501026e-05,
+      "loss": 1.2887,
+      "step": 11960
+    },
+    {
+      "epoch": 1.91376,
+      "grad_norm": 1.800757295650315,
+      "learning_rate": 6.150635081317879e-05,
+      "loss": 1.2886,
+      "step": 11961
+    },
+    {
+      "epoch": 1.91392,
+      "grad_norm": 2.219961077114597,
+      "learning_rate": 6.149040861994178e-05,
+      "loss": 1.1953,
+      "step": 11962
+    },
+    {
+      "epoch": 1.91408,
+      "grad_norm": 2.175518362024765,
+      "learning_rate": 6.14744675757748e-05,
+      "loss": 1.2779,
+      "step": 11963
+    },
+    {
+      "epoch": 1.91424,
+      "grad_norm": 2.0455707997837846,
+      "learning_rate": 6.145852768115361e-05,
+      "loss": 1.2779,
+      "step": 11964
+    },
+    {
+      "epoch": 1.9144,
+      "grad_norm": 1.9641855297570152,
+      "learning_rate": 6.144258893655377e-05,
+      "loss": 1.2514,
+      "step": 11965
+    },
+    {
+      "epoch": 1.91456,
+      "grad_norm": 2.1522183231397958,
+      "learning_rate": 6.142665134245093e-05,
+      "loss": 1.2595,
+      "step": 11966
+    },
+    {
+      "epoch": 1.91472,
+      "grad_norm": 2.0378597239124194,
+      "learning_rate": 6.141071489932055e-05,
+      "loss": 1.2101,
+      "step": 11967
+    },
+    {
+      "epoch": 1.9148800000000001,
+      "grad_norm": 1.9223143658420208,
+      "learning_rate": 6.139477960763827e-05,
+      "loss": 1.2449,
+      "step": 11968
+    },
+    {
+      "epoch": 1.9150399999999999,
+      "grad_norm": 2.1520176393847414,
+      "learning_rate": 6.137884546787947e-05,
+      "loss": 1.2589,
+      "step": 11969
+    },
+    {
+      "epoch": 1.9152,
+      "grad_norm": 1.8648181663396985,
+      "learning_rate": 6.136291248051968e-05,
+      "loss": 1.283,
+      "step": 11970
+    },
+    {
+      "epoch": 1.91536,
+      "grad_norm": 1.7789968337628417,
+      "learning_rate": 6.134698064603423e-05,
+      "loss": 1.1802,
+      "step": 11971
+    },
+    {
+      "epoch": 1.91552,
+      "grad_norm": 2.02514981381802,
+      "learning_rate": 6.133104996489859e-05,
+      "loss": 1.1905,
+      "step": 11972
+    },
+    {
+      "epoch": 1.91568,
+      "grad_norm": 2.180468945907282,
+      "learning_rate": 6.131512043758805e-05,
+      "loss": 1.285,
+      "step": 11973
+    },
+    {
+      "epoch": 1.91584,
+      "grad_norm": 2.3207285263791166,
+      "learning_rate": 6.129919206457798e-05,
+      "loss": 1.2076,
+      "step": 11974
+    },
+    {
+      "epoch": 1.916,
+      "grad_norm": 1.7802582251068315,
+      "learning_rate": 6.128326484634359e-05,
+      "loss": 1.2655,
+      "step": 11975
+    },
+    {
+      "epoch": 1.91616,
+      "grad_norm": 2.115483753556383,
+      "learning_rate": 6.126733878336019e-05,
+      "loss": 1.2562,
+      "step": 11976
+    },
+    {
+      "epoch": 1.91632,
+      "grad_norm": 2.126592108545692,
+      "learning_rate": 6.125141387610293e-05,
+      "loss": 1.2585,
+      "step": 11977
+    },
+    {
+      "epoch": 1.91648,
+      "grad_norm": 1.9426987319602047,
+      "learning_rate": 6.123549012504707e-05,
+      "loss": 1.2284,
+      "step": 11978
+    },
+    {
+      "epoch": 1.9166400000000001,
+      "grad_norm": 1.9687135057718885,
+      "learning_rate": 6.121956753066764e-05,
+      "loss": 1.229,
+      "step": 11979
+    },
+    {
+      "epoch": 1.9167999999999998,
+      "grad_norm": 1.9332337528395542,
+      "learning_rate": 6.120364609343985e-05,
+      "loss": 1.2575,
+      "step": 11980
+    },
+    {
+      "epoch": 1.91696,
+      "grad_norm": 2.0531518753080937,
+      "learning_rate": 6.118772581383869e-05,
+      "loss": 1.2772,
+      "step": 11981
+    },
+    {
+      "epoch": 1.9171200000000002,
+      "grad_norm": 1.8048707992770587,
+      "learning_rate": 6.117180669233927e-05,
+      "loss": 1.2083,
+      "step": 11982
+    },
+    {
+      "epoch": 1.9172799999999999,
+      "grad_norm": 2.039181992021956,
+      "learning_rate": 6.115588872941654e-05,
+      "loss": 1.2179,
+      "step": 11983
+    },
+    {
+      "epoch": 1.91744,
+      "grad_norm": 2.1225893761789862,
+      "learning_rate": 6.113997192554546e-05,
+      "loss": 1.2689,
+      "step": 11984
+    },
+    {
+      "epoch": 1.9176,
+      "grad_norm": 1.8406900029406281,
+      "learning_rate": 6.112405628120106e-05,
+      "loss": 1.1377,
+      "step": 11985
+    },
+    {
+      "epoch": 1.91776,
+      "grad_norm": 1.8901190442041351,
+      "learning_rate": 6.110814179685811e-05,
+      "loss": 1.2767,
+      "step": 11986
+    },
+    {
+      "epoch": 1.91792,
+      "grad_norm": 1.9208392020315852,
+      "learning_rate": 6.109222847299159e-05,
+      "loss": 1.2334,
+      "step": 11987
+    },
+    {
+      "epoch": 1.91808,
+      "grad_norm": 2.184333271348253,
+      "learning_rate": 6.107631631007624e-05,
+      "loss": 1.1895,
+      "step": 11988
+    },
+    {
+      "epoch": 1.91824,
+      "grad_norm": 1.96986463611998,
+      "learning_rate": 6.106040530858689e-05,
+      "loss": 1.2364,
+      "step": 11989
+    },
+    {
+      "epoch": 1.9184,
+      "grad_norm": 2.2586884609421167,
+      "learning_rate": 6.104449546899834e-05,
+      "loss": 1.308,
+      "step": 11990
+    },
+    {
+      "epoch": 1.91856,
+      "grad_norm": 1.9864611053829424,
+      "learning_rate": 6.1028586791785266e-05,
+      "loss": 1.2047,
+      "step": 11991
+    },
+    {
+      "epoch": 1.91872,
+      "grad_norm": 1.8928812640936898,
+      "learning_rate": 6.101267927742236e-05,
+      "loss": 1.1723,
+      "step": 11992
+    },
+    {
+      "epoch": 1.9188800000000001,
+      "grad_norm": 2.123817389595316,
+      "learning_rate": 6.0996772926384324e-05,
+      "loss": 1.2707,
+      "step": 11993
+    },
+    {
+      "epoch": 1.9190399999999999,
+      "grad_norm": 2.100536245820889,
+      "learning_rate": 6.098086773914571e-05,
+      "loss": 1.2164,
+      "step": 11994
+    },
+    {
+      "epoch": 1.9192,
+      "grad_norm": 1.859427115102647,
+      "learning_rate": 6.096496371618119e-05,
+      "loss": 1.2683,
+      "step": 11995
+    },
+    {
+      "epoch": 1.91936,
+      "grad_norm": 2.057326460377133,
+      "learning_rate": 6.0949060857965235e-05,
+      "loss": 1.182,
+      "step": 11996
+    },
+    {
+      "epoch": 1.91952,
+      "grad_norm": 2.103239669679571,
+      "learning_rate": 6.093315916497242e-05,
+      "loss": 1.2636,
+      "step": 11997
+    },
+    {
+      "epoch": 1.91968,
+      "grad_norm": 2.2497423876629674,
+      "learning_rate": 6.091725863767721e-05,
+      "loss": 1.212,
+      "step": 11998
+    },
+    {
+      "epoch": 1.91984,
+      "grad_norm": 1.7371021376646725,
+      "learning_rate": 6.090135927655405e-05,
+      "loss": 1.2334,
+      "step": 11999
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 2.014291125540509,
+      "learning_rate": 6.0885461082077335e-05,
+      "loss": 1.2451,
+      "step": 12000
+    },
+    {
+      "epoch": 1.92016,
+      "grad_norm": 1.6644215257536426,
+      "learning_rate": 6.0869564054721505e-05,
+      "loss": 1.3012,
+      "step": 12001
+    },
+    {
+      "epoch": 1.92032,
+      "grad_norm": 1.6640054341152641,
+      "learning_rate": 6.0853668194960824e-05,
+      "loss": 1.2952,
+      "step": 12002
+    },
+    {
+      "epoch": 1.92048,
+      "grad_norm": 1.8594537823745927,
+      "learning_rate": 6.083777350326969e-05,
+      "loss": 1.2173,
+      "step": 12003
+    },
+    {
+      "epoch": 1.9206400000000001,
+      "grad_norm": 2.055450395460042,
+      "learning_rate": 6.082187998012228e-05,
+      "loss": 1.2831,
+      "step": 12004
+    },
+    {
+      "epoch": 1.9207999999999998,
+      "grad_norm": 1.8047275952814381,
+      "learning_rate": 6.08059876259929e-05,
+      "loss": 1.2296,
+      "step": 12005
+    },
+    {
+      "epoch": 1.92096,
+      "grad_norm": 1.816922267341314,
+      "learning_rate": 6.079009644135575e-05,
+      "loss": 1.2121,
+      "step": 12006
+    },
+    {
+      "epoch": 1.9211200000000002,
+      "grad_norm": 1.6902652228684032,
+      "learning_rate": 6.0774206426684975e-05,
+      "loss": 1.2255,
+      "step": 12007
+    },
+    {
+      "epoch": 1.9212799999999999,
+      "grad_norm": 1.994516789176983,
+      "learning_rate": 6.0758317582454716e-05,
+      "loss": 1.2855,
+      "step": 12008
+    },
+    {
+      "epoch": 1.92144,
+      "grad_norm": 1.8835158935789702,
+      "learning_rate": 6.07424299091391e-05,
+      "loss": 1.222,
+      "step": 12009
+    },
+    {
+      "epoch": 1.9216,
+      "grad_norm": 2.1892544328204435,
+      "learning_rate": 6.072654340721215e-05,
+      "loss": 1.2777,
+      "step": 12010
+    },
+    {
+      "epoch": 1.92176,
+      "grad_norm": 2.502203927739345,
+      "learning_rate": 6.071065807714795e-05,
+      "loss": 1.1762,
+      "step": 12011
+    },
+    {
+      "epoch": 1.92192,
+      "grad_norm": 2.5781026087914296,
+      "learning_rate": 6.0694773919420423e-05,
+      "loss": 1.2458,
+      "step": 12012
+    },
+    {
+      "epoch": 1.92208,
+      "grad_norm": 1.8883856456291546,
+      "learning_rate": 6.0678890934503585e-05,
+      "loss": 1.2428,
+      "step": 12013
+    },
+    {
+      "epoch": 1.92224,
+      "grad_norm": 1.8997822501521495,
+      "learning_rate": 6.066300912287135e-05,
+      "loss": 1.2105,
+      "step": 12014
+    },
+    {
+      "epoch": 1.9224,
+      "grad_norm": 2.197677824113116,
+      "learning_rate": 6.0647128484997595e-05,
+      "loss": 1.3016,
+      "step": 12015
+    },
+    {
+      "epoch": 1.92256,
+      "grad_norm": 2.1856698428207806,
+      "learning_rate": 6.063124902135617e-05,
+      "loss": 1.2375,
+      "step": 12016
+    },
+    {
+      "epoch": 1.92272,
+      "grad_norm": 2.0117513586194806,
+      "learning_rate": 6.061537073242094e-05,
+      "loss": 1.2793,
+      "step": 12017
+    },
+    {
+      "epoch": 1.9228800000000001,
+      "grad_norm": 2.745329470680726,
+      "learning_rate": 6.059949361866564e-05,
+      "loss": 1.243,
+      "step": 12018
+    },
+    {
+      "epoch": 1.9230399999999999,
+      "grad_norm": 2.26705753523481,
+      "learning_rate": 6.058361768056405e-05,
+      "loss": 1.2806,
+      "step": 12019
+    },
+    {
+      "epoch": 1.9232,
+      "grad_norm": 2.0352425533504457,
+      "learning_rate": 6.0567742918589844e-05,
+      "loss": 1.2356,
+      "step": 12020
+    },
+    {
+      "epoch": 1.92336,
+      "grad_norm": 1.8437397809045029,
+      "learning_rate": 6.055186933321675e-05,
+      "loss": 1.2268,
+      "step": 12021
+    },
+    {
+      "epoch": 1.92352,
+      "grad_norm": 2.1230684555004444,
+      "learning_rate": 6.0535996924918406e-05,
+      "loss": 1.2722,
+      "step": 12022
+    },
+    {
+      "epoch": 1.92368,
+      "grad_norm": 1.7491413109566145,
+      "learning_rate": 6.052012569416838e-05,
+      "loss": 1.2401,
+      "step": 12023
+    },
+    {
+      "epoch": 1.92384,
+      "grad_norm": 1.886614536699479,
+      "learning_rate": 6.0504255641440276e-05,
+      "loss": 1.2245,
+      "step": 12024
+    },
+    {
+      "epoch": 1.924,
+      "grad_norm": 2.1120158574425614,
+      "learning_rate": 6.048838676720767e-05,
+      "loss": 1.2246,
+      "step": 12025
+    },
+    {
+      "epoch": 1.92416,
+      "grad_norm": 2.141101170415076,
+      "learning_rate": 6.0472519071943976e-05,
+      "loss": 1.2258,
+      "step": 12026
+    },
+    {
+      "epoch": 1.92432,
+      "grad_norm": 2.036663090869736,
+      "learning_rate": 6.045665255612277e-05,
+      "loss": 1.2438,
+      "step": 12027
+    },
+    {
+      "epoch": 1.92448,
+      "grad_norm": 2.0906331425170883,
+      "learning_rate": 6.044078722021739e-05,
+      "loss": 1.2279,
+      "step": 12028
+    },
+    {
+      "epoch": 1.9246400000000001,
+      "grad_norm": 2.177148813681901,
+      "learning_rate": 6.042492306470129e-05,
+      "loss": 1.2374,
+      "step": 12029
+    },
+    {
+      "epoch": 1.9247999999999998,
+      "grad_norm": 1.8600446422215107,
+      "learning_rate": 6.040906009004782e-05,
+      "loss": 1.2206,
+      "step": 12030
+    },
+    {
+      "epoch": 1.92496,
+      "grad_norm": 2.106785077796803,
+      "learning_rate": 6.0393198296730316e-05,
+      "loss": 1.2517,
+      "step": 12031
+    },
+    {
+      "epoch": 1.9251200000000002,
+      "grad_norm": 2.2922582089942205,
+      "learning_rate": 6.0377337685222024e-05,
+      "loss": 1.2738,
+      "step": 12032
+    },
+    {
+      "epoch": 1.9252799999999999,
+      "grad_norm": 2.0025744724538597,
+      "learning_rate": 6.0361478255996295e-05,
+      "loss": 1.272,
+      "step": 12033
+    },
+    {
+      "epoch": 1.92544,
+      "grad_norm": 1.738597222439128,
+      "learning_rate": 6.0345620009526236e-05,
+      "loss": 1.2484,
+      "step": 12034
+    },
+    {
+      "epoch": 1.9256,
+      "grad_norm": 2.294271233603972,
+      "learning_rate": 6.032976294628515e-05,
+      "loss": 1.282,
+      "step": 12035
+    },
+    {
+      "epoch": 1.92576,
+      "grad_norm": 1.8777777499044517,
+      "learning_rate": 6.031390706674609e-05,
+      "loss": 1.25,
+      "step": 12036
+    },
+    {
+      "epoch": 1.92592,
+      "grad_norm": 2.194946426490301,
+      "learning_rate": 6.029805237138224e-05,
+      "loss": 1.3054,
+      "step": 12037
+    },
+    {
+      "epoch": 1.92608,
+      "grad_norm": 1.8863076222756656,
+      "learning_rate": 6.028219886066664e-05,
+      "loss": 1.2064,
+      "step": 12038
+    },
+    {
+      "epoch": 1.92624,
+      "grad_norm": 2.055886358823254,
+      "learning_rate": 6.026634653507236e-05,
+      "loss": 1.253,
+      "step": 12039
+    },
+    {
+      "epoch": 1.9264000000000001,
+      "grad_norm": 1.8738306251137282,
+      "learning_rate": 6.0250495395072384e-05,
+      "loss": 1.2372,
+      "step": 12040
+    },
+    {
+      "epoch": 1.92656,
+      "grad_norm": 2.3628449980495887,
+      "learning_rate": 6.023464544113974e-05,
+      "loss": 1.2091,
+      "step": 12041
+    },
+    {
+      "epoch": 1.92672,
+      "grad_norm": 1.7473550889123173,
+      "learning_rate": 6.021879667374728e-05,
+      "loss": 1.271,
+      "step": 12042
+    },
+    {
+      "epoch": 1.9268800000000001,
+      "grad_norm": 2.0882328723948893,
+      "learning_rate": 6.0202949093367974e-05,
+      "loss": 1.2151,
+      "step": 12043
+    },
+    {
+      "epoch": 1.9270399999999999,
+      "grad_norm": 2.363491180685284,
+      "learning_rate": 6.0187102700474716e-05,
+      "loss": 1.2857,
+      "step": 12044
+    },
+    {
+      "epoch": 1.9272,
+      "grad_norm": 2.062223676666028,
+      "learning_rate": 6.017125749554027e-05,
+      "loss": 1.2683,
+      "step": 12045
+    },
+    {
+      "epoch": 1.92736,
+      "grad_norm": 1.6672187675431973,
+      "learning_rate": 6.0155413479037456e-05,
+      "loss": 1.2644,
+      "step": 12046
+    },
+    {
+      "epoch": 1.92752,
+      "grad_norm": 1.8738095785545823,
+      "learning_rate": 6.013957065143907e-05,
+      "loss": 1.2319,
+      "step": 12047
+    },
+    {
+      "epoch": 1.92768,
+      "grad_norm": 1.9397578118134224,
+      "learning_rate": 6.01237290132178e-05,
+      "loss": 1.2283,
+      "step": 12048
+    },
+    {
+      "epoch": 1.92784,
+      "grad_norm": 2.108702413217494,
+      "learning_rate": 6.0107888564846335e-05,
+      "loss": 1.2183,
+      "step": 12049
+    },
+    {
+      "epoch": 1.928,
+      "grad_norm": 2.1039138958471684,
+      "learning_rate": 6.0092049306797394e-05,
+      "loss": 1.2984,
+      "step": 12050
+    },
+    {
+      "epoch": 1.92816,
+      "grad_norm": 2.003485704695514,
+      "learning_rate": 6.007621123954349e-05,
+      "loss": 1.1981,
+      "step": 12051
+    },
+    {
+      "epoch": 1.92832,
+      "grad_norm": 1.9223272066801638,
+      "learning_rate": 6.0060374363557316e-05,
+      "loss": 1.1818,
+      "step": 12052
+    },
+    {
+      "epoch": 1.92848,
+      "grad_norm": 2.4295548186060643,
+      "learning_rate": 6.0044538679311345e-05,
+      "loss": 1.2446,
+      "step": 12053
+    },
+    {
+      "epoch": 1.9286400000000001,
+      "grad_norm": 1.9653363632997014,
+      "learning_rate": 6.002870418727813e-05,
+      "loss": 1.2377,
+      "step": 12054
+    },
+    {
+      "epoch": 1.9287999999999998,
+      "grad_norm": 1.779387532757738,
+      "learning_rate": 6.001287088793013e-05,
+      "loss": 1.27,
+      "step": 12055
+    },
+    {
+      "epoch": 1.92896,
+      "grad_norm": 1.891531355110823,
+      "learning_rate": 5.99970387817398e-05,
+      "loss": 1.1891,
+      "step": 12056
+    },
+    {
+      "epoch": 1.92912,
+      "grad_norm": 2.044030675618269,
+      "learning_rate": 5.9981207869179525e-05,
+      "loss": 1.2383,
+      "step": 12057
+    },
+    {
+      "epoch": 1.9292799999999999,
+      "grad_norm": 2.033368826052685,
+      "learning_rate": 5.996537815072173e-05,
+      "loss": 1.1984,
+      "step": 12058
+    },
+    {
+      "epoch": 1.92944,
+      "grad_norm": 2.1442144571807646,
+      "learning_rate": 5.9949549626838676e-05,
+      "loss": 1.1941,
+      "step": 12059
+    },
+    {
+      "epoch": 1.9296,
+      "grad_norm": 2.1303283122540018,
+      "learning_rate": 5.993372229800274e-05,
+      "loss": 1.1812,
+      "step": 12060
+    },
+    {
+      "epoch": 1.92976,
+      "grad_norm": 1.8665877839558123,
+      "learning_rate": 5.9917896164686116e-05,
+      "loss": 1.2119,
+      "step": 12061
+    },
+    {
+      "epoch": 1.92992,
+      "grad_norm": 2.2348239126995297,
+      "learning_rate": 5.9902071227361076e-05,
+      "loss": 1.2653,
+      "step": 12062
+    },
+    {
+      "epoch": 1.93008,
+      "grad_norm": 2.237551679189263,
+      "learning_rate": 5.98862474864998e-05,
+      "loss": 1.2139,
+      "step": 12063
+    },
+    {
+      "epoch": 1.93024,
+      "grad_norm": 2.1561933550224537,
+      "learning_rate": 5.9870424942574456e-05,
+      "loss": 1.2792,
+      "step": 12064
+    },
+    {
+      "epoch": 1.9304000000000001,
+      "grad_norm": 2.3571464322153703,
+      "learning_rate": 5.9854603596057135e-05,
+      "loss": 1.2779,
+      "step": 12065
+    },
+    {
+      "epoch": 1.93056,
+      "grad_norm": 1.9455154014552627,
+      "learning_rate": 5.9838783447419974e-05,
+      "loss": 1.2617,
+      "step": 12066
+    },
+    {
+      "epoch": 1.93072,
+      "grad_norm": 2.338010391092208,
+      "learning_rate": 5.982296449713497e-05,
+      "loss": 1.2422,
+      "step": 12067
+    },
+    {
+      "epoch": 1.9308800000000002,
+      "grad_norm": 1.9919411397637092,
+      "learning_rate": 5.980714674567419e-05,
+      "loss": 1.2772,
+      "step": 12068
+    },
+    {
+      "epoch": 1.9310399999999999,
+      "grad_norm": 1.9205102816685584,
+      "learning_rate": 5.9791330193509556e-05,
+      "loss": 1.2527,
+      "step": 12069
+    },
+    {
+      "epoch": 1.9312,
+      "grad_norm": 2.172238520832009,
+      "learning_rate": 5.9775514841113036e-05,
+      "loss": 1.2496,
+      "step": 12070
+    },
+    {
+      "epoch": 1.93136,
+      "grad_norm": 2.159501682679179,
+      "learning_rate": 5.975970068895655e-05,
+      "loss": 1.2802,
+      "step": 12071
+    },
+    {
+      "epoch": 1.93152,
+      "grad_norm": 1.704001224314882,
+      "learning_rate": 5.9743887737511974e-05,
+      "loss": 1.2221,
+      "step": 12072
+    },
+    {
+      "epoch": 1.93168,
+      "grad_norm": 1.8503416093669098,
+      "learning_rate": 5.97280759872511e-05,
+      "loss": 1.2255,
+      "step": 12073
+    },
+    {
+      "epoch": 1.93184,
+      "grad_norm": 2.300564432606861,
+      "learning_rate": 5.971226543864579e-05,
+      "loss": 1.2429,
+      "step": 12074
+    },
+    {
+      "epoch": 1.932,
+      "grad_norm": 2.113181954964018,
+      "learning_rate": 5.969645609216774e-05,
+      "loss": 1.2077,
+      "step": 12075
+    },
+    {
+      "epoch": 1.93216,
+      "grad_norm": 1.8906406898493426,
+      "learning_rate": 5.968064794828875e-05,
+      "loss": 1.2262,
+      "step": 12076
+    },
+    {
+      "epoch": 1.93232,
+      "grad_norm": 1.9937833292158111,
+      "learning_rate": 5.966484100748042e-05,
+      "loss": 1.2234,
+      "step": 12077
+    },
+    {
+      "epoch": 1.93248,
+      "grad_norm": 1.7748886420486505,
+      "learning_rate": 5.96490352702145e-05,
+      "loss": 1.2509,
+      "step": 12078
+    },
+    {
+      "epoch": 1.9326400000000001,
+      "grad_norm": 1.9987392861778388,
+      "learning_rate": 5.9633230736962566e-05,
+      "loss": 1.2276,
+      "step": 12079
+    },
+    {
+      "epoch": 1.9327999999999999,
+      "grad_norm": 2.035188062479214,
+      "learning_rate": 5.9617427408196206e-05,
+      "loss": 1.269,
+      "step": 12080
+    },
+    {
+      "epoch": 1.93296,
+      "grad_norm": 1.8273921302579537,
+      "learning_rate": 5.960162528438694e-05,
+      "loss": 1.2587,
+      "step": 12081
+    },
+    {
+      "epoch": 1.93312,
+      "grad_norm": 1.9187154041508845,
+      "learning_rate": 5.958582436600635e-05,
+      "loss": 1.2558,
+      "step": 12082
+    },
+    {
+      "epoch": 1.9332799999999999,
+      "grad_norm": 2.050538496128149,
+      "learning_rate": 5.957002465352584e-05,
+      "loss": 1.2801,
+      "step": 12083
+    },
+    {
+      "epoch": 1.93344,
+      "grad_norm": 2.1012314011925497,
+      "learning_rate": 5.955422614741693e-05,
+      "loss": 1.2301,
+      "step": 12084
+    },
+    {
+      "epoch": 1.9336,
+      "grad_norm": 2.16741795681604,
+      "learning_rate": 5.9538428848150904e-05,
+      "loss": 1.2397,
+      "step": 12085
+    },
+    {
+      "epoch": 1.93376,
+      "grad_norm": 2.1521498850047216,
+      "learning_rate": 5.952263275619925e-05,
+      "loss": 1.2621,
+      "step": 12086
+    },
+    {
+      "epoch": 1.93392,
+      "grad_norm": 1.6510657814179281,
+      "learning_rate": 5.950683787203324e-05,
+      "loss": 1.2424,
+      "step": 12087
+    },
+    {
+      "epoch": 1.93408,
+      "grad_norm": 1.950309044046298,
+      "learning_rate": 5.949104419612419e-05,
+      "loss": 1.3072,
+      "step": 12088
+    },
+    {
+      "epoch": 1.93424,
+      "grad_norm": 1.8687611533855735,
+      "learning_rate": 5.947525172894333e-05,
+      "loss": 1.2253,
+      "step": 12089
+    },
+    {
+      "epoch": 1.9344000000000001,
+      "grad_norm": 2.1570187821221887,
+      "learning_rate": 5.9459460470961946e-05,
+      "loss": 1.2045,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9345599999999998,
+      "grad_norm": 2.3475610549397645,
+      "learning_rate": 5.944367042265115e-05,
+      "loss": 1.2625,
+      "step": 12091
+    },
+    {
+      "epoch": 1.93472,
+      "grad_norm": 2.189284825308847,
+      "learning_rate": 5.942788158448218e-05,
+      "loss": 1.2445,
+      "step": 12092
+    },
+    {
+      "epoch": 1.9348800000000002,
+      "grad_norm": 1.8907071324620897,
+      "learning_rate": 5.9412093956926054e-05,
+      "loss": 1.2642,
+      "step": 12093
+    },
+    {
+      "epoch": 1.9350399999999999,
+      "grad_norm": 1.8518321548205476,
+      "learning_rate": 5.9396307540453924e-05,
+      "loss": 1.1937,
+      "step": 12094
+    },
+    {
+      "epoch": 1.9352,
+      "grad_norm": 2.733111245487975,
+      "learning_rate": 5.9380522335536807e-05,
+      "loss": 1.2852,
+      "step": 12095
+    },
+    {
+      "epoch": 1.93536,
+      "grad_norm": 1.8301651183207353,
+      "learning_rate": 5.9364738342645706e-05,
+      "loss": 1.1771,
+      "step": 12096
+    },
+    {
+      "epoch": 1.93552,
+      "grad_norm": 1.8611444435203808,
+      "learning_rate": 5.9348955562251594e-05,
+      "loss": 1.2623,
+      "step": 12097
+    },
+    {
+      "epoch": 1.93568,
+      "grad_norm": 1.9447762814566105,
+      "learning_rate": 5.933317399482546e-05,
+      "loss": 1.2522,
+      "step": 12098
+    },
+    {
+      "epoch": 1.93584,
+      "grad_norm": 1.802441188837548,
+      "learning_rate": 5.93173936408381e-05,
+      "loss": 1.2789,
+      "step": 12099
+    },
+    {
+      "epoch": 1.936,
+      "grad_norm": 2.0621202330453374,
+      "learning_rate": 5.930161450076049e-05,
+      "loss": 1.1775,
+      "step": 12100
+    },
+    {
+      "epoch": 1.93616,
+      "grad_norm": 2.2093109554585233,
+      "learning_rate": 5.9285836575063346e-05,
+      "loss": 1.2102,
+      "step": 12101
+    },
+    {
+      "epoch": 1.93632,
+      "grad_norm": 1.7771129126986944,
+      "learning_rate": 5.927005986421752e-05,
+      "loss": 1.214,
+      "step": 12102
+    },
+    {
+      "epoch": 1.93648,
+      "grad_norm": 1.9632032911204316,
+      "learning_rate": 5.92542843686938e-05,
+      "loss": 1.1986,
+      "step": 12103
+    },
+    {
+      "epoch": 1.9366400000000001,
+      "grad_norm": 2.227588902491908,
+      "learning_rate": 5.923851008896282e-05,
+      "loss": 1.2189,
+      "step": 12104
+    },
+    {
+      "epoch": 1.9367999999999999,
+      "grad_norm": 1.9621867774147945,
+      "learning_rate": 5.922273702549531e-05,
+      "loss": 1.2551,
+      "step": 12105
+    },
+    {
+      "epoch": 1.93696,
+      "grad_norm": 1.88817879796683,
+      "learning_rate": 5.9206965178761894e-05,
+      "loss": 1.1593,
+      "step": 12106
+    },
+    {
+      "epoch": 1.93712,
+      "grad_norm": 1.9647159998806605,
+      "learning_rate": 5.919119454923325e-05,
+      "loss": 1.2413,
+      "step": 12107
+    },
+    {
+      "epoch": 1.93728,
+      "grad_norm": 2.192030486011756,
+      "learning_rate": 5.917542513737984e-05,
+      "loss": 1.2648,
+      "step": 12108
+    },
+    {
+      "epoch": 1.93744,
+      "grad_norm": 1.8579027214446668,
+      "learning_rate": 5.915965694367231e-05,
+      "loss": 1.2375,
+      "step": 12109
+    },
+    {
+      "epoch": 1.9376,
+      "grad_norm": 1.9664317175573864,
+      "learning_rate": 5.914388996858105e-05,
+      "loss": 1.1928,
+      "step": 12110
+    },
+    {
+      "epoch": 1.93776,
+      "grad_norm": 1.9916503453006857,
+      "learning_rate": 5.912812421257663e-05,
+      "loss": 1.2285,
+      "step": 12111
+    },
+    {
+      "epoch": 1.93792,
+      "grad_norm": 2.832489614687507,
+      "learning_rate": 5.911235967612938e-05,
+      "loss": 1.1848,
+      "step": 12112
+    },
+    {
+      "epoch": 1.93808,
+      "grad_norm": 1.894351537013268,
+      "learning_rate": 5.909659635970977e-05,
+      "loss": 1.2158,
+      "step": 12113
+    },
+    {
+      "epoch": 1.93824,
+      "grad_norm": 1.9180494928220002,
+      "learning_rate": 5.9080834263788074e-05,
+      "loss": 1.2533,
+      "step": 12114
+    },
+    {
+      "epoch": 1.9384000000000001,
+      "grad_norm": 1.9159519366215927,
+      "learning_rate": 5.9065073388834734e-05,
+      "loss": 1.272,
+      "step": 12115
+    },
+    {
+      "epoch": 1.9385599999999998,
+      "grad_norm": 2.0320352810184827,
+      "learning_rate": 5.90493137353199e-05,
+      "loss": 1.1892,
+      "step": 12116
+    },
+    {
+      "epoch": 1.93872,
+      "grad_norm": 2.092955400094437,
+      "learning_rate": 5.90335553037139e-05,
+      "loss": 1.1988,
+      "step": 12117
+    },
+    {
+      "epoch": 1.9388800000000002,
+      "grad_norm": 2.1352161052303265,
+      "learning_rate": 5.901779809448689e-05,
+      "loss": 1.2365,
+      "step": 12118
+    },
+    {
+      "epoch": 1.9390399999999999,
+      "grad_norm": 1.9890820875536885,
+      "learning_rate": 5.9002042108109116e-05,
+      "loss": 1.332,
+      "step": 12119
+    },
+    {
+      "epoch": 1.9392,
+      "grad_norm": 2.412684993846126,
+      "learning_rate": 5.898628734505061e-05,
+      "loss": 1.2411,
+      "step": 12120
+    },
+    {
+      "epoch": 1.93936,
+      "grad_norm": 1.9634943113104213,
+      "learning_rate": 5.8970533805781546e-05,
+      "loss": 1.1974,
+      "step": 12121
+    },
+    {
+      "epoch": 1.93952,
+      "grad_norm": 1.8865383483420943,
+      "learning_rate": 5.895478149077195e-05,
+      "loss": 1.1925,
+      "step": 12122
+    },
+    {
+      "epoch": 1.93968,
+      "grad_norm": 2.245397695322255,
+      "learning_rate": 5.893903040049191e-05,
+      "loss": 1.2454,
+      "step": 12123
+    },
+    {
+      "epoch": 1.93984,
+      "grad_norm": 1.857631956385269,
+      "learning_rate": 5.892328053541132e-05,
+      "loss": 1.187,
+      "step": 12124
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 1.9161541019551904,
+      "learning_rate": 5.8907531896000237e-05,
+      "loss": 1.2361,
+      "step": 12125
+    },
+    {
+      "epoch": 1.94016,
+      "grad_norm": 2.1513634050841315,
+      "learning_rate": 5.8891784482728475e-05,
+      "loss": 1.2083,
+      "step": 12126
+    },
+    {
+      "epoch": 1.94032,
+      "grad_norm": 2.1154798215819683,
+      "learning_rate": 5.8876038296066e-05,
+      "loss": 1.2408,
+      "step": 12127
+    },
+    {
+      "epoch": 1.94048,
+      "grad_norm": 1.870494293639263,
+      "learning_rate": 5.8860293336482595e-05,
+      "loss": 1.229,
+      "step": 12128
+    },
+    {
+      "epoch": 1.9406400000000001,
+      "grad_norm": 1.9210746892726527,
+      "learning_rate": 5.8844549604448094e-05,
+      "loss": 1.197,
+      "step": 12129
+    },
+    {
+      "epoch": 1.9407999999999999,
+      "grad_norm": 1.8833188138172166,
+      "learning_rate": 5.882880710043225e-05,
+      "loss": 1.1983,
+      "step": 12130
+    },
+    {
+      "epoch": 1.94096,
+      "grad_norm": 2.3317499042845298,
+      "learning_rate": 5.881306582490485e-05,
+      "loss": 1.2637,
+      "step": 12131
+    },
+    {
+      "epoch": 1.94112,
+      "grad_norm": 1.815450826874383,
+      "learning_rate": 5.87973257783355e-05,
+      "loss": 1.2318,
+      "step": 12132
+    },
+    {
+      "epoch": 1.94128,
+      "grad_norm": 1.870152242670611,
+      "learning_rate": 5.8781586961193956e-05,
+      "loss": 1.2137,
+      "step": 12133
+    },
+    {
+      "epoch": 1.94144,
+      "grad_norm": 1.9575449641220635,
+      "learning_rate": 5.876584937394975e-05,
+      "loss": 1.213,
+      "step": 12134
+    },
+    {
+      "epoch": 1.9416,
+      "grad_norm": 1.8973002462787276,
+      "learning_rate": 5.875011301707257e-05,
+      "loss": 1.2469,
+      "step": 12135
+    },
+    {
+      "epoch": 1.94176,
+      "grad_norm": 1.8265660958426406,
+      "learning_rate": 5.8734377891031865e-05,
+      "loss": 1.2011,
+      "step": 12136
+    },
+    {
+      "epoch": 1.94192,
+      "grad_norm": 2.1192234065918596,
+      "learning_rate": 5.871864399629721e-05,
+      "loss": 1.2133,
+      "step": 12137
+    },
+    {
+      "epoch": 1.94208,
+      "grad_norm": 1.83142276621561,
+      "learning_rate": 5.870291133333805e-05,
+      "loss": 1.1634,
+      "step": 12138
+    },
+    {
+      "epoch": 1.94224,
+      "grad_norm": 1.8107868846254338,
+      "learning_rate": 5.868717990262387e-05,
+      "loss": 1.2288,
+      "step": 12139
+    },
+    {
+      "epoch": 1.9424000000000001,
+      "grad_norm": 2.0525410899564642,
+      "learning_rate": 5.8671449704624004e-05,
+      "loss": 1.2652,
+      "step": 12140
+    },
+    {
+      "epoch": 1.9425599999999998,
+      "grad_norm": 1.9429521955836098,
+      "learning_rate": 5.8655720739807896e-05,
+      "loss": 1.2269,
+      "step": 12141
+    },
+    {
+      "epoch": 1.94272,
+      "grad_norm": 1.465421850561763,
+      "learning_rate": 5.86399930086448e-05,
+      "loss": 1.2103,
+      "step": 12142
+    },
+    {
+      "epoch": 1.9428800000000002,
+      "grad_norm": 1.8966551421386375,
+      "learning_rate": 5.8624266511604066e-05,
+      "loss": 1.2076,
+      "step": 12143
+    },
+    {
+      "epoch": 1.9430399999999999,
+      "grad_norm": 2.033856938956495,
+      "learning_rate": 5.8608541249154936e-05,
+      "loss": 1.2459,
+      "step": 12144
+    },
+    {
+      "epoch": 1.9432,
+      "grad_norm": 2.0137611738723584,
+      "learning_rate": 5.8592817221766614e-05,
+      "loss": 1.232,
+      "step": 12145
+    },
+    {
+      "epoch": 1.94336,
+      "grad_norm": 1.9488613211421657,
+      "learning_rate": 5.857709442990826e-05,
+      "loss": 1.2126,
+      "step": 12146
+    },
+    {
+      "epoch": 1.94352,
+      "grad_norm": 2.0801003597502774,
+      "learning_rate": 5.8561372874049105e-05,
+      "loss": 1.2243,
+      "step": 12147
+    },
+    {
+      "epoch": 1.94368,
+      "grad_norm": 1.8454089607255941,
+      "learning_rate": 5.8545652554658146e-05,
+      "loss": 1.1982,
+      "step": 12148
+    },
+    {
+      "epoch": 1.94384,
+      "grad_norm": 2.0374251850677787,
+      "learning_rate": 5.852993347220456e-05,
+      "loss": 1.2409,
+      "step": 12149
+    },
+    {
+      "epoch": 1.944,
+      "grad_norm": 2.3492087786113696,
+      "learning_rate": 5.8514215627157286e-05,
+      "loss": 1.2125,
+      "step": 12150
+    },
+    {
+      "epoch": 1.94416,
+      "grad_norm": 1.9420569876523164,
+      "learning_rate": 5.849849901998539e-05,
+      "loss": 1.2379,
+      "step": 12151
+    },
+    {
+      "epoch": 1.94432,
+      "grad_norm": 2.1653390274140527,
+      "learning_rate": 5.84827836511578e-05,
+      "loss": 1.2087,
+      "step": 12152
+    },
+    {
+      "epoch": 1.94448,
+      "grad_norm": 1.887800322194097,
+      "learning_rate": 5.846706952114346e-05,
+      "loss": 1.2461,
+      "step": 12153
+    },
+    {
+      "epoch": 1.9446400000000001,
+      "grad_norm": 1.8446446927886428,
+      "learning_rate": 5.845135663041121e-05,
+      "loss": 1.275,
+      "step": 12154
+    },
+    {
+      "epoch": 1.9447999999999999,
+      "grad_norm": 1.8869394382995361,
+      "learning_rate": 5.843564497942999e-05,
+      "loss": 1.2848,
+      "step": 12155
+    },
+    {
+      "epoch": 1.94496,
+      "grad_norm": 1.956212628998272,
+      "learning_rate": 5.8419934568668525e-05,
+      "loss": 1.2065,
+      "step": 12156
+    },
+    {
+      "epoch": 1.94512,
+      "grad_norm": 2.1207441053631095,
+      "learning_rate": 5.840422539859565e-05,
+      "loss": 1.269,
+      "step": 12157
+    },
+    {
+      "epoch": 1.94528,
+      "grad_norm": 1.8480400264007115,
+      "learning_rate": 5.838851746968004e-05,
+      "loss": 1.2933,
+      "step": 12158
+    },
+    {
+      "epoch": 1.94544,
+      "grad_norm": 1.9379883358237655,
+      "learning_rate": 5.837281078239043e-05,
+      "loss": 1.248,
+      "step": 12159
+    },
+    {
+      "epoch": 1.9456,
+      "grad_norm": 2.257445867739843,
+      "learning_rate": 5.835710533719552e-05,
+      "loss": 1.3332,
+      "step": 12160
+    },
+    {
+      "epoch": 1.94576,
+      "grad_norm": 1.9474859044476016,
+      "learning_rate": 5.83414011345639e-05,
+      "loss": 1.2719,
+      "step": 12161
+    },
+    {
+      "epoch": 1.94592,
+      "grad_norm": 1.8705328481902117,
+      "learning_rate": 5.832569817496416e-05,
+      "loss": 1.1947,
+      "step": 12162
+    },
+    {
+      "epoch": 1.94608,
+      "grad_norm": 1.695651461848682,
+      "learning_rate": 5.830999645886488e-05,
+      "loss": 1.2172,
+      "step": 12163
+    },
+    {
+      "epoch": 1.94624,
+      "grad_norm": 2.060861244248941,
+      "learning_rate": 5.829429598673455e-05,
+      "loss": 1.218,
+      "step": 12164
+    },
+    {
+      "epoch": 1.9464000000000001,
+      "grad_norm": 1.8394362624560594,
+      "learning_rate": 5.827859675904164e-05,
+      "loss": 1.2537,
+      "step": 12165
+    },
+    {
+      "epoch": 1.9465599999999998,
+      "grad_norm": 1.7452538665800557,
+      "learning_rate": 5.826289877625461e-05,
+      "loss": 1.2285,
+      "step": 12166
+    },
+    {
+      "epoch": 1.94672,
+      "grad_norm": 1.959181015754981,
+      "learning_rate": 5.824720203884184e-05,
+      "loss": 1.1859,
+      "step": 12167
+    },
+    {
+      "epoch": 1.9468800000000002,
+      "grad_norm": 1.9452682886425192,
+      "learning_rate": 5.8231506547271805e-05,
+      "loss": 1.2858,
+      "step": 12168
+    },
+    {
+      "epoch": 1.9470399999999999,
+      "grad_norm": 1.8343594845756304,
+      "learning_rate": 5.821581230201265e-05,
+      "loss": 1.2591,
+      "step": 12169
+    },
+    {
+      "epoch": 1.9472,
+      "grad_norm": 2.242108432697394,
+      "learning_rate": 5.8200119303532865e-05,
+      "loss": 1.2642,
+      "step": 12170
+    },
+    {
+      "epoch": 1.94736,
+      "grad_norm": 1.8212702020305862,
+      "learning_rate": 5.8184427552300534e-05,
+      "loss": 1.2906,
+      "step": 12171
+    },
+    {
+      "epoch": 1.94752,
+      "grad_norm": 1.903227187903569,
+      "learning_rate": 5.8168737048784014e-05,
+      "loss": 1.2222,
+      "step": 12172
+    },
+    {
+      "epoch": 1.94768,
+      "grad_norm": 1.9155101866972324,
+      "learning_rate": 5.815304779345136e-05,
+      "loss": 1.2486,
+      "step": 12173
+    },
+    {
+      "epoch": 1.94784,
+      "grad_norm": 1.9055354356081728,
+      "learning_rate": 5.813735978677082e-05,
+      "loss": 1.2412,
+      "step": 12174
+    },
+    {
+      "epoch": 1.948,
+      "grad_norm": 2.1880965237001,
+      "learning_rate": 5.812167302921047e-05,
+      "loss": 1.2287,
+      "step": 12175
+    },
+    {
+      "epoch": 1.9481600000000001,
+      "grad_norm": 1.9264228978674798,
+      "learning_rate": 5.810598752123837e-05,
+      "loss": 1.1912,
+      "step": 12176
+    },
+    {
+      "epoch": 1.94832,
+      "grad_norm": 1.9514162617542108,
+      "learning_rate": 5.809030326332254e-05,
+      "loss": 1.2548,
+      "step": 12177
+    },
+    {
+      "epoch": 1.94848,
+      "grad_norm": 1.8659249135787974,
+      "learning_rate": 5.8074620255931014e-05,
+      "loss": 1.1737,
+      "step": 12178
+    },
+    {
+      "epoch": 1.9486400000000001,
+      "grad_norm": 2.2263696583558694,
+      "learning_rate": 5.805893849953172e-05,
+      "loss": 1.2686,
+      "step": 12179
+    },
+    {
+      "epoch": 1.9487999999999999,
+      "grad_norm": 2.0403792945047816,
+      "learning_rate": 5.804325799459258e-05,
+      "loss": 1.2536,
+      "step": 12180
+    },
+    {
+      "epoch": 1.94896,
+      "grad_norm": 1.961716792315799,
+      "learning_rate": 5.802757874158149e-05,
+      "loss": 1.2464,
+      "step": 12181
+    },
+    {
+      "epoch": 1.94912,
+      "grad_norm": 2.1222429391408566,
+      "learning_rate": 5.8011900740966295e-05,
+      "loss": 1.2606,
+      "step": 12182
+    },
+    {
+      "epoch": 1.94928,
+      "grad_norm": 2.2237843070198777,
+      "learning_rate": 5.799622399321475e-05,
+      "loss": 1.2619,
+      "step": 12183
+    },
+    {
+      "epoch": 1.94944,
+      "grad_norm": 2.037362704062041,
+      "learning_rate": 5.798054849879476e-05,
+      "loss": 1.2164,
+      "step": 12184
+    },
+    {
+      "epoch": 1.9496,
+      "grad_norm": 2.3005444236876125,
+      "learning_rate": 5.7964874258173894e-05,
+      "loss": 1.1813,
+      "step": 12185
+    },
+    {
+      "epoch": 1.94976,
+      "grad_norm": 1.872412560701496,
+      "learning_rate": 5.794920127182001e-05,
+      "loss": 1.1793,
+      "step": 12186
+    },
+    {
+      "epoch": 1.94992,
+      "grad_norm": 2.1404765721371524,
+      "learning_rate": 5.793352954020059e-05,
+      "loss": 1.2238,
+      "step": 12187
+    },
+    {
+      "epoch": 1.95008,
+      "grad_norm": 2.374061603523279,
+      "learning_rate": 5.7917859063783456e-05,
+      "loss": 1.3101,
+      "step": 12188
+    },
+    {
+      "epoch": 1.95024,
+      "grad_norm": 1.8589676907496655,
+      "learning_rate": 5.7902189843035995e-05,
+      "loss": 1.2364,
+      "step": 12189
+    },
+    {
+      "epoch": 1.9504000000000001,
+      "grad_norm": 1.9857669350858431,
+      "learning_rate": 5.788652187842588e-05,
+      "loss": 1.1784,
+      "step": 12190
+    },
+    {
+      "epoch": 1.9505599999999998,
+      "grad_norm": 1.963899356828068,
+      "learning_rate": 5.787085517042059e-05,
+      "loss": 1.192,
+      "step": 12191
+    },
+    {
+      "epoch": 1.95072,
+      "grad_norm": 1.722207900424737,
+      "learning_rate": 5.78551897194876e-05,
+      "loss": 1.2716,
+      "step": 12192
+    },
+    {
+      "epoch": 1.95088,
+      "grad_norm": 1.8399119188445974,
+      "learning_rate": 5.783952552609433e-05,
+      "loss": 1.2404,
+      "step": 12193
+    },
+    {
+      "epoch": 1.9510399999999999,
+      "grad_norm": 1.942113509088583,
+      "learning_rate": 5.78238625907082e-05,
+      "loss": 1.221,
+      "step": 12194
+    },
+    {
+      "epoch": 1.9512,
+      "grad_norm": 2.07148847568198,
+      "learning_rate": 5.780820091379653e-05,
+      "loss": 1.2274,
+      "step": 12195
+    },
+    {
+      "epoch": 1.95136,
+      "grad_norm": 1.93297460920355,
+      "learning_rate": 5.7792540495826676e-05,
+      "loss": 1.2224,
+      "step": 12196
+    },
+    {
+      "epoch": 1.95152,
+      "grad_norm": 2.036658229305831,
+      "learning_rate": 5.777688133726591e-05,
+      "loss": 1.2183,
+      "step": 12197
+    },
+    {
+      "epoch": 1.95168,
+      "grad_norm": 2.484451725866535,
+      "learning_rate": 5.776122343858148e-05,
+      "loss": 1.1986,
+      "step": 12198
+    },
+    {
+      "epoch": 1.95184,
+      "grad_norm": 1.8976635461079678,
+      "learning_rate": 5.774556680024055e-05,
+      "loss": 1.2352,
+      "step": 12199
+    },
+    {
+      "epoch": 1.952,
+      "grad_norm": 1.9759415746034048,
+      "learning_rate": 5.7729911422710425e-05,
+      "loss": 1.2404,
+      "step": 12200
+    },
+    {
+      "epoch": 1.9521600000000001,
+      "grad_norm": 2.1119740513025205,
+      "learning_rate": 5.771425730645805e-05,
+      "loss": 1.2353,
+      "step": 12201
+    },
+    {
+      "epoch": 1.95232,
+      "grad_norm": 2.16178207268473,
+      "learning_rate": 5.769860445195072e-05,
+      "loss": 1.2649,
+      "step": 12202
+    },
+    {
+      "epoch": 1.95248,
+      "grad_norm": 2.0313011765838986,
+      "learning_rate": 5.76829528596553e-05,
+      "loss": 1.1718,
+      "step": 12203
+    },
+    {
+      "epoch": 1.9526400000000002,
+      "grad_norm": 1.9548978908394699,
+      "learning_rate": 5.7667302530039e-05,
+      "loss": 1.2442,
+      "step": 12204
+    },
+    {
+      "epoch": 1.9527999999999999,
+      "grad_norm": 1.8165083511527178,
+      "learning_rate": 5.7651653463568623e-05,
+      "loss": 1.3101,
+      "step": 12205
+    },
+    {
+      "epoch": 1.95296,
+      "grad_norm": 2.157434401857494,
+      "learning_rate": 5.7636005660711234e-05,
+      "loss": 1.1687,
+      "step": 12206
+    },
+    {
+      "epoch": 1.95312,
+      "grad_norm": 2.0220800099656295,
+      "learning_rate": 5.7620359121933695e-05,
+      "loss": 1.2594,
+      "step": 12207
+    },
+    {
+      "epoch": 1.95328,
+      "grad_norm": 1.9443648309736632,
+      "learning_rate": 5.760471384770289e-05,
+      "loss": 1.2172,
+      "step": 12208
+    },
+    {
+      "epoch": 1.95344,
+      "grad_norm": 2.062681548689023,
+      "learning_rate": 5.758906983848566e-05,
+      "loss": 1.2186,
+      "step": 12209
+    },
+    {
+      "epoch": 1.9536,
+      "grad_norm": 2.013103030692419,
+      "learning_rate": 5.757342709474878e-05,
+      "loss": 1.2428,
+      "step": 12210
+    },
+    {
+      "epoch": 1.95376,
+      "grad_norm": 1.8157442690096843,
+      "learning_rate": 5.755778561695902e-05,
+      "loss": 1.2455,
+      "step": 12211
+    },
+    {
+      "epoch": 1.95392,
+      "grad_norm": 1.914823979515427,
+      "learning_rate": 5.754214540558308e-05,
+      "loss": 1.1837,
+      "step": 12212
+    },
+    {
+      "epoch": 1.95408,
+      "grad_norm": 1.9328974717967151,
+      "learning_rate": 5.752650646108766e-05,
+      "loss": 1.282,
+      "step": 12213
+    },
+    {
+      "epoch": 1.95424,
+      "grad_norm": 2.0827093938549037,
+      "learning_rate": 5.75108687839394e-05,
+      "loss": 1.2526,
+      "step": 12214
+    },
+    {
+      "epoch": 1.9544000000000001,
+      "grad_norm": 1.9228077020096979,
+      "learning_rate": 5.749523237460486e-05,
+      "loss": 1.2323,
+      "step": 12215
+    },
+    {
+      "epoch": 1.9545599999999999,
+      "grad_norm": 2.00033207344718,
+      "learning_rate": 5.747959723355072e-05,
+      "loss": 1.1716,
+      "step": 12216
+    },
+    {
+      "epoch": 1.95472,
+      "grad_norm": 1.8845856648284929,
+      "learning_rate": 5.746396336124338e-05,
+      "loss": 1.3222,
+      "step": 12217
+    },
+    {
+      "epoch": 1.95488,
+      "grad_norm": 1.885141842752306,
+      "learning_rate": 5.7448330758149415e-05,
+      "loss": 1.1638,
+      "step": 12218
+    },
+    {
+      "epoch": 1.95504,
+      "grad_norm": 1.7113503142814006,
+      "learning_rate": 5.743269942473526e-05,
+      "loss": 1.2301,
+      "step": 12219
+    },
+    {
+      "epoch": 1.9552,
+      "grad_norm": 1.9688670997467013,
+      "learning_rate": 5.741706936146734e-05,
+      "loss": 1.2122,
+      "step": 12220
+    },
+    {
+      "epoch": 1.95536,
+      "grad_norm": 1.829833015300241,
+      "learning_rate": 5.740144056881199e-05,
+      "loss": 1.219,
+      "step": 12221
+    },
+    {
+      "epoch": 1.95552,
+      "grad_norm": 1.8633831955516147,
+      "learning_rate": 5.73858130472356e-05,
+      "loss": 1.1726,
+      "step": 12222
+    },
+    {
+      "epoch": 1.95568,
+      "grad_norm": 2.121593421289797,
+      "learning_rate": 5.737018679720445e-05,
+      "loss": 1.248,
+      "step": 12223
+    },
+    {
+      "epoch": 1.95584,
+      "grad_norm": 2.1386575870316977,
+      "learning_rate": 5.735456181918477e-05,
+      "loss": 1.2345,
+      "step": 12224
+    },
+    {
+      "epoch": 1.956,
+      "grad_norm": 1.7842280487508142,
+      "learning_rate": 5.7338938113642905e-05,
+      "loss": 1.1874,
+      "step": 12225
+    },
+    {
+      "epoch": 1.9561600000000001,
+      "grad_norm": 2.4034324972306003,
+      "learning_rate": 5.732331568104489e-05,
+      "loss": 1.2472,
+      "step": 12226
+    },
+    {
+      "epoch": 1.9563199999999998,
+      "grad_norm": 1.9668944470516678,
+      "learning_rate": 5.730769452185702e-05,
+      "loss": 1.275,
+      "step": 12227
+    },
+    {
+      "epoch": 1.95648,
+      "grad_norm": 1.7902219097083554,
+      "learning_rate": 5.7292074636545255e-05,
+      "loss": 1.2609,
+      "step": 12228
+    },
+    {
+      "epoch": 1.9566400000000002,
+      "grad_norm": 2.2656130935812375,
+      "learning_rate": 5.7276456025575834e-05,
+      "loss": 1.1934,
+      "step": 12229
+    },
+    {
+      "epoch": 1.9567999999999999,
+      "grad_norm": 1.736514056729276,
+      "learning_rate": 5.7260838689414634e-05,
+      "loss": 1.2337,
+      "step": 12230
+    },
+    {
+      "epoch": 1.95696,
+      "grad_norm": 2.289314253289351,
+      "learning_rate": 5.724522262852776e-05,
+      "loss": 1.2932,
+      "step": 12231
+    },
+    {
+      "epoch": 1.95712,
+      "grad_norm": 2.032328219610249,
+      "learning_rate": 5.722960784338114e-05,
+      "loss": 1.2467,
+      "step": 12232
+    },
+    {
+      "epoch": 1.95728,
+      "grad_norm": 1.7432469157372212,
+      "learning_rate": 5.7213994334440715e-05,
+      "loss": 1.1763,
+      "step": 12233
+    },
+    {
+      "epoch": 1.95744,
+      "grad_norm": 1.8823394226429333,
+      "learning_rate": 5.719838210217234e-05,
+      "loss": 1.2267,
+      "step": 12234
+    },
+    {
+      "epoch": 1.9576,
+      "grad_norm": 2.133599658370982,
+      "learning_rate": 5.718277114704188e-05,
+      "loss": 1.2364,
+      "step": 12235
+    },
+    {
+      "epoch": 1.95776,
+      "grad_norm": 1.9015633727522094,
+      "learning_rate": 5.716716146951514e-05,
+      "loss": 1.2227,
+      "step": 12236
+    },
+    {
+      "epoch": 1.95792,
+      "grad_norm": 1.957628300482707,
+      "learning_rate": 5.7151553070057887e-05,
+      "loss": 1.2207,
+      "step": 12237
+    },
+    {
+      "epoch": 1.95808,
+      "grad_norm": 1.7493168597656548,
+      "learning_rate": 5.713594594913585e-05,
+      "loss": 1.1764,
+      "step": 12238
+    },
+    {
+      "epoch": 1.95824,
+      "grad_norm": 2.1677155209801486,
+      "learning_rate": 5.712034010721472e-05,
+      "loss": 1.2672,
+      "step": 12239
+    },
+    {
+      "epoch": 1.9584000000000001,
+      "grad_norm": 1.9276918298771941,
+      "learning_rate": 5.710473554476014e-05,
+      "loss": 1.2385,
+      "step": 12240
+    },
+    {
+      "epoch": 1.9585599999999999,
+      "grad_norm": 1.9133495122055328,
+      "learning_rate": 5.70891322622378e-05,
+      "loss": 1.235,
+      "step": 12241
+    },
+    {
+      "epoch": 1.95872,
+      "grad_norm": 1.9460243785258622,
+      "learning_rate": 5.707353026011316e-05,
+      "loss": 1.2318,
+      "step": 12242
+    },
+    {
+      "epoch": 1.95888,
+      "grad_norm": 2.0455348253404346,
+      "learning_rate": 5.705792953885191e-05,
+      "loss": 1.2065,
+      "step": 12243
+    },
+    {
+      "epoch": 1.95904,
+      "grad_norm": 1.6190057048763686,
+      "learning_rate": 5.7042330098919384e-05,
+      "loss": 1.2198,
+      "step": 12244
+    },
+    {
+      "epoch": 1.9592,
+      "grad_norm": 1.9405284425415632,
+      "learning_rate": 5.702673194078121e-05,
+      "loss": 1.1642,
+      "step": 12245
+    },
+    {
+      "epoch": 1.95936,
+      "grad_norm": 2.219802446805258,
+      "learning_rate": 5.7011135064902656e-05,
+      "loss": 1.2464,
+      "step": 12246
+    },
+    {
+      "epoch": 1.95952,
+      "grad_norm": 1.9368814564121923,
+      "learning_rate": 5.699553947174924e-05,
+      "loss": 1.2351,
+      "step": 12247
+    },
+    {
+      "epoch": 1.95968,
+      "grad_norm": 2.0708734121219754,
+      "learning_rate": 5.697994516178625e-05,
+      "loss": 1.2008,
+      "step": 12248
+    },
+    {
+      "epoch": 1.95984,
+      "grad_norm": 2.019724356437286,
+      "learning_rate": 5.696435213547902e-05,
+      "loss": 1.1825,
+      "step": 12249
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 2.005848954984117,
+      "learning_rate": 5.6948760393292796e-05,
+      "loss": 1.2138,
+      "step": 12250
+    },
+    {
+      "epoch": 1.9601600000000001,
+      "grad_norm": 2.0965759546124856,
+      "learning_rate": 5.693316993569284e-05,
+      "loss": 1.2165,
+      "step": 12251
+    },
+    {
+      "epoch": 1.9603199999999998,
+      "grad_norm": 1.7756883933833583,
+      "learning_rate": 5.691758076314435e-05,
+      "loss": 1.2031,
+      "step": 12252
+    },
+    {
+      "epoch": 1.96048,
+      "grad_norm": 2.2004340397627358,
+      "learning_rate": 5.6901992876112466e-05,
+      "loss": 1.2671,
+      "step": 12253
+    },
+    {
+      "epoch": 1.9606400000000002,
+      "grad_norm": 2.12841784431043,
+      "learning_rate": 5.6886406275062296e-05,
+      "loss": 1.2503,
+      "step": 12254
+    },
+    {
+      "epoch": 1.9607999999999999,
+      "grad_norm": 2.0476093878785773,
+      "learning_rate": 5.687082096045896e-05,
+      "loss": 1.24,
+      "step": 12255
+    },
+    {
+      "epoch": 1.96096,
+      "grad_norm": 1.826079600499125,
+      "learning_rate": 5.685523693276743e-05,
+      "loss": 1.1988,
+      "step": 12256
+    },
+    {
+      "epoch": 1.96112,
+      "grad_norm": 2.2471667335982306,
+      "learning_rate": 5.683965419245285e-05,
+      "loss": 1.2316,
+      "step": 12257
+    },
+    {
+      "epoch": 1.96128,
+      "grad_norm": 1.9689383407225651,
+      "learning_rate": 5.6824072739980004e-05,
+      "loss": 1.1666,
+      "step": 12258
+    },
+    {
+      "epoch": 1.96144,
+      "grad_norm": 1.8829977957715351,
+      "learning_rate": 5.6808492575814e-05,
+      "loss": 1.2093,
+      "step": 12259
+    },
+    {
+      "epoch": 1.9616,
+      "grad_norm": 2.4128134147107527,
+      "learning_rate": 5.679291370041957e-05,
+      "loss": 1.2142,
+      "step": 12260
+    },
+    {
+      "epoch": 1.96176,
+      "grad_norm": 2.717571152507479,
+      "learning_rate": 5.67773361142617e-05,
+      "loss": 1.2422,
+      "step": 12261
+    },
+    {
+      "epoch": 1.96192,
+      "grad_norm": 1.7914426665968652,
+      "learning_rate": 5.676175981780507e-05,
+      "loss": 1.2316,
+      "step": 12262
+    },
+    {
+      "epoch": 1.96208,
+      "grad_norm": 2.252465690050173,
+      "learning_rate": 5.674618481151457e-05,
+      "loss": 1.1855,
+      "step": 12263
+    },
+    {
+      "epoch": 1.96224,
+      "grad_norm": 2.7094494594923946,
+      "learning_rate": 5.673061109585487e-05,
+      "loss": 1.2342,
+      "step": 12264
+    },
+    {
+      "epoch": 1.9624000000000001,
+      "grad_norm": 1.5759628455518981,
+      "learning_rate": 5.6715038671290686e-05,
+      "loss": 1.1713,
+      "step": 12265
+    },
+    {
+      "epoch": 1.9625599999999999,
+      "grad_norm": 1.8746362118964,
+      "learning_rate": 5.6699467538286685e-05,
+      "loss": 1.2303,
+      "step": 12266
+    },
+    {
+      "epoch": 1.96272,
+      "grad_norm": 2.052287195243602,
+      "learning_rate": 5.668389769730747e-05,
+      "loss": 1.1672,
+      "step": 12267
+    },
+    {
+      "epoch": 1.96288,
+      "grad_norm": 1.595586179392782,
+      "learning_rate": 5.666832914881763e-05,
+      "loss": 1.2108,
+      "step": 12268
+    },
+    {
+      "epoch": 1.96304,
+      "grad_norm": 1.9521260068522182,
+      "learning_rate": 5.6652761893281726e-05,
+      "loss": 1.2821,
+      "step": 12269
+    },
+    {
+      "epoch": 1.9632,
+      "grad_norm": 1.9395869720089032,
+      "learning_rate": 5.663719593116421e-05,
+      "loss": 1.2202,
+      "step": 12270
+    },
+    {
+      "epoch": 1.96336,
+      "grad_norm": 2.0351963258637324,
+      "learning_rate": 5.662163126292961e-05,
+      "loss": 1.1619,
+      "step": 12271
+    },
+    {
+      "epoch": 1.96352,
+      "grad_norm": 2.1807573056737946,
+      "learning_rate": 5.660606788904227e-05,
+      "loss": 1.1971,
+      "step": 12272
+    },
+    {
+      "epoch": 1.96368,
+      "grad_norm": 2.0762662234132976,
+      "learning_rate": 5.659050580996672e-05,
+      "loss": 1.2471,
+      "step": 12273
+    },
+    {
+      "epoch": 1.96384,
+      "grad_norm": 1.728910196569553,
+      "learning_rate": 5.6574945026167134e-05,
+      "loss": 1.2395,
+      "step": 12274
+    },
+    {
+      "epoch": 1.964,
+      "grad_norm": 2.4352490521492083,
+      "learning_rate": 5.655938553810799e-05,
+      "loss": 1.1918,
+      "step": 12275
+    },
+    {
+      "epoch": 1.9641600000000001,
+      "grad_norm": 2.724768237880957,
+      "learning_rate": 5.654382734625341e-05,
+      "loss": 1.2156,
+      "step": 12276
+    },
+    {
+      "epoch": 1.9643199999999998,
+      "grad_norm": 1.9630033826289826,
+      "learning_rate": 5.6528270451067735e-05,
+      "loss": 1.226,
+      "step": 12277
+    },
+    {
+      "epoch": 1.96448,
+      "grad_norm": 2.463656957065451,
+      "learning_rate": 5.6512714853015124e-05,
+      "loss": 1.2445,
+      "step": 12278
+    },
+    {
+      "epoch": 1.9646400000000002,
+      "grad_norm": 2.071385265478601,
+      "learning_rate": 5.6497160552559736e-05,
+      "loss": 1.2049,
+      "step": 12279
+    },
+    {
+      "epoch": 1.9647999999999999,
+      "grad_norm": 2.003679900937499,
+      "learning_rate": 5.648160755016569e-05,
+      "loss": 1.2054,
+      "step": 12280
+    },
+    {
+      "epoch": 1.96496,
+      "grad_norm": 1.8902285396807186,
+      "learning_rate": 5.646605584629702e-05,
+      "loss": 1.1911,
+      "step": 12281
+    },
+    {
+      "epoch": 1.96512,
+      "grad_norm": 2.5746225835914314,
+      "learning_rate": 5.6450505441417876e-05,
+      "loss": 1.2493,
+      "step": 12282
+    },
+    {
+      "epoch": 1.96528,
+      "grad_norm": 2.280677762014503,
+      "learning_rate": 5.643495633599213e-05,
+      "loss": 1.2643,
+      "step": 12283
+    },
+    {
+      "epoch": 1.96544,
+      "grad_norm": 2.2664714067823084,
+      "learning_rate": 5.641940853048387e-05,
+      "loss": 1.2109,
+      "step": 12284
+    },
+    {
+      "epoch": 1.9656,
+      "grad_norm": 2.0698108085402973,
+      "learning_rate": 5.640386202535689e-05,
+      "loss": 1.2519,
+      "step": 12285
+    },
+    {
+      "epoch": 1.96576,
+      "grad_norm": 2.170533522927101,
+      "learning_rate": 5.638831682107522e-05,
+      "loss": 1.1811,
+      "step": 12286
+    },
+    {
+      "epoch": 1.9659200000000001,
+      "grad_norm": 2.212001085220815,
+      "learning_rate": 5.637277291810254e-05,
+      "loss": 1.2238,
+      "step": 12287
+    },
+    {
+      "epoch": 1.96608,
+      "grad_norm": 2.271951283541751,
+      "learning_rate": 5.6357230316902786e-05,
+      "loss": 1.2925,
+      "step": 12288
+    },
+    {
+      "epoch": 1.96624,
+      "grad_norm": 2.372963082795009,
+      "learning_rate": 5.634168901793968e-05,
+      "loss": 1.2361,
+      "step": 12289
+    },
+    {
+      "epoch": 1.9664000000000001,
+      "grad_norm": 2.2929553604044743,
+      "learning_rate": 5.6326149021676966e-05,
+      "loss": 1.261,
+      "step": 12290
+    },
+    {
+      "epoch": 1.9665599999999999,
+      "grad_norm": 1.9199937993977236,
+      "learning_rate": 5.631061032857832e-05,
+      "loss": 1.1952,
+      "step": 12291
+    },
+    {
+      "epoch": 1.96672,
+      "grad_norm": 1.9682158596500536,
+      "learning_rate": 5.6295072939107406e-05,
+      "loss": 1.2191,
+      "step": 12292
+    },
+    {
+      "epoch": 1.96688,
+      "grad_norm": 2.1806552627026,
+      "learning_rate": 5.627953685372782e-05,
+      "loss": 1.2134,
+      "step": 12293
+    },
+    {
+      "epoch": 1.96704,
+      "grad_norm": 2.273581659273718,
+      "learning_rate": 5.626400207290316e-05,
+      "loss": 1.1963,
+      "step": 12294
+    },
+    {
+      "epoch": 1.9672,
+      "grad_norm": 2.2287744633807076,
+      "learning_rate": 5.624846859709694e-05,
+      "loss": 1.2341,
+      "step": 12295
+    },
+    {
+      "epoch": 1.96736,
+      "grad_norm": 2.244785927521898,
+      "learning_rate": 5.623293642677265e-05,
+      "loss": 1.2018,
+      "step": 12296
+    },
+    {
+      "epoch": 1.96752,
+      "grad_norm": 2.0034468853868024,
+      "learning_rate": 5.621740556239374e-05,
+      "loss": 1.2632,
+      "step": 12297
+    },
+    {
+      "epoch": 1.96768,
+      "grad_norm": 2.217438647408264,
+      "learning_rate": 5.620187600442374e-05,
+      "loss": 1.2165,
+      "step": 12298
+    },
+    {
+      "epoch": 1.96784,
+      "grad_norm": 1.9361770273414536,
+      "learning_rate": 5.618634775332584e-05,
+      "loss": 1.2862,
+      "step": 12299
+    },
+    {
+      "epoch": 1.968,
+      "grad_norm": 1.9302074430288878,
+      "learning_rate": 5.617082080956357e-05,
+      "loss": 1.2282,
+      "step": 12300
+    },
+    {
+      "epoch": 1.9681600000000001,
+      "grad_norm": 2.0507632371843685,
+      "learning_rate": 5.6155295173600054e-05,
+      "loss": 1.1906,
+      "step": 12301
+    },
+    {
+      "epoch": 1.9683199999999998,
+      "grad_norm": 2.2771428425900857,
+      "learning_rate": 5.613977084589872e-05,
+      "loss": 1.2021,
+      "step": 12302
+    },
+    {
+      "epoch": 1.96848,
+      "grad_norm": 2.117444976354279,
+      "learning_rate": 5.6124247826922646e-05,
+      "loss": 1.2472,
+      "step": 12303
+    },
+    {
+      "epoch": 1.96864,
+      "grad_norm": 2.5223325307064757,
+      "learning_rate": 5.6108726117135114e-05,
+      "loss": 1.2611,
+      "step": 12304
+    },
+    {
+      "epoch": 1.9687999999999999,
+      "grad_norm": 2.3247276483313404,
+      "learning_rate": 5.609320571699923e-05,
+      "loss": 1.3001,
+      "step": 12305
+    },
+    {
+      "epoch": 1.96896,
+      "grad_norm": 2.065826835985614,
+      "learning_rate": 5.607768662697812e-05,
+      "loss": 1.1997,
+      "step": 12306
+    },
+    {
+      "epoch": 1.96912,
+      "grad_norm": 2.203618074102745,
+      "learning_rate": 5.606216884753484e-05,
+      "loss": 1.241,
+      "step": 12307
+    },
+    {
+      "epoch": 1.96928,
+      "grad_norm": 1.88105976176387,
+      "learning_rate": 5.6046652379132424e-05,
+      "loss": 1.2242,
+      "step": 12308
+    },
+    {
+      "epoch": 1.96944,
+      "grad_norm": 2.7253219602403367,
+      "learning_rate": 5.603113722223383e-05,
+      "loss": 1.249,
+      "step": 12309
+    },
+    {
+      "epoch": 1.9696,
+      "grad_norm": 1.990987459779516,
+      "learning_rate": 5.601562337730206e-05,
+      "loss": 1.226,
+      "step": 12310
+    },
+    {
+      "epoch": 1.96976,
+      "grad_norm": 2.6525675777121873,
+      "learning_rate": 5.600011084479997e-05,
+      "loss": 1.1744,
+      "step": 12311
+    },
+    {
+      "epoch": 1.9699200000000001,
+      "grad_norm": 2.113824829482814,
+      "learning_rate": 5.598459962519047e-05,
+      "loss": 1.2486,
+      "step": 12312
+    },
+    {
+      "epoch": 1.97008,
+      "grad_norm": 1.7539231220821871,
+      "learning_rate": 5.5969089718936354e-05,
+      "loss": 1.1859,
+      "step": 12313
+    },
+    {
+      "epoch": 1.97024,
+      "grad_norm": 1.9076114608801087,
+      "learning_rate": 5.595358112650051e-05,
+      "loss": 1.2545,
+      "step": 12314
+    },
+    {
+      "epoch": 1.9704000000000002,
+      "grad_norm": 1.97910900152215,
+      "learning_rate": 5.593807384834553e-05,
+      "loss": 1.222,
+      "step": 12315
+    },
+    {
+      "epoch": 1.9705599999999999,
+      "grad_norm": 2.1890563218719987,
+      "learning_rate": 5.5922567884934306e-05,
+      "loss": 1.2133,
+      "step": 12316
+    },
+    {
+      "epoch": 1.97072,
+      "grad_norm": 2.179501072973133,
+      "learning_rate": 5.590706323672936e-05,
+      "loss": 1.2056,
+      "step": 12317
+    },
+    {
+      "epoch": 1.97088,
+      "grad_norm": 1.9828987873029054,
+      "learning_rate": 5.589155990419348e-05,
+      "loss": 1.2675,
+      "step": 12318
+    },
+    {
+      "epoch": 1.97104,
+      "grad_norm": 1.7153153293635617,
+      "learning_rate": 5.587605788778909e-05,
+      "loss": 1.1668,
+      "step": 12319
+    },
+    {
+      "epoch": 1.9712,
+      "grad_norm": 1.8176527850340658,
+      "learning_rate": 5.5860557187978865e-05,
+      "loss": 1.2386,
+      "step": 12320
+    },
+    {
+      "epoch": 1.97136,
+      "grad_norm": 1.8140413219324905,
+      "learning_rate": 5.584505780522532e-05,
+      "loss": 1.2386,
+      "step": 12321
+    },
+    {
+      "epoch": 1.97152,
+      "grad_norm": 1.924880689723719,
+      "learning_rate": 5.5829559739990886e-05,
+      "loss": 1.2208,
+      "step": 12322
+    },
+    {
+      "epoch": 1.97168,
+      "grad_norm": 1.9767520016310476,
+      "learning_rate": 5.5814062992738045e-05,
+      "loss": 1.2192,
+      "step": 12323
+    },
+    {
+      "epoch": 1.97184,
+      "grad_norm": 2.186911904555544,
+      "learning_rate": 5.579856756392918e-05,
+      "loss": 1.1924,
+      "step": 12324
+    },
+    {
+      "epoch": 1.972,
+      "grad_norm": 1.6189990479107332,
+      "learning_rate": 5.5783073454026645e-05,
+      "loss": 1.2252,
+      "step": 12325
+    },
+    {
+      "epoch": 1.9721600000000001,
+      "grad_norm": 1.8079853209166659,
+      "learning_rate": 5.5767580663492756e-05,
+      "loss": 1.2276,
+      "step": 12326
+    },
+    {
+      "epoch": 1.9723199999999999,
+      "grad_norm": 2.0748276251117552,
+      "learning_rate": 5.575208919278982e-05,
+      "loss": 1.2474,
+      "step": 12327
+    },
+    {
+      "epoch": 1.97248,
+      "grad_norm": 1.8095998521463428,
+      "learning_rate": 5.573659904238007e-05,
+      "loss": 1.2203,
+      "step": 12328
+    },
+    {
+      "epoch": 1.97264,
+      "grad_norm": 1.7476854876731918,
+      "learning_rate": 5.572111021272565e-05,
+      "loss": 1.2506,
+      "step": 12329
+    },
+    {
+      "epoch": 1.9727999999999999,
+      "grad_norm": 1.9769867595622255,
+      "learning_rate": 5.570562270428887e-05,
+      "loss": 1.2768,
+      "step": 12330
+    },
+    {
+      "epoch": 1.97296,
+      "grad_norm": 2.0880123363674095,
+      "learning_rate": 5.569013651753168e-05,
+      "loss": 1.2511,
+      "step": 12331
+    },
+    {
+      "epoch": 1.97312,
+      "grad_norm": 1.7833207882720565,
+      "learning_rate": 5.567465165291633e-05,
+      "loss": 1.1637,
+      "step": 12332
+    },
+    {
+      "epoch": 1.97328,
+      "grad_norm": 1.7550728599777194,
+      "learning_rate": 5.565916811090471e-05,
+      "loss": 1.2131,
+      "step": 12333
+    },
+    {
+      "epoch": 1.97344,
+      "grad_norm": 1.8743692582200504,
+      "learning_rate": 5.564368589195899e-05,
+      "loss": 1.2611,
+      "step": 12334
+    },
+    {
+      "epoch": 1.9736,
+      "grad_norm": 2.2533214239731674,
+      "learning_rate": 5.562820499654096e-05,
+      "loss": 1.2566,
+      "step": 12335
+    },
+    {
+      "epoch": 1.97376,
+      "grad_norm": 1.9398024124655324,
+      "learning_rate": 5.561272542511267e-05,
+      "loss": 1.2029,
+      "step": 12336
+    },
+    {
+      "epoch": 1.9739200000000001,
+      "grad_norm": 2.2318223374275914,
+      "learning_rate": 5.559724717813598e-05,
+      "loss": 1.2653,
+      "step": 12337
+    },
+    {
+      "epoch": 1.9740799999999998,
+      "grad_norm": 2.095686614431,
+      "learning_rate": 5.558177025607274e-05,
+      "loss": 1.266,
+      "step": 12338
+    },
+    {
+      "epoch": 1.97424,
+      "grad_norm": 2.219176643637414,
+      "learning_rate": 5.5566294659384745e-05,
+      "loss": 1.2317,
+      "step": 12339
+    },
+    {
+      "epoch": 1.9744000000000002,
+      "grad_norm": 2.491636491243517,
+      "learning_rate": 5.555082038853372e-05,
+      "loss": 1.1632,
+      "step": 12340
+    },
+    {
+      "epoch": 1.9745599999999999,
+      "grad_norm": 1.7630026581737497,
+      "learning_rate": 5.553534744398155e-05,
+      "loss": 1.2109,
+      "step": 12341
+    },
+    {
+      "epoch": 1.97472,
+      "grad_norm": 2.3284358362485404,
+      "learning_rate": 5.551987582618972e-05,
+      "loss": 1.2163,
+      "step": 12342
+    },
+    {
+      "epoch": 1.97488,
+      "grad_norm": 2.2405379939838315,
+      "learning_rate": 5.550440553562006e-05,
+      "loss": 1.1763,
+      "step": 12343
+    },
+    {
+      "epoch": 1.97504,
+      "grad_norm": 1.9293844858687637,
+      "learning_rate": 5.548893657273403e-05,
+      "loss": 1.2086,
+      "step": 12344
+    },
+    {
+      "epoch": 1.9752,
+      "grad_norm": 1.9734742034683281,
+      "learning_rate": 5.547346893799331e-05,
+      "loss": 1.2632,
+      "step": 12345
+    },
+    {
+      "epoch": 1.97536,
+      "grad_norm": 1.9793392084917056,
+      "learning_rate": 5.5458002631859394e-05,
+      "loss": 1.2038,
+      "step": 12346
+    },
+    {
+      "epoch": 1.97552,
+      "grad_norm": 2.0462587210753633,
+      "learning_rate": 5.544253765479378e-05,
+      "loss": 1.2073,
+      "step": 12347
+    },
+    {
+      "epoch": 1.97568,
+      "grad_norm": 1.9106816250020549,
+      "learning_rate": 5.5427074007257906e-05,
+      "loss": 1.2724,
+      "step": 12348
+    },
+    {
+      "epoch": 1.97584,
+      "grad_norm": 1.9636554310572125,
+      "learning_rate": 5.5411611689713206e-05,
+      "loss": 1.2579,
+      "step": 12349
+    },
+    {
+      "epoch": 1.976,
+      "grad_norm": 1.8639004372694545,
+      "learning_rate": 5.539615070262103e-05,
+      "loss": 1.2052,
+      "step": 12350
+    },
+    {
+      "epoch": 1.9761600000000001,
+      "grad_norm": 2.0285230677252537,
+      "learning_rate": 5.5380691046442734e-05,
+      "loss": 1.2271,
+      "step": 12351
+    },
+    {
+      "epoch": 1.9763199999999999,
+      "grad_norm": 1.7342560183230147,
+      "learning_rate": 5.53652327216396e-05,
+      "loss": 1.1591,
+      "step": 12352
+    },
+    {
+      "epoch": 1.97648,
+      "grad_norm": 2.1219340820338735,
+      "learning_rate": 5.534977572867287e-05,
+      "loss": 1.2524,
+      "step": 12353
+    },
+    {
+      "epoch": 1.97664,
+      "grad_norm": 2.0420396494878235,
+      "learning_rate": 5.533432006800379e-05,
+      "loss": 1.1789,
+      "step": 12354
+    },
+    {
+      "epoch": 1.9768,
+      "grad_norm": 1.8760657809065746,
+      "learning_rate": 5.531886574009349e-05,
+      "loss": 1.1485,
+      "step": 12355
+    },
+    {
+      "epoch": 1.97696,
+      "grad_norm": 2.238402887214194,
+      "learning_rate": 5.530341274540309e-05,
+      "loss": 1.2219,
+      "step": 12356
+    },
+    {
+      "epoch": 1.97712,
+      "grad_norm": 1.9890715590753525,
+      "learning_rate": 5.5287961084393826e-05,
+      "loss": 1.2259,
+      "step": 12357
+    },
+    {
+      "epoch": 1.97728,
+      "grad_norm": 2.255042036492094,
+      "learning_rate": 5.527251075752655e-05,
+      "loss": 1.231,
+      "step": 12358
+    },
+    {
+      "epoch": 1.97744,
+      "grad_norm": 2.30076742354035,
+      "learning_rate": 5.525706176526246e-05,
+      "loss": 1.3202,
+      "step": 12359
+    },
+    {
+      "epoch": 1.9776,
+      "grad_norm": 2.076134986679542,
+      "learning_rate": 5.524161410806239e-05,
+      "loss": 1.2861,
+      "step": 12360
+    },
+    {
+      "epoch": 1.97776,
+      "grad_norm": 1.8321245278936265,
+      "learning_rate": 5.522616778638736e-05,
+      "loss": 1.2526,
+      "step": 12361
+    },
+    {
+      "epoch": 1.9779200000000001,
+      "grad_norm": 2.451267528543566,
+      "learning_rate": 5.5210722800698236e-05,
+      "loss": 1.221,
+      "step": 12362
+    },
+    {
+      "epoch": 1.9780799999999998,
+      "grad_norm": 2.6075689833174587,
+      "learning_rate": 5.519527915145588e-05,
+      "loss": 1.1962,
+      "step": 12363
+    },
+    {
+      "epoch": 1.97824,
+      "grad_norm": 1.8986797346396418,
+      "learning_rate": 5.517983683912111e-05,
+      "loss": 1.1998,
+      "step": 12364
+    },
+    {
+      "epoch": 1.9784000000000002,
+      "grad_norm": 2.250598754978113,
+      "learning_rate": 5.516439586415472e-05,
+      "loss": 1.2406,
+      "step": 12365
+    },
+    {
+      "epoch": 1.9785599999999999,
+      "grad_norm": 1.9772352007104812,
+      "learning_rate": 5.514895622701739e-05,
+      "loss": 1.2965,
+      "step": 12366
+    },
+    {
+      "epoch": 1.97872,
+      "grad_norm": 1.911840727834075,
+      "learning_rate": 5.513351792816988e-05,
+      "loss": 1.2043,
+      "step": 12367
+    },
+    {
+      "epoch": 1.97888,
+      "grad_norm": 2.016153945679344,
+      "learning_rate": 5.5118080968072805e-05,
+      "loss": 1.2125,
+      "step": 12368
+    },
+    {
+      "epoch": 1.97904,
+      "grad_norm": 1.8703851885500435,
+      "learning_rate": 5.510264534718679e-05,
+      "loss": 1.1702,
+      "step": 12369
+    },
+    {
+      "epoch": 1.9792,
+      "grad_norm": 2.1921429081686394,
+      "learning_rate": 5.5087211065972374e-05,
+      "loss": 1.1778,
+      "step": 12370
+    },
+    {
+      "epoch": 1.97936,
+      "grad_norm": 1.9845187771516875,
+      "learning_rate": 5.5071778124890214e-05,
+      "loss": 1.2849,
+      "step": 12371
+    },
+    {
+      "epoch": 1.97952,
+      "grad_norm": 1.9369855707505774,
+      "learning_rate": 5.505634652440066e-05,
+      "loss": 1.2282,
+      "step": 12372
+    },
+    {
+      "epoch": 1.97968,
+      "grad_norm": 1.9189832058470262,
+      "learning_rate": 5.504091626496431e-05,
+      "loss": 1.2409,
+      "step": 12373
+    },
+    {
+      "epoch": 1.97984,
+      "grad_norm": 1.7699607912544186,
+      "learning_rate": 5.502548734704143e-05,
+      "loss": 1.1753,
+      "step": 12374
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 1.948340819740486,
+      "learning_rate": 5.5010059771092545e-05,
+      "loss": 1.2512,
+      "step": 12375
+    },
+    {
+      "epoch": 1.9801600000000001,
+      "grad_norm": 1.8629167627338246,
+      "learning_rate": 5.4994633537577836e-05,
+      "loss": 1.1753,
+      "step": 12376
+    },
+    {
+      "epoch": 1.9803199999999999,
+      "grad_norm": 1.8950507644177597,
+      "learning_rate": 5.497920864695773e-05,
+      "loss": 1.2818,
+      "step": 12377
+    },
+    {
+      "epoch": 1.98048,
+      "grad_norm": 1.926429893651703,
+      "learning_rate": 5.496378509969245e-05,
+      "loss": 1.1979,
+      "step": 12378
+    },
+    {
+      "epoch": 1.98064,
+      "grad_norm": 1.9132649218451325,
+      "learning_rate": 5.4948362896242167e-05,
+      "loss": 1.2334,
+      "step": 12379
+    },
+    {
+      "epoch": 1.9808,
+      "grad_norm": 2.2537670979123408,
+      "learning_rate": 5.4932942037067116e-05,
+      "loss": 1.2303,
+      "step": 12380
+    },
+    {
+      "epoch": 1.98096,
+      "grad_norm": 1.8996251557887451,
+      "learning_rate": 5.491752252262738e-05,
+      "loss": 1.2518,
+      "step": 12381
+    },
+    {
+      "epoch": 1.98112,
+      "grad_norm": 2.080854503154333,
+      "learning_rate": 5.4902104353383096e-05,
+      "loss": 1.2006,
+      "step": 12382
+    },
+    {
+      "epoch": 1.98128,
+      "grad_norm": 2.028699588963817,
+      "learning_rate": 5.48866875297943e-05,
+      "loss": 1.2137,
+      "step": 12383
+    },
+    {
+      "epoch": 1.98144,
+      "grad_norm": 1.8719223096691224,
+      "learning_rate": 5.487127205232101e-05,
+      "loss": 1.2213,
+      "step": 12384
+    },
+    {
+      "epoch": 1.9816,
+      "grad_norm": 2.127719995003123,
+      "learning_rate": 5.48558579214232e-05,
+      "loss": 1.1987,
+      "step": 12385
+    },
+    {
+      "epoch": 1.98176,
+      "grad_norm": 2.105975097674776,
+      "learning_rate": 5.484044513756077e-05,
+      "loss": 1.236,
+      "step": 12386
+    },
+    {
+      "epoch": 1.9819200000000001,
+      "grad_norm": 1.6097742667817552,
+      "learning_rate": 5.482503370119374e-05,
+      "loss": 1.1727,
+      "step": 12387
+    },
+    {
+      "epoch": 1.9820799999999998,
+      "grad_norm": 1.9119003965829855,
+      "learning_rate": 5.480962361278179e-05,
+      "loss": 1.3108,
+      "step": 12388
+    },
+    {
+      "epoch": 1.98224,
+      "grad_norm": 2.139943320667753,
+      "learning_rate": 5.4794214872784913e-05,
+      "loss": 1.2334,
+      "step": 12389
+    },
+    {
+      "epoch": 1.9824000000000002,
+      "grad_norm": 1.6547484925326208,
+      "learning_rate": 5.477880748166269e-05,
+      "loss": 1.2383,
+      "step": 12390
+    },
+    {
+      "epoch": 1.9825599999999999,
+      "grad_norm": 2.0529724819872914,
+      "learning_rate": 5.476340143987506e-05,
+      "loss": 1.1858,
+      "step": 12391
+    },
+    {
+      "epoch": 1.98272,
+      "grad_norm": 1.8016142999038098,
+      "learning_rate": 5.4747996747881514e-05,
+      "loss": 1.1809,
+      "step": 12392
+    },
+    {
+      "epoch": 1.98288,
+      "grad_norm": 1.9700206248875256,
+      "learning_rate": 5.473259340614186e-05,
+      "loss": 1.2489,
+      "step": 12393
+    },
+    {
+      "epoch": 1.98304,
+      "grad_norm": 1.9655033580815764,
+      "learning_rate": 5.4717191415115656e-05,
+      "loss": 1.2273,
+      "step": 12394
+    },
+    {
+      "epoch": 1.9832,
+      "grad_norm": 2.146771080145431,
+      "learning_rate": 5.4701790775262475e-05,
+      "loss": 1.239,
+      "step": 12395
+    },
+    {
+      "epoch": 1.98336,
+      "grad_norm": 2.2525565139359225,
+      "learning_rate": 5.4686391487041845e-05,
+      "loss": 1.208,
+      "step": 12396
+    },
+    {
+      "epoch": 1.98352,
+      "grad_norm": 1.9610103733684114,
+      "learning_rate": 5.467099355091323e-05,
+      "loss": 1.2295,
+      "step": 12397
+    },
+    {
+      "epoch": 1.98368,
+      "grad_norm": 1.8573411752485227,
+      "learning_rate": 5.46555969673362e-05,
+      "loss": 1.2123,
+      "step": 12398
+    },
+    {
+      "epoch": 1.98384,
+      "grad_norm": 2.0406407587237854,
+      "learning_rate": 5.4640201736769995e-05,
+      "loss": 1.2328,
+      "step": 12399
+    },
+    {
+      "epoch": 1.984,
+      "grad_norm": 1.9767925204867167,
+      "learning_rate": 5.462480785967415e-05,
+      "loss": 1.1509,
+      "step": 12400
+    },
+    {
+      "epoch": 1.9841600000000001,
+      "grad_norm": 1.98878152912539,
+      "learning_rate": 5.4609415336507854e-05,
+      "loss": 1.1978,
+      "step": 12401
+    },
+    {
+      "epoch": 1.9843199999999999,
+      "grad_norm": 2.2415152524228454,
+      "learning_rate": 5.459402416773049e-05,
+      "loss": 1.2867,
+      "step": 12402
+    },
+    {
+      "epoch": 1.98448,
+      "grad_norm": 1.9534222477831136,
+      "learning_rate": 5.4578634353801284e-05,
+      "loss": 1.2297,
+      "step": 12403
+    },
+    {
+      "epoch": 1.98464,
+      "grad_norm": 1.7777265945870626,
+      "learning_rate": 5.456324589517945e-05,
+      "loss": 1.2648,
+      "step": 12404
+    },
+    {
+      "epoch": 1.9848,
+      "grad_norm": 1.8501860695736994,
+      "learning_rate": 5.4547858792324145e-05,
+      "loss": 1.2156,
+      "step": 12405
+    },
+    {
+      "epoch": 1.98496,
+      "grad_norm": 2.170575335937072,
+      "learning_rate": 5.453247304569451e-05,
+      "loss": 1.2611,
+      "step": 12406
+    },
+    {
+      "epoch": 1.98512,
+      "grad_norm": 1.9937818581297284,
+      "learning_rate": 5.451708865574962e-05,
+      "loss": 1.1843,
+      "step": 12407
+    },
+    {
+      "epoch": 1.98528,
+      "grad_norm": 1.9582293171772838,
+      "learning_rate": 5.450170562294852e-05,
+      "loss": 1.227,
+      "step": 12408
+    },
+    {
+      "epoch": 1.98544,
+      "grad_norm": 1.9764386680898638,
+      "learning_rate": 5.448632394775024e-05,
+      "loss": 1.2205,
+      "step": 12409
+    },
+    {
+      "epoch": 1.9856,
+      "grad_norm": 1.7222279194000623,
+      "learning_rate": 5.447094363061374e-05,
+      "loss": 1.1397,
+      "step": 12410
+    },
+    {
+      "epoch": 1.98576,
+      "grad_norm": 2.162821235220171,
+      "learning_rate": 5.4455564671997914e-05,
+      "loss": 1.2061,
+      "step": 12411
+    },
+    {
+      "epoch": 1.9859200000000001,
+      "grad_norm": 1.8691936655613175,
+      "learning_rate": 5.444018707236169e-05,
+      "loss": 1.196,
+      "step": 12412
+    },
+    {
+      "epoch": 1.9860799999999998,
+      "grad_norm": 1.775341009882368,
+      "learning_rate": 5.4424810832163855e-05,
+      "loss": 1.2177,
+      "step": 12413
+    },
+    {
+      "epoch": 1.98624,
+      "grad_norm": 1.9962224607101189,
+      "learning_rate": 5.440943595186332e-05,
+      "loss": 1.2317,
+      "step": 12414
+    },
+    {
+      "epoch": 1.9864000000000002,
+      "grad_norm": 2.179425921337779,
+      "learning_rate": 5.4394062431918726e-05,
+      "loss": 1.2102,
+      "step": 12415
+    },
+    {
+      "epoch": 1.9865599999999999,
+      "grad_norm": 2.0852306267272804,
+      "learning_rate": 5.437869027278893e-05,
+      "loss": 1.2406,
+      "step": 12416
+    },
+    {
+      "epoch": 1.98672,
+      "grad_norm": 1.9652311510056961,
+      "learning_rate": 5.436331947493245e-05,
+      "loss": 1.2295,
+      "step": 12417
+    },
+    {
+      "epoch": 1.98688,
+      "grad_norm": 2.1760968524715274,
+      "learning_rate": 5.434795003880806e-05,
+      "loss": 1.22,
+      "step": 12418
+    },
+    {
+      "epoch": 1.98704,
+      "grad_norm": 2.0873910862217557,
+      "learning_rate": 5.4332581964874316e-05,
+      "loss": 1.2044,
+      "step": 12419
+    },
+    {
+      "epoch": 1.9872,
+      "grad_norm": 2.091662620022506,
+      "learning_rate": 5.431721525358978e-05,
+      "loss": 1.259,
+      "step": 12420
+    },
+    {
+      "epoch": 1.98736,
+      "grad_norm": 2.044369381776714,
+      "learning_rate": 5.430184990541298e-05,
+      "loss": 1.2385,
+      "step": 12421
+    },
+    {
+      "epoch": 1.98752,
+      "grad_norm": 1.8306795333241455,
+      "learning_rate": 5.428648592080239e-05,
+      "loss": 1.1505,
+      "step": 12422
+    },
+    {
+      "epoch": 1.9876800000000001,
+      "grad_norm": 1.8756608067800928,
+      "learning_rate": 5.427112330021644e-05,
+      "loss": 1.1826,
+      "step": 12423
+    },
+    {
+      "epoch": 1.98784,
+      "grad_norm": 1.966589976143976,
+      "learning_rate": 5.4255762044113556e-05,
+      "loss": 1.2553,
+      "step": 12424
+    },
+    {
+      "epoch": 1.988,
+      "grad_norm": 1.9537708640483473,
+      "learning_rate": 5.4240402152952055e-05,
+      "loss": 1.2029,
+      "step": 12425
+    },
+    {
+      "epoch": 1.9881600000000001,
+      "grad_norm": 1.9234624138835255,
+      "learning_rate": 5.4225043627190283e-05,
+      "loss": 1.2225,
+      "step": 12426
+    },
+    {
+      "epoch": 1.9883199999999999,
+      "grad_norm": 1.6683774837543497,
+      "learning_rate": 5.42096864672865e-05,
+      "loss": 1.2021,
+      "step": 12427
+    },
+    {
+      "epoch": 1.98848,
+      "grad_norm": 2.0645694542702775,
+      "learning_rate": 5.419433067369897e-05,
+      "loss": 1.1482,
+      "step": 12428
+    },
+    {
+      "epoch": 1.98864,
+      "grad_norm": 1.7354208850572155,
+      "learning_rate": 5.4178976246885816e-05,
+      "loss": 1.2609,
+      "step": 12429
+    },
+    {
+      "epoch": 1.9888,
+      "grad_norm": 1.9169614667375392,
+      "learning_rate": 5.4163623187305325e-05,
+      "loss": 1.2175,
+      "step": 12430
+    },
+    {
+      "epoch": 1.98896,
+      "grad_norm": 2.1277011357168334,
+      "learning_rate": 5.4148271495415434e-05,
+      "loss": 1.2219,
+      "step": 12431
+    },
+    {
+      "epoch": 1.98912,
+      "grad_norm": 1.98725547146719,
+      "learning_rate": 5.41329211716744e-05,
+      "loss": 1.2303,
+      "step": 12432
+    },
+    {
+      "epoch": 1.98928,
+      "grad_norm": 2.0729719240745523,
+      "learning_rate": 5.411757221654008e-05,
+      "loss": 1.2543,
+      "step": 12433
+    },
+    {
+      "epoch": 1.98944,
+      "grad_norm": 1.8160439871798122,
+      "learning_rate": 5.410222463047059e-05,
+      "loss": 1.2312,
+      "step": 12434
+    },
+    {
+      "epoch": 1.9896,
+      "grad_norm": 2.1989077461219018,
+      "learning_rate": 5.408687841392384e-05,
+      "loss": 1.2112,
+      "step": 12435
+    },
+    {
+      "epoch": 1.98976,
+      "grad_norm": 2.0960389736232075,
+      "learning_rate": 5.407153356735772e-05,
+      "loss": 1.1939,
+      "step": 12436
+    },
+    {
+      "epoch": 1.9899200000000001,
+      "grad_norm": 2.013532758069967,
+      "learning_rate": 5.405619009123013e-05,
+      "loss": 1.2498,
+      "step": 12437
+    },
+    {
+      "epoch": 1.9900799999999998,
+      "grad_norm": 1.8553634333131765,
+      "learning_rate": 5.404084798599887e-05,
+      "loss": 1.3032,
+      "step": 12438
+    },
+    {
+      "epoch": 1.99024,
+      "grad_norm": 2.2857909226619713,
+      "learning_rate": 5.402550725212174e-05,
+      "loss": 1.257,
+      "step": 12439
+    },
+    {
+      "epoch": 1.9904,
+      "grad_norm": 2.2123672883968384,
+      "learning_rate": 5.4010167890056485e-05,
+      "loss": 1.2054,
+      "step": 12440
+    },
+    {
+      "epoch": 1.9905599999999999,
+      "grad_norm": 1.9875129431911072,
+      "learning_rate": 5.3994829900260816e-05,
+      "loss": 1.2627,
+      "step": 12441
+    },
+    {
+      "epoch": 1.99072,
+      "grad_norm": 2.0250974854301,
+      "learning_rate": 5.397949328319236e-05,
+      "loss": 1.2574,
+      "step": 12442
+    },
+    {
+      "epoch": 1.99088,
+      "grad_norm": 2.0577294946249003,
+      "learning_rate": 5.3964158039308786e-05,
+      "loss": 1.2173,
+      "step": 12443
+    },
+    {
+      "epoch": 1.99104,
+      "grad_norm": 1.977495757146848,
+      "learning_rate": 5.394882416906765e-05,
+      "loss": 1.2476,
+      "step": 12444
+    },
+    {
+      "epoch": 1.9912,
+      "grad_norm": 1.7785881486075485,
+      "learning_rate": 5.393349167292645e-05,
+      "loss": 1.2481,
+      "step": 12445
+    },
+    {
+      "epoch": 1.99136,
+      "grad_norm": 1.829305896899639,
+      "learning_rate": 5.3918160551342824e-05,
+      "loss": 1.1682,
+      "step": 12446
+    },
+    {
+      "epoch": 1.99152,
+      "grad_norm": 2.1117801403193766,
+      "learning_rate": 5.390283080477405e-05,
+      "loss": 1.2479,
+      "step": 12447
+    },
+    {
+      "epoch": 1.9916800000000001,
+      "grad_norm": 1.9653360325020286,
+      "learning_rate": 5.388750243367772e-05,
+      "loss": 1.1873,
+      "step": 12448
+    },
+    {
+      "epoch": 1.99184,
+      "grad_norm": 1.838917803745115,
+      "learning_rate": 5.387217543851104e-05,
+      "loss": 1.2835,
+      "step": 12449
+    },
+    {
+      "epoch": 1.992,
+      "grad_norm": 2.1699640793925226,
+      "learning_rate": 5.385684981973147e-05,
+      "loss": 1.2052,
+      "step": 12450
+    },
+    {
+      "epoch": 1.9921600000000002,
+      "grad_norm": 1.841661449966954,
+      "learning_rate": 5.384152557779626e-05,
+      "loss": 1.2638,
+      "step": 12451
+    },
+    {
+      "epoch": 1.9923199999999999,
+      "grad_norm": 2.4064696110450203,
+      "learning_rate": 5.382620271316268e-05,
+      "loss": 1.226,
+      "step": 12452
+    },
+    {
+      "epoch": 1.99248,
+      "grad_norm": 2.44562381140598,
+      "learning_rate": 5.381088122628791e-05,
+      "loss": 1.37,
+      "step": 12453
+    },
+    {
+      "epoch": 1.99264,
+      "grad_norm": 1.8723199992244466,
+      "learning_rate": 5.379556111762911e-05,
+      "loss": 1.2127,
+      "step": 12454
+    },
+    {
+      "epoch": 1.9928,
+      "grad_norm": 2.039612460714932,
+      "learning_rate": 5.378024238764352e-05,
+      "loss": 1.266,
+      "step": 12455
+    },
+    {
+      "epoch": 1.99296,
+      "grad_norm": 1.8682788041334653,
+      "learning_rate": 5.376492503678807e-05,
+      "loss": 1.2147,
+      "step": 12456
+    },
+    {
+      "epoch": 1.99312,
+      "grad_norm": 2.120132790900788,
+      "learning_rate": 5.374960906551996e-05,
+      "loss": 1.2468,
+      "step": 12457
+    },
+    {
+      "epoch": 1.99328,
+      "grad_norm": 2.1380658613716874,
+      "learning_rate": 5.3734294474296046e-05,
+      "loss": 1.262,
+      "step": 12458
+    },
+    {
+      "epoch": 1.99344,
+      "grad_norm": 2.156744368991557,
+      "learning_rate": 5.37189812635734e-05,
+      "loss": 1.2889,
+      "step": 12459
+    },
+    {
+      "epoch": 1.9936,
+      "grad_norm": 1.9053162374198795,
+      "learning_rate": 5.370366943380892e-05,
+      "loss": 1.2357,
+      "step": 12460
+    },
+    {
+      "epoch": 1.99376,
+      "grad_norm": 2.1047309671847434,
+      "learning_rate": 5.368835898545949e-05,
+      "loss": 1.2401,
+      "step": 12461
+    },
+    {
+      "epoch": 1.9939200000000001,
+      "grad_norm": 1.9101953460035486,
+      "learning_rate": 5.367304991898193e-05,
+      "loss": 1.2347,
+      "step": 12462
+    },
+    {
+      "epoch": 1.9940799999999999,
+      "grad_norm": 1.9309689067922045,
+      "learning_rate": 5.365774223483306e-05,
+      "loss": 1.2341,
+      "step": 12463
+    },
+    {
+      "epoch": 1.99424,
+      "grad_norm": 1.7113184201305462,
+      "learning_rate": 5.3642435933469646e-05,
+      "loss": 1.2013,
+      "step": 12464
+    },
+    {
+      "epoch": 1.9944,
+      "grad_norm": 1.8139602044410088,
+      "learning_rate": 5.3627131015348384e-05,
+      "loss": 1.1729,
+      "step": 12465
+    },
+    {
+      "epoch": 1.9945599999999999,
+      "grad_norm": 2.134548739292335,
+      "learning_rate": 5.3611827480925956e-05,
+      "loss": 1.2159,
+      "step": 12466
+    },
+    {
+      "epoch": 1.99472,
+      "grad_norm": 1.9120909444959506,
+      "learning_rate": 5.3596525330659005e-05,
+      "loss": 1.1836,
+      "step": 12467
+    },
+    {
+      "epoch": 1.99488,
+      "grad_norm": 1.8608717230469916,
+      "learning_rate": 5.358122456500413e-05,
+      "loss": 1.1867,
+      "step": 12468
+    },
+    {
+      "epoch": 1.99504,
+      "grad_norm": 1.8060582357722335,
+      "learning_rate": 5.3565925184417867e-05,
+      "loss": 1.1327,
+      "step": 12469
+    },
+    {
+      "epoch": 1.9952,
+      "grad_norm": 1.9477316726641811,
+      "learning_rate": 5.35506271893567e-05,
+      "loss": 1.2524,
+      "step": 12470
+    },
+    {
+      "epoch": 1.99536,
+      "grad_norm": 1.9951923299421563,
+      "learning_rate": 5.35353305802772e-05,
+      "loss": 1.1989,
+      "step": 12471
+    },
+    {
+      "epoch": 1.99552,
+      "grad_norm": 1.8373417845517162,
+      "learning_rate": 5.352003535763566e-05,
+      "loss": 1.1986,
+      "step": 12472
+    },
+    {
+      "epoch": 1.9956800000000001,
+      "grad_norm": 1.987019799091858,
+      "learning_rate": 5.350474152188861e-05,
+      "loss": 1.208,
+      "step": 12473
+    },
+    {
+      "epoch": 1.9958399999999998,
+      "grad_norm": 1.7636588805207492,
+      "learning_rate": 5.348944907349225e-05,
+      "loss": 1.1803,
+      "step": 12474
+    },
+    {
+      "epoch": 1.996,
+      "grad_norm": 1.9471734503407319,
+      "learning_rate": 5.3474158012902984e-05,
+      "loss": 1.2979,
+      "step": 12475
+    },
+    {
+      "epoch": 1.9961600000000002,
+      "grad_norm": 2.24152200413729,
+      "learning_rate": 5.345886834057707e-05,
+      "loss": 1.1835,
+      "step": 12476
+    },
+    {
+      "epoch": 1.9963199999999999,
+      "grad_norm": 2.291461800287325,
+      "learning_rate": 5.34435800569707e-05,
+      "loss": 1.2392,
+      "step": 12477
+    },
+    {
+      "epoch": 1.99648,
+      "grad_norm": 1.9457899433304413,
+      "learning_rate": 5.3428293162540045e-05,
+      "loss": 1.2462,
+      "step": 12478
+    },
+    {
+      "epoch": 1.99664,
+      "grad_norm": 1.9798073587254366,
+      "learning_rate": 5.341300765774127e-05,
+      "loss": 1.2665,
+      "step": 12479
+    },
+    {
+      "epoch": 1.9968,
+      "grad_norm": 1.9738604798068937,
+      "learning_rate": 5.3397723543030466e-05,
+      "loss": 1.2725,
+      "step": 12480
+    },
+    {
+      "epoch": 1.99696,
+      "grad_norm": 2.0864771242308575,
+      "learning_rate": 5.3382440818863676e-05,
+      "loss": 1.1944,
+      "step": 12481
+    },
+    {
+      "epoch": 1.99712,
+      "grad_norm": 1.8692168506511102,
+      "learning_rate": 5.3367159485696926e-05,
+      "loss": 1.1999,
+      "step": 12482
+    },
+    {
+      "epoch": 1.99728,
+      "grad_norm": 1.5999064983651632,
+      "learning_rate": 5.335187954398619e-05,
+      "loss": 1.2217,
+      "step": 12483
+    },
+    {
+      "epoch": 1.99744,
+      "grad_norm": 2.126056737764069,
+      "learning_rate": 5.3336600994187405e-05,
+      "loss": 1.1568,
+      "step": 12484
+    },
+    {
+      "epoch": 1.9976,
+      "grad_norm": 1.9596882579677342,
+      "learning_rate": 5.332132383675643e-05,
+      "loss": 1.2412,
+      "step": 12485
+    },
+    {
+      "epoch": 1.99776,
+      "grad_norm": 2.046019191851334,
+      "learning_rate": 5.330604807214909e-05,
+      "loss": 1.2157,
+      "step": 12486
+    },
+    {
+      "epoch": 1.9979200000000001,
+      "grad_norm": 2.143472072994583,
+      "learning_rate": 5.329077370082134e-05,
+      "loss": 1.3074,
+      "step": 12487
+    },
+    {
+      "epoch": 1.9980799999999999,
+      "grad_norm": 1.8376740170365622,
+      "learning_rate": 5.327550072322874e-05,
+      "loss": 1.2312,
+      "step": 12488
+    },
+    {
+      "epoch": 1.99824,
+      "grad_norm": 2.2870310519404087,
+      "learning_rate": 5.32602291398272e-05,
+      "loss": 1.2355,
+      "step": 12489
+    },
+    {
+      "epoch": 1.9984,
+      "grad_norm": 2.0057020058830366,
+      "learning_rate": 5.3244958951072243e-05,
+      "loss": 1.2001,
+      "step": 12490
+    },
+    {
+      "epoch": 1.99856,
+      "grad_norm": 1.8590072376894264,
+      "learning_rate": 5.322969015741961e-05,
+      "loss": 1.1869,
+      "step": 12491
+    },
+    {
+      "epoch": 1.99872,
+      "grad_norm": 1.7917174360021162,
+      "learning_rate": 5.321442275932488e-05,
+      "loss": 1.1618,
+      "step": 12492
+    },
+    {
+      "epoch": 1.99888,
+      "grad_norm": 1.6985293915879474,
+      "learning_rate": 5.3199156757243596e-05,
+      "loss": 1.1777,
+      "step": 12493
+    },
+    {
+      "epoch": 1.99904,
+      "grad_norm": 2.0267511524089583,
+      "learning_rate": 5.318389215163129e-05,
+      "loss": 1.219,
+      "step": 12494
+    },
+    {
+      "epoch": 1.9992,
+      "grad_norm": 1.9973224236802936,
+      "learning_rate": 5.316862894294341e-05,
+      "loss": 1.2831,
+      "step": 12495
+    },
+    {
+      "epoch": 1.99936,
+      "grad_norm": 2.1250018208375727,
+      "learning_rate": 5.31533671316354e-05,
+      "loss": 1.2278,
+      "step": 12496
+    },
+    {
+      "epoch": 1.99952,
+      "grad_norm": 1.9796441736123223,
+      "learning_rate": 5.313810671816266e-05,
+      "loss": 1.2401,
+      "step": 12497
+    },
+    {
+      "epoch": 1.9996800000000001,
+      "grad_norm": 2.1405184911918766,
+      "learning_rate": 5.312284770298053e-05,
+      "loss": 1.259,
+      "step": 12498
+    },
+    {
+      "epoch": 1.9998399999999998,
+      "grad_norm": 1.8822540372475811,
+      "learning_rate": 5.310759008654431e-05,
+      "loss": 1.2582,
+      "step": 12499
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.740690399932073,
+      "learning_rate": 5.3092333869309274e-05,
+      "loss": 1.2149,
+      "step": 12500
+    },
+    {
+      "epoch": 2.00016,
+      "grad_norm": 1.8805809881963473,
+      "learning_rate": 5.307707905173065e-05,
+      "loss": 1.1801,
+      "step": 12501
+    },
+    {
+      "epoch": 2.00032,
+      "grad_norm": 1.9306280126893833,
+      "learning_rate": 5.306182563426357e-05,
+      "loss": 1.1807,
+      "step": 12502
+    },
+    {
+      "epoch": 2.00048,
+      "grad_norm": 1.9498803941009422,
+      "learning_rate": 5.304657361736329e-05,
+      "loss": 1.1656,
+      "step": 12503
+    },
+    {
+      "epoch": 2.00064,
+      "grad_norm": 1.7280021785750002,
+      "learning_rate": 5.3031323001484765e-05,
+      "loss": 1.125,
+      "step": 12504
+    },
+    {
+      "epoch": 2.0008,
+      "grad_norm": 1.7775361420131077,
+      "learning_rate": 5.3016073787083196e-05,
+      "loss": 1.1364,
+      "step": 12505
+    },
+    {
+      "epoch": 2.00096,
+      "grad_norm": 1.842896839628593,
+      "learning_rate": 5.3000825974613454e-05,
+      "loss": 1.1052,
+      "step": 12506
+    },
+    {
+      "epoch": 2.00112,
+      "grad_norm": 1.6956339320897535,
+      "learning_rate": 5.2985579564530606e-05,
+      "loss": 1.1275,
+      "step": 12507
+    },
+    {
+      "epoch": 2.00128,
+      "grad_norm": 1.821870292801444,
+      "learning_rate": 5.297033455728957e-05,
+      "loss": 1.1908,
+      "step": 12508
+    },
+    {
+      "epoch": 2.00144,
+      "grad_norm": 2.1667056368317414,
+      "learning_rate": 5.295509095334521e-05,
+      "loss": 1.1241,
+      "step": 12509
+    },
+    {
+      "epoch": 2.0016,
+      "grad_norm": 1.7945672940623445,
+      "learning_rate": 5.2939848753152385e-05,
+      "loss": 1.1703,
+      "step": 12510
+    },
+    {
+      "epoch": 2.00176,
+      "grad_norm": 1.746657554542582,
+      "learning_rate": 5.292460795716586e-05,
+      "loss": 1.1475,
+      "step": 12511
+    },
+    {
+      "epoch": 2.00192,
+      "grad_norm": 1.8763154226322942,
+      "learning_rate": 5.2909368565840524e-05,
+      "loss": 1.1811,
+      "step": 12512
+    },
+    {
+      "epoch": 2.00208,
+      "grad_norm": 1.9389470450699453,
+      "learning_rate": 5.289413057963092e-05,
+      "loss": 1.094,
+      "step": 12513
+    },
+    {
+      "epoch": 2.00224,
+      "grad_norm": 2.1096477667875337,
+      "learning_rate": 5.2878893998991905e-05,
+      "loss": 1.1248,
+      "step": 12514
+    },
+    {
+      "epoch": 2.0024,
+      "grad_norm": 1.7984623165485534,
+      "learning_rate": 5.286365882437795e-05,
+      "loss": 1.1321,
+      "step": 12515
+    },
+    {
+      "epoch": 2.00256,
+      "grad_norm": 2.122068429192626,
+      "learning_rate": 5.284842505624381e-05,
+      "loss": 1.1689,
+      "step": 12516
+    },
+    {
+      "epoch": 2.00272,
+      "grad_norm": 2.290431191930499,
+      "learning_rate": 5.2833192695043875e-05,
+      "loss": 1.1491,
+      "step": 12517
+    },
+    {
+      "epoch": 2.00288,
+      "grad_norm": 1.75616387928476,
+      "learning_rate": 5.2817961741232794e-05,
+      "loss": 1.1331,
+      "step": 12518
+    },
+    {
+      "epoch": 2.00304,
+      "grad_norm": 2.5745008225493105,
+      "learning_rate": 5.280273219526497e-05,
+      "loss": 1.1446,
+      "step": 12519
+    },
+    {
+      "epoch": 2.0032,
+      "grad_norm": 2.214592369833346,
+      "learning_rate": 5.278750405759485e-05,
+      "loss": 1.1325,
+      "step": 12520
+    },
+    {
+      "epoch": 2.00336,
+      "grad_norm": 1.7944387032150653,
+      "learning_rate": 5.2772277328676824e-05,
+      "loss": 1.185,
+      "step": 12521
+    },
+    {
+      "epoch": 2.00352,
+      "grad_norm": 1.8209616902013306,
+      "learning_rate": 5.27570520089652e-05,
+      "loss": 1.1081,
+      "step": 12522
+    },
+    {
+      "epoch": 2.00368,
+      "grad_norm": 1.9553302036243465,
+      "learning_rate": 5.2741828098914326e-05,
+      "loss": 1.1409,
+      "step": 12523
+    },
+    {
+      "epoch": 2.00384,
+      "grad_norm": 2.259761536076456,
+      "learning_rate": 5.272660559897842e-05,
+      "loss": 1.1455,
+      "step": 12524
+    },
+    {
+      "epoch": 2.004,
+      "grad_norm": 1.627877843594939,
+      "learning_rate": 5.271138450961173e-05,
+      "loss": 1.1067,
+      "step": 12525
+    },
+    {
+      "epoch": 2.00416,
+      "grad_norm": 1.77958783642418,
+      "learning_rate": 5.2696164831268424e-05,
+      "loss": 1.1576,
+      "step": 12526
+    },
+    {
+      "epoch": 2.00432,
+      "grad_norm": 1.8502897363849624,
+      "learning_rate": 5.268094656440258e-05,
+      "loss": 1.1528,
+      "step": 12527
+    },
+    {
+      "epoch": 2.00448,
+      "grad_norm": 2.570180757784531,
+      "learning_rate": 5.266572970946841e-05,
+      "loss": 1.1191,
+      "step": 12528
+    },
+    {
+      "epoch": 2.00464,
+      "grad_norm": 2.048301083583937,
+      "learning_rate": 5.265051426691984e-05,
+      "loss": 1.1317,
+      "step": 12529
+    },
+    {
+      "epoch": 2.0048,
+      "grad_norm": 1.6822933407147977,
+      "learning_rate": 5.2635300237210995e-05,
+      "loss": 1.1466,
+      "step": 12530
+    },
+    {
+      "epoch": 2.00496,
+      "grad_norm": 3.0206866688823797,
+      "learning_rate": 5.262008762079569e-05,
+      "loss": 1.1456,
+      "step": 12531
+    },
+    {
+      "epoch": 2.00512,
+      "grad_norm": 2.241198545809431,
+      "learning_rate": 5.260487641812801e-05,
+      "loss": 1.1212,
+      "step": 12532
+    },
+    {
+      "epoch": 2.00528,
+      "grad_norm": 1.768569981089679,
+      "learning_rate": 5.258966662966167e-05,
+      "loss": 1.1618,
+      "step": 12533
+    },
+    {
+      "epoch": 2.00544,
+      "grad_norm": 1.9233488023723129,
+      "learning_rate": 5.2574458255850653e-05,
+      "loss": 1.16,
+      "step": 12534
+    },
+    {
+      "epoch": 2.0056,
+      "grad_norm": 2.124924061805262,
+      "learning_rate": 5.255925129714868e-05,
+      "loss": 1.1338,
+      "step": 12535
+    },
+    {
+      "epoch": 2.00576,
+      "grad_norm": 2.1563510513145046,
+      "learning_rate": 5.254404575400953e-05,
+      "loss": 1.1276,
+      "step": 12536
+    },
+    {
+      "epoch": 2.00592,
+      "grad_norm": 1.7078210990686737,
+      "learning_rate": 5.252884162688689e-05,
+      "loss": 1.12,
+      "step": 12537
+    },
+    {
+      "epoch": 2.00608,
+      "grad_norm": 2.496404391081529,
+      "learning_rate": 5.251363891623445e-05,
+      "loss": 1.1383,
+      "step": 12538
+    },
+    {
+      "epoch": 2.00624,
+      "grad_norm": 2.219138319235716,
+      "learning_rate": 5.249843762250584e-05,
+      "loss": 1.1267,
+      "step": 12539
+    },
+    {
+      "epoch": 2.0064,
+      "grad_norm": 1.8521663590249349,
+      "learning_rate": 5.2483237746154624e-05,
+      "loss": 1.1741,
+      "step": 12540
+    },
+    {
+      "epoch": 2.00656,
+      "grad_norm": 2.0649943263076787,
+      "learning_rate": 5.246803928763435e-05,
+      "loss": 1.1682,
+      "step": 12541
+    },
+    {
+      "epoch": 2.00672,
+      "grad_norm": 2.614986240142889,
+      "learning_rate": 5.2452842247398526e-05,
+      "loss": 1.1109,
+      "step": 12542
+    },
+    {
+      "epoch": 2.00688,
+      "grad_norm": 2.7089797013171295,
+      "learning_rate": 5.243764662590057e-05,
+      "loss": 1.1201,
+      "step": 12543
+    },
+    {
+      "epoch": 2.00704,
+      "grad_norm": 1.6992414670152245,
+      "learning_rate": 5.2422452423594005e-05,
+      "loss": 1.151,
+      "step": 12544
+    },
+    {
+      "epoch": 2.0072,
+      "grad_norm": 2.5272863421753926,
+      "learning_rate": 5.2407259640932074e-05,
+      "loss": 1.1399,
+      "step": 12545
+    },
+    {
+      "epoch": 2.00736,
+      "grad_norm": 2.392670078116263,
+      "learning_rate": 5.2392068278368225e-05,
+      "loss": 1.1472,
+      "step": 12546
+    },
+    {
+      "epoch": 2.00752,
+      "grad_norm": 1.847285499351649,
+      "learning_rate": 5.237687833635561e-05,
+      "loss": 1.0796,
+      "step": 12547
+    },
+    {
+      "epoch": 2.00768,
+      "grad_norm": 2.6827508800448556,
+      "learning_rate": 5.236168981534763e-05,
+      "loss": 1.2063,
+      "step": 12548
+    },
+    {
+      "epoch": 2.00784,
+      "grad_norm": 1.8453557422524276,
+      "learning_rate": 5.234650271579733e-05,
+      "loss": 1.126,
+      "step": 12549
+    },
+    {
+      "epoch": 2.008,
+      "grad_norm": 2.0339551164145013,
+      "learning_rate": 5.233131703815799e-05,
+      "loss": 1.1197,
+      "step": 12550
+    },
+    {
+      "epoch": 2.00816,
+      "grad_norm": 1.7767169121953978,
+      "learning_rate": 5.231613278288269e-05,
+      "loss": 1.1599,
+      "step": 12551
+    },
+    {
+      "epoch": 2.00832,
+      "grad_norm": 1.9057363846604665,
+      "learning_rate": 5.2300949950424506e-05,
+      "loss": 1.113,
+      "step": 12552
+    },
+    {
+      "epoch": 2.00848,
+      "grad_norm": 1.8362830357794107,
+      "learning_rate": 5.228576854123647e-05,
+      "loss": 1.1473,
+      "step": 12553
+    },
+    {
+      "epoch": 2.00864,
+      "grad_norm": 2.4134644401013086,
+      "learning_rate": 5.227058855577157e-05,
+      "loss": 1.1766,
+      "step": 12554
+    },
+    {
+      "epoch": 2.0088,
+      "grad_norm": 2.1592513813968055,
+      "learning_rate": 5.2255409994482775e-05,
+      "loss": 1.1105,
+      "step": 12555
+    },
+    {
+      "epoch": 2.00896,
+      "grad_norm": 1.679131784576301,
+      "learning_rate": 5.224023285782296e-05,
+      "loss": 1.14,
+      "step": 12556
+    },
+    {
+      "epoch": 2.00912,
+      "grad_norm": 2.2267606069611943,
+      "learning_rate": 5.2225057146244993e-05,
+      "loss": 1.1781,
+      "step": 12557
+    },
+    {
+      "epoch": 2.00928,
+      "grad_norm": 2.0550857331285206,
+      "learning_rate": 5.220988286020172e-05,
+      "loss": 1.1218,
+      "step": 12558
+    },
+    {
+      "epoch": 2.00944,
+      "grad_norm": 2.0078532122557466,
+      "learning_rate": 5.219471000014586e-05,
+      "loss": 1.1111,
+      "step": 12559
+    },
+    {
+      "epoch": 2.0096,
+      "grad_norm": 2.1468661868738113,
+      "learning_rate": 5.2179538566530284e-05,
+      "loss": 1.1487,
+      "step": 12560
+    },
+    {
+      "epoch": 2.00976,
+      "grad_norm": 1.8084523838129376,
+      "learning_rate": 5.216436855980751e-05,
+      "loss": 1.1213,
+      "step": 12561
+    },
+    {
+      "epoch": 2.00992,
+      "grad_norm": 1.7218477557483973,
+      "learning_rate": 5.214919998043035e-05,
+      "loss": 1.0973,
+      "step": 12562
+    },
+    {
+      "epoch": 2.01008,
+      "grad_norm": 1.8936382391725968,
+      "learning_rate": 5.2134032828851256e-05,
+      "loss": 1.138,
+      "step": 12563
+    },
+    {
+      "epoch": 2.01024,
+      "grad_norm": 1.8930423799679503,
+      "learning_rate": 5.211886710552292e-05,
+      "loss": 1.117,
+      "step": 12564
+    },
+    {
+      "epoch": 2.0104,
+      "grad_norm": 2.142013583392089,
+      "learning_rate": 5.2103702810897816e-05,
+      "loss": 1.119,
+      "step": 12565
+    },
+    {
+      "epoch": 2.01056,
+      "grad_norm": 1.8026830455970153,
+      "learning_rate": 5.2088539945428416e-05,
+      "loss": 1.1692,
+      "step": 12566
+    },
+    {
+      "epoch": 2.01072,
+      "grad_norm": 2.1356750097944404,
+      "learning_rate": 5.2073378509567196e-05,
+      "loss": 1.0941,
+      "step": 12567
+    },
+    {
+      "epoch": 2.0108800000000002,
+      "grad_norm": 1.729069287858422,
+      "learning_rate": 5.205821850376651e-05,
+      "loss": 1.1249,
+      "step": 12568
+    },
+    {
+      "epoch": 2.01104,
+      "grad_norm": 2.00190133475559,
+      "learning_rate": 5.204305992847872e-05,
+      "loss": 1.1252,
+      "step": 12569
+    },
+    {
+      "epoch": 2.0112,
+      "grad_norm": 1.7466091804048007,
+      "learning_rate": 5.20279027841561e-05,
+      "loss": 1.1083,
+      "step": 12570
+    },
+    {
+      "epoch": 2.01136,
+      "grad_norm": 1.7846015369872903,
+      "learning_rate": 5.2012747071251056e-05,
+      "loss": 1.1331,
+      "step": 12571
+    },
+    {
+      "epoch": 2.01152,
+      "grad_norm": 1.7716768064489012,
+      "learning_rate": 5.199759279021562e-05,
+      "loss": 1.1183,
+      "step": 12572
+    },
+    {
+      "epoch": 2.01168,
+      "grad_norm": 1.8982087179004195,
+      "learning_rate": 5.198243994150216e-05,
+      "loss": 1.0755,
+      "step": 12573
+    },
+    {
+      "epoch": 2.01184,
+      "grad_norm": 1.8885444833810467,
+      "learning_rate": 5.196728852556263e-05,
+      "loss": 1.1545,
+      "step": 12574
+    },
+    {
+      "epoch": 2.012,
+      "grad_norm": 1.7747906551841945,
+      "learning_rate": 5.195213854284926e-05,
+      "loss": 1.1011,
+      "step": 12575
+    },
+    {
+      "epoch": 2.01216,
+      "grad_norm": 1.8995615702314563,
+      "learning_rate": 5.1936989993814064e-05,
+      "loss": 1.15,
+      "step": 12576
+    },
+    {
+      "epoch": 2.01232,
+      "grad_norm": 1.7544131620069745,
+      "learning_rate": 5.1921842878909063e-05,
+      "loss": 1.1412,
+      "step": 12577
+    },
+    {
+      "epoch": 2.01248,
+      "grad_norm": 1.777930290413517,
+      "learning_rate": 5.1906697198586204e-05,
+      "loss": 1.0791,
+      "step": 12578
+    },
+    {
+      "epoch": 2.01264,
+      "grad_norm": 1.9169166844330587,
+      "learning_rate": 5.1891552953297417e-05,
+      "loss": 1.1329,
+      "step": 12579
+    },
+    {
+      "epoch": 2.0128,
+      "grad_norm": 1.785897008414572,
+      "learning_rate": 5.18764101434946e-05,
+      "loss": 1.1701,
+      "step": 12580
+    },
+    {
+      "epoch": 2.01296,
+      "grad_norm": 1.8751576640362804,
+      "learning_rate": 5.1861268769629576e-05,
+      "loss": 1.1238,
+      "step": 12581
+    },
+    {
+      "epoch": 2.01312,
+      "grad_norm": 1.8512372879572003,
+      "learning_rate": 5.184612883215414e-05,
+      "loss": 1.0992,
+      "step": 12582
+    },
+    {
+      "epoch": 2.01328,
+      "grad_norm": 1.7027000862651482,
+      "learning_rate": 5.1830990331520054e-05,
+      "loss": 1.1349,
+      "step": 12583
+    },
+    {
+      "epoch": 2.01344,
+      "grad_norm": 1.7936778206724688,
+      "learning_rate": 5.181585326817899e-05,
+      "loss": 1.1367,
+      "step": 12584
+    },
+    {
+      "epoch": 2.0136,
+      "grad_norm": 1.8270372400029105,
+      "learning_rate": 5.180071764258274e-05,
+      "loss": 1.1539,
+      "step": 12585
+    },
+    {
+      "epoch": 2.01376,
+      "grad_norm": 1.716346695969295,
+      "learning_rate": 5.1785583455182754e-05,
+      "loss": 1.162,
+      "step": 12586
+    },
+    {
+      "epoch": 2.01392,
+      "grad_norm": 1.9863739038946124,
+      "learning_rate": 5.1770450706430775e-05,
+      "loss": 1.136,
+      "step": 12587
+    },
+    {
+      "epoch": 2.01408,
+      "grad_norm": 1.8663797849078543,
+      "learning_rate": 5.17553193967782e-05,
+      "loss": 1.1187,
+      "step": 12588
+    },
+    {
+      "epoch": 2.01424,
+      "grad_norm": 1.7259731674728864,
+      "learning_rate": 5.174018952667668e-05,
+      "loss": 1.1471,
+      "step": 12589
+    },
+    {
+      "epoch": 2.0144,
+      "grad_norm": 1.9717757172483596,
+      "learning_rate": 5.172506109657749e-05,
+      "loss": 1.1267,
+      "step": 12590
+    },
+    {
+      "epoch": 2.01456,
+      "grad_norm": 1.9545073750681123,
+      "learning_rate": 5.170993410693217e-05,
+      "loss": 1.124,
+      "step": 12591
+    },
+    {
+      "epoch": 2.01472,
+      "grad_norm": 1.9020285441170852,
+      "learning_rate": 5.1694808558192056e-05,
+      "loss": 1.1635,
+      "step": 12592
+    },
+    {
+      "epoch": 2.01488,
+      "grad_norm": 2.072211266065848,
+      "learning_rate": 5.167968445080846e-05,
+      "loss": 1.1161,
+      "step": 12593
+    },
+    {
+      "epoch": 2.01504,
+      "grad_norm": 1.7810851416702675,
+      "learning_rate": 5.166456178523268e-05,
+      "loss": 1.0995,
+      "step": 12594
+    },
+    {
+      "epoch": 2.0152,
+      "grad_norm": 1.8237132976761237,
+      "learning_rate": 5.164944056191594e-05,
+      "loss": 1.1054,
+      "step": 12595
+    },
+    {
+      "epoch": 2.01536,
+      "grad_norm": 2.257174791096831,
+      "learning_rate": 5.1634320781309434e-05,
+      "loss": 1.0885,
+      "step": 12596
+    },
+    {
+      "epoch": 2.01552,
+      "grad_norm": 1.9852730881798784,
+      "learning_rate": 5.1619202443864334e-05,
+      "loss": 1.1618,
+      "step": 12597
+    },
+    {
+      "epoch": 2.01568,
+      "grad_norm": 1.777186952533389,
+      "learning_rate": 5.160408555003172e-05,
+      "loss": 1.1476,
+      "step": 12598
+    },
+    {
+      "epoch": 2.01584,
+      "grad_norm": 2.27029024593019,
+      "learning_rate": 5.158897010026268e-05,
+      "loss": 1.1536,
+      "step": 12599
+    },
+    {
+      "epoch": 2.016,
+      "grad_norm": 2.2834529089600477,
+      "learning_rate": 5.15738560950082e-05,
+      "loss": 1.0931,
+      "step": 12600
+    },
+    {
+      "epoch": 2.01616,
+      "grad_norm": 1.7079305882936229,
+      "learning_rate": 5.1558743534719346e-05,
+      "loss": 1.0845,
+      "step": 12601
+    },
+    {
+      "epoch": 2.01632,
+      "grad_norm": 1.5870059231341764,
+      "learning_rate": 5.1543632419846944e-05,
+      "loss": 1.107,
+      "step": 12602
+    },
+    {
+      "epoch": 2.01648,
+      "grad_norm": 2.001668556489411,
+      "learning_rate": 5.1528522750842e-05,
+      "loss": 1.1362,
+      "step": 12603
+    },
+    {
+      "epoch": 2.01664,
+      "grad_norm": 1.660533735058852,
+      "learning_rate": 5.1513414528155236e-05,
+      "loss": 1.1118,
+      "step": 12604
+    },
+    {
+      "epoch": 2.0168,
+      "grad_norm": 1.8826169630077092,
+      "learning_rate": 5.149830775223761e-05,
+      "loss": 1.1276,
+      "step": 12605
+    },
+    {
+      "epoch": 2.01696,
+      "grad_norm": 1.7990670093474668,
+      "learning_rate": 5.1483202423539724e-05,
+      "loss": 1.1929,
+      "step": 12606
+    },
+    {
+      "epoch": 2.01712,
+      "grad_norm": 1.9976516327291318,
+      "learning_rate": 5.146809854251242e-05,
+      "loss": 1.136,
+      "step": 12607
+    },
+    {
+      "epoch": 2.01728,
+      "grad_norm": 1.7951666908442523,
+      "learning_rate": 5.145299610960635e-05,
+      "loss": 1.1436,
+      "step": 12608
+    },
+    {
+      "epoch": 2.01744,
+      "grad_norm": 1.9179655289944613,
+      "learning_rate": 5.143789512527213e-05,
+      "loss": 1.1688,
+      "step": 12609
+    },
+    {
+      "epoch": 2.0176,
+      "grad_norm": 2.0085208018017804,
+      "learning_rate": 5.142279558996035e-05,
+      "loss": 1.1086,
+      "step": 12610
+    },
+    {
+      "epoch": 2.01776,
+      "grad_norm": 1.9239650994451718,
+      "learning_rate": 5.140769750412156e-05,
+      "loss": 1.1251,
+      "step": 12611
+    },
+    {
+      "epoch": 2.01792,
+      "grad_norm": 1.966427465093524,
+      "learning_rate": 5.1392600868206284e-05,
+      "loss": 1.1325,
+      "step": 12612
+    },
+    {
+      "epoch": 2.01808,
+      "grad_norm": 1.7242306213766667,
+      "learning_rate": 5.137750568266497e-05,
+      "loss": 1.1001,
+      "step": 12613
+    },
+    {
+      "epoch": 2.01824,
+      "grad_norm": 1.702834852333757,
+      "learning_rate": 5.136241194794802e-05,
+      "loss": 1.1774,
+      "step": 12614
+    },
+    {
+      "epoch": 2.0184,
+      "grad_norm": 2.138188839428707,
+      "learning_rate": 5.134731966450583e-05,
+      "loss": 1.1607,
+      "step": 12615
+    },
+    {
+      "epoch": 2.01856,
+      "grad_norm": 1.9957147030667257,
+      "learning_rate": 5.13322288327887e-05,
+      "loss": 1.1241,
+      "step": 12616
+    },
+    {
+      "epoch": 2.01872,
+      "grad_norm": 1.860567014713556,
+      "learning_rate": 5.131713945324702e-05,
+      "loss": 1.1641,
+      "step": 12617
+    },
+    {
+      "epoch": 2.01888,
+      "grad_norm": 1.8997013463016545,
+      "learning_rate": 5.1302051526330875e-05,
+      "loss": 1.1083,
+      "step": 12618
+    },
+    {
+      "epoch": 2.01904,
+      "grad_norm": 1.9292546637408938,
+      "learning_rate": 5.128696505249064e-05,
+      "loss": 1.1772,
+      "step": 12619
+    },
+    {
+      "epoch": 2.0192,
+      "grad_norm": 1.7237681255705803,
+      "learning_rate": 5.12718800321763e-05,
+      "loss": 1.1241,
+      "step": 12620
+    },
+    {
+      "epoch": 2.01936,
+      "grad_norm": 1.7856289164517198,
+      "learning_rate": 5.125679646583813e-05,
+      "loss": 1.1549,
+      "step": 12621
+    },
+    {
+      "epoch": 2.01952,
+      "grad_norm": 1.8106893250919338,
+      "learning_rate": 5.124171435392606e-05,
+      "loss": 1.1272,
+      "step": 12622
+    },
+    {
+      "epoch": 2.01968,
+      "grad_norm": 2.0762372256843435,
+      "learning_rate": 5.122663369689022e-05,
+      "loss": 1.1017,
+      "step": 12623
+    },
+    {
+      "epoch": 2.01984,
+      "grad_norm": 1.9666283081345795,
+      "learning_rate": 5.121155449518056e-05,
+      "loss": 1.2176,
+      "step": 12624
+    },
+    {
+      "epoch": 2.02,
+      "grad_norm": 1.8916660839199215,
+      "learning_rate": 5.1196476749247016e-05,
+      "loss": 1.1527,
+      "step": 12625
+    },
+    {
+      "epoch": 2.02016,
+      "grad_norm": 1.8406785923534932,
+      "learning_rate": 5.1181400459539495e-05,
+      "loss": 1.1304,
+      "step": 12626
+    },
+    {
+      "epoch": 2.02032,
+      "grad_norm": 2.1347010584236914,
+      "learning_rate": 5.1166325626507804e-05,
+      "loss": 1.1401,
+      "step": 12627
+    },
+    {
+      "epoch": 2.02048,
+      "grad_norm": 1.8936924448693278,
+      "learning_rate": 5.115125225060189e-05,
+      "loss": 1.1624,
+      "step": 12628
+    },
+    {
+      "epoch": 2.02064,
+      "grad_norm": 2.069127006828981,
+      "learning_rate": 5.113618033227133e-05,
+      "loss": 1.1315,
+      "step": 12629
+    },
+    {
+      "epoch": 2.0208,
+      "grad_norm": 1.7525657877241234,
+      "learning_rate": 5.112110987196603e-05,
+      "loss": 1.1333,
+      "step": 12630
+    },
+    {
+      "epoch": 2.02096,
+      "grad_norm": 1.9084604266122989,
+      "learning_rate": 5.110604087013553e-05,
+      "loss": 1.0955,
+      "step": 12631
+    },
+    {
+      "epoch": 2.02112,
+      "grad_norm": 1.7367391071000453,
+      "learning_rate": 5.109097332722954e-05,
+      "loss": 1.0844,
+      "step": 12632
+    },
+    {
+      "epoch": 2.02128,
+      "grad_norm": 1.7712637544827847,
+      "learning_rate": 5.107590724369765e-05,
+      "loss": 1.1615,
+      "step": 12633
+    },
+    {
+      "epoch": 2.02144,
+      "grad_norm": 2.2090413494809,
+      "learning_rate": 5.1060842619989404e-05,
+      "loss": 1.1617,
+      "step": 12634
+    },
+    {
+      "epoch": 2.0216,
+      "grad_norm": 1.770087053651189,
+      "learning_rate": 5.104577945655429e-05,
+      "loss": 1.1248,
+      "step": 12635
+    },
+    {
+      "epoch": 2.02176,
+      "grad_norm": 1.929989251495988,
+      "learning_rate": 5.1030717753841785e-05,
+      "loss": 1.1609,
+      "step": 12636
+    },
+    {
+      "epoch": 2.02192,
+      "grad_norm": 2.0538718377760476,
+      "learning_rate": 5.1015657512301304e-05,
+      "loss": 1.1022,
+      "step": 12637
+    },
+    {
+      "epoch": 2.02208,
+      "grad_norm": 1.7319660710871552,
+      "learning_rate": 5.100059873238223e-05,
+      "loss": 1.1141,
+      "step": 12638
+    },
+    {
+      "epoch": 2.02224,
+      "grad_norm": 1.8393311291580585,
+      "learning_rate": 5.0985541414533886e-05,
+      "loss": 1.2057,
+      "step": 12639
+    },
+    {
+      "epoch": 2.0224,
+      "grad_norm": 1.877021942375122,
+      "learning_rate": 5.097048555920558e-05,
+      "loss": 1.1356,
+      "step": 12640
+    },
+    {
+      "epoch": 2.02256,
+      "grad_norm": 1.8809969155185822,
+      "learning_rate": 5.095543116684649e-05,
+      "loss": 1.1843,
+      "step": 12641
+    },
+    {
+      "epoch": 2.02272,
+      "grad_norm": 1.8205691419469312,
+      "learning_rate": 5.094037823790594e-05,
+      "loss": 1.1166,
+      "step": 12642
+    },
+    {
+      "epoch": 2.02288,
+      "grad_norm": 1.998928436904079,
+      "learning_rate": 5.0925326772832945e-05,
+      "loss": 1.1002,
+      "step": 12643
+    },
+    {
+      "epoch": 2.02304,
+      "grad_norm": 1.9066079411271435,
+      "learning_rate": 5.091027677207677e-05,
+      "loss": 1.142,
+      "step": 12644
+    },
+    {
+      "epoch": 2.0232,
+      "grad_norm": 1.7961882522262806,
+      "learning_rate": 5.089522823608632e-05,
+      "loss": 1.1136,
+      "step": 12645
+    },
+    {
+      "epoch": 2.02336,
+      "grad_norm": 1.768336295507302,
+      "learning_rate": 5.0880181165310795e-05,
+      "loss": 1.1326,
+      "step": 12646
+    },
+    {
+      "epoch": 2.02352,
+      "grad_norm": 2.014192853806705,
+      "learning_rate": 5.086513556019901e-05,
+      "loss": 1.1476,
+      "step": 12647
+    },
+    {
+      "epoch": 2.02368,
+      "grad_norm": 2.045236045778119,
+      "learning_rate": 5.0850091421200006e-05,
+      "loss": 1.0725,
+      "step": 12648
+    },
+    {
+      "epoch": 2.02384,
+      "grad_norm": 1.7898059280733818,
+      "learning_rate": 5.083504874876267e-05,
+      "loss": 1.1481,
+      "step": 12649
+    },
+    {
+      "epoch": 2.024,
+      "grad_norm": 1.944354376720142,
+      "learning_rate": 5.0820007543335834e-05,
+      "loss": 1.1383,
+      "step": 12650
+    },
+    {
+      "epoch": 2.02416,
+      "grad_norm": 2.114659429036422,
+      "learning_rate": 5.080496780536831e-05,
+      "loss": 1.1389,
+      "step": 12651
+    },
+    {
+      "epoch": 2.02432,
+      "grad_norm": 2.0754956741064032,
+      "learning_rate": 5.078992953530886e-05,
+      "loss": 1.108,
+      "step": 12652
+    },
+    {
+      "epoch": 2.02448,
+      "grad_norm": 1.6980588193774417,
+      "learning_rate": 5.077489273360621e-05,
+      "loss": 1.106,
+      "step": 12653
+    },
+    {
+      "epoch": 2.02464,
+      "grad_norm": 2.043662775705973,
+      "learning_rate": 5.075985740070903e-05,
+      "loss": 1.1882,
+      "step": 12654
+    },
+    {
+      "epoch": 2.0248,
+      "grad_norm": 2.008159034750406,
+      "learning_rate": 5.0744823537065955e-05,
+      "loss": 1.1433,
+      "step": 12655
+    },
+    {
+      "epoch": 2.02496,
+      "grad_norm": 2.108072784998321,
+      "learning_rate": 5.072979114312557e-05,
+      "loss": 1.1526,
+      "step": 12656
+    },
+    {
+      "epoch": 2.02512,
+      "grad_norm": 1.9853979185330957,
+      "learning_rate": 5.0714760219336385e-05,
+      "loss": 1.169,
+      "step": 12657
+    },
+    {
+      "epoch": 2.02528,
+      "grad_norm": 1.8360333872371224,
+      "learning_rate": 5.069973076614703e-05,
+      "loss": 1.1559,
+      "step": 12658
+    },
+    {
+      "epoch": 2.02544,
+      "grad_norm": 1.9665529142396845,
+      "learning_rate": 5.068470278400578e-05,
+      "loss": 1.0932,
+      "step": 12659
+    },
+    {
+      "epoch": 2.0256,
+      "grad_norm": 1.8980817577543914,
+      "learning_rate": 5.066967627336121e-05,
+      "loss": 1.0844,
+      "step": 12660
+    },
+    {
+      "epoch": 2.02576,
+      "grad_norm": 1.8485976104464432,
+      "learning_rate": 5.0654651234661554e-05,
+      "loss": 1.1341,
+      "step": 12661
+    },
+    {
+      "epoch": 2.02592,
+      "grad_norm": 1.7372333971231562,
+      "learning_rate": 5.063962766835528e-05,
+      "loss": 1.0718,
+      "step": 12662
+    },
+    {
+      "epoch": 2.02608,
+      "grad_norm": 1.7509990511996916,
+      "learning_rate": 5.062460557489051e-05,
+      "loss": 1.1118,
+      "step": 12663
+    },
+    {
+      "epoch": 2.02624,
+      "grad_norm": 2.2949239136681734,
+      "learning_rate": 5.0609584954715596e-05,
+      "loss": 1.1149,
+      "step": 12664
+    },
+    {
+      "epoch": 2.0264,
+      "grad_norm": 1.8648489257750689,
+      "learning_rate": 5.059456580827869e-05,
+      "loss": 1.1334,
+      "step": 12665
+    },
+    {
+      "epoch": 2.02656,
+      "grad_norm": 1.9192513208316204,
+      "learning_rate": 5.0579548136027966e-05,
+      "loss": 1.163,
+      "step": 12666
+    },
+    {
+      "epoch": 2.02672,
+      "grad_norm": 1.9910243480735854,
+      "learning_rate": 5.056453193841151e-05,
+      "loss": 1.1353,
+      "step": 12667
+    },
+    {
+      "epoch": 2.02688,
+      "grad_norm": 1.9852037012004926,
+      "learning_rate": 5.054951721587737e-05,
+      "loss": 1.1356,
+      "step": 12668
+    },
+    {
+      "epoch": 2.02704,
+      "grad_norm": 1.8770646540946831,
+      "learning_rate": 5.053450396887359e-05,
+      "loss": 1.1142,
+      "step": 12669
+    },
+    {
+      "epoch": 2.0272,
+      "grad_norm": 2.0063811488052976,
+      "learning_rate": 5.051949219784813e-05,
+      "loss": 1.1683,
+      "step": 12670
+    },
+    {
+      "epoch": 2.02736,
+      "grad_norm": 2.14236214339397,
+      "learning_rate": 5.0504481903248923e-05,
+      "loss": 1.0971,
+      "step": 12671
+    },
+    {
+      "epoch": 2.02752,
+      "grad_norm": 1.730953711890211,
+      "learning_rate": 5.0489473085523864e-05,
+      "loss": 1.1536,
+      "step": 12672
+    },
+    {
+      "epoch": 2.02768,
+      "grad_norm": 1.9099933249387222,
+      "learning_rate": 5.047446574512074e-05,
+      "loss": 1.1128,
+      "step": 12673
+    },
+    {
+      "epoch": 2.02784,
+      "grad_norm": 1.8070559206923786,
+      "learning_rate": 5.045945988248746e-05,
+      "loss": 1.1397,
+      "step": 12674
+    },
+    {
+      "epoch": 2.028,
+      "grad_norm": 2.122098116883452,
+      "learning_rate": 5.044445549807164e-05,
+      "loss": 1.1607,
+      "step": 12675
+    },
+    {
+      "epoch": 2.02816,
+      "grad_norm": 1.9003124881977707,
+      "learning_rate": 5.0429452592321125e-05,
+      "loss": 1.166,
+      "step": 12676
+    },
+    {
+      "epoch": 2.02832,
+      "grad_norm": 1.6936790873227685,
+      "learning_rate": 5.041445116568345e-05,
+      "loss": 1.1353,
+      "step": 12677
+    },
+    {
+      "epoch": 2.02848,
+      "grad_norm": 1.8700795856485386,
+      "learning_rate": 5.0399451218606366e-05,
+      "loss": 1.05,
+      "step": 12678
+    },
+    {
+      "epoch": 2.02864,
+      "grad_norm": 1.9640565668444674,
+      "learning_rate": 5.0384452751537315e-05,
+      "loss": 1.0973,
+      "step": 12679
+    },
+    {
+      "epoch": 2.0288,
+      "grad_norm": 1.7869028542824248,
+      "learning_rate": 5.036945576492392e-05,
+      "loss": 1.1315,
+      "step": 12680
+    },
+    {
+      "epoch": 2.02896,
+      "grad_norm": 2.1359870623063975,
+      "learning_rate": 5.0354460259213655e-05,
+      "loss": 1.1382,
+      "step": 12681
+    },
+    {
+      "epoch": 2.02912,
+      "grad_norm": 1.6809368229713668,
+      "learning_rate": 5.033946623485395e-05,
+      "loss": 1.1522,
+      "step": 12682
+    },
+    {
+      "epoch": 2.02928,
+      "grad_norm": 2.0669301961904467,
+      "learning_rate": 5.032447369229221e-05,
+      "loss": 1.1304,
+      "step": 12683
+    },
+    {
+      "epoch": 2.02944,
+      "grad_norm": 1.971283522400671,
+      "learning_rate": 5.030948263197579e-05,
+      "loss": 1.147,
+      "step": 12684
+    },
+    {
+      "epoch": 2.0296,
+      "grad_norm": 1.6948229771492482,
+      "learning_rate": 5.029449305435201e-05,
+      "loss": 1.1698,
+      "step": 12685
+    },
+    {
+      "epoch": 2.02976,
+      "grad_norm": 2.0366439758674493,
+      "learning_rate": 5.0279504959868086e-05,
+      "loss": 1.1435,
+      "step": 12686
+    },
+    {
+      "epoch": 2.02992,
+      "grad_norm": 1.9522797664085163,
+      "learning_rate": 5.0264518348971356e-05,
+      "loss": 1.1565,
+      "step": 12687
+    },
+    {
+      "epoch": 2.03008,
+      "grad_norm": 1.7617249618362807,
+      "learning_rate": 5.0249533222108855e-05,
+      "loss": 1.2047,
+      "step": 12688
+    },
+    {
+      "epoch": 2.03024,
+      "grad_norm": 1.9808301357470108,
+      "learning_rate": 5.023454957972782e-05,
+      "loss": 1.1236,
+      "step": 12689
+    },
+    {
+      "epoch": 2.0304,
+      "grad_norm": 1.6255318979302904,
+      "learning_rate": 5.021956742227532e-05,
+      "loss": 1.0881,
+      "step": 12690
+    },
+    {
+      "epoch": 2.03056,
+      "grad_norm": 2.3149200364718445,
+      "learning_rate": 5.020458675019838e-05,
+      "loss": 1.1487,
+      "step": 12691
+    },
+    {
+      "epoch": 2.03072,
+      "grad_norm": 1.7703654695877171,
+      "learning_rate": 5.018960756394402e-05,
+      "loss": 1.1757,
+      "step": 12692
+    },
+    {
+      "epoch": 2.03088,
+      "grad_norm": 2.0945103094235535,
+      "learning_rate": 5.017462986395918e-05,
+      "loss": 1.1058,
+      "step": 12693
+    },
+    {
+      "epoch": 2.03104,
+      "grad_norm": 1.8294423285320494,
+      "learning_rate": 5.015965365069078e-05,
+      "loss": 1.1138,
+      "step": 12694
+    },
+    {
+      "epoch": 2.0312,
+      "grad_norm": 2.0183330381625972,
+      "learning_rate": 5.01446789245857e-05,
+      "loss": 1.0758,
+      "step": 12695
+    },
+    {
+      "epoch": 2.03136,
+      "grad_norm": 1.8070610833144427,
+      "learning_rate": 5.012970568609075e-05,
+      "loss": 1.174,
+      "step": 12696
+    },
+    {
+      "epoch": 2.03152,
+      "grad_norm": 2.0646761412699717,
+      "learning_rate": 5.011473393565271e-05,
+      "loss": 1.1342,
+      "step": 12697
+    },
+    {
+      "epoch": 2.03168,
+      "grad_norm": 1.7803397996430368,
+      "learning_rate": 5.009976367371829e-05,
+      "loss": 1.085,
+      "step": 12698
+    },
+    {
+      "epoch": 2.03184,
+      "grad_norm": 1.87342922660537,
+      "learning_rate": 5.0084794900734276e-05,
+      "loss": 1.085,
+      "step": 12699
+    },
+    {
+      "epoch": 2.032,
+      "grad_norm": 1.9825040564156937,
+      "learning_rate": 5.006982761714718e-05,
+      "loss": 1.1634,
+      "step": 12700
+    },
+    {
+      "epoch": 2.03216,
+      "grad_norm": 1.9737937225185107,
+      "learning_rate": 5.0054861823403744e-05,
+      "loss": 1.1022,
+      "step": 12701
+    },
+    {
+      "epoch": 2.03232,
+      "grad_norm": 1.925663525135137,
+      "learning_rate": 5.0039897519950375e-05,
+      "loss": 1.1645,
+      "step": 12702
+    },
+    {
+      "epoch": 2.03248,
+      "grad_norm": 1.9196552249730554,
+      "learning_rate": 5.002493470723375e-05,
+      "loss": 1.1727,
+      "step": 12703
+    },
+    {
+      "epoch": 2.03264,
+      "grad_norm": 2.049526509855201,
+      "learning_rate": 5.000997338570017e-05,
+      "loss": 1.1523,
+      "step": 12704
+    },
+    {
+      "epoch": 2.0328,
+      "grad_norm": 1.9948359287675128,
+      "learning_rate": 4.9995013555796166e-05,
+      "loss": 1.1291,
+      "step": 12705
+    },
+    {
+      "epoch": 2.03296,
+      "grad_norm": 1.7882165251043114,
+      "learning_rate": 4.9980055217968116e-05,
+      "loss": 1.1045,
+      "step": 12706
+    },
+    {
+      "epoch": 2.03312,
+      "grad_norm": 2.0687628894285455,
+      "learning_rate": 4.996509837266231e-05,
+      "loss": 1.1237,
+      "step": 12707
+    },
+    {
+      "epoch": 2.03328,
+      "grad_norm": 1.9093550415292966,
+      "learning_rate": 4.995014302032507e-05,
+      "loss": 1.0658,
+      "step": 12708
+    },
+    {
+      "epoch": 2.03344,
+      "grad_norm": 2.210098280863122,
+      "learning_rate": 4.993518916140264e-05,
+      "loss": 1.1722,
+      "step": 12709
+    },
+    {
+      "epoch": 2.0336,
+      "grad_norm": 1.780674528458593,
+      "learning_rate": 4.9920236796341205e-05,
+      "loss": 1.0998,
+      "step": 12710
+    },
+    {
+      "epoch": 2.03376,
+      "grad_norm": 1.9223236475034944,
+      "learning_rate": 4.990528592558693e-05,
+      "loss": 1.1603,
+      "step": 12711
+    },
+    {
+      "epoch": 2.03392,
+      "grad_norm": 1.720459503693182,
+      "learning_rate": 4.989033654958593e-05,
+      "loss": 1.1495,
+      "step": 12712
+    },
+    {
+      "epoch": 2.03408,
+      "grad_norm": 1.735943632407312,
+      "learning_rate": 4.987538866878426e-05,
+      "loss": 1.1559,
+      "step": 12713
+    },
+    {
+      "epoch": 2.03424,
+      "grad_norm": 2.079398902154809,
+      "learning_rate": 4.986044228362793e-05,
+      "loss": 1.1137,
+      "step": 12714
+    },
+    {
+      "epoch": 2.0344,
+      "grad_norm": 1.887419074982742,
+      "learning_rate": 4.9845497394563e-05,
+      "loss": 1.1087,
+      "step": 12715
+    },
+    {
+      "epoch": 2.03456,
+      "grad_norm": 1.7764182208519688,
+      "learning_rate": 4.983055400203528e-05,
+      "loss": 1.157,
+      "step": 12716
+    },
+    {
+      "epoch": 2.03472,
+      "grad_norm": 2.0484703012404073,
+      "learning_rate": 4.98156121064908e-05,
+      "loss": 1.1407,
+      "step": 12717
+    },
+    {
+      "epoch": 2.03488,
+      "grad_norm": 1.84930654078138,
+      "learning_rate": 4.9800671708375246e-05,
+      "loss": 1.1432,
+      "step": 12718
+    },
+    {
+      "epoch": 2.03504,
+      "grad_norm": 1.6244076300922439,
+      "learning_rate": 4.9785732808134586e-05,
+      "loss": 1.1145,
+      "step": 12719
+    },
+    {
+      "epoch": 2.0352,
+      "grad_norm": 1.7831905529455412,
+      "learning_rate": 4.977079540621441e-05,
+      "loss": 1.13,
+      "step": 12720
+    },
+    {
+      "epoch": 2.03536,
+      "grad_norm": 1.959123713737054,
+      "learning_rate": 4.975585950306054e-05,
+      "loss": 1.1551,
+      "step": 12721
+    },
+    {
+      "epoch": 2.03552,
+      "grad_norm": 2.0468536493115144,
+      "learning_rate": 4.9740925099118615e-05,
+      "loss": 1.1271,
+      "step": 12722
+    },
+    {
+      "epoch": 2.03568,
+      "grad_norm": 1.8730955627460422,
+      "learning_rate": 4.972599219483425e-05,
+      "loss": 1.1614,
+      "step": 12723
+    },
+    {
+      "epoch": 2.03584,
+      "grad_norm": 1.997459523314318,
+      "learning_rate": 4.971106079065301e-05,
+      "loss": 1.1592,
+      "step": 12724
+    },
+    {
+      "epoch": 2.036,
+      "grad_norm": 1.9703544378904614,
+      "learning_rate": 4.969613088702045e-05,
+      "loss": 1.175,
+      "step": 12725
+    },
+    {
+      "epoch": 2.03616,
+      "grad_norm": 1.738635651887743,
+      "learning_rate": 4.968120248438204e-05,
+      "loss": 1.1164,
+      "step": 12726
+    },
+    {
+      "epoch": 2.03632,
+      "grad_norm": 1.8120451272068034,
+      "learning_rate": 4.966627558318322e-05,
+      "loss": 1.1304,
+      "step": 12727
+    },
+    {
+      "epoch": 2.03648,
+      "grad_norm": 1.6941366263190094,
+      "learning_rate": 4.9651350183869406e-05,
+      "loss": 1.0699,
+      "step": 12728
+    },
+    {
+      "epoch": 2.03664,
+      "grad_norm": 1.8349291850355876,
+      "learning_rate": 4.9636426286885915e-05,
+      "loss": 1.1204,
+      "step": 12729
+    },
+    {
+      "epoch": 2.0368,
+      "grad_norm": 1.8351503635718167,
+      "learning_rate": 4.9621503892678056e-05,
+      "loss": 1.2061,
+      "step": 12730
+    },
+    {
+      "epoch": 2.03696,
+      "grad_norm": 2.367002078855624,
+      "learning_rate": 4.9606583001691187e-05,
+      "loss": 1.1512,
+      "step": 12731
+    },
+    {
+      "epoch": 2.03712,
+      "grad_norm": 1.7813597778015648,
+      "learning_rate": 4.9591663614370365e-05,
+      "loss": 1.1116,
+      "step": 12732
+    },
+    {
+      "epoch": 2.03728,
+      "grad_norm": 1.769821238048886,
+      "learning_rate": 4.957674573116093e-05,
+      "loss": 1.1428,
+      "step": 12733
+    },
+    {
+      "epoch": 2.03744,
+      "grad_norm": 1.867704947974715,
+      "learning_rate": 4.956182935250785e-05,
+      "loss": 1.1402,
+      "step": 12734
+    },
+    {
+      "epoch": 2.0376,
+      "grad_norm": 1.8451243708135703,
+      "learning_rate": 4.9546914478856365e-05,
+      "loss": 1.1491,
+      "step": 12735
+    },
+    {
+      "epoch": 2.03776,
+      "grad_norm": 2.083226798695409,
+      "learning_rate": 4.953200111065136e-05,
+      "loss": 1.1628,
+      "step": 12736
+    },
+    {
+      "epoch": 2.03792,
+      "grad_norm": 1.8413850703346393,
+      "learning_rate": 4.9517089248337935e-05,
+      "loss": 1.1416,
+      "step": 12737
+    },
+    {
+      "epoch": 2.03808,
+      "grad_norm": 1.9613772593455634,
+      "learning_rate": 4.950217889236101e-05,
+      "loss": 1.0512,
+      "step": 12738
+    },
+    {
+      "epoch": 2.03824,
+      "grad_norm": 2.1692858981455454,
+      "learning_rate": 4.948727004316548e-05,
+      "loss": 1.1558,
+      "step": 12739
+    },
+    {
+      "epoch": 2.0384,
+      "grad_norm": 1.7654294106883655,
+      "learning_rate": 4.947236270119621e-05,
+      "loss": 1.1722,
+      "step": 12740
+    },
+    {
+      "epoch": 2.03856,
+      "grad_norm": 1.6902826290070134,
+      "learning_rate": 4.9457456866898e-05,
+      "loss": 1.1024,
+      "step": 12741
+    },
+    {
+      "epoch": 2.03872,
+      "grad_norm": 1.7845650550612038,
+      "learning_rate": 4.9442552540715634e-05,
+      "loss": 1.1191,
+      "step": 12742
+    },
+    {
+      "epoch": 2.03888,
+      "grad_norm": 1.984611089677538,
+      "learning_rate": 4.942764972309384e-05,
+      "loss": 1.0602,
+      "step": 12743
+    },
+    {
+      "epoch": 2.03904,
+      "grad_norm": 1.7958677271677341,
+      "learning_rate": 4.9412748414477284e-05,
+      "loss": 1.1628,
+      "step": 12744
+    },
+    {
+      "epoch": 2.0392,
+      "grad_norm": 2.2507256298770106,
+      "learning_rate": 4.939784861531055e-05,
+      "loss": 1.1439,
+      "step": 12745
+    },
+    {
+      "epoch": 2.03936,
+      "grad_norm": 1.808082108741239,
+      "learning_rate": 4.9382950326038324e-05,
+      "loss": 1.1489,
+      "step": 12746
+    },
+    {
+      "epoch": 2.03952,
+      "grad_norm": 1.5874566248736823,
+      "learning_rate": 4.936805354710511e-05,
+      "loss": 1.1552,
+      "step": 12747
+    },
+    {
+      "epoch": 2.03968,
+      "grad_norm": 2.3034946669664933,
+      "learning_rate": 4.935315827895539e-05,
+      "loss": 1.1181,
+      "step": 12748
+    },
+    {
+      "epoch": 2.03984,
+      "grad_norm": 1.9975222698742423,
+      "learning_rate": 4.933826452203364e-05,
+      "loss": 1.1323,
+      "step": 12749
+    },
+    {
+      "epoch": 2.04,
+      "grad_norm": 2.078184058491704,
+      "learning_rate": 4.9323372276784244e-05,
+      "loss": 1.1403,
+      "step": 12750
+    },
+    {
+      "epoch": 2.04016,
+      "grad_norm": 1.9882960436509283,
+      "learning_rate": 4.9308481543651575e-05,
+      "loss": 1.1686,
+      "step": 12751
+    },
+    {
+      "epoch": 2.04032,
+      "grad_norm": 2.0313244574596716,
+      "learning_rate": 4.929359232307995e-05,
+      "loss": 1.1423,
+      "step": 12752
+    },
+    {
+      "epoch": 2.04048,
+      "grad_norm": 1.986263493143512,
+      "learning_rate": 4.9278704615513646e-05,
+      "loss": 1.1041,
+      "step": 12753
+    },
+    {
+      "epoch": 2.04064,
+      "grad_norm": 1.8144668827055648,
+      "learning_rate": 4.92638184213969e-05,
+      "loss": 1.1281,
+      "step": 12754
+    },
+    {
+      "epoch": 2.0408,
+      "grad_norm": 1.8893535785334836,
+      "learning_rate": 4.9248933741173865e-05,
+      "loss": 1.0988,
+      "step": 12755
+    },
+    {
+      "epoch": 2.04096,
+      "grad_norm": 2.0733379830202305,
+      "learning_rate": 4.9234050575288704e-05,
+      "loss": 1.0816,
+      "step": 12756
+    },
+    {
+      "epoch": 2.04112,
+      "grad_norm": 1.8586985358033363,
+      "learning_rate": 4.9219168924185455e-05,
+      "loss": 1.1617,
+      "step": 12757
+    },
+    {
+      "epoch": 2.04128,
+      "grad_norm": 1.6365584032994551,
+      "learning_rate": 4.9204288788308295e-05,
+      "loss": 1.1192,
+      "step": 12758
+    },
+    {
+      "epoch": 2.04144,
+      "grad_norm": 1.586095894626809,
+      "learning_rate": 4.9189410168101055e-05,
+      "loss": 1.0916,
+      "step": 12759
+    },
+    {
+      "epoch": 2.0416,
+      "grad_norm": 1.854698256865999,
+      "learning_rate": 4.917453306400785e-05,
+      "loss": 1.1219,
+      "step": 12760
+    },
+    {
+      "epoch": 2.04176,
+      "grad_norm": 1.7909522169667051,
+      "learning_rate": 4.915965747647246e-05,
+      "loss": 1.112,
+      "step": 12761
+    },
+    {
+      "epoch": 2.04192,
+      "grad_norm": 1.9537710107390185,
+      "learning_rate": 4.914478340593883e-05,
+      "loss": 1.1759,
+      "step": 12762
+    },
+    {
+      "epoch": 2.04208,
+      "grad_norm": 1.8462689172869104,
+      "learning_rate": 4.912991085285077e-05,
+      "loss": 1.1504,
+      "step": 12763
+    },
+    {
+      "epoch": 2.04224,
+      "grad_norm": 2.0268370619307725,
+      "learning_rate": 4.911503981765203e-05,
+      "loss": 1.1083,
+      "step": 12764
+    },
+    {
+      "epoch": 2.0424,
+      "grad_norm": 1.6728431640174508,
+      "learning_rate": 4.9100170300786366e-05,
+      "loss": 1.1262,
+      "step": 12765
+    },
+    {
+      "epoch": 2.04256,
+      "grad_norm": 1.9370616585572733,
+      "learning_rate": 4.908530230269746e-05,
+      "loss": 1.1653,
+      "step": 12766
+    },
+    {
+      "epoch": 2.04272,
+      "grad_norm": 1.8128774571524258,
+      "learning_rate": 4.907043582382892e-05,
+      "loss": 1.121,
+      "step": 12767
+    },
+    {
+      "epoch": 2.04288,
+      "grad_norm": 2.122138118276405,
+      "learning_rate": 4.905557086462438e-05,
+      "loss": 1.1566,
+      "step": 12768
+    },
+    {
+      "epoch": 2.04304,
+      "grad_norm": 2.0711297649073237,
+      "learning_rate": 4.904070742552736e-05,
+      "loss": 1.14,
+      "step": 12769
+    },
+    {
+      "epoch": 2.0432,
+      "grad_norm": 1.7677900970405573,
+      "learning_rate": 4.902584550698138e-05,
+      "loss": 1.0751,
+      "step": 12770
+    },
+    {
+      "epoch": 2.04336,
+      "grad_norm": 1.8236659783763067,
+      "learning_rate": 4.9010985109429895e-05,
+      "loss": 1.1179,
+      "step": 12771
+    },
+    {
+      "epoch": 2.04352,
+      "grad_norm": 1.9710010233199489,
+      "learning_rate": 4.899612623331632e-05,
+      "loss": 1.1419,
+      "step": 12772
+    },
+    {
+      "epoch": 2.04368,
+      "grad_norm": 1.704869669523425,
+      "learning_rate": 4.898126887908397e-05,
+      "loss": 1.1498,
+      "step": 12773
+    },
+    {
+      "epoch": 2.04384,
+      "grad_norm": 2.1636635539372344,
+      "learning_rate": 4.8966413047176294e-05,
+      "loss": 1.168,
+      "step": 12774
+    },
+    {
+      "epoch": 2.044,
+      "grad_norm": 1.9735358142702446,
+      "learning_rate": 4.895155873803642e-05,
+      "loss": 1.1409,
+      "step": 12775
+    },
+    {
+      "epoch": 2.04416,
+      "grad_norm": 1.7423254669361339,
+      "learning_rate": 4.893670595210771e-05,
+      "loss": 1.1056,
+      "step": 12776
+    },
+    {
+      "epoch": 2.04432,
+      "grad_norm": 1.688871165112059,
+      "learning_rate": 4.8921854689833225e-05,
+      "loss": 1.0901,
+      "step": 12777
+    },
+    {
+      "epoch": 2.04448,
+      "grad_norm": 1.8383995182280057,
+      "learning_rate": 4.8907004951656196e-05,
+      "loss": 1.1197,
+      "step": 12778
+    },
+    {
+      "epoch": 2.04464,
+      "grad_norm": 1.775244860553942,
+      "learning_rate": 4.889215673801969e-05,
+      "loss": 1.1357,
+      "step": 12779
+    },
+    {
+      "epoch": 2.0448,
+      "grad_norm": 1.9351761366721676,
+      "learning_rate": 4.8877310049366755e-05,
+      "loss": 1.1058,
+      "step": 12780
+    },
+    {
+      "epoch": 2.04496,
+      "grad_norm": 1.8639108688975812,
+      "learning_rate": 4.88624648861404e-05,
+      "loss": 1.181,
+      "step": 12781
+    },
+    {
+      "epoch": 2.04512,
+      "grad_norm": 2.435135270404393,
+      "learning_rate": 4.884762124878357e-05,
+      "loss": 1.1301,
+      "step": 12782
+    },
+    {
+      "epoch": 2.04528,
+      "grad_norm": 2.0668609274185634,
+      "learning_rate": 4.883277913773919e-05,
+      "loss": 1.1513,
+      "step": 12783
+    },
+    {
+      "epoch": 2.04544,
+      "grad_norm": 1.949234825591589,
+      "learning_rate": 4.881793855345012e-05,
+      "loss": 1.1607,
+      "step": 12784
+    },
+    {
+      "epoch": 2.0456,
+      "grad_norm": 1.976574919246906,
+      "learning_rate": 4.8803099496359184e-05,
+      "loss": 1.1539,
+      "step": 12785
+    },
+    {
+      "epoch": 2.04576,
+      "grad_norm": 1.542790085065,
+      "learning_rate": 4.878826196690915e-05,
+      "loss": 1.1216,
+      "step": 12786
+    },
+    {
+      "epoch": 2.04592,
+      "grad_norm": 1.7920408491919209,
+      "learning_rate": 4.877342596554273e-05,
+      "loss": 1.1108,
+      "step": 12787
+    },
+    {
+      "epoch": 2.04608,
+      "grad_norm": 2.345470713171099,
+      "learning_rate": 4.875859149270271e-05,
+      "loss": 1.0956,
+      "step": 12788
+    },
+    {
+      "epoch": 2.04624,
+      "grad_norm": 1.799591193618586,
+      "learning_rate": 4.8743758548831574e-05,
+      "loss": 1.1773,
+      "step": 12789
+    },
+    {
+      "epoch": 2.0464,
+      "grad_norm": 1.875146002970931,
+      "learning_rate": 4.8728927134372074e-05,
+      "loss": 1.1213,
+      "step": 12790
+    },
+    {
+      "epoch": 2.04656,
+      "grad_norm": 1.7913527516241186,
+      "learning_rate": 4.871409724976659e-05,
+      "loss": 1.1434,
+      "step": 12791
+    },
+    {
+      "epoch": 2.04672,
+      "grad_norm": 1.923808240609987,
+      "learning_rate": 4.8699268895457796e-05,
+      "loss": 1.1318,
+      "step": 12792
+    },
+    {
+      "epoch": 2.04688,
+      "grad_norm": 1.8205854680789024,
+      "learning_rate": 4.868444207188799e-05,
+      "loss": 1.1123,
+      "step": 12793
+    },
+    {
+      "epoch": 2.04704,
+      "grad_norm": 1.961118444381336,
+      "learning_rate": 4.866961677949969e-05,
+      "loss": 1.1307,
+      "step": 12794
+    },
+    {
+      "epoch": 2.0472,
+      "grad_norm": 1.6537186954564425,
+      "learning_rate": 4.865479301873523e-05,
+      "loss": 1.1539,
+      "step": 12795
+    },
+    {
+      "epoch": 2.04736,
+      "grad_norm": 1.9016601296300812,
+      "learning_rate": 4.863997079003694e-05,
+      "loss": 1.2038,
+      "step": 12796
+    },
+    {
+      "epoch": 2.04752,
+      "grad_norm": 1.7673358988530865,
+      "learning_rate": 4.862515009384707e-05,
+      "loss": 1.1445,
+      "step": 12797
+    },
+    {
+      "epoch": 2.04768,
+      "grad_norm": 1.8274070638357687,
+      "learning_rate": 4.861033093060787e-05,
+      "loss": 1.0957,
+      "step": 12798
+    },
+    {
+      "epoch": 2.04784,
+      "grad_norm": 2.2927231409460975,
+      "learning_rate": 4.8595513300761506e-05,
+      "loss": 1.1603,
+      "step": 12799
+    },
+    {
+      "epoch": 2.048,
+      "grad_norm": 1.7640321397506271,
+      "learning_rate": 4.858069720475012e-05,
+      "loss": 1.1299,
+      "step": 12800
+    },
+    {
+      "epoch": 2.04816,
+      "grad_norm": 1.76845539071376,
+      "learning_rate": 4.856588264301581e-05,
+      "loss": 1.1756,
+      "step": 12801
+    },
+    {
+      "epoch": 2.04832,
+      "grad_norm": 2.025288308283933,
+      "learning_rate": 4.855106961600061e-05,
+      "loss": 1.1283,
+      "step": 12802
+    },
+    {
+      "epoch": 2.04848,
+      "grad_norm": 1.8689266241685878,
+      "learning_rate": 4.8536258124146485e-05,
+      "loss": 1.1351,
+      "step": 12803
+    },
+    {
+      "epoch": 2.04864,
+      "grad_norm": 1.6848723550151399,
+      "learning_rate": 4.8521448167895466e-05,
+      "loss": 1.1448,
+      "step": 12804
+    },
+    {
+      "epoch": 2.0488,
+      "grad_norm": 2.140370950139357,
+      "learning_rate": 4.850663974768942e-05,
+      "loss": 1.2011,
+      "step": 12805
+    },
+    {
+      "epoch": 2.04896,
+      "grad_norm": 1.7475638457248603,
+      "learning_rate": 4.849183286397021e-05,
+      "loss": 1.1547,
+      "step": 12806
+    },
+    {
+      "epoch": 2.04912,
+      "grad_norm": 1.869401686837673,
+      "learning_rate": 4.847702751717965e-05,
+      "loss": 1.1352,
+      "step": 12807
+    },
+    {
+      "epoch": 2.04928,
+      "grad_norm": 1.9816201633535584,
+      "learning_rate": 4.846222370775952e-05,
+      "loss": 1.1598,
+      "step": 12808
+    },
+    {
+      "epoch": 2.04944,
+      "grad_norm": 1.7616135461227937,
+      "learning_rate": 4.844742143615153e-05,
+      "loss": 1.1397,
+      "step": 12809
+    },
+    {
+      "epoch": 2.0496,
+      "grad_norm": 1.786081607318871,
+      "learning_rate": 4.8432620702797374e-05,
+      "loss": 1.1121,
+      "step": 12810
+    },
+    {
+      "epoch": 2.04976,
+      "grad_norm": 1.8400588302604952,
+      "learning_rate": 4.841782150813866e-05,
+      "loss": 1.181,
+      "step": 12811
+    },
+    {
+      "epoch": 2.04992,
+      "grad_norm": 1.9304328033327576,
+      "learning_rate": 4.840302385261699e-05,
+      "loss": 1.1621,
+      "step": 12812
+    },
+    {
+      "epoch": 2.05008,
+      "grad_norm": 1.8311830464036358,
+      "learning_rate": 4.838822773667392e-05,
+      "loss": 1.1834,
+      "step": 12813
+    },
+    {
+      "epoch": 2.05024,
+      "grad_norm": 1.983670593452692,
+      "learning_rate": 4.837343316075087e-05,
+      "loss": 1.1713,
+      "step": 12814
+    },
+    {
+      "epoch": 2.0504,
+      "grad_norm": 2.1428884612412475,
+      "learning_rate": 4.835864012528943e-05,
+      "loss": 1.1456,
+      "step": 12815
+    },
+    {
+      "epoch": 2.05056,
+      "grad_norm": 1.7756965858545393,
+      "learning_rate": 4.834384863073085e-05,
+      "loss": 1.1573,
+      "step": 12816
+    },
+    {
+      "epoch": 2.05072,
+      "grad_norm": 1.646426528056101,
+      "learning_rate": 4.8329058677516626e-05,
+      "loss": 1.1425,
+      "step": 12817
+    },
+    {
+      "epoch": 2.05088,
+      "grad_norm": 1.776047753560271,
+      "learning_rate": 4.831427026608792e-05,
+      "loss": 1.1399,
+      "step": 12818
+    },
+    {
+      "epoch": 2.05104,
+      "grad_norm": 1.582778254875306,
+      "learning_rate": 4.8299483396886124e-05,
+      "loss": 1.1511,
+      "step": 12819
+    },
+    {
+      "epoch": 2.0512,
+      "grad_norm": 1.6969718743893136,
+      "learning_rate": 4.82846980703524e-05,
+      "loss": 1.1269,
+      "step": 12820
+    },
+    {
+      "epoch": 2.05136,
+      "grad_norm": 1.9327198886942827,
+      "learning_rate": 4.8269914286927934e-05,
+      "loss": 1.1157,
+      "step": 12821
+    },
+    {
+      "epoch": 2.05152,
+      "grad_norm": 1.920057432209592,
+      "learning_rate": 4.825513204705383e-05,
+      "loss": 1.1525,
+      "step": 12822
+    },
+    {
+      "epoch": 2.05168,
+      "grad_norm": 1.8287260213408871,
+      "learning_rate": 4.82403513511712e-05,
+      "loss": 1.124,
+      "step": 12823
+    },
+    {
+      "epoch": 2.05184,
+      "grad_norm": 1.9194947566556648,
+      "learning_rate": 4.8225572199721056e-05,
+      "loss": 1.1452,
+      "step": 12824
+    },
+    {
+      "epoch": 2.052,
+      "grad_norm": 1.6950096066344749,
+      "learning_rate": 4.82107945931444e-05,
+      "loss": 1.255,
+      "step": 12825
+    },
+    {
+      "epoch": 2.05216,
+      "grad_norm": 1.8766035596393664,
+      "learning_rate": 4.819601853188217e-05,
+      "loss": 1.1634,
+      "step": 12826
+    },
+    {
+      "epoch": 2.05232,
+      "grad_norm": 1.9599763463946278,
+      "learning_rate": 4.818124401637526e-05,
+      "loss": 1.1386,
+      "step": 12827
+    },
+    {
+      "epoch": 2.05248,
+      "grad_norm": 1.967815961153488,
+      "learning_rate": 4.816647104706452e-05,
+      "loss": 1.0824,
+      "step": 12828
+    },
+    {
+      "epoch": 2.05264,
+      "grad_norm": 1.8535275065103547,
+      "learning_rate": 4.8151699624390756e-05,
+      "loss": 1.097,
+      "step": 12829
+    },
+    {
+      "epoch": 2.0528,
+      "grad_norm": 1.8466937305262834,
+      "learning_rate": 4.813692974879468e-05,
+      "loss": 1.1658,
+      "step": 12830
+    },
+    {
+      "epoch": 2.05296,
+      "grad_norm": 2.0818048847924193,
+      "learning_rate": 4.812216142071713e-05,
+      "loss": 1.1375,
+      "step": 12831
+    },
+    {
+      "epoch": 2.05312,
+      "grad_norm": 1.7281924024337176,
+      "learning_rate": 4.8107394640598614e-05,
+      "loss": 1.1302,
+      "step": 12832
+    },
+    {
+      "epoch": 2.05328,
+      "grad_norm": 1.7780279310037899,
+      "learning_rate": 4.8092629408879894e-05,
+      "loss": 1.1102,
+      "step": 12833
+    },
+    {
+      "epoch": 2.05344,
+      "grad_norm": 1.82250054436789,
+      "learning_rate": 4.8077865726001406e-05,
+      "loss": 1.1408,
+      "step": 12834
+    },
+    {
+      "epoch": 2.0536,
+      "grad_norm": 1.9233108867133941,
+      "learning_rate": 4.8063103592403777e-05,
+      "loss": 1.1448,
+      "step": 12835
+    },
+    {
+      "epoch": 2.05376,
+      "grad_norm": 1.722599140783903,
+      "learning_rate": 4.804834300852746e-05,
+      "loss": 1.0905,
+      "step": 12836
+    },
+    {
+      "epoch": 2.05392,
+      "grad_norm": 1.939897606048198,
+      "learning_rate": 4.80335839748129e-05,
+      "loss": 1.1433,
+      "step": 12837
+    },
+    {
+      "epoch": 2.05408,
+      "grad_norm": 1.7911574061874977,
+      "learning_rate": 4.801882649170045e-05,
+      "loss": 1.0833,
+      "step": 12838
+    },
+    {
+      "epoch": 2.05424,
+      "grad_norm": 1.9341602770979696,
+      "learning_rate": 4.800407055963049e-05,
+      "loss": 1.1268,
+      "step": 12839
+    },
+    {
+      "epoch": 2.0544,
+      "grad_norm": 1.78831572606232,
+      "learning_rate": 4.7989316179043285e-05,
+      "loss": 1.1416,
+      "step": 12840
+    },
+    {
+      "epoch": 2.05456,
+      "grad_norm": 1.7830971256138268,
+      "learning_rate": 4.797456335037911e-05,
+      "loss": 1.1261,
+      "step": 12841
+    },
+    {
+      "epoch": 2.05472,
+      "grad_norm": 1.8404340469749503,
+      "learning_rate": 4.795981207407815e-05,
+      "loss": 1.14,
+      "step": 12842
+    },
+    {
+      "epoch": 2.05488,
+      "grad_norm": 1.949524579433655,
+      "learning_rate": 4.794506235058056e-05,
+      "loss": 1.1327,
+      "step": 12843
+    },
+    {
+      "epoch": 2.05504,
+      "grad_norm": 1.943420071094316,
+      "learning_rate": 4.793031418032647e-05,
+      "loss": 1.1193,
+      "step": 12844
+    },
+    {
+      "epoch": 2.0552,
+      "grad_norm": 1.6250911840498707,
+      "learning_rate": 4.7915567563755926e-05,
+      "loss": 1.0872,
+      "step": 12845
+    },
+    {
+      "epoch": 2.05536,
+      "grad_norm": 1.8506035649982753,
+      "learning_rate": 4.790082250130891e-05,
+      "loss": 1.1582,
+      "step": 12846
+    },
+    {
+      "epoch": 2.05552,
+      "grad_norm": 1.9252034291086573,
+      "learning_rate": 4.788607899342551e-05,
+      "loss": 1.1196,
+      "step": 12847
+    },
+    {
+      "epoch": 2.05568,
+      "grad_norm": 2.1026105170947,
+      "learning_rate": 4.787133704054551e-05,
+      "loss": 1.1243,
+      "step": 12848
+    },
+    {
+      "epoch": 2.05584,
+      "grad_norm": 1.8914255940032065,
+      "learning_rate": 4.7856596643108906e-05,
+      "loss": 1.095,
+      "step": 12849
+    },
+    {
+      "epoch": 2.056,
+      "grad_norm": 1.739195338414446,
+      "learning_rate": 4.7841857801555426e-05,
+      "loss": 1.1327,
+      "step": 12850
+    },
+    {
+      "epoch": 2.05616,
+      "grad_norm": 2.1372020773792637,
+      "learning_rate": 4.782712051632493e-05,
+      "loss": 1.0985,
+      "step": 12851
+    },
+    {
+      "epoch": 2.05632,
+      "grad_norm": 2.1948113551614177,
+      "learning_rate": 4.781238478785713e-05,
+      "loss": 1.1325,
+      "step": 12852
+    },
+    {
+      "epoch": 2.05648,
+      "grad_norm": 1.7845210704818772,
+      "learning_rate": 4.779765061659174e-05,
+      "loss": 1.1398,
+      "step": 12853
+    },
+    {
+      "epoch": 2.05664,
+      "grad_norm": 1.9559164001036813,
+      "learning_rate": 4.778291800296838e-05,
+      "loss": 1.1664,
+      "step": 12854
+    },
+    {
+      "epoch": 2.0568,
+      "grad_norm": 1.9484960688035773,
+      "learning_rate": 4.776818694742665e-05,
+      "loss": 1.141,
+      "step": 12855
+    },
+    {
+      "epoch": 2.05696,
+      "grad_norm": 2.123231190339718,
+      "learning_rate": 4.775345745040612e-05,
+      "loss": 1.0752,
+      "step": 12856
+    },
+    {
+      "epoch": 2.05712,
+      "grad_norm": 1.9950266888648232,
+      "learning_rate": 4.773872951234629e-05,
+      "loss": 1.1803,
+      "step": 12857
+    },
+    {
+      "epoch": 2.05728,
+      "grad_norm": 1.8148030533564987,
+      "learning_rate": 4.772400313368662e-05,
+      "loss": 1.1221,
+      "step": 12858
+    },
+    {
+      "epoch": 2.05744,
+      "grad_norm": 1.9230883282675848,
+      "learning_rate": 4.770927831486652e-05,
+      "loss": 1.147,
+      "step": 12859
+    },
+    {
+      "epoch": 2.0576,
+      "grad_norm": 2.1935144228427674,
+      "learning_rate": 4.769455505632535e-05,
+      "loss": 1.1766,
+      "step": 12860
+    },
+    {
+      "epoch": 2.05776,
+      "grad_norm": 1.7759744896733607,
+      "learning_rate": 4.767983335850241e-05,
+      "loss": 1.0803,
+      "step": 12861
+    },
+    {
+      "epoch": 2.05792,
+      "grad_norm": 1.680020444802472,
+      "learning_rate": 4.7665113221837056e-05,
+      "loss": 1.1441,
+      "step": 12862
+    },
+    {
+      "epoch": 2.05808,
+      "grad_norm": 2.010145688476872,
+      "learning_rate": 4.765039464676845e-05,
+      "loss": 1.1709,
+      "step": 12863
+    },
+    {
+      "epoch": 2.05824,
+      "grad_norm": 1.842286202581359,
+      "learning_rate": 4.763567763373577e-05,
+      "loss": 1.1442,
+      "step": 12864
+    },
+    {
+      "epoch": 2.0584,
+      "grad_norm": 1.7356370474333964,
+      "learning_rate": 4.76209621831782e-05,
+      "loss": 1.112,
+      "step": 12865
+    },
+    {
+      "epoch": 2.05856,
+      "grad_norm": 1.9416160687785455,
+      "learning_rate": 4.760624829553477e-05,
+      "loss": 1.1235,
+      "step": 12866
+    },
+    {
+      "epoch": 2.05872,
+      "grad_norm": 2.0117844461288263,
+      "learning_rate": 4.759153597124455e-05,
+      "loss": 1.1431,
+      "step": 12867
+    },
+    {
+      "epoch": 2.05888,
+      "grad_norm": 1.6608642955605015,
+      "learning_rate": 4.757682521074654e-05,
+      "loss": 1.1692,
+      "step": 12868
+    },
+    {
+      "epoch": 2.05904,
+      "grad_norm": 1.7742529721426161,
+      "learning_rate": 4.756211601447966e-05,
+      "loss": 1.0946,
+      "step": 12869
+    },
+    {
+      "epoch": 2.0592,
+      "grad_norm": 1.6583896559850158,
+      "learning_rate": 4.754740838288284e-05,
+      "loss": 1.1475,
+      "step": 12870
+    },
+    {
+      "epoch": 2.05936,
+      "grad_norm": 1.7529021954057475,
+      "learning_rate": 4.753270231639487e-05,
+      "loss": 1.1112,
+      "step": 12871
+    },
+    {
+      "epoch": 2.05952,
+      "grad_norm": 1.6579815903466288,
+      "learning_rate": 4.7517997815454705e-05,
+      "loss": 1.1265,
+      "step": 12872
+    },
+    {
+      "epoch": 2.05968,
+      "grad_norm": 1.8184645746670864,
+      "learning_rate": 4.7503294880500927e-05,
+      "loss": 1.1063,
+      "step": 12873
+    },
+    {
+      "epoch": 2.05984,
+      "grad_norm": 2.196540637508271,
+      "learning_rate": 4.7488593511972415e-05,
+      "loss": 1.1733,
+      "step": 12874
+    },
+    {
+      "epoch": 2.06,
+      "grad_norm": 1.819053088490086,
+      "learning_rate": 4.747389371030767e-05,
+      "loss": 1.1153,
+      "step": 12875
+    },
+    {
+      "epoch": 2.06016,
+      "grad_norm": 1.9732059922321552,
+      "learning_rate": 4.745919547594547e-05,
+      "loss": 1.1243,
+      "step": 12876
+    },
+    {
+      "epoch": 2.06032,
+      "grad_norm": 1.866778854586452,
+      "learning_rate": 4.744449880932426e-05,
+      "loss": 1.1718,
+      "step": 12877
+    },
+    {
+      "epoch": 2.06048,
+      "grad_norm": 1.9838430649237397,
+      "learning_rate": 4.7429803710882646e-05,
+      "loss": 1.1401,
+      "step": 12878
+    },
+    {
+      "epoch": 2.06064,
+      "grad_norm": 1.9443027901980838,
+      "learning_rate": 4.7415110181059086e-05,
+      "loss": 1.1231,
+      "step": 12879
+    },
+    {
+      "epoch": 2.0608,
+      "grad_norm": 1.6575608870468943,
+      "learning_rate": 4.740041822029202e-05,
+      "loss": 1.1347,
+      "step": 12880
+    },
+    {
+      "epoch": 2.06096,
+      "grad_norm": 1.656420564229768,
+      "learning_rate": 4.738572782901983e-05,
+      "loss": 1.1677,
+      "step": 12881
+    },
+    {
+      "epoch": 2.06112,
+      "grad_norm": 2.077573126750962,
+      "learning_rate": 4.737103900768085e-05,
+      "loss": 1.1316,
+      "step": 12882
+    },
+    {
+      "epoch": 2.06128,
+      "grad_norm": 1.8846496172488165,
+      "learning_rate": 4.735635175671338e-05,
+      "loss": 1.1078,
+      "step": 12883
+    },
+    {
+      "epoch": 2.06144,
+      "grad_norm": 1.8638572522552241,
+      "learning_rate": 4.7341666076555666e-05,
+      "loss": 1.0872,
+      "step": 12884
+    },
+    {
+      "epoch": 2.0616,
+      "grad_norm": 1.8587249313719456,
+      "learning_rate": 4.73269819676459e-05,
+      "loss": 1.1071,
+      "step": 12885
+    },
+    {
+      "epoch": 2.06176,
+      "grad_norm": 1.9712390706828111,
+      "learning_rate": 4.731229943042225e-05,
+      "loss": 1.1355,
+      "step": 12886
+    },
+    {
+      "epoch": 2.06192,
+      "grad_norm": 1.6982940710940222,
+      "learning_rate": 4.7297618465322756e-05,
+      "loss": 1.0869,
+      "step": 12887
+    },
+    {
+      "epoch": 2.06208,
+      "grad_norm": 1.8442439928795178,
+      "learning_rate": 4.728293907278562e-05,
+      "loss": 1.1544,
+      "step": 12888
+    },
+    {
+      "epoch": 2.06224,
+      "grad_norm": 2.0883366249250837,
+      "learning_rate": 4.726826125324868e-05,
+      "loss": 1.1186,
+      "step": 12889
+    },
+    {
+      "epoch": 2.0624,
+      "grad_norm": 2.0541043052492256,
+      "learning_rate": 4.725358500715007e-05,
+      "loss": 1.1473,
+      "step": 12890
+    },
+    {
+      "epoch": 2.06256,
+      "grad_norm": 1.7573380962348146,
+      "learning_rate": 4.723891033492755e-05,
+      "loss": 1.0951,
+      "step": 12891
+    },
+    {
+      "epoch": 2.06272,
+      "grad_norm": 1.880706314318855,
+      "learning_rate": 4.72242372370191e-05,
+      "loss": 1.1275,
+      "step": 12892
+    },
+    {
+      "epoch": 2.06288,
+      "grad_norm": 2.1917082270433426,
+      "learning_rate": 4.72095657138625e-05,
+      "loss": 1.1337,
+      "step": 12893
+    },
+    {
+      "epoch": 2.06304,
+      "grad_norm": 1.9888871257180007,
+      "learning_rate": 4.719489576589554e-05,
+      "loss": 1.119,
+      "step": 12894
+    },
+    {
+      "epoch": 2.0632,
+      "grad_norm": 2.0059861840883495,
+      "learning_rate": 4.7180227393555956e-05,
+      "loss": 1.1339,
+      "step": 12895
+    },
+    {
+      "epoch": 2.06336,
+      "grad_norm": 2.030904607224169,
+      "learning_rate": 4.71655605972814e-05,
+      "loss": 1.1449,
+      "step": 12896
+    },
+    {
+      "epoch": 2.06352,
+      "grad_norm": 2.030904607224169,
+      "learning_rate": 4.71655605972814e-05,
+      "loss": 1.1346,
+      "step": 12897
+    },
+    {
+      "epoch": 2.06368,
+      "grad_norm": 2.1546715363811764,
+      "learning_rate": 4.715089537750954e-05,
+      "loss": 1.1169,
+      "step": 12898
+    },
+    {
+      "epoch": 2.06384,
+      "grad_norm": 2.688735704240298,
+      "learning_rate": 4.7136231734677946e-05,
+      "loss": 1.1886,
+      "step": 12899
+    },
+    {
+      "epoch": 2.064,
+      "grad_norm": 1.926509754385764,
+      "learning_rate": 4.712156966922417e-05,
+      "loss": 1.1097,
+      "step": 12900
+    },
+    {
+      "epoch": 2.06416,
+      "grad_norm": 1.6681307418437996,
+      "learning_rate": 4.710690918158571e-05,
+      "loss": 1.1527,
+      "step": 12901
+    },
+    {
+      "epoch": 2.06432,
+      "grad_norm": 2.33513511508292,
+      "learning_rate": 4.70922502722e-05,
+      "loss": 1.0962,
+      "step": 12902
+    },
+    {
+      "epoch": 2.06448,
+      "grad_norm": 2.826014932549772,
+      "learning_rate": 4.707759294150446e-05,
+      "loss": 1.1961,
+      "step": 12903
+    },
+    {
+      "epoch": 2.06464,
+      "grad_norm": 1.8223250827136623,
+      "learning_rate": 4.706293718993637e-05,
+      "loss": 1.1104,
+      "step": 12904
+    },
+    {
+      "epoch": 2.0648,
+      "grad_norm": 1.8730763259268859,
+      "learning_rate": 4.7048283017933193e-05,
+      "loss": 1.1442,
+      "step": 12905
+    },
+    {
+      "epoch": 2.06496,
+      "grad_norm": 2.30219123992797,
+      "learning_rate": 4.703363042593201e-05,
+      "loss": 1.1313,
+      "step": 12906
+    },
+    {
+      "epoch": 2.06512,
+      "grad_norm": 2.443666123889598,
+      "learning_rate": 4.701897941437018e-05,
+      "loss": 1.1503,
+      "step": 12907
+    },
+    {
+      "epoch": 2.06528,
+      "grad_norm": 1.9328298034252767,
+      "learning_rate": 4.700432998368474e-05,
+      "loss": 1.1731,
+      "step": 12908
+    },
+    {
+      "epoch": 2.06544,
+      "grad_norm": 1.4357718604535794,
+      "learning_rate": 4.6989682134312896e-05,
+      "loss": 1.1112,
+      "step": 12909
+    },
+    {
+      "epoch": 2.0656,
+      "grad_norm": 2.151870938770897,
+      "learning_rate": 4.69750358666917e-05,
+      "loss": 1.1495,
+      "step": 12910
+    },
+    {
+      "epoch": 2.06576,
+      "grad_norm": 2.587851388449393,
+      "learning_rate": 4.6960391181258167e-05,
+      "loss": 1.0917,
+      "step": 12911
+    },
+    {
+      "epoch": 2.06592,
+      "grad_norm": 2.0667527103330903,
+      "learning_rate": 4.694574807844928e-05,
+      "loss": 1.1177,
+      "step": 12912
+    },
+    {
+      "epoch": 2.06608,
+      "grad_norm": 1.9978147178955028,
+      "learning_rate": 4.6931106558701955e-05,
+      "loss": 1.2003,
+      "step": 12913
+    },
+    {
+      "epoch": 2.06624,
+      "grad_norm": 2.239193495766661,
+      "learning_rate": 4.691646662245308e-05,
+      "loss": 1.071,
+      "step": 12914
+    },
+    {
+      "epoch": 2.0664,
+      "grad_norm": 1.826315413575402,
+      "learning_rate": 4.690182827013949e-05,
+      "loss": 1.1806,
+      "step": 12915
+    },
+    {
+      "epoch": 2.06656,
+      "grad_norm": 1.7204310656525803,
+      "learning_rate": 4.6887191502197983e-05,
+      "loss": 1.084,
+      "step": 12916
+    },
+    {
+      "epoch": 2.06672,
+      "grad_norm": 1.8457806730898807,
+      "learning_rate": 4.687255631906527e-05,
+      "loss": 1.1344,
+      "step": 12917
+    },
+    {
+      "epoch": 2.06688,
+      "grad_norm": 1.9057511620822611,
+      "learning_rate": 4.6857922721178086e-05,
+      "loss": 1.0788,
+      "step": 12918
+    },
+    {
+      "epoch": 2.06704,
+      "grad_norm": 1.8229102205442091,
+      "learning_rate": 4.684329070897303e-05,
+      "loss": 1.1663,
+      "step": 12919
+    },
+    {
+      "epoch": 2.0672,
+      "grad_norm": 1.857209047211745,
+      "learning_rate": 4.6828660282886695e-05,
+      "loss": 1.1195,
+      "step": 12920
+    },
+    {
+      "epoch": 2.06736,
+      "grad_norm": 1.8256930750883529,
+      "learning_rate": 4.681403144335569e-05,
+      "loss": 1.1484,
+      "step": 12921
+    },
+    {
+      "epoch": 2.06752,
+      "grad_norm": 1.9329269984727044,
+      "learning_rate": 4.679940419081649e-05,
+      "loss": 1.1118,
+      "step": 12922
+    },
+    {
+      "epoch": 2.06768,
+      "grad_norm": 2.3067265679342968,
+      "learning_rate": 4.678477852570555e-05,
+      "loss": 1.1539,
+      "step": 12923
+    },
+    {
+      "epoch": 2.06784,
+      "grad_norm": 2.162828945815701,
+      "learning_rate": 4.677015444845927e-05,
+      "loss": 1.1551,
+      "step": 12924
+    },
+    {
+      "epoch": 2.068,
+      "grad_norm": 1.7510607381917356,
+      "learning_rate": 4.675553195951402e-05,
+      "loss": 1.1393,
+      "step": 12925
+    },
+    {
+      "epoch": 2.0681599999999998,
+      "grad_norm": 1.7515726650425312,
+      "learning_rate": 4.674091105930611e-05,
+      "loss": 1.1214,
+      "step": 12926
+    },
+    {
+      "epoch": 2.06832,
+      "grad_norm": 1.6875055835369215,
+      "learning_rate": 4.672629174827181e-05,
+      "loss": 1.1193,
+      "step": 12927
+    },
+    {
+      "epoch": 2.06848,
+      "grad_norm": 1.7752560621567455,
+      "learning_rate": 4.671167402684733e-05,
+      "loss": 1.1165,
+      "step": 12928
+    },
+    {
+      "epoch": 2.06864,
+      "grad_norm": 2.04021658903627,
+      "learning_rate": 4.669705789546882e-05,
+      "loss": 1.1176,
+      "step": 12929
+    },
+    {
+      "epoch": 2.0688,
+      "grad_norm": 1.9229394456181486,
+      "learning_rate": 4.66824433545725e-05,
+      "loss": 1.1133,
+      "step": 12930
+    },
+    {
+      "epoch": 2.06896,
+      "grad_norm": 2.2636222074513137,
+      "learning_rate": 4.66678304045943e-05,
+      "loss": 1.1714,
+      "step": 12931
+    },
+    {
+      "epoch": 2.06912,
+      "grad_norm": 1.7458959956088838,
+      "learning_rate": 4.665321904597041e-05,
+      "loss": 1.1137,
+      "step": 12932
+    },
+    {
+      "epoch": 2.06928,
+      "grad_norm": 1.6555427939280107,
+      "learning_rate": 4.663860927913666e-05,
+      "loss": 1.1002,
+      "step": 12933
+    },
+    {
+      "epoch": 2.06944,
+      "grad_norm": 1.9510592929496837,
+      "learning_rate": 4.662400110452911e-05,
+      "loss": 1.1209,
+      "step": 12934
+    },
+    {
+      "epoch": 2.0696,
+      "grad_norm": 1.795001229185224,
+      "learning_rate": 4.660939452258353e-05,
+      "loss": 1.1304,
+      "step": 12935
+    },
+    {
+      "epoch": 2.06976,
+      "grad_norm": 2.0865165913336106,
+      "learning_rate": 4.659478953373585e-05,
+      "loss": 1.1766,
+      "step": 12936
+    },
+    {
+      "epoch": 2.06992,
+      "grad_norm": 2.082747774156536,
+      "learning_rate": 4.658018613842183e-05,
+      "loss": 1.0893,
+      "step": 12937
+    },
+    {
+      "epoch": 2.07008,
+      "grad_norm": 1.7618729138072835,
+      "learning_rate": 4.656558433707722e-05,
+      "loss": 1.1365,
+      "step": 12938
+    },
+    {
+      "epoch": 2.07024,
+      "grad_norm": 1.9417396806691776,
+      "learning_rate": 4.6550984130137707e-05,
+      "loss": 1.1076,
+      "step": 12939
+    },
+    {
+      "epoch": 2.0704,
+      "grad_norm": 2.3862405903338897,
+      "learning_rate": 4.653638551803895e-05,
+      "loss": 1.1695,
+      "step": 12940
+    },
+    {
+      "epoch": 2.07056,
+      "grad_norm": 1.9613525447077171,
+      "learning_rate": 4.652178850121653e-05,
+      "loss": 1.1171,
+      "step": 12941
+    },
+    {
+      "epoch": 2.07072,
+      "grad_norm": 1.8864923344572757,
+      "learning_rate": 4.6507193080106035e-05,
+      "loss": 1.1596,
+      "step": 12942
+    },
+    {
+      "epoch": 2.07088,
+      "grad_norm": 2.4794462500693504,
+      "learning_rate": 4.6492599255142946e-05,
+      "loss": 1.1724,
+      "step": 12943
+    },
+    {
+      "epoch": 2.07104,
+      "grad_norm": 2.591905063751335,
+      "learning_rate": 4.6478007026762714e-05,
+      "loss": 1.1148,
+      "step": 12944
+    },
+    {
+      "epoch": 2.0712,
+      "grad_norm": 1.9877912238016626,
+      "learning_rate": 4.6463416395400736e-05,
+      "loss": 1.1444,
+      "step": 12945
+    },
+    {
+      "epoch": 2.07136,
+      "grad_norm": 2.4168145020994927,
+      "learning_rate": 4.6448827361492475e-05,
+      "loss": 1.1038,
+      "step": 12946
+    },
+    {
+      "epoch": 2.07152,
+      "grad_norm": 2.7356794167990968,
+      "learning_rate": 4.643423992547309e-05,
+      "loss": 1.1502,
+      "step": 12947
+    },
+    {
+      "epoch": 2.07168,
+      "grad_norm": 1.7534364651015746,
+      "learning_rate": 4.641965408777802e-05,
+      "loss": 1.1566,
+      "step": 12948
+    },
+    {
+      "epoch": 2.07184,
+      "grad_norm": 1.8401297667965888,
+      "learning_rate": 4.640506984884232e-05,
+      "loss": 1.187,
+      "step": 12949
+    },
+    {
+      "epoch": 2.072,
+      "grad_norm": 2.2965672863999393,
+      "learning_rate": 4.6390487209101305e-05,
+      "loss": 1.1147,
+      "step": 12950
+    },
+    {
+      "epoch": 2.0721600000000002,
+      "grad_norm": 1.6567690166981341,
+      "learning_rate": 4.637590616898997e-05,
+      "loss": 1.1299,
+      "step": 12951
+    },
+    {
+      "epoch": 2.07232,
+      "grad_norm": 1.539308139466018,
+      "learning_rate": 4.636132672894349e-05,
+      "loss": 1.1222,
+      "step": 12952
+    },
+    {
+      "epoch": 2.07248,
+      "grad_norm": 2.0290615053182957,
+      "learning_rate": 4.634674888939685e-05,
+      "loss": 1.1391,
+      "step": 12953
+    },
+    {
+      "epoch": 2.07264,
+      "grad_norm": 2.1749470475176382,
+      "learning_rate": 4.6332172650785056e-05,
+      "loss": 1.1891,
+      "step": 12954
+    },
+    {
+      "epoch": 2.0728,
+      "grad_norm": 1.7577279612913566,
+      "learning_rate": 4.631759801354302e-05,
+      "loss": 1.1112,
+      "step": 12955
+    },
+    {
+      "epoch": 2.07296,
+      "grad_norm": 1.9461946568756705,
+      "learning_rate": 4.6303024978105644e-05,
+      "loss": 1.1244,
+      "step": 12956
+    },
+    {
+      "epoch": 2.07312,
+      "grad_norm": 1.8342658669757217,
+      "learning_rate": 4.628845354490775e-05,
+      "loss": 1.1432,
+      "step": 12957
+    },
+    {
+      "epoch": 2.07328,
+      "grad_norm": 2.0429983343282805,
+      "learning_rate": 4.6273883714384136e-05,
+      "loss": 1.1382,
+      "step": 12958
+    },
+    {
+      "epoch": 2.07344,
+      "grad_norm": 1.8633872026626694,
+      "learning_rate": 4.6259315486969555e-05,
+      "loss": 1.0952,
+      "step": 12959
+    },
+    {
+      "epoch": 2.0736,
+      "grad_norm": 2.072136806300078,
+      "learning_rate": 4.624474886309869e-05,
+      "loss": 1.1536,
+      "step": 12960
+    },
+    {
+      "epoch": 2.07376,
+      "grad_norm": 2.0351848670284087,
+      "learning_rate": 4.6230183843206144e-05,
+      "loss": 1.0829,
+      "step": 12961
+    },
+    {
+      "epoch": 2.07392,
+      "grad_norm": 1.9713931249477217,
+      "learning_rate": 4.6215620427726634e-05,
+      "loss": 1.169,
+      "step": 12962
+    },
+    {
+      "epoch": 2.07408,
+      "grad_norm": 2.015181260742333,
+      "learning_rate": 4.620105861709457e-05,
+      "loss": 1.1322,
+      "step": 12963
+    },
+    {
+      "epoch": 2.07424,
+      "grad_norm": 1.9271332213028918,
+      "learning_rate": 4.6186498411744596e-05,
+      "loss": 1.151,
+      "step": 12964
+    },
+    {
+      "epoch": 2.0744,
+      "grad_norm": 1.682436151864785,
+      "learning_rate": 4.617193981211102e-05,
+      "loss": 1.1105,
+      "step": 12965
+    },
+    {
+      "epoch": 2.07456,
+      "grad_norm": 1.8070144467698306,
+      "learning_rate": 4.615738281862839e-05,
+      "loss": 1.1436,
+      "step": 12966
+    },
+    {
+      "epoch": 2.07472,
+      "grad_norm": 1.7634332102492571,
+      "learning_rate": 4.614282743173094e-05,
+      "loss": 1.1138,
+      "step": 12967
+    },
+    {
+      "epoch": 2.07488,
+      "grad_norm": 1.848998981528906,
+      "learning_rate": 4.612827365185307e-05,
+      "loss": 1.1038,
+      "step": 12968
+    },
+    {
+      "epoch": 2.07504,
+      "grad_norm": 2.0655341691012303,
+      "learning_rate": 4.611372147942902e-05,
+      "loss": 1.1371,
+      "step": 12969
+    },
+    {
+      "epoch": 2.0752,
+      "grad_norm": 1.8331872897400379,
+      "learning_rate": 4.6099170914893e-05,
+      "loss": 1.1266,
+      "step": 12970
+    },
+    {
+      "epoch": 2.07536,
+      "grad_norm": 2.034497431441384,
+      "learning_rate": 4.6084621958679174e-05,
+      "loss": 1.1478,
+      "step": 12971
+    },
+    {
+      "epoch": 2.07552,
+      "grad_norm": 1.7236642111641622,
+      "learning_rate": 4.607007461122167e-05,
+      "loss": 1.112,
+      "step": 12972
+    },
+    {
+      "epoch": 2.07568,
+      "grad_norm": 2.2333954474141833,
+      "learning_rate": 4.605552887295456e-05,
+      "loss": 1.0874,
+      "step": 12973
+    },
+    {
+      "epoch": 2.07584,
+      "grad_norm": 1.8529813623974425,
+      "learning_rate": 4.604098474431186e-05,
+      "loss": 1.0995,
+      "step": 12974
+    },
+    {
+      "epoch": 2.076,
+      "grad_norm": 1.8002206067396453,
+      "learning_rate": 4.6026442225727537e-05,
+      "loss": 1.091,
+      "step": 12975
+    },
+    {
+      "epoch": 2.07616,
+      "grad_norm": 1.9016710865396502,
+      "learning_rate": 4.601190131763554e-05,
+      "loss": 1.1228,
+      "step": 12976
+    },
+    {
+      "epoch": 2.07632,
+      "grad_norm": 1.8293444715763265,
+      "learning_rate": 4.5997362020469693e-05,
+      "loss": 1.1855,
+      "step": 12977
+    },
+    {
+      "epoch": 2.07648,
+      "grad_norm": 1.7380003400202106,
+      "learning_rate": 4.598282433466395e-05,
+      "loss": 1.1422,
+      "step": 12978
+    },
+    {
+      "epoch": 2.07664,
+      "grad_norm": 1.8021744973391964,
+      "learning_rate": 4.596828826065196e-05,
+      "loss": 1.1541,
+      "step": 12979
+    },
+    {
+      "epoch": 2.0768,
+      "grad_norm": 1.5993440805366663,
+      "learning_rate": 4.5953753798867535e-05,
+      "loss": 1.1216,
+      "step": 12980
+    },
+    {
+      "epoch": 2.07696,
+      "grad_norm": 2.0733993539795983,
+      "learning_rate": 4.593922094974434e-05,
+      "loss": 1.1454,
+      "step": 12981
+    },
+    {
+      "epoch": 2.07712,
+      "grad_norm": 1.651991550424121,
+      "learning_rate": 4.592468971371603e-05,
+      "loss": 1.1597,
+      "step": 12982
+    },
+    {
+      "epoch": 2.07728,
+      "grad_norm": 1.8918534499592032,
+      "learning_rate": 4.591016009121618e-05,
+      "loss": 1.1397,
+      "step": 12983
+    },
+    {
+      "epoch": 2.07744,
+      "grad_norm": 1.8064634837639104,
+      "learning_rate": 4.589563208267833e-05,
+      "loss": 1.1042,
+      "step": 12984
+    },
+    {
+      "epoch": 2.0776,
+      "grad_norm": 2.020240358624395,
+      "learning_rate": 4.5881105688535983e-05,
+      "loss": 1.1301,
+      "step": 12985
+    },
+    {
+      "epoch": 2.07776,
+      "grad_norm": 1.6397716263631854,
+      "learning_rate": 4.586658090922256e-05,
+      "loss": 1.0929,
+      "step": 12986
+    },
+    {
+      "epoch": 2.07792,
+      "grad_norm": 1.6661413758701396,
+      "learning_rate": 4.585205774517154e-05,
+      "loss": 1.098,
+      "step": 12987
+    },
+    {
+      "epoch": 2.07808,
+      "grad_norm": 1.8643679251957375,
+      "learning_rate": 4.583753619681614e-05,
+      "loss": 1.0918,
+      "step": 12988
+    },
+    {
+      "epoch": 2.07824,
+      "grad_norm": 1.7088039497688698,
+      "learning_rate": 4.582301626458981e-05,
+      "loss": 1.1645,
+      "step": 12989
+    },
+    {
+      "epoch": 2.0784,
+      "grad_norm": 1.9871403196498727,
+      "learning_rate": 4.580849794892566e-05,
+      "loss": 1.2208,
+      "step": 12990
+    },
+    {
+      "epoch": 2.07856,
+      "grad_norm": 1.7086352644417762,
+      "learning_rate": 4.579398125025705e-05,
+      "loss": 1.1045,
+      "step": 12991
+    },
+    {
+      "epoch": 2.07872,
+      "grad_norm": 1.9660548958452235,
+      "learning_rate": 4.577946616901697e-05,
+      "loss": 1.1159,
+      "step": 12992
+    },
+    {
+      "epoch": 2.07888,
+      "grad_norm": 1.827093317974206,
+      "learning_rate": 4.576495270563864e-05,
+      "loss": 1.1514,
+      "step": 12993
+    },
+    {
+      "epoch": 2.07904,
+      "grad_norm": 2.0154819669956656,
+      "learning_rate": 4.575044086055511e-05,
+      "loss": 1.1394,
+      "step": 12994
+    },
+    {
+      "epoch": 2.0792,
+      "grad_norm": 2.0163342386309706,
+      "learning_rate": 4.5735930634199364e-05,
+      "loss": 1.1659,
+      "step": 12995
+    },
+    {
+      "epoch": 2.07936,
+      "grad_norm": 1.9468155797765574,
+      "learning_rate": 4.572142202700438e-05,
+      "loss": 1.1812,
+      "step": 12996
+    },
+    {
+      "epoch": 2.07952,
+      "grad_norm": 1.8704579188439951,
+      "learning_rate": 4.570691503940307e-05,
+      "loss": 1.1179,
+      "step": 12997
+    },
+    {
+      "epoch": 2.07968,
+      "grad_norm": 1.7031725402609093,
+      "learning_rate": 4.5692409671828296e-05,
+      "loss": 1.1492,
+      "step": 12998
+    },
+    {
+      "epoch": 2.07984,
+      "grad_norm": 1.6964597824579195,
+      "learning_rate": 4.567790592471288e-05,
+      "loss": 1.1714,
+      "step": 12999
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 2.1565859291632217,
+      "learning_rate": 4.566340379848961e-05,
+      "loss": 1.082,
+      "step": 13000
+    },
+    {
+      "epoch": 2.08016,
+      "grad_norm": 1.9563354876283259,
+      "learning_rate": 4.564890329359119e-05,
+      "loss": 1.184,
+      "step": 13001
+    },
+    {
+      "epoch": 2.08032,
+      "grad_norm": 2.265733800534703,
+      "learning_rate": 4.563440441045026e-05,
+      "loss": 1.1871,
+      "step": 13002
+    },
+    {
+      "epoch": 2.08048,
+      "grad_norm": 1.9606356283651125,
+      "learning_rate": 4.561990714949955e-05,
+      "loss": 1.111,
+      "step": 13003
+    },
+    {
+      "epoch": 2.08064,
+      "grad_norm": 1.88501466545856,
+      "learning_rate": 4.56054115111715e-05,
+      "loss": 1.1006,
+      "step": 13004
+    },
+    {
+      "epoch": 2.0808,
+      "grad_norm": 1.8051892680564166,
+      "learning_rate": 4.559091749589879e-05,
+      "loss": 1.1182,
+      "step": 13005
+    },
+    {
+      "epoch": 2.08096,
+      "grad_norm": 1.7036284868754064,
+      "learning_rate": 4.557642510411374e-05,
+      "loss": 1.1356,
+      "step": 13006
+    },
+    {
+      "epoch": 2.08112,
+      "grad_norm": 1.9171591292807453,
+      "learning_rate": 4.5561934336248934e-05,
+      "loss": 1.1694,
+      "step": 13007
+    },
+    {
+      "epoch": 2.08128,
+      "grad_norm": 1.899434849919824,
+      "learning_rate": 4.5547445192736615e-05,
+      "loss": 1.1445,
+      "step": 13008
+    },
+    {
+      "epoch": 2.08144,
+      "grad_norm": 1.8934105475851042,
+      "learning_rate": 4.553295767400921e-05,
+      "loss": 1.1326,
+      "step": 13009
+    },
+    {
+      "epoch": 2.0816,
+      "grad_norm": 2.012642401432302,
+      "learning_rate": 4.551847178049899e-05,
+      "loss": 1.1107,
+      "step": 13010
+    },
+    {
+      "epoch": 2.08176,
+      "grad_norm": 1.6693222386009297,
+      "learning_rate": 4.550398751263819e-05,
+      "loss": 1.1284,
+      "step": 13011
+    },
+    {
+      "epoch": 2.08192,
+      "grad_norm": 2.057930768026357,
+      "learning_rate": 4.548950487085899e-05,
+      "loss": 1.158,
+      "step": 13012
+    },
+    {
+      "epoch": 2.08208,
+      "grad_norm": 2.1239638688514773,
+      "learning_rate": 4.547502385559353e-05,
+      "loss": 1.0984,
+      "step": 13013
+    },
+    {
+      "epoch": 2.08224,
+      "grad_norm": 1.7566695945476511,
+      "learning_rate": 4.54605444672739e-05,
+      "loss": 1.0899,
+      "step": 13014
+    },
+    {
+      "epoch": 2.0824,
+      "grad_norm": 1.9454243267914282,
+      "learning_rate": 4.544606670633217e-05,
+      "loss": 1.1394,
+      "step": 13015
+    },
+    {
+      "epoch": 2.08256,
+      "grad_norm": 1.9185156900694382,
+      "learning_rate": 4.54315905732003e-05,
+      "loss": 1.1232,
+      "step": 13016
+    },
+    {
+      "epoch": 2.08272,
+      "grad_norm": 2.1746921148181104,
+      "learning_rate": 4.541711606831026e-05,
+      "loss": 1.1767,
+      "step": 13017
+    },
+    {
+      "epoch": 2.08288,
+      "grad_norm": 1.6720858417657107,
+      "learning_rate": 4.5402643192093895e-05,
+      "loss": 1.1095,
+      "step": 13018
+    },
+    {
+      "epoch": 2.08304,
+      "grad_norm": 1.8683549591018649,
+      "learning_rate": 4.538817194498318e-05,
+      "loss": 1.1121,
+      "step": 13019
+    },
+    {
+      "epoch": 2.0832,
+      "grad_norm": 1.902779130428826,
+      "learning_rate": 4.537370232740977e-05,
+      "loss": 1.0905,
+      "step": 13020
+    },
+    {
+      "epoch": 2.08336,
+      "grad_norm": 1.873572133328256,
+      "learning_rate": 4.535923433980555e-05,
+      "loss": 1.1864,
+      "step": 13021
+    },
+    {
+      "epoch": 2.08352,
+      "grad_norm": 1.7505000483810351,
+      "learning_rate": 4.5344767982602085e-05,
+      "loss": 1.1771,
+      "step": 13022
+    },
+    {
+      "epoch": 2.08368,
+      "grad_norm": 1.8078947790811433,
+      "learning_rate": 4.533030325623118e-05,
+      "loss": 1.1112,
+      "step": 13023
+    },
+    {
+      "epoch": 2.08384,
+      "grad_norm": 1.7783847389132559,
+      "learning_rate": 4.531584016112428e-05,
+      "loss": 1.0777,
+      "step": 13024
+    },
+    {
+      "epoch": 2.084,
+      "grad_norm": 2.1412356280728497,
+      "learning_rate": 4.530137869771307e-05,
+      "loss": 1.1616,
+      "step": 13025
+    },
+    {
+      "epoch": 2.08416,
+      "grad_norm": 1.719212103646325,
+      "learning_rate": 4.528691886642902e-05,
+      "loss": 1.0842,
+      "step": 13026
+    },
+    {
+      "epoch": 2.08432,
+      "grad_norm": 2.0685523180837286,
+      "learning_rate": 4.527246066770358e-05,
+      "loss": 1.12,
+      "step": 13027
+    },
+    {
+      "epoch": 2.08448,
+      "grad_norm": 1.8564605288591187,
+      "learning_rate": 4.525800410196817e-05,
+      "loss": 1.1593,
+      "step": 13028
+    },
+    {
+      "epoch": 2.08464,
+      "grad_norm": 1.947545747010916,
+      "learning_rate": 4.524354916965414e-05,
+      "loss": 1.0905,
+      "step": 13029
+    },
+    {
+      "epoch": 2.0848,
+      "grad_norm": 1.862697408679243,
+      "learning_rate": 4.522909587119282e-05,
+      "loss": 1.1364,
+      "step": 13030
+    },
+    {
+      "epoch": 2.08496,
+      "grad_norm": 1.7389912024272371,
+      "learning_rate": 4.521464420701547e-05,
+      "loss": 1.14,
+      "step": 13031
+    },
+    {
+      "epoch": 2.08512,
+      "grad_norm": 1.7987991128564145,
+      "learning_rate": 4.520019417755331e-05,
+      "loss": 1.1262,
+      "step": 13032
+    },
+    {
+      "epoch": 2.08528,
+      "grad_norm": 2.1049773719469935,
+      "learning_rate": 4.51857457832375e-05,
+      "loss": 1.1278,
+      "step": 13033
+    },
+    {
+      "epoch": 2.08544,
+      "grad_norm": 1.93298570124944,
+      "learning_rate": 4.517129902449912e-05,
+      "loss": 1.1428,
+      "step": 13034
+    },
+    {
+      "epoch": 2.0856,
+      "grad_norm": 1.9579432449492087,
+      "learning_rate": 4.515685390176937e-05,
+      "loss": 1.1562,
+      "step": 13035
+    },
+    {
+      "epoch": 2.08576,
+      "grad_norm": 2.09443134042586,
+      "learning_rate": 4.514241041547911e-05,
+      "loss": 1.1174,
+      "step": 13036
+    },
+    {
+      "epoch": 2.0859199999999998,
+      "grad_norm": 2.3148455850543916,
+      "learning_rate": 4.512796856605945e-05,
+      "loss": 1.0979,
+      "step": 13037
+    },
+    {
+      "epoch": 2.08608,
+      "grad_norm": 1.8744899461056967,
+      "learning_rate": 4.511352835394118e-05,
+      "loss": 1.1194,
+      "step": 13038
+    },
+    {
+      "epoch": 2.08624,
+      "grad_norm": 1.7559091326038376,
+      "learning_rate": 4.5099089779555294e-05,
+      "loss": 1.0729,
+      "step": 13039
+    },
+    {
+      "epoch": 2.0864,
+      "grad_norm": 2.181042744059968,
+      "learning_rate": 4.508465284333255e-05,
+      "loss": 1.1659,
+      "step": 13040
+    },
+    {
+      "epoch": 2.08656,
+      "grad_norm": 2.047559572310379,
+      "learning_rate": 4.507021754570376e-05,
+      "loss": 1.0901,
+      "step": 13041
+    },
+    {
+      "epoch": 2.08672,
+      "grad_norm": 1.808897537760957,
+      "learning_rate": 4.5055783887099635e-05,
+      "loss": 1.1003,
+      "step": 13042
+    },
+    {
+      "epoch": 2.08688,
+      "grad_norm": 1.8477421377770789,
+      "learning_rate": 4.5041351867950824e-05,
+      "loss": 1.1392,
+      "step": 13043
+    },
+    {
+      "epoch": 2.08704,
+      "grad_norm": 1.7441930722106205,
+      "learning_rate": 4.5026921488688054e-05,
+      "loss": 1.113,
+      "step": 13044
+    },
+    {
+      "epoch": 2.0872,
+      "grad_norm": 1.8251916297089772,
+      "learning_rate": 4.5012492749741775e-05,
+      "loss": 1.1275,
+      "step": 13045
+    },
+    {
+      "epoch": 2.08736,
+      "grad_norm": 2.050389209979113,
+      "learning_rate": 4.4998065651542666e-05,
+      "loss": 1.0813,
+      "step": 13046
+    },
+    {
+      "epoch": 2.08752,
+      "grad_norm": 2.0568286910818365,
+      "learning_rate": 4.4983640194521045e-05,
+      "loss": 1.1033,
+      "step": 13047
+    },
+    {
+      "epoch": 2.08768,
+      "grad_norm": 1.7004450684295191,
+      "learning_rate": 4.496921637910751e-05,
+      "loss": 1.0908,
+      "step": 13048
+    },
+    {
+      "epoch": 2.08784,
+      "grad_norm": 2.285420203286047,
+      "learning_rate": 4.49547942057323e-05,
+      "loss": 1.1425,
+      "step": 13049
+    },
+    {
+      "epoch": 2.088,
+      "grad_norm": 2.0229438950756666,
+      "learning_rate": 4.4940373674825864e-05,
+      "loss": 1.1307,
+      "step": 13050
+    },
+    {
+      "epoch": 2.08816,
+      "grad_norm": 1.7420617447165503,
+      "learning_rate": 4.4925954786818436e-05,
+      "loss": 1.112,
+      "step": 13051
+    },
+    {
+      "epoch": 2.08832,
+      "grad_norm": 1.8407204591501425,
+      "learning_rate": 4.4911537542140266e-05,
+      "loss": 1.1251,
+      "step": 13052
+    },
+    {
+      "epoch": 2.08848,
+      "grad_norm": 1.957536174694674,
+      "learning_rate": 4.489712194122155e-05,
+      "loss": 1.1272,
+      "step": 13053
+    },
+    {
+      "epoch": 2.08864,
+      "grad_norm": 1.7523008376348939,
+      "learning_rate": 4.4882707984492424e-05,
+      "loss": 1.1093,
+      "step": 13054
+    },
+    {
+      "epoch": 2.0888,
+      "grad_norm": 1.9046324428411248,
+      "learning_rate": 4.486829567238298e-05,
+      "loss": 1.1236,
+      "step": 13055
+    },
+    {
+      "epoch": 2.08896,
+      "grad_norm": 1.7634127013642966,
+      "learning_rate": 4.485388500532325e-05,
+      "loss": 1.1296,
+      "step": 13056
+    },
+    {
+      "epoch": 2.08912,
+      "grad_norm": 1.7653339853295118,
+      "learning_rate": 4.4839475983743226e-05,
+      "loss": 1.0709,
+      "step": 13057
+    },
+    {
+      "epoch": 2.08928,
+      "grad_norm": 1.707357078673208,
+      "learning_rate": 4.4825068608072875e-05,
+      "loss": 1.1395,
+      "step": 13058
+    },
+    {
+      "epoch": 2.08944,
+      "grad_norm": 1.7666241363052628,
+      "learning_rate": 4.4810662878742036e-05,
+      "loss": 1.1089,
+      "step": 13059
+    },
+    {
+      "epoch": 2.0896,
+      "grad_norm": 1.8149967589983567,
+      "learning_rate": 4.479625879618068e-05,
+      "loss": 1.1394,
+      "step": 13060
+    },
+    {
+      "epoch": 2.08976,
+      "grad_norm": 1.8118266703153656,
+      "learning_rate": 4.478185636081844e-05,
+      "loss": 1.178,
+      "step": 13061
+    },
+    {
+      "epoch": 2.08992,
+      "grad_norm": 1.9777564336009648,
+      "learning_rate": 4.476745557308522e-05,
+      "loss": 1.1315,
+      "step": 13062
+    },
+    {
+      "epoch": 2.09008,
+      "grad_norm": 1.9442778300438581,
+      "learning_rate": 4.475305643341057e-05,
+      "loss": 1.1264,
+      "step": 13063
+    },
+    {
+      "epoch": 2.09024,
+      "grad_norm": 2.1729293354586976,
+      "learning_rate": 4.473865894222429e-05,
+      "loss": 1.1251,
+      "step": 13064
+    },
+    {
+      "epoch": 2.0904,
+      "grad_norm": 1.8290041459393271,
+      "learning_rate": 4.4724263099955824e-05,
+      "loss": 1.0767,
+      "step": 13065
+    },
+    {
+      "epoch": 2.09056,
+      "grad_norm": 1.8448835458407025,
+      "learning_rate": 4.470986890703483e-05,
+      "loss": 1.1242,
+      "step": 13066
+    },
+    {
+      "epoch": 2.09072,
+      "grad_norm": 1.5847906533344642,
+      "learning_rate": 4.469547636389079e-05,
+      "loss": 1.1237,
+      "step": 13067
+    },
+    {
+      "epoch": 2.09088,
+      "grad_norm": 1.8545335884978889,
+      "learning_rate": 4.468108547095315e-05,
+      "loss": 1.0872,
+      "step": 13068
+    },
+    {
+      "epoch": 2.09104,
+      "grad_norm": 1.780008982260065,
+      "learning_rate": 4.466669622865132e-05,
+      "loss": 1.0565,
+      "step": 13069
+    },
+    {
+      "epoch": 2.0912,
+      "grad_norm": 2.1466075276051306,
+      "learning_rate": 4.465230863741463e-05,
+      "loss": 1.1408,
+      "step": 13070
+    },
+    {
+      "epoch": 2.09136,
+      "grad_norm": 1.9023997315920378,
+      "learning_rate": 4.463792269767241e-05,
+      "loss": 1.1246,
+      "step": 13071
+    },
+    {
+      "epoch": 2.09152,
+      "grad_norm": 1.9784855677505038,
+      "learning_rate": 4.4623538409853894e-05,
+      "loss": 1.1101,
+      "step": 13072
+    },
+    {
+      "epoch": 2.09168,
+      "grad_norm": 1.8361714975919248,
+      "learning_rate": 4.460915577438831e-05,
+      "loss": 1.0899,
+      "step": 13073
+    },
+    {
+      "epoch": 2.09184,
+      "grad_norm": 1.9548927270521574,
+      "learning_rate": 4.459477479170479e-05,
+      "loss": 1.196,
+      "step": 13074
+    },
+    {
+      "epoch": 2.092,
+      "grad_norm": 1.7098486286617154,
+      "learning_rate": 4.4580395462232436e-05,
+      "loss": 1.0863,
+      "step": 13075
+    },
+    {
+      "epoch": 2.09216,
+      "grad_norm": 2.844065376815792,
+      "learning_rate": 4.4566017786400404e-05,
+      "loss": 1.1434,
+      "step": 13076
+    },
+    {
+      "epoch": 2.09232,
+      "grad_norm": 1.8843303038023407,
+      "learning_rate": 4.455164176463754e-05,
+      "loss": 1.1301,
+      "step": 13077
+    },
+    {
+      "epoch": 2.09248,
+      "grad_norm": 2.101246178866824,
+      "learning_rate": 4.453726739737296e-05,
+      "loss": 1.1269,
+      "step": 13078
+    },
+    {
+      "epoch": 2.09264,
+      "grad_norm": 2.0354516599257817,
+      "learning_rate": 4.452289468503544e-05,
+      "loss": 1.128,
+      "step": 13079
+    },
+    {
+      "epoch": 2.0928,
+      "grad_norm": 1.9835933360260087,
+      "learning_rate": 4.450852362805398e-05,
+      "loss": 1.126,
+      "step": 13080
+    },
+    {
+      "epoch": 2.09296,
+      "grad_norm": 2.166477294997129,
+      "learning_rate": 4.4494154226857235e-05,
+      "loss": 1.1244,
+      "step": 13081
+    },
+    {
+      "epoch": 2.09312,
+      "grad_norm": 2.064371538487105,
+      "learning_rate": 4.447978648187408e-05,
+      "loss": 1.1662,
+      "step": 13082
+    },
+    {
+      "epoch": 2.09328,
+      "grad_norm": 1.7405661473631786,
+      "learning_rate": 4.44654203935332e-05,
+      "loss": 1.121,
+      "step": 13083
+    },
+    {
+      "epoch": 2.09344,
+      "grad_norm": 1.7546110241335118,
+      "learning_rate": 4.4451055962263244e-05,
+      "loss": 1.1241,
+      "step": 13084
+    },
+    {
+      "epoch": 2.0936,
+      "grad_norm": 1.6254915021357106,
+      "learning_rate": 4.443669318849284e-05,
+      "loss": 1.1049,
+      "step": 13085
+    },
+    {
+      "epoch": 2.09376,
+      "grad_norm": 1.9772210571007067,
+      "learning_rate": 4.4422332072650555e-05,
+      "loss": 1.1518,
+      "step": 13086
+    },
+    {
+      "epoch": 2.09392,
+      "grad_norm": 1.915475833940429,
+      "learning_rate": 4.4407972615164884e-05,
+      "loss": 1.1322,
+      "step": 13087
+    },
+    {
+      "epoch": 2.09408,
+      "grad_norm": 1.882010990921692,
+      "learning_rate": 4.4393614816464314e-05,
+      "loss": 1.1314,
+      "step": 13088
+    },
+    {
+      "epoch": 2.09424,
+      "grad_norm": 1.8878067247396368,
+      "learning_rate": 4.437925867697724e-05,
+      "loss": 1.106,
+      "step": 13089
+    },
+    {
+      "epoch": 2.0944,
+      "grad_norm": 1.9265494473670244,
+      "learning_rate": 4.436490419713205e-05,
+      "loss": 1.1164,
+      "step": 13090
+    },
+    {
+      "epoch": 2.09456,
+      "grad_norm": 1.6888603171970051,
+      "learning_rate": 4.435055137735701e-05,
+      "loss": 1.11,
+      "step": 13091
+    },
+    {
+      "epoch": 2.09472,
+      "grad_norm": 1.8086329141563167,
+      "learning_rate": 4.43362002180805e-05,
+      "loss": 1.1016,
+      "step": 13092
+    },
+    {
+      "epoch": 2.09488,
+      "grad_norm": 1.7552076907795953,
+      "learning_rate": 4.4321850719730593e-05,
+      "loss": 1.1212,
+      "step": 13093
+    },
+    {
+      "epoch": 2.09504,
+      "grad_norm": 1.8148147944403625,
+      "learning_rate": 4.43075028827356e-05,
+      "loss": 1.1207,
+      "step": 13094
+    },
+    {
+      "epoch": 2.0952,
+      "grad_norm": 1.7550035819407492,
+      "learning_rate": 4.4293156707523495e-05,
+      "loss": 1.1722,
+      "step": 13095
+    },
+    {
+      "epoch": 2.09536,
+      "grad_norm": 2.130180861602515,
+      "learning_rate": 4.427881219452246e-05,
+      "loss": 1.0774,
+      "step": 13096
+    },
+    {
+      "epoch": 2.09552,
+      "grad_norm": 2.1145293610500717,
+      "learning_rate": 4.426446934416046e-05,
+      "loss": 1.1741,
+      "step": 13097
+    },
+    {
+      "epoch": 2.09568,
+      "grad_norm": 1.8368621868842212,
+      "learning_rate": 4.42501281568655e-05,
+      "loss": 1.0627,
+      "step": 13098
+    },
+    {
+      "epoch": 2.09584,
+      "grad_norm": 1.8024615561829826,
+      "learning_rate": 4.423578863306547e-05,
+      "loss": 1.1256,
+      "step": 13099
+    },
+    {
+      "epoch": 2.096,
+      "grad_norm": 1.65265803657825,
+      "learning_rate": 4.422145077318821e-05,
+      "loss": 1.1333,
+      "step": 13100
+    },
+    {
+      "epoch": 2.09616,
+      "grad_norm": 1.8006372364337477,
+      "learning_rate": 4.420711457766167e-05,
+      "loss": 1.1199,
+      "step": 13101
+    },
+    {
+      "epoch": 2.09632,
+      "grad_norm": 1.690871343384997,
+      "learning_rate": 4.419278004691344e-05,
+      "loss": 1.1161,
+      "step": 13102
+    },
+    {
+      "epoch": 2.09648,
+      "grad_norm": 2.2758067775212654,
+      "learning_rate": 4.417844718137142e-05,
+      "loss": 1.169,
+      "step": 13103
+    },
+    {
+      "epoch": 2.09664,
+      "grad_norm": 1.8431212402806223,
+      "learning_rate": 4.4164115981463114e-05,
+      "loss": 1.0943,
+      "step": 13104
+    },
+    {
+      "epoch": 2.0968,
+      "grad_norm": 1.6242480769458534,
+      "learning_rate": 4.4149786447616306e-05,
+      "loss": 1.0848,
+      "step": 13105
+    },
+    {
+      "epoch": 2.09696,
+      "grad_norm": 2.0548250688143317,
+      "learning_rate": 4.4135458580258406e-05,
+      "loss": 1.0723,
+      "step": 13106
+    },
+    {
+      "epoch": 2.09712,
+      "grad_norm": 2.4120322465148574,
+      "learning_rate": 4.412113237981706e-05,
+      "loss": 1.1669,
+      "step": 13107
+    },
+    {
+      "epoch": 2.09728,
+      "grad_norm": 1.8964350199317461,
+      "learning_rate": 4.41068078467197e-05,
+      "loss": 1.1246,
+      "step": 13108
+    },
+    {
+      "epoch": 2.09744,
+      "grad_norm": 1.9728458241083242,
+      "learning_rate": 4.409248498139375e-05,
+      "loss": 1.1719,
+      "step": 13109
+    },
+    {
+      "epoch": 2.0976,
+      "grad_norm": 1.9664331335410647,
+      "learning_rate": 4.407816378426658e-05,
+      "loss": 1.1135,
+      "step": 13110
+    },
+    {
+      "epoch": 2.09776,
+      "grad_norm": 1.9448854285395274,
+      "learning_rate": 4.406384425576552e-05,
+      "loss": 1.1293,
+      "step": 13111
+    },
+    {
+      "epoch": 2.09792,
+      "grad_norm": 2.0465201034113676,
+      "learning_rate": 4.404952639631783e-05,
+      "loss": 1.2344,
+      "step": 13112
+    },
+    {
+      "epoch": 2.09808,
+      "grad_norm": 1.8516222425698299,
+      "learning_rate": 4.4035210206350754e-05,
+      "loss": 1.1619,
+      "step": 13113
+    },
+    {
+      "epoch": 2.09824,
+      "grad_norm": 1.6690293871544355,
+      "learning_rate": 4.4020895686291454e-05,
+      "loss": 1.1456,
+      "step": 13114
+    },
+    {
+      "epoch": 2.0984,
+      "grad_norm": 1.9198718119028666,
+      "learning_rate": 4.400658283656706e-05,
+      "loss": 1.1342,
+      "step": 13115
+    },
+    {
+      "epoch": 2.09856,
+      "grad_norm": 1.9443399989864893,
+      "learning_rate": 4.39922716576046e-05,
+      "loss": 1.1107,
+      "step": 13116
+    },
+    {
+      "epoch": 2.09872,
+      "grad_norm": 2.1471212449318418,
+      "learning_rate": 4.397796214983121e-05,
+      "loss": 1.2073,
+      "step": 13117
+    },
+    {
+      "epoch": 2.09888,
+      "grad_norm": 1.7830632522863237,
+      "learning_rate": 4.396365431367373e-05,
+      "loss": 1.1239,
+      "step": 13118
+    },
+    {
+      "epoch": 2.09904,
+      "grad_norm": 1.9356124733485043,
+      "learning_rate": 4.394934814955923e-05,
+      "loss": 1.1409,
+      "step": 13119
+    },
+    {
+      "epoch": 2.0992,
+      "grad_norm": 2.0986093942480335,
+      "learning_rate": 4.3935043657914434e-05,
+      "loss": 1.1339,
+      "step": 13120
+    },
+    {
+      "epoch": 2.09936,
+      "grad_norm": 1.714819077326692,
+      "learning_rate": 4.3920740839166306e-05,
+      "loss": 1.117,
+      "step": 13121
+    },
+    {
+      "epoch": 2.09952,
+      "grad_norm": 1.9641249127656142,
+      "learning_rate": 4.390643969374149e-05,
+      "loss": 1.1238,
+      "step": 13122
+    },
+    {
+      "epoch": 2.09968,
+      "grad_norm": 1.753634789182818,
+      "learning_rate": 4.38921402220668e-05,
+      "loss": 1.1227,
+      "step": 13123
+    },
+    {
+      "epoch": 2.09984,
+      "grad_norm": 1.7476564995970967,
+      "learning_rate": 4.387784242456889e-05,
+      "loss": 1.2188,
+      "step": 13124
+    },
+    {
+      "epoch": 2.1,
+      "grad_norm": 1.8770644888582146,
+      "learning_rate": 4.38635463016744e-05,
+      "loss": 1.1224,
+      "step": 13125
+    },
+    {
+      "epoch": 2.10016,
+      "grad_norm": 1.9163596442924324,
+      "learning_rate": 4.384925185380987e-05,
+      "loss": 1.1807,
+      "step": 13126
+    },
+    {
+      "epoch": 2.10032,
+      "grad_norm": 1.549016823031963,
+      "learning_rate": 4.383495908140185e-05,
+      "loss": 1.1359,
+      "step": 13127
+    },
+    {
+      "epoch": 2.10048,
+      "grad_norm": 1.7060984116805553,
+      "learning_rate": 4.382066798487682e-05,
+      "loss": 1.1165,
+      "step": 13128
+    },
+    {
+      "epoch": 2.10064,
+      "grad_norm": 1.9765948831532079,
+      "learning_rate": 4.380637856466118e-05,
+      "loss": 1.1195,
+      "step": 13129
+    },
+    {
+      "epoch": 2.1008,
+      "grad_norm": 1.8515173265630636,
+      "learning_rate": 4.379209082118133e-05,
+      "loss": 1.1648,
+      "step": 13130
+    },
+    {
+      "epoch": 2.10096,
+      "grad_norm": 1.775621702964584,
+      "learning_rate": 4.377780475486358e-05,
+      "loss": 1.1546,
+      "step": 13131
+    },
+    {
+      "epoch": 2.10112,
+      "grad_norm": 1.7155392109988181,
+      "learning_rate": 4.376352036613418e-05,
+      "loss": 1.1078,
+      "step": 13132
+    },
+    {
+      "epoch": 2.10128,
+      "grad_norm": 1.9187898397101677,
+      "learning_rate": 4.3749237655419464e-05,
+      "loss": 1.1218,
+      "step": 13133
+    },
+    {
+      "epoch": 2.10144,
+      "grad_norm": 1.9144126844458937,
+      "learning_rate": 4.373495662314545e-05,
+      "loss": 1.1522,
+      "step": 13134
+    },
+    {
+      "epoch": 2.1016,
+      "grad_norm": 2.008162191067798,
+      "learning_rate": 4.372067726973842e-05,
+      "loss": 1.1255,
+      "step": 13135
+    },
+    {
+      "epoch": 2.10176,
+      "grad_norm": 1.716448525202295,
+      "learning_rate": 4.37063995956243e-05,
+      "loss": 1.1146,
+      "step": 13136
+    },
+    {
+      "epoch": 2.10192,
+      "grad_norm": 1.7114083050619173,
+      "learning_rate": 4.369212360122926e-05,
+      "loss": 1.1158,
+      "step": 13137
+    },
+    {
+      "epoch": 2.10208,
+      "grad_norm": 1.9685654685802072,
+      "learning_rate": 4.367784928697913e-05,
+      "loss": 1.1623,
+      "step": 13138
+    },
+    {
+      "epoch": 2.10224,
+      "grad_norm": 1.8341450853173165,
+      "learning_rate": 4.3663576653299945e-05,
+      "loss": 1.1096,
+      "step": 13139
+    },
+    {
+      "epoch": 2.1024,
+      "grad_norm": 1.8365296566325908,
+      "learning_rate": 4.364930570061754e-05,
+      "loss": 1.1196,
+      "step": 13140
+    },
+    {
+      "epoch": 2.10256,
+      "grad_norm": 1.7771707879776388,
+      "learning_rate": 4.363503642935773e-05,
+      "loss": 1.1416,
+      "step": 13141
+    },
+    {
+      "epoch": 2.10272,
+      "grad_norm": 1.9293148123479475,
+      "learning_rate": 4.3620768839946324e-05,
+      "loss": 1.1369,
+      "step": 13142
+    },
+    {
+      "epoch": 2.10288,
+      "grad_norm": 1.8153751172568586,
+      "learning_rate": 4.3606502932809e-05,
+      "loss": 1.1837,
+      "step": 13143
+    },
+    {
+      "epoch": 2.10304,
+      "grad_norm": 1.759434908822508,
+      "learning_rate": 4.359223870837147e-05,
+      "loss": 1.0697,
+      "step": 13144
+    },
+    {
+      "epoch": 2.1032,
+      "grad_norm": 1.7381299009562337,
+      "learning_rate": 4.357797616705934e-05,
+      "loss": 1.1072,
+      "step": 13145
+    },
+    {
+      "epoch": 2.10336,
+      "grad_norm": 1.8488274259288757,
+      "learning_rate": 4.356371530929818e-05,
+      "loss": 1.1332,
+      "step": 13146
+    },
+    {
+      "epoch": 2.10352,
+      "grad_norm": 1.782956679192643,
+      "learning_rate": 4.354945613551352e-05,
+      "loss": 1.155,
+      "step": 13147
+    },
+    {
+      "epoch": 2.1036799999999998,
+      "grad_norm": 1.7005570128689573,
+      "learning_rate": 4.353519864613079e-05,
+      "loss": 1.134,
+      "step": 13148
+    },
+    {
+      "epoch": 2.10384,
+      "grad_norm": 1.765812548313569,
+      "learning_rate": 4.352094284157554e-05,
+      "loss": 1.0425,
+      "step": 13149
+    },
+    {
+      "epoch": 2.104,
+      "grad_norm": 1.7101413852874947,
+      "learning_rate": 4.350668872227298e-05,
+      "loss": 1.1113,
+      "step": 13150
+    },
+    {
+      "epoch": 2.10416,
+      "grad_norm": 1.8922441803759769,
+      "learning_rate": 4.3492436288648577e-05,
+      "loss": 1.1415,
+      "step": 13151
+    },
+    {
+      "epoch": 2.10432,
+      "grad_norm": 1.8877528231527851,
+      "learning_rate": 4.347818554112747e-05,
+      "loss": 1.1464,
+      "step": 13152
+    },
+    {
+      "epoch": 2.10448,
+      "grad_norm": 1.910049974125479,
+      "learning_rate": 4.3463936480135016e-05,
+      "loss": 1.1053,
+      "step": 13153
+    },
+    {
+      "epoch": 2.10464,
+      "grad_norm": 1.7396252077338337,
+      "learning_rate": 4.344968910609624e-05,
+      "loss": 1.1638,
+      "step": 13154
+    },
+    {
+      "epoch": 2.1048,
+      "grad_norm": 1.8223809435737748,
+      "learning_rate": 4.3435443419436375e-05,
+      "loss": 1.0898,
+      "step": 13155
+    },
+    {
+      "epoch": 2.10496,
+      "grad_norm": 1.73350274264125,
+      "learning_rate": 4.342119942058045e-05,
+      "loss": 1.1277,
+      "step": 13156
+    },
+    {
+      "epoch": 2.10512,
+      "grad_norm": 2.0057622930509287,
+      "learning_rate": 4.340695710995349e-05,
+      "loss": 1.0814,
+      "step": 13157
+    },
+    {
+      "epoch": 2.10528,
+      "grad_norm": 1.762219796507081,
+      "learning_rate": 4.339271648798048e-05,
+      "loss": 1.1259,
+      "step": 13158
+    },
+    {
+      "epoch": 2.10544,
+      "grad_norm": 1.9498942160774915,
+      "learning_rate": 4.337847755508627e-05,
+      "loss": 1.1102,
+      "step": 13159
+    },
+    {
+      "epoch": 2.1056,
+      "grad_norm": 2.0639874595277177,
+      "learning_rate": 4.336424031169587e-05,
+      "loss": 1.1256,
+      "step": 13160
+    },
+    {
+      "epoch": 2.10576,
+      "grad_norm": 1.9056983485441372,
+      "learning_rate": 4.335000475823393e-05,
+      "loss": 1.1044,
+      "step": 13161
+    },
+    {
+      "epoch": 2.10592,
+      "grad_norm": 2.21554993213269,
+      "learning_rate": 4.333577089512538e-05,
+      "loss": 1.066,
+      "step": 13162
+    },
+    {
+      "epoch": 2.10608,
+      "grad_norm": 2.0817730185152747,
+      "learning_rate": 4.332153872279479e-05,
+      "loss": 1.1477,
+      "step": 13163
+    },
+    {
+      "epoch": 2.10624,
+      "grad_norm": 1.8329298239029947,
+      "learning_rate": 4.330730824166692e-05,
+      "loss": 1.1213,
+      "step": 13164
+    },
+    {
+      "epoch": 2.1064,
+      "grad_norm": 1.700453712233466,
+      "learning_rate": 4.329307945216636e-05,
+      "loss": 1.1092,
+      "step": 13165
+    },
+    {
+      "epoch": 2.10656,
+      "grad_norm": 1.8457018011929796,
+      "learning_rate": 4.327885235471769e-05,
+      "loss": 1.1184,
+      "step": 13166
+    },
+    {
+      "epoch": 2.10672,
+      "grad_norm": 1.9014598879278877,
+      "learning_rate": 4.32646269497454e-05,
+      "loss": 1.1148,
+      "step": 13167
+    },
+    {
+      "epoch": 2.10688,
+      "grad_norm": 2.127956053638647,
+      "learning_rate": 4.3250403237673986e-05,
+      "loss": 1.1151,
+      "step": 13168
+    },
+    {
+      "epoch": 2.10704,
+      "grad_norm": 2.511084993293082,
+      "learning_rate": 4.323618121892784e-05,
+      "loss": 1.1053,
+      "step": 13169
+    },
+    {
+      "epoch": 2.1072,
+      "grad_norm": 1.7354545936253174,
+      "learning_rate": 4.3221960893931336e-05,
+      "loss": 1.1413,
+      "step": 13170
+    },
+    {
+      "epoch": 2.10736,
+      "grad_norm": 1.861193848857811,
+      "learning_rate": 4.3207742263108774e-05,
+      "loss": 1.1737,
+      "step": 13171
+    },
+    {
+      "epoch": 2.10752,
+      "grad_norm": 1.7503441833264919,
+      "learning_rate": 4.3193525326884435e-05,
+      "loss": 1.1217,
+      "step": 13172
+    },
+    {
+      "epoch": 2.10768,
+      "grad_norm": 1.7721464042965047,
+      "learning_rate": 4.317931008568252e-05,
+      "loss": 1.1124,
+      "step": 13173
+    },
+    {
+      "epoch": 2.10784,
+      "grad_norm": 2.020044536371121,
+      "learning_rate": 4.31650965399272e-05,
+      "loss": 1.088,
+      "step": 13174
+    },
+    {
+      "epoch": 2.108,
+      "grad_norm": 1.7922971941757706,
+      "learning_rate": 4.315088469004255e-05,
+      "loss": 1.1244,
+      "step": 13175
+    },
+    {
+      "epoch": 2.10816,
+      "grad_norm": 1.8205999989987687,
+      "learning_rate": 4.3136674536452724e-05,
+      "loss": 1.1689,
+      "step": 13176
+    },
+    {
+      "epoch": 2.10832,
+      "grad_norm": 1.7229989132469672,
+      "learning_rate": 4.312246607958161e-05,
+      "loss": 1.116,
+      "step": 13177
+    },
+    {
+      "epoch": 2.10848,
+      "grad_norm": 1.8056333756872684,
+      "learning_rate": 4.310825931985329e-05,
+      "loss": 1.1175,
+      "step": 13178
+    },
+    {
+      "epoch": 2.10864,
+      "grad_norm": 1.904926164649663,
+      "learning_rate": 4.309405425769155e-05,
+      "loss": 1.1649,
+      "step": 13179
+    },
+    {
+      "epoch": 2.1088,
+      "grad_norm": 1.8100786253622017,
+      "learning_rate": 4.307985089352035e-05,
+      "loss": 1.1892,
+      "step": 13180
+    },
+    {
+      "epoch": 2.10896,
+      "grad_norm": 2.0446858445476415,
+      "learning_rate": 4.3065649227763436e-05,
+      "loss": 1.1728,
+      "step": 13181
+    },
+    {
+      "epoch": 2.10912,
+      "grad_norm": 1.8933931976597775,
+      "learning_rate": 4.305144926084461e-05,
+      "loss": 1.1364,
+      "step": 13182
+    },
+    {
+      "epoch": 2.10928,
+      "grad_norm": 1.9375999326987805,
+      "learning_rate": 4.303725099318755e-05,
+      "loss": 1.1563,
+      "step": 13183
+    },
+    {
+      "epoch": 2.10944,
+      "grad_norm": 1.843331861732407,
+      "learning_rate": 4.3023054425215905e-05,
+      "loss": 1.1162,
+      "step": 13184
+    },
+    {
+      "epoch": 2.1096,
+      "grad_norm": 1.991005524164962,
+      "learning_rate": 4.30088595573533e-05,
+      "loss": 1.1162,
+      "step": 13185
+    },
+    {
+      "epoch": 2.10976,
+      "grad_norm": 1.8036406742439615,
+      "learning_rate": 4.2994666390023284e-05,
+      "loss": 1.1135,
+      "step": 13186
+    },
+    {
+      "epoch": 2.10992,
+      "grad_norm": 1.8263452686764807,
+      "learning_rate": 4.298047492364935e-05,
+      "loss": 1.1426,
+      "step": 13187
+    },
+    {
+      "epoch": 2.11008,
+      "grad_norm": 1.9421138550397414,
+      "learning_rate": 4.2966285158654964e-05,
+      "loss": 1.1006,
+      "step": 13188
+    },
+    {
+      "epoch": 2.11024,
+      "grad_norm": 1.6900879171573584,
+      "learning_rate": 4.295209709546353e-05,
+      "loss": 1.1163,
+      "step": 13189
+    },
+    {
+      "epoch": 2.1104,
+      "grad_norm": 1.8284428847915706,
+      "learning_rate": 4.2937910734498374e-05,
+      "loss": 1.0981,
+      "step": 13190
+    },
+    {
+      "epoch": 2.11056,
+      "grad_norm": 1.7234195318296515,
+      "learning_rate": 4.292372607618279e-05,
+      "loss": 1.1468,
+      "step": 13191
+    },
+    {
+      "epoch": 2.11072,
+      "grad_norm": 2.080477060543137,
+      "learning_rate": 4.290954312094011e-05,
+      "loss": 1.1232,
+      "step": 13192
+    },
+    {
+      "epoch": 2.11088,
+      "grad_norm": 1.6056022545299264,
+      "learning_rate": 4.289536186919342e-05,
+      "loss": 1.1102,
+      "step": 13193
+    },
+    {
+      "epoch": 2.11104,
+      "grad_norm": 2.0460437312591075,
+      "learning_rate": 4.288118232136598e-05,
+      "loss": 1.2044,
+      "step": 13194
+    },
+    {
+      "epoch": 2.1112,
+      "grad_norm": 2.124737311175092,
+      "learning_rate": 4.2867004477880755e-05,
+      "loss": 1.1953,
+      "step": 13195
+    },
+    {
+      "epoch": 2.11136,
+      "grad_norm": 1.6775950261214876,
+      "learning_rate": 4.285282833916091e-05,
+      "loss": 1.1022,
+      "step": 13196
+    },
+    {
+      "epoch": 2.11152,
+      "grad_norm": 1.790564674471561,
+      "learning_rate": 4.283865390562939e-05,
+      "loss": 1.1059,
+      "step": 13197
+    },
+    {
+      "epoch": 2.11168,
+      "grad_norm": 1.7719889514673928,
+      "learning_rate": 4.282448117770914e-05,
+      "loss": 1.1455,
+      "step": 13198
+    },
+    {
+      "epoch": 2.11184,
+      "grad_norm": 1.883439566957886,
+      "learning_rate": 4.281031015582306e-05,
+      "loss": 1.1308,
+      "step": 13199
+    },
+    {
+      "epoch": 2.112,
+      "grad_norm": 1.7554715988195244,
+      "learning_rate": 4.2796140840393996e-05,
+      "loss": 1.1304,
+      "step": 13200
+    },
+    {
+      "epoch": 2.11216,
+      "grad_norm": 1.8392632409399146,
+      "learning_rate": 4.278197323184474e-05,
+      "loss": 1.1612,
+      "step": 13201
+    },
+    {
+      "epoch": 2.11232,
+      "grad_norm": 1.9508591685098884,
+      "learning_rate": 4.276780733059802e-05,
+      "loss": 1.1279,
+      "step": 13202
+    },
+    {
+      "epoch": 2.11248,
+      "grad_norm": 2.0618774996934506,
+      "learning_rate": 4.275364313707653e-05,
+      "loss": 1.1045,
+      "step": 13203
+    },
+    {
+      "epoch": 2.11264,
+      "grad_norm": 1.6541771405580548,
+      "learning_rate": 4.273948065170291e-05,
+      "loss": 1.1838,
+      "step": 13204
+    },
+    {
+      "epoch": 2.1128,
+      "grad_norm": 1.683286119505814,
+      "learning_rate": 4.272531987489973e-05,
+      "loss": 1.0961,
+      "step": 13205
+    },
+    {
+      "epoch": 2.11296,
+      "grad_norm": 2.022443909221696,
+      "learning_rate": 4.27111608070896e-05,
+      "loss": 1.1436,
+      "step": 13206
+    },
+    {
+      "epoch": 2.11312,
+      "grad_norm": 1.8057120194653766,
+      "learning_rate": 4.26970034486949e-05,
+      "loss": 1.0796,
+      "step": 13207
+    },
+    {
+      "epoch": 2.11328,
+      "grad_norm": 1.686405657706876,
+      "learning_rate": 4.2682847800138174e-05,
+      "loss": 1.1128,
+      "step": 13208
+    },
+    {
+      "epoch": 2.11344,
+      "grad_norm": 1.8801934464334715,
+      "learning_rate": 4.266869386184168e-05,
+      "loss": 1.0899,
+      "step": 13209
+    },
+    {
+      "epoch": 2.1136,
+      "grad_norm": 1.9077527401425232,
+      "learning_rate": 4.2654541634227895e-05,
+      "loss": 1.124,
+      "step": 13210
+    },
+    {
+      "epoch": 2.11376,
+      "grad_norm": 1.7526141557116846,
+      "learning_rate": 4.2640391117718946e-05,
+      "loss": 1.0969,
+      "step": 13211
+    },
+    {
+      "epoch": 2.11392,
+      "grad_norm": 1.8417168903166097,
+      "learning_rate": 4.26262423127372e-05,
+      "loss": 1.1127,
+      "step": 13212
+    },
+    {
+      "epoch": 2.11408,
+      "grad_norm": 1.8074605568224802,
+      "learning_rate": 4.261209521970476e-05,
+      "loss": 1.1032,
+      "step": 13213
+    },
+    {
+      "epoch": 2.11424,
+      "grad_norm": 1.9850964571264473,
+      "learning_rate": 4.2597949839043795e-05,
+      "loss": 1.1875,
+      "step": 13214
+    },
+    {
+      "epoch": 2.1144,
+      "grad_norm": 1.8149175547422385,
+      "learning_rate": 4.2583806171176346e-05,
+      "loss": 1.1188,
+      "step": 13215
+    },
+    {
+      "epoch": 2.11456,
+      "grad_norm": 1.9207153241594745,
+      "learning_rate": 4.256966421652443e-05,
+      "loss": 1.064,
+      "step": 13216
+    },
+    {
+      "epoch": 2.11472,
+      "grad_norm": 2.287220750553043,
+      "learning_rate": 4.255552397551012e-05,
+      "loss": 1.1177,
+      "step": 13217
+    },
+    {
+      "epoch": 2.11488,
+      "grad_norm": 1.8105645039362692,
+      "learning_rate": 4.254138544855519e-05,
+      "loss": 1.1294,
+      "step": 13218
+    },
+    {
+      "epoch": 2.11504,
+      "grad_norm": 1.9592019015135618,
+      "learning_rate": 4.2527248636081675e-05,
+      "loss": 1.1127,
+      "step": 13219
+    },
+    {
+      "epoch": 2.1152,
+      "grad_norm": 1.656178635799968,
+      "learning_rate": 4.251311353851124e-05,
+      "loss": 1.1399,
+      "step": 13220
+    },
+    {
+      "epoch": 2.11536,
+      "grad_norm": 1.9344206743294448,
+      "learning_rate": 4.2498980156265754e-05,
+      "loss": 1.1133,
+      "step": 13221
+    },
+    {
+      "epoch": 2.11552,
+      "grad_norm": 1.7179557558408158,
+      "learning_rate": 4.248484848976693e-05,
+      "loss": 1.1524,
+      "step": 13222
+    },
+    {
+      "epoch": 2.1156800000000002,
+      "grad_norm": 2.125697980331787,
+      "learning_rate": 4.2470718539436404e-05,
+      "loss": 1.0849,
+      "step": 13223
+    },
+    {
+      "epoch": 2.11584,
+      "grad_norm": 1.9528510763186595,
+      "learning_rate": 4.245659030569582e-05,
+      "loss": 1.1057,
+      "step": 13224
+    },
+    {
+      "epoch": 2.116,
+      "grad_norm": 1.8452509752145736,
+      "learning_rate": 4.244246378896672e-05,
+      "loss": 1.123,
+      "step": 13225
+    },
+    {
+      "epoch": 2.11616,
+      "grad_norm": 1.687050793351543,
+      "learning_rate": 4.242833898967064e-05,
+      "loss": 1.0819,
+      "step": 13226
+    },
+    {
+      "epoch": 2.11632,
+      "grad_norm": 1.8758851172205717,
+      "learning_rate": 4.241421590822903e-05,
+      "loss": 1.0997,
+      "step": 13227
+    },
+    {
+      "epoch": 2.11648,
+      "grad_norm": 1.9278328811837862,
+      "learning_rate": 4.240009454506331e-05,
+      "loss": 1.1052,
+      "step": 13228
+    },
+    {
+      "epoch": 2.11664,
+      "grad_norm": 1.803180866243989,
+      "learning_rate": 4.2385974900594824e-05,
+      "loss": 1.1095,
+      "step": 13229
+    },
+    {
+      "epoch": 2.1168,
+      "grad_norm": 1.7780070684196296,
+      "learning_rate": 4.237185697524491e-05,
+      "loss": 1.0609,
+      "step": 13230
+    },
+    {
+      "epoch": 2.11696,
+      "grad_norm": 1.8030233528451332,
+      "learning_rate": 4.2357740769434794e-05,
+      "loss": 1.1108,
+      "step": 13231
+    },
+    {
+      "epoch": 2.11712,
+      "grad_norm": 2.0295973288647167,
+      "learning_rate": 4.234362628358568e-05,
+      "loss": 1.1632,
+      "step": 13232
+    },
+    {
+      "epoch": 2.11728,
+      "grad_norm": 1.7365082831086396,
+      "learning_rate": 4.23295135181188e-05,
+      "loss": 1.161,
+      "step": 13233
+    },
+    {
+      "epoch": 2.11744,
+      "grad_norm": 1.659697866970156,
+      "learning_rate": 4.231540247345515e-05,
+      "loss": 1.0959,
+      "step": 13234
+    },
+    {
+      "epoch": 2.1176,
+      "grad_norm": 2.248508121180914,
+      "learning_rate": 4.2301293150015886e-05,
+      "loss": 1.1804,
+      "step": 13235
+    },
+    {
+      "epoch": 2.11776,
+      "grad_norm": 2.043768496807719,
+      "learning_rate": 4.22871855482219e-05,
+      "loss": 1.1281,
+      "step": 13236
+    },
+    {
+      "epoch": 2.11792,
+      "grad_norm": 1.741847632629191,
+      "learning_rate": 4.2273079668494234e-05,
+      "loss": 1.11,
+      "step": 13237
+    },
+    {
+      "epoch": 2.11808,
+      "grad_norm": 1.7507613806157478,
+      "learning_rate": 4.2258975511253754e-05,
+      "loss": 1.0919,
+      "step": 13238
+    },
+    {
+      "epoch": 2.11824,
+      "grad_norm": 2.4234341043626086,
+      "learning_rate": 4.22448730769213e-05,
+      "loss": 1.1663,
+      "step": 13239
+    },
+    {
+      "epoch": 2.1184,
+      "grad_norm": 1.9939670783316215,
+      "learning_rate": 4.223077236591768e-05,
+      "loss": 1.2296,
+      "step": 13240
+    },
+    {
+      "epoch": 2.11856,
+      "grad_norm": 1.6839188718630727,
+      "learning_rate": 4.221667337866364e-05,
+      "loss": 1.1178,
+      "step": 13241
+    },
+    {
+      "epoch": 2.11872,
+      "grad_norm": 2.194144291245529,
+      "learning_rate": 4.2202576115579864e-05,
+      "loss": 1.0944,
+      "step": 13242
+    },
+    {
+      "epoch": 2.11888,
+      "grad_norm": 1.6218334324186965,
+      "learning_rate": 4.2188480577087e-05,
+      "loss": 1.1097,
+      "step": 13243
+    },
+    {
+      "epoch": 2.11904,
+      "grad_norm": 1.9443050713444583,
+      "learning_rate": 4.2174386763605625e-05,
+      "loss": 1.1805,
+      "step": 13244
+    },
+    {
+      "epoch": 2.1192,
+      "grad_norm": 2.152786770212296,
+      "learning_rate": 4.216029467555629e-05,
+      "loss": 1.1191,
+      "step": 13245
+    },
+    {
+      "epoch": 2.11936,
+      "grad_norm": 2.014879219468663,
+      "learning_rate": 4.2146204313359484e-05,
+      "loss": 1.1367,
+      "step": 13246
+    },
+    {
+      "epoch": 2.11952,
+      "grad_norm": 2.244836323922865,
+      "learning_rate": 4.213211567743564e-05,
+      "loss": 1.1234,
+      "step": 13247
+    },
+    {
+      "epoch": 2.11968,
+      "grad_norm": 2.2893360968958443,
+      "learning_rate": 4.211802876820511e-05,
+      "loss": 1.0913,
+      "step": 13248
+    },
+    {
+      "epoch": 2.11984,
+      "grad_norm": 1.9182754024033064,
+      "learning_rate": 4.2103943586088327e-05,
+      "loss": 1.2105,
+      "step": 13249
+    },
+    {
+      "epoch": 2.12,
+      "grad_norm": 1.7650585604621338,
+      "learning_rate": 4.208986013150542e-05,
+      "loss": 1.152,
+      "step": 13250
+    },
+    {
+      "epoch": 2.12016,
+      "grad_norm": 2.1526735163516633,
+      "learning_rate": 4.207577840487679e-05,
+      "loss": 1.1183,
+      "step": 13251
+    },
+    {
+      "epoch": 2.12032,
+      "grad_norm": 2.477565035375623,
+      "learning_rate": 4.2061698406622454e-05,
+      "loss": 1.1699,
+      "step": 13252
+    },
+    {
+      "epoch": 2.12048,
+      "grad_norm": 1.7918230858865085,
+      "learning_rate": 4.204762013716265e-05,
+      "loss": 1.1181,
+      "step": 13253
+    },
+    {
+      "epoch": 2.12064,
+      "grad_norm": 1.7574679848246972,
+      "learning_rate": 4.20335435969174e-05,
+      "loss": 1.0899,
+      "step": 13254
+    },
+    {
+      "epoch": 2.1208,
+      "grad_norm": 2.2294451023750135,
+      "learning_rate": 4.201946878630676e-05,
+      "loss": 1.1802,
+      "step": 13255
+    },
+    {
+      "epoch": 2.12096,
+      "grad_norm": 2.172026546959274,
+      "learning_rate": 4.200539570575068e-05,
+      "loss": 1.1196,
+      "step": 13256
+    },
+    {
+      "epoch": 2.12112,
+      "grad_norm": 1.8677022403279575,
+      "learning_rate": 4.1991324355669085e-05,
+      "loss": 1.0936,
+      "step": 13257
+    },
+    {
+      "epoch": 2.12128,
+      "grad_norm": 1.847133848895918,
+      "learning_rate": 4.197725473648184e-05,
+      "loss": 1.1177,
+      "step": 13258
+    },
+    {
+      "epoch": 2.12144,
+      "grad_norm": 2.8314524016772937,
+      "learning_rate": 4.1963186848608774e-05,
+      "loss": 1.1135,
+      "step": 13259
+    },
+    {
+      "epoch": 2.1216,
+      "grad_norm": 1.7281359914207666,
+      "learning_rate": 4.194912069246964e-05,
+      "loss": 1.1491,
+      "step": 13260
+    },
+    {
+      "epoch": 2.12176,
+      "grad_norm": 1.8692730741093457,
+      "learning_rate": 4.193505626848415e-05,
+      "loss": 1.1366,
+      "step": 13261
+    },
+    {
+      "epoch": 2.12192,
+      "grad_norm": 1.968771772952243,
+      "learning_rate": 4.1920993577071975e-05,
+      "loss": 1.0861,
+      "step": 13262
+    },
+    {
+      "epoch": 2.12208,
+      "grad_norm": 2.1184047675604973,
+      "learning_rate": 4.1906932618652717e-05,
+      "loss": 1.1686,
+      "step": 13263
+    },
+    {
+      "epoch": 2.12224,
+      "grad_norm": 2.165513206987352,
+      "learning_rate": 4.189287339364592e-05,
+      "loss": 1.2123,
+      "step": 13264
+    },
+    {
+      "epoch": 2.1224,
+      "grad_norm": 1.794471308318345,
+      "learning_rate": 4.1878815902471156e-05,
+      "loss": 1.1545,
+      "step": 13265
+    },
+    {
+      "epoch": 2.12256,
+      "grad_norm": 1.850623571372095,
+      "learning_rate": 4.1864760145547775e-05,
+      "loss": 1.1765,
+      "step": 13266
+    },
+    {
+      "epoch": 2.12272,
+      "grad_norm": 2.069343176216883,
+      "learning_rate": 4.1850706123295314e-05,
+      "loss": 1.1445,
+      "step": 13267
+    },
+    {
+      "epoch": 2.12288,
+      "grad_norm": 2.074301171118888,
+      "learning_rate": 4.183665383613298e-05,
+      "loss": 1.15,
+      "step": 13268
+    },
+    {
+      "epoch": 2.12304,
+      "grad_norm": 1.828830445533629,
+      "learning_rate": 4.182260328448018e-05,
+      "loss": 1.1811,
+      "step": 13269
+    },
+    {
+      "epoch": 2.1232,
+      "grad_norm": 1.8761988729539103,
+      "learning_rate": 4.180855446875611e-05,
+      "loss": 1.1178,
+      "step": 13270
+    },
+    {
+      "epoch": 2.12336,
+      "grad_norm": 1.9263125430301267,
+      "learning_rate": 4.179450738937999e-05,
+      "loss": 1.119,
+      "step": 13271
+    },
+    {
+      "epoch": 2.12352,
+      "grad_norm": 2.0395185373795024,
+      "learning_rate": 4.1780462046770964e-05,
+      "loss": 1.1357,
+      "step": 13272
+    },
+    {
+      "epoch": 2.12368,
+      "grad_norm": 1.6096061436047986,
+      "learning_rate": 4.1766418441348065e-05,
+      "loss": 1.077,
+      "step": 13273
+    },
+    {
+      "epoch": 2.12384,
+      "grad_norm": 1.843345011903914,
+      "learning_rate": 4.175237657353046e-05,
+      "loss": 1.1322,
+      "step": 13274
+    },
+    {
+      "epoch": 2.124,
+      "grad_norm": 2.0141503715999955,
+      "learning_rate": 4.173833644373699e-05,
+      "loss": 1.1783,
+      "step": 13275
+    },
+    {
+      "epoch": 2.12416,
+      "grad_norm": 1.6999283849427649,
+      "learning_rate": 4.172429805238674e-05,
+      "loss": 1.1835,
+      "step": 13276
+    },
+    {
+      "epoch": 2.12432,
+      "grad_norm": 1.7698965168249983,
+      "learning_rate": 4.171026139989844e-05,
+      "loss": 1.0965,
+      "step": 13277
+    },
+    {
+      "epoch": 2.12448,
+      "grad_norm": 1.9505968242570921,
+      "learning_rate": 4.169622648669109e-05,
+      "loss": 1.1179,
+      "step": 13278
+    },
+    {
+      "epoch": 2.12464,
+      "grad_norm": 1.9279876634859705,
+      "learning_rate": 4.16821933131833e-05,
+      "loss": 1.1194,
+      "step": 13279
+    },
+    {
+      "epoch": 2.1248,
+      "grad_norm": 1.7658988641555649,
+      "learning_rate": 4.166816187979392e-05,
+      "loss": 1.1165,
+      "step": 13280
+    },
+    {
+      "epoch": 2.12496,
+      "grad_norm": 1.990103695605489,
+      "learning_rate": 4.1654132186941594e-05,
+      "loss": 1.1575,
+      "step": 13281
+    },
+    {
+      "epoch": 2.12512,
+      "grad_norm": 1.8281105243640714,
+      "learning_rate": 4.1640104235044954e-05,
+      "loss": 1.111,
+      "step": 13282
+    },
+    {
+      "epoch": 2.12528,
+      "grad_norm": 1.9135201545050056,
+      "learning_rate": 4.162607802452255e-05,
+      "loss": 1.1641,
+      "step": 13283
+    },
+    {
+      "epoch": 2.12544,
+      "grad_norm": 1.6493442131645202,
+      "learning_rate": 4.161205355579292e-05,
+      "loss": 1.1024,
+      "step": 13284
+    },
+    {
+      "epoch": 2.1256,
+      "grad_norm": 1.8191164600234784,
+      "learning_rate": 4.1598030829274535e-05,
+      "loss": 1.2224,
+      "step": 13285
+    },
+    {
+      "epoch": 2.12576,
+      "grad_norm": 1.8169605130984887,
+      "learning_rate": 4.1584009845385805e-05,
+      "loss": 1.119,
+      "step": 13286
+    },
+    {
+      "epoch": 2.12592,
+      "grad_norm": 1.8662951120897093,
+      "learning_rate": 4.15699906045451e-05,
+      "loss": 1.1348,
+      "step": 13287
+    },
+    {
+      "epoch": 2.12608,
+      "grad_norm": 1.5568306247575576,
+      "learning_rate": 4.155597310717074e-05,
+      "loss": 1.1289,
+      "step": 13288
+    },
+    {
+      "epoch": 2.12624,
+      "grad_norm": 1.8232001410503347,
+      "learning_rate": 4.154195735368095e-05,
+      "loss": 1.1691,
+      "step": 13289
+    },
+    {
+      "epoch": 2.1264,
+      "grad_norm": 2.0286759562442103,
+      "learning_rate": 4.152794334449402e-05,
+      "loss": 1.1233,
+      "step": 13290
+    },
+    {
+      "epoch": 2.12656,
+      "grad_norm": 1.8080550992383135,
+      "learning_rate": 4.151393108002801e-05,
+      "loss": 1.0933,
+      "step": 13291
+    },
+    {
+      "epoch": 2.12672,
+      "grad_norm": 1.7540504512376975,
+      "learning_rate": 4.149992056070113e-05,
+      "loss": 1.0859,
+      "step": 13292
+    },
+    {
+      "epoch": 2.12688,
+      "grad_norm": 1.9382361755600224,
+      "learning_rate": 4.148591178693131e-05,
+      "loss": 1.1204,
+      "step": 13293
+    },
+    {
+      "epoch": 2.12704,
+      "grad_norm": 1.6518425019755616,
+      "learning_rate": 4.1471904759136693e-05,
+      "loss": 1.1204,
+      "step": 13294
+    },
+    {
+      "epoch": 2.1272,
+      "grad_norm": 1.7732756276904944,
+      "learning_rate": 4.145789947773507e-05,
+      "loss": 1.0906,
+      "step": 13295
+    },
+    {
+      "epoch": 2.12736,
+      "grad_norm": 1.7910135713766866,
+      "learning_rate": 4.144389594314447e-05,
+      "loss": 1.1089,
+      "step": 13296
+    },
+    {
+      "epoch": 2.12752,
+      "grad_norm": 1.919755713040857,
+      "learning_rate": 4.1429894155782686e-05,
+      "loss": 1.119,
+      "step": 13297
+    },
+    {
+      "epoch": 2.12768,
+      "grad_norm": 1.7057629515353228,
+      "learning_rate": 4.141589411606752e-05,
+      "loss": 1.1458,
+      "step": 13298
+    },
+    {
+      "epoch": 2.12784,
+      "grad_norm": 1.8987123951156943,
+      "learning_rate": 4.14018958244167e-05,
+      "loss": 1.1335,
+      "step": 13299
+    },
+    {
+      "epoch": 2.128,
+      "grad_norm": 1.6443939158188765,
+      "learning_rate": 4.138789928124793e-05,
+      "loss": 1.1359,
+      "step": 13300
+    },
+    {
+      "epoch": 2.12816,
+      "grad_norm": 1.9307123105003714,
+      "learning_rate": 4.137390448697883e-05,
+      "loss": 1.0781,
+      "step": 13301
+    },
+    {
+      "epoch": 2.12832,
+      "grad_norm": 1.8961269304338344,
+      "learning_rate": 4.1359911442027e-05,
+      "loss": 1.1245,
+      "step": 13302
+    },
+    {
+      "epoch": 2.12848,
+      "grad_norm": 2.300242114747396,
+      "learning_rate": 4.134592014680996e-05,
+      "loss": 1.0941,
+      "step": 13303
+    },
+    {
+      "epoch": 2.12864,
+      "grad_norm": 1.6453017164987944,
+      "learning_rate": 4.1331930601745194e-05,
+      "loss": 1.1172,
+      "step": 13304
+    },
+    {
+      "epoch": 2.1288,
+      "grad_norm": 2.0057003974391967,
+      "learning_rate": 4.13179428072501e-05,
+      "loss": 1.0771,
+      "step": 13305
+    },
+    {
+      "epoch": 2.12896,
+      "grad_norm": 2.267692943882667,
+      "learning_rate": 4.130395676374215e-05,
+      "loss": 1.1789,
+      "step": 13306
+    },
+    {
+      "epoch": 2.12912,
+      "grad_norm": 1.7940753738402242,
+      "learning_rate": 4.1289972471638526e-05,
+      "loss": 1.1718,
+      "step": 13307
+    },
+    {
+      "epoch": 2.12928,
+      "grad_norm": 2.115962344526453,
+      "learning_rate": 4.127598993135665e-05,
+      "loss": 1.2173,
+      "step": 13308
+    },
+    {
+      "epoch": 2.1294399999999998,
+      "grad_norm": 1.6284575467182811,
+      "learning_rate": 4.1262009143313594e-05,
+      "loss": 1.1722,
+      "step": 13309
+    },
+    {
+      "epoch": 2.1296,
+      "grad_norm": 1.9604976562781575,
+      "learning_rate": 4.124803010792664e-05,
+      "loss": 1.1111,
+      "step": 13310
+    },
+    {
+      "epoch": 2.12976,
+      "grad_norm": 1.831514375939566,
+      "learning_rate": 4.123405282561285e-05,
+      "loss": 1.119,
+      "step": 13311
+    },
+    {
+      "epoch": 2.12992,
+      "grad_norm": 1.9434229521454587,
+      "learning_rate": 4.1220077296789285e-05,
+      "loss": 1.059,
+      "step": 13312
+    },
+    {
+      "epoch": 2.13008,
+      "grad_norm": 1.8894705606049913,
+      "learning_rate": 4.1206103521872964e-05,
+      "loss": 1.1113,
+      "step": 13313
+    },
+    {
+      "epoch": 2.13024,
+      "grad_norm": 2.105276093997548,
+      "learning_rate": 4.1192131501280854e-05,
+      "loss": 1.1225,
+      "step": 13314
+    },
+    {
+      "epoch": 2.1304,
+      "grad_norm": 2.055084328682511,
+      "learning_rate": 4.117816123542984e-05,
+      "loss": 1.109,
+      "step": 13315
+    },
+    {
+      "epoch": 2.13056,
+      "grad_norm": 2.0185982063803856,
+      "learning_rate": 4.116419272473677e-05,
+      "loss": 1.1211,
+      "step": 13316
+    },
+    {
+      "epoch": 2.13072,
+      "grad_norm": 1.8463087231913766,
+      "learning_rate": 4.1150225969618475e-05,
+      "loss": 1.1041,
+      "step": 13317
+    },
+    {
+      "epoch": 2.13088,
+      "grad_norm": 2.2599694428422117,
+      "learning_rate": 4.113626097049167e-05,
+      "loss": 1.1283,
+      "step": 13318
+    },
+    {
+      "epoch": 2.13104,
+      "grad_norm": 2.0520939132873037,
+      "learning_rate": 4.1122297727773076e-05,
+      "loss": 1.153,
+      "step": 13319
+    },
+    {
+      "epoch": 2.1312,
+      "grad_norm": 2.3436567998214435,
+      "learning_rate": 4.110833624187931e-05,
+      "loss": 1.1834,
+      "step": 13320
+    },
+    {
+      "epoch": 2.13136,
+      "grad_norm": 1.947497174882054,
+      "learning_rate": 4.109437651322695e-05,
+      "loss": 1.1051,
+      "step": 13321
+    },
+    {
+      "epoch": 2.13152,
+      "grad_norm": 1.691581805441337,
+      "learning_rate": 4.108041854223264e-05,
+      "loss": 1.1151,
+      "step": 13322
+    },
+    {
+      "epoch": 2.13168,
+      "grad_norm": 1.7321742297443559,
+      "learning_rate": 4.1066462329312695e-05,
+      "loss": 1.1137,
+      "step": 13323
+    },
+    {
+      "epoch": 2.13184,
+      "grad_norm": 1.9040380975303999,
+      "learning_rate": 4.105250787488373e-05,
+      "loss": 1.099,
+      "step": 13324
+    },
+    {
+      "epoch": 2.132,
+      "grad_norm": 2.0300949290582415,
+      "learning_rate": 4.103855517936195e-05,
+      "loss": 1.0917,
+      "step": 13325
+    },
+    {
+      "epoch": 2.13216,
+      "grad_norm": 1.9706712078830888,
+      "learning_rate": 4.10246042431638e-05,
+      "loss": 1.1519,
+      "step": 13326
+    },
+    {
+      "epoch": 2.13232,
+      "grad_norm": 1.7883205279869938,
+      "learning_rate": 4.101065506670554e-05,
+      "loss": 1.1546,
+      "step": 13327
+    },
+    {
+      "epoch": 2.13248,
+      "grad_norm": 2.0532948894869505,
+      "learning_rate": 4.0996707650403355e-05,
+      "loss": 1.1128,
+      "step": 13328
+    },
+    {
+      "epoch": 2.13264,
+      "grad_norm": 1.83156002114475,
+      "learning_rate": 4.0982761994673456e-05,
+      "loss": 1.1255,
+      "step": 13329
+    },
+    {
+      "epoch": 2.1328,
+      "grad_norm": 1.7155452763812253,
+      "learning_rate": 4.096881809993189e-05,
+      "loss": 1.0804,
+      "step": 13330
+    },
+    {
+      "epoch": 2.13296,
+      "grad_norm": 2.0010090228782564,
+      "learning_rate": 4.095487596659486e-05,
+      "loss": 1.1302,
+      "step": 13331
+    },
+    {
+      "epoch": 2.13312,
+      "grad_norm": 2.271481017892748,
+      "learning_rate": 4.094093559507821e-05,
+      "loss": 1.1035,
+      "step": 13332
+    },
+    {
+      "epoch": 2.13328,
+      "grad_norm": 2.075149225641839,
+      "learning_rate": 4.092699698579806e-05,
+      "loss": 1.111,
+      "step": 13333
+    },
+    {
+      "epoch": 2.1334400000000002,
+      "grad_norm": 1.8184670559453116,
+      "learning_rate": 4.091306013917018e-05,
+      "loss": 1.0726,
+      "step": 13334
+    },
+    {
+      "epoch": 2.1336,
+      "grad_norm": 1.8742737867338974,
+      "learning_rate": 4.089912505561054e-05,
+      "loss": 1.1595,
+      "step": 13335
+    },
+    {
+      "epoch": 2.13376,
+      "grad_norm": 2.09917157214896,
+      "learning_rate": 4.088519173553483e-05,
+      "loss": 1.1343,
+      "step": 13336
+    },
+    {
+      "epoch": 2.13392,
+      "grad_norm": 1.8757901200137679,
+      "learning_rate": 4.0871260179358895e-05,
+      "loss": 1.1442,
+      "step": 13337
+    },
+    {
+      "epoch": 2.13408,
+      "grad_norm": 1.7931013322378329,
+      "learning_rate": 4.085733038749839e-05,
+      "loss": 1.1162,
+      "step": 13338
+    },
+    {
+      "epoch": 2.13424,
+      "grad_norm": 1.943867933349439,
+      "learning_rate": 4.0843402360368976e-05,
+      "loss": 1.1007,
+      "step": 13339
+    },
+    {
+      "epoch": 2.1344,
+      "grad_norm": 1.9880836922270728,
+      "learning_rate": 4.082947609838622e-05,
+      "loss": 1.2049,
+      "step": 13340
+    },
+    {
+      "epoch": 2.13456,
+      "grad_norm": 2.114250399025095,
+      "learning_rate": 4.08155516019657e-05,
+      "loss": 1.1323,
+      "step": 13341
+    },
+    {
+      "epoch": 2.13472,
+      "grad_norm": 2.033006462137685,
+      "learning_rate": 4.0801628871522854e-05,
+      "loss": 1.144,
+      "step": 13342
+    },
+    {
+      "epoch": 2.13488,
+      "grad_norm": 1.7218172612129519,
+      "learning_rate": 4.078770790747315e-05,
+      "loss": 1.0866,
+      "step": 13343
+    },
+    {
+      "epoch": 2.13504,
+      "grad_norm": 1.7086799522448661,
+      "learning_rate": 4.077378871023196e-05,
+      "loss": 1.1099,
+      "step": 13344
+    },
+    {
+      "epoch": 2.1352,
+      "grad_norm": 1.8973219379049455,
+      "learning_rate": 4.075987128021462e-05,
+      "loss": 1.0686,
+      "step": 13345
+    },
+    {
+      "epoch": 2.13536,
+      "grad_norm": 1.8821605279700928,
+      "learning_rate": 4.0745955617836365e-05,
+      "loss": 1.1109,
+      "step": 13346
+    },
+    {
+      "epoch": 2.13552,
+      "grad_norm": 1.9123383195267043,
+      "learning_rate": 4.073204172351251e-05,
+      "loss": 1.1543,
+      "step": 13347
+    },
+    {
+      "epoch": 2.13568,
+      "grad_norm": 1.5778002840023517,
+      "learning_rate": 4.0718129597658104e-05,
+      "loss": 1.1158,
+      "step": 13348
+    },
+    {
+      "epoch": 2.13584,
+      "grad_norm": 1.7849096906770436,
+      "learning_rate": 4.0704219240688404e-05,
+      "loss": 1.0996,
+      "step": 13349
+    },
+    {
+      "epoch": 2.136,
+      "grad_norm": 1.8494887300448324,
+      "learning_rate": 4.069031065301833e-05,
+      "loss": 1.1297,
+      "step": 13350
+    },
+    {
+      "epoch": 2.13616,
+      "grad_norm": 1.8454883680245049,
+      "learning_rate": 4.0676403835063014e-05,
+      "loss": 1.2189,
+      "step": 13351
+    },
+    {
+      "epoch": 2.13632,
+      "grad_norm": 2.5828568705773653,
+      "learning_rate": 4.0662498787237314e-05,
+      "loss": 1.1076,
+      "step": 13352
+    },
+    {
+      "epoch": 2.13648,
+      "grad_norm": 1.9208702538452798,
+      "learning_rate": 4.064859550995621e-05,
+      "loss": 1.1267,
+      "step": 13353
+    },
+    {
+      "epoch": 2.13664,
+      "grad_norm": 1.9832609281046396,
+      "learning_rate": 4.063469400363453e-05,
+      "loss": 1.1066,
+      "step": 13354
+    },
+    {
+      "epoch": 2.1368,
+      "grad_norm": 2.0480435762778657,
+      "learning_rate": 4.062079426868709e-05,
+      "loss": 1.1636,
+      "step": 13355
+    },
+    {
+      "epoch": 2.13696,
+      "grad_norm": 2.0642730953901625,
+      "learning_rate": 4.060689630552861e-05,
+      "loss": 1.1818,
+      "step": 13356
+    },
+    {
+      "epoch": 2.13712,
+      "grad_norm": 1.6657849965001057,
+      "learning_rate": 4.059300011457381e-05,
+      "loss": 1.1263,
+      "step": 13357
+    },
+    {
+      "epoch": 2.13728,
+      "grad_norm": 1.9281461626159333,
+      "learning_rate": 4.057910569623732e-05,
+      "loss": 1.0822,
+      "step": 13358
+    },
+    {
+      "epoch": 2.13744,
+      "grad_norm": 1.8176450251234242,
+      "learning_rate": 4.056521305093373e-05,
+      "loss": 1.156,
+      "step": 13359
+    },
+    {
+      "epoch": 2.1376,
+      "grad_norm": 1.7062523181633389,
+      "learning_rate": 4.0551322179077564e-05,
+      "loss": 1.0843,
+      "step": 13360
+    },
+    {
+      "epoch": 2.13776,
+      "grad_norm": 1.8227122060729173,
+      "learning_rate": 4.053743308108333e-05,
+      "loss": 1.1442,
+      "step": 13361
+    },
+    {
+      "epoch": 2.13792,
+      "grad_norm": 2.153385385664839,
+      "learning_rate": 4.052354575736541e-05,
+      "loss": 1.1474,
+      "step": 13362
+    },
+    {
+      "epoch": 2.13808,
+      "grad_norm": 1.8213345242967727,
+      "learning_rate": 4.05096602083383e-05,
+      "loss": 1.1129,
+      "step": 13363
+    },
+    {
+      "epoch": 2.13824,
+      "grad_norm": 2.0236630091440566,
+      "learning_rate": 4.0495776434416155e-05,
+      "loss": 1.0925,
+      "step": 13364
+    },
+    {
+      "epoch": 2.1384,
+      "grad_norm": 1.6852529507576042,
+      "learning_rate": 4.0481894436013414e-05,
+      "loss": 1.0802,
+      "step": 13365
+    },
+    {
+      "epoch": 2.13856,
+      "grad_norm": 1.5882705479701313,
+      "learning_rate": 4.046801421354414e-05,
+      "loss": 1.1212,
+      "step": 13366
+    },
+    {
+      "epoch": 2.13872,
+      "grad_norm": 1.7922654321035518,
+      "learning_rate": 4.045413576742264e-05,
+      "loss": 1.1553,
+      "step": 13367
+    },
+    {
+      "epoch": 2.13888,
+      "grad_norm": 1.8789282230006703,
+      "learning_rate": 4.04402590980629e-05,
+      "loss": 1.1367,
+      "step": 13368
+    },
+    {
+      "epoch": 2.13904,
+      "grad_norm": 1.7778268971070446,
+      "learning_rate": 4.0426384205879075e-05,
+      "loss": 1.0974,
+      "step": 13369
+    },
+    {
+      "epoch": 2.1391999999999998,
+      "grad_norm": 1.756302909158267,
+      "learning_rate": 4.0412511091285144e-05,
+      "loss": 1.1227,
+      "step": 13370
+    },
+    {
+      "epoch": 2.13936,
+      "grad_norm": 1.7097870108027453,
+      "learning_rate": 4.039863975469504e-05,
+      "loss": 1.0913,
+      "step": 13371
+    },
+    {
+      "epoch": 2.13952,
+      "grad_norm": 2.4590833445207454,
+      "learning_rate": 4.0384770196522694e-05,
+      "loss": 1.0959,
+      "step": 13372
+    },
+    {
+      "epoch": 2.13968,
+      "grad_norm": 1.8890340591204404,
+      "learning_rate": 4.037090241718192e-05,
+      "loss": 1.1394,
+      "step": 13373
+    },
+    {
+      "epoch": 2.13984,
+      "grad_norm": 1.8386306060596493,
+      "learning_rate": 4.0357036417086537e-05,
+      "loss": 1.1183,
+      "step": 13374
+    },
+    {
+      "epoch": 2.14,
+      "grad_norm": 1.9241299670097516,
+      "learning_rate": 4.034317219665028e-05,
+      "loss": 1.1082,
+      "step": 13375
+    },
+    {
+      "epoch": 2.14016,
+      "grad_norm": 1.9411119573070528,
+      "learning_rate": 4.032930975628682e-05,
+      "loss": 1.1194,
+      "step": 13376
+    },
+    {
+      "epoch": 2.14032,
+      "grad_norm": 1.8893216071030101,
+      "learning_rate": 4.031544909640983e-05,
+      "loss": 1.1168,
+      "step": 13377
+    },
+    {
+      "epoch": 2.14048,
+      "grad_norm": 1.7562460933867943,
+      "learning_rate": 4.030159021743283e-05,
+      "loss": 1.0705,
+      "step": 13378
+    },
+    {
+      "epoch": 2.14064,
+      "grad_norm": 2.267305880601187,
+      "learning_rate": 4.028773311976944e-05,
+      "loss": 1.1548,
+      "step": 13379
+    },
+    {
+      "epoch": 2.1408,
+      "grad_norm": 1.95257969728081,
+      "learning_rate": 4.027387780383304e-05,
+      "loss": 1.125,
+      "step": 13380
+    },
+    {
+      "epoch": 2.14096,
+      "grad_norm": 1.7571051335708274,
+      "learning_rate": 4.026002427003714e-05,
+      "loss": 1.1358,
+      "step": 13381
+    },
+    {
+      "epoch": 2.14112,
+      "grad_norm": 1.7702091135120535,
+      "learning_rate": 4.024617251879501e-05,
+      "loss": 1.0828,
+      "step": 13382
+    },
+    {
+      "epoch": 2.14128,
+      "grad_norm": 2.0738814655151736,
+      "learning_rate": 4.02323225505201e-05,
+      "loss": 1.1188,
+      "step": 13383
+    },
+    {
+      "epoch": 2.1414400000000002,
+      "grad_norm": 1.891962017698281,
+      "learning_rate": 4.02184743656255e-05,
+      "loss": 1.104,
+      "step": 13384
+    },
+    {
+      "epoch": 2.1416,
+      "grad_norm": 1.7279065476752744,
+      "learning_rate": 4.020462796452457e-05,
+      "loss": 1.1312,
+      "step": 13385
+    },
+    {
+      "epoch": 2.14176,
+      "grad_norm": 1.9409457208040894,
+      "learning_rate": 4.01907833476304e-05,
+      "loss": 1.0786,
+      "step": 13386
+    },
+    {
+      "epoch": 2.14192,
+      "grad_norm": 2.0404260412161173,
+      "learning_rate": 4.017694051535611e-05,
+      "loss": 1.1676,
+      "step": 13387
+    },
+    {
+      "epoch": 2.14208,
+      "grad_norm": 1.8370556631019033,
+      "learning_rate": 4.016309946811474e-05,
+      "loss": 1.151,
+      "step": 13388
+    },
+    {
+      "epoch": 2.14224,
+      "grad_norm": 1.943015019098331,
+      "learning_rate": 4.0149260206319265e-05,
+      "loss": 1.1304,
+      "step": 13389
+    },
+    {
+      "epoch": 2.1424,
+      "grad_norm": 1.7290150651385499,
+      "learning_rate": 4.013542273038271e-05,
+      "loss": 1.1281,
+      "step": 13390
+    },
+    {
+      "epoch": 2.14256,
+      "grad_norm": 1.9125339408934559,
+      "learning_rate": 4.0121587040717846e-05,
+      "loss": 1.1618,
+      "step": 13391
+    },
+    {
+      "epoch": 2.14272,
+      "grad_norm": 1.995303323655967,
+      "learning_rate": 4.0107753137737634e-05,
+      "loss": 1.1541,
+      "step": 13392
+    },
+    {
+      "epoch": 2.14288,
+      "grad_norm": 1.7063682986277569,
+      "learning_rate": 4.009392102185473e-05,
+      "loss": 1.119,
+      "step": 13393
+    },
+    {
+      "epoch": 2.14304,
+      "grad_norm": 1.9881312196243073,
+      "learning_rate": 4.0080090693481976e-05,
+      "loss": 1.1287,
+      "step": 13394
+    },
+    {
+      "epoch": 2.1432,
+      "grad_norm": 2.0069634718916576,
+      "learning_rate": 4.006626215303199e-05,
+      "loss": 1.1197,
+      "step": 13395
+    },
+    {
+      "epoch": 2.14336,
+      "grad_norm": 1.8416265318226712,
+      "learning_rate": 4.005243540091741e-05,
+      "loss": 1.1646,
+      "step": 13396
+    },
+    {
+      "epoch": 2.14352,
+      "grad_norm": 1.9524733318343899,
+      "learning_rate": 4.0038610437550814e-05,
+      "loss": 1.1177,
+      "step": 13397
+    },
+    {
+      "epoch": 2.14368,
+      "grad_norm": 2.0992303202325964,
+      "learning_rate": 4.00247872633447e-05,
+      "loss": 1.1682,
+      "step": 13398
+    },
+    {
+      "epoch": 2.14384,
+      "grad_norm": 1.671627610775067,
+      "learning_rate": 4.001096587871155e-05,
+      "loss": 1.1057,
+      "step": 13399
+    },
+    {
+      "epoch": 2.144,
+      "grad_norm": 1.8320422152996052,
+      "learning_rate": 3.999714628406376e-05,
+      "loss": 1.0881,
+      "step": 13400
+    },
+    {
+      "epoch": 2.14416,
+      "grad_norm": 1.853279083280283,
+      "learning_rate": 3.998332847981369e-05,
+      "loss": 1.0965,
+      "step": 13401
+    },
+    {
+      "epoch": 2.14432,
+      "grad_norm": 1.876073157515853,
+      "learning_rate": 3.996951246637365e-05,
+      "loss": 1.1832,
+      "step": 13402
+    },
+    {
+      "epoch": 2.14448,
+      "grad_norm": 1.87359229137846,
+      "learning_rate": 3.995569824415584e-05,
+      "loss": 1.1414,
+      "step": 13403
+    },
+    {
+      "epoch": 2.14464,
+      "grad_norm": 1.898430927549399,
+      "learning_rate": 3.9941885813572574e-05,
+      "loss": 1.1083,
+      "step": 13404
+    },
+    {
+      "epoch": 2.1448,
+      "grad_norm": 1.7435864468451425,
+      "learning_rate": 3.992807517503585e-05,
+      "loss": 1.1172,
+      "step": 13405
+    },
+    {
+      "epoch": 2.14496,
+      "grad_norm": 1.8967607034366816,
+      "learning_rate": 3.991426632895791e-05,
+      "loss": 1.1482,
+      "step": 13406
+    },
+    {
+      "epoch": 2.14512,
+      "grad_norm": 2.078293986015007,
+      "learning_rate": 3.9900459275750624e-05,
+      "loss": 1.1458,
+      "step": 13407
+    },
+    {
+      "epoch": 2.14528,
+      "grad_norm": 1.8314918977104895,
+      "learning_rate": 3.988665401582614e-05,
+      "loss": 1.1005,
+      "step": 13408
+    },
+    {
+      "epoch": 2.14544,
+      "grad_norm": 1.907841151936977,
+      "learning_rate": 3.987285054959623e-05,
+      "loss": 1.1217,
+      "step": 13409
+    },
+    {
+      "epoch": 2.1456,
+      "grad_norm": 1.9256563299523743,
+      "learning_rate": 3.985904887747288e-05,
+      "loss": 1.1768,
+      "step": 13410
+    },
+    {
+      "epoch": 2.14576,
+      "grad_norm": 1.7165611435007793,
+      "learning_rate": 3.984524899986788e-05,
+      "loss": 1.1245,
+      "step": 13411
+    },
+    {
+      "epoch": 2.14592,
+      "grad_norm": 1.6974573548195149,
+      "learning_rate": 3.9831450917192994e-05,
+      "loss": 1.1266,
+      "step": 13412
+    },
+    {
+      "epoch": 2.14608,
+      "grad_norm": 1.876170305918239,
+      "learning_rate": 3.981765462985994e-05,
+      "loss": 1.1446,
+      "step": 13413
+    },
+    {
+      "epoch": 2.14624,
+      "grad_norm": 1.8968995659426806,
+      "learning_rate": 3.980386013828037e-05,
+      "loss": 1.1195,
+      "step": 13414
+    },
+    {
+      "epoch": 2.1464,
+      "grad_norm": 1.791187057433684,
+      "learning_rate": 3.97900674428659e-05,
+      "loss": 1.1407,
+      "step": 13415
+    },
+    {
+      "epoch": 2.14656,
+      "grad_norm": 1.959565011663683,
+      "learning_rate": 3.977627654402809e-05,
+      "loss": 1.1258,
+      "step": 13416
+    },
+    {
+      "epoch": 2.14672,
+      "grad_norm": 1.7732746394708694,
+      "learning_rate": 3.976248744217842e-05,
+      "loss": 1.0941,
+      "step": 13417
+    },
+    {
+      "epoch": 2.14688,
+      "grad_norm": 1.9903988612365886,
+      "learning_rate": 3.974870013772836e-05,
+      "loss": 1.0777,
+      "step": 13418
+    },
+    {
+      "epoch": 2.14704,
+      "grad_norm": 2.006435663769099,
+      "learning_rate": 3.9734914631089246e-05,
+      "loss": 1.1088,
+      "step": 13419
+    },
+    {
+      "epoch": 2.1471999999999998,
+      "grad_norm": 1.956652034170063,
+      "learning_rate": 3.972113092267255e-05,
+      "loss": 1.1029,
+      "step": 13420
+    },
+    {
+      "epoch": 2.14736,
+      "grad_norm": 1.625616214121025,
+      "learning_rate": 3.9707349012889394e-05,
+      "loss": 1.1718,
+      "step": 13421
+    },
+    {
+      "epoch": 2.14752,
+      "grad_norm": 2.0009574647636006,
+      "learning_rate": 3.969356890215116e-05,
+      "loss": 1.1271,
+      "step": 13422
+    },
+    {
+      "epoch": 2.14768,
+      "grad_norm": 2.0986937105750965,
+      "learning_rate": 3.9679790590868885e-05,
+      "loss": 1.1206,
+      "step": 13423
+    },
+    {
+      "epoch": 2.14784,
+      "grad_norm": 1.8442008430616355,
+      "learning_rate": 3.9666014079453826e-05,
+      "loss": 1.1141,
+      "step": 13424
+    },
+    {
+      "epoch": 2.148,
+      "grad_norm": 1.980158064590901,
+      "learning_rate": 3.965223936831692e-05,
+      "loss": 1.1383,
+      "step": 13425
+    },
+    {
+      "epoch": 2.14816,
+      "grad_norm": 1.7310339892743098,
+      "learning_rate": 3.9638466457869294e-05,
+      "loss": 1.1121,
+      "step": 13426
+    },
+    {
+      "epoch": 2.14832,
+      "grad_norm": 1.9032359200089601,
+      "learning_rate": 3.962469534852188e-05,
+      "loss": 1.1123,
+      "step": 13427
+    },
+    {
+      "epoch": 2.14848,
+      "grad_norm": 1.8654627319425063,
+      "learning_rate": 3.961092604068557e-05,
+      "loss": 1.1159,
+      "step": 13428
+    },
+    {
+      "epoch": 2.14864,
+      "grad_norm": 1.8365387848912025,
+      "learning_rate": 3.959715853477124e-05,
+      "loss": 1.1571,
+      "step": 13429
+    },
+    {
+      "epoch": 2.1488,
+      "grad_norm": 1.6203164051824477,
+      "learning_rate": 3.958339283118968e-05,
+      "loss": 1.1057,
+      "step": 13430
+    },
+    {
+      "epoch": 2.14896,
+      "grad_norm": 1.7384028609578035,
+      "learning_rate": 3.956962893035163e-05,
+      "loss": 1.1467,
+      "step": 13431
+    },
+    {
+      "epoch": 2.14912,
+      "grad_norm": 1.599616596347503,
+      "learning_rate": 3.955586683266781e-05,
+      "loss": 1.0906,
+      "step": 13432
+    },
+    {
+      "epoch": 2.14928,
+      "grad_norm": 1.8262005030020358,
+      "learning_rate": 3.954210653854885e-05,
+      "loss": 1.1406,
+      "step": 13433
+    },
+    {
+      "epoch": 2.14944,
+      "grad_norm": 1.7508573287063744,
+      "learning_rate": 3.952834804840533e-05,
+      "loss": 1.1704,
+      "step": 13434
+    },
+    {
+      "epoch": 2.1496,
+      "grad_norm": 1.8041347760414457,
+      "learning_rate": 3.951459136264775e-05,
+      "loss": 1.1403,
+      "step": 13435
+    },
+    {
+      "epoch": 2.14976,
+      "grad_norm": 1.863277555289344,
+      "learning_rate": 3.9500836481686697e-05,
+      "loss": 1.1121,
+      "step": 13436
+    },
+    {
+      "epoch": 2.14992,
+      "grad_norm": 1.8872144154182728,
+      "learning_rate": 3.948708340593246e-05,
+      "loss": 1.1256,
+      "step": 13437
+    },
+    {
+      "epoch": 2.15008,
+      "grad_norm": 1.657960549858527,
+      "learning_rate": 3.947333213579554e-05,
+      "loss": 1.1133,
+      "step": 13438
+    },
+    {
+      "epoch": 2.15024,
+      "grad_norm": 1.9193502198172823,
+      "learning_rate": 3.9459582671686124e-05,
+      "loss": 1.1114,
+      "step": 13439
+    },
+    {
+      "epoch": 2.1504,
+      "grad_norm": 1.6779670412927237,
+      "learning_rate": 3.944583501401462e-05,
+      "loss": 1.0943,
+      "step": 13440
+    },
+    {
+      "epoch": 2.15056,
+      "grad_norm": 1.6758911867877826,
+      "learning_rate": 3.943208916319109e-05,
+      "loss": 1.1114,
+      "step": 13441
+    },
+    {
+      "epoch": 2.15072,
+      "grad_norm": 1.649112227090345,
+      "learning_rate": 3.94183451196258e-05,
+      "loss": 1.1312,
+      "step": 13442
+    },
+    {
+      "epoch": 2.15088,
+      "grad_norm": 1.6612257190711794,
+      "learning_rate": 3.940460288372881e-05,
+      "loss": 1.161,
+      "step": 13443
+    },
+    {
+      "epoch": 2.15104,
+      "grad_norm": 1.4707017238449265,
+      "learning_rate": 3.9390862455910183e-05,
+      "loss": 1.1296,
+      "step": 13444
+    },
+    {
+      "epoch": 2.1512000000000002,
+      "grad_norm": 1.8857213292083856,
+      "learning_rate": 3.9377123836579885e-05,
+      "loss": 1.1649,
+      "step": 13445
+    },
+    {
+      "epoch": 2.15136,
+      "grad_norm": 2.0240657138392883,
+      "learning_rate": 3.936338702614789e-05,
+      "loss": 1.1353,
+      "step": 13446
+    },
+    {
+      "epoch": 2.15152,
+      "grad_norm": 1.989061229360138,
+      "learning_rate": 3.9349652025024066e-05,
+      "loss": 1.134,
+      "step": 13447
+    },
+    {
+      "epoch": 2.15168,
+      "grad_norm": 1.7053434331978958,
+      "learning_rate": 3.9335918833618215e-05,
+      "loss": 1.1196,
+      "step": 13448
+    },
+    {
+      "epoch": 2.15184,
+      "grad_norm": 1.6677233107767981,
+      "learning_rate": 3.932218745234022e-05,
+      "loss": 1.1346,
+      "step": 13449
+    },
+    {
+      "epoch": 2.152,
+      "grad_norm": 2.514218539415607,
+      "learning_rate": 3.930845788159966e-05,
+      "loss": 1.1259,
+      "step": 13450
+    },
+    {
+      "epoch": 2.15216,
+      "grad_norm": 1.888576386946842,
+      "learning_rate": 3.929473012180632e-05,
+      "loss": 1.1089,
+      "step": 13451
+    },
+    {
+      "epoch": 2.15232,
+      "grad_norm": 2.0600330733596124,
+      "learning_rate": 3.928100417336977e-05,
+      "loss": 1.143,
+      "step": 13452
+    },
+    {
+      "epoch": 2.15248,
+      "grad_norm": 2.0482916614312567,
+      "learning_rate": 3.926728003669958e-05,
+      "loss": 1.1163,
+      "step": 13453
+    },
+    {
+      "epoch": 2.15264,
+      "grad_norm": 2.1202294823811414,
+      "learning_rate": 3.925355771220526e-05,
+      "loss": 1.1373,
+      "step": 13454
+    },
+    {
+      "epoch": 2.1528,
+      "grad_norm": 1.9143623754206613,
+      "learning_rate": 3.923983720029626e-05,
+      "loss": 1.1465,
+      "step": 13455
+    },
+    {
+      "epoch": 2.15296,
+      "grad_norm": 1.913881664939006,
+      "learning_rate": 3.922611850138198e-05,
+      "loss": 1.1048,
+      "step": 13456
+    },
+    {
+      "epoch": 2.15312,
+      "grad_norm": 2.004195519662234,
+      "learning_rate": 3.921240161587176e-05,
+      "loss": 1.1118,
+      "step": 13457
+    },
+    {
+      "epoch": 2.15328,
+      "grad_norm": 1.8943693222771254,
+      "learning_rate": 3.919868654417491e-05,
+      "loss": 1.1038,
+      "step": 13458
+    },
+    {
+      "epoch": 2.15344,
+      "grad_norm": 1.9496562962820796,
+      "learning_rate": 3.9184973286700654e-05,
+      "loss": 1.1252,
+      "step": 13459
+    },
+    {
+      "epoch": 2.1536,
+      "grad_norm": 1.7930017638760873,
+      "learning_rate": 3.917126184385814e-05,
+      "loss": 1.1876,
+      "step": 13460
+    },
+    {
+      "epoch": 2.15376,
+      "grad_norm": 1.9994612601906523,
+      "learning_rate": 3.915755221605661e-05,
+      "loss": 1.1212,
+      "step": 13461
+    },
+    {
+      "epoch": 2.15392,
+      "grad_norm": 2.016531938682038,
+      "learning_rate": 3.9143844403705e-05,
+      "loss": 1.0932,
+      "step": 13462
+    },
+    {
+      "epoch": 2.15408,
+      "grad_norm": 2.009727607897262,
+      "learning_rate": 3.9130138407212455e-05,
+      "loss": 1.1378,
+      "step": 13463
+    },
+    {
+      "epoch": 2.15424,
+      "grad_norm": 1.6314594835044736,
+      "learning_rate": 3.911643422698782e-05,
+      "loss": 1.0755,
+      "step": 13464
+    },
+    {
+      "epoch": 2.1544,
+      "grad_norm": 1.9789455342391786,
+      "learning_rate": 3.910273186344014e-05,
+      "loss": 1.156,
+      "step": 13465
+    },
+    {
+      "epoch": 2.15456,
+      "grad_norm": 1.6465414199919686,
+      "learning_rate": 3.908903131697813e-05,
+      "loss": 1.0774,
+      "step": 13466
+    },
+    {
+      "epoch": 2.15472,
+      "grad_norm": 1.762318667973258,
+      "learning_rate": 3.90753325880107e-05,
+      "loss": 1.1384,
+      "step": 13467
+    },
+    {
+      "epoch": 2.15488,
+      "grad_norm": 2.2048760437726336,
+      "learning_rate": 3.9061635676946585e-05,
+      "loss": 1.1226,
+      "step": 13468
+    },
+    {
+      "epoch": 2.15504,
+      "grad_norm": 1.7902367686594176,
+      "learning_rate": 3.9047940584194455e-05,
+      "loss": 1.1356,
+      "step": 13469
+    },
+    {
+      "epoch": 2.1552,
+      "grad_norm": 2.203915455526219,
+      "learning_rate": 3.9034247310162955e-05,
+      "loss": 1.0942,
+      "step": 13470
+    },
+    {
+      "epoch": 2.15536,
+      "grad_norm": 1.8821927714929145,
+      "learning_rate": 3.902055585526069e-05,
+      "loss": 1.1024,
+      "step": 13471
+    },
+    {
+      "epoch": 2.15552,
+      "grad_norm": 1.85781611256323,
+      "learning_rate": 3.900686621989618e-05,
+      "loss": 1.0753,
+      "step": 13472
+    },
+    {
+      "epoch": 2.15568,
+      "grad_norm": 1.9635679568335025,
+      "learning_rate": 3.89931784044779e-05,
+      "loss": 1.1538,
+      "step": 13473
+    },
+    {
+      "epoch": 2.15584,
+      "grad_norm": 2.026987941494708,
+      "learning_rate": 3.897949240941429e-05,
+      "loss": 1.117,
+      "step": 13474
+    },
+    {
+      "epoch": 2.156,
+      "grad_norm": 1.9031182256490786,
+      "learning_rate": 3.89658082351137e-05,
+      "loss": 1.1184,
+      "step": 13475
+    },
+    {
+      "epoch": 2.15616,
+      "grad_norm": 1.8201281146049508,
+      "learning_rate": 3.895212588198441e-05,
+      "loss": 1.1422,
+      "step": 13476
+    },
+    {
+      "epoch": 2.15632,
+      "grad_norm": 1.9745969344588745,
+      "learning_rate": 3.893844535043481e-05,
+      "loss": 1.1455,
+      "step": 13477
+    },
+    {
+      "epoch": 2.15648,
+      "grad_norm": 1.7880359032988018,
+      "learning_rate": 3.8924766640872965e-05,
+      "loss": 1.0831,
+      "step": 13478
+    },
+    {
+      "epoch": 2.15664,
+      "grad_norm": 1.734296193611604,
+      "learning_rate": 3.891108975370715e-05,
+      "loss": 1.1479,
+      "step": 13479
+    },
+    {
+      "epoch": 2.1568,
+      "grad_norm": 1.8196748857356941,
+      "learning_rate": 3.889741468934533e-05,
+      "loss": 1.1319,
+      "step": 13480
+    },
+    {
+      "epoch": 2.1569599999999998,
+      "grad_norm": 1.7630603538109952,
+      "learning_rate": 3.888374144819569e-05,
+      "loss": 1.1803,
+      "step": 13481
+    },
+    {
+      "epoch": 2.15712,
+      "grad_norm": 1.7329501975294006,
+      "learning_rate": 3.8870070030666086e-05,
+      "loss": 1.1039,
+      "step": 13482
+    },
+    {
+      "epoch": 2.15728,
+      "grad_norm": 1.796758563736949,
+      "learning_rate": 3.885640043716454e-05,
+      "loss": 1.0941,
+      "step": 13483
+    },
+    {
+      "epoch": 2.15744,
+      "grad_norm": 1.9578829629494117,
+      "learning_rate": 3.884273266809891e-05,
+      "loss": 1.1388,
+      "step": 13484
+    },
+    {
+      "epoch": 2.1576,
+      "grad_norm": 1.8523287201734562,
+      "learning_rate": 3.882906672387703e-05,
+      "loss": 1.0644,
+      "step": 13485
+    },
+    {
+      "epoch": 2.15776,
+      "grad_norm": 1.845094556086755,
+      "learning_rate": 3.881540260490666e-05,
+      "loss": 1.1177,
+      "step": 13486
+    },
+    {
+      "epoch": 2.15792,
+      "grad_norm": 1.920029030708955,
+      "learning_rate": 3.880174031159551e-05,
+      "loss": 1.1265,
+      "step": 13487
+    },
+    {
+      "epoch": 2.15808,
+      "grad_norm": 2.0599750294865804,
+      "learning_rate": 3.8788079844351266e-05,
+      "loss": 1.1261,
+      "step": 13488
+    },
+    {
+      "epoch": 2.15824,
+      "grad_norm": 1.997859198215956,
+      "learning_rate": 3.8774421203581526e-05,
+      "loss": 1.1369,
+      "step": 13489
+    },
+    {
+      "epoch": 2.1584,
+      "grad_norm": 1.8314691908641574,
+      "learning_rate": 3.8760764389693836e-05,
+      "loss": 1.1053,
+      "step": 13490
+    },
+    {
+      "epoch": 2.15856,
+      "grad_norm": 1.9248168331025655,
+      "learning_rate": 3.874710940309571e-05,
+      "loss": 1.139,
+      "step": 13491
+    },
+    {
+      "epoch": 2.15872,
+      "grad_norm": 2.0020567509897838,
+      "learning_rate": 3.873345624419454e-05,
+      "loss": 1.1071,
+      "step": 13492
+    },
+    {
+      "epoch": 2.15888,
+      "grad_norm": 1.8350889308504672,
+      "learning_rate": 3.871980491339783e-05,
+      "loss": 1.0932,
+      "step": 13493
+    },
+    {
+      "epoch": 2.15904,
+      "grad_norm": 1.7290586938670767,
+      "learning_rate": 3.870615541111279e-05,
+      "loss": 1.1021,
+      "step": 13494
+    },
+    {
+      "epoch": 2.1592000000000002,
+      "grad_norm": 1.6859993361621821,
+      "learning_rate": 3.869250773774681e-05,
+      "loss": 1.1229,
+      "step": 13495
+    },
+    {
+      "epoch": 2.15936,
+      "grad_norm": 1.744334969684078,
+      "learning_rate": 3.867886189370701e-05,
+      "loss": 1.0889,
+      "step": 13496
+    },
+    {
+      "epoch": 2.15952,
+      "grad_norm": 1.7197376824111654,
+      "learning_rate": 3.8665217879400675e-05,
+      "loss": 1.135,
+      "step": 13497
+    },
+    {
+      "epoch": 2.15968,
+      "grad_norm": 1.7287256569619422,
+      "learning_rate": 3.8651575695234797e-05,
+      "loss": 1.0618,
+      "step": 13498
+    },
+    {
+      "epoch": 2.15984,
+      "grad_norm": 1.8105304976134309,
+      "learning_rate": 3.863793534161655e-05,
+      "loss": 1.117,
+      "step": 13499
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 1.855211674384679,
+      "learning_rate": 3.862429681895289e-05,
+      "loss": 1.1372,
+      "step": 13500
+    },
+    {
+      "epoch": 2.16016,
+      "grad_norm": 2.0339901643438605,
+      "learning_rate": 3.8610660127650776e-05,
+      "loss": 1.1529,
+      "step": 13501
+    },
+    {
+      "epoch": 2.16032,
+      "grad_norm": 2.0867853016661004,
+      "learning_rate": 3.859702526811712e-05,
+      "loss": 1.1048,
+      "step": 13502
+    },
+    {
+      "epoch": 2.16048,
+      "grad_norm": 1.9408126675933244,
+      "learning_rate": 3.858339224075875e-05,
+      "loss": 1.1457,
+      "step": 13503
+    },
+    {
+      "epoch": 2.16064,
+      "grad_norm": 2.123582955108023,
+      "learning_rate": 3.856976104598247e-05,
+      "loss": 1.1383,
+      "step": 13504
+    },
+    {
+      "epoch": 2.1608,
+      "grad_norm": 1.9676318745490966,
+      "learning_rate": 3.855613168419496e-05,
+      "loss": 1.18,
+      "step": 13505
+    },
+    {
+      "epoch": 2.16096,
+      "grad_norm": 1.9502651011419438,
+      "learning_rate": 3.854250415580303e-05,
+      "loss": 1.1314,
+      "step": 13506
+    },
+    {
+      "epoch": 2.16112,
+      "grad_norm": 1.963376605503434,
+      "learning_rate": 3.8528878461213136e-05,
+      "loss": 1.1879,
+      "step": 13507
+    },
+    {
+      "epoch": 2.16128,
+      "grad_norm": 1.8291409736299793,
+      "learning_rate": 3.851525460083199e-05,
+      "loss": 1.1044,
+      "step": 13508
+    },
+    {
+      "epoch": 2.16144,
+      "grad_norm": 1.9329322043069626,
+      "learning_rate": 3.8501632575066035e-05,
+      "loss": 1.1101,
+      "step": 13509
+    },
+    {
+      "epoch": 2.1616,
+      "grad_norm": 1.9031162971803317,
+      "learning_rate": 3.848801238432177e-05,
+      "loss": 1.124,
+      "step": 13510
+    },
+    {
+      "epoch": 2.16176,
+      "grad_norm": 1.571060996003012,
+      "learning_rate": 3.847439402900557e-05,
+      "loss": 1.1247,
+      "step": 13511
+    },
+    {
+      "epoch": 2.16192,
+      "grad_norm": 1.6592944194485078,
+      "learning_rate": 3.8460777509523806e-05,
+      "loss": 1.1105,
+      "step": 13512
+    },
+    {
+      "epoch": 2.16208,
+      "grad_norm": 1.7219308352735163,
+      "learning_rate": 3.844716282628277e-05,
+      "loss": 1.1508,
+      "step": 13513
+    },
+    {
+      "epoch": 2.16224,
+      "grad_norm": 1.7746500767194953,
+      "learning_rate": 3.84335499796887e-05,
+      "loss": 1.1368,
+      "step": 13514
+    },
+    {
+      "epoch": 2.1624,
+      "grad_norm": 1.979420449510644,
+      "learning_rate": 3.841993897014778e-05,
+      "loss": 1.1171,
+      "step": 13515
+    },
+    {
+      "epoch": 2.16256,
+      "grad_norm": 1.7957627200665485,
+      "learning_rate": 3.840632979806615e-05,
+      "loss": 1.1495,
+      "step": 13516
+    },
+    {
+      "epoch": 2.16272,
+      "grad_norm": 1.98957948729489,
+      "learning_rate": 3.839272246384991e-05,
+      "loss": 1.1002,
+      "step": 13517
+    },
+    {
+      "epoch": 2.16288,
+      "grad_norm": 1.6537106868704397,
+      "learning_rate": 3.837911696790504e-05,
+      "loss": 1.1451,
+      "step": 13518
+    },
+    {
+      "epoch": 2.16304,
+      "grad_norm": 2.1455799766469257,
+      "learning_rate": 3.836551331063749e-05,
+      "loss": 1.1697,
+      "step": 13519
+    },
+    {
+      "epoch": 2.1632,
+      "grad_norm": 1.7719708807812975,
+      "learning_rate": 3.83519114924533e-05,
+      "loss": 1.1746,
+      "step": 13520
+    },
+    {
+      "epoch": 2.16336,
+      "grad_norm": 1.8525169078717065,
+      "learning_rate": 3.833831151375815e-05,
+      "loss": 1.111,
+      "step": 13521
+    },
+    {
+      "epoch": 2.16352,
+      "grad_norm": 1.854404635375237,
+      "learning_rate": 3.8324713374958024e-05,
+      "loss": 1.138,
+      "step": 13522
+    },
+    {
+      "epoch": 2.16368,
+      "grad_norm": 1.986845747302309,
+      "learning_rate": 3.831111707645849e-05,
+      "loss": 1.1177,
+      "step": 13523
+    },
+    {
+      "epoch": 2.16384,
+      "grad_norm": 1.834531536138945,
+      "learning_rate": 3.829752261866538e-05,
+      "loss": 1.0671,
+      "step": 13524
+    },
+    {
+      "epoch": 2.164,
+      "grad_norm": 1.752850502671773,
+      "learning_rate": 3.828393000198428e-05,
+      "loss": 1.06,
+      "step": 13525
+    },
+    {
+      "epoch": 2.16416,
+      "grad_norm": 1.7740973675418807,
+      "learning_rate": 3.8270339226820785e-05,
+      "loss": 1.0823,
+      "step": 13526
+    },
+    {
+      "epoch": 2.16432,
+      "grad_norm": 1.744100870993831,
+      "learning_rate": 3.825675029358041e-05,
+      "loss": 1.0874,
+      "step": 13527
+    },
+    {
+      "epoch": 2.16448,
+      "grad_norm": 1.7151699324836016,
+      "learning_rate": 3.824316320266865e-05,
+      "loss": 1.1363,
+      "step": 13528
+    },
+    {
+      "epoch": 2.16464,
+      "grad_norm": 1.8527846628530926,
+      "learning_rate": 3.82295779544909e-05,
+      "loss": 1.1383,
+      "step": 13529
+    },
+    {
+      "epoch": 2.1648,
+      "grad_norm": 1.9163126443228307,
+      "learning_rate": 3.821599454945254e-05,
+      "loss": 1.1511,
+      "step": 13530
+    },
+    {
+      "epoch": 2.1649599999999998,
+      "grad_norm": 1.822709628139173,
+      "learning_rate": 3.820241298795887e-05,
+      "loss": 1.1408,
+      "step": 13531
+    },
+    {
+      "epoch": 2.16512,
+      "grad_norm": 1.7065797564425678,
+      "learning_rate": 3.8188833270415146e-05,
+      "loss": 1.0848,
+      "step": 13532
+    },
+    {
+      "epoch": 2.16528,
+      "grad_norm": 2.1075162125060056,
+      "learning_rate": 3.817525539722654e-05,
+      "loss": 1.1257,
+      "step": 13533
+    },
+    {
+      "epoch": 2.16544,
+      "grad_norm": 1.8705038601570758,
+      "learning_rate": 3.816167936879829e-05,
+      "loss": 1.1554,
+      "step": 13534
+    },
+    {
+      "epoch": 2.1656,
+      "grad_norm": 1.8465170858127635,
+      "learning_rate": 3.814810518553535e-05,
+      "loss": 1.0981,
+      "step": 13535
+    },
+    {
+      "epoch": 2.16576,
+      "grad_norm": 1.9725292613963032,
+      "learning_rate": 3.8134532847842893e-05,
+      "loss": 1.1388,
+      "step": 13536
+    },
+    {
+      "epoch": 2.16592,
+      "grad_norm": 1.9558376186024256,
+      "learning_rate": 3.8120962356125766e-05,
+      "loss": 1.0996,
+      "step": 13537
+    },
+    {
+      "epoch": 2.16608,
+      "grad_norm": 1.6957253520311963,
+      "learning_rate": 3.810739371078902e-05,
+      "loss": 1.1173,
+      "step": 13538
+    },
+    {
+      "epoch": 2.16624,
+      "grad_norm": 1.9007467267469682,
+      "learning_rate": 3.809382691223738e-05,
+      "loss": 1.1153,
+      "step": 13539
+    },
+    {
+      "epoch": 2.1664,
+      "grad_norm": 1.9878988914929023,
+      "learning_rate": 3.808026196087578e-05,
+      "loss": 1.1115,
+      "step": 13540
+    },
+    {
+      "epoch": 2.16656,
+      "grad_norm": 2.0119047559089958,
+      "learning_rate": 3.806669885710894e-05,
+      "loss": 1.1269,
+      "step": 13541
+    },
+    {
+      "epoch": 2.16672,
+      "grad_norm": 1.8296488955699157,
+      "learning_rate": 3.805313760134156e-05,
+      "loss": 1.13,
+      "step": 13542
+    },
+    {
+      "epoch": 2.16688,
+      "grad_norm": 1.9552584153040922,
+      "learning_rate": 3.803957819397828e-05,
+      "loss": 1.1082,
+      "step": 13543
+    },
+    {
+      "epoch": 2.16704,
+      "grad_norm": 1.949349240916335,
+      "learning_rate": 3.8026020635423706e-05,
+      "loss": 1.1368,
+      "step": 13544
+    },
+    {
+      "epoch": 2.1672,
+      "grad_norm": 1.8137196972351632,
+      "learning_rate": 3.801246492608237e-05,
+      "loss": 1.1439,
+      "step": 13545
+    },
+    {
+      "epoch": 2.16736,
+      "grad_norm": 1.8258011486015864,
+      "learning_rate": 3.799891106635875e-05,
+      "loss": 1.1008,
+      "step": 13546
+    },
+    {
+      "epoch": 2.16752,
+      "grad_norm": 2.0238568425085317,
+      "learning_rate": 3.798535905665729e-05,
+      "loss": 1.1385,
+      "step": 13547
+    },
+    {
+      "epoch": 2.16768,
+      "grad_norm": 1.8664842188383441,
+      "learning_rate": 3.7971808897382344e-05,
+      "loss": 1.0777,
+      "step": 13548
+    },
+    {
+      "epoch": 2.16784,
+      "grad_norm": 1.8790533157507439,
+      "learning_rate": 3.7958260588938196e-05,
+      "loss": 1.1406,
+      "step": 13549
+    },
+    {
+      "epoch": 2.168,
+      "grad_norm": 1.7075850318337351,
+      "learning_rate": 3.794471413172923e-05,
+      "loss": 1.1453,
+      "step": 13550
+    },
+    {
+      "epoch": 2.16816,
+      "grad_norm": 1.8245146169607809,
+      "learning_rate": 3.7931169526159495e-05,
+      "loss": 1.1737,
+      "step": 13551
+    },
+    {
+      "epoch": 2.16832,
+      "grad_norm": 1.9868881480949443,
+      "learning_rate": 3.791762677263329e-05,
+      "loss": 1.1394,
+      "step": 13552
+    },
+    {
+      "epoch": 2.16848,
+      "grad_norm": 1.5775504088957153,
+      "learning_rate": 3.7904085871554565e-05,
+      "loss": 1.1383,
+      "step": 13553
+    },
+    {
+      "epoch": 2.16864,
+      "grad_norm": 1.948240924433846,
+      "learning_rate": 3.78905468233275e-05,
+      "loss": 1.1298,
+      "step": 13554
+    },
+    {
+      "epoch": 2.1688,
+      "grad_norm": 1.71718484101623,
+      "learning_rate": 3.7877009628355943e-05,
+      "loss": 1.1623,
+      "step": 13555
+    },
+    {
+      "epoch": 2.16896,
+      "grad_norm": 1.8061468207806495,
+      "learning_rate": 3.7863474287043945e-05,
+      "loss": 1.1112,
+      "step": 13556
+    },
+    {
+      "epoch": 2.16912,
+      "grad_norm": 1.9775455346928754,
+      "learning_rate": 3.784994079979532e-05,
+      "loss": 1.089,
+      "step": 13557
+    },
+    {
+      "epoch": 2.16928,
+      "grad_norm": 2.1679999349862675,
+      "learning_rate": 3.78364091670139e-05,
+      "loss": 1.1534,
+      "step": 13558
+    },
+    {
+      "epoch": 2.16944,
+      "grad_norm": 1.9459195264535472,
+      "learning_rate": 3.782287938910345e-05,
+      "loss": 1.1392,
+      "step": 13559
+    },
+    {
+      "epoch": 2.1696,
+      "grad_norm": 2.0361028559428447,
+      "learning_rate": 3.780935146646768e-05,
+      "loss": 1.1549,
+      "step": 13560
+    },
+    {
+      "epoch": 2.16976,
+      "grad_norm": 2.0222245282727465,
+      "learning_rate": 3.779582539951024e-05,
+      "loss": 1.1244,
+      "step": 13561
+    },
+    {
+      "epoch": 2.16992,
+      "grad_norm": 1.8817062888795564,
+      "learning_rate": 3.778230118863473e-05,
+      "loss": 1.0934,
+      "step": 13562
+    },
+    {
+      "epoch": 2.17008,
+      "grad_norm": 1.82553342697076,
+      "learning_rate": 3.776877883424469e-05,
+      "loss": 1.1483,
+      "step": 13563
+    },
+    {
+      "epoch": 2.17024,
+      "grad_norm": 1.8778466492804349,
+      "learning_rate": 3.775525833674357e-05,
+      "loss": 1.1653,
+      "step": 13564
+    },
+    {
+      "epoch": 2.1704,
+      "grad_norm": 1.978268764126892,
+      "learning_rate": 3.774173969653488e-05,
+      "loss": 1.101,
+      "step": 13565
+    },
+    {
+      "epoch": 2.17056,
+      "grad_norm": 1.789368961827698,
+      "learning_rate": 3.772822291402196e-05,
+      "loss": 1.1162,
+      "step": 13566
+    },
+    {
+      "epoch": 2.17072,
+      "grad_norm": 1.6937294214351288,
+      "learning_rate": 3.771470798960812e-05,
+      "loss": 1.0636,
+      "step": 13567
+    },
+    {
+      "epoch": 2.17088,
+      "grad_norm": 1.6041722911386018,
+      "learning_rate": 3.770119492369665e-05,
+      "loss": 1.1691,
+      "step": 13568
+    },
+    {
+      "epoch": 2.17104,
+      "grad_norm": 1.6377783697151302,
+      "learning_rate": 3.768768371669073e-05,
+      "loss": 1.1147,
+      "step": 13569
+    },
+    {
+      "epoch": 2.1712,
+      "grad_norm": 1.8674590045047101,
+      "learning_rate": 3.767417436899354e-05,
+      "loss": 1.093,
+      "step": 13570
+    },
+    {
+      "epoch": 2.17136,
+      "grad_norm": 2.0613679691641726,
+      "learning_rate": 3.766066688100817e-05,
+      "loss": 1.0985,
+      "step": 13571
+    },
+    {
+      "epoch": 2.17152,
+      "grad_norm": 1.7561010647189115,
+      "learning_rate": 3.764716125313764e-05,
+      "loss": 1.1451,
+      "step": 13572
+    },
+    {
+      "epoch": 2.17168,
+      "grad_norm": 2.14463099859104,
+      "learning_rate": 3.763365748578498e-05,
+      "loss": 1.1521,
+      "step": 13573
+    },
+    {
+      "epoch": 2.17184,
+      "grad_norm": 1.6778138246531469,
+      "learning_rate": 3.76201555793531e-05,
+      "loss": 1.0916,
+      "step": 13574
+    },
+    {
+      "epoch": 2.172,
+      "grad_norm": 1.6241034013808546,
+      "learning_rate": 3.760665553424489e-05,
+      "loss": 1.0774,
+      "step": 13575
+    },
+    {
+      "epoch": 2.17216,
+      "grad_norm": 1.8218836124132678,
+      "learning_rate": 3.7593157350863104e-05,
+      "loss": 1.1138,
+      "step": 13576
+    },
+    {
+      "epoch": 2.17232,
+      "grad_norm": 1.5760973620502308,
+      "learning_rate": 3.7579661029610654e-05,
+      "loss": 1.1069,
+      "step": 13577
+    },
+    {
+      "epoch": 2.17248,
+      "grad_norm": 1.761696259816349,
+      "learning_rate": 3.756616657089008e-05,
+      "loss": 1.1249,
+      "step": 13578
+    },
+    {
+      "epoch": 2.17264,
+      "grad_norm": 1.9167182463783785,
+      "learning_rate": 3.7552673975104203e-05,
+      "loss": 1.1072,
+      "step": 13579
+    },
+    {
+      "epoch": 2.1728,
+      "grad_norm": 1.8070127117548755,
+      "learning_rate": 3.753918324265546e-05,
+      "loss": 1.1215,
+      "step": 13580
+    },
+    {
+      "epoch": 2.17296,
+      "grad_norm": 1.8681208862331335,
+      "learning_rate": 3.7525694373946516e-05,
+      "loss": 1.0927,
+      "step": 13581
+    },
+    {
+      "epoch": 2.17312,
+      "grad_norm": 1.908255051522941,
+      "learning_rate": 3.7512207369379815e-05,
+      "loss": 1.1293,
+      "step": 13582
+    },
+    {
+      "epoch": 2.17328,
+      "grad_norm": 1.9553471382305456,
+      "learning_rate": 3.749872222935779e-05,
+      "loss": 1.0894,
+      "step": 13583
+    },
+    {
+      "epoch": 2.17344,
+      "grad_norm": 1.858535620952099,
+      "learning_rate": 3.748523895428282e-05,
+      "loss": 1.2194,
+      "step": 13584
+    },
+    {
+      "epoch": 2.1736,
+      "grad_norm": 1.9522834662732063,
+      "learning_rate": 3.747175754455723e-05,
+      "loss": 1.115,
+      "step": 13585
+    },
+    {
+      "epoch": 2.17376,
+      "grad_norm": 1.9740017669300225,
+      "learning_rate": 3.745827800058328e-05,
+      "loss": 1.1284,
+      "step": 13586
+    },
+    {
+      "epoch": 2.17392,
+      "grad_norm": 1.7921129654578007,
+      "learning_rate": 3.7444800322763176e-05,
+      "loss": 1.0759,
+      "step": 13587
+    },
+    {
+      "epoch": 2.17408,
+      "grad_norm": 1.931843076159031,
+      "learning_rate": 3.743132451149909e-05,
+      "loss": 1.1523,
+      "step": 13588
+    },
+    {
+      "epoch": 2.17424,
+      "grad_norm": 1.7559111882126102,
+      "learning_rate": 3.741785056719309e-05,
+      "loss": 1.13,
+      "step": 13589
+    },
+    {
+      "epoch": 2.1744,
+      "grad_norm": 1.7010410311691313,
+      "learning_rate": 3.740437849024725e-05,
+      "loss": 1.1068,
+      "step": 13590
+    },
+    {
+      "epoch": 2.17456,
+      "grad_norm": 1.821550890863256,
+      "learning_rate": 3.739090828106354e-05,
+      "loss": 1.0852,
+      "step": 13591
+    },
+    {
+      "epoch": 2.1747199999999998,
+      "grad_norm": 2.0863642664038555,
+      "learning_rate": 3.7377439940043866e-05,
+      "loss": 1.1932,
+      "step": 13592
+    },
+    {
+      "epoch": 2.17488,
+      "grad_norm": 1.8462636903875822,
+      "learning_rate": 3.7363973467590196e-05,
+      "loss": 1.1402,
+      "step": 13593
+    },
+    {
+      "epoch": 2.17504,
+      "grad_norm": 1.9712724520291587,
+      "learning_rate": 3.735050886410423e-05,
+      "loss": 1.1567,
+      "step": 13594
+    },
+    {
+      "epoch": 2.1752,
+      "grad_norm": 1.9194584603859255,
+      "learning_rate": 3.733704612998784e-05,
+      "loss": 1.1137,
+      "step": 13595
+    },
+    {
+      "epoch": 2.17536,
+      "grad_norm": 1.9002669105297838,
+      "learning_rate": 3.732358526564263e-05,
+      "loss": 1.0982,
+      "step": 13596
+    },
+    {
+      "epoch": 2.17552,
+      "grad_norm": 1.728809661947192,
+      "learning_rate": 3.731012627147034e-05,
+      "loss": 1.1652,
+      "step": 13597
+    },
+    {
+      "epoch": 2.17568,
+      "grad_norm": 1.8953039686013522,
+      "learning_rate": 3.729666914787253e-05,
+      "loss": 1.0659,
+      "step": 13598
+    },
+    {
+      "epoch": 2.17584,
+      "grad_norm": 1.7612608039397408,
+      "learning_rate": 3.7283213895250736e-05,
+      "loss": 1.1191,
+      "step": 13599
+    },
+    {
+      "epoch": 2.176,
+      "grad_norm": 1.9375113295059956,
+      "learning_rate": 3.726976051400646e-05,
+      "loss": 1.085,
+      "step": 13600
+    },
+    {
+      "epoch": 2.17616,
+      "grad_norm": 1.8518942777633354,
+      "learning_rate": 3.725630900454112e-05,
+      "loss": 1.1243,
+      "step": 13601
+    },
+    {
+      "epoch": 2.17632,
+      "grad_norm": 2.025221050417636,
+      "learning_rate": 3.72428593672561e-05,
+      "loss": 1.1463,
+      "step": 13602
+    },
+    {
+      "epoch": 2.17648,
+      "grad_norm": 1.647016780608129,
+      "learning_rate": 3.72294116025527e-05,
+      "loss": 1.0858,
+      "step": 13603
+    },
+    {
+      "epoch": 2.17664,
+      "grad_norm": 1.79389713016904,
+      "learning_rate": 3.72159657108322e-05,
+      "loss": 1.1233,
+      "step": 13604
+    },
+    {
+      "epoch": 2.1768,
+      "grad_norm": 1.8048288503709722,
+      "learning_rate": 3.720252169249578e-05,
+      "loss": 1.0955,
+      "step": 13605
+    },
+    {
+      "epoch": 2.1769600000000002,
+      "grad_norm": 2.113187488095717,
+      "learning_rate": 3.7189079547944626e-05,
+      "loss": 1.088,
+      "step": 13606
+    },
+    {
+      "epoch": 2.17712,
+      "grad_norm": 1.812441668891399,
+      "learning_rate": 3.717563927757981e-05,
+      "loss": 1.0734,
+      "step": 13607
+    },
+    {
+      "epoch": 2.17728,
+      "grad_norm": 1.7591857398204613,
+      "learning_rate": 3.7162200881802337e-05,
+      "loss": 1.1278,
+      "step": 13608
+    },
+    {
+      "epoch": 2.17744,
+      "grad_norm": 1.8585023083993713,
+      "learning_rate": 3.714876436101329e-05,
+      "loss": 1.1035,
+      "step": 13609
+    },
+    {
+      "epoch": 2.1776,
+      "grad_norm": 1.8373703514030602,
+      "learning_rate": 3.7135329715613464e-05,
+      "loss": 1.069,
+      "step": 13610
+    },
+    {
+      "epoch": 2.17776,
+      "grad_norm": 1.9814896703882208,
+      "learning_rate": 3.7121896946003865e-05,
+      "loss": 1.1601,
+      "step": 13611
+    },
+    {
+      "epoch": 2.17792,
+      "grad_norm": 2.1013911647683603,
+      "learning_rate": 3.7108466052585167e-05,
+      "loss": 1.0977,
+      "step": 13612
+    },
+    {
+      "epoch": 2.17808,
+      "grad_norm": 1.8616130730502534,
+      "learning_rate": 3.709503703575823e-05,
+      "loss": 1.1346,
+      "step": 13613
+    },
+    {
+      "epoch": 2.17824,
+      "grad_norm": 1.8293981290538386,
+      "learning_rate": 3.708160989592373e-05,
+      "loss": 1.1064,
+      "step": 13614
+    },
+    {
+      "epoch": 2.1784,
+      "grad_norm": 2.2842295754185526,
+      "learning_rate": 3.7068184633482316e-05,
+      "loss": 1.1129,
+      "step": 13615
+    },
+    {
+      "epoch": 2.17856,
+      "grad_norm": 2.1493302091259037,
+      "learning_rate": 3.7054761248834566e-05,
+      "loss": 1.1168,
+      "step": 13616
+    },
+    {
+      "epoch": 2.17872,
+      "grad_norm": 1.8514575681226735,
+      "learning_rate": 3.704133974238102e-05,
+      "loss": 1.1117,
+      "step": 13617
+    },
+    {
+      "epoch": 2.17888,
+      "grad_norm": 1.8256924449465572,
+      "learning_rate": 3.7027920114522154e-05,
+      "loss": 1.1173,
+      "step": 13618
+    },
+    {
+      "epoch": 2.17904,
+      "grad_norm": 2.438862982356466,
+      "learning_rate": 3.7014502365658385e-05,
+      "loss": 1.1801,
+      "step": 13619
+    },
+    {
+      "epoch": 2.1792,
+      "grad_norm": 2.2082937961381317,
+      "learning_rate": 3.7001086496190093e-05,
+      "loss": 1.1521,
+      "step": 13620
+    },
+    {
+      "epoch": 2.17936,
+      "grad_norm": 2.134612705647295,
+      "learning_rate": 3.698767250651758e-05,
+      "loss": 1.1886,
+      "step": 13621
+    },
+    {
+      "epoch": 2.17952,
+      "grad_norm": 1.9623796293918245,
+      "learning_rate": 3.697426039704107e-05,
+      "loss": 1.1054,
+      "step": 13622
+    },
+    {
+      "epoch": 2.17968,
+      "grad_norm": 2.1814259223067043,
+      "learning_rate": 3.696085016816082e-05,
+      "loss": 1.1204,
+      "step": 13623
+    },
+    {
+      "epoch": 2.17984,
+      "grad_norm": 1.6757072387613905,
+      "learning_rate": 3.694744182027695e-05,
+      "loss": 1.0987,
+      "step": 13624
+    },
+    {
+      "epoch": 2.18,
+      "grad_norm": 1.8396549642577749,
+      "learning_rate": 3.693403535378953e-05,
+      "loss": 1.0825,
+      "step": 13625
+    },
+    {
+      "epoch": 2.18016,
+      "grad_norm": 2.533877659289989,
+      "learning_rate": 3.692063076909861e-05,
+      "loss": 1.1542,
+      "step": 13626
+    },
+    {
+      "epoch": 2.18032,
+      "grad_norm": 2.0874256832089566,
+      "learning_rate": 3.690722806660414e-05,
+      "loss": 1.1531,
+      "step": 13627
+    },
+    {
+      "epoch": 2.18048,
+      "grad_norm": 1.867573266231174,
+      "learning_rate": 3.689382724670606e-05,
+      "loss": 1.1373,
+      "step": 13628
+    },
+    {
+      "epoch": 2.18064,
+      "grad_norm": 1.8381056465976706,
+      "learning_rate": 3.688042830980421e-05,
+      "loss": 1.1379,
+      "step": 13629
+    },
+    {
+      "epoch": 2.1808,
+      "grad_norm": 1.6418643044327381,
+      "learning_rate": 3.68670312562984e-05,
+      "loss": 1.1097,
+      "step": 13630
+    },
+    {
+      "epoch": 2.18096,
+      "grad_norm": 1.8874289948694272,
+      "learning_rate": 3.68536360865884e-05,
+      "loss": 1.0904,
+      "step": 13631
+    },
+    {
+      "epoch": 2.18112,
+      "grad_norm": 1.6187401330236486,
+      "learning_rate": 3.684024280107386e-05,
+      "loss": 1.1185,
+      "step": 13632
+    },
+    {
+      "epoch": 2.18128,
+      "grad_norm": 1.8827903530421082,
+      "learning_rate": 3.682685140015443e-05,
+      "loss": 1.1082,
+      "step": 13633
+    },
+    {
+      "epoch": 2.18144,
+      "grad_norm": 1.8286327899769488,
+      "learning_rate": 3.681346188422976e-05,
+      "loss": 1.1338,
+      "step": 13634
+    },
+    {
+      "epoch": 2.1816,
+      "grad_norm": 1.6363441638881604,
+      "learning_rate": 3.680007425369926e-05,
+      "loss": 1.1086,
+      "step": 13635
+    },
+    {
+      "epoch": 2.18176,
+      "grad_norm": 1.8156131705175589,
+      "learning_rate": 3.67866885089625e-05,
+      "loss": 1.1325,
+      "step": 13636
+    },
+    {
+      "epoch": 2.18192,
+      "grad_norm": 1.7857933060364326,
+      "learning_rate": 3.6773304650418785e-05,
+      "loss": 1.0951,
+      "step": 13637
+    },
+    {
+      "epoch": 2.18208,
+      "grad_norm": 1.8758036990567692,
+      "learning_rate": 3.6759922678467564e-05,
+      "loss": 1.1198,
+      "step": 13638
+    },
+    {
+      "epoch": 2.18224,
+      "grad_norm": 2.044917591118037,
+      "learning_rate": 3.674654259350809e-05,
+      "loss": 1.1602,
+      "step": 13639
+    },
+    {
+      "epoch": 2.1824,
+      "grad_norm": 1.8084607005952804,
+      "learning_rate": 3.6733164395939626e-05,
+      "loss": 1.101,
+      "step": 13640
+    },
+    {
+      "epoch": 2.18256,
+      "grad_norm": 2.103880533661203,
+      "learning_rate": 3.6719788086161344e-05,
+      "loss": 1.0797,
+      "step": 13641
+    },
+    {
+      "epoch": 2.1827199999999998,
+      "grad_norm": 1.8835801804929835,
+      "learning_rate": 3.670641366457238e-05,
+      "loss": 1.1518,
+      "step": 13642
+    },
+    {
+      "epoch": 2.18288,
+      "grad_norm": 1.7029473354667906,
+      "learning_rate": 3.669304113157179e-05,
+      "loss": 1.1004,
+      "step": 13643
+    },
+    {
+      "epoch": 2.18304,
+      "grad_norm": 1.7716113709582568,
+      "learning_rate": 3.6679670487558614e-05,
+      "loss": 1.1594,
+      "step": 13644
+    },
+    {
+      "epoch": 2.1832,
+      "grad_norm": 1.770307152316985,
+      "learning_rate": 3.666630173293181e-05,
+      "loss": 1.1416,
+      "step": 13645
+    },
+    {
+      "epoch": 2.18336,
+      "grad_norm": 1.854411461020954,
+      "learning_rate": 3.665293486809027e-05,
+      "loss": 1.0915,
+      "step": 13646
+    },
+    {
+      "epoch": 2.18352,
+      "grad_norm": 1.873202992994016,
+      "learning_rate": 3.6639569893432845e-05,
+      "loss": 1.1482,
+      "step": 13647
+    },
+    {
+      "epoch": 2.18368,
+      "grad_norm": 1.8860921911354682,
+      "learning_rate": 3.662620680935832e-05,
+      "loss": 1.1446,
+      "step": 13648
+    },
+    {
+      "epoch": 2.18384,
+      "grad_norm": 1.847812988690069,
+      "learning_rate": 3.661284561626542e-05,
+      "loss": 1.1348,
+      "step": 13649
+    },
+    {
+      "epoch": 2.184,
+      "grad_norm": 1.859899343987299,
+      "learning_rate": 3.65994863145529e-05,
+      "loss": 1.1202,
+      "step": 13650
+    },
+    {
+      "epoch": 2.18416,
+      "grad_norm": 1.8862724180803532,
+      "learning_rate": 3.658612890461926e-05,
+      "loss": 1.1603,
+      "step": 13651
+    },
+    {
+      "epoch": 2.18432,
+      "grad_norm": 1.799847870126666,
+      "learning_rate": 3.65727733868632e-05,
+      "loss": 1.1295,
+      "step": 13652
+    },
+    {
+      "epoch": 2.18448,
+      "grad_norm": 1.9652513278286914,
+      "learning_rate": 3.655941976168309e-05,
+      "loss": 1.0978,
+      "step": 13653
+    },
+    {
+      "epoch": 2.18464,
+      "grad_norm": 1.8008946054564245,
+      "learning_rate": 3.6546068029477476e-05,
+      "loss": 1.1307,
+      "step": 13654
+    },
+    {
+      "epoch": 2.1848,
+      "grad_norm": 1.7930427301256868,
+      "learning_rate": 3.653271819064474e-05,
+      "loss": 1.1463,
+      "step": 13655
+    },
+    {
+      "epoch": 2.1849600000000002,
+      "grad_norm": 2.0239823776644625,
+      "learning_rate": 3.651937024558321e-05,
+      "loss": 1.1035,
+      "step": 13656
+    },
+    {
+      "epoch": 2.18512,
+      "grad_norm": 1.8516440710341755,
+      "learning_rate": 3.650602419469117e-05,
+      "loss": 1.1359,
+      "step": 13657
+    },
+    {
+      "epoch": 2.18528,
+      "grad_norm": 2.2169888755715568,
+      "learning_rate": 3.649268003836685e-05,
+      "loss": 1.0871,
+      "step": 13658
+    },
+    {
+      "epoch": 2.18544,
+      "grad_norm": 1.6786638784619086,
+      "learning_rate": 3.647933777700842e-05,
+      "loss": 1.1018,
+      "step": 13659
+    },
+    {
+      "epoch": 2.1856,
+      "grad_norm": 1.8241648403280897,
+      "learning_rate": 3.6465997411013974e-05,
+      "loss": 1.1363,
+      "step": 13660
+    },
+    {
+      "epoch": 2.18576,
+      "grad_norm": 1.6986149690026704,
+      "learning_rate": 3.64526589407816e-05,
+      "loss": 1.0786,
+      "step": 13661
+    },
+    {
+      "epoch": 2.18592,
+      "grad_norm": 1.8886842199630216,
+      "learning_rate": 3.6439322366709296e-05,
+      "loss": 1.1054,
+      "step": 13662
+    },
+    {
+      "epoch": 2.18608,
+      "grad_norm": 1.890006784453863,
+      "learning_rate": 3.6425987689194976e-05,
+      "loss": 1.1549,
+      "step": 13663
+    },
+    {
+      "epoch": 2.18624,
+      "grad_norm": 1.8917619793107188,
+      "learning_rate": 3.6412654908636554e-05,
+      "loss": 1.1977,
+      "step": 13664
+    },
+    {
+      "epoch": 2.1864,
+      "grad_norm": 1.711441204847404,
+      "learning_rate": 3.639932402543181e-05,
+      "loss": 1.153,
+      "step": 13665
+    },
+    {
+      "epoch": 2.18656,
+      "grad_norm": 1.8298674728288769,
+      "learning_rate": 3.638599503997865e-05,
+      "loss": 1.1297,
+      "step": 13666
+    },
+    {
+      "epoch": 2.18672,
+      "grad_norm": 2.0500688085948635,
+      "learning_rate": 3.637266795267461e-05,
+      "loss": 1.0986,
+      "step": 13667
+    },
+    {
+      "epoch": 2.18688,
+      "grad_norm": 1.968905559013215,
+      "learning_rate": 3.6359342763917525e-05,
+      "loss": 1.073,
+      "step": 13668
+    },
+    {
+      "epoch": 2.18704,
+      "grad_norm": 1.9083239385092803,
+      "learning_rate": 3.634601947410485e-05,
+      "loss": 1.1579,
+      "step": 13669
+    },
+    {
+      "epoch": 2.1872,
+      "grad_norm": 1.9926987023318832,
+      "learning_rate": 3.633269808363423e-05,
+      "loss": 1.1084,
+      "step": 13670
+    },
+    {
+      "epoch": 2.18736,
+      "grad_norm": 2.1202667365801915,
+      "learning_rate": 3.631937859290313e-05,
+      "loss": 1.083,
+      "step": 13671
+    },
+    {
+      "epoch": 2.18752,
+      "grad_norm": 1.9975635785045545,
+      "learning_rate": 3.630606100230898e-05,
+      "loss": 1.1373,
+      "step": 13672
+    },
+    {
+      "epoch": 2.18768,
+      "grad_norm": 2.054580616049187,
+      "learning_rate": 3.629274531224916e-05,
+      "loss": 1.0924,
+      "step": 13673
+    },
+    {
+      "epoch": 2.18784,
+      "grad_norm": 1.938626832281331,
+      "learning_rate": 3.627943152312098e-05,
+      "loss": 1.1227,
+      "step": 13674
+    },
+    {
+      "epoch": 2.188,
+      "grad_norm": 1.8637821517763886,
+      "learning_rate": 3.626611963532172e-05,
+      "loss": 1.1248,
+      "step": 13675
+    },
+    {
+      "epoch": 2.18816,
+      "grad_norm": 2.1641334704109108,
+      "learning_rate": 3.6252809649248576e-05,
+      "loss": 1.078,
+      "step": 13676
+    },
+    {
+      "epoch": 2.18832,
+      "grad_norm": 1.844752672959717,
+      "learning_rate": 3.62395015652987e-05,
+      "loss": 1.0787,
+      "step": 13677
+    },
+    {
+      "epoch": 2.18848,
+      "grad_norm": 1.9372078567223228,
+      "learning_rate": 3.62261953838692e-05,
+      "loss": 1.0968,
+      "step": 13678
+    },
+    {
+      "epoch": 2.18864,
+      "grad_norm": 1.8303575101672411,
+      "learning_rate": 3.621289110535709e-05,
+      "loss": 1.0696,
+      "step": 13679
+    },
+    {
+      "epoch": 2.1888,
+      "grad_norm": 1.77136317857246,
+      "learning_rate": 3.619958873015936e-05,
+      "loss": 1.1473,
+      "step": 13680
+    },
+    {
+      "epoch": 2.18896,
+      "grad_norm": 1.6426938126113253,
+      "learning_rate": 3.6186288258672915e-05,
+      "loss": 1.1097,
+      "step": 13681
+    },
+    {
+      "epoch": 2.18912,
+      "grad_norm": 1.9631010507809643,
+      "learning_rate": 3.617298969129467e-05,
+      "loss": 1.2098,
+      "step": 13682
+    },
+    {
+      "epoch": 2.18928,
+      "grad_norm": 1.8269083027758781,
+      "learning_rate": 3.6159693028421404e-05,
+      "loss": 1.1569,
+      "step": 13683
+    },
+    {
+      "epoch": 2.18944,
+      "grad_norm": 1.7921390607003032,
+      "learning_rate": 3.6146398270449866e-05,
+      "loss": 1.0903,
+      "step": 13684
+    },
+    {
+      "epoch": 2.1896,
+      "grad_norm": 2.0122036516160735,
+      "learning_rate": 3.613310541777677e-05,
+      "loss": 1.1584,
+      "step": 13685
+    },
+    {
+      "epoch": 2.18976,
+      "grad_norm": 1.8663664306745136,
+      "learning_rate": 3.611981447079874e-05,
+      "loss": 1.1028,
+      "step": 13686
+    },
+    {
+      "epoch": 2.18992,
+      "grad_norm": 1.8338569916019922,
+      "learning_rate": 3.6106525429912364e-05,
+      "loss": 1.1258,
+      "step": 13687
+    },
+    {
+      "epoch": 2.19008,
+      "grad_norm": 1.8998925023711284,
+      "learning_rate": 3.609323829551416e-05,
+      "loss": 1.141,
+      "step": 13688
+    },
+    {
+      "epoch": 2.19024,
+      "grad_norm": 1.7324440515574056,
+      "learning_rate": 3.6079953068000616e-05,
+      "loss": 1.0834,
+      "step": 13689
+    },
+    {
+      "epoch": 2.1904,
+      "grad_norm": 1.8205070437294122,
+      "learning_rate": 3.60666697477681e-05,
+      "loss": 1.1121,
+      "step": 13690
+    },
+    {
+      "epoch": 2.19056,
+      "grad_norm": 1.7569740584049216,
+      "learning_rate": 3.6053388335213055e-05,
+      "loss": 1.1642,
+      "step": 13691
+    },
+    {
+      "epoch": 2.19072,
+      "grad_norm": 1.7837829258470026,
+      "learning_rate": 3.604010883073166e-05,
+      "loss": 1.1523,
+      "step": 13692
+    },
+    {
+      "epoch": 2.19088,
+      "grad_norm": 1.8915863947033753,
+      "learning_rate": 3.6026831234720274e-05,
+      "loss": 1.1027,
+      "step": 13693
+    },
+    {
+      "epoch": 2.19104,
+      "grad_norm": 2.0775278654058567,
+      "learning_rate": 3.601355554757496e-05,
+      "loss": 1.0992,
+      "step": 13694
+    },
+    {
+      "epoch": 2.1912,
+      "grad_norm": 1.7325121711175338,
+      "learning_rate": 3.600028176969199e-05,
+      "loss": 1.1145,
+      "step": 13695
+    },
+    {
+      "epoch": 2.19136,
+      "grad_norm": 1.8623228205621076,
+      "learning_rate": 3.598700990146729e-05,
+      "loss": 1.1035,
+      "step": 13696
+    },
+    {
+      "epoch": 2.19152,
+      "grad_norm": 1.8049901765001084,
+      "learning_rate": 3.597373994329696e-05,
+      "loss": 1.1364,
+      "step": 13697
+    },
+    {
+      "epoch": 2.19168,
+      "grad_norm": 1.9053175598238534,
+      "learning_rate": 3.596047189557693e-05,
+      "loss": 1.0887,
+      "step": 13698
+    },
+    {
+      "epoch": 2.19184,
+      "grad_norm": 1.6741146949366403,
+      "learning_rate": 3.5947205758703105e-05,
+      "loss": 1.1433,
+      "step": 13699
+    },
+    {
+      "epoch": 2.192,
+      "grad_norm": 1.7311872405298048,
+      "learning_rate": 3.593394153307133e-05,
+      "loss": 1.0704,
+      "step": 13700
+    },
+    {
+      "epoch": 2.19216,
+      "grad_norm": 2.118343428131631,
+      "learning_rate": 3.592067921907739e-05,
+      "loss": 1.1251,
+      "step": 13701
+    },
+    {
+      "epoch": 2.19232,
+      "grad_norm": 1.8282872722462653,
+      "learning_rate": 3.5907418817117e-05,
+      "loss": 1.1071,
+      "step": 13702
+    },
+    {
+      "epoch": 2.19248,
+      "grad_norm": 1.813786642531233,
+      "learning_rate": 3.589416032758585e-05,
+      "loss": 1.1113,
+      "step": 13703
+    },
+    {
+      "epoch": 2.19264,
+      "grad_norm": 1.6219750274960925,
+      "learning_rate": 3.5880903750879535e-05,
+      "loss": 1.0812,
+      "step": 13704
+    },
+    {
+      "epoch": 2.1928,
+      "grad_norm": 2.19390866586306,
+      "learning_rate": 3.5867649087393616e-05,
+      "loss": 1.1268,
+      "step": 13705
+    },
+    {
+      "epoch": 2.19296,
+      "grad_norm": 1.7519091608273796,
+      "learning_rate": 3.585439633752358e-05,
+      "loss": 1.101,
+      "step": 13706
+    },
+    {
+      "epoch": 2.19312,
+      "grad_norm": 2.1907178583237887,
+      "learning_rate": 3.584114550166494e-05,
+      "loss": 1.0844,
+      "step": 13707
+    },
+    {
+      "epoch": 2.19328,
+      "grad_norm": 1.8811981923661407,
+      "learning_rate": 3.582789658021297e-05,
+      "loss": 1.1025,
+      "step": 13708
+    },
+    {
+      "epoch": 2.19344,
+      "grad_norm": 2.0555425112034005,
+      "learning_rate": 3.581464957356312e-05,
+      "loss": 1.104,
+      "step": 13709
+    },
+    {
+      "epoch": 2.1936,
+      "grad_norm": 1.7552206355621462,
+      "learning_rate": 3.580140448211054e-05,
+      "loss": 1.1251,
+      "step": 13710
+    },
+    {
+      "epoch": 2.19376,
+      "grad_norm": 1.77067990441132,
+      "learning_rate": 3.578816130625055e-05,
+      "loss": 1.1319,
+      "step": 13711
+    },
+    {
+      "epoch": 2.19392,
+      "grad_norm": 1.9667176011263279,
+      "learning_rate": 3.5774920046378215e-05,
+      "loss": 1.175,
+      "step": 13712
+    },
+    {
+      "epoch": 2.19408,
+      "grad_norm": 1.8794901690219663,
+      "learning_rate": 3.57616807028887e-05,
+      "loss": 1.2011,
+      "step": 13713
+    },
+    {
+      "epoch": 2.19424,
+      "grad_norm": 1.822361592866605,
+      "learning_rate": 3.574844327617703e-05,
+      "loss": 1.1226,
+      "step": 13714
+    },
+    {
+      "epoch": 2.1944,
+      "grad_norm": 1.8619047008052976,
+      "learning_rate": 3.573520776663819e-05,
+      "loss": 1.1278,
+      "step": 13715
+    },
+    {
+      "epoch": 2.19456,
+      "grad_norm": 1.8971706261227543,
+      "learning_rate": 3.5721974174667115e-05,
+      "loss": 1.1122,
+      "step": 13716
+    },
+    {
+      "epoch": 2.1947200000000002,
+      "grad_norm": 1.8935787900708956,
+      "learning_rate": 3.570874250065865e-05,
+      "loss": 1.1292,
+      "step": 13717
+    },
+    {
+      "epoch": 2.19488,
+      "grad_norm": 1.8043595686215317,
+      "learning_rate": 3.569551274500764e-05,
+      "loss": 1.0844,
+      "step": 13718
+    },
+    {
+      "epoch": 2.19504,
+      "grad_norm": 1.9365598719767203,
+      "learning_rate": 3.568228490810883e-05,
+      "loss": 1.0632,
+      "step": 13719
+    },
+    {
+      "epoch": 2.1952,
+      "grad_norm": 1.6137089012312422,
+      "learning_rate": 3.5669058990356904e-05,
+      "loss": 1.1259,
+      "step": 13720
+    },
+    {
+      "epoch": 2.19536,
+      "grad_norm": 1.623079388443219,
+      "learning_rate": 3.565583499214653e-05,
+      "loss": 1.1366,
+      "step": 13721
+    },
+    {
+      "epoch": 2.19552,
+      "grad_norm": 1.8883686138871933,
+      "learning_rate": 3.5642612913872245e-05,
+      "loss": 1.1547,
+      "step": 13722
+    },
+    {
+      "epoch": 2.19568,
+      "grad_norm": 1.9098363281658568,
+      "learning_rate": 3.562939275592868e-05,
+      "loss": 1.1312,
+      "step": 13723
+    },
+    {
+      "epoch": 2.19584,
+      "grad_norm": 1.9298450186174636,
+      "learning_rate": 3.5616174518710165e-05,
+      "loss": 1.105,
+      "step": 13724
+    },
+    {
+      "epoch": 2.196,
+      "grad_norm": 1.6940553900114375,
+      "learning_rate": 3.560295820261126e-05,
+      "loss": 1.1442,
+      "step": 13725
+    },
+    {
+      "epoch": 2.19616,
+      "grad_norm": 1.8103012151512414,
+      "learning_rate": 3.5589743808026165e-05,
+      "loss": 1.1463,
+      "step": 13726
+    },
+    {
+      "epoch": 2.19632,
+      "grad_norm": 1.7082027415879342,
+      "learning_rate": 3.557653133534931e-05,
+      "loss": 1.1063,
+      "step": 13727
+    },
+    {
+      "epoch": 2.19648,
+      "grad_norm": 1.8105893357708314,
+      "learning_rate": 3.5563320784974875e-05,
+      "loss": 1.1249,
+      "step": 13728
+    },
+    {
+      "epoch": 2.19664,
+      "grad_norm": 1.7805317315361349,
+      "learning_rate": 3.5550112157297044e-05,
+      "loss": 1.0988,
+      "step": 13729
+    },
+    {
+      "epoch": 2.1968,
+      "grad_norm": 1.712028053846014,
+      "learning_rate": 3.553690545270997e-05,
+      "loss": 1.1072,
+      "step": 13730
+    },
+    {
+      "epoch": 2.19696,
+      "grad_norm": 1.8058172291256145,
+      "learning_rate": 3.5523700671607694e-05,
+      "loss": 1.0652,
+      "step": 13731
+    },
+    {
+      "epoch": 2.19712,
+      "grad_norm": 2.013719672631666,
+      "learning_rate": 3.551049781438425e-05,
+      "loss": 1.068,
+      "step": 13732
+    },
+    {
+      "epoch": 2.19728,
+      "grad_norm": 1.8277941186507278,
+      "learning_rate": 3.549729688143356e-05,
+      "loss": 1.1,
+      "step": 13733
+    },
+    {
+      "epoch": 2.19744,
+      "grad_norm": 1.755032147370505,
+      "learning_rate": 3.5484097873149555e-05,
+      "loss": 1.1357,
+      "step": 13734
+    },
+    {
+      "epoch": 2.1976,
+      "grad_norm": 1.8477154052689404,
+      "learning_rate": 3.547090078992607e-05,
+      "loss": 1.1154,
+      "step": 13735
+    },
+    {
+      "epoch": 2.19776,
+      "grad_norm": 1.907038391867785,
+      "learning_rate": 3.5457705632156855e-05,
+      "loss": 1.1129,
+      "step": 13736
+    },
+    {
+      "epoch": 2.19792,
+      "grad_norm": 1.7529113828761287,
+      "learning_rate": 3.544451240023566e-05,
+      "loss": 1.107,
+      "step": 13737
+    },
+    {
+      "epoch": 2.19808,
+      "grad_norm": 1.8229743151715245,
+      "learning_rate": 3.543132109455612e-05,
+      "loss": 1.1147,
+      "step": 13738
+    },
+    {
+      "epoch": 2.19824,
+      "grad_norm": 1.6021271471816132,
+      "learning_rate": 3.54181317155119e-05,
+      "loss": 1.065,
+      "step": 13739
+    },
+    {
+      "epoch": 2.1984,
+      "grad_norm": 1.906008104736456,
+      "learning_rate": 3.5404944263496545e-05,
+      "loss": 1.1517,
+      "step": 13740
+    },
+    {
+      "epoch": 2.19856,
+      "grad_norm": 1.7744612524681365,
+      "learning_rate": 3.53917587389035e-05,
+      "loss": 1.1033,
+      "step": 13741
+    },
+    {
+      "epoch": 2.19872,
+      "grad_norm": 1.7008621902124061,
+      "learning_rate": 3.5378575142126245e-05,
+      "loss": 1.1333,
+      "step": 13742
+    },
+    {
+      "epoch": 2.19888,
+      "grad_norm": 1.744314310018782,
+      "learning_rate": 3.5365393473558136e-05,
+      "loss": 1.1225,
+      "step": 13743
+    },
+    {
+      "epoch": 2.19904,
+      "grad_norm": 1.693277711844562,
+      "learning_rate": 3.535221373359251e-05,
+      "loss": 1.145,
+      "step": 13744
+    },
+    {
+      "epoch": 2.1992,
+      "grad_norm": 1.9559706206140748,
+      "learning_rate": 3.533903592262263e-05,
+      "loss": 1.1217,
+      "step": 13745
+    },
+    {
+      "epoch": 2.19936,
+      "grad_norm": 1.9998941385343092,
+      "learning_rate": 3.532586004104169e-05,
+      "loss": 1.1287,
+      "step": 13746
+    },
+    {
+      "epoch": 2.19952,
+      "grad_norm": 1.8019417656413674,
+      "learning_rate": 3.531268608924281e-05,
+      "loss": 1.1291,
+      "step": 13747
+    },
+    {
+      "epoch": 2.19968,
+      "grad_norm": 1.7439221978695063,
+      "learning_rate": 3.529951406761919e-05,
+      "loss": 1.1663,
+      "step": 13748
+    },
+    {
+      "epoch": 2.19984,
+      "grad_norm": 1.9491102920238341,
+      "learning_rate": 3.528634397656372e-05,
+      "loss": 1.1086,
+      "step": 13749
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6716818825008857,
+      "learning_rate": 3.5273175816469516e-05,
+      "loss": 1.1168,
+      "step": 13750
+    },
+    {
+      "epoch": 2.20016,
+      "grad_norm": 1.6329148954613282,
+      "learning_rate": 3.526000958772936e-05,
+      "loss": 1.1061,
+      "step": 13751
+    },
+    {
+      "epoch": 2.20032,
+      "grad_norm": 1.744931468423266,
+      "learning_rate": 3.524684529073625e-05,
+      "loss": 1.0781,
+      "step": 13752
+    },
+    {
+      "epoch": 2.2004799999999998,
+      "grad_norm": 1.8238535135289666,
+      "learning_rate": 3.523368292588286e-05,
+      "loss": 1.1014,
+      "step": 13753
+    },
+    {
+      "epoch": 2.20064,
+      "grad_norm": 1.6896052254512735,
+      "learning_rate": 3.5220522493562025e-05,
+      "loss": 1.1014,
+      "step": 13754
+    },
+    {
+      "epoch": 2.2008,
+      "grad_norm": 1.7284435930538398,
+      "learning_rate": 3.520736399416641e-05,
+      "loss": 1.0722,
+      "step": 13755
+    },
+    {
+      "epoch": 2.20096,
+      "grad_norm": 1.7326732508343823,
+      "learning_rate": 3.519420742808863e-05,
+      "loss": 1.0967,
+      "step": 13756
+    },
+    {
+      "epoch": 2.20112,
+      "grad_norm": 1.6765117129407747,
+      "learning_rate": 3.518105279572128e-05,
+      "loss": 1.0583,
+      "step": 13757
+    },
+    {
+      "epoch": 2.20128,
+      "grad_norm": 1.8981231070747058,
+      "learning_rate": 3.5167900097456866e-05,
+      "loss": 1.1474,
+      "step": 13758
+    },
+    {
+      "epoch": 2.20144,
+      "grad_norm": 1.8424001943778454,
+      "learning_rate": 3.515474933368783e-05,
+      "loss": 1.1567,
+      "step": 13759
+    },
+    {
+      "epoch": 2.2016,
+      "grad_norm": 1.9155597347337934,
+      "learning_rate": 3.514160050480658e-05,
+      "loss": 1.1598,
+      "step": 13760
+    },
+    {
+      "epoch": 2.20176,
+      "grad_norm": 2.135134776167753,
+      "learning_rate": 3.512845361120548e-05,
+      "loss": 1.1377,
+      "step": 13761
+    },
+    {
+      "epoch": 2.20192,
+      "grad_norm": 1.7971046996271665,
+      "learning_rate": 3.511530865327679e-05,
+      "loss": 1.0847,
+      "step": 13762
+    },
+    {
+      "epoch": 2.20208,
+      "grad_norm": 1.860562004820456,
+      "learning_rate": 3.510216563141271e-05,
+      "loss": 1.1283,
+      "step": 13763
+    },
+    {
+      "epoch": 2.20224,
+      "grad_norm": 1.8590390744880678,
+      "learning_rate": 3.50890245460055e-05,
+      "loss": 1.1041,
+      "step": 13764
+    },
+    {
+      "epoch": 2.2024,
+      "grad_norm": 1.6737906762416261,
+      "learning_rate": 3.507588539744715e-05,
+      "loss": 1.1287,
+      "step": 13765
+    },
+    {
+      "epoch": 2.20256,
+      "grad_norm": 2.278601209013381,
+      "learning_rate": 3.506274818612984e-05,
+      "loss": 1.1437,
+      "step": 13766
+    },
+    {
+      "epoch": 2.2027200000000002,
+      "grad_norm": 2.069290828486021,
+      "learning_rate": 3.504961291244544e-05,
+      "loss": 1.1324,
+      "step": 13767
+    },
+    {
+      "epoch": 2.20288,
+      "grad_norm": 2.283520782019977,
+      "learning_rate": 3.503647957678602e-05,
+      "loss": 1.1016,
+      "step": 13768
+    },
+    {
+      "epoch": 2.20304,
+      "grad_norm": 1.814277962955189,
+      "learning_rate": 3.5023348179543316e-05,
+      "loss": 1.1196,
+      "step": 13769
+    },
+    {
+      "epoch": 2.2032,
+      "grad_norm": 1.9131541118284774,
+      "learning_rate": 3.501021872110924e-05,
+      "loss": 1.0927,
+      "step": 13770
+    },
+    {
+      "epoch": 2.20336,
+      "grad_norm": 1.843499354366293,
+      "learning_rate": 3.499709120187555e-05,
+      "loss": 1.1599,
+      "step": 13771
+    },
+    {
+      "epoch": 2.20352,
+      "grad_norm": 2.037426639733691,
+      "learning_rate": 3.498396562223394e-05,
+      "loss": 1.1642,
+      "step": 13772
+    },
+    {
+      "epoch": 2.20368,
+      "grad_norm": 1.6192480879612565,
+      "learning_rate": 3.4970841982576053e-05,
+      "loss": 1.1193,
+      "step": 13773
+    },
+    {
+      "epoch": 2.20384,
+      "grad_norm": 1.7158125095489896,
+      "learning_rate": 3.49577202832935e-05,
+      "loss": 1.1415,
+      "step": 13774
+    },
+    {
+      "epoch": 2.204,
+      "grad_norm": 1.6260137899491294,
+      "learning_rate": 3.494460052477777e-05,
+      "loss": 1.1205,
+      "step": 13775
+    },
+    {
+      "epoch": 2.20416,
+      "grad_norm": 2.0737227106022216,
+      "learning_rate": 3.493148270742038e-05,
+      "loss": 1.159,
+      "step": 13776
+    },
+    {
+      "epoch": 2.20432,
+      "grad_norm": 1.7763432013919003,
+      "learning_rate": 3.491836683161273e-05,
+      "loss": 1.1097,
+      "step": 13777
+    },
+    {
+      "epoch": 2.20448,
+      "grad_norm": 1.9244640725020348,
+      "learning_rate": 3.490525289774618e-05,
+      "loss": 1.0987,
+      "step": 13778
+    },
+    {
+      "epoch": 2.20464,
+      "grad_norm": 1.9555749984229798,
+      "learning_rate": 3.4892140906212e-05,
+      "loss": 1.1203,
+      "step": 13779
+    },
+    {
+      "epoch": 2.2048,
+      "grad_norm": 1.6765255961380172,
+      "learning_rate": 3.4879030857401516e-05,
+      "loss": 1.0948,
+      "step": 13780
+    },
+    {
+      "epoch": 2.20496,
+      "grad_norm": 1.577776808506685,
+      "learning_rate": 3.486592275170581e-05,
+      "loss": 1.1106,
+      "step": 13781
+    },
+    {
+      "epoch": 2.20512,
+      "grad_norm": 1.777462350271684,
+      "learning_rate": 3.485281658951611e-05,
+      "loss": 1.0941,
+      "step": 13782
+    },
+    {
+      "epoch": 2.20528,
+      "grad_norm": 1.7023558495569648,
+      "learning_rate": 3.483971237122339e-05,
+      "loss": 1.1405,
+      "step": 13783
+    },
+    {
+      "epoch": 2.20544,
+      "grad_norm": 1.9086278390526037,
+      "learning_rate": 3.4826610097218746e-05,
+      "loss": 1.1301,
+      "step": 13784
+    },
+    {
+      "epoch": 2.2056,
+      "grad_norm": 1.6586678213648518,
+      "learning_rate": 3.481350976789304e-05,
+      "loss": 1.0616,
+      "step": 13785
+    },
+    {
+      "epoch": 2.20576,
+      "grad_norm": 1.9995336600418239,
+      "learning_rate": 3.480041138363723e-05,
+      "loss": 1.1429,
+      "step": 13786
+    },
+    {
+      "epoch": 2.20592,
+      "grad_norm": 1.7118847885445239,
+      "learning_rate": 3.4787314944842156e-05,
+      "loss": 1.0986,
+      "step": 13787
+    },
+    {
+      "epoch": 2.20608,
+      "grad_norm": 1.9709622866322658,
+      "learning_rate": 3.4774220451898574e-05,
+      "loss": 1.0913,
+      "step": 13788
+    },
+    {
+      "epoch": 2.20624,
+      "grad_norm": 1.621449472595123,
+      "learning_rate": 3.476112790519721e-05,
+      "loss": 1.0878,
+      "step": 13789
+    },
+    {
+      "epoch": 2.2064,
+      "grad_norm": 2.0541814639556475,
+      "learning_rate": 3.474803730512872e-05,
+      "loss": 1.1062,
+      "step": 13790
+    },
+    {
+      "epoch": 2.20656,
+      "grad_norm": 1.9657454626089663,
+      "learning_rate": 3.473494865208372e-05,
+      "loss": 1.1299,
+      "step": 13791
+    },
+    {
+      "epoch": 2.20672,
+      "grad_norm": 1.7862994385149222,
+      "learning_rate": 3.472186194645276e-05,
+      "loss": 1.1168,
+      "step": 13792
+    },
+    {
+      "epoch": 2.20688,
+      "grad_norm": 1.783512772476382,
+      "learning_rate": 3.470877718862631e-05,
+      "loss": 1.1311,
+      "step": 13793
+    },
+    {
+      "epoch": 2.20704,
+      "grad_norm": 1.8196426929306053,
+      "learning_rate": 3.46956943789948e-05,
+      "loss": 1.1104,
+      "step": 13794
+    },
+    {
+      "epoch": 2.2072,
+      "grad_norm": 2.300286237660115,
+      "learning_rate": 3.4682613517948593e-05,
+      "loss": 1.1159,
+      "step": 13795
+    },
+    {
+      "epoch": 2.20736,
+      "grad_norm": 1.983872430278381,
+      "learning_rate": 3.466953460587808e-05,
+      "loss": 1.1679,
+      "step": 13796
+    },
+    {
+      "epoch": 2.20752,
+      "grad_norm": 1.9054941734633677,
+      "learning_rate": 3.46564576431734e-05,
+      "loss": 1.1014,
+      "step": 13797
+    },
+    {
+      "epoch": 2.20768,
+      "grad_norm": 1.7928908531288872,
+      "learning_rate": 3.464338263022483e-05,
+      "loss": 1.1286,
+      "step": 13798
+    },
+    {
+      "epoch": 2.20784,
+      "grad_norm": 2.255387864433241,
+      "learning_rate": 3.463030956742249e-05,
+      "loss": 1.0907,
+      "step": 13799
+    },
+    {
+      "epoch": 2.208,
+      "grad_norm": 2.0705229729969403,
+      "learning_rate": 3.461723845515646e-05,
+      "loss": 1.1185,
+      "step": 13800
+    },
+    {
+      "epoch": 2.20816,
+      "grad_norm": 1.7851375986456934,
+      "learning_rate": 3.460416929381677e-05,
+      "loss": 1.1262,
+      "step": 13801
+    },
+    {
+      "epoch": 2.20832,
+      "grad_norm": 2.20497240526691,
+      "learning_rate": 3.459110208379337e-05,
+      "loss": 1.1397,
+      "step": 13802
+    },
+    {
+      "epoch": 2.2084799999999998,
+      "grad_norm": 2.1911030167126575,
+      "learning_rate": 3.457803682547617e-05,
+      "loss": 1.0884,
+      "step": 13803
+    },
+    {
+      "epoch": 2.20864,
+      "grad_norm": 2.200981063733549,
+      "learning_rate": 3.456497351925499e-05,
+      "loss": 1.1156,
+      "step": 13804
+    },
+    {
+      "epoch": 2.2088,
+      "grad_norm": 1.646059325526788,
+      "learning_rate": 3.4551912165519714e-05,
+      "loss": 1.1023,
+      "step": 13805
+    },
+    {
+      "epoch": 2.20896,
+      "grad_norm": 2.058909481492411,
+      "learning_rate": 3.453885276465996e-05,
+      "loss": 1.116,
+      "step": 13806
+    },
+    {
+      "epoch": 2.20912,
+      "grad_norm": 2.4075331224564964,
+      "learning_rate": 3.45257953170655e-05,
+      "loss": 1.1293,
+      "step": 13807
+    },
+    {
+      "epoch": 2.20928,
+      "grad_norm": 2.0603869837295514,
+      "learning_rate": 3.451273982312585e-05,
+      "loss": 1.1525,
+      "step": 13808
+    },
+    {
+      "epoch": 2.20944,
+      "grad_norm": 1.8040774326474591,
+      "learning_rate": 3.449968628323068e-05,
+      "loss": 1.1318,
+      "step": 13809
+    },
+    {
+      "epoch": 2.2096,
+      "grad_norm": 1.7386012346548665,
+      "learning_rate": 3.448663469776936e-05,
+      "loss": 1.1402,
+      "step": 13810
+    },
+    {
+      "epoch": 2.20976,
+      "grad_norm": 1.8884025363067043,
+      "learning_rate": 3.447358506713142e-05,
+      "loss": 1.0905,
+      "step": 13811
+    },
+    {
+      "epoch": 2.20992,
+      "grad_norm": 1.950715819693604,
+      "learning_rate": 3.446053739170624e-05,
+      "loss": 1.112,
+      "step": 13812
+    },
+    {
+      "epoch": 2.21008,
+      "grad_norm": 2.3841978948850473,
+      "learning_rate": 3.44474916718831e-05,
+      "loss": 1.0983,
+      "step": 13813
+    },
+    {
+      "epoch": 2.21024,
+      "grad_norm": 1.8532213614989037,
+      "learning_rate": 3.44344479080513e-05,
+      "loss": 1.1009,
+      "step": 13814
+    },
+    {
+      "epoch": 2.2104,
+      "grad_norm": 2.127656235181109,
+      "learning_rate": 3.442140610060004e-05,
+      "loss": 1.1004,
+      "step": 13815
+    },
+    {
+      "epoch": 2.21056,
+      "grad_norm": 1.8821376131089333,
+      "learning_rate": 3.440836624991847e-05,
+      "loss": 1.1597,
+      "step": 13816
+    },
+    {
+      "epoch": 2.21072,
+      "grad_norm": 1.6528491459880263,
+      "learning_rate": 3.439532835639566e-05,
+      "loss": 1.1214,
+      "step": 13817
+    },
+    {
+      "epoch": 2.21088,
+      "grad_norm": 2.119748710472671,
+      "learning_rate": 3.438229242042066e-05,
+      "loss": 1.0586,
+      "step": 13818
+    },
+    {
+      "epoch": 2.21104,
+      "grad_norm": 1.8640508340091146,
+      "learning_rate": 3.436925844238243e-05,
+      "loss": 1.147,
+      "step": 13819
+    },
+    {
+      "epoch": 2.2112,
+      "grad_norm": 1.5723900091148912,
+      "learning_rate": 3.435622642266988e-05,
+      "loss": 1.1387,
+      "step": 13820
+    },
+    {
+      "epoch": 2.21136,
+      "grad_norm": 1.8801135802440443,
+      "learning_rate": 3.434319636167194e-05,
+      "loss": 1.1,
+      "step": 13821
+    },
+    {
+      "epoch": 2.21152,
+      "grad_norm": 1.7582157629677344,
+      "learning_rate": 3.433016825977727e-05,
+      "loss": 1.1007,
+      "step": 13822
+    },
+    {
+      "epoch": 2.21168,
+      "grad_norm": 2.0518480446570746,
+      "learning_rate": 3.431714211737477e-05,
+      "loss": 1.0644,
+      "step": 13823
+    },
+    {
+      "epoch": 2.21184,
+      "grad_norm": 2.020428562076435,
+      "learning_rate": 3.4304117934852966e-05,
+      "loss": 1.174,
+      "step": 13824
+    },
+    {
+      "epoch": 2.212,
+      "grad_norm": 1.7533105460714848,
+      "learning_rate": 3.429109571260062e-05,
+      "loss": 1.0961,
+      "step": 13825
+    },
+    {
+      "epoch": 2.21216,
+      "grad_norm": 2.2890642768356395,
+      "learning_rate": 3.427807545100617e-05,
+      "loss": 1.099,
+      "step": 13826
+    },
+    {
+      "epoch": 2.21232,
+      "grad_norm": 2.0625294582070435,
+      "learning_rate": 3.42650571504582e-05,
+      "loss": 1.1197,
+      "step": 13827
+    },
+    {
+      "epoch": 2.2124800000000002,
+      "grad_norm": 1.785652709181677,
+      "learning_rate": 3.4252040811345145e-05,
+      "loss": 1.068,
+      "step": 13828
+    },
+    {
+      "epoch": 2.21264,
+      "grad_norm": 1.7968883396076594,
+      "learning_rate": 3.4239026434055386e-05,
+      "loss": 1.1035,
+      "step": 13829
+    },
+    {
+      "epoch": 2.2128,
+      "grad_norm": 1.9186862461510303,
+      "learning_rate": 3.422601401897726e-05,
+      "loss": 1.0918,
+      "step": 13830
+    },
+    {
+      "epoch": 2.21296,
+      "grad_norm": 2.0494300162072974,
+      "learning_rate": 3.421300356649904e-05,
+      "loss": 1.1249,
+      "step": 13831
+    },
+    {
+      "epoch": 2.21312,
+      "grad_norm": 1.6275468953626353,
+      "learning_rate": 3.419999507700892e-05,
+      "loss": 1.0907,
+      "step": 13832
+    },
+    {
+      "epoch": 2.21328,
+      "grad_norm": 2.055606629734252,
+      "learning_rate": 3.4186988550895074e-05,
+      "loss": 1.0519,
+      "step": 13833
+    },
+    {
+      "epoch": 2.21344,
+      "grad_norm": 1.9509180591245716,
+      "learning_rate": 3.417398398854558e-05,
+      "loss": 1.1311,
+      "step": 13834
+    },
+    {
+      "epoch": 2.2136,
+      "grad_norm": 1.6389787342980726,
+      "learning_rate": 3.416098139034849e-05,
+      "loss": 1.0665,
+      "step": 13835
+    },
+    {
+      "epoch": 2.21376,
+      "grad_norm": 2.0429272036870665,
+      "learning_rate": 3.4147980756691745e-05,
+      "loss": 1.1122,
+      "step": 13836
+    },
+    {
+      "epoch": 2.21392,
+      "grad_norm": 2.0662333349442505,
+      "learning_rate": 3.4134982087963364e-05,
+      "loss": 1.1129,
+      "step": 13837
+    },
+    {
+      "epoch": 2.21408,
+      "grad_norm": 1.826619144965938,
+      "learning_rate": 3.412198538455107e-05,
+      "loss": 1.1329,
+      "step": 13838
+    },
+    {
+      "epoch": 2.21424,
+      "grad_norm": 2.0595559451798473,
+      "learning_rate": 3.410899064684281e-05,
+      "loss": 1.1087,
+      "step": 13839
+    },
+    {
+      "epoch": 2.2144,
+      "grad_norm": 1.9080768652668563,
+      "learning_rate": 3.4095997875226194e-05,
+      "loss": 1.1157,
+      "step": 13840
+    },
+    {
+      "epoch": 2.21456,
+      "grad_norm": 1.8356318356330203,
+      "learning_rate": 3.408300707008903e-05,
+      "loss": 1.0839,
+      "step": 13841
+    },
+    {
+      "epoch": 2.21472,
+      "grad_norm": 1.8293842939930034,
+      "learning_rate": 3.407001823181881e-05,
+      "loss": 1.0995,
+      "step": 13842
+    },
+    {
+      "epoch": 2.21488,
+      "grad_norm": 1.7459635655024492,
+      "learning_rate": 3.405703136080321e-05,
+      "loss": 1.101,
+      "step": 13843
+    },
+    {
+      "epoch": 2.21504,
+      "grad_norm": 1.8426371216104864,
+      "learning_rate": 3.404404645742972e-05,
+      "loss": 1.1298,
+      "step": 13844
+    },
+    {
+      "epoch": 2.2152,
+      "grad_norm": 1.7321816484845265,
+      "learning_rate": 3.403106352208577e-05,
+      "loss": 1.1184,
+      "step": 13845
+    },
+    {
+      "epoch": 2.21536,
+      "grad_norm": 1.7027474264709543,
+      "learning_rate": 3.4018082555158746e-05,
+      "loss": 1.1388,
+      "step": 13846
+    },
+    {
+      "epoch": 2.21552,
+      "grad_norm": 1.9509457519270852,
+      "learning_rate": 3.400510355703601e-05,
+      "loss": 1.1438,
+      "step": 13847
+    },
+    {
+      "epoch": 2.21568,
+      "grad_norm": 1.974961187430551,
+      "learning_rate": 3.399212652810481e-05,
+      "loss": 1.1285,
+      "step": 13848
+    },
+    {
+      "epoch": 2.21584,
+      "grad_norm": 1.7023002697649976,
+      "learning_rate": 3.397915146875237e-05,
+      "loss": 1.1206,
+      "step": 13849
+    },
+    {
+      "epoch": 2.216,
+      "grad_norm": 1.847565486538425,
+      "learning_rate": 3.396617837936586e-05,
+      "loss": 1.1433,
+      "step": 13850
+    },
+    {
+      "epoch": 2.21616,
+      "grad_norm": 1.850009323833172,
+      "learning_rate": 3.395320726033234e-05,
+      "loss": 1.1087,
+      "step": 13851
+    },
+    {
+      "epoch": 2.21632,
+      "grad_norm": 1.8740445437073343,
+      "learning_rate": 3.394023811203886e-05,
+      "loss": 1.1463,
+      "step": 13852
+    },
+    {
+      "epoch": 2.21648,
+      "grad_norm": 2.1817748444080785,
+      "learning_rate": 3.392727093487248e-05,
+      "loss": 1.1081,
+      "step": 13853
+    },
+    {
+      "epoch": 2.21664,
+      "grad_norm": 1.721187715796985,
+      "learning_rate": 3.3914305729219984e-05,
+      "loss": 1.112,
+      "step": 13854
+    },
+    {
+      "epoch": 2.2168,
+      "grad_norm": 2.075111970202638,
+      "learning_rate": 3.390134249546838e-05,
+      "loss": 1.1611,
+      "step": 13855
+    },
+    {
+      "epoch": 2.21696,
+      "grad_norm": 1.7875339889139026,
+      "learning_rate": 3.388838123400434e-05,
+      "loss": 1.1249,
+      "step": 13856
+    },
+    {
+      "epoch": 2.21712,
+      "grad_norm": 1.9938594311503204,
+      "learning_rate": 3.387542194521469e-05,
+      "loss": 1.1287,
+      "step": 13857
+    },
+    {
+      "epoch": 2.21728,
+      "grad_norm": 1.7504384371412636,
+      "learning_rate": 3.38624646294861e-05,
+      "loss": 1.0884,
+      "step": 13858
+    },
+    {
+      "epoch": 2.21744,
+      "grad_norm": 1.9433230574158016,
+      "learning_rate": 3.3849509287205196e-05,
+      "loss": 1.0739,
+      "step": 13859
+    },
+    {
+      "epoch": 2.2176,
+      "grad_norm": 1.655780624730917,
+      "learning_rate": 3.383655591875855e-05,
+      "loss": 1.1282,
+      "step": 13860
+    },
+    {
+      "epoch": 2.21776,
+      "grad_norm": 1.966279081492007,
+      "learning_rate": 3.382360452453262e-05,
+      "loss": 1.169,
+      "step": 13861
+    },
+    {
+      "epoch": 2.21792,
+      "grad_norm": 1.784889478250415,
+      "learning_rate": 3.381065510491397e-05,
+      "loss": 1.1136,
+      "step": 13862
+    },
+    {
+      "epoch": 2.21808,
+      "grad_norm": 1.7506766423098403,
+      "learning_rate": 3.3797707660288856e-05,
+      "loss": 1.2098,
+      "step": 13863
+    },
+    {
+      "epoch": 2.2182399999999998,
+      "grad_norm": 2.1563803497174945,
+      "learning_rate": 3.378476219104375e-05,
+      "loss": 1.094,
+      "step": 13864
+    },
+    {
+      "epoch": 2.2184,
+      "grad_norm": 1.8647117686580457,
+      "learning_rate": 3.3771818697564785e-05,
+      "loss": 1.1652,
+      "step": 13865
+    },
+    {
+      "epoch": 2.21856,
+      "grad_norm": 2.044789538031203,
+      "learning_rate": 3.3758877180238315e-05,
+      "loss": 1.0891,
+      "step": 13866
+    },
+    {
+      "epoch": 2.21872,
+      "grad_norm": 2.176074083002377,
+      "learning_rate": 3.374593763945035e-05,
+      "loss": 1.1012,
+      "step": 13867
+    },
+    {
+      "epoch": 2.21888,
+      "grad_norm": 1.8480715891737558,
+      "learning_rate": 3.373300007558711e-05,
+      "loss": 1.1217,
+      "step": 13868
+    },
+    {
+      "epoch": 2.21904,
+      "grad_norm": 1.8050565637392275,
+      "learning_rate": 3.372006448903458e-05,
+      "loss": 1.0886,
+      "step": 13869
+    },
+    {
+      "epoch": 2.2192,
+      "grad_norm": 1.8377193799719858,
+      "learning_rate": 3.370713088017875e-05,
+      "loss": 1.1134,
+      "step": 13870
+    },
+    {
+      "epoch": 2.21936,
+      "grad_norm": 2.168129460168348,
+      "learning_rate": 3.3694199249405535e-05,
+      "loss": 1.1612,
+      "step": 13871
+    },
+    {
+      "epoch": 2.21952,
+      "grad_norm": 1.6463074227574013,
+      "learning_rate": 3.3681269597100786e-05,
+      "loss": 1.1207,
+      "step": 13872
+    },
+    {
+      "epoch": 2.21968,
+      "grad_norm": 1.7902224222379888,
+      "learning_rate": 3.3668341923650326e-05,
+      "loss": 1.1278,
+      "step": 13873
+    },
+    {
+      "epoch": 2.21984,
+      "grad_norm": 1.715175490127633,
+      "learning_rate": 3.365541622943989e-05,
+      "loss": 1.061,
+      "step": 13874
+    },
+    {
+      "epoch": 2.22,
+      "grad_norm": 1.945261597131947,
+      "learning_rate": 3.364249251485515e-05,
+      "loss": 1.0937,
+      "step": 13875
+    },
+    {
+      "epoch": 2.22016,
+      "grad_norm": 1.7030690432296687,
+      "learning_rate": 3.3629570780281735e-05,
+      "loss": 1.1296,
+      "step": 13876
+    },
+    {
+      "epoch": 2.22032,
+      "grad_norm": 1.8720432504431408,
+      "learning_rate": 3.3616651026105195e-05,
+      "loss": 1.1272,
+      "step": 13877
+    },
+    {
+      "epoch": 2.2204800000000002,
+      "grad_norm": 1.9383987250483135,
+      "learning_rate": 3.360373325271112e-05,
+      "loss": 1.0587,
+      "step": 13878
+    },
+    {
+      "epoch": 2.22064,
+      "grad_norm": 1.7557677026440002,
+      "learning_rate": 3.359081746048484e-05,
+      "loss": 1.1306,
+      "step": 13879
+    },
+    {
+      "epoch": 2.2208,
+      "grad_norm": 1.929380493166814,
+      "learning_rate": 3.3577903649811846e-05,
+      "loss": 1.1391,
+      "step": 13880
+    },
+    {
+      "epoch": 2.22096,
+      "grad_norm": 1.7944274720141058,
+      "learning_rate": 3.356499182107736e-05,
+      "loss": 1.1221,
+      "step": 13881
+    },
+    {
+      "epoch": 2.22112,
+      "grad_norm": 1.7298567102931572,
+      "learning_rate": 3.355208197466678e-05,
+      "loss": 1.1085,
+      "step": 13882
+    },
+    {
+      "epoch": 2.22128,
+      "grad_norm": 1.6853729191917404,
+      "learning_rate": 3.3539174110965187e-05,
+      "loss": 1.0792,
+      "step": 13883
+    },
+    {
+      "epoch": 2.22144,
+      "grad_norm": 1.8750386882928114,
+      "learning_rate": 3.3526268230357824e-05,
+      "loss": 1.1138,
+      "step": 13884
+    },
+    {
+      "epoch": 2.2216,
+      "grad_norm": 1.5544707884846671,
+      "learning_rate": 3.351336433322976e-05,
+      "loss": 1.0704,
+      "step": 13885
+    },
+    {
+      "epoch": 2.22176,
+      "grad_norm": 1.7435959182778074,
+      "learning_rate": 3.3500462419966015e-05,
+      "loss": 1.1393,
+      "step": 13886
+    },
+    {
+      "epoch": 2.22192,
+      "grad_norm": 1.5752555691398131,
+      "learning_rate": 3.3487562490951576e-05,
+      "loss": 1.0902,
+      "step": 13887
+    },
+    {
+      "epoch": 2.22208,
+      "grad_norm": 1.93859927819847,
+      "learning_rate": 3.3474664546571355e-05,
+      "loss": 1.1065,
+      "step": 13888
+    },
+    {
+      "epoch": 2.22224,
+      "grad_norm": 1.986144663079287,
+      "learning_rate": 3.3461768587210196e-05,
+      "loss": 1.1158,
+      "step": 13889
+    },
+    {
+      "epoch": 2.2224,
+      "grad_norm": 1.8654021488967936,
+      "learning_rate": 3.344887461325291e-05,
+      "loss": 1.0853,
+      "step": 13890
+    },
+    {
+      "epoch": 2.22256,
+      "grad_norm": 1.7642346938058158,
+      "learning_rate": 3.343598262508424e-05,
+      "loss": 1.1295,
+      "step": 13891
+    },
+    {
+      "epoch": 2.22272,
+      "grad_norm": 2.119152723932392,
+      "learning_rate": 3.342309262308885e-05,
+      "loss": 1.1118,
+      "step": 13892
+    },
+    {
+      "epoch": 2.22288,
+      "grad_norm": 1.8131284300149975,
+      "learning_rate": 3.341020460765133e-05,
+      "loss": 1.1173,
+      "step": 13893
+    },
+    {
+      "epoch": 2.22304,
+      "grad_norm": 1.7300815706160602,
+      "learning_rate": 3.3397318579156334e-05,
+      "loss": 1.0907,
+      "step": 13894
+    },
+    {
+      "epoch": 2.2232,
+      "grad_norm": 1.7928461073091224,
+      "learning_rate": 3.338443453798824e-05,
+      "loss": 1.0978,
+      "step": 13895
+    },
+    {
+      "epoch": 2.22336,
+      "grad_norm": 1.8182321079793902,
+      "learning_rate": 3.337155248453161e-05,
+      "loss": 1.1134,
+      "step": 13896
+    },
+    {
+      "epoch": 2.22352,
+      "grad_norm": 1.7712124065921713,
+      "learning_rate": 3.33586724191707e-05,
+      "loss": 1.117,
+      "step": 13897
+    },
+    {
+      "epoch": 2.22368,
+      "grad_norm": 2.0032808488710034,
+      "learning_rate": 3.334579434228997e-05,
+      "loss": 1.1148,
+      "step": 13898
+    },
+    {
+      "epoch": 2.22384,
+      "grad_norm": 2.218607495271537,
+      "learning_rate": 3.333291825427355e-05,
+      "loss": 1.1983,
+      "step": 13899
+    },
+    {
+      "epoch": 2.224,
+      "grad_norm": 1.767065369854084,
+      "learning_rate": 3.332004415550573e-05,
+      "loss": 1.1393,
+      "step": 13900
+    },
+    {
+      "epoch": 2.22416,
+      "grad_norm": 1.7684602384232997,
+      "learning_rate": 3.3307172046370624e-05,
+      "loss": 1.1434,
+      "step": 13901
+    },
+    {
+      "epoch": 2.22432,
+      "grad_norm": 1.702968040938399,
+      "learning_rate": 3.3294301927252334e-05,
+      "loss": 1.1083,
+      "step": 13902
+    },
+    {
+      "epoch": 2.22448,
+      "grad_norm": 1.674474032517394,
+      "learning_rate": 3.3281433798534876e-05,
+      "loss": 1.0892,
+      "step": 13903
+    },
+    {
+      "epoch": 2.22464,
+      "grad_norm": 1.8663197016567585,
+      "learning_rate": 3.32685676606022e-05,
+      "loss": 1.1116,
+      "step": 13904
+    },
+    {
+      "epoch": 2.2248,
+      "grad_norm": 1.8251668223576871,
+      "learning_rate": 3.325570351383824e-05,
+      "loss": 1.1026,
+      "step": 13905
+    },
+    {
+      "epoch": 2.22496,
+      "grad_norm": 1.7789336189583678,
+      "learning_rate": 3.324284135862683e-05,
+      "loss": 1.1718,
+      "step": 13906
+    },
+    {
+      "epoch": 2.22512,
+      "grad_norm": 1.691521999945804,
+      "learning_rate": 3.322998119535176e-05,
+      "loss": 1.113,
+      "step": 13907
+    },
+    {
+      "epoch": 2.22528,
+      "grad_norm": 1.774746017166777,
+      "learning_rate": 3.3217123024396745e-05,
+      "loss": 1.1712,
+      "step": 13908
+    },
+    {
+      "epoch": 2.22544,
+      "grad_norm": 1.9913355076545056,
+      "learning_rate": 3.3204266846145437e-05,
+      "loss": 1.1352,
+      "step": 13909
+    },
+    {
+      "epoch": 2.2256,
+      "grad_norm": 1.7117395526221642,
+      "learning_rate": 3.3191412660981545e-05,
+      "loss": 1.1341,
+      "step": 13910
+    },
+    {
+      "epoch": 2.22576,
+      "grad_norm": 1.6086773753229129,
+      "learning_rate": 3.317856046928848e-05,
+      "loss": 1.0626,
+      "step": 13911
+    },
+    {
+      "epoch": 2.22592,
+      "grad_norm": 1.797091498804963,
+      "learning_rate": 3.316571027144987e-05,
+      "loss": 1.1296,
+      "step": 13912
+    },
+    {
+      "epoch": 2.22608,
+      "grad_norm": 1.6007978631765913,
+      "learning_rate": 3.315286206784901e-05,
+      "loss": 1.1275,
+      "step": 13913
+    },
+    {
+      "epoch": 2.2262399999999998,
+      "grad_norm": 1.7388177668574294,
+      "learning_rate": 3.31400158588694e-05,
+      "loss": 1.0837,
+      "step": 13914
+    },
+    {
+      "epoch": 2.2264,
+      "grad_norm": 1.589304671385862,
+      "learning_rate": 3.312717164489423e-05,
+      "loss": 1.0625,
+      "step": 13915
+    },
+    {
+      "epoch": 2.22656,
+      "grad_norm": 1.589304671385862,
+      "learning_rate": 3.312717164489423e-05,
+      "loss": 1.1052,
+      "step": 13916
+    },
+    {
+      "epoch": 2.22672,
+      "grad_norm": 2.2805497450847407,
+      "learning_rate": 3.3114329426306846e-05,
+      "loss": 1.1271,
+      "step": 13917
+    },
+    {
+      "epoch": 2.22688,
+      "grad_norm": 1.8048813192201725,
+      "learning_rate": 3.310148920349041e-05,
+      "loss": 1.0838,
+      "step": 13918
+    },
+    {
+      "epoch": 2.22704,
+      "grad_norm": 1.6710194139238748,
+      "learning_rate": 3.3088650976828065e-05,
+      "loss": 1.1463,
+      "step": 13919
+    },
+    {
+      "epoch": 2.2272,
+      "grad_norm": 1.663153158648352,
+      "learning_rate": 3.307581474670286e-05,
+      "loss": 1.1206,
+      "step": 13920
+    },
+    {
+      "epoch": 2.22736,
+      "grad_norm": 2.0986736537515958,
+      "learning_rate": 3.30629805134978e-05,
+      "loss": 1.1218,
+      "step": 13921
+    },
+    {
+      "epoch": 2.22752,
+      "grad_norm": 2.072225144385388,
+      "learning_rate": 3.3050148277595914e-05,
+      "loss": 1.0967,
+      "step": 13922
+    },
+    {
+      "epoch": 2.22768,
+      "grad_norm": 1.8295417462337706,
+      "learning_rate": 3.3037318039380004e-05,
+      "loss": 1.1414,
+      "step": 13923
+    },
+    {
+      "epoch": 2.22784,
+      "grad_norm": 2.1741498546085793,
+      "learning_rate": 3.3024489799233006e-05,
+      "loss": 1.1289,
+      "step": 13924
+    },
+    {
+      "epoch": 2.228,
+      "grad_norm": 2.2284614111213528,
+      "learning_rate": 3.3011663557537564e-05,
+      "loss": 1.1289,
+      "step": 13925
+    },
+    {
+      "epoch": 2.22816,
+      "grad_norm": 2.071573716754016,
+      "learning_rate": 3.299883931467651e-05,
+      "loss": 1.1365,
+      "step": 13926
+    },
+    {
+      "epoch": 2.22832,
+      "grad_norm": 1.8047057482825843,
+      "learning_rate": 3.298601707103246e-05,
+      "loss": 1.106,
+      "step": 13927
+    },
+    {
+      "epoch": 2.22848,
+      "grad_norm": 1.7606635093135596,
+      "learning_rate": 3.297319682698801e-05,
+      "loss": 1.0805,
+      "step": 13928
+    },
+    {
+      "epoch": 2.22864,
+      "grad_norm": 1.9074604079873665,
+      "learning_rate": 3.29603785829257e-05,
+      "loss": 1.0798,
+      "step": 13929
+    },
+    {
+      "epoch": 2.2288,
+      "grad_norm": 1.897146068214173,
+      "learning_rate": 3.2947562339228e-05,
+      "loss": 1.1664,
+      "step": 13930
+    },
+    {
+      "epoch": 2.22896,
+      "grad_norm": 2.162807680095441,
+      "learning_rate": 3.2934748096277335e-05,
+      "loss": 1.128,
+      "step": 13931
+    },
+    {
+      "epoch": 2.22912,
+      "grad_norm": 2.2286750995854123,
+      "learning_rate": 3.292193585445606e-05,
+      "loss": 1.1386,
+      "step": 13932
+    },
+    {
+      "epoch": 2.22928,
+      "grad_norm": 1.8112748654622794,
+      "learning_rate": 3.29091256141465e-05,
+      "loss": 1.1344,
+      "step": 13933
+    },
+    {
+      "epoch": 2.22944,
+      "grad_norm": 2.06224520766159,
+      "learning_rate": 3.2896317375730844e-05,
+      "loss": 1.1214,
+      "step": 13934
+    },
+    {
+      "epoch": 2.2296,
+      "grad_norm": 1.7270542356702174,
+      "learning_rate": 3.28835111395913e-05,
+      "loss": 1.0954,
+      "step": 13935
+    },
+    {
+      "epoch": 2.22976,
+      "grad_norm": 1.8559833842584772,
+      "learning_rate": 3.287070690610999e-05,
+      "loss": 1.1312,
+      "step": 13936
+    },
+    {
+      "epoch": 2.22992,
+      "grad_norm": 1.8283404242203338,
+      "learning_rate": 3.285790467566894e-05,
+      "loss": 1.1479,
+      "step": 13937
+    },
+    {
+      "epoch": 2.23008,
+      "grad_norm": 1.5388794306349318,
+      "learning_rate": 3.284510444865023e-05,
+      "loss": 1.1113,
+      "step": 13938
+    },
+    {
+      "epoch": 2.2302400000000002,
+      "grad_norm": 1.7736851005830576,
+      "learning_rate": 3.2832306225435695e-05,
+      "loss": 1.1064,
+      "step": 13939
+    },
+    {
+      "epoch": 2.2304,
+      "grad_norm": 1.808860061341401,
+      "learning_rate": 3.281951000640733e-05,
+      "loss": 1.1927,
+      "step": 13940
+    },
+    {
+      "epoch": 2.23056,
+      "grad_norm": 1.7364448922885845,
+      "learning_rate": 3.280671579194683e-05,
+      "loss": 1.1234,
+      "step": 13941
+    },
+    {
+      "epoch": 2.23072,
+      "grad_norm": 1.9312582855004439,
+      "learning_rate": 3.279392358243605e-05,
+      "loss": 1.145,
+      "step": 13942
+    },
+    {
+      "epoch": 2.23088,
+      "grad_norm": 1.8079321077903963,
+      "learning_rate": 3.2781133378256665e-05,
+      "loss": 1.0982,
+      "step": 13943
+    },
+    {
+      "epoch": 2.23104,
+      "grad_norm": 1.9911857798217136,
+      "learning_rate": 3.2768345179790304e-05,
+      "loss": 1.181,
+      "step": 13944
+    },
+    {
+      "epoch": 2.2312,
+      "grad_norm": 1.9851240927425264,
+      "learning_rate": 3.275555898741857e-05,
+      "loss": 1.1082,
+      "step": 13945
+    },
+    {
+      "epoch": 2.23136,
+      "grad_norm": 1.9442384030319046,
+      "learning_rate": 3.2742774801522954e-05,
+      "loss": 1.1486,
+      "step": 13946
+    },
+    {
+      "epoch": 2.23152,
+      "grad_norm": 1.844053567581617,
+      "learning_rate": 3.272999262248494e-05,
+      "loss": 1.1336,
+      "step": 13947
+    },
+    {
+      "epoch": 2.23168,
+      "grad_norm": 1.9069495068621927,
+      "learning_rate": 3.271721245068593e-05,
+      "loss": 1.1155,
+      "step": 13948
+    },
+    {
+      "epoch": 2.23184,
+      "grad_norm": 1.9730727425730425,
+      "learning_rate": 3.2704434286507246e-05,
+      "loss": 1.0765,
+      "step": 13949
+    },
+    {
+      "epoch": 2.232,
+      "grad_norm": 1.7951712958891994,
+      "learning_rate": 3.269165813033019e-05,
+      "loss": 1.0885,
+      "step": 13950
+    },
+    {
+      "epoch": 2.23216,
+      "grad_norm": 1.6927560140029787,
+      "learning_rate": 3.267888398253595e-05,
+      "loss": 1.1214,
+      "step": 13951
+    },
+    {
+      "epoch": 2.23232,
+      "grad_norm": 1.9492081964031867,
+      "learning_rate": 3.266611184350577e-05,
+      "loss": 1.0964,
+      "step": 13952
+    },
+    {
+      "epoch": 2.23248,
+      "grad_norm": 1.9630839162410232,
+      "learning_rate": 3.265334171362063e-05,
+      "loss": 1.1224,
+      "step": 13953
+    },
+    {
+      "epoch": 2.23264,
+      "grad_norm": 1.7787591291750988,
+      "learning_rate": 3.2640573593261714e-05,
+      "loss": 1.0582,
+      "step": 13954
+    },
+    {
+      "epoch": 2.2328,
+      "grad_norm": 1.8006148128957384,
+      "learning_rate": 3.2627807482809856e-05,
+      "loss": 1.1126,
+      "step": 13955
+    },
+    {
+      "epoch": 2.23296,
+      "grad_norm": 1.8190738379852756,
+      "learning_rate": 3.261504338264613e-05,
+      "loss": 1.0847,
+      "step": 13956
+    },
+    {
+      "epoch": 2.23312,
+      "grad_norm": 2.1784143681976214,
+      "learning_rate": 3.260228129315124e-05,
+      "loss": 1.1706,
+      "step": 13957
+    },
+    {
+      "epoch": 2.23328,
+      "grad_norm": 1.73371746430037,
+      "learning_rate": 3.2589521214706095e-05,
+      "loss": 1.1131,
+      "step": 13958
+    },
+    {
+      "epoch": 2.23344,
+      "grad_norm": 1.7550330645760208,
+      "learning_rate": 3.2576763147691424e-05,
+      "loss": 1.0887,
+      "step": 13959
+    },
+    {
+      "epoch": 2.2336,
+      "grad_norm": 2.337206334622198,
+      "learning_rate": 3.2564007092487895e-05,
+      "loss": 1.1103,
+      "step": 13960
+    },
+    {
+      "epoch": 2.23376,
+      "grad_norm": 2.1175486966027735,
+      "learning_rate": 3.2551253049476115e-05,
+      "loss": 1.1072,
+      "step": 13961
+    },
+    {
+      "epoch": 2.23392,
+      "grad_norm": 1.8319114524274402,
+      "learning_rate": 3.253850101903667e-05,
+      "loss": 1.1191,
+      "step": 13962
+    },
+    {
+      "epoch": 2.23408,
+      "grad_norm": 1.760929835923097,
+      "learning_rate": 3.2525751001550056e-05,
+      "loss": 1.099,
+      "step": 13963
+    },
+    {
+      "epoch": 2.23424,
+      "grad_norm": 2.1038207484064926,
+      "learning_rate": 3.25130029973967e-05,
+      "loss": 1.1328,
+      "step": 13964
+    },
+    {
+      "epoch": 2.2344,
+      "grad_norm": 1.9979475974945555,
+      "learning_rate": 3.2500257006957014e-05,
+      "loss": 1.067,
+      "step": 13965
+    },
+    {
+      "epoch": 2.23456,
+      "grad_norm": 1.722567294976537,
+      "learning_rate": 3.248751303061128e-05,
+      "loss": 1.118,
+      "step": 13966
+    },
+    {
+      "epoch": 2.23472,
+      "grad_norm": 2.040549385596892,
+      "learning_rate": 3.247477106873976e-05,
+      "loss": 1.1052,
+      "step": 13967
+    },
+    {
+      "epoch": 2.23488,
+      "grad_norm": 1.6530744908312587,
+      "learning_rate": 3.246203112172275e-05,
+      "loss": 1.0647,
+      "step": 13968
+    },
+    {
+      "epoch": 2.23504,
+      "grad_norm": 1.5936659614713973,
+      "learning_rate": 3.244929318994024e-05,
+      "loss": 1.076,
+      "step": 13969
+    },
+    {
+      "epoch": 2.2352,
+      "grad_norm": 2.0614693302571787,
+      "learning_rate": 3.243655727377246e-05,
+      "loss": 1.1391,
+      "step": 13970
+    },
+    {
+      "epoch": 2.23536,
+      "grad_norm": 1.5322999497116918,
+      "learning_rate": 3.242382337359929e-05,
+      "loss": 1.101,
+      "step": 13971
+    },
+    {
+      "epoch": 2.23552,
+      "grad_norm": 1.7813985620662647,
+      "learning_rate": 3.2411091489800834e-05,
+      "loss": 1.1135,
+      "step": 13972
+    },
+    {
+      "epoch": 2.23568,
+      "grad_norm": 1.6974130896889474,
+      "learning_rate": 3.2398361622756856e-05,
+      "loss": 1.1002,
+      "step": 13973
+    },
+    {
+      "epoch": 2.23584,
+      "grad_norm": 1.8293854421751798,
+      "learning_rate": 3.238563377284729e-05,
+      "loss": 1.1056,
+      "step": 13974
+    },
+    {
+      "epoch": 2.2359999999999998,
+      "grad_norm": 1.7689866733401975,
+      "learning_rate": 3.237290794045189e-05,
+      "loss": 1.1762,
+      "step": 13975
+    },
+    {
+      "epoch": 2.23616,
+      "grad_norm": 1.8143836332589205,
+      "learning_rate": 3.236018412595037e-05,
+      "loss": 1.0754,
+      "step": 13976
+    },
+    {
+      "epoch": 2.23632,
+      "grad_norm": 1.8472443887847267,
+      "learning_rate": 3.234746232972239e-05,
+      "loss": 1.1421,
+      "step": 13977
+    },
+    {
+      "epoch": 2.23648,
+      "grad_norm": 1.8209544280769814,
+      "learning_rate": 3.2334742552147536e-05,
+      "loss": 1.1454,
+      "step": 13978
+    },
+    {
+      "epoch": 2.23664,
+      "grad_norm": 1.8183734172214883,
+      "learning_rate": 3.232202479360542e-05,
+      "loss": 1.1819,
+      "step": 13979
+    },
+    {
+      "epoch": 2.2368,
+      "grad_norm": 1.7202942737804157,
+      "learning_rate": 3.23093090544754e-05,
+      "loss": 1.0427,
+      "step": 13980
+    },
+    {
+      "epoch": 2.23696,
+      "grad_norm": 1.7296814877519513,
+      "learning_rate": 3.229659533513703e-05,
+      "loss": 1.1179,
+      "step": 13981
+    },
+    {
+      "epoch": 2.23712,
+      "grad_norm": 1.5065406145470566,
+      "learning_rate": 3.228388363596954e-05,
+      "loss": 1.0957,
+      "step": 13982
+    },
+    {
+      "epoch": 2.23728,
+      "grad_norm": 1.788106249148507,
+      "learning_rate": 3.227117395735232e-05,
+      "loss": 1.1187,
+      "step": 13983
+    },
+    {
+      "epoch": 2.23744,
+      "grad_norm": 1.6869766362212897,
+      "learning_rate": 3.2258466299664566e-05,
+      "loss": 1.1009,
+      "step": 13984
+    },
+    {
+      "epoch": 2.2376,
+      "grad_norm": 1.7487185931006552,
+      "learning_rate": 3.2245760663285486e-05,
+      "loss": 1.1396,
+      "step": 13985
+    },
+    {
+      "epoch": 2.23776,
+      "grad_norm": 1.8190857043037558,
+      "learning_rate": 3.2233057048594176e-05,
+      "loss": 1.1158,
+      "step": 13986
+    },
+    {
+      "epoch": 2.23792,
+      "grad_norm": 1.8870262161081448,
+      "learning_rate": 3.2220355455969685e-05,
+      "loss": 1.1849,
+      "step": 13987
+    },
+    {
+      "epoch": 2.23808,
+      "grad_norm": 1.9857430067992952,
+      "learning_rate": 3.220765588579103e-05,
+      "loss": 1.1282,
+      "step": 13988
+    },
+    {
+      "epoch": 2.2382400000000002,
+      "grad_norm": 1.8334017730689582,
+      "learning_rate": 3.219495833843714e-05,
+      "loss": 1.1181,
+      "step": 13989
+    },
+    {
+      "epoch": 2.2384,
+      "grad_norm": 1.6690548454495375,
+      "learning_rate": 3.218226281428689e-05,
+      "loss": 1.1159,
+      "step": 13990
+    },
+    {
+      "epoch": 2.23856,
+      "grad_norm": 1.8326841407763232,
+      "learning_rate": 3.2169569313719096e-05,
+      "loss": 1.0931,
+      "step": 13991
+    },
+    {
+      "epoch": 2.23872,
+      "grad_norm": 1.6887705204555366,
+      "learning_rate": 3.215687783711251e-05,
+      "loss": 1.1072,
+      "step": 13992
+    },
+    {
+      "epoch": 2.23888,
+      "grad_norm": 1.6775259807759657,
+      "learning_rate": 3.214418838484584e-05,
+      "loss": 1.0592,
+      "step": 13993
+    },
+    {
+      "epoch": 2.23904,
+      "grad_norm": 1.9533730481335392,
+      "learning_rate": 3.213150095729768e-05,
+      "loss": 1.0833,
+      "step": 13994
+    },
+    {
+      "epoch": 2.2392,
+      "grad_norm": 1.8051327332126892,
+      "learning_rate": 3.211881555484669e-05,
+      "loss": 1.109,
+      "step": 13995
+    },
+    {
+      "epoch": 2.23936,
+      "grad_norm": 1.8034376467365454,
+      "learning_rate": 3.2106132177871276e-05,
+      "loss": 1.081,
+      "step": 13996
+    },
+    {
+      "epoch": 2.23952,
+      "grad_norm": 1.9121169812133483,
+      "learning_rate": 3.209345082675e-05,
+      "loss": 1.0989,
+      "step": 13997
+    },
+    {
+      "epoch": 2.23968,
+      "grad_norm": 1.8439953611743098,
+      "learning_rate": 3.208077150186115e-05,
+      "loss": 1.1239,
+      "step": 13998
+    },
+    {
+      "epoch": 2.23984,
+      "grad_norm": 1.7018684628411271,
+      "learning_rate": 3.206809420358313e-05,
+      "loss": 1.1369,
+      "step": 13999
+    },
+    {
+      "epoch": 2.24,
+      "grad_norm": 2.0054362625041504,
+      "learning_rate": 3.20554189322942e-05,
+      "loss": 1.0968,
+      "step": 14000
+    },
+    {
+      "epoch": 2.24016,
+      "grad_norm": 1.7615762799778039,
+      "learning_rate": 3.2042745688372564e-05,
+      "loss": 1.076,
+      "step": 14001
+    },
+    {
+      "epoch": 2.24032,
+      "grad_norm": 1.6604660377962464,
+      "learning_rate": 3.203007447219637e-05,
+      "loss": 1.1215,
+      "step": 14002
+    },
+    {
+      "epoch": 2.24048,
+      "grad_norm": 1.8941764613783936,
+      "learning_rate": 3.201740528414371e-05,
+      "loss": 1.1069,
+      "step": 14003
+    },
+    {
+      "epoch": 2.24064,
+      "grad_norm": 2.073489194662187,
+      "learning_rate": 3.200473812459262e-05,
+      "loss": 1.0451,
+      "step": 14004
+    },
+    {
+      "epoch": 2.2408,
+      "grad_norm": 1.9190320928834057,
+      "learning_rate": 3.199207299392105e-05,
+      "loss": 1.1057,
+      "step": 14005
+    },
+    {
+      "epoch": 2.24096,
+      "grad_norm": 2.129407410659505,
+      "learning_rate": 3.197940989250694e-05,
+      "loss": 1.1091,
+      "step": 14006
+    },
+    {
+      "epoch": 2.24112,
+      "grad_norm": 1.9034190652775973,
+      "learning_rate": 3.196674882072812e-05,
+      "loss": 1.1192,
+      "step": 14007
+    },
+    {
+      "epoch": 2.24128,
+      "grad_norm": 1.7783857000480787,
+      "learning_rate": 3.1954089778962374e-05,
+      "loss": 1.1238,
+      "step": 14008
+    },
+    {
+      "epoch": 2.24144,
+      "grad_norm": 2.058743596709524,
+      "learning_rate": 3.194143276758744e-05,
+      "loss": 1.0766,
+      "step": 14009
+    },
+    {
+      "epoch": 2.2416,
+      "grad_norm": 2.058743596709524,
+      "learning_rate": 3.194143276758744e-05,
+      "loss": 1.1396,
+      "step": 14010
+    },
+    {
+      "epoch": 2.24176,
+      "grad_norm": 1.8653784168570313,
+      "learning_rate": 3.192877778698094e-05,
+      "loss": 1.0996,
+      "step": 14011
+    },
+    {
+      "epoch": 2.24192,
+      "grad_norm": 1.888555451070093,
+      "learning_rate": 3.19161248375206e-05,
+      "loss": 1.1549,
+      "step": 14012
+    },
+    {
+      "epoch": 2.24208,
+      "grad_norm": 1.8361969905360533,
+      "learning_rate": 3.1903473919583816e-05,
+      "loss": 1.1601,
+      "step": 14013
+    },
+    {
+      "epoch": 2.24224,
+      "grad_norm": 1.8440981440960507,
+      "learning_rate": 3.18908250335482e-05,
+      "loss": 1.133,
+      "step": 14014
+    },
+    {
+      "epoch": 2.2424,
+      "grad_norm": 1.736107411071958,
+      "learning_rate": 3.187817817979106e-05,
+      "loss": 1.1396,
+      "step": 14015
+    },
+    {
+      "epoch": 2.24256,
+      "grad_norm": 1.9213785392258127,
+      "learning_rate": 3.186553335868985e-05,
+      "loss": 1.0736,
+      "step": 14016
+    },
+    {
+      "epoch": 2.24272,
+      "grad_norm": 1.6743491741173293,
+      "learning_rate": 3.185289057062184e-05,
+      "loss": 1.1023,
+      "step": 14017
+    },
+    {
+      "epoch": 2.24288,
+      "grad_norm": 1.8862284928542192,
+      "learning_rate": 3.184024981596428e-05,
+      "loss": 1.1529,
+      "step": 14018
+    },
+    {
+      "epoch": 2.24304,
+      "grad_norm": 1.9385336182816097,
+      "learning_rate": 3.1827611095094346e-05,
+      "loss": 1.1565,
+      "step": 14019
+    },
+    {
+      "epoch": 2.2432,
+      "grad_norm": 1.9712681639256597,
+      "learning_rate": 3.181497440838915e-05,
+      "loss": 1.1276,
+      "step": 14020
+    },
+    {
+      "epoch": 2.24336,
+      "grad_norm": 2.0164623973593088,
+      "learning_rate": 3.1802339756225776e-05,
+      "loss": 1.1113,
+      "step": 14021
+    },
+    {
+      "epoch": 2.24352,
+      "grad_norm": 1.8790408806415742,
+      "learning_rate": 3.1789707138981204e-05,
+      "loss": 1.0851,
+      "step": 14022
+    },
+    {
+      "epoch": 2.24368,
+      "grad_norm": 1.7986126214846103,
+      "learning_rate": 3.177707655703238e-05,
+      "loss": 1.0672,
+      "step": 14023
+    },
+    {
+      "epoch": 2.24384,
+      "grad_norm": 1.8786783310556696,
+      "learning_rate": 3.1764448010756175e-05,
+      "loss": 1.111,
+      "step": 14024
+    },
+    {
+      "epoch": 2.2439999999999998,
+      "grad_norm": 1.910812416769067,
+      "learning_rate": 3.175182150052941e-05,
+      "loss": 1.1668,
+      "step": 14025
+    },
+    {
+      "epoch": 2.24416,
+      "grad_norm": 1.8017906906793595,
+      "learning_rate": 3.173919702672885e-05,
+      "loss": 1.0881,
+      "step": 14026
+    },
+    {
+      "epoch": 2.24432,
+      "grad_norm": 1.6663716570084524,
+      "learning_rate": 3.172657458973115e-05,
+      "loss": 1.0758,
+      "step": 14027
+    },
+    {
+      "epoch": 2.24448,
+      "grad_norm": 1.7608424440797965,
+      "learning_rate": 3.171395418991304e-05,
+      "loss": 1.1159,
+      "step": 14028
+    },
+    {
+      "epoch": 2.24464,
+      "grad_norm": 1.837530639892322,
+      "learning_rate": 3.1701335827650967e-05,
+      "loss": 1.1271,
+      "step": 14029
+    },
+    {
+      "epoch": 2.2448,
+      "grad_norm": 1.7599169884715329,
+      "learning_rate": 3.1688719503321586e-05,
+      "loss": 1.077,
+      "step": 14030
+    },
+    {
+      "epoch": 2.24496,
+      "grad_norm": 1.8527080656073207,
+      "learning_rate": 3.16761052173012e-05,
+      "loss": 1.1515,
+      "step": 14031
+    },
+    {
+      "epoch": 2.24512,
+      "grad_norm": 1.98433394395756,
+      "learning_rate": 3.166349296996631e-05,
+      "loss": 1.1106,
+      "step": 14032
+    },
+    {
+      "epoch": 2.24528,
+      "grad_norm": 1.6511026590566171,
+      "learning_rate": 3.165088276169322e-05,
+      "loss": 1.1112,
+      "step": 14033
+    },
+    {
+      "epoch": 2.24544,
+      "grad_norm": 2.110513565292021,
+      "learning_rate": 3.163827459285819e-05,
+      "loss": 1.1154,
+      "step": 14034
+    },
+    {
+      "epoch": 2.2456,
+      "grad_norm": 1.7660241888202792,
+      "learning_rate": 3.1625668463837434e-05,
+      "loss": 1.1012,
+      "step": 14035
+    },
+    {
+      "epoch": 2.24576,
+      "grad_norm": 1.8181564429882342,
+      "learning_rate": 3.1613064375007065e-05,
+      "loss": 1.0786,
+      "step": 14036
+    },
+    {
+      "epoch": 2.24592,
+      "grad_norm": 1.7416672460120797,
+      "learning_rate": 3.1600462326743264e-05,
+      "loss": 1.1168,
+      "step": 14037
+    },
+    {
+      "epoch": 2.24608,
+      "grad_norm": 1.8626088064679611,
+      "learning_rate": 3.1587862319421944e-05,
+      "loss": 1.1016,
+      "step": 14038
+    },
+    {
+      "epoch": 2.24624,
+      "grad_norm": 1.8050413289702834,
+      "learning_rate": 3.157526435341919e-05,
+      "loss": 1.0839,
+      "step": 14039
+    },
+    {
+      "epoch": 2.2464,
+      "grad_norm": 1.8511643564741462,
+      "learning_rate": 3.156266842911076e-05,
+      "loss": 1.1529,
+      "step": 14040
+    },
+    {
+      "epoch": 2.24656,
+      "grad_norm": 1.8175992297445038,
+      "learning_rate": 3.155007454687263e-05,
+      "loss": 1.0706,
+      "step": 14041
+    },
+    {
+      "epoch": 2.24672,
+      "grad_norm": 2.0069131186069487,
+      "learning_rate": 3.1537482707080524e-05,
+      "loss": 1.1463,
+      "step": 14042
+    },
+    {
+      "epoch": 2.24688,
+      "grad_norm": 2.0380455555885675,
+      "learning_rate": 3.152489291011018e-05,
+      "loss": 1.0872,
+      "step": 14043
+    },
+    {
+      "epoch": 2.24704,
+      "grad_norm": 2.018139128274044,
+      "learning_rate": 3.151230515633725e-05,
+      "loss": 1.0999,
+      "step": 14044
+    },
+    {
+      "epoch": 2.2472,
+      "grad_norm": 1.9270194621080874,
+      "learning_rate": 3.149971944613734e-05,
+      "loss": 1.1398,
+      "step": 14045
+    },
+    {
+      "epoch": 2.24736,
+      "grad_norm": 1.7772408683369993,
+      "learning_rate": 3.148713577988599e-05,
+      "loss": 1.0886,
+      "step": 14046
+    },
+    {
+      "epoch": 2.24752,
+      "grad_norm": 1.7270109327344327,
+      "learning_rate": 3.147455415795867e-05,
+      "loss": 1.1222,
+      "step": 14047
+    },
+    {
+      "epoch": 2.24768,
+      "grad_norm": 1.752202651577799,
+      "learning_rate": 3.146197458073079e-05,
+      "loss": 1.1029,
+      "step": 14048
+    },
+    {
+      "epoch": 2.24784,
+      "grad_norm": 1.9167064070741426,
+      "learning_rate": 3.1449397048577735e-05,
+      "loss": 1.0727,
+      "step": 14049
+    },
+    {
+      "epoch": 2.248,
+      "grad_norm": 1.7954179477451262,
+      "learning_rate": 3.143682156187477e-05,
+      "loss": 1.1132,
+      "step": 14050
+    },
+    {
+      "epoch": 2.24816,
+      "grad_norm": 1.658327325151597,
+      "learning_rate": 3.142424812099716e-05,
+      "loss": 1.0813,
+      "step": 14051
+    },
+    {
+      "epoch": 2.24832,
+      "grad_norm": 1.7457331004449956,
+      "learning_rate": 3.1411676726320016e-05,
+      "loss": 1.1209,
+      "step": 14052
+    },
+    {
+      "epoch": 2.24848,
+      "grad_norm": 2.0056953495582284,
+      "learning_rate": 3.139910737821856e-05,
+      "loss": 1.0948,
+      "step": 14053
+    },
+    {
+      "epoch": 2.24864,
+      "grad_norm": 1.754459242469359,
+      "learning_rate": 3.138654007706772e-05,
+      "loss": 1.1245,
+      "step": 14054
+    },
+    {
+      "epoch": 2.2488,
+      "grad_norm": 1.9076606559432197,
+      "learning_rate": 3.1373974823242594e-05,
+      "loss": 1.1151,
+      "step": 14055
+    },
+    {
+      "epoch": 2.24896,
+      "grad_norm": 1.9248130402064427,
+      "learning_rate": 3.136141161711801e-05,
+      "loss": 1.1006,
+      "step": 14056
+    },
+    {
+      "epoch": 2.24912,
+      "grad_norm": 1.8473997622555987,
+      "learning_rate": 3.134885045906891e-05,
+      "loss": 1.1386,
+      "step": 14057
+    },
+    {
+      "epoch": 2.24928,
+      "grad_norm": 1.7850961161133585,
+      "learning_rate": 3.133629134947008e-05,
+      "loss": 1.1129,
+      "step": 14058
+    },
+    {
+      "epoch": 2.24944,
+      "grad_norm": 1.889830776715013,
+      "learning_rate": 3.132373428869626e-05,
+      "loss": 1.1126,
+      "step": 14059
+    },
+    {
+      "epoch": 2.2496,
+      "grad_norm": 2.0023043822052977,
+      "learning_rate": 3.131117927712214e-05,
+      "loss": 1.1017,
+      "step": 14060
+    },
+    {
+      "epoch": 2.24976,
+      "grad_norm": 1.8225800804229493,
+      "learning_rate": 3.1298626315122335e-05,
+      "loss": 1.107,
+      "step": 14061
+    },
+    {
+      "epoch": 2.24992,
+      "grad_norm": 2.1006830616398644,
+      "learning_rate": 3.1286075403071426e-05,
+      "loss": 1.1128,
+      "step": 14062
+    },
+    {
+      "epoch": 2.25008,
+      "grad_norm": 1.8089868991243068,
+      "learning_rate": 3.127352654134389e-05,
+      "loss": 1.0909,
+      "step": 14063
+    },
+    {
+      "epoch": 2.25024,
+      "grad_norm": 1.6805717344082596,
+      "learning_rate": 3.126097973031418e-05,
+      "loss": 1.0543,
+      "step": 14064
+    },
+    {
+      "epoch": 2.2504,
+      "grad_norm": 2.1721872988369726,
+      "learning_rate": 3.1248434970356665e-05,
+      "loss": 1.1253,
+      "step": 14065
+    },
+    {
+      "epoch": 2.25056,
+      "grad_norm": 2.7623531561318355,
+      "learning_rate": 3.1235892261845675e-05,
+      "loss": 1.1105,
+      "step": 14066
+    },
+    {
+      "epoch": 2.25072,
+      "grad_norm": 1.8152915959205627,
+      "learning_rate": 3.122335160515546e-05,
+      "loss": 1.1199,
+      "step": 14067
+    },
+    {
+      "epoch": 2.25088,
+      "grad_norm": 1.7877341044107617,
+      "learning_rate": 3.1210813000660176e-05,
+      "loss": 1.0675,
+      "step": 14068
+    },
+    {
+      "epoch": 2.25104,
+      "grad_norm": 1.8048938977921025,
+      "learning_rate": 3.1198276448734064e-05,
+      "loss": 1.1164,
+      "step": 14069
+    },
+    {
+      "epoch": 2.2512,
+      "grad_norm": 2.3879356371327742,
+      "learning_rate": 3.118574194975107e-05,
+      "loss": 1.1498,
+      "step": 14070
+    },
+    {
+      "epoch": 2.25136,
+      "grad_norm": 1.733005041719143,
+      "learning_rate": 3.117320950408532e-05,
+      "loss": 1.1174,
+      "step": 14071
+    },
+    {
+      "epoch": 2.25152,
+      "grad_norm": 1.6002813250153936,
+      "learning_rate": 3.116067911211065e-05,
+      "loss": 1.0669,
+      "step": 14072
+    },
+    {
+      "epoch": 2.25168,
+      "grad_norm": 1.9033062045191391,
+      "learning_rate": 3.1148150774201034e-05,
+      "loss": 1.1317,
+      "step": 14073
+    },
+    {
+      "epoch": 2.25184,
+      "grad_norm": 1.6626811348049158,
+      "learning_rate": 3.1135624490730265e-05,
+      "loss": 1.1098,
+      "step": 14074
+    },
+    {
+      "epoch": 2.252,
+      "grad_norm": 1.9026783150629138,
+      "learning_rate": 3.112310026207212e-05,
+      "loss": 1.0899,
+      "step": 14075
+    },
+    {
+      "epoch": 2.25216,
+      "grad_norm": 1.9108185143971461,
+      "learning_rate": 3.11105780886003e-05,
+      "loss": 1.1289,
+      "step": 14076
+    },
+    {
+      "epoch": 2.25232,
+      "grad_norm": 1.977758638442251,
+      "learning_rate": 3.109805797068846e-05,
+      "loss": 1.117,
+      "step": 14077
+    },
+    {
+      "epoch": 2.25248,
+      "grad_norm": 1.7322526031678132,
+      "learning_rate": 3.1085539908710145e-05,
+      "loss": 1.1072,
+      "step": 14078
+    },
+    {
+      "epoch": 2.25264,
+      "grad_norm": 1.8407426489423602,
+      "learning_rate": 3.107302390303891e-05,
+      "loss": 1.1271,
+      "step": 14079
+    },
+    {
+      "epoch": 2.2528,
+      "grad_norm": 2.0126983754852166,
+      "learning_rate": 3.106050995404821e-05,
+      "loss": 1.1357,
+      "step": 14080
+    },
+    {
+      "epoch": 2.25296,
+      "grad_norm": 1.7183754249061485,
+      "learning_rate": 3.104799806211143e-05,
+      "loss": 1.1478,
+      "step": 14081
+    },
+    {
+      "epoch": 2.25312,
+      "grad_norm": 1.6706471458353367,
+      "learning_rate": 3.103548822760192e-05,
+      "loss": 1.1005,
+      "step": 14082
+    },
+    {
+      "epoch": 2.25328,
+      "grad_norm": 1.9050286453181302,
+      "learning_rate": 3.102298045089295e-05,
+      "loss": 1.1347,
+      "step": 14083
+    },
+    {
+      "epoch": 2.25344,
+      "grad_norm": 1.9714003034665624,
+      "learning_rate": 3.1010474732357696e-05,
+      "loss": 1.1159,
+      "step": 14084
+    },
+    {
+      "epoch": 2.2536,
+      "grad_norm": 1.973488586507523,
+      "learning_rate": 3.0997971072369415e-05,
+      "loss": 1.1017,
+      "step": 14085
+    },
+    {
+      "epoch": 2.2537599999999998,
+      "grad_norm": 1.729011189258202,
+      "learning_rate": 3.0985469471301063e-05,
+      "loss": 1.0947,
+      "step": 14086
+    },
+    {
+      "epoch": 2.25392,
+      "grad_norm": 1.819926515109175,
+      "learning_rate": 3.097296992952581e-05,
+      "loss": 1.0952,
+      "step": 14087
+    },
+    {
+      "epoch": 2.25408,
+      "grad_norm": 2.1290848787610237,
+      "learning_rate": 3.096047244741648e-05,
+      "loss": 1.0941,
+      "step": 14088
+    },
+    {
+      "epoch": 2.2542400000000002,
+      "grad_norm": 1.856319535979766,
+      "learning_rate": 3.094797702534608e-05,
+      "loss": 1.0694,
+      "step": 14089
+    },
+    {
+      "epoch": 2.2544,
+      "grad_norm": 1.9833380043955104,
+      "learning_rate": 3.093548366368742e-05,
+      "loss": 1.1014,
+      "step": 14090
+    },
+    {
+      "epoch": 2.25456,
+      "grad_norm": 1.7098352282066593,
+      "learning_rate": 3.092299236281331e-05,
+      "loss": 1.0579,
+      "step": 14091
+    },
+    {
+      "epoch": 2.25472,
+      "grad_norm": 1.6497647075592632,
+      "learning_rate": 3.091050312309645e-05,
+      "loss": 1.0826,
+      "step": 14092
+    },
+    {
+      "epoch": 2.25488,
+      "grad_norm": 1.8225650033305767,
+      "learning_rate": 3.089801594490945e-05,
+      "loss": 1.106,
+      "step": 14093
+    },
+    {
+      "epoch": 2.25504,
+      "grad_norm": 1.9728937152106556,
+      "learning_rate": 3.088553082862504e-05,
+      "loss": 1.0769,
+      "step": 14094
+    },
+    {
+      "epoch": 2.2552,
+      "grad_norm": 1.7104512064820747,
+      "learning_rate": 3.087304777461562e-05,
+      "loss": 1.1288,
+      "step": 14095
+    },
+    {
+      "epoch": 2.25536,
+      "grad_norm": 1.8929433562332239,
+      "learning_rate": 3.0860566783253786e-05,
+      "loss": 1.1102,
+      "step": 14096
+    },
+    {
+      "epoch": 2.25552,
+      "grad_norm": 1.8313979693131912,
+      "learning_rate": 3.084808785491182e-05,
+      "loss": 1.0846,
+      "step": 14097
+    },
+    {
+      "epoch": 2.25568,
+      "grad_norm": 1.8510724998525008,
+      "learning_rate": 3.0835610989962216e-05,
+      "loss": 1.15,
+      "step": 14098
+    },
+    {
+      "epoch": 2.25584,
+      "grad_norm": 1.9212119356183224,
+      "learning_rate": 3.082313618877713e-05,
+      "loss": 1.0965,
+      "step": 14099
+    },
+    {
+      "epoch": 2.2560000000000002,
+      "grad_norm": 1.9069979717942682,
+      "learning_rate": 3.08106634517289e-05,
+      "loss": 1.1242,
+      "step": 14100
+    },
+    {
+      "epoch": 2.25616,
+      "grad_norm": 1.8890978740966138,
+      "learning_rate": 3.079819277918965e-05,
+      "loss": 1.1518,
+      "step": 14101
+    },
+    {
+      "epoch": 2.25632,
+      "grad_norm": 2.5823429596051515,
+      "learning_rate": 3.0785724171531495e-05,
+      "loss": 1.1151,
+      "step": 14102
+    },
+    {
+      "epoch": 2.25648,
+      "grad_norm": 1.6246919396012742,
+      "learning_rate": 3.077325762912646e-05,
+      "loss": 1.133,
+      "step": 14103
+    },
+    {
+      "epoch": 2.25664,
+      "grad_norm": 1.746543736378937,
+      "learning_rate": 3.076079315234656e-05,
+      "loss": 1.1308,
+      "step": 14104
+    },
+    {
+      "epoch": 2.2568,
+      "grad_norm": 1.662062057527673,
+      "learning_rate": 3.074833074156369e-05,
+      "loss": 1.0579,
+      "step": 14105
+    },
+    {
+      "epoch": 2.25696,
+      "grad_norm": 1.7031288493654568,
+      "learning_rate": 3.073587039714972e-05,
+      "loss": 1.1189,
+      "step": 14106
+    },
+    {
+      "epoch": 2.25712,
+      "grad_norm": 2.543910314997392,
+      "learning_rate": 3.072341211947646e-05,
+      "loss": 1.0878,
+      "step": 14107
+    },
+    {
+      "epoch": 2.25728,
+      "grad_norm": 1.9529204807862426,
+      "learning_rate": 3.071095590891562e-05,
+      "loss": 1.1782,
+      "step": 14108
+    },
+    {
+      "epoch": 2.25744,
+      "grad_norm": 1.7557846604592187,
+      "learning_rate": 3.069850176583886e-05,
+      "loss": 1.0893,
+      "step": 14109
+    },
+    {
+      "epoch": 2.2576,
+      "grad_norm": 2.3341851431145906,
+      "learning_rate": 3.06860496906179e-05,
+      "loss": 1.1247,
+      "step": 14110
+    },
+    {
+      "epoch": 2.25776,
+      "grad_norm": 1.9858539568032234,
+      "learning_rate": 3.0673599683624145e-05,
+      "loss": 1.1005,
+      "step": 14111
+    },
+    {
+      "epoch": 2.25792,
+      "grad_norm": 1.8904335072945024,
+      "learning_rate": 3.066115174522922e-05,
+      "loss": 1.1451,
+      "step": 14112
+    },
+    {
+      "epoch": 2.25808,
+      "grad_norm": 1.8285136119917629,
+      "learning_rate": 3.064870587580443e-05,
+      "loss": 1.1321,
+      "step": 14113
+    },
+    {
+      "epoch": 2.25824,
+      "grad_norm": 1.760562915346532,
+      "learning_rate": 3.0636262075721255e-05,
+      "loss": 1.1162,
+      "step": 14114
+    },
+    {
+      "epoch": 2.2584,
+      "grad_norm": 1.824720758664831,
+      "learning_rate": 3.0623820345350895e-05,
+      "loss": 1.0883,
+      "step": 14115
+    },
+    {
+      "epoch": 2.25856,
+      "grad_norm": 1.9562104410389978,
+      "learning_rate": 3.061138068506467e-05,
+      "loss": 1.0629,
+      "step": 14116
+    },
+    {
+      "epoch": 2.25872,
+      "grad_norm": 1.544139796860865,
+      "learning_rate": 3.059894309523374e-05,
+      "loss": 1.119,
+      "step": 14117
+    },
+    {
+      "epoch": 2.25888,
+      "grad_norm": 1.8826965505331297,
+      "learning_rate": 3.0586507576229204e-05,
+      "loss": 1.0938,
+      "step": 14118
+    },
+    {
+      "epoch": 2.25904,
+      "grad_norm": 1.618004323704532,
+      "learning_rate": 3.057407412842216e-05,
+      "loss": 1.1097,
+      "step": 14119
+    },
+    {
+      "epoch": 2.2592,
+      "grad_norm": 1.754480567540998,
+      "learning_rate": 3.0561642752183564e-05,
+      "loss": 1.1704,
+      "step": 14120
+    },
+    {
+      "epoch": 2.25936,
+      "grad_norm": 1.8200799629831672,
+      "learning_rate": 3.054921344788436e-05,
+      "loss": 1.1334,
+      "step": 14121
+    },
+    {
+      "epoch": 2.25952,
+      "grad_norm": 1.8407695123887624,
+      "learning_rate": 3.053678621589544e-05,
+      "loss": 1.1212,
+      "step": 14122
+    },
+    {
+      "epoch": 2.25968,
+      "grad_norm": 1.7118779969238123,
+      "learning_rate": 3.05243610565876e-05,
+      "loss": 1.1035,
+      "step": 14123
+    },
+    {
+      "epoch": 2.25984,
+      "grad_norm": 1.9475492180979277,
+      "learning_rate": 3.0511937970331582e-05,
+      "loss": 1.0827,
+      "step": 14124
+    },
+    {
+      "epoch": 2.26,
+      "grad_norm": 1.8401312522835962,
+      "learning_rate": 3.0499516957498053e-05,
+      "loss": 1.1207,
+      "step": 14125
+    },
+    {
+      "epoch": 2.26016,
+      "grad_norm": 1.7604827251330992,
+      "learning_rate": 3.0487098018457728e-05,
+      "loss": 1.0932,
+      "step": 14126
+    },
+    {
+      "epoch": 2.26032,
+      "grad_norm": 1.6973037052618785,
+      "learning_rate": 3.0474681153581042e-05,
+      "loss": 1.1246,
+      "step": 14127
+    },
+    {
+      "epoch": 2.26048,
+      "grad_norm": 1.941292454899441,
+      "learning_rate": 3.0462266363238622e-05,
+      "loss": 1.1001,
+      "step": 14128
+    },
+    {
+      "epoch": 2.26064,
+      "grad_norm": 2.1211644393544895,
+      "learning_rate": 3.0449853647800786e-05,
+      "loss": 1.1549,
+      "step": 14129
+    },
+    {
+      "epoch": 2.2608,
+      "grad_norm": 1.72194317102081,
+      "learning_rate": 3.0437443007638044e-05,
+      "loss": 1.0818,
+      "step": 14130
+    },
+    {
+      "epoch": 2.26096,
+      "grad_norm": 1.6022614877115637,
+      "learning_rate": 3.042503444312057e-05,
+      "loss": 1.0501,
+      "step": 14131
+    },
+    {
+      "epoch": 2.26112,
+      "grad_norm": 1.6802431710820862,
+      "learning_rate": 3.041262795461871e-05,
+      "loss": 1.0718,
+      "step": 14132
+    },
+    {
+      "epoch": 2.26128,
+      "grad_norm": 1.898440301876149,
+      "learning_rate": 3.0400223542502638e-05,
+      "loss": 1.1111,
+      "step": 14133
+    },
+    {
+      "epoch": 2.26144,
+      "grad_norm": 2.0855682446078285,
+      "learning_rate": 3.0387821207142485e-05,
+      "loss": 1.1484,
+      "step": 14134
+    },
+    {
+      "epoch": 2.2616,
+      "grad_norm": 1.9652978146207436,
+      "learning_rate": 3.0375420948908295e-05,
+      "loss": 1.1329,
+      "step": 14135
+    },
+    {
+      "epoch": 2.2617599999999998,
+      "grad_norm": 1.8584220467091115,
+      "learning_rate": 3.03630227681701e-05,
+      "loss": 1.102,
+      "step": 14136
+    },
+    {
+      "epoch": 2.26192,
+      "grad_norm": 1.7717190038825334,
+      "learning_rate": 3.0350626665297842e-05,
+      "loss": 1.0712,
+      "step": 14137
+    },
+    {
+      "epoch": 2.26208,
+      "grad_norm": 1.8036413324659544,
+      "learning_rate": 3.033823264066138e-05,
+      "loss": 1.1076,
+      "step": 14138
+    },
+    {
+      "epoch": 2.26224,
+      "grad_norm": 1.9407795677758244,
+      "learning_rate": 3.0325840694630557e-05,
+      "loss": 1.1102,
+      "step": 14139
+    },
+    {
+      "epoch": 2.2624,
+      "grad_norm": 1.687750582346107,
+      "learning_rate": 3.0313450827575117e-05,
+      "loss": 1.138,
+      "step": 14140
+    },
+    {
+      "epoch": 2.26256,
+      "grad_norm": 1.7308483524591058,
+      "learning_rate": 3.030106303986474e-05,
+      "loss": 1.0951,
+      "step": 14141
+    },
+    {
+      "epoch": 2.26272,
+      "grad_norm": 1.8684302310742944,
+      "learning_rate": 3.0288677331869132e-05,
+      "loss": 1.131,
+      "step": 14142
+    },
+    {
+      "epoch": 2.26288,
+      "grad_norm": 1.6853899108425412,
+      "learning_rate": 3.0276293703957747e-05,
+      "loss": 1.1103,
+      "step": 14143
+    },
+    {
+      "epoch": 2.26304,
+      "grad_norm": 1.6701059063303587,
+      "learning_rate": 3.026391215650023e-05,
+      "loss": 1.1327,
+      "step": 14144
+    },
+    {
+      "epoch": 2.2632,
+      "grad_norm": 1.665258168103434,
+      "learning_rate": 3.025153268986589e-05,
+      "loss": 1.0848,
+      "step": 14145
+    },
+    {
+      "epoch": 2.26336,
+      "grad_norm": 1.761556588802756,
+      "learning_rate": 3.0239155304424218e-05,
+      "loss": 1.0777,
+      "step": 14146
+    },
+    {
+      "epoch": 2.2635199999999998,
+      "grad_norm": 1.8276550340604216,
+      "learning_rate": 3.0226780000544495e-05,
+      "loss": 1.1214,
+      "step": 14147
+    },
+    {
+      "epoch": 2.26368,
+      "grad_norm": 1.712378362612673,
+      "learning_rate": 3.021440677859598e-05,
+      "loss": 1.1098,
+      "step": 14148
+    },
+    {
+      "epoch": 2.26384,
+      "grad_norm": 1.9348841233600942,
+      "learning_rate": 3.020203563894789e-05,
+      "loss": 1.0882,
+      "step": 14149
+    },
+    {
+      "epoch": 2.2640000000000002,
+      "grad_norm": 1.9522840944228066,
+      "learning_rate": 3.0189666581969345e-05,
+      "loss": 1.099,
+      "step": 14150
+    },
+    {
+      "epoch": 2.26416,
+      "grad_norm": 1.8183785690908325,
+      "learning_rate": 3.017729960802942e-05,
+      "loss": 1.15,
+      "step": 14151
+    },
+    {
+      "epoch": 2.26432,
+      "grad_norm": 1.731877762110283,
+      "learning_rate": 3.0164934717497106e-05,
+      "loss": 1.1112,
+      "step": 14152
+    },
+    {
+      "epoch": 2.26448,
+      "grad_norm": 1.9504614892980243,
+      "learning_rate": 3.0152571910741443e-05,
+      "loss": 1.1508,
+      "step": 14153
+    },
+    {
+      "epoch": 2.26464,
+      "grad_norm": 1.7853554249684176,
+      "learning_rate": 3.0140211188131195e-05,
+      "loss": 1.1512,
+      "step": 14154
+    },
+    {
+      "epoch": 2.2648,
+      "grad_norm": 1.8388061407817113,
+      "learning_rate": 3.0127852550035307e-05,
+      "loss": 1.0945,
+      "step": 14155
+    },
+    {
+      "epoch": 2.26496,
+      "grad_norm": 1.833795140191581,
+      "learning_rate": 3.011549599682242e-05,
+      "loss": 1.0722,
+      "step": 14156
+    },
+    {
+      "epoch": 2.26512,
+      "grad_norm": 1.8141154425960906,
+      "learning_rate": 3.0103141528861333e-05,
+      "loss": 1.1084,
+      "step": 14157
+    },
+    {
+      "epoch": 2.26528,
+      "grad_norm": 1.721004576515317,
+      "learning_rate": 3.009078914652066e-05,
+      "loss": 1.1116,
+      "step": 14158
+    },
+    {
+      "epoch": 2.26544,
+      "grad_norm": 1.6538549201469785,
+      "learning_rate": 3.0078438850168965e-05,
+      "loss": 1.0778,
+      "step": 14159
+    },
+    {
+      "epoch": 2.2656,
+      "grad_norm": 1.9316530748273224,
+      "learning_rate": 3.006609064017476e-05,
+      "loss": 1.1454,
+      "step": 14160
+    },
+    {
+      "epoch": 2.26576,
+      "grad_norm": 1.7746471451620673,
+      "learning_rate": 3.0053744516906522e-05,
+      "loss": 1.1044,
+      "step": 14161
+    },
+    {
+      "epoch": 2.26592,
+      "grad_norm": 1.964008324163678,
+      "learning_rate": 3.0041400480732618e-05,
+      "loss": 1.0519,
+      "step": 14162
+    },
+    {
+      "epoch": 2.26608,
+      "grad_norm": 1.6825109512081797,
+      "learning_rate": 3.0029058532021383e-05,
+      "loss": 1.1269,
+      "step": 14163
+    },
+    {
+      "epoch": 2.26624,
+      "grad_norm": 1.9024154972103564,
+      "learning_rate": 3.0016718671141076e-05,
+      "loss": 1.0987,
+      "step": 14164
+    },
+    {
+      "epoch": 2.2664,
+      "grad_norm": 1.8733685888775238,
+      "learning_rate": 3.0004380898459915e-05,
+      "loss": 1.1039,
+      "step": 14165
+    },
+    {
+      "epoch": 2.26656,
+      "grad_norm": 1.9209219477673645,
+      "learning_rate": 2.9992045214345997e-05,
+      "loss": 1.1573,
+      "step": 14166
+    },
+    {
+      "epoch": 2.26672,
+      "grad_norm": 1.8326419892765196,
+      "learning_rate": 2.9979711619167505e-05,
+      "loss": 1.1054,
+      "step": 14167
+    },
+    {
+      "epoch": 2.26688,
+      "grad_norm": 2.2358089767333835,
+      "learning_rate": 2.996738011329232e-05,
+      "loss": 1.1054,
+      "step": 14168
+    },
+    {
+      "epoch": 2.26704,
+      "grad_norm": 1.9003975852231654,
+      "learning_rate": 2.995505069708854e-05,
+      "loss": 1.1124,
+      "step": 14169
+    },
+    {
+      "epoch": 2.2672,
+      "grad_norm": 2.0411469793430803,
+      "learning_rate": 2.9942723370923896e-05,
+      "loss": 1.1186,
+      "step": 14170
+    },
+    {
+      "epoch": 2.26736,
+      "grad_norm": 1.7238985122628607,
+      "learning_rate": 2.9930398135166382e-05,
+      "loss": 1.1489,
+      "step": 14171
+    },
+    {
+      "epoch": 2.26752,
+      "grad_norm": 1.9708106071906017,
+      "learning_rate": 2.991807499018362e-05,
+      "loss": 1.1393,
+      "step": 14172
+    },
+    {
+      "epoch": 2.26768,
+      "grad_norm": 1.7915112722252697,
+      "learning_rate": 2.9905753936343406e-05,
+      "loss": 1.1167,
+      "step": 14173
+    },
+    {
+      "epoch": 2.26784,
+      "grad_norm": 1.8536201099865404,
+      "learning_rate": 2.9893434974013368e-05,
+      "loss": 1.1252,
+      "step": 14174
+    },
+    {
+      "epoch": 2.268,
+      "grad_norm": 1.7295626206658772,
+      "learning_rate": 2.9881118103561067e-05,
+      "loss": 1.098,
+      "step": 14175
+    },
+    {
+      "epoch": 2.26816,
+      "grad_norm": 1.8572857390193,
+      "learning_rate": 2.986880332535402e-05,
+      "loss": 1.1136,
+      "step": 14176
+    },
+    {
+      "epoch": 2.26832,
+      "grad_norm": 1.7322193661103007,
+      "learning_rate": 2.9856490639759703e-05,
+      "loss": 1.0876,
+      "step": 14177
+    },
+    {
+      "epoch": 2.26848,
+      "grad_norm": 2.3099882002849155,
+      "learning_rate": 2.9844180047145488e-05,
+      "loss": 1.1311,
+      "step": 14178
+    },
+    {
+      "epoch": 2.26864,
+      "grad_norm": 1.8563847168836147,
+      "learning_rate": 2.983187154787871e-05,
+      "loss": 1.128,
+      "step": 14179
+    },
+    {
+      "epoch": 2.2688,
+      "grad_norm": 1.671107931383991,
+      "learning_rate": 2.9819565142326644e-05,
+      "loss": 1.1116,
+      "step": 14180
+    },
+    {
+      "epoch": 2.26896,
+      "grad_norm": 1.97992356789063,
+      "learning_rate": 2.9807260830856487e-05,
+      "loss": 1.0924,
+      "step": 14181
+    },
+    {
+      "epoch": 2.26912,
+      "grad_norm": 1.838446707234354,
+      "learning_rate": 2.9794958613835365e-05,
+      "loss": 1.0836,
+      "step": 14182
+    },
+    {
+      "epoch": 2.26928,
+      "grad_norm": 1.9907201799876006,
+      "learning_rate": 2.9782658491630432e-05,
+      "loss": 1.1205,
+      "step": 14183
+    },
+    {
+      "epoch": 2.26944,
+      "grad_norm": 1.7266242813719395,
+      "learning_rate": 2.9770360464608593e-05,
+      "loss": 1.135,
+      "step": 14184
+    },
+    {
+      "epoch": 2.2696,
+      "grad_norm": 1.8166147820398832,
+      "learning_rate": 2.975806453313692e-05,
+      "loss": 1.1095,
+      "step": 14185
+    },
+    {
+      "epoch": 2.2697599999999998,
+      "grad_norm": 1.8120744677743719,
+      "learning_rate": 2.9745770697582187e-05,
+      "loss": 1.0571,
+      "step": 14186
+    },
+    {
+      "epoch": 2.26992,
+      "grad_norm": 1.7844263493939843,
+      "learning_rate": 2.9733478958311344e-05,
+      "loss": 1.1468,
+      "step": 14187
+    },
+    {
+      "epoch": 2.27008,
+      "grad_norm": 2.243828713797851,
+      "learning_rate": 2.9721189315691046e-05,
+      "loss": 1.1058,
+      "step": 14188
+    },
+    {
+      "epoch": 2.27024,
+      "grad_norm": 1.9106939689922182,
+      "learning_rate": 2.9708901770088084e-05,
+      "loss": 1.1622,
+      "step": 14189
+    },
+    {
+      "epoch": 2.2704,
+      "grad_norm": 1.690050704838226,
+      "learning_rate": 2.9696616321869054e-05,
+      "loss": 1.0491,
+      "step": 14190
+    },
+    {
+      "epoch": 2.27056,
+      "grad_norm": 1.7839550685822663,
+      "learning_rate": 2.9684332971400563e-05,
+      "loss": 1.1236,
+      "step": 14191
+    },
+    {
+      "epoch": 2.27072,
+      "grad_norm": 2.1408411671443557,
+      "learning_rate": 2.967205171904911e-05,
+      "loss": 1.1547,
+      "step": 14192
+    },
+    {
+      "epoch": 2.27088,
+      "grad_norm": 1.8053364711062316,
+      "learning_rate": 2.9659772565181155e-05,
+      "loss": 1.1144,
+      "step": 14193
+    },
+    {
+      "epoch": 2.27104,
+      "grad_norm": 1.7147975578402306,
+      "learning_rate": 2.9647495510163092e-05,
+      "loss": 1.0756,
+      "step": 14194
+    },
+    {
+      "epoch": 2.2712,
+      "grad_norm": 1.6010835141300825,
+      "learning_rate": 2.9635220554361253e-05,
+      "loss": 1.0914,
+      "step": 14195
+    },
+    {
+      "epoch": 2.27136,
+      "grad_norm": 1.700367965461236,
+      "learning_rate": 2.9622947698141888e-05,
+      "loss": 1.1164,
+      "step": 14196
+    },
+    {
+      "epoch": 2.2715199999999998,
+      "grad_norm": 1.8240742839393977,
+      "learning_rate": 2.961067694187123e-05,
+      "loss": 1.0931,
+      "step": 14197
+    },
+    {
+      "epoch": 2.27168,
+      "grad_norm": 1.6708213106190475,
+      "learning_rate": 2.9598408285915358e-05,
+      "loss": 1.1316,
+      "step": 14198
+    },
+    {
+      "epoch": 2.27184,
+      "grad_norm": 1.5174768954755318,
+      "learning_rate": 2.958614173064046e-05,
+      "loss": 1.0909,
+      "step": 14199
+    },
+    {
+      "epoch": 2.2720000000000002,
+      "grad_norm": 1.9068088722807666,
+      "learning_rate": 2.9573877276412433e-05,
+      "loss": 1.0803,
+      "step": 14200
+    },
+    {
+      "epoch": 2.27216,
+      "grad_norm": 1.776215985855388,
+      "learning_rate": 2.9561614923597346e-05,
+      "loss": 1.0723,
+      "step": 14201
+    },
+    {
+      "epoch": 2.27232,
+      "grad_norm": 1.9195856037973815,
+      "learning_rate": 2.9549354672560968e-05,
+      "loss": 1.1427,
+      "step": 14202
+    },
+    {
+      "epoch": 2.27248,
+      "grad_norm": 1.7859897426604314,
+      "learning_rate": 2.953709652366926e-05,
+      "loss": 1.1618,
+      "step": 14203
+    },
+    {
+      "epoch": 2.27264,
+      "grad_norm": 1.7512074318799347,
+      "learning_rate": 2.952484047728784e-05,
+      "loss": 1.1011,
+      "step": 14204
+    },
+    {
+      "epoch": 2.2728,
+      "grad_norm": 2.0323301707904826,
+      "learning_rate": 2.9512586533782526e-05,
+      "loss": 1.1041,
+      "step": 14205
+    },
+    {
+      "epoch": 2.27296,
+      "grad_norm": 1.8376578670768489,
+      "learning_rate": 2.9500334693518918e-05,
+      "loss": 1.1156,
+      "step": 14206
+    },
+    {
+      "epoch": 2.27312,
+      "grad_norm": 1.7899352708627398,
+      "learning_rate": 2.9488084956862593e-05,
+      "loss": 1.0743,
+      "step": 14207
+    },
+    {
+      "epoch": 2.27328,
+      "grad_norm": 1.9385239641882341,
+      "learning_rate": 2.9475837324179066e-05,
+      "loss": 1.1293,
+      "step": 14208
+    },
+    {
+      "epoch": 2.27344,
+      "grad_norm": 1.9011590229551687,
+      "learning_rate": 2.9463591795833755e-05,
+      "loss": 1.1483,
+      "step": 14209
+    },
+    {
+      "epoch": 2.2736,
+      "grad_norm": 1.8533853193039314,
+      "learning_rate": 2.9451348372192157e-05,
+      "loss": 1.1054,
+      "step": 14210
+    },
+    {
+      "epoch": 2.2737600000000002,
+      "grad_norm": 1.7292676493876584,
+      "learning_rate": 2.9439107053619452e-05,
+      "loss": 1.1012,
+      "step": 14211
+    },
+    {
+      "epoch": 2.27392,
+      "grad_norm": 1.767871255469777,
+      "learning_rate": 2.9426867840481033e-05,
+      "loss": 1.1354,
+      "step": 14212
+    },
+    {
+      "epoch": 2.27408,
+      "grad_norm": 1.877607906984433,
+      "learning_rate": 2.9414630733141978e-05,
+      "loss": 1.1725,
+      "step": 14213
+    },
+    {
+      "epoch": 2.27424,
+      "grad_norm": 1.7985632000335587,
+      "learning_rate": 2.9402395731967524e-05,
+      "loss": 1.0832,
+      "step": 14214
+    },
+    {
+      "epoch": 2.2744,
+      "grad_norm": 1.758626863356638,
+      "learning_rate": 2.939016283732271e-05,
+      "loss": 1.103,
+      "step": 14215
+    },
+    {
+      "epoch": 2.27456,
+      "grad_norm": 1.9170847842902987,
+      "learning_rate": 2.9377932049572555e-05,
+      "loss": 1.1233,
+      "step": 14216
+    },
+    {
+      "epoch": 2.27472,
+      "grad_norm": 1.9416030598006568,
+      "learning_rate": 2.9365703369081987e-05,
+      "loss": 1.152,
+      "step": 14217
+    },
+    {
+      "epoch": 2.27488,
+      "grad_norm": 1.9643072810853899,
+      "learning_rate": 2.9353476796215918e-05,
+      "loss": 1.1619,
+      "step": 14218
+    },
+    {
+      "epoch": 2.27504,
+      "grad_norm": 1.832489657525409,
+      "learning_rate": 2.9341252331339152e-05,
+      "loss": 1.0689,
+      "step": 14219
+    },
+    {
+      "epoch": 2.2752,
+      "grad_norm": 1.6045326618264604,
+      "learning_rate": 2.9329029974816445e-05,
+      "loss": 1.086,
+      "step": 14220
+    },
+    {
+      "epoch": 2.27536,
+      "grad_norm": 1.805312789362136,
+      "learning_rate": 2.9316809727012517e-05,
+      "loss": 1.1197,
+      "step": 14221
+    },
+    {
+      "epoch": 2.27552,
+      "grad_norm": 1.8962299225477266,
+      "learning_rate": 2.9304591588291998e-05,
+      "loss": 1.1459,
+      "step": 14222
+    },
+    {
+      "epoch": 2.27568,
+      "grad_norm": 1.798492772570301,
+      "learning_rate": 2.9292375559019403e-05,
+      "loss": 1.0717,
+      "step": 14223
+    },
+    {
+      "epoch": 2.27584,
+      "grad_norm": 1.7326058188348468,
+      "learning_rate": 2.9280161639559367e-05,
+      "loss": 1.1342,
+      "step": 14224
+    },
+    {
+      "epoch": 2.276,
+      "grad_norm": 1.8885508097099237,
+      "learning_rate": 2.9267949830276188e-05,
+      "loss": 1.0834,
+      "step": 14225
+    },
+    {
+      "epoch": 2.27616,
+      "grad_norm": 1.8808573477756168,
+      "learning_rate": 2.9255740131534382e-05,
+      "loss": 1.1011,
+      "step": 14226
+    },
+    {
+      "epoch": 2.27632,
+      "grad_norm": 1.6838634169265179,
+      "learning_rate": 2.924353254369814e-05,
+      "loss": 1.0605,
+      "step": 14227
+    },
+    {
+      "epoch": 2.27648,
+      "grad_norm": 1.878012783571601,
+      "learning_rate": 2.9231327067131853e-05,
+      "loss": 1.0863,
+      "step": 14228
+    },
+    {
+      "epoch": 2.27664,
+      "grad_norm": 1.9715173128314216,
+      "learning_rate": 2.921912370219958e-05,
+      "loss": 1.1164,
+      "step": 14229
+    },
+    {
+      "epoch": 2.2768,
+      "grad_norm": 1.682408970896776,
+      "learning_rate": 2.920692244926555e-05,
+      "loss": 1.0848,
+      "step": 14230
+    },
+    {
+      "epoch": 2.27696,
+      "grad_norm": 1.8734893683522864,
+      "learning_rate": 2.9194723308693795e-05,
+      "loss": 1.0758,
+      "step": 14231
+    },
+    {
+      "epoch": 2.27712,
+      "grad_norm": 1.9107907832419615,
+      "learning_rate": 2.918252628084832e-05,
+      "loss": 1.0939,
+      "step": 14232
+    },
+    {
+      "epoch": 2.27728,
+      "grad_norm": 1.5780295092436498,
+      "learning_rate": 2.9170331366093073e-05,
+      "loss": 1.0974,
+      "step": 14233
+    },
+    {
+      "epoch": 2.27744,
+      "grad_norm": 1.6585661187575562,
+      "learning_rate": 2.9158138564791925e-05,
+      "loss": 1.07,
+      "step": 14234
+    },
+    {
+      "epoch": 2.2776,
+      "grad_norm": 1.7458225211376324,
+      "learning_rate": 2.91459478773087e-05,
+      "loss": 1.0971,
+      "step": 14235
+    },
+    {
+      "epoch": 2.27776,
+      "grad_norm": 1.9322704455954491,
+      "learning_rate": 2.9133759304007135e-05,
+      "loss": 1.1074,
+      "step": 14236
+    },
+    {
+      "epoch": 2.27792,
+      "grad_norm": 1.8728026719614532,
+      "learning_rate": 2.912157284525093e-05,
+      "loss": 1.0726,
+      "step": 14237
+    },
+    {
+      "epoch": 2.27808,
+      "grad_norm": 1.6161580799929844,
+      "learning_rate": 2.910938850140372e-05,
+      "loss": 1.0891,
+      "step": 14238
+    },
+    {
+      "epoch": 2.27824,
+      "grad_norm": 1.7828680470739586,
+      "learning_rate": 2.9097206272829024e-05,
+      "loss": 1.0987,
+      "step": 14239
+    },
+    {
+      "epoch": 2.2784,
+      "grad_norm": 1.8061778352995728,
+      "learning_rate": 2.908502615989044e-05,
+      "loss": 1.0822,
+      "step": 14240
+    },
+    {
+      "epoch": 2.27856,
+      "grad_norm": 1.84058470030824,
+      "learning_rate": 2.9072848162951282e-05,
+      "loss": 1.0741,
+      "step": 14241
+    },
+    {
+      "epoch": 2.27872,
+      "grad_norm": 2.531388409250441,
+      "learning_rate": 2.9060672282375036e-05,
+      "loss": 1.076,
+      "step": 14242
+    },
+    {
+      "epoch": 2.27888,
+      "grad_norm": 1.7595327191737176,
+      "learning_rate": 2.9048498518524903e-05,
+      "loss": 1.0792,
+      "step": 14243
+    },
+    {
+      "epoch": 2.27904,
+      "grad_norm": 1.8353425228986484,
+      "learning_rate": 2.9036326871764252e-05,
+      "loss": 1.152,
+      "step": 14244
+    },
+    {
+      "epoch": 2.2792,
+      "grad_norm": 2.5092489695459537,
+      "learning_rate": 2.9024157342456137e-05,
+      "loss": 1.1138,
+      "step": 14245
+    },
+    {
+      "epoch": 2.27936,
+      "grad_norm": 1.8952132250355616,
+      "learning_rate": 2.9011989930963768e-05,
+      "loss": 1.0878,
+      "step": 14246
+    },
+    {
+      "epoch": 2.2795199999999998,
+      "grad_norm": 1.7982869563398654,
+      "learning_rate": 2.8999824637650186e-05,
+      "loss": 1.1116,
+      "step": 14247
+    },
+    {
+      "epoch": 2.27968,
+      "grad_norm": 1.9322614506543592,
+      "learning_rate": 2.898766146287838e-05,
+      "loss": 1.0372,
+      "step": 14248
+    },
+    {
+      "epoch": 2.27984,
+      "grad_norm": 1.6915574946956742,
+      "learning_rate": 2.8975500407011282e-05,
+      "loss": 1.1016,
+      "step": 14249
+    },
+    {
+      "epoch": 2.2800000000000002,
+      "grad_norm": 1.8309110321595112,
+      "learning_rate": 2.8963341470411755e-05,
+      "loss": 1.0909,
+      "step": 14250
+    },
+    {
+      "epoch": 2.28016,
+      "grad_norm": 1.746148953329229,
+      "learning_rate": 2.895118465344261e-05,
+      "loss": 1.085,
+      "step": 14251
+    },
+    {
+      "epoch": 2.28032,
+      "grad_norm": 1.8201408779301649,
+      "learning_rate": 2.893902995646658e-05,
+      "loss": 1.11,
+      "step": 14252
+    },
+    {
+      "epoch": 2.28048,
+      "grad_norm": 1.8361188194485716,
+      "learning_rate": 2.892687737984635e-05,
+      "loss": 1.0604,
+      "step": 14253
+    },
+    {
+      "epoch": 2.28064,
+      "grad_norm": 1.7412824575933599,
+      "learning_rate": 2.8914726923944536e-05,
+      "loss": 1.0839,
+      "step": 14254
+    },
+    {
+      "epoch": 2.2808,
+      "grad_norm": 2.540536374416152,
+      "learning_rate": 2.8902578589123653e-05,
+      "loss": 1.1097,
+      "step": 14255
+    },
+    {
+      "epoch": 2.28096,
+      "grad_norm": 1.614863105467984,
+      "learning_rate": 2.88904323757463e-05,
+      "loss": 1.0873,
+      "step": 14256
+    },
+    {
+      "epoch": 2.28112,
+      "grad_norm": 1.6681300970515665,
+      "learning_rate": 2.8878288284174748e-05,
+      "loss": 1.1174,
+      "step": 14257
+    },
+    {
+      "epoch": 2.2812799999999998,
+      "grad_norm": 1.7928897200148062,
+      "learning_rate": 2.8866146314771515e-05,
+      "loss": 1.1169,
+      "step": 14258
+    },
+    {
+      "epoch": 2.28144,
+      "grad_norm": 1.6932212255631767,
+      "learning_rate": 2.8854006467898752e-05,
+      "loss": 1.0815,
+      "step": 14259
+    },
+    {
+      "epoch": 2.2816,
+      "grad_norm": 1.770566811246949,
+      "learning_rate": 2.884186874391883e-05,
+      "loss": 1.0691,
+      "step": 14260
+    },
+    {
+      "epoch": 2.2817600000000002,
+      "grad_norm": 1.9646028076537883,
+      "learning_rate": 2.8829733143193806e-05,
+      "loss": 1.1238,
+      "step": 14261
+    },
+    {
+      "epoch": 2.28192,
+      "grad_norm": 1.7720873974383606,
+      "learning_rate": 2.881759966608585e-05,
+      "loss": 1.0418,
+      "step": 14262
+    },
+    {
+      "epoch": 2.28208,
+      "grad_norm": 1.7780950327371408,
+      "learning_rate": 2.8805468312957007e-05,
+      "loss": 1.1231,
+      "step": 14263
+    },
+    {
+      "epoch": 2.28224,
+      "grad_norm": 1.8625245787912954,
+      "learning_rate": 2.8793339084169245e-05,
+      "loss": 1.127,
+      "step": 14264
+    },
+    {
+      "epoch": 2.2824,
+      "grad_norm": 1.9494387625453393,
+      "learning_rate": 2.8781211980084488e-05,
+      "loss": 1.1115,
+      "step": 14265
+    },
+    {
+      "epoch": 2.28256,
+      "grad_norm": 1.9695122654218866,
+      "learning_rate": 2.8769087001064588e-05,
+      "loss": 1.0919,
+      "step": 14266
+    },
+    {
+      "epoch": 2.28272,
+      "grad_norm": 1.791674534724661,
+      "learning_rate": 2.8756964147471344e-05,
+      "loss": 1.1359,
+      "step": 14267
+    },
+    {
+      "epoch": 2.28288,
+      "grad_norm": 1.6613914989157705,
+      "learning_rate": 2.8744843419666435e-05,
+      "loss": 1.1058,
+      "step": 14268
+    },
+    {
+      "epoch": 2.28304,
+      "grad_norm": 2.1582214452694966,
+      "learning_rate": 2.8732724818011636e-05,
+      "loss": 1.1207,
+      "step": 14269
+    },
+    {
+      "epoch": 2.2832,
+      "grad_norm": 1.743266730864685,
+      "learning_rate": 2.8720608342868416e-05,
+      "loss": 1.0825,
+      "step": 14270
+    },
+    {
+      "epoch": 2.28336,
+      "grad_norm": 1.8209974325515954,
+      "learning_rate": 2.8708493994598418e-05,
+      "loss": 1.0721,
+      "step": 14271
+    },
+    {
+      "epoch": 2.28352,
+      "grad_norm": 1.96252464924817,
+      "learning_rate": 2.8696381773563064e-05,
+      "loss": 1.0715,
+      "step": 14272
+    },
+    {
+      "epoch": 2.28368,
+      "grad_norm": 1.6993952977506388,
+      "learning_rate": 2.8684271680123787e-05,
+      "loss": 1.1296,
+      "step": 14273
+    },
+    {
+      "epoch": 2.28384,
+      "grad_norm": 1.7952808038781478,
+      "learning_rate": 2.867216371464192e-05,
+      "loss": 1.0851,
+      "step": 14274
+    },
+    {
+      "epoch": 2.284,
+      "grad_norm": 1.7407476611969737,
+      "learning_rate": 2.8660057877478753e-05,
+      "loss": 1.0931,
+      "step": 14275
+    },
+    {
+      "epoch": 2.28416,
+      "grad_norm": 1.8977379832766885,
+      "learning_rate": 2.8647954168995506e-05,
+      "loss": 1.1064,
+      "step": 14276
+    },
+    {
+      "epoch": 2.28432,
+      "grad_norm": 2.0027222864326832,
+      "learning_rate": 2.8635852589553336e-05,
+      "loss": 1.098,
+      "step": 14277
+    },
+    {
+      "epoch": 2.28448,
+      "grad_norm": 1.9913797063732646,
+      "learning_rate": 2.8623753139513343e-05,
+      "loss": 1.1102,
+      "step": 14278
+    },
+    {
+      "epoch": 2.28464,
+      "grad_norm": 1.9615290125762934,
+      "learning_rate": 2.8611655819236537e-05,
+      "loss": 1.1031,
+      "step": 14279
+    },
+    {
+      "epoch": 2.2848,
+      "grad_norm": 1.8703026587521068,
+      "learning_rate": 2.8599560629083876e-05,
+      "loss": 1.0685,
+      "step": 14280
+    },
+    {
+      "epoch": 2.28496,
+      "grad_norm": 1.991233305259693,
+      "learning_rate": 2.8587467569416347e-05,
+      "loss": 1.1535,
+      "step": 14281
+    },
+    {
+      "epoch": 2.28512,
+      "grad_norm": 1.8034218671417943,
+      "learning_rate": 2.8575376640594652e-05,
+      "loss": 1.1172,
+      "step": 14282
+    },
+    {
+      "epoch": 2.28528,
+      "grad_norm": 2.2645583557664932,
+      "learning_rate": 2.8563287842979714e-05,
+      "loss": 1.1304,
+      "step": 14283
+    },
+    {
+      "epoch": 2.28544,
+      "grad_norm": 2.444091935363063,
+      "learning_rate": 2.855120117693212e-05,
+      "loss": 1.0957,
+      "step": 14284
+    },
+    {
+      "epoch": 2.2856,
+      "grad_norm": 1.882563175162864,
+      "learning_rate": 2.8539116642812612e-05,
+      "loss": 1.1362,
+      "step": 14285
+    },
+    {
+      "epoch": 2.28576,
+      "grad_norm": 1.594128390001747,
+      "learning_rate": 2.8527034240981686e-05,
+      "loss": 1.0906,
+      "step": 14286
+    },
+    {
+      "epoch": 2.28592,
+      "grad_norm": 1.950054230433158,
+      "learning_rate": 2.851495397179994e-05,
+      "loss": 1.1038,
+      "step": 14287
+    },
+    {
+      "epoch": 2.28608,
+      "grad_norm": 2.234996960533912,
+      "learning_rate": 2.8502875835627797e-05,
+      "loss": 1.0889,
+      "step": 14288
+    },
+    {
+      "epoch": 2.28624,
+      "grad_norm": 1.9535211063619586,
+      "learning_rate": 2.8490799832825666e-05,
+      "loss": 1.1124,
+      "step": 14289
+    },
+    {
+      "epoch": 2.2864,
+      "grad_norm": 2.5282486659996564,
+      "learning_rate": 2.847872596375386e-05,
+      "loss": 1.1235,
+      "step": 14290
+    },
+    {
+      "epoch": 2.28656,
+      "grad_norm": 1.8011176183998525,
+      "learning_rate": 2.846665422877265e-05,
+      "loss": 1.1715,
+      "step": 14291
+    },
+    {
+      "epoch": 2.28672,
+      "grad_norm": 1.8330710538146806,
+      "learning_rate": 2.8454584628242253e-05,
+      "loss": 1.1016,
+      "step": 14292
+    },
+    {
+      "epoch": 2.28688,
+      "grad_norm": 2.132424841736967,
+      "learning_rate": 2.8442517162522797e-05,
+      "loss": 1.1198,
+      "step": 14293
+    },
+    {
+      "epoch": 2.28704,
+      "grad_norm": 2.372207950857883,
+      "learning_rate": 2.843045183197436e-05,
+      "loss": 1.1006,
+      "step": 14294
+    },
+    {
+      "epoch": 2.2872,
+      "grad_norm": 1.8953321326425465,
+      "learning_rate": 2.8418388636956938e-05,
+      "loss": 1.1203,
+      "step": 14295
+    },
+    {
+      "epoch": 2.28736,
+      "grad_norm": 1.7294408844138334,
+      "learning_rate": 2.840632757783047e-05,
+      "loss": 1.1029,
+      "step": 14296
+    },
+    {
+      "epoch": 2.2875199999999998,
+      "grad_norm": 1.703784794488712,
+      "learning_rate": 2.8394268654954938e-05,
+      "loss": 1.1025,
+      "step": 14297
+    },
+    {
+      "epoch": 2.28768,
+      "grad_norm": 1.9356260205461717,
+      "learning_rate": 2.8382211868690012e-05,
+      "loss": 1.0519,
+      "step": 14298
+    },
+    {
+      "epoch": 2.28784,
+      "grad_norm": 1.7419603034353561,
+      "learning_rate": 2.8370157219395598e-05,
+      "loss": 1.1018,
+      "step": 14299
+    },
+    {
+      "epoch": 2.288,
+      "grad_norm": 1.8105442505612226,
+      "learning_rate": 2.8358104707431245e-05,
+      "loss": 1.1282,
+      "step": 14300
+    },
+    {
+      "epoch": 2.28816,
+      "grad_norm": 1.7241710056338215,
+      "learning_rate": 2.8346054333156724e-05,
+      "loss": 1.0686,
+      "step": 14301
+    },
+    {
+      "epoch": 2.28832,
+      "grad_norm": 2.0319764233601125,
+      "learning_rate": 2.8334006096931465e-05,
+      "loss": 1.1313,
+      "step": 14302
+    },
+    {
+      "epoch": 2.28848,
+      "grad_norm": 1.631136259782853,
+      "learning_rate": 2.8321959999115066e-05,
+      "loss": 1.1271,
+      "step": 14303
+    },
+    {
+      "epoch": 2.28864,
+      "grad_norm": 1.8074012722678763,
+      "learning_rate": 2.830991604006694e-05,
+      "loss": 1.119,
+      "step": 14304
+    },
+    {
+      "epoch": 2.2888,
+      "grad_norm": 2.012233546008447,
+      "learning_rate": 2.8297874220146446e-05,
+      "loss": 1.1214,
+      "step": 14305
+    },
+    {
+      "epoch": 2.28896,
+      "grad_norm": 1.597293033763766,
+      "learning_rate": 2.828583453971292e-05,
+      "loss": 1.1327,
+      "step": 14306
+    },
+    {
+      "epoch": 2.28912,
+      "grad_norm": 1.7049697706737776,
+      "learning_rate": 2.827379699912558e-05,
+      "loss": 1.133,
+      "step": 14307
+    },
+    {
+      "epoch": 2.2892799999999998,
+      "grad_norm": 1.6414379050474674,
+      "learning_rate": 2.826176159874363e-05,
+      "loss": 1.0932,
+      "step": 14308
+    },
+    {
+      "epoch": 2.28944,
+      "grad_norm": 1.7928982143718535,
+      "learning_rate": 2.8249728338926175e-05,
+      "loss": 1.0631,
+      "step": 14309
+    },
+    {
+      "epoch": 2.2896,
+      "grad_norm": 1.9363803288993584,
+      "learning_rate": 2.8237697220032266e-05,
+      "loss": 1.1111,
+      "step": 14310
+    },
+    {
+      "epoch": 2.2897600000000002,
+      "grad_norm": 1.7691361848655853,
+      "learning_rate": 2.8225668242420923e-05,
+      "loss": 1.1215,
+      "step": 14311
+    },
+    {
+      "epoch": 2.28992,
+      "grad_norm": 2.070072390589471,
+      "learning_rate": 2.821364140645101e-05,
+      "loss": 1.1339,
+      "step": 14312
+    },
+    {
+      "epoch": 2.29008,
+      "grad_norm": 2.024525117211653,
+      "learning_rate": 2.8201616712481494e-05,
+      "loss": 1.171,
+      "step": 14313
+    },
+    {
+      "epoch": 2.29024,
+      "grad_norm": 2.0968486229111436,
+      "learning_rate": 2.8189594160871058e-05,
+      "loss": 1.1308,
+      "step": 14314
+    },
+    {
+      "epoch": 2.2904,
+      "grad_norm": 1.68370393455643,
+      "learning_rate": 2.8177573751978546e-05,
+      "loss": 1.098,
+      "step": 14315
+    },
+    {
+      "epoch": 2.29056,
+      "grad_norm": 1.8884928956491749,
+      "learning_rate": 2.8165555486162533e-05,
+      "loss": 1.1031,
+      "step": 14316
+    },
+    {
+      "epoch": 2.29072,
+      "grad_norm": 1.9013979163849737,
+      "learning_rate": 2.8153539363781712e-05,
+      "loss": 1.0852,
+      "step": 14317
+    },
+    {
+      "epoch": 2.29088,
+      "grad_norm": 1.6127495725019585,
+      "learning_rate": 2.8141525385194535e-05,
+      "loss": 1.1058,
+      "step": 14318
+    },
+    {
+      "epoch": 2.29104,
+      "grad_norm": 1.8182876465897775,
+      "learning_rate": 2.8129513550759556e-05,
+      "loss": 1.0828,
+      "step": 14319
+    },
+    {
+      "epoch": 2.2912,
+      "grad_norm": 1.8529100532455636,
+      "learning_rate": 2.811750386083517e-05,
+      "loss": 1.0963,
+      "step": 14320
+    },
+    {
+      "epoch": 2.29136,
+      "grad_norm": 2.013024493786357,
+      "learning_rate": 2.8105496315779734e-05,
+      "loss": 1.0359,
+      "step": 14321
+    },
+    {
+      "epoch": 2.2915200000000002,
+      "grad_norm": 1.6970904317684081,
+      "learning_rate": 2.8093490915951516e-05,
+      "loss": 1.0686,
+      "step": 14322
+    },
+    {
+      "epoch": 2.29168,
+      "grad_norm": 1.726474049852487,
+      "learning_rate": 2.808148766170875e-05,
+      "loss": 1.1322,
+      "step": 14323
+    },
+    {
+      "epoch": 2.29184,
+      "grad_norm": 1.859239264594475,
+      "learning_rate": 2.8069486553409597e-05,
+      "loss": 1.0631,
+      "step": 14324
+    },
+    {
+      "epoch": 2.292,
+      "grad_norm": 1.7906669762214553,
+      "learning_rate": 2.8057487591412158e-05,
+      "loss": 1.1084,
+      "step": 14325
+    },
+    {
+      "epoch": 2.29216,
+      "grad_norm": 1.807606754419064,
+      "learning_rate": 2.8045490776074446e-05,
+      "loss": 1.1274,
+      "step": 14326
+    },
+    {
+      "epoch": 2.29232,
+      "grad_norm": 1.8947177879927395,
+      "learning_rate": 2.803349610775442e-05,
+      "loss": 1.0774,
+      "step": 14327
+    },
+    {
+      "epoch": 2.29248,
+      "grad_norm": 1.7144980686507783,
+      "learning_rate": 2.8021503586810027e-05,
+      "loss": 1.0601,
+      "step": 14328
+    },
+    {
+      "epoch": 2.29264,
+      "grad_norm": 1.7517092103991214,
+      "learning_rate": 2.8009513213599092e-05,
+      "loss": 1.0857,
+      "step": 14329
+    },
+    {
+      "epoch": 2.2928,
+      "grad_norm": 1.8879027627220273,
+      "learning_rate": 2.7997524988479372e-05,
+      "loss": 1.1258,
+      "step": 14330
+    },
+    {
+      "epoch": 2.29296,
+      "grad_norm": 1.9310138999544544,
+      "learning_rate": 2.7985538911808595e-05,
+      "loss": 1.07,
+      "step": 14331
+    },
+    {
+      "epoch": 2.29312,
+      "grad_norm": 1.631784914903473,
+      "learning_rate": 2.7973554983944395e-05,
+      "loss": 1.1075,
+      "step": 14332
+    },
+    {
+      "epoch": 2.29328,
+      "grad_norm": 1.766054752745892,
+      "learning_rate": 2.7961573205244363e-05,
+      "loss": 1.0798,
+      "step": 14333
+    },
+    {
+      "epoch": 2.29344,
+      "grad_norm": 1.7799125862632803,
+      "learning_rate": 2.7949593576066015e-05,
+      "loss": 1.0732,
+      "step": 14334
+    },
+    {
+      "epoch": 2.2936,
+      "grad_norm": 1.7483259589628264,
+      "learning_rate": 2.7937616096766804e-05,
+      "loss": 1.0766,
+      "step": 14335
+    },
+    {
+      "epoch": 2.29376,
+      "grad_norm": 1.7466227195943773,
+      "learning_rate": 2.792564076770413e-05,
+      "loss": 1.09,
+      "step": 14336
+    },
+    {
+      "epoch": 2.29392,
+      "grad_norm": 1.5903142753902024,
+      "learning_rate": 2.7913667589235305e-05,
+      "loss": 1.0593,
+      "step": 14337
+    },
+    {
+      "epoch": 2.29408,
+      "grad_norm": 1.7674474564979499,
+      "learning_rate": 2.7901696561717604e-05,
+      "loss": 1.1104,
+      "step": 14338
+    },
+    {
+      "epoch": 2.29424,
+      "grad_norm": 1.8184925626790842,
+      "learning_rate": 2.788972768550818e-05,
+      "loss": 1.2133,
+      "step": 14339
+    },
+    {
+      "epoch": 2.2944,
+      "grad_norm": 1.9134040753548143,
+      "learning_rate": 2.7877760960964272e-05,
+      "loss": 1.0872,
+      "step": 14340
+    },
+    {
+      "epoch": 2.29456,
+      "grad_norm": 2.2956507087144167,
+      "learning_rate": 2.786579638844282e-05,
+      "loss": 1.1123,
+      "step": 14341
+    },
+    {
+      "epoch": 2.29472,
+      "grad_norm": 2.125856478255092,
+      "learning_rate": 2.7853833968300968e-05,
+      "loss": 1.1513,
+      "step": 14342
+    },
+    {
+      "epoch": 2.29488,
+      "grad_norm": 2.213759376465757,
+      "learning_rate": 2.7841873700895503e-05,
+      "loss": 1.1357,
+      "step": 14343
+    },
+    {
+      "epoch": 2.29504,
+      "grad_norm": 1.6791700987644258,
+      "learning_rate": 2.782991558658342e-05,
+      "loss": 1.0701,
+      "step": 14344
+    },
+    {
+      "epoch": 2.2952,
+      "grad_norm": 1.8677126576641432,
+      "learning_rate": 2.78179596257215e-05,
+      "loss": 1.1191,
+      "step": 14345
+    },
+    {
+      "epoch": 2.29536,
+      "grad_norm": 1.8343007021966666,
+      "learning_rate": 2.780600581866647e-05,
+      "loss": 1.1145,
+      "step": 14346
+    },
+    {
+      "epoch": 2.29552,
+      "grad_norm": 1.7271607063267347,
+      "learning_rate": 2.779405416577504e-05,
+      "loss": 1.1363,
+      "step": 14347
+    },
+    {
+      "epoch": 2.29568,
+      "grad_norm": 1.8064381053903866,
+      "learning_rate": 2.778210466740381e-05,
+      "loss": 1.0928,
+      "step": 14348
+    },
+    {
+      "epoch": 2.29584,
+      "grad_norm": 1.7296681284415367,
+      "learning_rate": 2.7770157323909353e-05,
+      "loss": 1.1084,
+      "step": 14349
+    },
+    {
+      "epoch": 2.296,
+      "grad_norm": 1.8176920085291364,
+      "learning_rate": 2.7758212135648143e-05,
+      "loss": 1.1362,
+      "step": 14350
+    },
+    {
+      "epoch": 2.29616,
+      "grad_norm": 1.8045372577473358,
+      "learning_rate": 2.774626910297662e-05,
+      "loss": 1.0942,
+      "step": 14351
+    },
+    {
+      "epoch": 2.29632,
+      "grad_norm": 1.63822495313963,
+      "learning_rate": 2.773432822625114e-05,
+      "loss": 1.0684,
+      "step": 14352
+    },
+    {
+      "epoch": 2.29648,
+      "grad_norm": 1.7240334063270844,
+      "learning_rate": 2.7722389505828005e-05,
+      "loss": 1.1565,
+      "step": 14353
+    },
+    {
+      "epoch": 2.29664,
+      "grad_norm": 1.6476711335704455,
+      "learning_rate": 2.7710452942063448e-05,
+      "loss": 1.063,
+      "step": 14354
+    },
+    {
+      "epoch": 2.2968,
+      "grad_norm": 1.9941502511706812,
+      "learning_rate": 2.769851853531361e-05,
+      "loss": 1.0918,
+      "step": 14355
+    },
+    {
+      "epoch": 2.29696,
+      "grad_norm": 1.6588369090009676,
+      "learning_rate": 2.768658628593468e-05,
+      "loss": 1.0496,
+      "step": 14356
+    },
+    {
+      "epoch": 2.29712,
+      "grad_norm": 1.7346138430559133,
+      "learning_rate": 2.767465619428259e-05,
+      "loss": 1.1052,
+      "step": 14357
+    },
+    {
+      "epoch": 2.2972799999999998,
+      "grad_norm": 1.8069205862440143,
+      "learning_rate": 2.7662728260713423e-05,
+      "loss": 1.0976,
+      "step": 14358
+    },
+    {
+      "epoch": 2.29744,
+      "grad_norm": 2.0061457052450837,
+      "learning_rate": 2.7650802485582984e-05,
+      "loss": 1.1016,
+      "step": 14359
+    },
+    {
+      "epoch": 2.2976,
+      "grad_norm": 1.7908914570626793,
+      "learning_rate": 2.7638878869247196e-05,
+      "loss": 1.1323,
+      "step": 14360
+    },
+    {
+      "epoch": 2.2977600000000002,
+      "grad_norm": 1.6189229316369387,
+      "learning_rate": 2.7626957412061828e-05,
+      "loss": 1.0482,
+      "step": 14361
+    },
+    {
+      "epoch": 2.29792,
+      "grad_norm": 1.727803014700048,
+      "learning_rate": 2.7615038114382587e-05,
+      "loss": 1.1069,
+      "step": 14362
+    },
+    {
+      "epoch": 2.29808,
+      "grad_norm": 1.8635660075813034,
+      "learning_rate": 2.7603120976565144e-05,
+      "loss": 1.1264,
+      "step": 14363
+    },
+    {
+      "epoch": 2.29824,
+      "grad_norm": 1.7690833151107546,
+      "learning_rate": 2.7591205998965065e-05,
+      "loss": 1.0687,
+      "step": 14364
+    },
+    {
+      "epoch": 2.2984,
+      "grad_norm": 1.6921853865833616,
+      "learning_rate": 2.7579293181937893e-05,
+      "loss": 1.0889,
+      "step": 14365
+    },
+    {
+      "epoch": 2.29856,
+      "grad_norm": 1.744354643288015,
+      "learning_rate": 2.756738252583909e-05,
+      "loss": 1.1452,
+      "step": 14366
+    },
+    {
+      "epoch": 2.29872,
+      "grad_norm": 1.9501763969399102,
+      "learning_rate": 2.755547403102404e-05,
+      "loss": 1.0754,
+      "step": 14367
+    },
+    {
+      "epoch": 2.29888,
+      "grad_norm": 1.6945231169905945,
+      "learning_rate": 2.754356769784808e-05,
+      "loss": 1.1263,
+      "step": 14368
+    },
+    {
+      "epoch": 2.2990399999999998,
+      "grad_norm": 1.8271197550881433,
+      "learning_rate": 2.753166352666645e-05,
+      "loss": 1.0962,
+      "step": 14369
+    },
+    {
+      "epoch": 2.2992,
+      "grad_norm": 1.7765782722339356,
+      "learning_rate": 2.7519761517834453e-05,
+      "loss": 1.1195,
+      "step": 14370
+    },
+    {
+      "epoch": 2.29936,
+      "grad_norm": 1.6894107812435744,
+      "learning_rate": 2.7507861671707093e-05,
+      "loss": 1.1097,
+      "step": 14371
+    },
+    {
+      "epoch": 2.2995200000000002,
+      "grad_norm": 1.676635608197468,
+      "learning_rate": 2.749596398863956e-05,
+      "loss": 1.1211,
+      "step": 14372
+    },
+    {
+      "epoch": 2.29968,
+      "grad_norm": 1.8442692756178793,
+      "learning_rate": 2.7484068468986757e-05,
+      "loss": 1.1061,
+      "step": 14373
+    },
+    {
+      "epoch": 2.29984,
+      "grad_norm": 1.7444234180122284,
+      "learning_rate": 2.747217511310375e-05,
+      "loss": 1.0563,
+      "step": 14374
+    },
+    {
+      "epoch": 2.3,
+      "grad_norm": 1.6702106636643748,
+      "learning_rate": 2.7460283921345288e-05,
+      "loss": 1.1168,
+      "step": 14375
+    },
+    {
+      "epoch": 2.30016,
+      "grad_norm": 1.843798190206972,
+      "learning_rate": 2.7448394894066288e-05,
+      "loss": 1.0894,
+      "step": 14376
+    },
+    {
+      "epoch": 2.30032,
+      "grad_norm": 1.964645682877025,
+      "learning_rate": 2.7436508031621453e-05,
+      "loss": 1.109,
+      "step": 14377
+    },
+    {
+      "epoch": 2.30048,
+      "grad_norm": 1.7866761539750136,
+      "learning_rate": 2.7424623334365495e-05,
+      "loss": 1.1029,
+      "step": 14378
+    },
+    {
+      "epoch": 2.30064,
+      "grad_norm": 1.8063663553654243,
+      "learning_rate": 2.7412740802653014e-05,
+      "loss": 1.0679,
+      "step": 14379
+    },
+    {
+      "epoch": 2.3008,
+      "grad_norm": 1.7668212628731568,
+      "learning_rate": 2.7400860436838582e-05,
+      "loss": 1.0745,
+      "step": 14380
+    },
+    {
+      "epoch": 2.30096,
+      "grad_norm": 2.115369534822095,
+      "learning_rate": 2.7388982237276682e-05,
+      "loss": 1.0791,
+      "step": 14381
+    },
+    {
+      "epoch": 2.30112,
+      "grad_norm": 1.9032988017380121,
+      "learning_rate": 2.7377106204321757e-05,
+      "loss": 1.1259,
+      "step": 14382
+    },
+    {
+      "epoch": 2.30128,
+      "grad_norm": 1.875495500951509,
+      "learning_rate": 2.7365232338328152e-05,
+      "loss": 1.1406,
+      "step": 14383
+    },
+    {
+      "epoch": 2.30144,
+      "grad_norm": 2.078883764693099,
+      "learning_rate": 2.7353360639650173e-05,
+      "loss": 1.1087,
+      "step": 14384
+    },
+    {
+      "epoch": 2.3016,
+      "grad_norm": 1.8629764990026079,
+      "learning_rate": 2.7341491108642026e-05,
+      "loss": 1.1163,
+      "step": 14385
+    },
+    {
+      "epoch": 2.30176,
+      "grad_norm": 1.747918487524492,
+      "learning_rate": 2.732962374565794e-05,
+      "loss": 1.1044,
+      "step": 14386
+    },
+    {
+      "epoch": 2.30192,
+      "grad_norm": 1.7734609165935338,
+      "learning_rate": 2.7317758551051985e-05,
+      "loss": 1.0844,
+      "step": 14387
+    },
+    {
+      "epoch": 2.30208,
+      "grad_norm": 1.87575360038038,
+      "learning_rate": 2.7305895525178217e-05,
+      "loss": 1.103,
+      "step": 14388
+    },
+    {
+      "epoch": 2.30224,
+      "grad_norm": 1.6344621983300154,
+      "learning_rate": 2.7294034668390588e-05,
+      "loss": 1.1095,
+      "step": 14389
+    },
+    {
+      "epoch": 2.3024,
+      "grad_norm": 1.6536248661206656,
+      "learning_rate": 2.7282175981043035e-05,
+      "loss": 1.1247,
+      "step": 14390
+    },
+    {
+      "epoch": 2.30256,
+      "grad_norm": 2.017991354829785,
+      "learning_rate": 2.727031946348938e-05,
+      "loss": 1.0994,
+      "step": 14391
+    },
+    {
+      "epoch": 2.30272,
+      "grad_norm": 1.66182190990233,
+      "learning_rate": 2.725846511608341e-05,
+      "loss": 1.1157,
+      "step": 14392
+    },
+    {
+      "epoch": 2.30288,
+      "grad_norm": 1.6977433568636586,
+      "learning_rate": 2.724661293917885e-05,
+      "loss": 1.1269,
+      "step": 14393
+    },
+    {
+      "epoch": 2.30304,
+      "grad_norm": 1.7406680024386298,
+      "learning_rate": 2.7234762933129342e-05,
+      "loss": 1.1142,
+      "step": 14394
+    },
+    {
+      "epoch": 2.3032,
+      "grad_norm": 1.969356427701291,
+      "learning_rate": 2.7222915098288495e-05,
+      "loss": 1.1275,
+      "step": 14395
+    },
+    {
+      "epoch": 2.30336,
+      "grad_norm": 1.6399542764590445,
+      "learning_rate": 2.721106943500976e-05,
+      "loss": 1.1271,
+      "step": 14396
+    },
+    {
+      "epoch": 2.30352,
+      "grad_norm": 1.6362200832613651,
+      "learning_rate": 2.719922594364672e-05,
+      "loss": 1.0798,
+      "step": 14397
+    },
+    {
+      "epoch": 2.30368,
+      "grad_norm": 2.259799647701043,
+      "learning_rate": 2.718738462455265e-05,
+      "loss": 1.1047,
+      "step": 14398
+    },
+    {
+      "epoch": 2.30384,
+      "grad_norm": 1.6163675835474856,
+      "learning_rate": 2.7175545478080965e-05,
+      "loss": 1.1241,
+      "step": 14399
+    },
+    {
+      "epoch": 2.304,
+      "grad_norm": 2.111789418252038,
+      "learning_rate": 2.716370850458483e-05,
+      "loss": 1.075,
+      "step": 14400
+    },
+    {
+      "epoch": 2.30416,
+      "grad_norm": 1.5247616430046405,
+      "learning_rate": 2.7151873704417528e-05,
+      "loss": 1.1226,
+      "step": 14401
+    },
+    {
+      "epoch": 2.30432,
+      "grad_norm": 1.7320416505879375,
+      "learning_rate": 2.7140041077932164e-05,
+      "loss": 1.1108,
+      "step": 14402
+    },
+    {
+      "epoch": 2.30448,
+      "grad_norm": 2.0472997920647638,
+      "learning_rate": 2.712821062548181e-05,
+      "loss": 1.0873,
+      "step": 14403
+    },
+    {
+      "epoch": 2.30464,
+      "grad_norm": 1.7288103421409184,
+      "learning_rate": 2.7116382347419468e-05,
+      "loss": 1.1095,
+      "step": 14404
+    },
+    {
+      "epoch": 2.3048,
+      "grad_norm": 1.817487047841519,
+      "learning_rate": 2.7104556244098068e-05,
+      "loss": 1.1446,
+      "step": 14405
+    },
+    {
+      "epoch": 2.30496,
+      "grad_norm": 1.729223174350231,
+      "learning_rate": 2.7092732315870495e-05,
+      "loss": 1.0984,
+      "step": 14406
+    },
+    {
+      "epoch": 2.30512,
+      "grad_norm": 1.887113934648954,
+      "learning_rate": 2.7080910563089544e-05,
+      "loss": 1.091,
+      "step": 14407
+    },
+    {
+      "epoch": 2.3052799999999998,
+      "grad_norm": 1.7577509848978705,
+      "learning_rate": 2.7069090986107982e-05,
+      "loss": 1.0985,
+      "step": 14408
+    },
+    {
+      "epoch": 2.30544,
+      "grad_norm": 1.690453220982819,
+      "learning_rate": 2.7057273585278464e-05,
+      "loss": 1.0635,
+      "step": 14409
+    },
+    {
+      "epoch": 2.3056,
+      "grad_norm": 2.176352282407821,
+      "learning_rate": 2.7045458360953613e-05,
+      "loss": 1.129,
+      "step": 14410
+    },
+    {
+      "epoch": 2.30576,
+      "grad_norm": 2.024287591107729,
+      "learning_rate": 2.7033645313485977e-05,
+      "loss": 1.1107,
+      "step": 14411
+    },
+    {
+      "epoch": 2.30592,
+      "grad_norm": 1.695770093388818,
+      "learning_rate": 2.702183444322801e-05,
+      "loss": 1.0635,
+      "step": 14412
+    },
+    {
+      "epoch": 2.30608,
+      "grad_norm": 1.5954002095372868,
+      "learning_rate": 2.701002575053223e-05,
+      "loss": 1.1504,
+      "step": 14413
+    },
+    {
+      "epoch": 2.30624,
+      "grad_norm": 2.0823322962166984,
+      "learning_rate": 2.6998219235750866e-05,
+      "loss": 1.0748,
+      "step": 14414
+    },
+    {
+      "epoch": 2.3064,
+      "grad_norm": 1.9419550190932708,
+      "learning_rate": 2.6986414899236324e-05,
+      "loss": 1.1437,
+      "step": 14415
+    },
+    {
+      "epoch": 2.30656,
+      "grad_norm": 1.8544065010971786,
+      "learning_rate": 2.6974612741340698e-05,
+      "loss": 1.117,
+      "step": 14416
+    },
+    {
+      "epoch": 2.30672,
+      "grad_norm": 1.765429186433176,
+      "learning_rate": 2.6962812762416255e-05,
+      "loss": 1.1011,
+      "step": 14417
+    },
+    {
+      "epoch": 2.30688,
+      "grad_norm": 1.9176224317107124,
+      "learning_rate": 2.6951014962815058e-05,
+      "loss": 1.0932,
+      "step": 14418
+    },
+    {
+      "epoch": 2.3070399999999998,
+      "grad_norm": 2.1831962506422693,
+      "learning_rate": 2.693921934288912e-05,
+      "loss": 1.1118,
+      "step": 14419
+    },
+    {
+      "epoch": 2.3072,
+      "grad_norm": 1.788329819737829,
+      "learning_rate": 2.692742590299043e-05,
+      "loss": 1.1327,
+      "step": 14420
+    },
+    {
+      "epoch": 2.30736,
+      "grad_norm": 1.6768301874227285,
+      "learning_rate": 2.6915634643470865e-05,
+      "loss": 1.1091,
+      "step": 14421
+    },
+    {
+      "epoch": 2.3075200000000002,
+      "grad_norm": 1.5855083035204736,
+      "learning_rate": 2.690384556468226e-05,
+      "loss": 1.0124,
+      "step": 14422
+    },
+    {
+      "epoch": 2.30768,
+      "grad_norm": 2.1078570775903205,
+      "learning_rate": 2.6892058666976404e-05,
+      "loss": 1.1507,
+      "step": 14423
+    },
+    {
+      "epoch": 2.30784,
+      "grad_norm": 2.1603512746741034,
+      "learning_rate": 2.6880273950704983e-05,
+      "loss": 1.0914,
+      "step": 14424
+    },
+    {
+      "epoch": 2.308,
+      "grad_norm": 1.7636640942004287,
+      "learning_rate": 2.6868491416219645e-05,
+      "loss": 1.1783,
+      "step": 14425
+    },
+    {
+      "epoch": 2.30816,
+      "grad_norm": 1.7830767942336425,
+      "learning_rate": 2.6856711063871953e-05,
+      "loss": 1.0806,
+      "step": 14426
+    },
+    {
+      "epoch": 2.30832,
+      "grad_norm": 1.809938254403475,
+      "learning_rate": 2.6844932894013418e-05,
+      "loss": 1.0923,
+      "step": 14427
+    },
+    {
+      "epoch": 2.30848,
+      "grad_norm": 1.7246974699717579,
+      "learning_rate": 2.6833156906995472e-05,
+      "loss": 1.0873,
+      "step": 14428
+    },
+    {
+      "epoch": 2.30864,
+      "grad_norm": 2.0129178380441837,
+      "learning_rate": 2.6821383103169552e-05,
+      "loss": 1.0608,
+      "step": 14429
+    },
+    {
+      "epoch": 2.3088,
+      "grad_norm": 2.0874951239094117,
+      "learning_rate": 2.6809611482886875e-05,
+      "loss": 1.126,
+      "step": 14430
+    },
+    {
+      "epoch": 2.30896,
+      "grad_norm": 2.0098467927591255,
+      "learning_rate": 2.67978420464988e-05,
+      "loss": 1.1239,
+      "step": 14431
+    },
+    {
+      "epoch": 2.30912,
+      "grad_norm": 1.6382343149718575,
+      "learning_rate": 2.67860747943564e-05,
+      "loss": 1.1519,
+      "step": 14432
+    },
+    {
+      "epoch": 2.30928,
+      "grad_norm": 2.1358379834383814,
+      "learning_rate": 2.6774309726810864e-05,
+      "loss": 1.0943,
+      "step": 14433
+    },
+    {
+      "epoch": 2.30944,
+      "grad_norm": 1.744813444014419,
+      "learning_rate": 2.676254684421323e-05,
+      "loss": 1.1272,
+      "step": 14434
+    },
+    {
+      "epoch": 2.3096,
+      "grad_norm": 2.254342115843279,
+      "learning_rate": 2.6750786146914476e-05,
+      "loss": 1.0626,
+      "step": 14435
+    },
+    {
+      "epoch": 2.30976,
+      "grad_norm": 1.7426372585961512,
+      "learning_rate": 2.6739027635265525e-05,
+      "loss": 1.1208,
+      "step": 14436
+    },
+    {
+      "epoch": 2.30992,
+      "grad_norm": 2.0897331823599754,
+      "learning_rate": 2.6727271309617242e-05,
+      "loss": 1.1217,
+      "step": 14437
+    },
+    {
+      "epoch": 2.31008,
+      "grad_norm": 2.6568398505637374,
+      "learning_rate": 2.6715517170320414e-05,
+      "loss": 1.1169,
+      "step": 14438
+    },
+    {
+      "epoch": 2.31024,
+      "grad_norm": 1.931136532291073,
+      "learning_rate": 2.6703765217725775e-05,
+      "loss": 1.0977,
+      "step": 14439
+    },
+    {
+      "epoch": 2.3104,
+      "grad_norm": 1.921452963861268,
+      "learning_rate": 2.669201545218396e-05,
+      "loss": 1.0688,
+      "step": 14440
+    },
+    {
+      "epoch": 2.31056,
+      "grad_norm": 1.9657774997170627,
+      "learning_rate": 2.6680267874045596e-05,
+      "loss": 1.0962,
+      "step": 14441
+    },
+    {
+      "epoch": 2.31072,
+      "grad_norm": 2.0780824364768833,
+      "learning_rate": 2.6668522483661197e-05,
+      "loss": 1.0747,
+      "step": 14442
+    },
+    {
+      "epoch": 2.31088,
+      "grad_norm": 2.0453673044155556,
+      "learning_rate": 2.6656779281381195e-05,
+      "loss": 1.1397,
+      "step": 14443
+    },
+    {
+      "epoch": 2.31104,
+      "grad_norm": 1.7407460930189178,
+      "learning_rate": 2.6645038267556065e-05,
+      "loss": 1.0904,
+      "step": 14444
+    },
+    {
+      "epoch": 2.3112,
+      "grad_norm": 1.7329912708442108,
+      "learning_rate": 2.663329944253611e-05,
+      "loss": 1.0763,
+      "step": 14445
+    },
+    {
+      "epoch": 2.31136,
+      "grad_norm": 2.086300125351116,
+      "learning_rate": 2.6621562806671575e-05,
+      "loss": 1.091,
+      "step": 14446
+    },
+    {
+      "epoch": 2.31152,
+      "grad_norm": 1.677627460432546,
+      "learning_rate": 2.6609828360312695e-05,
+      "loss": 1.1093,
+      "step": 14447
+    },
+    {
+      "epoch": 2.31168,
+      "grad_norm": 1.716100135245494,
+      "learning_rate": 2.6598096103809588e-05,
+      "loss": 1.0803,
+      "step": 14448
+    },
+    {
+      "epoch": 2.31184,
+      "grad_norm": 1.6754585883902744,
+      "learning_rate": 2.6586366037512345e-05,
+      "loss": 1.1394,
+      "step": 14449
+    },
+    {
+      "epoch": 2.312,
+      "grad_norm": 1.5867263923869461,
+      "learning_rate": 2.657463816177096e-05,
+      "loss": 1.1345,
+      "step": 14450
+    },
+    {
+      "epoch": 2.31216,
+      "grad_norm": 1.9393857242110595,
+      "learning_rate": 2.6562912476935374e-05,
+      "loss": 1.1126,
+      "step": 14451
+    },
+    {
+      "epoch": 2.31232,
+      "grad_norm": 1.8847636870450841,
+      "learning_rate": 2.6551188983355478e-05,
+      "loss": 1.1247,
+      "step": 14452
+    },
+    {
+      "epoch": 2.31248,
+      "grad_norm": 2.349375044363579,
+      "learning_rate": 2.6539467681381047e-05,
+      "loss": 1.1058,
+      "step": 14453
+    },
+    {
+      "epoch": 2.31264,
+      "grad_norm": 1.5521511226366729,
+      "learning_rate": 2.6527748571361922e-05,
+      "loss": 1.142,
+      "step": 14454
+    },
+    {
+      "epoch": 2.3128,
+      "grad_norm": 1.701361570431045,
+      "learning_rate": 2.6516031653647644e-05,
+      "loss": 1.1023,
+      "step": 14455
+    },
+    {
+      "epoch": 2.31296,
+      "grad_norm": 1.8228758318157696,
+      "learning_rate": 2.650431692858798e-05,
+      "loss": 1.0959,
+      "step": 14456
+    },
+    {
+      "epoch": 2.31312,
+      "grad_norm": 1.7027915576964578,
+      "learning_rate": 2.6492604396532327e-05,
+      "loss": 1.0926,
+      "step": 14457
+    },
+    {
+      "epoch": 2.31328,
+      "grad_norm": 1.7945864341928002,
+      "learning_rate": 2.6480894057830296e-05,
+      "loss": 1.067,
+      "step": 14458
+    },
+    {
+      "epoch": 2.31344,
+      "grad_norm": 1.8913524637730343,
+      "learning_rate": 2.646918591283124e-05,
+      "loss": 1.1059,
+      "step": 14459
+    },
+    {
+      "epoch": 2.3136,
+      "grad_norm": 2.0479836365059296,
+      "learning_rate": 2.6457479961884545e-05,
+      "loss": 1.1007,
+      "step": 14460
+    },
+    {
+      "epoch": 2.31376,
+      "grad_norm": 1.917721356210675,
+      "learning_rate": 2.644577620533948e-05,
+      "loss": 1.1514,
+      "step": 14461
+    },
+    {
+      "epoch": 2.31392,
+      "grad_norm": 1.9614314670567283,
+      "learning_rate": 2.643407464354527e-05,
+      "loss": 1.1098,
+      "step": 14462
+    },
+    {
+      "epoch": 2.31408,
+      "grad_norm": 1.9970704017758059,
+      "learning_rate": 2.6422375276851073e-05,
+      "loss": 1.1136,
+      "step": 14463
+    },
+    {
+      "epoch": 2.31424,
+      "grad_norm": 2.154600381553379,
+      "learning_rate": 2.6410678105605992e-05,
+      "loss": 1.0815,
+      "step": 14464
+    },
+    {
+      "epoch": 2.3144,
+      "grad_norm": 2.046285130728198,
+      "learning_rate": 2.639898313015905e-05,
+      "loss": 1.115,
+      "step": 14465
+    },
+    {
+      "epoch": 2.31456,
+      "grad_norm": 1.8065905540284002,
+      "learning_rate": 2.6387290350859193e-05,
+      "loss": 1.1265,
+      "step": 14466
+    },
+    {
+      "epoch": 2.31472,
+      "grad_norm": 1.6474448528501662,
+      "learning_rate": 2.6375599768055327e-05,
+      "loss": 1.1348,
+      "step": 14467
+    },
+    {
+      "epoch": 2.31488,
+      "grad_norm": 2.1175594248804255,
+      "learning_rate": 2.6363911382096285e-05,
+      "loss": 1.1016,
+      "step": 14468
+    },
+    {
+      "epoch": 2.3150399999999998,
+      "grad_norm": 1.7588547120263938,
+      "learning_rate": 2.6352225193330794e-05,
+      "loss": 1.0574,
+      "step": 14469
+    },
+    {
+      "epoch": 2.3152,
+      "grad_norm": 2.413733811114081,
+      "learning_rate": 2.634054120210765e-05,
+      "loss": 1.0975,
+      "step": 14470
+    },
+    {
+      "epoch": 2.31536,
+      "grad_norm": 1.6755669178243413,
+      "learning_rate": 2.6328859408775365e-05,
+      "loss": 1.0791,
+      "step": 14471
+    },
+    {
+      "epoch": 2.3155200000000002,
+      "grad_norm": 1.7046815520525016,
+      "learning_rate": 2.6317179813682623e-05,
+      "loss": 1.1153,
+      "step": 14472
+    },
+    {
+      "epoch": 2.31568,
+      "grad_norm": 2.509453499842488,
+      "learning_rate": 2.6305502417177817e-05,
+      "loss": 1.0963,
+      "step": 14473
+    },
+    {
+      "epoch": 2.31584,
+      "grad_norm": 1.883990298599894,
+      "learning_rate": 2.6293827219609456e-05,
+      "loss": 1.0985,
+      "step": 14474
+    },
+    {
+      "epoch": 2.316,
+      "grad_norm": 1.8409516137588358,
+      "learning_rate": 2.6282154221325882e-05,
+      "loss": 1.1399,
+      "step": 14475
+    },
+    {
+      "epoch": 2.31616,
+      "grad_norm": 1.6471531550057115,
+      "learning_rate": 2.627048342267542e-05,
+      "loss": 1.1059,
+      "step": 14476
+    },
+    {
+      "epoch": 2.31632,
+      "grad_norm": 2.024888935053627,
+      "learning_rate": 2.6258814824006294e-05,
+      "loss": 1.1036,
+      "step": 14477
+    },
+    {
+      "epoch": 2.31648,
+      "grad_norm": 1.8229205424780777,
+      "learning_rate": 2.624714842566668e-05,
+      "loss": 1.1181,
+      "step": 14478
+    },
+    {
+      "epoch": 2.31664,
+      "grad_norm": 1.8823731346954868,
+      "learning_rate": 2.623548422800469e-05,
+      "loss": 1.1643,
+      "step": 14479
+    },
+    {
+      "epoch": 2.3168,
+      "grad_norm": 1.8141998003796949,
+      "learning_rate": 2.6223822231368366e-05,
+      "loss": 1.0883,
+      "step": 14480
+    },
+    {
+      "epoch": 2.31696,
+      "grad_norm": 1.7893327939988384,
+      "learning_rate": 2.621216243610568e-05,
+      "loss": 1.1827,
+      "step": 14481
+    },
+    {
+      "epoch": 2.31712,
+      "grad_norm": 1.7779315798181274,
+      "learning_rate": 2.6200504842564545e-05,
+      "loss": 1.118,
+      "step": 14482
+    },
+    {
+      "epoch": 2.3172800000000002,
+      "grad_norm": 1.669337784244151,
+      "learning_rate": 2.6188849451092813e-05,
+      "loss": 1.1195,
+      "step": 14483
+    },
+    {
+      "epoch": 2.31744,
+      "grad_norm": 1.788683618606938,
+      "learning_rate": 2.617719626203826e-05,
+      "loss": 1.0925,
+      "step": 14484
+    },
+    {
+      "epoch": 2.3176,
+      "grad_norm": 2.1459784898697722,
+      "learning_rate": 2.616554527574856e-05,
+      "loss": 1.1036,
+      "step": 14485
+    },
+    {
+      "epoch": 2.31776,
+      "grad_norm": 1.8554083995980335,
+      "learning_rate": 2.615389649257147e-05,
+      "loss": 1.1089,
+      "step": 14486
+    },
+    {
+      "epoch": 2.31792,
+      "grad_norm": 1.8451468068219956,
+      "learning_rate": 2.6142249912854444e-05,
+      "loss": 1.0867,
+      "step": 14487
+    },
+    {
+      "epoch": 2.31808,
+      "grad_norm": 1.7610857349506646,
+      "learning_rate": 2.613060553694511e-05,
+      "loss": 1.0649,
+      "step": 14488
+    },
+    {
+      "epoch": 2.31824,
+      "grad_norm": 1.8027787346944804,
+      "learning_rate": 2.611896336519081e-05,
+      "loss": 1.1137,
+      "step": 14489
+    },
+    {
+      "epoch": 2.3184,
+      "grad_norm": 1.6391187207460522,
+      "learning_rate": 2.6107323397939022e-05,
+      "loss": 1.074,
+      "step": 14490
+    },
+    {
+      "epoch": 2.31856,
+      "grad_norm": 1.8871509636048929,
+      "learning_rate": 2.6095685635537027e-05,
+      "loss": 1.1053,
+      "step": 14491
+    },
+    {
+      "epoch": 2.31872,
+      "grad_norm": 1.7937010551198247,
+      "learning_rate": 2.608405007833209e-05,
+      "loss": 1.0531,
+      "step": 14492
+    },
+    {
+      "epoch": 2.31888,
+      "grad_norm": 1.8305565461302502,
+      "learning_rate": 2.6072416726671376e-05,
+      "loss": 1.1155,
+      "step": 14493
+    },
+    {
+      "epoch": 2.31904,
+      "grad_norm": 1.8638877964721066,
+      "learning_rate": 2.6060785580902038e-05,
+      "loss": 1.1204,
+      "step": 14494
+    },
+    {
+      "epoch": 2.3192,
+      "grad_norm": 1.9974139031882696,
+      "learning_rate": 2.604915664137111e-05,
+      "loss": 1.0814,
+      "step": 14495
+    },
+    {
+      "epoch": 2.31936,
+      "grad_norm": 1.8233270443180707,
+      "learning_rate": 2.6037529908425596e-05,
+      "loss": 1.089,
+      "step": 14496
+    },
+    {
+      "epoch": 2.31952,
+      "grad_norm": 2.0225321197635115,
+      "learning_rate": 2.602590538241241e-05,
+      "loss": 1.1068,
+      "step": 14497
+    },
+    {
+      "epoch": 2.31968,
+      "grad_norm": 1.6835760784438722,
+      "learning_rate": 2.6014283063678412e-05,
+      "loss": 1.087,
+      "step": 14498
+    },
+    {
+      "epoch": 2.31984,
+      "grad_norm": 2.0202688271551827,
+      "learning_rate": 2.600266295257041e-05,
+      "loss": 1.1471,
+      "step": 14499
+    },
+    {
+      "epoch": 2.32,
+      "grad_norm": 1.826352542228207,
+      "learning_rate": 2.5991045049435115e-05,
+      "loss": 1.085,
+      "step": 14500
+    },
+    {
+      "epoch": 2.32016,
+      "grad_norm": 1.7198938660721683,
+      "learning_rate": 2.5979429354619156e-05,
+      "loss": 1.0822,
+      "step": 14501
+    },
+    {
+      "epoch": 2.32032,
+      "grad_norm": 1.8299922384086733,
+      "learning_rate": 2.5967815868469203e-05,
+      "loss": 1.0789,
+      "step": 14502
+    },
+    {
+      "epoch": 2.32048,
+      "grad_norm": 1.6522275503888868,
+      "learning_rate": 2.5956204591331744e-05,
+      "loss": 1.0945,
+      "step": 14503
+    },
+    {
+      "epoch": 2.32064,
+      "grad_norm": 1.8420660894259795,
+      "learning_rate": 2.594459552355325e-05,
+      "loss": 1.1208,
+      "step": 14504
+    },
+    {
+      "epoch": 2.3208,
+      "grad_norm": 2.038654022417923,
+      "learning_rate": 2.5932988665480116e-05,
+      "loss": 1.1104,
+      "step": 14505
+    },
+    {
+      "epoch": 2.32096,
+      "grad_norm": 2.0163651555393733,
+      "learning_rate": 2.5921384017458672e-05,
+      "loss": 1.0967,
+      "step": 14506
+    },
+    {
+      "epoch": 2.32112,
+      "grad_norm": 1.832089054536778,
+      "learning_rate": 2.590978157983519e-05,
+      "loss": 1.131,
+      "step": 14507
+    },
+    {
+      "epoch": 2.32128,
+      "grad_norm": 2.010047981026224,
+      "learning_rate": 2.5898181352955865e-05,
+      "loss": 1.0779,
+      "step": 14508
+    },
+    {
+      "epoch": 2.32144,
+      "grad_norm": 1.639808146296029,
+      "learning_rate": 2.5886583337166815e-05,
+      "loss": 1.0918,
+      "step": 14509
+    },
+    {
+      "epoch": 2.3216,
+      "grad_norm": 1.7544996397312187,
+      "learning_rate": 2.587498753281411e-05,
+      "loss": 1.0739,
+      "step": 14510
+    },
+    {
+      "epoch": 2.32176,
+      "grad_norm": 1.850661148855394,
+      "learning_rate": 2.586339394024382e-05,
+      "loss": 1.1201,
+      "step": 14511
+    },
+    {
+      "epoch": 2.32192,
+      "grad_norm": 1.7205217916420383,
+      "learning_rate": 2.585180255980176e-05,
+      "loss": 1.1701,
+      "step": 14512
+    },
+    {
+      "epoch": 2.32208,
+      "grad_norm": 1.8080327121151334,
+      "learning_rate": 2.5840213391833934e-05,
+      "loss": 1.0958,
+      "step": 14513
+    },
+    {
+      "epoch": 2.32224,
+      "grad_norm": 1.8125787890283767,
+      "learning_rate": 2.582862643668602e-05,
+      "loss": 1.0887,
+      "step": 14514
+    },
+    {
+      "epoch": 2.3224,
+      "grad_norm": 1.936588770884828,
+      "learning_rate": 2.581704169470387e-05,
+      "loss": 1.1146,
+      "step": 14515
+    },
+    {
+      "epoch": 2.32256,
+      "grad_norm": 1.8990604667438589,
+      "learning_rate": 2.5805459166233038e-05,
+      "loss": 1.0807,
+      "step": 14516
+    },
+    {
+      "epoch": 2.32272,
+      "grad_norm": 1.7657424764181138,
+      "learning_rate": 2.5793878851619224e-05,
+      "loss": 1.0636,
+      "step": 14517
+    },
+    {
+      "epoch": 2.32288,
+      "grad_norm": 1.902163203744316,
+      "learning_rate": 2.5782300751207933e-05,
+      "loss": 1.0894,
+      "step": 14518
+    },
+    {
+      "epoch": 2.3230399999999998,
+      "grad_norm": 1.7381616806528934,
+      "learning_rate": 2.5770724865344644e-05,
+      "loss": 1.0643,
+      "step": 14519
+    },
+    {
+      "epoch": 2.3232,
+      "grad_norm": 1.7923151388956158,
+      "learning_rate": 2.5759151194374764e-05,
+      "loss": 1.1336,
+      "step": 14520
+    },
+    {
+      "epoch": 2.32336,
+      "grad_norm": 1.8926108776841877,
+      "learning_rate": 2.5747579738643623e-05,
+      "loss": 1.1079,
+      "step": 14521
+    },
+    {
+      "epoch": 2.32352,
+      "grad_norm": 1.62274088596611,
+      "learning_rate": 2.573601049849652e-05,
+      "loss": 1.0963,
+      "step": 14522
+    },
+    {
+      "epoch": 2.32368,
+      "grad_norm": 2.061198664156345,
+      "learning_rate": 2.5724443474278638e-05,
+      "loss": 1.1429,
+      "step": 14523
+    },
+    {
+      "epoch": 2.32384,
+      "grad_norm": 2.043143777774344,
+      "learning_rate": 2.571287866633514e-05,
+      "loss": 1.0651,
+      "step": 14524
+    },
+    {
+      "epoch": 2.324,
+      "grad_norm": 1.6985684701852517,
+      "learning_rate": 2.5701316075011084e-05,
+      "loss": 1.0661,
+      "step": 14525
+    },
+    {
+      "epoch": 2.32416,
+      "grad_norm": 1.7932772981663188,
+      "learning_rate": 2.5689755700651464e-05,
+      "loss": 1.1192,
+      "step": 14526
+    },
+    {
+      "epoch": 2.32432,
+      "grad_norm": 1.55259263870421,
+      "learning_rate": 2.5678197543601323e-05,
+      "loss": 1.0762,
+      "step": 14527
+    },
+    {
+      "epoch": 2.32448,
+      "grad_norm": 1.7259185901650809,
+      "learning_rate": 2.5666641604205398e-05,
+      "loss": 1.1045,
+      "step": 14528
+    },
+    {
+      "epoch": 2.32464,
+      "grad_norm": 2.029396663988026,
+      "learning_rate": 2.5655087882808636e-05,
+      "loss": 1.1413,
+      "step": 14529
+    },
+    {
+      "epoch": 2.3247999999999998,
+      "grad_norm": 1.627672029867602,
+      "learning_rate": 2.5643536379755663e-05,
+      "loss": 1.1384,
+      "step": 14530
+    },
+    {
+      "epoch": 2.32496,
+      "grad_norm": 1.8325445237097322,
+      "learning_rate": 2.5631987095391264e-05,
+      "loss": 1.1409,
+      "step": 14531
+    },
+    {
+      "epoch": 2.32512,
+      "grad_norm": 1.761741531117991,
+      "learning_rate": 2.5620440030059955e-05,
+      "loss": 1.0741,
+      "step": 14532
+    },
+    {
+      "epoch": 2.3252800000000002,
+      "grad_norm": 1.662479907616567,
+      "learning_rate": 2.5608895184106363e-05,
+      "loss": 1.1196,
+      "step": 14533
+    },
+    {
+      "epoch": 2.32544,
+      "grad_norm": 1.6174221139663265,
+      "learning_rate": 2.5597352557874932e-05,
+      "loss": 1.1116,
+      "step": 14534
+    },
+    {
+      "epoch": 2.3256,
+      "grad_norm": 1.6562408485794478,
+      "learning_rate": 2.5585812151710097e-05,
+      "loss": 1.0783,
+      "step": 14535
+    },
+    {
+      "epoch": 2.32576,
+      "grad_norm": 1.761218430214821,
+      "learning_rate": 2.5574273965956186e-05,
+      "loss": 1.1299,
+      "step": 14536
+    },
+    {
+      "epoch": 2.32592,
+      "grad_norm": 1.713071799605913,
+      "learning_rate": 2.556273800095749e-05,
+      "loss": 1.1198,
+      "step": 14537
+    },
+    {
+      "epoch": 2.32608,
+      "grad_norm": 1.9443208092224793,
+      "learning_rate": 2.5551204257058224e-05,
+      "loss": 1.0983,
+      "step": 14538
+    },
+    {
+      "epoch": 2.32624,
+      "grad_norm": 1.5162329178377822,
+      "learning_rate": 2.5539672734602536e-05,
+      "loss": 1.1174,
+      "step": 14539
+    },
+    {
+      "epoch": 2.3264,
+      "grad_norm": 1.701192834500345,
+      "learning_rate": 2.5528143433934526e-05,
+      "loss": 1.0943,
+      "step": 14540
+    },
+    {
+      "epoch": 2.32656,
+      "grad_norm": 1.8520048846883994,
+      "learning_rate": 2.5516616355398192e-05,
+      "loss": 1.1012,
+      "step": 14541
+    },
+    {
+      "epoch": 2.32672,
+      "grad_norm": 2.0559550259580077,
+      "learning_rate": 2.5505091499337462e-05,
+      "loss": 1.0968,
+      "step": 14542
+    },
+    {
+      "epoch": 2.32688,
+      "grad_norm": 2.029775022782697,
+      "learning_rate": 2.5493568866096317e-05,
+      "loss": 1.085,
+      "step": 14543
+    },
+    {
+      "epoch": 2.32704,
+      "grad_norm": 1.7460926664568936,
+      "learning_rate": 2.5482048456018438e-05,
+      "loss": 1.0909,
+      "step": 14544
+    },
+    {
+      "epoch": 2.3272,
+      "grad_norm": 1.7513120096924968,
+      "learning_rate": 2.5470530269447725e-05,
+      "loss": 1.0452,
+      "step": 14545
+    },
+    {
+      "epoch": 2.32736,
+      "grad_norm": 1.8145846260379397,
+      "learning_rate": 2.5459014306727713e-05,
+      "loss": 1.11,
+      "step": 14546
+    },
+    {
+      "epoch": 2.32752,
+      "grad_norm": 1.7336806851858335,
+      "learning_rate": 2.5447500568202166e-05,
+      "loss": 1.106,
+      "step": 14547
+    },
+    {
+      "epoch": 2.32768,
+      "grad_norm": 1.8808895621235162,
+      "learning_rate": 2.5435989054214505e-05,
+      "loss": 1.1291,
+      "step": 14548
+    },
+    {
+      "epoch": 2.32784,
+      "grad_norm": 1.7796635512576013,
+      "learning_rate": 2.54244797651083e-05,
+      "loss": 1.1134,
+      "step": 14549
+    },
+    {
+      "epoch": 2.328,
+      "grad_norm": 1.8533015747319823,
+      "learning_rate": 2.5412972701226954e-05,
+      "loss": 1.135,
+      "step": 14550
+    },
+    {
+      "epoch": 2.32816,
+      "grad_norm": 2.1456800648400276,
+      "learning_rate": 2.540146786291383e-05,
+      "loss": 1.0905,
+      "step": 14551
+    },
+    {
+      "epoch": 2.32832,
+      "grad_norm": 1.920663080649052,
+      "learning_rate": 2.5389965250512193e-05,
+      "loss": 1.0777,
+      "step": 14552
+    },
+    {
+      "epoch": 2.32848,
+      "grad_norm": 1.6364402776998477,
+      "learning_rate": 2.537846486436528e-05,
+      "loss": 1.0966,
+      "step": 14553
+    },
+    {
+      "epoch": 2.32864,
+      "grad_norm": 1.9374233371500136,
+      "learning_rate": 2.536696670481623e-05,
+      "loss": 1.1424,
+      "step": 14554
+    },
+    {
+      "epoch": 2.3288,
+      "grad_norm": 2.0690718461231055,
+      "learning_rate": 2.5355470772208157e-05,
+      "loss": 1.1326,
+      "step": 14555
+    },
+    {
+      "epoch": 2.32896,
+      "grad_norm": 2.0308069793314494,
+      "learning_rate": 2.5343977066884063e-05,
+      "loss": 1.0895,
+      "step": 14556
+    },
+    {
+      "epoch": 2.32912,
+      "grad_norm": 2.2670393164528058,
+      "learning_rate": 2.5332485589186907e-05,
+      "loss": 1.1042,
+      "step": 14557
+    },
+    {
+      "epoch": 2.32928,
+      "grad_norm": 1.8507484942600267,
+      "learning_rate": 2.5320996339459546e-05,
+      "loss": 1.1566,
+      "step": 14558
+    },
+    {
+      "epoch": 2.32944,
+      "grad_norm": 1.8870782660582028,
+      "learning_rate": 2.530950931804489e-05,
+      "loss": 1.1195,
+      "step": 14559
+    },
+    {
+      "epoch": 2.3296,
+      "grad_norm": 1.734435515832063,
+      "learning_rate": 2.529802452528559e-05,
+      "loss": 1.0877,
+      "step": 14560
+    },
+    {
+      "epoch": 2.32976,
+      "grad_norm": 1.931879704688391,
+      "learning_rate": 2.5286541961524413e-05,
+      "loss": 1.1175,
+      "step": 14561
+    },
+    {
+      "epoch": 2.32992,
+      "grad_norm": 1.8986339225639854,
+      "learning_rate": 2.5275061627103958e-05,
+      "loss": 1.087,
+      "step": 14562
+    },
+    {
+      "epoch": 2.33008,
+      "grad_norm": 2.1337463981530815,
+      "learning_rate": 2.5263583522366774e-05,
+      "loss": 1.0905,
+      "step": 14563
+    },
+    {
+      "epoch": 2.33024,
+      "grad_norm": 2.025808918219634,
+      "learning_rate": 2.5252107647655355e-05,
+      "loss": 1.1103,
+      "step": 14564
+    },
+    {
+      "epoch": 2.3304,
+      "grad_norm": 1.888807097480549,
+      "learning_rate": 2.5240634003312126e-05,
+      "loss": 1.0615,
+      "step": 14565
+    },
+    {
+      "epoch": 2.33056,
+      "grad_norm": 1.9386544092342595,
+      "learning_rate": 2.5229162589679433e-05,
+      "loss": 1.1296,
+      "step": 14566
+    },
+    {
+      "epoch": 2.33072,
+      "grad_norm": 2.109901514307117,
+      "learning_rate": 2.5217693407099564e-05,
+      "loss": 1.0808,
+      "step": 14567
+    },
+    {
+      "epoch": 2.33088,
+      "grad_norm": 1.7648941507543878,
+      "learning_rate": 2.5206226455914807e-05,
+      "loss": 1.0516,
+      "step": 14568
+    },
+    {
+      "epoch": 2.33104,
+      "grad_norm": 1.713339262434674,
+      "learning_rate": 2.5194761736467197e-05,
+      "loss": 1.0416,
+      "step": 14569
+    },
+    {
+      "epoch": 2.3312,
+      "grad_norm": 1.8047725333195697,
+      "learning_rate": 2.518329924909897e-05,
+      "loss": 1.0893,
+      "step": 14570
+    },
+    {
+      "epoch": 2.33136,
+      "grad_norm": 1.545227054854246,
+      "learning_rate": 2.5171838994151998e-05,
+      "loss": 1.0736,
+      "step": 14571
+    },
+    {
+      "epoch": 2.33152,
+      "grad_norm": 1.5076363938150097,
+      "learning_rate": 2.5160380971968378e-05,
+      "loss": 1.118,
+      "step": 14572
+    },
+    {
+      "epoch": 2.33168,
+      "grad_norm": 1.7901544632496427,
+      "learning_rate": 2.514892518288988e-05,
+      "loss": 1.1557,
+      "step": 14573
+    },
+    {
+      "epoch": 2.33184,
+      "grad_norm": 1.8391825395005144,
+      "learning_rate": 2.5137471627258423e-05,
+      "loss": 1.1031,
+      "step": 14574
+    },
+    {
+      "epoch": 2.332,
+      "grad_norm": 2.117991302067901,
+      "learning_rate": 2.512602030541572e-05,
+      "loss": 1.1418,
+      "step": 14575
+    },
+    {
+      "epoch": 2.33216,
+      "grad_norm": 1.9642817878265715,
+      "learning_rate": 2.5114571217703465e-05,
+      "loss": 1.0998,
+      "step": 14576
+    },
+    {
+      "epoch": 2.33232,
+      "grad_norm": 1.7866665534705066,
+      "learning_rate": 2.5103124364463294e-05,
+      "loss": 1.0659,
+      "step": 14577
+    },
+    {
+      "epoch": 2.33248,
+      "grad_norm": 1.7938747781703082,
+      "learning_rate": 2.5091679746036745e-05,
+      "loss": 1.0969,
+      "step": 14578
+    },
+    {
+      "epoch": 2.33264,
+      "grad_norm": 1.8499473316570392,
+      "learning_rate": 2.5080237362765325e-05,
+      "loss": 1.121,
+      "step": 14579
+    },
+    {
+      "epoch": 2.3327999999999998,
+      "grad_norm": 2.2709220567373034,
+      "learning_rate": 2.5068797214990448e-05,
+      "loss": 1.1042,
+      "step": 14580
+    },
+    {
+      "epoch": 2.33296,
+      "grad_norm": 1.994340434146735,
+      "learning_rate": 2.5057359303053484e-05,
+      "loss": 1.0904,
+      "step": 14581
+    },
+    {
+      "epoch": 2.33312,
+      "grad_norm": 1.7909988522629845,
+      "learning_rate": 2.5045923627295707e-05,
+      "loss": 1.1619,
+      "step": 14582
+    },
+    {
+      "epoch": 2.3332800000000002,
+      "grad_norm": 2.168974829117712,
+      "learning_rate": 2.5034490188058324e-05,
+      "loss": 1.1214,
+      "step": 14583
+    },
+    {
+      "epoch": 2.33344,
+      "grad_norm": 1.7889067559715575,
+      "learning_rate": 2.502305898568258e-05,
+      "loss": 1.1772,
+      "step": 14584
+    },
+    {
+      "epoch": 2.3336,
+      "grad_norm": 2.2346441910150343,
+      "learning_rate": 2.5011630020509434e-05,
+      "loss": 1.105,
+      "step": 14585
+    },
+    {
+      "epoch": 2.33376,
+      "grad_norm": 1.9882427527188835,
+      "learning_rate": 2.5000203292880042e-05,
+      "loss": 1.0711,
+      "step": 14586
+    },
+    {
+      "epoch": 2.33392,
+      "grad_norm": 1.9280856041056007,
+      "learning_rate": 2.4988778803135238e-05,
+      "loss": 1.1052,
+      "step": 14587
+    },
+    {
+      "epoch": 2.33408,
+      "grad_norm": 2.0334198007513327,
+      "learning_rate": 2.4977356551616036e-05,
+      "loss": 1.1246,
+      "step": 14588
+    },
+    {
+      "epoch": 2.33424,
+      "grad_norm": 1.7634670789266162,
+      "learning_rate": 2.4965936538663138e-05,
+      "loss": 1.0872,
+      "step": 14589
+    },
+    {
+      "epoch": 2.3344,
+      "grad_norm": 2.380995407512156,
+      "learning_rate": 2.4954518764617372e-05,
+      "loss": 1.0898,
+      "step": 14590
+    },
+    {
+      "epoch": 2.33456,
+      "grad_norm": 2.229642245732855,
+      "learning_rate": 2.4943103229819432e-05,
+      "loss": 1.1272,
+      "step": 14591
+    },
+    {
+      "epoch": 2.33472,
+      "grad_norm": 1.7468098619187469,
+      "learning_rate": 2.4931689934609903e-05,
+      "loss": 1.1033,
+      "step": 14592
+    },
+    {
+      "epoch": 2.33488,
+      "grad_norm": 1.9200103494094563,
+      "learning_rate": 2.4920278879329374e-05,
+      "loss": 1.1473,
+      "step": 14593
+    },
+    {
+      "epoch": 2.3350400000000002,
+      "grad_norm": 1.817906564455451,
+      "learning_rate": 2.490887006431831e-05,
+      "loss": 1.0529,
+      "step": 14594
+    },
+    {
+      "epoch": 2.3352,
+      "grad_norm": 2.2917831416014325,
+      "learning_rate": 2.489746348991716e-05,
+      "loss": 1.1065,
+      "step": 14595
+    },
+    {
+      "epoch": 2.33536,
+      "grad_norm": 2.1225422057413588,
+      "learning_rate": 2.4886059156466257e-05,
+      "loss": 1.1086,
+      "step": 14596
+    },
+    {
+      "epoch": 2.33552,
+      "grad_norm": 1.6340569374812133,
+      "learning_rate": 2.4874657064305884e-05,
+      "loss": 1.1116,
+      "step": 14597
+    },
+    {
+      "epoch": 2.33568,
+      "grad_norm": 1.6143097221095382,
+      "learning_rate": 2.486325721377629e-05,
+      "loss": 1.0398,
+      "step": 14598
+    },
+    {
+      "epoch": 2.33584,
+      "grad_norm": 1.9131017591208976,
+      "learning_rate": 2.4851859605217576e-05,
+      "loss": 1.0671,
+      "step": 14599
+    },
+    {
+      "epoch": 2.336,
+      "grad_norm": 1.7721050837353651,
+      "learning_rate": 2.4840464238969928e-05,
+      "loss": 1.0936,
+      "step": 14600
+    },
+    {
+      "epoch": 2.33616,
+      "grad_norm": 1.7275412904875018,
+      "learning_rate": 2.482907111537326e-05,
+      "loss": 1.0416,
+      "step": 14601
+    },
+    {
+      "epoch": 2.33632,
+      "grad_norm": 1.70121629127452,
+      "learning_rate": 2.4817680234767626e-05,
+      "loss": 1.1194,
+      "step": 14602
+    },
+    {
+      "epoch": 2.33648,
+      "grad_norm": 1.8727990393573566,
+      "learning_rate": 2.4806291597492805e-05,
+      "loss": 1.0961,
+      "step": 14603
+    },
+    {
+      "epoch": 2.33664,
+      "grad_norm": 1.6982016643926108,
+      "learning_rate": 2.479490520388873e-05,
+      "loss": 1.0884,
+      "step": 14604
+    },
+    {
+      "epoch": 2.3368,
+      "grad_norm": 1.8711458176969331,
+      "learning_rate": 2.4783521054295034e-05,
+      "loss": 1.119,
+      "step": 14605
+    },
+    {
+      "epoch": 2.33696,
+      "grad_norm": 2.2435773250109086,
+      "learning_rate": 2.477213914905151e-05,
+      "loss": 1.0814,
+      "step": 14606
+    },
+    {
+      "epoch": 2.33712,
+      "grad_norm": 1.9294484332146578,
+      "learning_rate": 2.476075948849772e-05,
+      "loss": 1.1343,
+      "step": 14607
+    },
+    {
+      "epoch": 2.33728,
+      "grad_norm": 2.143721441802775,
+      "learning_rate": 2.474938207297325e-05,
+      "loss": 1.1466,
+      "step": 14608
+    },
+    {
+      "epoch": 2.33744,
+      "grad_norm": 1.4869697403453286,
+      "learning_rate": 2.4738006902817547e-05,
+      "loss": 1.0994,
+      "step": 14609
+    },
+    {
+      "epoch": 2.3376,
+      "grad_norm": 1.8568099029083283,
+      "learning_rate": 2.4726633978370063e-05,
+      "loss": 1.1155,
+      "step": 14610
+    },
+    {
+      "epoch": 2.33776,
+      "grad_norm": 2.195604829175605,
+      "learning_rate": 2.4715263299970138e-05,
+      "loss": 1.1113,
+      "step": 14611
+    },
+    {
+      "epoch": 2.33792,
+      "grad_norm": 1.8796231788692772,
+      "learning_rate": 2.4703894867957046e-05,
+      "loss": 1.1149,
+      "step": 14612
+    },
+    {
+      "epoch": 2.33808,
+      "grad_norm": 1.6264602005218611,
+      "learning_rate": 2.4692528682670013e-05,
+      "loss": 1.1008,
+      "step": 14613
+    },
+    {
+      "epoch": 2.33824,
+      "grad_norm": 1.7339721262586318,
+      "learning_rate": 2.4681164744448192e-05,
+      "loss": 1.1015,
+      "step": 14614
+    },
+    {
+      "epoch": 2.3384,
+      "grad_norm": 1.9574464710432176,
+      "learning_rate": 2.4669803053630637e-05,
+      "loss": 1.092,
+      "step": 14615
+    },
+    {
+      "epoch": 2.33856,
+      "grad_norm": 2.065620383165926,
+      "learning_rate": 2.4658443610556448e-05,
+      "loss": 1.1219,
+      "step": 14616
+    },
+    {
+      "epoch": 2.33872,
+      "grad_norm": 2.469658338656351,
+      "learning_rate": 2.464708641556447e-05,
+      "loss": 1.0751,
+      "step": 14617
+    },
+    {
+      "epoch": 2.33888,
+      "grad_norm": 2.343419432915458,
+      "learning_rate": 2.463573146899365e-05,
+      "loss": 1.1476,
+      "step": 14618
+    },
+    {
+      "epoch": 2.33904,
+      "grad_norm": 1.9846321435457328,
+      "learning_rate": 2.4624378771182787e-05,
+      "loss": 1.1771,
+      "step": 14619
+    },
+    {
+      "epoch": 2.3392,
+      "grad_norm": 2.4894943398051255,
+      "learning_rate": 2.4613028322470623e-05,
+      "loss": 1.1566,
+      "step": 14620
+    },
+    {
+      "epoch": 2.33936,
+      "grad_norm": 1.864111494644854,
+      "learning_rate": 2.4601680123195847e-05,
+      "loss": 1.0873,
+      "step": 14621
+    },
+    {
+      "epoch": 2.33952,
+      "grad_norm": 1.7784180619137167,
+      "learning_rate": 2.4590334173697073e-05,
+      "loss": 1.1124,
+      "step": 14622
+    },
+    {
+      "epoch": 2.33968,
+      "grad_norm": 1.845786133134565,
+      "learning_rate": 2.4578990474312835e-05,
+      "loss": 1.089,
+      "step": 14623
+    },
+    {
+      "epoch": 2.33984,
+      "grad_norm": 2.057596397120892,
+      "learning_rate": 2.45676490253816e-05,
+      "loss": 1.0696,
+      "step": 14624
+    },
+    {
+      "epoch": 2.34,
+      "grad_norm": 1.6537276390524667,
+      "learning_rate": 2.4556309827241862e-05,
+      "loss": 1.0942,
+      "step": 14625
+    },
+    {
+      "epoch": 2.34016,
+      "grad_norm": 1.773750898860905,
+      "learning_rate": 2.454497288023184e-05,
+      "loss": 1.1734,
+      "step": 14626
+    },
+    {
+      "epoch": 2.34032,
+      "grad_norm": 1.7980138474049525,
+      "learning_rate": 2.4533638184689943e-05,
+      "loss": 1.1305,
+      "step": 14627
+    },
+    {
+      "epoch": 2.34048,
+      "grad_norm": 1.8113678006527396,
+      "learning_rate": 2.4522305740954255e-05,
+      "loss": 1.0972,
+      "step": 14628
+    },
+    {
+      "epoch": 2.34064,
+      "grad_norm": 1.627116134213522,
+      "learning_rate": 2.4510975549363036e-05,
+      "loss": 1.1005,
+      "step": 14629
+    },
+    {
+      "epoch": 2.3407999999999998,
+      "grad_norm": 1.761228659978622,
+      "learning_rate": 2.4499647610254252e-05,
+      "loss": 1.0837,
+      "step": 14630
+    },
+    {
+      "epoch": 2.34096,
+      "grad_norm": 1.5048496079274396,
+      "learning_rate": 2.4488321923965997e-05,
+      "loss": 1.0583,
+      "step": 14631
+    },
+    {
+      "epoch": 2.34112,
+      "grad_norm": 1.9495082233976904,
+      "learning_rate": 2.447699849083618e-05,
+      "loss": 1.0986,
+      "step": 14632
+    },
+    {
+      "epoch": 2.34128,
+      "grad_norm": 1.7078478432925892,
+      "learning_rate": 2.4465677311202685e-05,
+      "loss": 1.1219,
+      "step": 14633
+    },
+    {
+      "epoch": 2.34144,
+      "grad_norm": 1.774363329102543,
+      "learning_rate": 2.4454358385403307e-05,
+      "loss": 1.1325,
+      "step": 14634
+    },
+    {
+      "epoch": 2.3416,
+      "grad_norm": 1.677718410324741,
+      "learning_rate": 2.4443041713775805e-05,
+      "loss": 1.0748,
+      "step": 14635
+    },
+    {
+      "epoch": 2.34176,
+      "grad_norm": 1.7963825320575195,
+      "learning_rate": 2.4431727296657836e-05,
+      "loss": 1.0862,
+      "step": 14636
+    },
+    {
+      "epoch": 2.34192,
+      "grad_norm": 1.7088319128182752,
+      "learning_rate": 2.4420415134386997e-05,
+      "loss": 1.0678,
+      "step": 14637
+    },
+    {
+      "epoch": 2.34208,
+      "grad_norm": 1.8259272997049483,
+      "learning_rate": 2.4409105227300854e-05,
+      "loss": 1.1122,
+      "step": 14638
+    },
+    {
+      "epoch": 2.34224,
+      "grad_norm": 1.7278300805188782,
+      "learning_rate": 2.4397797575736857e-05,
+      "loss": 1.0912,
+      "step": 14639
+    },
+    {
+      "epoch": 2.3424,
+      "grad_norm": 2.1834189567682816,
+      "learning_rate": 2.4386492180032393e-05,
+      "loss": 1.1468,
+      "step": 14640
+    },
+    {
+      "epoch": 2.3425599999999998,
+      "grad_norm": 1.8006750494279824,
+      "learning_rate": 2.437518904052487e-05,
+      "loss": 1.0977,
+      "step": 14641
+    },
+    {
+      "epoch": 2.34272,
+      "grad_norm": 1.942532117730374,
+      "learning_rate": 2.436388815755146e-05,
+      "loss": 1.1523,
+      "step": 14642
+    },
+    {
+      "epoch": 2.34288,
+      "grad_norm": 1.5492589506839753,
+      "learning_rate": 2.4352589531449465e-05,
+      "loss": 1.0645,
+      "step": 14643
+    },
+    {
+      "epoch": 2.3430400000000002,
+      "grad_norm": 2.004681953707832,
+      "learning_rate": 2.4341293162555924e-05,
+      "loss": 1.1371,
+      "step": 14644
+    },
+    {
+      "epoch": 2.3432,
+      "grad_norm": 1.7655727820026748,
+      "learning_rate": 2.4329999051207986e-05,
+      "loss": 1.0581,
+      "step": 14645
+    },
+    {
+      "epoch": 2.34336,
+      "grad_norm": 1.7667060486340753,
+      "learning_rate": 2.4318707197742573e-05,
+      "loss": 1.0786,
+      "step": 14646
+    },
+    {
+      "epoch": 2.34352,
+      "grad_norm": 1.7338954009899143,
+      "learning_rate": 2.4307417602496674e-05,
+      "loss": 1.0932,
+      "step": 14647
+    },
+    {
+      "epoch": 2.34368,
+      "grad_norm": 2.0822359904079812,
+      "learning_rate": 2.429613026580715e-05,
+      "loss": 1.0845,
+      "step": 14648
+    },
+    {
+      "epoch": 2.34384,
+      "grad_norm": 2.045028900735999,
+      "learning_rate": 2.4284845188010775e-05,
+      "loss": 1.0939,
+      "step": 14649
+    },
+    {
+      "epoch": 2.344,
+      "grad_norm": 1.748740997983318,
+      "learning_rate": 2.4273562369444303e-05,
+      "loss": 1.0816,
+      "step": 14650
+    },
+    {
+      "epoch": 2.34416,
+      "grad_norm": 1.6416094649275441,
+      "learning_rate": 2.4262281810444378e-05,
+      "loss": 1.0949,
+      "step": 14651
+    },
+    {
+      "epoch": 2.34432,
+      "grad_norm": 1.7963801778371353,
+      "learning_rate": 2.4251003511347604e-05,
+      "loss": 1.1121,
+      "step": 14652
+    },
+    {
+      "epoch": 2.34448,
+      "grad_norm": 1.6292192049939032,
+      "learning_rate": 2.4239727472490504e-05,
+      "loss": 1.0585,
+      "step": 14653
+    },
+    {
+      "epoch": 2.34464,
+      "grad_norm": 1.7594292557166071,
+      "learning_rate": 2.422845369420954e-05,
+      "loss": 1.0475,
+      "step": 14654
+    },
+    {
+      "epoch": 2.3448,
+      "grad_norm": 1.766774589967642,
+      "learning_rate": 2.4217182176841113e-05,
+      "loss": 1.0884,
+      "step": 14655
+    },
+    {
+      "epoch": 2.34496,
+      "grad_norm": 1.8671146161096657,
+      "learning_rate": 2.4205912920721508e-05,
+      "loss": 1.1207,
+      "step": 14656
+    },
+    {
+      "epoch": 2.34512,
+      "grad_norm": 1.670989581114923,
+      "learning_rate": 2.419464592618709e-05,
+      "loss": 1.0907,
+      "step": 14657
+    },
+    {
+      "epoch": 2.34528,
+      "grad_norm": 1.6935142189828745,
+      "learning_rate": 2.4183381193573907e-05,
+      "loss": 1.097,
+      "step": 14658
+    },
+    {
+      "epoch": 2.34544,
+      "grad_norm": 1.9278975969194327,
+      "learning_rate": 2.4172118723218205e-05,
+      "loss": 1.1327,
+      "step": 14659
+    },
+    {
+      "epoch": 2.3456,
+      "grad_norm": 1.7227437674030204,
+      "learning_rate": 2.4160858515455943e-05,
+      "loss": 1.095,
+      "step": 14660
+    },
+    {
+      "epoch": 2.34576,
+      "grad_norm": 1.5938087889043366,
+      "learning_rate": 2.41496005706232e-05,
+      "loss": 1.0604,
+      "step": 14661
+    },
+    {
+      "epoch": 2.34592,
+      "grad_norm": 1.5711199845134287,
+      "learning_rate": 2.4138344889055796e-05,
+      "loss": 1.0428,
+      "step": 14662
+    },
+    {
+      "epoch": 2.34608,
+      "grad_norm": 1.680749910230647,
+      "learning_rate": 2.412709147108967e-05,
+      "loss": 1.1031,
+      "step": 14663
+    },
+    {
+      "epoch": 2.34624,
+      "grad_norm": 1.715913595049828,
+      "learning_rate": 2.411584031706058e-05,
+      "loss": 1.0876,
+      "step": 14664
+    },
+    {
+      "epoch": 2.3464,
+      "grad_norm": 1.6429871172021864,
+      "learning_rate": 2.4104591427304234e-05,
+      "loss": 1.0957,
+      "step": 14665
+    },
+    {
+      "epoch": 2.34656,
+      "grad_norm": 1.731491407699875,
+      "learning_rate": 2.4093344802156294e-05,
+      "loss": 1.0736,
+      "step": 14666
+    },
+    {
+      "epoch": 2.34672,
+      "grad_norm": 1.75568555384445,
+      "learning_rate": 2.4082100441952326e-05,
+      "loss": 1.1158,
+      "step": 14667
+    },
+    {
+      "epoch": 2.34688,
+      "grad_norm": 1.7890058343463118,
+      "learning_rate": 2.407085834702787e-05,
+      "loss": 1.0938,
+      "step": 14668
+    },
+    {
+      "epoch": 2.34704,
+      "grad_norm": 1.774610630908486,
+      "learning_rate": 2.4059618517718354e-05,
+      "loss": 1.1017,
+      "step": 14669
+    },
+    {
+      "epoch": 2.3472,
+      "grad_norm": 1.9656692680091885,
+      "learning_rate": 2.404838095435916e-05,
+      "loss": 1.0722,
+      "step": 14670
+    },
+    {
+      "epoch": 2.34736,
+      "grad_norm": 1.717016410188602,
+      "learning_rate": 2.403714565728561e-05,
+      "loss": 1.0749,
+      "step": 14671
+    },
+    {
+      "epoch": 2.34752,
+      "grad_norm": 1.9959438850873379,
+      "learning_rate": 2.4025912626832915e-05,
+      "loss": 1.1241,
+      "step": 14672
+    },
+    {
+      "epoch": 2.34768,
+      "grad_norm": 1.701576167299052,
+      "learning_rate": 2.401468186333634e-05,
+      "loss": 1.0571,
+      "step": 14673
+    },
+    {
+      "epoch": 2.34784,
+      "grad_norm": 1.740806542248874,
+      "learning_rate": 2.400345336713088e-05,
+      "loss": 1.0455,
+      "step": 14674
+    },
+    {
+      "epoch": 2.348,
+      "grad_norm": 1.92915314570272,
+      "learning_rate": 2.399222713855168e-05,
+      "loss": 1.1079,
+      "step": 14675
+    },
+    {
+      "epoch": 2.34816,
+      "grad_norm": 1.7905495566395158,
+      "learning_rate": 2.3981003177933624e-05,
+      "loss": 1.1019,
+      "step": 14676
+    },
+    {
+      "epoch": 2.34832,
+      "grad_norm": 2.1619829839572566,
+      "learning_rate": 2.3969781485611676e-05,
+      "loss": 1.1955,
+      "step": 14677
+    },
+    {
+      "epoch": 2.34848,
+      "grad_norm": 1.806895706422675,
+      "learning_rate": 2.3958562061920674e-05,
+      "loss": 1.1047,
+      "step": 14678
+    },
+    {
+      "epoch": 2.34864,
+      "grad_norm": 1.6806507588880446,
+      "learning_rate": 2.3947344907195367e-05,
+      "loss": 1.116,
+      "step": 14679
+    },
+    {
+      "epoch": 2.3487999999999998,
+      "grad_norm": 1.7104676025919359,
+      "learning_rate": 2.3936130021770476e-05,
+      "loss": 1.0953,
+      "step": 14680
+    },
+    {
+      "epoch": 2.34896,
+      "grad_norm": 1.8755604548129594,
+      "learning_rate": 2.3924917405980596e-05,
+      "loss": 1.1281,
+      "step": 14681
+    },
+    {
+      "epoch": 2.34912,
+      "grad_norm": 1.6260004047245606,
+      "learning_rate": 2.3913707060160383e-05,
+      "loss": 1.1002,
+      "step": 14682
+    },
+    {
+      "epoch": 2.34928,
+      "grad_norm": 1.7754081348927748,
+      "learning_rate": 2.390249898464422e-05,
+      "loss": 1.1497,
+      "step": 14683
+    },
+    {
+      "epoch": 2.34944,
+      "grad_norm": 1.700301070557511,
+      "learning_rate": 2.389129317976667e-05,
+      "loss": 1.1143,
+      "step": 14684
+    },
+    {
+      "epoch": 2.3496,
+      "grad_norm": 1.690787040848225,
+      "learning_rate": 2.3880089645861957e-05,
+      "loss": 1.127,
+      "step": 14685
+    },
+    {
+      "epoch": 2.34976,
+      "grad_norm": 1.8628998324694264,
+      "learning_rate": 2.386888838326452e-05,
+      "loss": 1.1276,
+      "step": 14686
+    },
+    {
+      "epoch": 2.34992,
+      "grad_norm": 1.8642008683376112,
+      "learning_rate": 2.3857689392308456e-05,
+      "loss": 1.1257,
+      "step": 14687
+    },
+    {
+      "epoch": 2.35008,
+      "grad_norm": 1.5933165231457578,
+      "learning_rate": 2.3846492673328026e-05,
+      "loss": 1.0793,
+      "step": 14688
+    },
+    {
+      "epoch": 2.35024,
+      "grad_norm": 1.5478138791669431,
+      "learning_rate": 2.3835298226657275e-05,
+      "loss": 1.0312,
+      "step": 14689
+    },
+    {
+      "epoch": 2.3504,
+      "grad_norm": 1.8699933148532975,
+      "learning_rate": 2.3824106052630258e-05,
+      "loss": 1.1133,
+      "step": 14690
+    },
+    {
+      "epoch": 2.3505599999999998,
+      "grad_norm": 2.248631307501112,
+      "learning_rate": 2.38129161515809e-05,
+      "loss": 1.0599,
+      "step": 14691
+    },
+    {
+      "epoch": 2.35072,
+      "grad_norm": 2.042660439646421,
+      "learning_rate": 2.3801728523843114e-05,
+      "loss": 1.1064,
+      "step": 14692
+    },
+    {
+      "epoch": 2.35088,
+      "grad_norm": 1.7058185298308592,
+      "learning_rate": 2.379054316975072e-05,
+      "loss": 1.0865,
+      "step": 14693
+    },
+    {
+      "epoch": 2.3510400000000002,
+      "grad_norm": 1.8432100093769466,
+      "learning_rate": 2.377936008963747e-05,
+      "loss": 1.0944,
+      "step": 14694
+    },
+    {
+      "epoch": 2.3512,
+      "grad_norm": 1.7940606891785325,
+      "learning_rate": 2.3768179283837044e-05,
+      "loss": 1.1147,
+      "step": 14695
+    },
+    {
+      "epoch": 2.35136,
+      "grad_norm": 1.950206299236228,
+      "learning_rate": 2.375700075268307e-05,
+      "loss": 1.0915,
+      "step": 14696
+    },
+    {
+      "epoch": 2.35152,
+      "grad_norm": 1.844301855325269,
+      "learning_rate": 2.374582449650906e-05,
+      "loss": 1.1072,
+      "step": 14697
+    },
+    {
+      "epoch": 2.35168,
+      "grad_norm": 1.7010956772407761,
+      "learning_rate": 2.3734650515648594e-05,
+      "loss": 1.1486,
+      "step": 14698
+    },
+    {
+      "epoch": 2.35184,
+      "grad_norm": 1.881876122061582,
+      "learning_rate": 2.3723478810434975e-05,
+      "loss": 1.1199,
+      "step": 14699
+    },
+    {
+      "epoch": 2.352,
+      "grad_norm": 1.9622226868191595,
+      "learning_rate": 2.3712309381201646e-05,
+      "loss": 1.134,
+      "step": 14700
+    },
+    {
+      "epoch": 2.35216,
+      "grad_norm": 1.9257762409439698,
+      "learning_rate": 2.370114222828179e-05,
+      "loss": 1.112,
+      "step": 14701
+    },
+    {
+      "epoch": 2.35232,
+      "grad_norm": 1.8502819564616657,
+      "learning_rate": 2.368997735200874e-05,
+      "loss": 1.1092,
+      "step": 14702
+    },
+    {
+      "epoch": 2.35248,
+      "grad_norm": 1.8113487530150179,
+      "learning_rate": 2.3678814752715506e-05,
+      "loss": 1.054,
+      "step": 14703
+    },
+    {
+      "epoch": 2.35264,
+      "grad_norm": 1.8023315662049006,
+      "learning_rate": 2.3667654430735252e-05,
+      "loss": 1.09,
+      "step": 14704
+    },
+    {
+      "epoch": 2.3528000000000002,
+      "grad_norm": 1.857944170276904,
+      "learning_rate": 2.3656496386400962e-05,
+      "loss": 1.0845,
+      "step": 14705
+    },
+    {
+      "epoch": 2.35296,
+      "grad_norm": 2.064359125862925,
+      "learning_rate": 2.3645340620045575e-05,
+      "loss": 1.1337,
+      "step": 14706
+    },
+    {
+      "epoch": 2.35312,
+      "grad_norm": 1.8181945773031525,
+      "learning_rate": 2.3634187132001973e-05,
+      "loss": 1.0876,
+      "step": 14707
+    },
+    {
+      "epoch": 2.35328,
+      "grad_norm": 1.87376985125848,
+      "learning_rate": 2.3623035922602943e-05,
+      "loss": 1.1194,
+      "step": 14708
+    },
+    {
+      "epoch": 2.35344,
+      "grad_norm": 1.8173777361292995,
+      "learning_rate": 2.361188699218123e-05,
+      "loss": 1.1569,
+      "step": 14709
+    },
+    {
+      "epoch": 2.3536,
+      "grad_norm": 1.8994939073035377,
+      "learning_rate": 2.3600740341069494e-05,
+      "loss": 1.1019,
+      "step": 14710
+    },
+    {
+      "epoch": 2.35376,
+      "grad_norm": 2.2504256240566467,
+      "learning_rate": 2.358959596960034e-05,
+      "loss": 1.0952,
+      "step": 14711
+    },
+    {
+      "epoch": 2.35392,
+      "grad_norm": 1.808907487427109,
+      "learning_rate": 2.357845387810631e-05,
+      "loss": 1.0453,
+      "step": 14712
+    },
+    {
+      "epoch": 2.35408,
+      "grad_norm": 1.828466501873767,
+      "learning_rate": 2.3567314066919833e-05,
+      "loss": 1.0575,
+      "step": 14713
+    },
+    {
+      "epoch": 2.35424,
+      "grad_norm": 2.045353475253683,
+      "learning_rate": 2.3556176536373376e-05,
+      "loss": 1.1077,
+      "step": 14714
+    },
+    {
+      "epoch": 2.3544,
+      "grad_norm": 1.8604951448769322,
+      "learning_rate": 2.3545041286799164e-05,
+      "loss": 1.1294,
+      "step": 14715
+    },
+    {
+      "epoch": 2.35456,
+      "grad_norm": 1.7376874701162293,
+      "learning_rate": 2.3533908318529573e-05,
+      "loss": 1.0997,
+      "step": 14716
+    },
+    {
+      "epoch": 2.35472,
+      "grad_norm": 2.162881654131694,
+      "learning_rate": 2.3522777631896687e-05,
+      "loss": 1.117,
+      "step": 14717
+    },
+    {
+      "epoch": 2.35488,
+      "grad_norm": 2.0255292474687328,
+      "learning_rate": 2.3511649227232723e-05,
+      "loss": 1.1057,
+      "step": 14718
+    },
+    {
+      "epoch": 2.35504,
+      "grad_norm": 1.9000872075769757,
+      "learning_rate": 2.350052310486963e-05,
+      "loss": 1.1014,
+      "step": 14719
+    },
+    {
+      "epoch": 2.3552,
+      "grad_norm": 1.6595095675422797,
+      "learning_rate": 2.34893992651395e-05,
+      "loss": 1.086,
+      "step": 14720
+    },
+    {
+      "epoch": 2.35536,
+      "grad_norm": 1.7282298002746026,
+      "learning_rate": 2.3478277708374198e-05,
+      "loss": 1.0895,
+      "step": 14721
+    },
+    {
+      "epoch": 2.35552,
+      "grad_norm": 1.8375532838104687,
+      "learning_rate": 2.346715843490559e-05,
+      "loss": 1.0469,
+      "step": 14722
+    },
+    {
+      "epoch": 2.35568,
+      "grad_norm": 2.169590229054902,
+      "learning_rate": 2.3456041445065445e-05,
+      "loss": 1.1378,
+      "step": 14723
+    },
+    {
+      "epoch": 2.35584,
+      "grad_norm": 1.810878241317556,
+      "learning_rate": 2.344492673918549e-05,
+      "loss": 1.1199,
+      "step": 14724
+    },
+    {
+      "epoch": 2.356,
+      "grad_norm": 1.7539282935615834,
+      "learning_rate": 2.343381431759738e-05,
+      "loss": 1.1114,
+      "step": 14725
+    },
+    {
+      "epoch": 2.35616,
+      "grad_norm": 1.656162874704451,
+      "learning_rate": 2.3422704180632672e-05,
+      "loss": 1.1076,
+      "step": 14726
+    },
+    {
+      "epoch": 2.35632,
+      "grad_norm": 2.0245536228134027,
+      "learning_rate": 2.341159632862289e-05,
+      "loss": 1.0705,
+      "step": 14727
+    },
+    {
+      "epoch": 2.35648,
+      "grad_norm": 2.0896858744643168,
+      "learning_rate": 2.340049076189946e-05,
+      "loss": 1.0636,
+      "step": 14728
+    },
+    {
+      "epoch": 2.35664,
+      "grad_norm": 2.047682651340784,
+      "learning_rate": 2.3389387480793757e-05,
+      "loss": 1.1286,
+      "step": 14729
+    },
+    {
+      "epoch": 2.3568,
+      "grad_norm": 2.0117031794343014,
+      "learning_rate": 2.3378286485637146e-05,
+      "loss": 1.0673,
+      "step": 14730
+    },
+    {
+      "epoch": 2.35696,
+      "grad_norm": 1.752695309455964,
+      "learning_rate": 2.336718777676077e-05,
+      "loss": 1.1046,
+      "step": 14731
+    },
+    {
+      "epoch": 2.35712,
+      "grad_norm": 1.7040908746192915,
+      "learning_rate": 2.335609135449589e-05,
+      "loss": 1.074,
+      "step": 14732
+    },
+    {
+      "epoch": 2.35728,
+      "grad_norm": 2.059190135774805,
+      "learning_rate": 2.3344997219173505e-05,
+      "loss": 1.1164,
+      "step": 14733
+    },
+    {
+      "epoch": 2.35744,
+      "grad_norm": 1.8936686820940174,
+      "learning_rate": 2.3333905371124766e-05,
+      "loss": 1.0896,
+      "step": 14734
+    },
+    {
+      "epoch": 2.3576,
+      "grad_norm": 1.6739792049412225,
+      "learning_rate": 2.332281581068052e-05,
+      "loss": 1.1555,
+      "step": 14735
+    },
+    {
+      "epoch": 2.35776,
+      "grad_norm": 1.6140133646464312,
+      "learning_rate": 2.331172853817175e-05,
+      "loss": 1.1019,
+      "step": 14736
+    },
+    {
+      "epoch": 2.35792,
+      "grad_norm": 1.9393960311033236,
+      "learning_rate": 2.3300643553929247e-05,
+      "loss": 1.0456,
+      "step": 14737
+    },
+    {
+      "epoch": 2.35808,
+      "grad_norm": 2.234615729124211,
+      "learning_rate": 2.3289560858283786e-05,
+      "loss": 1.1101,
+      "step": 14738
+    },
+    {
+      "epoch": 2.35824,
+      "grad_norm": 1.9715763046515546,
+      "learning_rate": 2.3278480451566044e-05,
+      "loss": 1.113,
+      "step": 14739
+    },
+    {
+      "epoch": 2.3584,
+      "grad_norm": 1.7093337403125728,
+      "learning_rate": 2.3267402334106626e-05,
+      "loss": 1.1121,
+      "step": 14740
+    },
+    {
+      "epoch": 2.3585599999999998,
+      "grad_norm": 1.8174813711521198,
+      "learning_rate": 2.325632650623616e-05,
+      "loss": 1.1387,
+      "step": 14741
+    },
+    {
+      "epoch": 2.35872,
+      "grad_norm": 1.9669932451735135,
+      "learning_rate": 2.3245252968285037e-05,
+      "loss": 1.0888,
+      "step": 14742
+    },
+    {
+      "epoch": 2.35888,
+      "grad_norm": 1.9504187092553564,
+      "learning_rate": 2.323418172058377e-05,
+      "loss": 1.1001,
+      "step": 14743
+    },
+    {
+      "epoch": 2.3590400000000002,
+      "grad_norm": 1.6754413006240825,
+      "learning_rate": 2.3223112763462607e-05,
+      "loss": 1.0406,
+      "step": 14744
+    },
+    {
+      "epoch": 2.3592,
+      "grad_norm": 1.7387240570482947,
+      "learning_rate": 2.3212046097251915e-05,
+      "loss": 1.1239,
+      "step": 14745
+    },
+    {
+      "epoch": 2.35936,
+      "grad_norm": 1.884009083897425,
+      "learning_rate": 2.320098172228188e-05,
+      "loss": 1.0302,
+      "step": 14746
+    },
+    {
+      "epoch": 2.35952,
+      "grad_norm": 1.8718770344309237,
+      "learning_rate": 2.318991963888263e-05,
+      "loss": 1.1087,
+      "step": 14747
+    },
+    {
+      "epoch": 2.35968,
+      "grad_norm": 1.6386772614412277,
+      "learning_rate": 2.317885984738426e-05,
+      "loss": 1.0921,
+      "step": 14748
+    },
+    {
+      "epoch": 2.35984,
+      "grad_norm": 1.8050868133292695,
+      "learning_rate": 2.3167802348116773e-05,
+      "loss": 1.1296,
+      "step": 14749
+    },
+    {
+      "epoch": 2.36,
+      "grad_norm": 1.7691656765396748,
+      "learning_rate": 2.3156747141410107e-05,
+      "loss": 1.0895,
+      "step": 14750
+    },
+    {
+      "epoch": 2.36016,
+      "grad_norm": 1.6002660431798448,
+      "learning_rate": 2.314569422759414e-05,
+      "loss": 1.0694,
+      "step": 14751
+    },
+    {
+      "epoch": 2.3603199999999998,
+      "grad_norm": 1.7230492754713482,
+      "learning_rate": 2.313464360699865e-05,
+      "loss": 1.0914,
+      "step": 14752
+    },
+    {
+      "epoch": 2.36048,
+      "grad_norm": 1.951966828772433,
+      "learning_rate": 2.3123595279953415e-05,
+      "loss": 1.1306,
+      "step": 14753
+    },
+    {
+      "epoch": 2.36064,
+      "grad_norm": 1.7414680595387402,
+      "learning_rate": 2.3112549246788064e-05,
+      "loss": 1.0752,
+      "step": 14754
+    },
+    {
+      "epoch": 2.3608000000000002,
+      "grad_norm": 2.3157199477289514,
+      "learning_rate": 2.31015055078322e-05,
+      "loss": 1.094,
+      "step": 14755
+    },
+    {
+      "epoch": 2.36096,
+      "grad_norm": 1.7347599246062797,
+      "learning_rate": 2.309046406341534e-05,
+      "loss": 1.0522,
+      "step": 14756
+    },
+    {
+      "epoch": 2.36112,
+      "grad_norm": 1.7125857795838186,
+      "learning_rate": 2.3079424913867008e-05,
+      "loss": 1.1051,
+      "step": 14757
+    },
+    {
+      "epoch": 2.36128,
+      "grad_norm": 1.9329975734352958,
+      "learning_rate": 2.306838805951651e-05,
+      "loss": 1.1418,
+      "step": 14758
+    },
+    {
+      "epoch": 2.36144,
+      "grad_norm": 1.7246360006297425,
+      "learning_rate": 2.3057353500693256e-05,
+      "loss": 1.0537,
+      "step": 14759
+    },
+    {
+      "epoch": 2.3616,
+      "grad_norm": 2.000350165500644,
+      "learning_rate": 2.304632123772641e-05,
+      "loss": 1.1542,
+      "step": 14760
+    },
+    {
+      "epoch": 2.36176,
+      "grad_norm": 1.790951229904245,
+      "learning_rate": 2.3035291270945225e-05,
+      "loss": 1.0301,
+      "step": 14761
+    },
+    {
+      "epoch": 2.36192,
+      "grad_norm": 2.0629803148939727,
+      "learning_rate": 2.3024263600678796e-05,
+      "loss": 1.1068,
+      "step": 14762
+    },
+    {
+      "epoch": 2.36208,
+      "grad_norm": 1.8307428981248537,
+      "learning_rate": 2.3013238227256184e-05,
+      "loss": 1.0743,
+      "step": 14763
+    },
+    {
+      "epoch": 2.36224,
+      "grad_norm": 1.7087260822797299,
+      "learning_rate": 2.3002215151006356e-05,
+      "loss": 1.102,
+      "step": 14764
+    },
+    {
+      "epoch": 2.3624,
+      "grad_norm": 2.0998258439742084,
+      "learning_rate": 2.299119437225824e-05,
+      "loss": 1.1179,
+      "step": 14765
+    },
+    {
+      "epoch": 2.36256,
+      "grad_norm": 1.775580532765374,
+      "learning_rate": 2.298017589134066e-05,
+      "loss": 1.0798,
+      "step": 14766
+    },
+    {
+      "epoch": 2.36272,
+      "grad_norm": 1.7043210031773943,
+      "learning_rate": 2.296915970858241e-05,
+      "loss": 1.0444,
+      "step": 14767
+    },
+    {
+      "epoch": 2.36288,
+      "grad_norm": 1.6534780244613572,
+      "learning_rate": 2.2958145824312182e-05,
+      "loss": 1.0898,
+      "step": 14768
+    },
+    {
+      "epoch": 2.36304,
+      "grad_norm": 1.7901707898688133,
+      "learning_rate": 2.2947134238858625e-05,
+      "loss": 1.0762,
+      "step": 14769
+    },
+    {
+      "epoch": 2.3632,
+      "grad_norm": 1.8279713581937909,
+      "learning_rate": 2.29361249525503e-05,
+      "loss": 1.0848,
+      "step": 14770
+    },
+    {
+      "epoch": 2.36336,
+      "grad_norm": 1.869824782708163,
+      "learning_rate": 2.2925117965715716e-05,
+      "loss": 1.109,
+      "step": 14771
+    },
+    {
+      "epoch": 2.36352,
+      "grad_norm": 1.7455009818983869,
+      "learning_rate": 2.291411327868327e-05,
+      "loss": 1.105,
+      "step": 14772
+    },
+    {
+      "epoch": 2.36368,
+      "grad_norm": 1.772511128321362,
+      "learning_rate": 2.2903110891781408e-05,
+      "loss": 1.1182,
+      "step": 14773
+    },
+    {
+      "epoch": 2.36384,
+      "grad_norm": 1.8799444728808294,
+      "learning_rate": 2.2892110805338328e-05,
+      "loss": 1.1807,
+      "step": 14774
+    },
+    {
+      "epoch": 2.364,
+      "grad_norm": 1.7836470767148271,
+      "learning_rate": 2.2881113019682365e-05,
+      "loss": 1.0565,
+      "step": 14775
+    },
+    {
+      "epoch": 2.36416,
+      "grad_norm": 1.931246834814149,
+      "learning_rate": 2.2870117535141555e-05,
+      "loss": 1.1275,
+      "step": 14776
+    },
+    {
+      "epoch": 2.36432,
+      "grad_norm": 1.7870522673659572,
+      "learning_rate": 2.285912435204407e-05,
+      "loss": 1.1096,
+      "step": 14777
+    },
+    {
+      "epoch": 2.36448,
+      "grad_norm": 1.880596386151311,
+      "learning_rate": 2.2848133470717912e-05,
+      "loss": 1.1031,
+      "step": 14778
+    },
+    {
+      "epoch": 2.36464,
+      "grad_norm": 1.7515151462446437,
+      "learning_rate": 2.2837144891491026e-05,
+      "loss": 1.0827,
+      "step": 14779
+    },
+    {
+      "epoch": 2.3648,
+      "grad_norm": 1.7901972023228565,
+      "learning_rate": 2.2826158614691305e-05,
+      "loss": 1.1049,
+      "step": 14780
+    },
+    {
+      "epoch": 2.36496,
+      "grad_norm": 1.9849207514804206,
+      "learning_rate": 2.2815174640646553e-05,
+      "loss": 1.1199,
+      "step": 14781
+    },
+    {
+      "epoch": 2.36512,
+      "grad_norm": 2.0461724665032803,
+      "learning_rate": 2.2804192969684513e-05,
+      "loss": 1.084,
+      "step": 14782
+    },
+    {
+      "epoch": 2.36528,
+      "grad_norm": 1.7847599078265508,
+      "learning_rate": 2.279321360213288e-05,
+      "loss": 1.1286,
+      "step": 14783
+    },
+    {
+      "epoch": 2.36544,
+      "grad_norm": 1.915760954164298,
+      "learning_rate": 2.2782236538319245e-05,
+      "loss": 1.1077,
+      "step": 14784
+    },
+    {
+      "epoch": 2.3656,
+      "grad_norm": 1.7239670679783599,
+      "learning_rate": 2.277126177857115e-05,
+      "loss": 1.0933,
+      "step": 14785
+    },
+    {
+      "epoch": 2.36576,
+      "grad_norm": 1.9166557509069564,
+      "learning_rate": 2.2760289323216056e-05,
+      "loss": 1.1031,
+      "step": 14786
+    },
+    {
+      "epoch": 2.36592,
+      "grad_norm": 1.9011378825869525,
+      "learning_rate": 2.274931917258142e-05,
+      "loss": 1.0755,
+      "step": 14787
+    },
+    {
+      "epoch": 2.36608,
+      "grad_norm": 2.1780985559646346,
+      "learning_rate": 2.273835132699448e-05,
+      "loss": 1.0809,
+      "step": 14788
+    },
+    {
+      "epoch": 2.36624,
+      "grad_norm": 1.7431934428176674,
+      "learning_rate": 2.2727385786782617e-05,
+      "loss": 1.0966,
+      "step": 14789
+    },
+    {
+      "epoch": 2.3664,
+      "grad_norm": 1.6992301258423679,
+      "learning_rate": 2.2716422552272897e-05,
+      "loss": 1.1179,
+      "step": 14790
+    },
+    {
+      "epoch": 2.3665599999999998,
+      "grad_norm": 1.8727896262879278,
+      "learning_rate": 2.2705461623792577e-05,
+      "loss": 1.0821,
+      "step": 14791
+    },
+    {
+      "epoch": 2.36672,
+      "grad_norm": 2.134037476393935,
+      "learning_rate": 2.26945030016686e-05,
+      "loss": 1.0893,
+      "step": 14792
+    },
+    {
+      "epoch": 2.36688,
+      "grad_norm": 1.9627549280820176,
+      "learning_rate": 2.268354668622803e-05,
+      "loss": 1.0818,
+      "step": 14793
+    },
+    {
+      "epoch": 2.36704,
+      "grad_norm": 1.7332011469787023,
+      "learning_rate": 2.2672592677797765e-05,
+      "loss": 1.0678,
+      "step": 14794
+    },
+    {
+      "epoch": 2.3672,
+      "grad_norm": 1.8569393932783655,
+      "learning_rate": 2.2661640976704658e-05,
+      "loss": 1.069,
+      "step": 14795
+    },
+    {
+      "epoch": 2.36736,
+      "grad_norm": 1.684798975227921,
+      "learning_rate": 2.2650691583275485e-05,
+      "loss": 1.0635,
+      "step": 14796
+    },
+    {
+      "epoch": 2.36752,
+      "grad_norm": 1.5919439515704275,
+      "learning_rate": 2.2639744497836947e-05,
+      "loss": 1.0533,
+      "step": 14797
+    },
+    {
+      "epoch": 2.36768,
+      "grad_norm": 1.8649259294753702,
+      "learning_rate": 2.2628799720715754e-05,
+      "loss": 1.1308,
+      "step": 14798
+    },
+    {
+      "epoch": 2.36784,
+      "grad_norm": 1.9093485413636389,
+      "learning_rate": 2.2617857252238383e-05,
+      "loss": 1.0263,
+      "step": 14799
+    },
+    {
+      "epoch": 2.368,
+      "grad_norm": 1.7276775378760174,
+      "learning_rate": 2.260691709273145e-05,
+      "loss": 1.0631,
+      "step": 14800
+    },
+    {
+      "epoch": 2.36816,
+      "grad_norm": 1.6797968969337582,
+      "learning_rate": 2.2595979242521282e-05,
+      "loss": 1.0868,
+      "step": 14801
+    },
+    {
+      "epoch": 2.3683199999999998,
+      "grad_norm": 1.6044122040814917,
+      "learning_rate": 2.258504370193434e-05,
+      "loss": 1.0838,
+      "step": 14802
+    },
+    {
+      "epoch": 2.36848,
+      "grad_norm": 1.6879823721296587,
+      "learning_rate": 2.2574110471296872e-05,
+      "loss": 1.0483,
+      "step": 14803
+    },
+    {
+      "epoch": 2.36864,
+      "grad_norm": 1.8858503277477614,
+      "learning_rate": 2.2563179550935142e-05,
+      "loss": 1.0835,
+      "step": 14804
+    },
+    {
+      "epoch": 2.3688000000000002,
+      "grad_norm": 2.334925242358678,
+      "learning_rate": 2.255225094117529e-05,
+      "loss": 1.0385,
+      "step": 14805
+    },
+    {
+      "epoch": 2.36896,
+      "grad_norm": 1.7895504229715673,
+      "learning_rate": 2.254132464234341e-05,
+      "loss": 1.077,
+      "step": 14806
+    },
+    {
+      "epoch": 2.36912,
+      "grad_norm": 1.8720717752501135,
+      "learning_rate": 2.253040065476554e-05,
+      "loss": 1.0718,
+      "step": 14807
+    },
+    {
+      "epoch": 2.36928,
+      "grad_norm": 1.8204891385761213,
+      "learning_rate": 2.2519478978767628e-05,
+      "loss": 1.0375,
+      "step": 14808
+    },
+    {
+      "epoch": 2.36944,
+      "grad_norm": 1.7202137013723788,
+      "learning_rate": 2.250855961467556e-05,
+      "loss": 1.1423,
+      "step": 14809
+    },
+    {
+      "epoch": 2.3696,
+      "grad_norm": 2.0600921238941554,
+      "learning_rate": 2.2497642562815156e-05,
+      "loss": 1.1382,
+      "step": 14810
+    },
+    {
+      "epoch": 2.36976,
+      "grad_norm": 2.012833467659333,
+      "learning_rate": 2.2486727823512177e-05,
+      "loss": 1.0449,
+      "step": 14811
+    },
+    {
+      "epoch": 2.36992,
+      "grad_norm": 1.5922871090000534,
+      "learning_rate": 2.247581539709228e-05,
+      "loss": 1.0812,
+      "step": 14812
+    },
+    {
+      "epoch": 2.37008,
+      "grad_norm": 1.7988835063395123,
+      "learning_rate": 2.2464905283881055e-05,
+      "loss": 1.0607,
+      "step": 14813
+    },
+    {
+      "epoch": 2.37024,
+      "grad_norm": 1.8673125132451942,
+      "learning_rate": 2.245399748420415e-05,
+      "loss": 1.1447,
+      "step": 14814
+    },
+    {
+      "epoch": 2.3704,
+      "grad_norm": 1.6949936840558955,
+      "learning_rate": 2.244309199838689e-05,
+      "loss": 1.0567,
+      "step": 14815
+    },
+    {
+      "epoch": 2.3705600000000002,
+      "grad_norm": 1.9094029695938837,
+      "learning_rate": 2.2432188826754808e-05,
+      "loss": 1.0627,
+      "step": 14816
+    },
+    {
+      "epoch": 2.37072,
+      "grad_norm": 1.814540553307605,
+      "learning_rate": 2.2421287969633132e-05,
+      "loss": 1.0641,
+      "step": 14817
+    },
+    {
+      "epoch": 2.37088,
+      "grad_norm": 1.8738608048849787,
+      "learning_rate": 2.2410389427347212e-05,
+      "loss": 1.1396,
+      "step": 14818
+    },
+    {
+      "epoch": 2.37104,
+      "grad_norm": 1.9122308120002027,
+      "learning_rate": 2.239949320022221e-05,
+      "loss": 1.0968,
+      "step": 14819
+    },
+    {
+      "epoch": 2.3712,
+      "grad_norm": 2.0238462042816954,
+      "learning_rate": 2.2388599288583246e-05,
+      "loss": 1.1097,
+      "step": 14820
+    },
+    {
+      "epoch": 2.37136,
+      "grad_norm": 1.9339659964109666,
+      "learning_rate": 2.23777076927554e-05,
+      "loss": 1.0627,
+      "step": 14821
+    },
+    {
+      "epoch": 2.37152,
+      "grad_norm": 2.1166019743294444,
+      "learning_rate": 2.236681841306364e-05,
+      "loss": 1.1244,
+      "step": 14822
+    },
+    {
+      "epoch": 2.37168,
+      "grad_norm": 1.65446299495542,
+      "learning_rate": 2.23559314498329e-05,
+      "loss": 1.0671,
+      "step": 14823
+    },
+    {
+      "epoch": 2.37184,
+      "grad_norm": 1.7488490138225985,
+      "learning_rate": 2.2345046803388026e-05,
+      "loss": 1.0693,
+      "step": 14824
+    },
+    {
+      "epoch": 2.372,
+      "grad_norm": 1.9684884331928472,
+      "learning_rate": 2.233416447405379e-05,
+      "loss": 1.1059,
+      "step": 14825
+    },
+    {
+      "epoch": 2.37216,
+      "grad_norm": 1.697655047748448,
+      "learning_rate": 2.2323284462154925e-05,
+      "loss": 1.0753,
+      "step": 14826
+    },
+    {
+      "epoch": 2.37232,
+      "grad_norm": 1.731514467592949,
+      "learning_rate": 2.231240676801606e-05,
+      "loss": 1.0936,
+      "step": 14827
+    },
+    {
+      "epoch": 2.37248,
+      "grad_norm": 1.7243009006066579,
+      "learning_rate": 2.2301531391961784e-05,
+      "loss": 1.0499,
+      "step": 14828
+    },
+    {
+      "epoch": 2.37264,
+      "grad_norm": 1.6245443912171704,
+      "learning_rate": 2.229065833431655e-05,
+      "loss": 1.0767,
+      "step": 14829
+    },
+    {
+      "epoch": 2.3728,
+      "grad_norm": 2.047790076227267,
+      "learning_rate": 2.22797875954049e-05,
+      "loss": 1.1525,
+      "step": 14830
+    },
+    {
+      "epoch": 2.37296,
+      "grad_norm": 1.992812215744522,
+      "learning_rate": 2.2268919175551083e-05,
+      "loss": 1.1519,
+      "step": 14831
+    },
+    {
+      "epoch": 2.37312,
+      "grad_norm": 1.5697502102192107,
+      "learning_rate": 2.225805307507951e-05,
+      "loss": 1.0604,
+      "step": 14832
+    },
+    {
+      "epoch": 2.37328,
+      "grad_norm": 1.9127619358321775,
+      "learning_rate": 2.2247189294314285e-05,
+      "loss": 1.1309,
+      "step": 14833
+    },
+    {
+      "epoch": 2.37344,
+      "grad_norm": 1.7475288834964764,
+      "learning_rate": 2.2236327833579672e-05,
+      "loss": 1.0825,
+      "step": 14834
+    },
+    {
+      "epoch": 2.3736,
+      "grad_norm": 1.7146018438093487,
+      "learning_rate": 2.2225468693199724e-05,
+      "loss": 1.0741,
+      "step": 14835
+    },
+    {
+      "epoch": 2.37376,
+      "grad_norm": 1.6858836085525177,
+      "learning_rate": 2.221461187349846e-05,
+      "loss": 1.0903,
+      "step": 14836
+    },
+    {
+      "epoch": 2.37392,
+      "grad_norm": 2.0590042819965584,
+      "learning_rate": 2.2203757374799826e-05,
+      "loss": 1.1269,
+      "step": 14837
+    },
+    {
+      "epoch": 2.37408,
+      "grad_norm": 1.8759240646022204,
+      "learning_rate": 2.2192905197427706e-05,
+      "loss": 1.1175,
+      "step": 14838
+    },
+    {
+      "epoch": 2.37424,
+      "grad_norm": 1.7769141583320254,
+      "learning_rate": 2.218205534170593e-05,
+      "loss": 1.0768,
+      "step": 14839
+    },
+    {
+      "epoch": 2.3744,
+      "grad_norm": 1.7393253499195425,
+      "learning_rate": 2.2171207807958228e-05,
+      "loss": 1.091,
+      "step": 14840
+    },
+    {
+      "epoch": 2.37456,
+      "grad_norm": 1.8229401053702132,
+      "learning_rate": 2.2160362596508267e-05,
+      "loss": 1.1346,
+      "step": 14841
+    },
+    {
+      "epoch": 2.37472,
+      "grad_norm": 1.8564010874510604,
+      "learning_rate": 2.2149519707679667e-05,
+      "loss": 1.1041,
+      "step": 14842
+    },
+    {
+      "epoch": 2.37488,
+      "grad_norm": 1.77852028783396,
+      "learning_rate": 2.213867914179597e-05,
+      "loss": 1.0419,
+      "step": 14843
+    },
+    {
+      "epoch": 2.37504,
+      "grad_norm": 1.735286199870491,
+      "learning_rate": 2.212784089918062e-05,
+      "loss": 1.1037,
+      "step": 14844
+    },
+    {
+      "epoch": 2.3752,
+      "grad_norm": 1.8468868935425529,
+      "learning_rate": 2.2117004980157e-05,
+      "loss": 1.0893,
+      "step": 14845
+    },
+    {
+      "epoch": 2.37536,
+      "grad_norm": 1.6283335620692507,
+      "learning_rate": 2.2106171385048512e-05,
+      "loss": 1.1,
+      "step": 14846
+    },
+    {
+      "epoch": 2.37552,
+      "grad_norm": 1.8366033947004186,
+      "learning_rate": 2.209534011417832e-05,
+      "loss": 1.1023,
+      "step": 14847
+    },
+    {
+      "epoch": 2.37568,
+      "grad_norm": 2.4319590764802905,
+      "learning_rate": 2.2084511167869714e-05,
+      "loss": 1.0796,
+      "step": 14848
+    },
+    {
+      "epoch": 2.37584,
+      "grad_norm": 1.587278800725822,
+      "learning_rate": 2.2073684546445706e-05,
+      "loss": 1.1076,
+      "step": 14849
+    },
+    {
+      "epoch": 2.376,
+      "grad_norm": 2.208548900540953,
+      "learning_rate": 2.2062860250229422e-05,
+      "loss": 1.0772,
+      "step": 14850
+    },
+    {
+      "epoch": 2.37616,
+      "grad_norm": 1.7776908080606804,
+      "learning_rate": 2.205203827954383e-05,
+      "loss": 1.0962,
+      "step": 14851
+    },
+    {
+      "epoch": 2.3763199999999998,
+      "grad_norm": 1.7736675090499223,
+      "learning_rate": 2.2041218634711834e-05,
+      "loss": 1.1269,
+      "step": 14852
+    },
+    {
+      "epoch": 2.37648,
+      "grad_norm": 1.8521143764833918,
+      "learning_rate": 2.203040131605627e-05,
+      "loss": 1.1161,
+      "step": 14853
+    },
+    {
+      "epoch": 2.37664,
+      "grad_norm": 1.8054682463368126,
+      "learning_rate": 2.2019586323899887e-05,
+      "loss": 1.0886,
+      "step": 14854
+    },
+    {
+      "epoch": 2.3768000000000002,
+      "grad_norm": 1.8108613259425654,
+      "learning_rate": 2.2008773658565485e-05,
+      "loss": 1.0869,
+      "step": 14855
+    },
+    {
+      "epoch": 2.37696,
+      "grad_norm": 1.6924386269141405,
+      "learning_rate": 2.199796332037556e-05,
+      "loss": 1.1007,
+      "step": 14856
+    },
+    {
+      "epoch": 2.37712,
+      "grad_norm": 1.625088380431296,
+      "learning_rate": 2.1987155309652818e-05,
+      "loss": 1.1354,
+      "step": 14857
+    },
+    {
+      "epoch": 2.37728,
+      "grad_norm": 1.8331188302196852,
+      "learning_rate": 2.197634962671963e-05,
+      "loss": 1.0661,
+      "step": 14858
+    },
+    {
+      "epoch": 2.37744,
+      "grad_norm": 1.707260797942631,
+      "learning_rate": 2.196554627189852e-05,
+      "loss": 1.1092,
+      "step": 14859
+    },
+    {
+      "epoch": 2.3776,
+      "grad_norm": 1.8614824487249764,
+      "learning_rate": 2.1954745245511752e-05,
+      "loss": 1.1144,
+      "step": 14860
+    },
+    {
+      "epoch": 2.37776,
+      "grad_norm": 1.7812385723575896,
+      "learning_rate": 2.1943946547881687e-05,
+      "loss": 1.0454,
+      "step": 14861
+    },
+    {
+      "epoch": 2.37792,
+      "grad_norm": 1.5867811817691893,
+      "learning_rate": 2.193315017933052e-05,
+      "loss": 1.1043,
+      "step": 14862
+    },
+    {
+      "epoch": 2.37808,
+      "grad_norm": 1.687792201882598,
+      "learning_rate": 2.19223561401804e-05,
+      "loss": 1.0843,
+      "step": 14863
+    },
+    {
+      "epoch": 2.37824,
+      "grad_norm": 1.8378446666789177,
+      "learning_rate": 2.1911564430753395e-05,
+      "loss": 1.1456,
+      "step": 14864
+    },
+    {
+      "epoch": 2.3784,
+      "grad_norm": 1.7220817185044313,
+      "learning_rate": 2.1900775051371525e-05,
+      "loss": 1.1093,
+      "step": 14865
+    },
+    {
+      "epoch": 2.3785600000000002,
+      "grad_norm": 1.7716020011077287,
+      "learning_rate": 2.188998800235672e-05,
+      "loss": 1.0995,
+      "step": 14866
+    },
+    {
+      "epoch": 2.37872,
+      "grad_norm": 1.7807242707698188,
+      "learning_rate": 2.187920328403087e-05,
+      "loss": 1.0701,
+      "step": 14867
+    },
+    {
+      "epoch": 2.37888,
+      "grad_norm": 1.7741095812715444,
+      "learning_rate": 2.1868420896715746e-05,
+      "loss": 1.1109,
+      "step": 14868
+    },
+    {
+      "epoch": 2.37904,
+      "grad_norm": 1.7573709457440199,
+      "learning_rate": 2.1857640840733097e-05,
+      "loss": 1.0991,
+      "step": 14869
+    },
+    {
+      "epoch": 2.3792,
+      "grad_norm": 1.9594134402868943,
+      "learning_rate": 2.1846863116404548e-05,
+      "loss": 1.1552,
+      "step": 14870
+    },
+    {
+      "epoch": 2.37936,
+      "grad_norm": 1.8104507685916313,
+      "learning_rate": 2.183608772405178e-05,
+      "loss": 1.08,
+      "step": 14871
+    },
+    {
+      "epoch": 2.37952,
+      "grad_norm": 1.9659169847891658,
+      "learning_rate": 2.18253146639962e-05,
+      "loss": 1.1355,
+      "step": 14872
+    },
+    {
+      "epoch": 2.37968,
+      "grad_norm": 1.8866952007273554,
+      "learning_rate": 2.1814543936559373e-05,
+      "loss": 1.1046,
+      "step": 14873
+    },
+    {
+      "epoch": 2.37984,
+      "grad_norm": 1.6220948009022946,
+      "learning_rate": 2.1803775542062578e-05,
+      "loss": 1.0751,
+      "step": 14874
+    },
+    {
+      "epoch": 2.38,
+      "grad_norm": 1.5491744850166753,
+      "learning_rate": 2.179300948082722e-05,
+      "loss": 1.0482,
+      "step": 14875
+    },
+    {
+      "epoch": 2.38016,
+      "grad_norm": 1.8814445700090072,
+      "learning_rate": 2.1782245753174456e-05,
+      "loss": 1.1233,
+      "step": 14876
+    },
+    {
+      "epoch": 2.38032,
+      "grad_norm": 1.842780789435564,
+      "learning_rate": 2.1771484359425532e-05,
+      "loss": 1.0981,
+      "step": 14877
+    },
+    {
+      "epoch": 2.38048,
+      "grad_norm": 1.7206245289077216,
+      "learning_rate": 2.176072529990152e-05,
+      "loss": 1.1102,
+      "step": 14878
+    },
+    {
+      "epoch": 2.38064,
+      "grad_norm": 1.7659189464513132,
+      "learning_rate": 2.1749968574923453e-05,
+      "loss": 1.0578,
+      "step": 14879
+    },
+    {
+      "epoch": 2.3808,
+      "grad_norm": 1.961220684625084,
+      "learning_rate": 2.1739214184812308e-05,
+      "loss": 1.0973,
+      "step": 14880
+    },
+    {
+      "epoch": 2.38096,
+      "grad_norm": 1.7233468858484187,
+      "learning_rate": 2.1728462129888972e-05,
+      "loss": 1.0997,
+      "step": 14881
+    },
+    {
+      "epoch": 2.38112,
+      "grad_norm": 1.8588628176935116,
+      "learning_rate": 2.1717712410474267e-05,
+      "loss": 1.0921,
+      "step": 14882
+    },
+    {
+      "epoch": 2.38128,
+      "grad_norm": 1.6951946253093515,
+      "learning_rate": 2.1706965026888958e-05,
+      "loss": 1.0933,
+      "step": 14883
+    },
+    {
+      "epoch": 2.38144,
+      "grad_norm": 1.8684297573969635,
+      "learning_rate": 2.1696219979453724e-05,
+      "loss": 1.1058,
+      "step": 14884
+    },
+    {
+      "epoch": 2.3816,
+      "grad_norm": 1.7225365011835732,
+      "learning_rate": 2.1685477268489195e-05,
+      "loss": 1.0527,
+      "step": 14885
+    },
+    {
+      "epoch": 2.38176,
+      "grad_norm": 1.9625430296759567,
+      "learning_rate": 2.167473689431587e-05,
+      "loss": 1.0976,
+      "step": 14886
+    },
+    {
+      "epoch": 2.38192,
+      "grad_norm": 1.9795106023825588,
+      "learning_rate": 2.166399885725433e-05,
+      "loss": 1.0597,
+      "step": 14887
+    },
+    {
+      "epoch": 2.38208,
+      "grad_norm": 1.746671531733859,
+      "learning_rate": 2.165326315762485e-05,
+      "loss": 1.0766,
+      "step": 14888
+    },
+    {
+      "epoch": 2.38224,
+      "grad_norm": 1.7539364215054734,
+      "learning_rate": 2.1642529795747902e-05,
+      "loss": 1.1081,
+      "step": 14889
+    },
+    {
+      "epoch": 2.3824,
+      "grad_norm": 1.7814958358931778,
+      "learning_rate": 2.1631798771943622e-05,
+      "loss": 1.0864,
+      "step": 14890
+    },
+    {
+      "epoch": 2.38256,
+      "grad_norm": 1.620999719904515,
+      "learning_rate": 2.1621070086532303e-05,
+      "loss": 1.0324,
+      "step": 14891
+    },
+    {
+      "epoch": 2.38272,
+      "grad_norm": 1.7990235665879573,
+      "learning_rate": 2.161034373983404e-05,
+      "loss": 1.0942,
+      "step": 14892
+    },
+    {
+      "epoch": 2.38288,
+      "grad_norm": 1.5958756863091113,
+      "learning_rate": 2.1599619732168897e-05,
+      "loss": 1.0644,
+      "step": 14893
+    },
+    {
+      "epoch": 2.38304,
+      "grad_norm": 1.5949148190280027,
+      "learning_rate": 2.1588898063856856e-05,
+      "loss": 1.1099,
+      "step": 14894
+    },
+    {
+      "epoch": 2.3832,
+      "grad_norm": 1.7184437157223078,
+      "learning_rate": 2.1578178735217835e-05,
+      "loss": 1.0969,
+      "step": 14895
+    },
+    {
+      "epoch": 2.38336,
+      "grad_norm": 1.7167829647774817,
+      "learning_rate": 2.1567461746571692e-05,
+      "loss": 1.0941,
+      "step": 14896
+    },
+    {
+      "epoch": 2.38352,
+      "grad_norm": 1.7690866733364805,
+      "learning_rate": 2.1556747098238205e-05,
+      "loss": 1.0904,
+      "step": 14897
+    },
+    {
+      "epoch": 2.38368,
+      "grad_norm": 1.945536466498522,
+      "learning_rate": 2.1546034790537083e-05,
+      "loss": 1.1096,
+      "step": 14898
+    },
+    {
+      "epoch": 2.38384,
+      "grad_norm": 1.836316822533666,
+      "learning_rate": 2.153532482378796e-05,
+      "loss": 1.1292,
+      "step": 14899
+    },
+    {
+      "epoch": 2.384,
+      "grad_norm": 1.4669137850996268,
+      "learning_rate": 2.1524617198310413e-05,
+      "loss": 1.0857,
+      "step": 14900
+    },
+    {
+      "epoch": 2.38416,
+      "grad_norm": 1.7477692063158423,
+      "learning_rate": 2.1513911914423933e-05,
+      "loss": 1.0781,
+      "step": 14901
+    },
+    {
+      "epoch": 2.3843199999999998,
+      "grad_norm": 1.7652568934103794,
+      "learning_rate": 2.150320897244793e-05,
+      "loss": 1.0427,
+      "step": 14902
+    },
+    {
+      "epoch": 2.38448,
+      "grad_norm": 1.7319991850249443,
+      "learning_rate": 2.149250837270185e-05,
+      "loss": 1.1011,
+      "step": 14903
+    },
+    {
+      "epoch": 2.38464,
+      "grad_norm": 2.0121005704377035,
+      "learning_rate": 2.1481810115504854e-05,
+      "loss": 1.1216,
+      "step": 14904
+    },
+    {
+      "epoch": 2.3848,
+      "grad_norm": 2.0231054558625465,
+      "learning_rate": 2.147111420117629e-05,
+      "loss": 1.1433,
+      "step": 14905
+    },
+    {
+      "epoch": 2.38496,
+      "grad_norm": 1.7164771731049935,
+      "learning_rate": 2.1460420630035205e-05,
+      "loss": 1.0479,
+      "step": 14906
+    },
+    {
+      "epoch": 2.38512,
+      "grad_norm": 1.6601590712958092,
+      "learning_rate": 2.1449729402400744e-05,
+      "loss": 1.1038,
+      "step": 14907
+    },
+    {
+      "epoch": 2.38528,
+      "grad_norm": 1.9344691225744695,
+      "learning_rate": 2.1439040518591892e-05,
+      "loss": 1.071,
+      "step": 14908
+    },
+    {
+      "epoch": 2.38544,
+      "grad_norm": 1.7737513995695249,
+      "learning_rate": 2.1428353978927606e-05,
+      "loss": 1.0699,
+      "step": 14909
+    },
+    {
+      "epoch": 2.3856,
+      "grad_norm": 1.6450421686119086,
+      "learning_rate": 2.1417669783726747e-05,
+      "loss": 1.0502,
+      "step": 14910
+    },
+    {
+      "epoch": 2.38576,
+      "grad_norm": 1.7836801781142362,
+      "learning_rate": 2.1406987933308076e-05,
+      "loss": 1.1488,
+      "step": 14911
+    },
+    {
+      "epoch": 2.38592,
+      "grad_norm": 1.94938508057917,
+      "learning_rate": 2.139630842799043e-05,
+      "loss": 1.0294,
+      "step": 14912
+    },
+    {
+      "epoch": 2.3860799999999998,
+      "grad_norm": 1.9369379648617746,
+      "learning_rate": 2.1385631268092344e-05,
+      "loss": 1.0518,
+      "step": 14913
+    },
+    {
+      "epoch": 2.38624,
+      "grad_norm": 1.7279113470790708,
+      "learning_rate": 2.1374956453932516e-05,
+      "loss": 1.1188,
+      "step": 14914
+    },
+    {
+      "epoch": 2.3864,
+      "grad_norm": 1.678442998648038,
+      "learning_rate": 2.136428398582937e-05,
+      "loss": 1.1296,
+      "step": 14915
+    },
+    {
+      "epoch": 2.3865600000000002,
+      "grad_norm": 1.94640576779645,
+      "learning_rate": 2.135361386410146e-05,
+      "loss": 1.0761,
+      "step": 14916
+    },
+    {
+      "epoch": 2.38672,
+      "grad_norm": 1.678739092908276,
+      "learning_rate": 2.134294608906705e-05,
+      "loss": 1.0908,
+      "step": 14917
+    },
+    {
+      "epoch": 2.38688,
+      "grad_norm": 1.666809570331356,
+      "learning_rate": 2.1332280661044557e-05,
+      "loss": 1.0714,
+      "step": 14918
+    },
+    {
+      "epoch": 2.38704,
+      "grad_norm": 1.7129375354261631,
+      "learning_rate": 2.132161758035216e-05,
+      "loss": 1.0977,
+      "step": 14919
+    },
+    {
+      "epoch": 2.3872,
+      "grad_norm": 1.5592896351715362,
+      "learning_rate": 2.131095684730806e-05,
+      "loss": 1.0994,
+      "step": 14920
+    },
+    {
+      "epoch": 2.38736,
+      "grad_norm": 1.987373495177525,
+      "learning_rate": 2.130029846223034e-05,
+      "loss": 1.0883,
+      "step": 14921
+    },
+    {
+      "epoch": 2.38752,
+      "grad_norm": 1.7669142812495204,
+      "learning_rate": 2.128964242543704e-05,
+      "loss": 1.0903,
+      "step": 14922
+    },
+    {
+      "epoch": 2.38768,
+      "grad_norm": 1.9600916751483564,
+      "learning_rate": 2.1278988737246118e-05,
+      "loss": 1.1494,
+      "step": 14923
+    },
+    {
+      "epoch": 2.38784,
+      "grad_norm": 1.613984467354421,
+      "learning_rate": 2.1268337397975458e-05,
+      "loss": 1.0642,
+      "step": 14924
+    },
+    {
+      "epoch": 2.388,
+      "grad_norm": 1.6811435981868417,
+      "learning_rate": 2.1257688407942887e-05,
+      "loss": 1.1034,
+      "step": 14925
+    },
+    {
+      "epoch": 2.38816,
+      "grad_norm": 1.9052473324954664,
+      "learning_rate": 2.124704176746616e-05,
+      "loss": 1.055,
+      "step": 14926
+    },
+    {
+      "epoch": 2.38832,
+      "grad_norm": 1.7482974498226123,
+      "learning_rate": 2.1236397476862914e-05,
+      "loss": 1.0736,
+      "step": 14927
+    },
+    {
+      "epoch": 2.38848,
+      "grad_norm": 1.7166857296335558,
+      "learning_rate": 2.1225755536450864e-05,
+      "loss": 1.0886,
+      "step": 14928
+    },
+    {
+      "epoch": 2.38864,
+      "grad_norm": 1.7088616988182495,
+      "learning_rate": 2.1215115946547413e-05,
+      "loss": 1.0931,
+      "step": 14929
+    },
+    {
+      "epoch": 2.3888,
+      "grad_norm": 1.671112815192732,
+      "learning_rate": 2.1204478707470164e-05,
+      "loss": 1.0972,
+      "step": 14930
+    },
+    {
+      "epoch": 2.38896,
+      "grad_norm": 1.8793447100031695,
+      "learning_rate": 2.119384381953641e-05,
+      "loss": 1.1299,
+      "step": 14931
+    },
+    {
+      "epoch": 2.38912,
+      "grad_norm": 1.7436197261341975,
+      "learning_rate": 2.1183211283063565e-05,
+      "loss": 1.0865,
+      "step": 14932
+    },
+    {
+      "epoch": 2.38928,
+      "grad_norm": 1.7108789199894363,
+      "learning_rate": 2.117258109836878e-05,
+      "loss": 1.1311,
+      "step": 14933
+    },
+    {
+      "epoch": 2.38944,
+      "grad_norm": 1.6882433040534344,
+      "learning_rate": 2.1161953265769364e-05,
+      "loss": 1.0975,
+      "step": 14934
+    },
+    {
+      "epoch": 2.3896,
+      "grad_norm": 1.7936960714560275,
+      "learning_rate": 2.1151327785582363e-05,
+      "loss": 1.0652,
+      "step": 14935
+    },
+    {
+      "epoch": 2.38976,
+      "grad_norm": 1.7357649607860355,
+      "learning_rate": 2.1140704658124854e-05,
+      "loss": 1.0554,
+      "step": 14936
+    },
+    {
+      "epoch": 2.38992,
+      "grad_norm": 1.9335400449687679,
+      "learning_rate": 2.1130083883713814e-05,
+      "loss": 1.1521,
+      "step": 14937
+    },
+    {
+      "epoch": 2.39008,
+      "grad_norm": 1.7912465197902514,
+      "learning_rate": 2.1119465462666144e-05,
+      "loss": 1.093,
+      "step": 14938
+    },
+    {
+      "epoch": 2.39024,
+      "grad_norm": 1.8984428970947982,
+      "learning_rate": 2.110884939529868e-05,
+      "loss": 1.099,
+      "step": 14939
+    },
+    {
+      "epoch": 2.3904,
+      "grad_norm": 1.9380768485145916,
+      "learning_rate": 2.109823568192819e-05,
+      "loss": 1.0762,
+      "step": 14940
+    },
+    {
+      "epoch": 2.39056,
+      "grad_norm": 1.5380292185475275,
+      "learning_rate": 2.108762432287138e-05,
+      "loss": 1.1008,
+      "step": 14941
+    },
+    {
+      "epoch": 2.39072,
+      "grad_norm": 1.6738718672565898,
+      "learning_rate": 2.1077015318444872e-05,
+      "loss": 1.0495,
+      "step": 14942
+    },
+    {
+      "epoch": 2.39088,
+      "grad_norm": 1.9397515339383908,
+      "learning_rate": 2.106640866896519e-05,
+      "loss": 1.1027,
+      "step": 14943
+    },
+    {
+      "epoch": 2.39104,
+      "grad_norm": 1.6734541333230397,
+      "learning_rate": 2.105580437474892e-05,
+      "loss": 1.0908,
+      "step": 14944
+    },
+    {
+      "epoch": 2.3912,
+      "grad_norm": 1.7700857920329642,
+      "learning_rate": 2.1045202436112353e-05,
+      "loss": 1.1031,
+      "step": 14945
+    },
+    {
+      "epoch": 2.39136,
+      "grad_norm": 1.7463290777981324,
+      "learning_rate": 2.1034602853371954e-05,
+      "loss": 1.0776,
+      "step": 14946
+    },
+    {
+      "epoch": 2.39152,
+      "grad_norm": 1.7595697143667257,
+      "learning_rate": 2.1024005626843877e-05,
+      "loss": 1.0857,
+      "step": 14947
+    },
+    {
+      "epoch": 2.39168,
+      "grad_norm": 2.01739816052209,
+      "learning_rate": 2.1013410756844455e-05,
+      "loss": 1.1141,
+      "step": 14948
+    },
+    {
+      "epoch": 2.39184,
+      "grad_norm": 2.087043027904906,
+      "learning_rate": 2.1002818243689703e-05,
+      "loss": 1.1006,
+      "step": 14949
+    },
+    {
+      "epoch": 2.392,
+      "grad_norm": 2.2940786432703772,
+      "learning_rate": 2.0992228087695776e-05,
+      "loss": 1.1155,
+      "step": 14950
+    },
+    {
+      "epoch": 2.39216,
+      "grad_norm": 2.0016666451465213,
+      "learning_rate": 2.098164028917863e-05,
+      "loss": 1.0844,
+      "step": 14951
+    },
+    {
+      "epoch": 2.39232,
+      "grad_norm": 2.2867550676387585,
+      "learning_rate": 2.0971054848454198e-05,
+      "loss": 1.0948,
+      "step": 14952
+    },
+    {
+      "epoch": 2.39248,
+      "grad_norm": 1.7656741227005879,
+      "learning_rate": 2.0960471765838342e-05,
+      "loss": 1.1151,
+      "step": 14953
+    },
+    {
+      "epoch": 2.39264,
+      "grad_norm": 1.8091437337786687,
+      "learning_rate": 2.0949891041646817e-05,
+      "loss": 1.0954,
+      "step": 14954
+    },
+    {
+      "epoch": 2.3928,
+      "grad_norm": 2.263773484078434,
+      "learning_rate": 2.0939312676195368e-05,
+      "loss": 1.0935,
+      "step": 14955
+    },
+    {
+      "epoch": 2.39296,
+      "grad_norm": 2.1263690730861065,
+      "learning_rate": 2.092873666979962e-05,
+      "loss": 1.0755,
+      "step": 14956
+    },
+    {
+      "epoch": 2.39312,
+      "grad_norm": 1.6100144938981809,
+      "learning_rate": 2.0918163022775152e-05,
+      "loss": 1.0886,
+      "step": 14957
+    },
+    {
+      "epoch": 2.39328,
+      "grad_norm": 1.672206804140631,
+      "learning_rate": 2.090759173543747e-05,
+      "loss": 1.0409,
+      "step": 14958
+    },
+    {
+      "epoch": 2.39344,
+      "grad_norm": 1.7632476869402316,
+      "learning_rate": 2.089702280810196e-05,
+      "loss": 1.131,
+      "step": 14959
+    },
+    {
+      "epoch": 2.3936,
+      "grad_norm": 1.9454816535374635,
+      "learning_rate": 2.0886456241084097e-05,
+      "loss": 1.0631,
+      "step": 14960
+    },
+    {
+      "epoch": 2.39376,
+      "grad_norm": 2.0541052318521964,
+      "learning_rate": 2.087589203469903e-05,
+      "loss": 1.0584,
+      "step": 14961
+    },
+    {
+      "epoch": 2.39392,
+      "grad_norm": 1.7831231864104786,
+      "learning_rate": 2.0865330189262113e-05,
+      "loss": 1.0767,
+      "step": 14962
+    },
+    {
+      "epoch": 2.3940799999999998,
+      "grad_norm": 1.8197309274212061,
+      "learning_rate": 2.0854770705088366e-05,
+      "loss": 1.0617,
+      "step": 14963
+    },
+    {
+      "epoch": 2.39424,
+      "grad_norm": 1.9600526658772197,
+      "learning_rate": 2.0844213582493e-05,
+      "loss": 1.0933,
+      "step": 14964
+    },
+    {
+      "epoch": 2.3944,
+      "grad_norm": 1.739471394036217,
+      "learning_rate": 2.0833658821790903e-05,
+      "loss": 1.1193,
+      "step": 14965
+    },
+    {
+      "epoch": 2.3945600000000002,
+      "grad_norm": 1.7449935258949498,
+      "learning_rate": 2.082310642329709e-05,
+      "loss": 1.0871,
+      "step": 14966
+    },
+    {
+      "epoch": 2.39472,
+      "grad_norm": 1.6814849544022223,
+      "learning_rate": 2.081255638732641e-05,
+      "loss": 1.113,
+      "step": 14967
+    },
+    {
+      "epoch": 2.39488,
+      "grad_norm": 2.1460543182422236,
+      "learning_rate": 2.0802008714193665e-05,
+      "loss": 1.1051,
+      "step": 14968
+    },
+    {
+      "epoch": 2.39504,
+      "grad_norm": 1.833565127081744,
+      "learning_rate": 2.0791463404213573e-05,
+      "loss": 1.0765,
+      "step": 14969
+    },
+    {
+      "epoch": 2.3952,
+      "grad_norm": 1.7536592865123517,
+      "learning_rate": 2.078092045770076e-05,
+      "loss": 1.0928,
+      "step": 14970
+    },
+    {
+      "epoch": 2.39536,
+      "grad_norm": 1.7020673511008848,
+      "learning_rate": 2.0770379874969903e-05,
+      "loss": 1.0945,
+      "step": 14971
+    },
+    {
+      "epoch": 2.39552,
+      "grad_norm": 1.5459810959942606,
+      "learning_rate": 2.075984165633541e-05,
+      "loss": 1.1124,
+      "step": 14972
+    },
+    {
+      "epoch": 2.39568,
+      "grad_norm": 2.002472218358637,
+      "learning_rate": 2.074930580211184e-05,
+      "loss": 1.0517,
+      "step": 14973
+    },
+    {
+      "epoch": 2.39584,
+      "grad_norm": 1.7808452078233294,
+      "learning_rate": 2.0738772312613442e-05,
+      "loss": 1.0998,
+      "step": 14974
+    },
+    {
+      "epoch": 2.396,
+      "grad_norm": 1.4688210240929347,
+      "learning_rate": 2.0728241188154608e-05,
+      "loss": 1.0929,
+      "step": 14975
+    },
+    {
+      "epoch": 2.39616,
+      "grad_norm": 1.7662706415386533,
+      "learning_rate": 2.071771242904954e-05,
+      "loss": 1.0451,
+      "step": 14976
+    },
+    {
+      "epoch": 2.3963200000000002,
+      "grad_norm": 1.723930889284857,
+      "learning_rate": 2.0707186035612415e-05,
+      "loss": 1.1108,
+      "step": 14977
+    },
+    {
+      "epoch": 2.39648,
+      "grad_norm": 1.7179370888952867,
+      "learning_rate": 2.0696662008157307e-05,
+      "loss": 1.112,
+      "step": 14978
+    },
+    {
+      "epoch": 2.39664,
+      "grad_norm": 1.9342571878415846,
+      "learning_rate": 2.0686140346998263e-05,
+      "loss": 1.1049,
+      "step": 14979
+    },
+    {
+      "epoch": 2.3968,
+      "grad_norm": 1.8026619397831831,
+      "learning_rate": 2.067562105244921e-05,
+      "loss": 1.1066,
+      "step": 14980
+    },
+    {
+      "epoch": 2.39696,
+      "grad_norm": 1.6104012960428118,
+      "learning_rate": 2.066510412482404e-05,
+      "loss": 1.0857,
+      "step": 14981
+    },
+    {
+      "epoch": 2.39712,
+      "grad_norm": 1.7426966180758798,
+      "learning_rate": 2.0654589564436556e-05,
+      "loss": 1.0629,
+      "step": 14982
+    },
+    {
+      "epoch": 2.39728,
+      "grad_norm": 1.6739856466088459,
+      "learning_rate": 2.06440773716005e-05,
+      "loss": 1.0749,
+      "step": 14983
+    },
+    {
+      "epoch": 2.39744,
+      "grad_norm": 1.7017580797333984,
+      "learning_rate": 2.063356754662952e-05,
+      "loss": 1.0966,
+      "step": 14984
+    },
+    {
+      "epoch": 2.3976,
+      "grad_norm": 1.8013506906783352,
+      "learning_rate": 2.0623060089837286e-05,
+      "loss": 1.1255,
+      "step": 14985
+    },
+    {
+      "epoch": 2.39776,
+      "grad_norm": 1.805972238723677,
+      "learning_rate": 2.0612555001537216e-05,
+      "loss": 1.1073,
+      "step": 14986
+    },
+    {
+      "epoch": 2.39792,
+      "grad_norm": 1.6347642745661313,
+      "learning_rate": 2.0602052282042882e-05,
+      "loss": 1.0742,
+      "step": 14987
+    },
+    {
+      "epoch": 2.39808,
+      "grad_norm": 1.9575117127597674,
+      "learning_rate": 2.059155193166756e-05,
+      "loss": 1.1223,
+      "step": 14988
+    },
+    {
+      "epoch": 2.39824,
+      "grad_norm": 1.8124511761734905,
+      "learning_rate": 2.058105395072467e-05,
+      "loss": 1.1132,
+      "step": 14989
+    },
+    {
+      "epoch": 2.3984,
+      "grad_norm": 1.846650249335859,
+      "learning_rate": 2.0570558339527357e-05,
+      "loss": 1.0876,
+      "step": 14990
+    },
+    {
+      "epoch": 2.39856,
+      "grad_norm": 1.7131683057145466,
+      "learning_rate": 2.056006509838887e-05,
+      "loss": 1.0415,
+      "step": 14991
+    },
+    {
+      "epoch": 2.39872,
+      "grad_norm": 1.7852012601140215,
+      "learning_rate": 2.0549574227622282e-05,
+      "loss": 1.1333,
+      "step": 14992
+    },
+    {
+      "epoch": 2.39888,
+      "grad_norm": 1.7689020570132026,
+      "learning_rate": 2.053908572754062e-05,
+      "loss": 1.0725,
+      "step": 14993
+    },
+    {
+      "epoch": 2.39904,
+      "grad_norm": 1.6012011468632168,
+      "learning_rate": 2.0528599598456854e-05,
+      "loss": 1.1122,
+      "step": 14994
+    },
+    {
+      "epoch": 2.3992,
+      "grad_norm": 1.8600040710038863,
+      "learning_rate": 2.0518115840683883e-05,
+      "loss": 1.0727,
+      "step": 14995
+    },
+    {
+      "epoch": 2.39936,
+      "grad_norm": 1.8957821698369004,
+      "learning_rate": 2.0507634454534508e-05,
+      "loss": 1.1201,
+      "step": 14996
+    },
+    {
+      "epoch": 2.39952,
+      "grad_norm": 1.876857782173582,
+      "learning_rate": 2.0497155440321493e-05,
+      "loss": 1.0914,
+      "step": 14997
+    },
+    {
+      "epoch": 2.39968,
+      "grad_norm": 1.8381623424740645,
+      "learning_rate": 2.0486678798357516e-05,
+      "loss": 1.0597,
+      "step": 14998
+    },
+    {
+      "epoch": 2.39984,
+      "grad_norm": 2.0198195502819116,
+      "learning_rate": 2.047620452895518e-05,
+      "loss": 1.101,
+      "step": 14999
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 1.79715032296043,
+      "learning_rate": 2.0465732632426993e-05,
+      "loss": 1.0962,
+      "step": 15000
+    },
+    {
+      "epoch": 2.40016,
+      "grad_norm": 1.940170875534107,
+      "learning_rate": 2.04552631090855e-05,
+      "loss": 1.097,
+      "step": 15001
+    },
+    {
+      "epoch": 2.40032,
+      "grad_norm": 1.6494578325539617,
+      "learning_rate": 2.044479595924299e-05,
+      "loss": 1.0406,
+      "step": 15002
+    },
+    {
+      "epoch": 2.40048,
+      "grad_norm": 1.8295019084330428,
+      "learning_rate": 2.043433118321191e-05,
+      "loss": 1.0793,
+      "step": 15003
+    },
+    {
+      "epoch": 2.40064,
+      "grad_norm": 2.041129744555703,
+      "learning_rate": 2.042386878130439e-05,
+      "loss": 1.0645,
+      "step": 15004
+    },
+    {
+      "epoch": 2.4008,
+      "grad_norm": 1.6325391032421481,
+      "learning_rate": 2.0413408753832718e-05,
+      "loss": 1.076,
+      "step": 15005
+    },
+    {
+      "epoch": 2.40096,
+      "grad_norm": 1.668369964532941,
+      "learning_rate": 2.0402951101108914e-05,
+      "loss": 1.1598,
+      "step": 15006
+    },
+    {
+      "epoch": 2.40112,
+      "grad_norm": 1.7780555337595043,
+      "learning_rate": 2.039249582344509e-05,
+      "loss": 1.0782,
+      "step": 15007
+    },
+    {
+      "epoch": 2.40128,
+      "grad_norm": 1.6553986501667417,
+      "learning_rate": 2.038204292115319e-05,
+      "loss": 1.1386,
+      "step": 15008
+    },
+    {
+      "epoch": 2.40144,
+      "grad_norm": 1.636076000566397,
+      "learning_rate": 2.0371592394545102e-05,
+      "loss": 1.116,
+      "step": 15009
+    },
+    {
+      "epoch": 2.4016,
+      "grad_norm": 1.5929814483210325,
+      "learning_rate": 2.0361144243932683e-05,
+      "loss": 1.1323,
+      "step": 15010
+    },
+    {
+      "epoch": 2.40176,
+      "grad_norm": 1.708307106634956,
+      "learning_rate": 2.035069846962766e-05,
+      "loss": 1.104,
+      "step": 15011
+    },
+    {
+      "epoch": 2.40192,
+      "grad_norm": 2.1235128970809503,
+      "learning_rate": 2.034025507194174e-05,
+      "loss": 1.1428,
+      "step": 15012
+    },
+    {
+      "epoch": 2.4020799999999998,
+      "grad_norm": 1.6900431903862874,
+      "learning_rate": 2.0329814051186534e-05,
+      "loss": 1.094,
+      "step": 15013
+    },
+    {
+      "epoch": 2.40224,
+      "grad_norm": 1.7019389003071286,
+      "learning_rate": 2.031937540767358e-05,
+      "loss": 1.1421,
+      "step": 15014
+    },
+    {
+      "epoch": 2.4024,
+      "grad_norm": 1.7441890447960087,
+      "learning_rate": 2.0308939141714366e-05,
+      "loss": 1.0917,
+      "step": 15015
+    },
+    {
+      "epoch": 2.40256,
+      "grad_norm": 1.6758460758571487,
+      "learning_rate": 2.0298505253620248e-05,
+      "loss": 1.0812,
+      "step": 15016
+    },
+    {
+      "epoch": 2.40272,
+      "grad_norm": 1.9105430722274954,
+      "learning_rate": 2.0288073743702652e-05,
+      "loss": 1.1235,
+      "step": 15017
+    },
+    {
+      "epoch": 2.40288,
+      "grad_norm": 1.663203249819538,
+      "learning_rate": 2.027764461227274e-05,
+      "loss": 1.093,
+      "step": 15018
+    },
+    {
+      "epoch": 2.40304,
+      "grad_norm": 1.74442503179755,
+      "learning_rate": 2.0267217859641786e-05,
+      "loss": 1.0836,
+      "step": 15019
+    },
+    {
+      "epoch": 2.4032,
+      "grad_norm": 1.7660915535424653,
+      "learning_rate": 2.0256793486120817e-05,
+      "loss": 1.0817,
+      "step": 15020
+    },
+    {
+      "epoch": 2.40336,
+      "grad_norm": 1.8133543565593426,
+      "learning_rate": 2.024637149202099e-05,
+      "loss": 1.0731,
+      "step": 15021
+    },
+    {
+      "epoch": 2.40352,
+      "grad_norm": 1.5288403245844295,
+      "learning_rate": 2.023595187765317e-05,
+      "loss": 1.0601,
+      "step": 15022
+    },
+    {
+      "epoch": 2.40368,
+      "grad_norm": 1.811451456914494,
+      "learning_rate": 2.0225534643328338e-05,
+      "loss": 1.0538,
+      "step": 15023
+    },
+    {
+      "epoch": 2.4038399999999998,
+      "grad_norm": 1.6881408877643151,
+      "learning_rate": 2.0215119789357316e-05,
+      "loss": 1.1287,
+      "step": 15024
+    },
+    {
+      "epoch": 2.404,
+      "grad_norm": 1.7840568291865972,
+      "learning_rate": 2.020470731605085e-05,
+      "loss": 1.0944,
+      "step": 15025
+    },
+    {
+      "epoch": 2.40416,
+      "grad_norm": 1.791696353203162,
+      "learning_rate": 2.019429722371966e-05,
+      "loss": 1.1367,
+      "step": 15026
+    },
+    {
+      "epoch": 2.4043200000000002,
+      "grad_norm": 2.0010119017498256,
+      "learning_rate": 2.0183889512674337e-05,
+      "loss": 1.1204,
+      "step": 15027
+    },
+    {
+      "epoch": 2.40448,
+      "grad_norm": 2.0397197385292323,
+      "learning_rate": 2.0173484183225456e-05,
+      "loss": 1.1536,
+      "step": 15028
+    },
+    {
+      "epoch": 2.40464,
+      "grad_norm": 2.000229815787172,
+      "learning_rate": 2.0163081235683456e-05,
+      "loss": 1.1217,
+      "step": 15029
+    },
+    {
+      "epoch": 2.4048,
+      "grad_norm": 1.599474046372426,
+      "learning_rate": 2.0152680670358835e-05,
+      "loss": 1.1166,
+      "step": 15030
+    },
+    {
+      "epoch": 2.40496,
+      "grad_norm": 1.7510672577932145,
+      "learning_rate": 2.0142282487561813e-05,
+      "loss": 1.0744,
+      "step": 15031
+    },
+    {
+      "epoch": 2.40512,
+      "grad_norm": 1.8120479764259112,
+      "learning_rate": 2.0131886687602754e-05,
+      "loss": 1.0857,
+      "step": 15032
+    },
+    {
+      "epoch": 2.40528,
+      "grad_norm": 1.8762041479452516,
+      "learning_rate": 2.0121493270791803e-05,
+      "loss": 1.1992,
+      "step": 15033
+    },
+    {
+      "epoch": 2.40544,
+      "grad_norm": 2.0977629650107263,
+      "learning_rate": 2.0111102237439107e-05,
+      "loss": 1.0738,
+      "step": 15034
+    },
+    {
+      "epoch": 2.4056,
+      "grad_norm": 1.5768587768918005,
+      "learning_rate": 2.0100713587854704e-05,
+      "loss": 1.1298,
+      "step": 15035
+    },
+    {
+      "epoch": 2.40576,
+      "grad_norm": 1.8094303536807954,
+      "learning_rate": 2.009032732234858e-05,
+      "loss": 1.0447,
+      "step": 15036
+    },
+    {
+      "epoch": 2.40592,
+      "grad_norm": 1.9361822852813009,
+      "learning_rate": 2.007994344123065e-05,
+      "loss": 1.0977,
+      "step": 15037
+    },
+    {
+      "epoch": 2.40608,
+      "grad_norm": 1.7070271092325149,
+      "learning_rate": 2.0069561944810756e-05,
+      "loss": 1.1279,
+      "step": 15038
+    },
+    {
+      "epoch": 2.40624,
+      "grad_norm": 1.8120415340939848,
+      "learning_rate": 2.005918283339866e-05,
+      "loss": 1.157,
+      "step": 15039
+    },
+    {
+      "epoch": 2.4064,
+      "grad_norm": 1.8834336722797003,
+      "learning_rate": 2.0048806107304065e-05,
+      "loss": 1.1091,
+      "step": 15040
+    },
+    {
+      "epoch": 2.40656,
+      "grad_norm": 1.7300431225695585,
+      "learning_rate": 2.003843176683656e-05,
+      "loss": 1.1495,
+      "step": 15041
+    },
+    {
+      "epoch": 2.40672,
+      "grad_norm": 1.5791457787960044,
+      "learning_rate": 2.00280598123058e-05,
+      "loss": 1.083,
+      "step": 15042
+    },
+    {
+      "epoch": 2.40688,
+      "grad_norm": 1.5703746595104005,
+      "learning_rate": 2.0017690244021148e-05,
+      "loss": 1.0882,
+      "step": 15043
+    },
+    {
+      "epoch": 2.40704,
+      "grad_norm": 1.7294914727826098,
+      "learning_rate": 2.0007323062292114e-05,
+      "loss": 1.187,
+      "step": 15044
+    },
+    {
+      "epoch": 2.4072,
+      "grad_norm": 1.96134629567343,
+      "learning_rate": 1.9996958267427956e-05,
+      "loss": 1.0592,
+      "step": 15045
+    },
+    {
+      "epoch": 2.40736,
+      "grad_norm": 1.8068878671215596,
+      "learning_rate": 1.9986595859738034e-05,
+      "loss": 1.0855,
+      "step": 15046
+    },
+    {
+      "epoch": 2.40752,
+      "grad_norm": 1.7807586653045577,
+      "learning_rate": 1.9976235839531453e-05,
+      "loss": 1.0524,
+      "step": 15047
+    },
+    {
+      "epoch": 2.40768,
+      "grad_norm": 1.7528133319182693,
+      "learning_rate": 1.9965878207117406e-05,
+      "loss": 1.0875,
+      "step": 15048
+    },
+    {
+      "epoch": 2.40784,
+      "grad_norm": 1.8323405571224167,
+      "learning_rate": 1.995552296280494e-05,
+      "loss": 1.153,
+      "step": 15049
+    },
+    {
+      "epoch": 2.408,
+      "grad_norm": 1.6914794020641528,
+      "learning_rate": 1.9945170106903034e-05,
+      "loss": 1.1051,
+      "step": 15050
+    },
+    {
+      "epoch": 2.40816,
+      "grad_norm": 1.6684057888323107,
+      "learning_rate": 1.9934819639720593e-05,
+      "loss": 1.0722,
+      "step": 15051
+    },
+    {
+      "epoch": 2.40832,
+      "grad_norm": 1.9347247020696139,
+      "learning_rate": 1.992447156156646e-05,
+      "loss": 1.1277,
+      "step": 15052
+    },
+    {
+      "epoch": 2.40848,
+      "grad_norm": 1.7359421730142908,
+      "learning_rate": 1.9914125872749425e-05,
+      "loss": 1.121,
+      "step": 15053
+    },
+    {
+      "epoch": 2.40864,
+      "grad_norm": 1.6908513857978338,
+      "learning_rate": 1.9903782573578167e-05,
+      "loss": 1.075,
+      "step": 15054
+    },
+    {
+      "epoch": 2.4088,
+      "grad_norm": 1.7512507820062369,
+      "learning_rate": 1.9893441664361334e-05,
+      "loss": 1.096,
+      "step": 15055
+    },
+    {
+      "epoch": 2.40896,
+      "grad_norm": 1.6803985588040933,
+      "learning_rate": 1.9883103145407467e-05,
+      "loss": 1.1126,
+      "step": 15056
+    },
+    {
+      "epoch": 2.40912,
+      "grad_norm": 1.9264090932349027,
+      "learning_rate": 1.987276701702503e-05,
+      "loss": 1.056,
+      "step": 15057
+    },
+    {
+      "epoch": 2.40928,
+      "grad_norm": 1.8330306565046233,
+      "learning_rate": 1.986243327952253e-05,
+      "loss": 1.0289,
+      "step": 15058
+    },
+    {
+      "epoch": 2.40944,
+      "grad_norm": 2.385417152246278,
+      "learning_rate": 1.985210193320819e-05,
+      "loss": 1.0798,
+      "step": 15059
+    },
+    {
+      "epoch": 2.4096,
+      "grad_norm": 1.8548556006023862,
+      "learning_rate": 1.9841772978390394e-05,
+      "loss": 1.0122,
+      "step": 15060
+    },
+    {
+      "epoch": 2.40976,
+      "grad_norm": 1.761158402153398,
+      "learning_rate": 1.9831446415377243e-05,
+      "loss": 1.0367,
+      "step": 15061
+    },
+    {
+      "epoch": 2.40992,
+      "grad_norm": 1.800933253287673,
+      "learning_rate": 1.9821122244476954e-05,
+      "loss": 1.0891,
+      "step": 15062
+    },
+    {
+      "epoch": 2.41008,
+      "grad_norm": 1.8349153492378905,
+      "learning_rate": 1.9810800465997494e-05,
+      "loss": 1.1092,
+      "step": 15063
+    },
+    {
+      "epoch": 2.41024,
+      "grad_norm": 2.024366284299209,
+      "learning_rate": 1.9800481080246935e-05,
+      "loss": 1.1801,
+      "step": 15064
+    },
+    {
+      "epoch": 2.4104,
+      "grad_norm": 2.1055672982925766,
+      "learning_rate": 1.9790164087533148e-05,
+      "loss": 1.1399,
+      "step": 15065
+    },
+    {
+      "epoch": 2.41056,
+      "grad_norm": 1.6990160348004286,
+      "learning_rate": 1.9779849488163994e-05,
+      "loss": 1.1139,
+      "step": 15066
+    },
+    {
+      "epoch": 2.41072,
+      "grad_norm": 1.74773389353024,
+      "learning_rate": 1.9769537282447238e-05,
+      "loss": 1.0592,
+      "step": 15067
+    },
+    {
+      "epoch": 2.41088,
+      "grad_norm": 2.0693656625926047,
+      "learning_rate": 1.9759227470690577e-05,
+      "loss": 1.1531,
+      "step": 15068
+    },
+    {
+      "epoch": 2.41104,
+      "grad_norm": 1.889252479347317,
+      "learning_rate": 1.9748920053201646e-05,
+      "loss": 1.114,
+      "step": 15069
+    },
+    {
+      "epoch": 2.4112,
+      "grad_norm": 1.7442645761969746,
+      "learning_rate": 1.9738615030288e-05,
+      "loss": 1.0416,
+      "step": 15070
+    },
+    {
+      "epoch": 2.41136,
+      "grad_norm": 1.7121657946729174,
+      "learning_rate": 1.972831240225713e-05,
+      "loss": 1.0779,
+      "step": 15071
+    },
+    {
+      "epoch": 2.41152,
+      "grad_norm": 1.7001394374654752,
+      "learning_rate": 1.9718012169416455e-05,
+      "loss": 1.097,
+      "step": 15072
+    },
+    {
+      "epoch": 2.41168,
+      "grad_norm": 1.9223078989941125,
+      "learning_rate": 1.970771433207328e-05,
+      "loss": 1.1301,
+      "step": 15073
+    },
+    {
+      "epoch": 2.4118399999999998,
+      "grad_norm": 2.1411771226837195,
+      "learning_rate": 1.969741889053497e-05,
+      "loss": 1.0949,
+      "step": 15074
+    },
+    {
+      "epoch": 2.412,
+      "grad_norm": 1.6947676466785766,
+      "learning_rate": 1.9687125845108612e-05,
+      "loss": 1.0604,
+      "step": 15075
+    },
+    {
+      "epoch": 2.41216,
+      "grad_norm": 2.013884149186384,
+      "learning_rate": 1.967683519610144e-05,
+      "loss": 1.1006,
+      "step": 15076
+    },
+    {
+      "epoch": 2.4123200000000002,
+      "grad_norm": 1.6394944655268746,
+      "learning_rate": 1.966654694382041e-05,
+      "loss": 1.0544,
+      "step": 15077
+    },
+    {
+      "epoch": 2.41248,
+      "grad_norm": 1.7858859310533268,
+      "learning_rate": 1.9656261088572613e-05,
+      "loss": 1.0974,
+      "step": 15078
+    },
+    {
+      "epoch": 2.41264,
+      "grad_norm": 1.6912673790874022,
+      "learning_rate": 1.9645977630664867e-05,
+      "loss": 1.0666,
+      "step": 15079
+    },
+    {
+      "epoch": 2.4128,
+      "grad_norm": 1.773012606077273,
+      "learning_rate": 1.9635696570404084e-05,
+      "loss": 1.04,
+      "step": 15080
+    },
+    {
+      "epoch": 2.41296,
+      "grad_norm": 1.8156325683665488,
+      "learning_rate": 1.962541790809701e-05,
+      "loss": 1.0731,
+      "step": 15081
+    },
+    {
+      "epoch": 2.41312,
+      "grad_norm": 1.6077425668854073,
+      "learning_rate": 1.9615141644050338e-05,
+      "loss": 1.0984,
+      "step": 15082
+    },
+    {
+      "epoch": 2.41328,
+      "grad_norm": 1.7448375605965722,
+      "learning_rate": 1.9604867778570722e-05,
+      "loss": 1.1449,
+      "step": 15083
+    },
+    {
+      "epoch": 2.41344,
+      "grad_norm": 1.7496810086903791,
+      "learning_rate": 1.9594596311964697e-05,
+      "loss": 1.0482,
+      "step": 15084
+    },
+    {
+      "epoch": 2.4136,
+      "grad_norm": 1.88421546488017,
+      "learning_rate": 1.958432724453876e-05,
+      "loss": 1.1173,
+      "step": 15085
+    },
+    {
+      "epoch": 2.41376,
+      "grad_norm": 1.8749167925415173,
+      "learning_rate": 1.9574060576599285e-05,
+      "loss": 1.0678,
+      "step": 15086
+    },
+    {
+      "epoch": 2.41392,
+      "grad_norm": 1.6668386655080025,
+      "learning_rate": 1.95637963084527e-05,
+      "loss": 1.0925,
+      "step": 15087
+    },
+    {
+      "epoch": 2.4140800000000002,
+      "grad_norm": 1.7329926648452194,
+      "learning_rate": 1.9553534440405187e-05,
+      "loss": 1.0962,
+      "step": 15088
+    },
+    {
+      "epoch": 2.41424,
+      "grad_norm": 1.8822855645786274,
+      "learning_rate": 1.9543274972762993e-05,
+      "loss": 1.1235,
+      "step": 15089
+    },
+    {
+      "epoch": 2.4144,
+      "grad_norm": 1.8226052317045487,
+      "learning_rate": 1.953301790583225e-05,
+      "loss": 1.113,
+      "step": 15090
+    },
+    {
+      "epoch": 2.41456,
+      "grad_norm": 1.7817051285408525,
+      "learning_rate": 1.9522763239918997e-05,
+      "loss": 1.0825,
+      "step": 15091
+    },
+    {
+      "epoch": 2.41472,
+      "grad_norm": 2.0798275030472047,
+      "learning_rate": 1.9512510975329213e-05,
+      "loss": 1.1326,
+      "step": 15092
+    },
+    {
+      "epoch": 2.41488,
+      "grad_norm": 2.057728136051462,
+      "learning_rate": 1.9502261112368825e-05,
+      "loss": 1.0881,
+      "step": 15093
+    },
+    {
+      "epoch": 2.41504,
+      "grad_norm": 2.082724858611973,
+      "learning_rate": 1.9492013651343665e-05,
+      "loss": 1.0532,
+      "step": 15094
+    },
+    {
+      "epoch": 2.4152,
+      "grad_norm": 1.9237859904463557,
+      "learning_rate": 1.9481768592559502e-05,
+      "loss": 1.0899,
+      "step": 15095
+    },
+    {
+      "epoch": 2.41536,
+      "grad_norm": 1.9039110718075705,
+      "learning_rate": 1.9471525936322044e-05,
+      "loss": 1.1289,
+      "step": 15096
+    },
+    {
+      "epoch": 2.41552,
+      "grad_norm": 1.7642110512670859,
+      "learning_rate": 1.9461285682936914e-05,
+      "loss": 1.1198,
+      "step": 15097
+    },
+    {
+      "epoch": 2.41568,
+      "grad_norm": 1.9720448728831466,
+      "learning_rate": 1.9451047832709623e-05,
+      "loss": 1.1663,
+      "step": 15098
+    },
+    {
+      "epoch": 2.41584,
+      "grad_norm": 1.8088636291820102,
+      "learning_rate": 1.944081238594575e-05,
+      "loss": 1.1065,
+      "step": 15099
+    },
+    {
+      "epoch": 2.416,
+      "grad_norm": 1.8892728264741732,
+      "learning_rate": 1.9430579342950595e-05,
+      "loss": 1.0869,
+      "step": 15100
+    },
+    {
+      "epoch": 2.41616,
+      "grad_norm": 2.039728168089262,
+      "learning_rate": 1.942034870402961e-05,
+      "loss": 1.0731,
+      "step": 15101
+    },
+    {
+      "epoch": 2.41632,
+      "grad_norm": 1.7365272480993297,
+      "learning_rate": 1.9410120469487947e-05,
+      "loss": 1.1332,
+      "step": 15102
+    },
+    {
+      "epoch": 2.41648,
+      "grad_norm": 2.0006008639018433,
+      "learning_rate": 1.93998946396309e-05,
+      "loss": 1.0728,
+      "step": 15103
+    },
+    {
+      "epoch": 2.41664,
+      "grad_norm": 1.625521498423363,
+      "learning_rate": 1.9389671214763505e-05,
+      "loss": 1.0467,
+      "step": 15104
+    },
+    {
+      "epoch": 2.4168,
+      "grad_norm": 2.196039432873421,
+      "learning_rate": 1.9379450195190895e-05,
+      "loss": 1.1662,
+      "step": 15105
+    },
+    {
+      "epoch": 2.41696,
+      "grad_norm": 2.195548055332278,
+      "learning_rate": 1.9369231581218005e-05,
+      "loss": 1.1666,
+      "step": 15106
+    },
+    {
+      "epoch": 2.41712,
+      "grad_norm": 1.9916850833538513,
+      "learning_rate": 1.9359015373149746e-05,
+      "loss": 1.1054,
+      "step": 15107
+    },
+    {
+      "epoch": 2.41728,
+      "grad_norm": 1.839932009129679,
+      "learning_rate": 1.9348801571290976e-05,
+      "loss": 1.1437,
+      "step": 15108
+    },
+    {
+      "epoch": 2.41744,
+      "grad_norm": 1.8436797094960473,
+      "learning_rate": 1.9338590175946437e-05,
+      "loss": 1.0885,
+      "step": 15109
+    },
+    {
+      "epoch": 2.4176,
+      "grad_norm": 1.9349227903673039,
+      "learning_rate": 1.932838118742083e-05,
+      "loss": 1.1021,
+      "step": 15110
+    },
+    {
+      "epoch": 2.41776,
+      "grad_norm": 1.6707270262896712,
+      "learning_rate": 1.9318174606018778e-05,
+      "loss": 1.1072,
+      "step": 15111
+    },
+    {
+      "epoch": 2.41792,
+      "grad_norm": 2.579540713485203,
+      "learning_rate": 1.9307970432044842e-05,
+      "loss": 1.1285,
+      "step": 15112
+    },
+    {
+      "epoch": 2.41808,
+      "grad_norm": 1.891650600078058,
+      "learning_rate": 1.9297768665803475e-05,
+      "loss": 1.1065,
+      "step": 15113
+    },
+    {
+      "epoch": 2.41824,
+      "grad_norm": 1.658676140477462,
+      "learning_rate": 1.9287569307599073e-05,
+      "loss": 1.1084,
+      "step": 15114
+    },
+    {
+      "epoch": 2.4184,
+      "grad_norm": 2.2399518105202314,
+      "learning_rate": 1.9277372357736047e-05,
+      "loss": 1.1129,
+      "step": 15115
+    },
+    {
+      "epoch": 2.41856,
+      "grad_norm": 1.8845388483272245,
+      "learning_rate": 1.9267177816518546e-05,
+      "loss": 1.0406,
+      "step": 15116
+    },
+    {
+      "epoch": 2.41872,
+      "grad_norm": 1.831144103378352,
+      "learning_rate": 1.925698568425087e-05,
+      "loss": 1.0982,
+      "step": 15117
+    },
+    {
+      "epoch": 2.41888,
+      "grad_norm": 1.7915394194630543,
+      "learning_rate": 1.924679596123704e-05,
+      "loss": 1.078,
+      "step": 15118
+    },
+    {
+      "epoch": 2.41904,
+      "grad_norm": 1.7277240661820286,
+      "learning_rate": 1.9236608647781206e-05,
+      "loss": 1.0695,
+      "step": 15119
+    },
+    {
+      "epoch": 2.4192,
+      "grad_norm": 1.799051977995196,
+      "learning_rate": 1.922642374418723e-05,
+      "loss": 1.1351,
+      "step": 15120
+    },
+    {
+      "epoch": 2.41936,
+      "grad_norm": 1.722340046166666,
+      "learning_rate": 1.9216241250759093e-05,
+      "loss": 1.0773,
+      "step": 15121
+    },
+    {
+      "epoch": 2.41952,
+      "grad_norm": 1.7916355130528625,
+      "learning_rate": 1.9206061167800594e-05,
+      "loss": 1.1135,
+      "step": 15122
+    },
+    {
+      "epoch": 2.41968,
+      "grad_norm": 1.6930073806711217,
+      "learning_rate": 1.919588349561551e-05,
+      "loss": 1.1077,
+      "step": 15123
+    },
+    {
+      "epoch": 2.4198399999999998,
+      "grad_norm": 1.831898202629371,
+      "learning_rate": 1.9185708234507516e-05,
+      "loss": 1.0603,
+      "step": 15124
+    },
+    {
+      "epoch": 2.42,
+      "grad_norm": 2.062037119113633,
+      "learning_rate": 1.917553538478023e-05,
+      "loss": 1.1127,
+      "step": 15125
+    },
+    {
+      "epoch": 2.42016,
+      "grad_norm": 1.9211514170515807,
+      "learning_rate": 1.9165364946737195e-05,
+      "loss": 1.1057,
+      "step": 15126
+    },
+    {
+      "epoch": 2.4203200000000002,
+      "grad_norm": 1.6300695988432166,
+      "learning_rate": 1.9155196920681884e-05,
+      "loss": 1.1125,
+      "step": 15127
+    },
+    {
+      "epoch": 2.42048,
+      "grad_norm": 1.6163582875197422,
+      "learning_rate": 1.9145031306917694e-05,
+      "loss": 1.1112,
+      "step": 15128
+    },
+    {
+      "epoch": 2.42064,
+      "grad_norm": 2.076779332114368,
+      "learning_rate": 1.9134868105747948e-05,
+      "loss": 1.1154,
+      "step": 15129
+    },
+    {
+      "epoch": 2.4208,
+      "grad_norm": 1.641170367186465,
+      "learning_rate": 1.9124707317475876e-05,
+      "loss": 1.0967,
+      "step": 15130
+    },
+    {
+      "epoch": 2.42096,
+      "grad_norm": 1.7375748345331758,
+      "learning_rate": 1.9114548942404753e-05,
+      "loss": 1.131,
+      "step": 15131
+    },
+    {
+      "epoch": 2.42112,
+      "grad_norm": 1.988209258493608,
+      "learning_rate": 1.910439298083756e-05,
+      "loss": 1.112,
+      "step": 15132
+    },
+    {
+      "epoch": 2.42128,
+      "grad_norm": 1.7830057863839295,
+      "learning_rate": 1.9094239433077464e-05,
+      "loss": 1.0206,
+      "step": 15133
+    },
+    {
+      "epoch": 2.42144,
+      "grad_norm": 1.9584041214526777,
+      "learning_rate": 1.9084088299427315e-05,
+      "loss": 1.1564,
+      "step": 15134
+    },
+    {
+      "epoch": 2.4215999999999998,
+      "grad_norm": 1.798126040637829,
+      "learning_rate": 1.9073939580190125e-05,
+      "loss": 1.1301,
+      "step": 15135
+    },
+    {
+      "epoch": 2.42176,
+      "grad_norm": 1.8249935793972933,
+      "learning_rate": 1.9063793275668597e-05,
+      "loss": 1.0671,
+      "step": 15136
+    },
+    {
+      "epoch": 2.42192,
+      "grad_norm": 2.059344972980852,
+      "learning_rate": 1.9053649386165574e-05,
+      "loss": 1.0478,
+      "step": 15137
+    },
+    {
+      "epoch": 2.4220800000000002,
+      "grad_norm": 1.919887088807059,
+      "learning_rate": 1.9043507911983692e-05,
+      "loss": 1.0813,
+      "step": 15138
+    },
+    {
+      "epoch": 2.42224,
+      "grad_norm": 1.7692732925416401,
+      "learning_rate": 1.903336885342557e-05,
+      "loss": 1.1298,
+      "step": 15139
+    },
+    {
+      "epoch": 2.4224,
+      "grad_norm": 1.7873296175279907,
+      "learning_rate": 1.902323221079374e-05,
+      "loss": 1.1069,
+      "step": 15140
+    },
+    {
+      "epoch": 2.42256,
+      "grad_norm": 2.122837659420738,
+      "learning_rate": 1.901309798439067e-05,
+      "loss": 1.0736,
+      "step": 15141
+    },
+    {
+      "epoch": 2.42272,
+      "grad_norm": 2.250185008832911,
+      "learning_rate": 1.9002966174518743e-05,
+      "loss": 1.1704,
+      "step": 15142
+    },
+    {
+      "epoch": 2.42288,
+      "grad_norm": 1.8520631242575107,
+      "learning_rate": 1.8992836781480282e-05,
+      "loss": 1.078,
+      "step": 15143
+    },
+    {
+      "epoch": 2.42304,
+      "grad_norm": 1.777529491096059,
+      "learning_rate": 1.8982709805577537e-05,
+      "loss": 1.0546,
+      "step": 15144
+    },
+    {
+      "epoch": 2.4232,
+      "grad_norm": 2.041011886188286,
+      "learning_rate": 1.897258524711264e-05,
+      "loss": 1.1165,
+      "step": 15145
+    },
+    {
+      "epoch": 2.42336,
+      "grad_norm": 1.496087094335833,
+      "learning_rate": 1.896246310638776e-05,
+      "loss": 1.091,
+      "step": 15146
+    },
+    {
+      "epoch": 2.42352,
+      "grad_norm": 1.9326793145588221,
+      "learning_rate": 1.8952343383704894e-05,
+      "loss": 1.1038,
+      "step": 15147
+    },
+    {
+      "epoch": 2.42368,
+      "grad_norm": 2.113662112028548,
+      "learning_rate": 1.894222607936601e-05,
+      "loss": 1.0679,
+      "step": 15148
+    },
+    {
+      "epoch": 2.42384,
+      "grad_norm": 2.190063754515941,
+      "learning_rate": 1.8932111193672973e-05,
+      "loss": 1.0709,
+      "step": 15149
+    },
+    {
+      "epoch": 2.424,
+      "grad_norm": 1.9051763795642571,
+      "learning_rate": 1.8921998726927627e-05,
+      "loss": 1.0618,
+      "step": 15150
+    },
+    {
+      "epoch": 2.42416,
+      "grad_norm": 1.9785083949374942,
+      "learning_rate": 1.8911888679431678e-05,
+      "loss": 1.089,
+      "step": 15151
+    },
+    {
+      "epoch": 2.42432,
+      "grad_norm": 1.7514591488883198,
+      "learning_rate": 1.8901781051486824e-05,
+      "loss": 1.1072,
+      "step": 15152
+    },
+    {
+      "epoch": 2.42448,
+      "grad_norm": 2.1979071027979713,
+      "learning_rate": 1.8891675843394652e-05,
+      "loss": 1.0962,
+      "step": 15153
+    },
+    {
+      "epoch": 2.42464,
+      "grad_norm": 1.8558862916002874,
+      "learning_rate": 1.8881573055456682e-05,
+      "loss": 1.0621,
+      "step": 15154
+    },
+    {
+      "epoch": 2.4248,
+      "grad_norm": 2.002388060890282,
+      "learning_rate": 1.8871472687974367e-05,
+      "loss": 1.0963,
+      "step": 15155
+    },
+    {
+      "epoch": 2.42496,
+      "grad_norm": 2.628950858926667,
+      "learning_rate": 1.8861374741249094e-05,
+      "loss": 1.0983,
+      "step": 15156
+    },
+    {
+      "epoch": 2.42512,
+      "grad_norm": 1.9207657546373575,
+      "learning_rate": 1.885127921558214e-05,
+      "loss": 1.1099,
+      "step": 15157
+    },
+    {
+      "epoch": 2.42528,
+      "grad_norm": 1.9225266605096554,
+      "learning_rate": 1.8841186111274823e-05,
+      "loss": 1.1386,
+      "step": 15158
+    },
+    {
+      "epoch": 2.42544,
+      "grad_norm": 1.7676315676101064,
+      "learning_rate": 1.88310954286282e-05,
+      "loss": 1.0802,
+      "step": 15159
+    },
+    {
+      "epoch": 2.4256,
+      "grad_norm": 1.7676315676101064,
+      "learning_rate": 1.88310954286282e-05,
+      "loss": 1.0826,
+      "step": 15160
+    },
+    {
+      "epoch": 2.42576,
+      "grad_norm": 1.7676315676101064,
+      "learning_rate": 1.88310954286282e-05,
+      "loss": 1.1251,
+      "step": 15161
+    },
+    {
+      "epoch": 2.42592,
+      "grad_norm": 2.4645735519287197,
+      "learning_rate": 1.882100716794346e-05,
+      "loss": 1.12,
+      "step": 15162
+    },
+    {
+      "epoch": 2.42608,
+      "grad_norm": 2.890861868265275,
+      "learning_rate": 1.8810921329521526e-05,
+      "loss": 1.1139,
+      "step": 15163
+    },
+    {
+      "epoch": 2.42624,
+      "grad_norm": 1.7735757275077826,
+      "learning_rate": 1.8800837913663428e-05,
+      "loss": 1.0581,
+      "step": 15164
+    },
+    {
+      "epoch": 2.4264,
+      "grad_norm": 1.8154158862709884,
+      "learning_rate": 1.8790756920669993e-05,
+      "loss": 1.0981,
+      "step": 15165
+    },
+    {
+      "epoch": 2.42656,
+      "grad_norm": 2.0892342182127726,
+      "learning_rate": 1.878067835084204e-05,
+      "loss": 1.1336,
+      "step": 15166
+    },
+    {
+      "epoch": 2.42672,
+      "grad_norm": 2.3236786463108645,
+      "learning_rate": 1.8770602204480304e-05,
+      "loss": 1.0658,
+      "step": 15167
+    },
+    {
+      "epoch": 2.42688,
+      "grad_norm": 2.0338882129243068,
+      "learning_rate": 1.876052848188543e-05,
+      "loss": 1.0845,
+      "step": 15168
+    },
+    {
+      "epoch": 2.42704,
+      "grad_norm": 1.9982432563465633,
+      "learning_rate": 1.8750457183358007e-05,
+      "loss": 1.1766,
+      "step": 15169
+    },
+    {
+      "epoch": 2.4272,
+      "grad_norm": 1.8566344265633814,
+      "learning_rate": 1.8740388309198554e-05,
+      "loss": 1.1128,
+      "step": 15170
+    },
+    {
+      "epoch": 2.42736,
+      "grad_norm": 2.018240011292465,
+      "learning_rate": 1.87303218597075e-05,
+      "loss": 1.0986,
+      "step": 15171
+    },
+    {
+      "epoch": 2.42752,
+      "grad_norm": 1.781886274196312,
+      "learning_rate": 1.872025783518523e-05,
+      "loss": 1.0566,
+      "step": 15172
+    },
+    {
+      "epoch": 2.42768,
+      "grad_norm": 2.3609145544998795,
+      "learning_rate": 1.871019623593201e-05,
+      "loss": 1.0468,
+      "step": 15173
+    },
+    {
+      "epoch": 2.4278399999999998,
+      "grad_norm": 2.054192375177543,
+      "learning_rate": 1.87001370622481e-05,
+      "loss": 1.0994,
+      "step": 15174
+    },
+    {
+      "epoch": 2.428,
+      "grad_norm": 1.727444104830747,
+      "learning_rate": 1.86900803144336e-05,
+      "loss": 1.0527,
+      "step": 15175
+    },
+    {
+      "epoch": 2.42816,
+      "grad_norm": 1.5861647274026578,
+      "learning_rate": 1.8680025992788675e-05,
+      "loss": 1.0747,
+      "step": 15176
+    },
+    {
+      "epoch": 2.42832,
+      "grad_norm": 1.86110442007064,
+      "learning_rate": 1.866997409761322e-05,
+      "loss": 1.0963,
+      "step": 15177
+    },
+    {
+      "epoch": 2.42848,
+      "grad_norm": 1.6486578983961842,
+      "learning_rate": 1.865992462920729e-05,
+      "loss": 1.0874,
+      "step": 15178
+    },
+    {
+      "epoch": 2.42864,
+      "grad_norm": 1.9033846495691307,
+      "learning_rate": 1.8649877587870623e-05,
+      "loss": 1.1234,
+      "step": 15179
+    },
+    {
+      "epoch": 2.4288,
+      "grad_norm": 1.7468222006266814,
+      "learning_rate": 1.8639832973903106e-05,
+      "loss": 1.0296,
+      "step": 15180
+    },
+    {
+      "epoch": 2.42896,
+      "grad_norm": 1.83182912186619,
+      "learning_rate": 1.8629790787604416e-05,
+      "loss": 1.0527,
+      "step": 15181
+    },
+    {
+      "epoch": 2.42912,
+      "grad_norm": 1.8963034747771244,
+      "learning_rate": 1.861975102927419e-05,
+      "loss": 1.1011,
+      "step": 15182
+    },
+    {
+      "epoch": 2.42928,
+      "grad_norm": 1.9614920945511536,
+      "learning_rate": 1.8609713699212018e-05,
+      "loss": 1.0911,
+      "step": 15183
+    },
+    {
+      "epoch": 2.42944,
+      "grad_norm": 1.702159596689129,
+      "learning_rate": 1.859967879771739e-05,
+      "loss": 1.0478,
+      "step": 15184
+    },
+    {
+      "epoch": 2.4295999999999998,
+      "grad_norm": 1.6447166254647465,
+      "learning_rate": 1.858964632508974e-05,
+      "loss": 1.105,
+      "step": 15185
+    },
+    {
+      "epoch": 2.42976,
+      "grad_norm": 1.9807714539345174,
+      "learning_rate": 1.857961628162841e-05,
+      "loss": 1.12,
+      "step": 15186
+    },
+    {
+      "epoch": 2.42992,
+      "grad_norm": 1.7457234104950716,
+      "learning_rate": 1.8569588667632687e-05,
+      "loss": 1.0539,
+      "step": 15187
+    },
+    {
+      "epoch": 2.4300800000000002,
+      "grad_norm": 1.7551105321136409,
+      "learning_rate": 1.8559563483401777e-05,
+      "loss": 1.0947,
+      "step": 15188
+    },
+    {
+      "epoch": 2.43024,
+      "grad_norm": 1.9173595654844775,
+      "learning_rate": 1.8549540729234827e-05,
+      "loss": 1.0833,
+      "step": 15189
+    },
+    {
+      "epoch": 2.4304,
+      "grad_norm": 2.1507710245228333,
+      "learning_rate": 1.85395204054309e-05,
+      "loss": 1.1091,
+      "step": 15190
+    },
+    {
+      "epoch": 2.43056,
+      "grad_norm": 2.1039849793870276,
+      "learning_rate": 1.8529502512288953e-05,
+      "loss": 1.0681,
+      "step": 15191
+    },
+    {
+      "epoch": 2.43072,
+      "grad_norm": 1.6693617960841067,
+      "learning_rate": 1.8519487050107988e-05,
+      "loss": 1.1059,
+      "step": 15192
+    },
+    {
+      "epoch": 2.43088,
+      "grad_norm": 1.9843726813280975,
+      "learning_rate": 1.8509474019186745e-05,
+      "loss": 1.1033,
+      "step": 15193
+    },
+    {
+      "epoch": 2.43104,
+      "grad_norm": 1.837776262116083,
+      "learning_rate": 1.8499463419824115e-05,
+      "loss": 1.0785,
+      "step": 15194
+    },
+    {
+      "epoch": 2.4312,
+      "grad_norm": 2.2846956463029713,
+      "learning_rate": 1.8489455252318677e-05,
+      "loss": 1.0896,
+      "step": 15195
+    },
+    {
+      "epoch": 2.43136,
+      "grad_norm": 1.9572675436135485,
+      "learning_rate": 1.847944951696915e-05,
+      "loss": 1.0917,
+      "step": 15196
+    },
+    {
+      "epoch": 2.43152,
+      "grad_norm": 1.8582993556910936,
+      "learning_rate": 1.846944621407406e-05,
+      "loss": 1.079,
+      "step": 15197
+    },
+    {
+      "epoch": 2.43168,
+      "grad_norm": 1.8939906390135333,
+      "learning_rate": 1.8459445343931892e-05,
+      "loss": 1.1057,
+      "step": 15198
+    },
+    {
+      "epoch": 2.4318400000000002,
+      "grad_norm": 1.8225017500561864,
+      "learning_rate": 1.8449446906841062e-05,
+      "loss": 1.1064,
+      "step": 15199
+    },
+    {
+      "epoch": 2.432,
+      "grad_norm": 1.8540707176381104,
+      "learning_rate": 1.8439450903099896e-05,
+      "loss": 1.1137,
+      "step": 15200
+    },
+    {
+      "epoch": 2.43216,
+      "grad_norm": 1.7819784164612347,
+      "learning_rate": 1.8429457333006673e-05,
+      "loss": 1.093,
+      "step": 15201
+    },
+    {
+      "epoch": 2.43232,
+      "grad_norm": 1.9495156735892611,
+      "learning_rate": 1.841946619685958e-05,
+      "loss": 1.1416,
+      "step": 15202
+    },
+    {
+      "epoch": 2.43248,
+      "grad_norm": 1.955494027935479,
+      "learning_rate": 1.840947749495675e-05,
+      "loss": 1.0676,
+      "step": 15203
+    },
+    {
+      "epoch": 2.43264,
+      "grad_norm": 1.619871551682588,
+      "learning_rate": 1.839949122759621e-05,
+      "loss": 1.0648,
+      "step": 15204
+    },
+    {
+      "epoch": 2.4328,
+      "grad_norm": 1.9685537437877536,
+      "learning_rate": 1.8389507395075933e-05,
+      "loss": 1.1217,
+      "step": 15205
+    },
+    {
+      "epoch": 2.43296,
+      "grad_norm": 1.728962001017333,
+      "learning_rate": 1.8379525997693848e-05,
+      "loss": 1.0718,
+      "step": 15206
+    },
+    {
+      "epoch": 2.43312,
+      "grad_norm": 1.7673706414936137,
+      "learning_rate": 1.836954703574778e-05,
+      "loss": 1.201,
+      "step": 15207
+    },
+    {
+      "epoch": 2.43328,
+      "grad_norm": 1.9872470196966903,
+      "learning_rate": 1.8359570509535486e-05,
+      "loss": 1.0877,
+      "step": 15208
+    },
+    {
+      "epoch": 2.43344,
+      "grad_norm": 1.8175376986742984,
+      "learning_rate": 1.8349596419354642e-05,
+      "loss": 1.1006,
+      "step": 15209
+    },
+    {
+      "epoch": 2.4336,
+      "grad_norm": 1.6491598368529772,
+      "learning_rate": 1.833962476550286e-05,
+      "loss": 1.0776,
+      "step": 15210
+    },
+    {
+      "epoch": 2.43376,
+      "grad_norm": 1.634245756317626,
+      "learning_rate": 1.832965554827769e-05,
+      "loss": 1.0463,
+      "step": 15211
+    },
+    {
+      "epoch": 2.43392,
+      "grad_norm": 2.0963400242053063,
+      "learning_rate": 1.8319688767976585e-05,
+      "loss": 1.067,
+      "step": 15212
+    },
+    {
+      "epoch": 2.43408,
+      "grad_norm": 2.0785560888404144,
+      "learning_rate": 1.8309724424896945e-05,
+      "loss": 1.1132,
+      "step": 15213
+    },
+    {
+      "epoch": 2.43424,
+      "grad_norm": 2.120903879557089,
+      "learning_rate": 1.8299762519336093e-05,
+      "loss": 1.0916,
+      "step": 15214
+    },
+    {
+      "epoch": 2.4344,
+      "grad_norm": 1.673332595930378,
+      "learning_rate": 1.8289803051591293e-05,
+      "loss": 1.0545,
+      "step": 15215
+    },
+    {
+      "epoch": 2.43456,
+      "grad_norm": 1.9458099691384667,
+      "learning_rate": 1.8279846021959668e-05,
+      "loss": 1.1168,
+      "step": 15216
+    },
+    {
+      "epoch": 2.43472,
+      "grad_norm": 1.9555706539581186,
+      "learning_rate": 1.8269891430738407e-05,
+      "loss": 1.046,
+      "step": 15217
+    },
+    {
+      "epoch": 2.43488,
+      "grad_norm": 1.8269022068161858,
+      "learning_rate": 1.8259939278224446e-05,
+      "loss": 1.0466,
+      "step": 15218
+    },
+    {
+      "epoch": 2.43504,
+      "grad_norm": 1.761949848978047,
+      "learning_rate": 1.8249989564714843e-05,
+      "loss": 1.0705,
+      "step": 15219
+    },
+    {
+      "epoch": 2.4352,
+      "grad_norm": 1.8649641644288932,
+      "learning_rate": 1.8240042290506365e-05,
+      "loss": 1.0719,
+      "step": 15220
+    },
+    {
+      "epoch": 2.43536,
+      "grad_norm": 1.8065264207999425,
+      "learning_rate": 1.8230097455895923e-05,
+      "loss": 1.1301,
+      "step": 15221
+    },
+    {
+      "epoch": 2.43552,
+      "grad_norm": 2.0516676108097056,
+      "learning_rate": 1.8220155061180223e-05,
+      "loss": 1.0943,
+      "step": 15222
+    },
+    {
+      "epoch": 2.43568,
+      "grad_norm": 1.8665915225776346,
+      "learning_rate": 1.8210215106655927e-05,
+      "loss": 1.0366,
+      "step": 15223
+    },
+    {
+      "epoch": 2.43584,
+      "grad_norm": 1.7789861901353348,
+      "learning_rate": 1.8200277592619642e-05,
+      "loss": 1.1085,
+      "step": 15224
+    },
+    {
+      "epoch": 2.436,
+      "grad_norm": 2.1606033904393773,
+      "learning_rate": 1.8190342519367863e-05,
+      "loss": 1.0741,
+      "step": 15225
+    },
+    {
+      "epoch": 2.43616,
+      "grad_norm": 1.9561718312527279,
+      "learning_rate": 1.818040988719706e-05,
+      "loss": 1.036,
+      "step": 15226
+    },
+    {
+      "epoch": 2.43632,
+      "grad_norm": 1.8938077722660709,
+      "learning_rate": 1.817047969640361e-05,
+      "loss": 1.0828,
+      "step": 15227
+    },
+    {
+      "epoch": 2.43648,
+      "grad_norm": 1.8141025790228114,
+      "learning_rate": 1.81605519472838e-05,
+      "loss": 1.0633,
+      "step": 15228
+    },
+    {
+      "epoch": 2.43664,
+      "grad_norm": 1.8081928700484429,
+      "learning_rate": 1.815062664013386e-05,
+      "loss": 1.1017,
+      "step": 15229
+    },
+    {
+      "epoch": 2.4368,
+      "grad_norm": 1.9449053110192525,
+      "learning_rate": 1.8140703775249967e-05,
+      "loss": 1.09,
+      "step": 15230
+    },
+    {
+      "epoch": 2.43696,
+      "grad_norm": 1.6124324701916866,
+      "learning_rate": 1.8130783352928183e-05,
+      "loss": 1.0703,
+      "step": 15231
+    },
+    {
+      "epoch": 2.43712,
+      "grad_norm": 1.7418818852925886,
+      "learning_rate": 1.8120865373464513e-05,
+      "loss": 1.0676,
+      "step": 15232
+    },
+    {
+      "epoch": 2.43728,
+      "grad_norm": 1.7737712215759618,
+      "learning_rate": 1.8110949837154954e-05,
+      "loss": 1.0517,
+      "step": 15233
+    },
+    {
+      "epoch": 2.43744,
+      "grad_norm": 1.7889843938911765,
+      "learning_rate": 1.8101036744295274e-05,
+      "loss": 1.1566,
+      "step": 15234
+    },
+    {
+      "epoch": 2.4375999999999998,
+      "grad_norm": 1.8429164922234225,
+      "learning_rate": 1.8091126095181365e-05,
+      "loss": 1.0748,
+      "step": 15235
+    },
+    {
+      "epoch": 2.43776,
+      "grad_norm": 1.8710017811092912,
+      "learning_rate": 1.808121789010886e-05,
+      "loss": 1.1196,
+      "step": 15236
+    },
+    {
+      "epoch": 2.43792,
+      "grad_norm": 1.5524689177180309,
+      "learning_rate": 1.8071312129373465e-05,
+      "loss": 1.0761,
+      "step": 15237
+    },
+    {
+      "epoch": 2.4380800000000002,
+      "grad_norm": 1.765972950273971,
+      "learning_rate": 1.806140881327073e-05,
+      "loss": 1.0634,
+      "step": 15238
+    },
+    {
+      "epoch": 2.43824,
+      "grad_norm": 1.897860038307566,
+      "learning_rate": 1.8051507942096167e-05,
+      "loss": 1.1497,
+      "step": 15239
+    },
+    {
+      "epoch": 2.4384,
+      "grad_norm": 1.6498761056966909,
+      "learning_rate": 1.8041609516145186e-05,
+      "loss": 1.0696,
+      "step": 15240
+    },
+    {
+      "epoch": 2.43856,
+      "grad_norm": 1.754253410897607,
+      "learning_rate": 1.8031713535713158e-05,
+      "loss": 1.0603,
+      "step": 15241
+    },
+    {
+      "epoch": 2.43872,
+      "grad_norm": 1.7739952641187053,
+      "learning_rate": 1.8021820001095358e-05,
+      "loss": 1.0928,
+      "step": 15242
+    },
+    {
+      "epoch": 2.43888,
+      "grad_norm": 1.8221641624686058,
+      "learning_rate": 1.8011928912586985e-05,
+      "loss": 1.1008,
+      "step": 15243
+    },
+    {
+      "epoch": 2.43904,
+      "grad_norm": 1.7203146984013509,
+      "learning_rate": 1.8002040270483188e-05,
+      "loss": 1.0661,
+      "step": 15244
+    },
+    {
+      "epoch": 2.4392,
+      "grad_norm": 1.6707090536316003,
+      "learning_rate": 1.7992154075079036e-05,
+      "loss": 1.0843,
+      "step": 15245
+    },
+    {
+      "epoch": 2.4393599999999998,
+      "grad_norm": 2.0055461112698985,
+      "learning_rate": 1.798227032666949e-05,
+      "loss": 1.1456,
+      "step": 15246
+    },
+    {
+      "epoch": 2.43952,
+      "grad_norm": 1.726135964145786,
+      "learning_rate": 1.7972389025549506e-05,
+      "loss": 1.1605,
+      "step": 15247
+    },
+    {
+      "epoch": 2.43968,
+      "grad_norm": 1.8610406523024505,
+      "learning_rate": 1.796251017201387e-05,
+      "loss": 1.0817,
+      "step": 15248
+    },
+    {
+      "epoch": 2.4398400000000002,
+      "grad_norm": 2.137567037733564,
+      "learning_rate": 1.7952633766357442e-05,
+      "loss": 1.1327,
+      "step": 15249
+    },
+    {
+      "epoch": 2.44,
+      "grad_norm": 1.8743251853113283,
+      "learning_rate": 1.794275980887481e-05,
+      "loss": 1.0815,
+      "step": 15250
+    },
+    {
+      "epoch": 2.44016,
+      "grad_norm": 1.7704887643985212,
+      "learning_rate": 1.7932888299860705e-05,
+      "loss": 1.0898,
+      "step": 15251
+    },
+    {
+      "epoch": 2.44032,
+      "grad_norm": 1.856473908764278,
+      "learning_rate": 1.7923019239609584e-05,
+      "loss": 1.1058,
+      "step": 15252
+    },
+    {
+      "epoch": 2.44048,
+      "grad_norm": 2.078006030758375,
+      "learning_rate": 1.7913152628415995e-05,
+      "loss": 1.1356,
+      "step": 15253
+    },
+    {
+      "epoch": 2.44064,
+      "grad_norm": 1.867392524887339,
+      "learning_rate": 1.790328846657432e-05,
+      "loss": 1.1178,
+      "step": 15254
+    },
+    {
+      "epoch": 2.4408,
+      "grad_norm": 1.8986218220645572,
+      "learning_rate": 1.7893426754378895e-05,
+      "loss": 1.1035,
+      "step": 15255
+    },
+    {
+      "epoch": 2.44096,
+      "grad_norm": 1.6548442652383735,
+      "learning_rate": 1.7883567492123966e-05,
+      "loss": 1.0867,
+      "step": 15256
+    },
+    {
+      "epoch": 2.44112,
+      "grad_norm": 2.15787219791641,
+      "learning_rate": 1.7873710680103728e-05,
+      "loss": 1.1047,
+      "step": 15257
+    },
+    {
+      "epoch": 2.44128,
+      "grad_norm": 2.3308033683282625,
+      "learning_rate": 1.7863856318612293e-05,
+      "loss": 1.0696,
+      "step": 15258
+    },
+    {
+      "epoch": 2.44144,
+      "grad_norm": 2.0134172416896745,
+      "learning_rate": 1.785400440794369e-05,
+      "loss": 1.0601,
+      "step": 15259
+    },
+    {
+      "epoch": 2.4416,
+      "grad_norm": 1.9913843272984353,
+      "learning_rate": 1.784415494839191e-05,
+      "loss": 1.1216,
+      "step": 15260
+    },
+    {
+      "epoch": 2.44176,
+      "grad_norm": 1.4630125974882697,
+      "learning_rate": 1.7834307940250826e-05,
+      "loss": 1.0754,
+      "step": 15261
+    },
+    {
+      "epoch": 2.44192,
+      "grad_norm": 1.7189754105865958,
+      "learning_rate": 1.782446338381427e-05,
+      "loss": 1.0992,
+      "step": 15262
+    },
+    {
+      "epoch": 2.44208,
+      "grad_norm": 2.1531058007554855,
+      "learning_rate": 1.7814621279375987e-05,
+      "loss": 1.078,
+      "step": 15263
+    },
+    {
+      "epoch": 2.44224,
+      "grad_norm": 1.9183481737416719,
+      "learning_rate": 1.7804781627229616e-05,
+      "loss": 1.075,
+      "step": 15264
+    },
+    {
+      "epoch": 2.4424,
+      "grad_norm": 1.8394029529676925,
+      "learning_rate": 1.7794944427668814e-05,
+      "loss": 1.0935,
+      "step": 15265
+    },
+    {
+      "epoch": 2.44256,
+      "grad_norm": 1.8121017177377063,
+      "learning_rate": 1.7785109680987088e-05,
+      "loss": 1.0827,
+      "step": 15266
+    },
+    {
+      "epoch": 2.44272,
+      "grad_norm": 1.917855188443172,
+      "learning_rate": 1.7775277387477894e-05,
+      "loss": 1.0839,
+      "step": 15267
+    },
+    {
+      "epoch": 2.44288,
+      "grad_norm": 1.830677611923429,
+      "learning_rate": 1.7765447547434598e-05,
+      "loss": 1.1134,
+      "step": 15268
+    },
+    {
+      "epoch": 2.44304,
+      "grad_norm": 2.1984591820561707,
+      "learning_rate": 1.775562016115052e-05,
+      "loss": 1.1226,
+      "step": 15269
+    },
+    {
+      "epoch": 2.4432,
+      "grad_norm": 2.0252545903679047,
+      "learning_rate": 1.7745795228918906e-05,
+      "loss": 1.1242,
+      "step": 15270
+    },
+    {
+      "epoch": 2.44336,
+      "grad_norm": 1.929023884926729,
+      "learning_rate": 1.7735972751032894e-05,
+      "loss": 1.1648,
+      "step": 15271
+    },
+    {
+      "epoch": 2.44352,
+      "grad_norm": 1.8173254251843858,
+      "learning_rate": 1.7726152727785594e-05,
+      "loss": 1.0515,
+      "step": 15272
+    },
+    {
+      "epoch": 2.44368,
+      "grad_norm": 1.7216003347777074,
+      "learning_rate": 1.7716335159469976e-05,
+      "loss": 1.0984,
+      "step": 15273
+    },
+    {
+      "epoch": 2.44384,
+      "grad_norm": 1.788400148867721,
+      "learning_rate": 1.770652004637907e-05,
+      "loss": 1.0994,
+      "step": 15274
+    },
+    {
+      "epoch": 2.444,
+      "grad_norm": 1.8976708819274843,
+      "learning_rate": 1.7696707388805645e-05,
+      "loss": 1.076,
+      "step": 15275
+    },
+    {
+      "epoch": 2.44416,
+      "grad_norm": 1.9458379902716154,
+      "learning_rate": 1.7686897187042585e-05,
+      "loss": 1.1304,
+      "step": 15276
+    },
+    {
+      "epoch": 2.44432,
+      "grad_norm": 1.6789165941794975,
+      "learning_rate": 1.7677089441382523e-05,
+      "loss": 1.06,
+      "step": 15277
+    },
+    {
+      "epoch": 2.44448,
+      "grad_norm": 1.9159413191543042,
+      "learning_rate": 1.7667284152118212e-05,
+      "loss": 1.1014,
+      "step": 15278
+    },
+    {
+      "epoch": 2.44464,
+      "grad_norm": 1.7253434259610212,
+      "learning_rate": 1.7657481319542112e-05,
+      "loss": 1.0877,
+      "step": 15279
+    },
+    {
+      "epoch": 2.4448,
+      "grad_norm": 1.7711925869148553,
+      "learning_rate": 1.764768094394681e-05,
+      "loss": 1.0834,
+      "step": 15280
+    },
+    {
+      "epoch": 2.44496,
+      "grad_norm": 1.7158811964279823,
+      "learning_rate": 1.7637883025624715e-05,
+      "loss": 1.0635,
+      "step": 15281
+    },
+    {
+      "epoch": 2.44512,
+      "grad_norm": 2.034092809743206,
+      "learning_rate": 1.762808756486817e-05,
+      "loss": 1.094,
+      "step": 15282
+    },
+    {
+      "epoch": 2.44528,
+      "grad_norm": 1.62457051705886,
+      "learning_rate": 1.761829456196946e-05,
+      "loss": 1.0611,
+      "step": 15283
+    },
+    {
+      "epoch": 2.44544,
+      "grad_norm": 1.7472108527046961,
+      "learning_rate": 1.7608504017220807e-05,
+      "loss": 1.0819,
+      "step": 15284
+    },
+    {
+      "epoch": 2.4455999999999998,
+      "grad_norm": 1.6992851124385362,
+      "learning_rate": 1.7598715930914335e-05,
+      "loss": 1.0554,
+      "step": 15285
+    },
+    {
+      "epoch": 2.44576,
+      "grad_norm": 1.8058188537535995,
+      "learning_rate": 1.7588930303342098e-05,
+      "loss": 1.101,
+      "step": 15286
+    },
+    {
+      "epoch": 2.44592,
+      "grad_norm": 1.9893664927487544,
+      "learning_rate": 1.7579147134796116e-05,
+      "loss": 1.1243,
+      "step": 15287
+    },
+    {
+      "epoch": 2.44608,
+      "grad_norm": 1.8709051137533876,
+      "learning_rate": 1.7569366425568267e-05,
+      "loss": 1.1238,
+      "step": 15288
+    },
+    {
+      "epoch": 2.44624,
+      "grad_norm": 1.9108770607947612,
+      "learning_rate": 1.755958817595038e-05,
+      "loss": 1.1101,
+      "step": 15289
+    },
+    {
+      "epoch": 2.4464,
+      "grad_norm": 1.7116001466146316,
+      "learning_rate": 1.754981238623432e-05,
+      "loss": 1.1008,
+      "step": 15290
+    },
+    {
+      "epoch": 2.44656,
+      "grad_norm": 1.715230685683673,
+      "learning_rate": 1.7540039056711656e-05,
+      "loss": 1.0764,
+      "step": 15291
+    },
+    {
+      "epoch": 2.44672,
+      "grad_norm": 1.7858727370501004,
+      "learning_rate": 1.7530268187674115e-05,
+      "loss": 1.0665,
+      "step": 15292
+    },
+    {
+      "epoch": 2.44688,
+      "grad_norm": 2.150222502221236,
+      "learning_rate": 1.752049977941316e-05,
+      "loss": 1.0545,
+      "step": 15293
+    },
+    {
+      "epoch": 2.44704,
+      "grad_norm": 1.5504383383106939,
+      "learning_rate": 1.751073383222035e-05,
+      "loss": 1.0512,
+      "step": 15294
+    },
+    {
+      "epoch": 2.4472,
+      "grad_norm": 1.7646008568900926,
+      "learning_rate": 1.7500970346387004e-05,
+      "loss": 1.077,
+      "step": 15295
+    },
+    {
+      "epoch": 2.4473599999999998,
+      "grad_norm": 1.9055834086493957,
+      "learning_rate": 1.7491209322204503e-05,
+      "loss": 1.1042,
+      "step": 15296
+    },
+    {
+      "epoch": 2.44752,
+      "grad_norm": 1.6598366075703455,
+      "learning_rate": 1.748145075996409e-05,
+      "loss": 1.0699,
+      "step": 15297
+    },
+    {
+      "epoch": 2.44768,
+      "grad_norm": 1.9001627956831308,
+      "learning_rate": 1.747169465995694e-05,
+      "loss": 1.0953,
+      "step": 15298
+    },
+    {
+      "epoch": 2.4478400000000002,
+      "grad_norm": 1.6055469078513358,
+      "learning_rate": 1.746194102247417e-05,
+      "loss": 1.0286,
+      "step": 15299
+    },
+    {
+      "epoch": 2.448,
+      "grad_norm": 1.6967314212059754,
+      "learning_rate": 1.745218984780681e-05,
+      "loss": 1.0647,
+      "step": 15300
+    },
+    {
+      "epoch": 2.44816,
+      "grad_norm": 1.6312090771247656,
+      "learning_rate": 1.7442441136245825e-05,
+      "loss": 1.0781,
+      "step": 15301
+    },
+    {
+      "epoch": 2.44832,
+      "grad_norm": 1.763006106345906,
+      "learning_rate": 1.7432694888082092e-05,
+      "loss": 1.1259,
+      "step": 15302
+    },
+    {
+      "epoch": 2.44848,
+      "grad_norm": 1.774878118092575,
+      "learning_rate": 1.7422951103606422e-05,
+      "loss": 1.0515,
+      "step": 15303
+    },
+    {
+      "epoch": 2.44864,
+      "grad_norm": 1.8667066742658718,
+      "learning_rate": 1.7413209783109576e-05,
+      "loss": 1.0813,
+      "step": 15304
+    },
+    {
+      "epoch": 2.4488,
+      "grad_norm": 1.604335118031338,
+      "learning_rate": 1.740347092688217e-05,
+      "loss": 1.0576,
+      "step": 15305
+    },
+    {
+      "epoch": 2.44896,
+      "grad_norm": 1.7429610556161494,
+      "learning_rate": 1.73937345352149e-05,
+      "loss": 1.058,
+      "step": 15306
+    },
+    {
+      "epoch": 2.44912,
+      "grad_norm": 1.9195982153790005,
+      "learning_rate": 1.7384000608398167e-05,
+      "loss": 1.0681,
+      "step": 15307
+    },
+    {
+      "epoch": 2.44928,
+      "grad_norm": 1.7223757277940943,
+      "learning_rate": 1.737426914672252e-05,
+      "loss": 1.0914,
+      "step": 15308
+    },
+    {
+      "epoch": 2.44944,
+      "grad_norm": 1.7782689165476357,
+      "learning_rate": 1.7364540150478238e-05,
+      "loss": 1.07,
+      "step": 15309
+    },
+    {
+      "epoch": 2.4496,
+      "grad_norm": 1.6985713282531247,
+      "learning_rate": 1.735481361995568e-05,
+      "loss": 1.1003,
+      "step": 15310
+    },
+    {
+      "epoch": 2.44976,
+      "grad_norm": 1.9103672115732315,
+      "learning_rate": 1.734508955544506e-05,
+      "loss": 1.1468,
+      "step": 15311
+    },
+    {
+      "epoch": 2.44992,
+      "grad_norm": 1.7085102462427688,
+      "learning_rate": 1.7335367957236524e-05,
+      "loss": 1.0766,
+      "step": 15312
+    },
+    {
+      "epoch": 2.45008,
+      "grad_norm": 1.658423586003805,
+      "learning_rate": 1.7325648825620168e-05,
+      "loss": 1.1117,
+      "step": 15313
+    },
+    {
+      "epoch": 2.45024,
+      "grad_norm": 1.7669232796097298,
+      "learning_rate": 1.7315932160885972e-05,
+      "loss": 1.0968,
+      "step": 15314
+    },
+    {
+      "epoch": 2.4504,
+      "grad_norm": 1.7168127105723585,
+      "learning_rate": 1.730621796332388e-05,
+      "loss": 1.1299,
+      "step": 15315
+    },
+    {
+      "epoch": 2.45056,
+      "grad_norm": 1.7697999078995448,
+      "learning_rate": 1.729650623322374e-05,
+      "loss": 1.0722,
+      "step": 15316
+    },
+    {
+      "epoch": 2.45072,
+      "grad_norm": 1.6553404867607957,
+      "learning_rate": 1.728679697087534e-05,
+      "loss": 1.0824,
+      "step": 15317
+    },
+    {
+      "epoch": 2.45088,
+      "grad_norm": 1.7654434557951517,
+      "learning_rate": 1.72770901765684e-05,
+      "loss": 1.1062,
+      "step": 15318
+    },
+    {
+      "epoch": 2.45104,
+      "grad_norm": 1.8269731725391907,
+      "learning_rate": 1.7267385850592555e-05,
+      "loss": 1.0569,
+      "step": 15319
+    },
+    {
+      "epoch": 2.4512,
+      "grad_norm": 1.7936103921094506,
+      "learning_rate": 1.725768399323735e-05,
+      "loss": 1.0807,
+      "step": 15320
+    },
+    {
+      "epoch": 2.45136,
+      "grad_norm": 1.9827517989230092,
+      "learning_rate": 1.7247984604792266e-05,
+      "loss": 1.1033,
+      "step": 15321
+    },
+    {
+      "epoch": 2.45152,
+      "grad_norm": 1.8864134238947767,
+      "learning_rate": 1.723828768554676e-05,
+      "loss": 1.1734,
+      "step": 15322
+    },
+    {
+      "epoch": 2.45168,
+      "grad_norm": 1.7348523552128052,
+      "learning_rate": 1.7228593235790158e-05,
+      "loss": 1.0967,
+      "step": 15323
+    },
+    {
+      "epoch": 2.45184,
+      "grad_norm": 1.7769974822135086,
+      "learning_rate": 1.7218901255811727e-05,
+      "loss": 1.058,
+      "step": 15324
+    },
+    {
+      "epoch": 2.452,
+      "grad_norm": 1.707568935763459,
+      "learning_rate": 1.7209211745900643e-05,
+      "loss": 1.0935,
+      "step": 15325
+    },
+    {
+      "epoch": 2.45216,
+      "grad_norm": 1.9573629575917264,
+      "learning_rate": 1.719952470634606e-05,
+      "loss": 1.1117,
+      "step": 15326
+    },
+    {
+      "epoch": 2.45232,
+      "grad_norm": 1.812368247839156,
+      "learning_rate": 1.7189840137436997e-05,
+      "loss": 1.112,
+      "step": 15327
+    },
+    {
+      "epoch": 2.45248,
+      "grad_norm": 1.813406809617385,
+      "learning_rate": 1.7180158039462435e-05,
+      "loss": 1.0573,
+      "step": 15328
+    },
+    {
+      "epoch": 2.45264,
+      "grad_norm": 1.7972418886320216,
+      "learning_rate": 1.717047841271128e-05,
+      "loss": 1.0918,
+      "step": 15329
+    },
+    {
+      "epoch": 2.4528,
+      "grad_norm": 1.6370765537676564,
+      "learning_rate": 1.7160801257472327e-05,
+      "loss": 1.075,
+      "step": 15330
+    },
+    {
+      "epoch": 2.45296,
+      "grad_norm": 1.9067865259403607,
+      "learning_rate": 1.7151126574034405e-05,
+      "loss": 1.0626,
+      "step": 15331
+    },
+    {
+      "epoch": 2.45312,
+      "grad_norm": 2.0374894703691995,
+      "learning_rate": 1.7141454362686094e-05,
+      "loss": 1.0655,
+      "step": 15332
+    },
+    {
+      "epoch": 2.45328,
+      "grad_norm": 1.7418211894693973,
+      "learning_rate": 1.7131784623716096e-05,
+      "loss": 1.0777,
+      "step": 15333
+    },
+    {
+      "epoch": 2.45344,
+      "grad_norm": 1.7154615221529474,
+      "learning_rate": 1.7122117357412836e-05,
+      "loss": 1.0667,
+      "step": 15334
+    },
+    {
+      "epoch": 2.4536,
+      "grad_norm": 1.773777895576418,
+      "learning_rate": 1.7112452564064896e-05,
+      "loss": 1.1229,
+      "step": 15335
+    },
+    {
+      "epoch": 2.45376,
+      "grad_norm": 2.0774948636638264,
+      "learning_rate": 1.7102790243960533e-05,
+      "loss": 1.0725,
+      "step": 15336
+    },
+    {
+      "epoch": 2.45392,
+      "grad_norm": 1.8683308142059756,
+      "learning_rate": 1.7093130397388136e-05,
+      "loss": 1.0824,
+      "step": 15337
+    },
+    {
+      "epoch": 2.45408,
+      "grad_norm": 2.079862583639984,
+      "learning_rate": 1.7083473024635922e-05,
+      "loss": 1.1278,
+      "step": 15338
+    },
+    {
+      "epoch": 2.45424,
+      "grad_norm": 2.0612014633666815,
+      "learning_rate": 1.7073818125992046e-05,
+      "loss": 1.1427,
+      "step": 15339
+    },
+    {
+      "epoch": 2.4544,
+      "grad_norm": 1.873065996230933,
+      "learning_rate": 1.7064165701744604e-05,
+      "loss": 1.1129,
+      "step": 15340
+    },
+    {
+      "epoch": 2.45456,
+      "grad_norm": 1.8594280510475396,
+      "learning_rate": 1.705451575218161e-05,
+      "loss": 1.1086,
+      "step": 15341
+    },
+    {
+      "epoch": 2.45472,
+      "grad_norm": 1.8092782540684402,
+      "learning_rate": 1.7044868277591e-05,
+      "loss": 1.0756,
+      "step": 15342
+    },
+    {
+      "epoch": 2.45488,
+      "grad_norm": 1.7908790840546256,
+      "learning_rate": 1.703522327826065e-05,
+      "loss": 1.0799,
+      "step": 15343
+    },
+    {
+      "epoch": 2.45504,
+      "grad_norm": 1.8610581779964563,
+      "learning_rate": 1.7025580754478332e-05,
+      "loss": 1.0788,
+      "step": 15344
+    },
+    {
+      "epoch": 2.4552,
+      "grad_norm": 1.6729238203061898,
+      "learning_rate": 1.7015940706531784e-05,
+      "loss": 1.0794,
+      "step": 15345
+    },
+    {
+      "epoch": 2.4553599999999998,
+      "grad_norm": 1.7203845966184452,
+      "learning_rate": 1.700630313470862e-05,
+      "loss": 1.077,
+      "step": 15346
+    },
+    {
+      "epoch": 2.45552,
+      "grad_norm": 1.6854912205017434,
+      "learning_rate": 1.6996668039296483e-05,
+      "loss": 1.0571,
+      "step": 15347
+    },
+    {
+      "epoch": 2.45568,
+      "grad_norm": 2.0747998946251096,
+      "learning_rate": 1.6987035420582776e-05,
+      "loss": 1.0801,
+      "step": 15348
+    },
+    {
+      "epoch": 2.4558400000000002,
+      "grad_norm": 1.6968711807783488,
+      "learning_rate": 1.6977405278855018e-05,
+      "loss": 1.1078,
+      "step": 15349
+    },
+    {
+      "epoch": 2.456,
+      "grad_norm": 1.8931940625267774,
+      "learning_rate": 1.696777761440046e-05,
+      "loss": 1.0792,
+      "step": 15350
+    },
+    {
+      "epoch": 2.45616,
+      "grad_norm": 1.854441428538191,
+      "learning_rate": 1.6958152427506478e-05,
+      "loss": 1.1038,
+      "step": 15351
+    },
+    {
+      "epoch": 2.45632,
+      "grad_norm": 1.8353389742067165,
+      "learning_rate": 1.694852971846017e-05,
+      "loss": 1.1287,
+      "step": 15352
+    },
+    {
+      "epoch": 2.45648,
+      "grad_norm": 1.6670491320263383,
+      "learning_rate": 1.693890948754875e-05,
+      "loss": 1.0635,
+      "step": 15353
+    },
+    {
+      "epoch": 2.45664,
+      "grad_norm": 1.7232519272983853,
+      "learning_rate": 1.6929291735059226e-05,
+      "loss": 1.0331,
+      "step": 15354
+    },
+    {
+      "epoch": 2.4568,
+      "grad_norm": 1.843817133497067,
+      "learning_rate": 1.69196764612786e-05,
+      "loss": 1.0852,
+      "step": 15355
+    },
+    {
+      "epoch": 2.45696,
+      "grad_norm": 1.6117018501637401,
+      "learning_rate": 1.6910063666493758e-05,
+      "loss": 1.0811,
+      "step": 15356
+    },
+    {
+      "epoch": 2.45712,
+      "grad_norm": 1.7243019067286165,
+      "learning_rate": 1.690045335099155e-05,
+      "loss": 1.1125,
+      "step": 15357
+    },
+    {
+      "epoch": 2.45728,
+      "grad_norm": 1.6564408856674484,
+      "learning_rate": 1.6890845515058716e-05,
+      "loss": 1.1297,
+      "step": 15358
+    },
+    {
+      "epoch": 2.45744,
+      "grad_norm": 1.8781927636341311,
+      "learning_rate": 1.6881240158981958e-05,
+      "loss": 1.0717,
+      "step": 15359
+    },
+    {
+      "epoch": 2.4576000000000002,
+      "grad_norm": 1.585654294044019,
+      "learning_rate": 1.687163728304787e-05,
+      "loss": 1.0452,
+      "step": 15360
+    },
+    {
+      "epoch": 2.45776,
+      "grad_norm": 1.5308176880853013,
+      "learning_rate": 1.6862036887542997e-05,
+      "loss": 1.1282,
+      "step": 15361
+    },
+    {
+      "epoch": 2.45792,
+      "grad_norm": 1.7165341739681281,
+      "learning_rate": 1.685243897275378e-05,
+      "loss": 1.1017,
+      "step": 15362
+    },
+    {
+      "epoch": 2.45808,
+      "grad_norm": 1.7043168907509456,
+      "learning_rate": 1.684284353896667e-05,
+      "loss": 1.1291,
+      "step": 15363
+    },
+    {
+      "epoch": 2.45824,
+      "grad_norm": 1.6238329663635656,
+      "learning_rate": 1.6833250586467886e-05,
+      "loss": 1.0743,
+      "step": 15364
+    },
+    {
+      "epoch": 2.4584,
+      "grad_norm": 1.8505958348090716,
+      "learning_rate": 1.682366011554377e-05,
+      "loss": 1.0631,
+      "step": 15365
+    },
+    {
+      "epoch": 2.45856,
+      "grad_norm": 1.769977588522472,
+      "learning_rate": 1.6814072126480385e-05,
+      "loss": 1.0983,
+      "step": 15366
+    },
+    {
+      "epoch": 2.45872,
+      "grad_norm": 1.768151734453956,
+      "learning_rate": 1.680448661956392e-05,
+      "loss": 1.0992,
+      "step": 15367
+    },
+    {
+      "epoch": 2.45888,
+      "grad_norm": 1.7163155137335648,
+      "learning_rate": 1.6794903595080292e-05,
+      "loss": 1.1073,
+      "step": 15368
+    },
+    {
+      "epoch": 2.45904,
+      "grad_norm": 1.7425981221159825,
+      "learning_rate": 1.6785323053315526e-05,
+      "loss": 1.0718,
+      "step": 15369
+    },
+    {
+      "epoch": 2.4592,
+      "grad_norm": 1.6516859083718736,
+      "learning_rate": 1.6775744994555464e-05,
+      "loss": 1.1565,
+      "step": 15370
+    },
+    {
+      "epoch": 2.45936,
+      "grad_norm": 1.9318820684771618,
+      "learning_rate": 1.676616941908591e-05,
+      "loss": 1.0836,
+      "step": 15371
+    },
+    {
+      "epoch": 2.45952,
+      "grad_norm": 1.975999377117749,
+      "learning_rate": 1.6756596327192565e-05,
+      "loss": 1.075,
+      "step": 15372
+    },
+    {
+      "epoch": 2.45968,
+      "grad_norm": 1.8179260198331362,
+      "learning_rate": 1.6747025719161092e-05,
+      "loss": 1.0486,
+      "step": 15373
+    },
+    {
+      "epoch": 2.45984,
+      "grad_norm": 1.9483217414611784,
+      "learning_rate": 1.673745759527705e-05,
+      "loss": 1.1153,
+      "step": 15374
+    },
+    {
+      "epoch": 2.46,
+      "grad_norm": 1.7914952119597847,
+      "learning_rate": 1.672789195582596e-05,
+      "loss": 1.0463,
+      "step": 15375
+    },
+    {
+      "epoch": 2.46016,
+      "grad_norm": 1.7040263317934046,
+      "learning_rate": 1.671832880109323e-05,
+      "loss": 1.107,
+      "step": 15376
+    },
+    {
+      "epoch": 2.46032,
+      "grad_norm": 1.7415870795531958,
+      "learning_rate": 1.6708768131364206e-05,
+      "loss": 1.1066,
+      "step": 15377
+    },
+    {
+      "epoch": 2.46048,
+      "grad_norm": 1.9537008592027119,
+      "learning_rate": 1.6699209946924155e-05,
+      "loss": 1.0999,
+      "step": 15378
+    },
+    {
+      "epoch": 2.46064,
+      "grad_norm": 1.9523757272594546,
+      "learning_rate": 1.668965424805835e-05,
+      "loss": 1.0557,
+      "step": 15379
+    },
+    {
+      "epoch": 2.4608,
+      "grad_norm": 1.796741985311621,
+      "learning_rate": 1.6680101035051808e-05,
+      "loss": 1.1072,
+      "step": 15380
+    },
+    {
+      "epoch": 2.46096,
+      "grad_norm": 1.746929430546287,
+      "learning_rate": 1.667055030818967e-05,
+      "loss": 1.0689,
+      "step": 15381
+    },
+    {
+      "epoch": 2.46112,
+      "grad_norm": 1.820805381918765,
+      "learning_rate": 1.6661002067756894e-05,
+      "loss": 1.111,
+      "step": 15382
+    },
+    {
+      "epoch": 2.46128,
+      "grad_norm": 1.8301787724834842,
+      "learning_rate": 1.665145631403837e-05,
+      "loss": 1.1109,
+      "step": 15383
+    },
+    {
+      "epoch": 2.46144,
+      "grad_norm": 1.9792628379839987,
+      "learning_rate": 1.6641913047318946e-05,
+      "loss": 1.1087,
+      "step": 15384
+    },
+    {
+      "epoch": 2.4616,
+      "grad_norm": 1.896119319838919,
+      "learning_rate": 1.6632372267883355e-05,
+      "loss": 1.1043,
+      "step": 15385
+    },
+    {
+      "epoch": 2.46176,
+      "grad_norm": 1.7411238437510776,
+      "learning_rate": 1.6622833976016315e-05,
+      "loss": 1.0605,
+      "step": 15386
+    },
+    {
+      "epoch": 2.46192,
+      "grad_norm": 1.451743849344359,
+      "learning_rate": 1.661329817200238e-05,
+      "loss": 1.0763,
+      "step": 15387
+    },
+    {
+      "epoch": 2.46208,
+      "grad_norm": 1.697534964499553,
+      "learning_rate": 1.6603764856126168e-05,
+      "loss": 1.0564,
+      "step": 15388
+    },
+    {
+      "epoch": 2.46224,
+      "grad_norm": 1.8984748294639866,
+      "learning_rate": 1.6594234028672053e-05,
+      "loss": 1.0851,
+      "step": 15389
+    },
+    {
+      "epoch": 2.4624,
+      "grad_norm": 1.8544113000986644,
+      "learning_rate": 1.6584705689924495e-05,
+      "loss": 1.0928,
+      "step": 15390
+    },
+    {
+      "epoch": 2.46256,
+      "grad_norm": 1.8396374632467825,
+      "learning_rate": 1.657517984016773e-05,
+      "loss": 1.0878,
+      "step": 15391
+    },
+    {
+      "epoch": 2.46272,
+      "grad_norm": 1.7547051288665338,
+      "learning_rate": 1.6565656479686087e-05,
+      "loss": 1.0639,
+      "step": 15392
+    },
+    {
+      "epoch": 2.46288,
+      "grad_norm": 1.8009506427243358,
+      "learning_rate": 1.6556135608763624e-05,
+      "loss": 1.0908,
+      "step": 15393
+    },
+    {
+      "epoch": 2.46304,
+      "grad_norm": 1.9082052116952446,
+      "learning_rate": 1.6546617227684514e-05,
+      "loss": 1.1017,
+      "step": 15394
+    },
+    {
+      "epoch": 2.4632,
+      "grad_norm": 1.7337895312949667,
+      "learning_rate": 1.653710133673273e-05,
+      "loss": 1.0729,
+      "step": 15395
+    },
+    {
+      "epoch": 2.4633599999999998,
+      "grad_norm": 1.9494400630992677,
+      "learning_rate": 1.6527587936192234e-05,
+      "loss": 1.0856,
+      "step": 15396
+    },
+    {
+      "epoch": 2.46352,
+      "grad_norm": 1.924956172224818,
+      "learning_rate": 1.651807702634688e-05,
+      "loss": 1.0911,
+      "step": 15397
+    },
+    {
+      "epoch": 2.46368,
+      "grad_norm": 1.934605613314632,
+      "learning_rate": 1.6508568607480458e-05,
+      "loss": 1.0808,
+      "step": 15398
+    },
+    {
+      "epoch": 2.46384,
+      "grad_norm": 1.8057275121977965,
+      "learning_rate": 1.6499062679876686e-05,
+      "loss": 1.0669,
+      "step": 15399
+    },
+    {
+      "epoch": 2.464,
+      "grad_norm": 1.773634704693597,
+      "learning_rate": 1.648955924381922e-05,
+      "loss": 1.1012,
+      "step": 15400
+    },
+    {
+      "epoch": 2.46416,
+      "grad_norm": 1.8039889757561172,
+      "learning_rate": 1.6480058299591617e-05,
+      "loss": 1.102,
+      "step": 15401
+    },
+    {
+      "epoch": 2.46432,
+      "grad_norm": 1.7031488404307744,
+      "learning_rate": 1.6470559847477364e-05,
+      "loss": 1.0708,
+      "step": 15402
+    },
+    {
+      "epoch": 2.46448,
+      "grad_norm": 1.7935734572213375,
+      "learning_rate": 1.6461063887759876e-05,
+      "loss": 1.0925,
+      "step": 15403
+    },
+    {
+      "epoch": 2.46464,
+      "grad_norm": 1.6448857839679452,
+      "learning_rate": 1.645157042072255e-05,
+      "loss": 1.139,
+      "step": 15404
+    },
+    {
+      "epoch": 2.4648,
+      "grad_norm": 1.7648882480059127,
+      "learning_rate": 1.6442079446648573e-05,
+      "loss": 1.0911,
+      "step": 15405
+    },
+    {
+      "epoch": 2.46496,
+      "grad_norm": 1.9349898443990876,
+      "learning_rate": 1.643259096582125e-05,
+      "loss": 1.1017,
+      "step": 15406
+    },
+    {
+      "epoch": 2.4651199999999998,
+      "grad_norm": 1.9659304728801574,
+      "learning_rate": 1.6423104978523584e-05,
+      "loss": 1.0905,
+      "step": 15407
+    },
+    {
+      "epoch": 2.46528,
+      "grad_norm": 1.6817756637137748,
+      "learning_rate": 1.6413621485038732e-05,
+      "loss": 1.0985,
+      "step": 15408
+    },
+    {
+      "epoch": 2.46544,
+      "grad_norm": 2.0326520964794526,
+      "learning_rate": 1.6404140485649566e-05,
+      "loss": 1.0344,
+      "step": 15409
+    },
+    {
+      "epoch": 2.4656000000000002,
+      "grad_norm": 1.725902145634253,
+      "learning_rate": 1.639466198063906e-05,
+      "loss": 1.0867,
+      "step": 15410
+    },
+    {
+      "epoch": 2.46576,
+      "grad_norm": 1.856148795069447,
+      "learning_rate": 1.6385185970290006e-05,
+      "loss": 1.0774,
+      "step": 15411
+    },
+    {
+      "epoch": 2.46592,
+      "grad_norm": 1.7547424776163332,
+      "learning_rate": 1.6375712454885173e-05,
+      "loss": 1.1056,
+      "step": 15412
+    },
+    {
+      "epoch": 2.46608,
+      "grad_norm": 1.7590036804236975,
+      "learning_rate": 1.636624143470723e-05,
+      "loss": 1.0587,
+      "step": 15413
+    },
+    {
+      "epoch": 2.46624,
+      "grad_norm": 2.00994082415617,
+      "learning_rate": 1.6356772910038777e-05,
+      "loss": 1.0888,
+      "step": 15414
+    },
+    {
+      "epoch": 2.4664,
+      "grad_norm": 2.21688067221646,
+      "learning_rate": 1.634730688116233e-05,
+      "loss": 1.0772,
+      "step": 15415
+    },
+    {
+      "epoch": 2.46656,
+      "grad_norm": 1.8247025893109194,
+      "learning_rate": 1.633784334836035e-05,
+      "loss": 1.1064,
+      "step": 15416
+    },
+    {
+      "epoch": 2.46672,
+      "grad_norm": 1.6353727713359367,
+      "learning_rate": 1.6328382311915226e-05,
+      "loss": 1.04,
+      "step": 15417
+    },
+    {
+      "epoch": 2.46688,
+      "grad_norm": 1.742106093121277,
+      "learning_rate": 1.631892377210924e-05,
+      "loss": 1.0955,
+      "step": 15418
+    },
+    {
+      "epoch": 2.46704,
+      "grad_norm": 1.9308498331565245,
+      "learning_rate": 1.6309467729224613e-05,
+      "loss": 1.0721,
+      "step": 15419
+    },
+    {
+      "epoch": 2.4672,
+      "grad_norm": 1.8940618190207816,
+      "learning_rate": 1.6300014183543578e-05,
+      "loss": 1.0734,
+      "step": 15420
+    },
+    {
+      "epoch": 2.46736,
+      "grad_norm": 1.7208539913409509,
+      "learning_rate": 1.629056313534809e-05,
+      "loss": 1.041,
+      "step": 15421
+    },
+    {
+      "epoch": 2.46752,
+      "grad_norm": 1.755109224944284,
+      "learning_rate": 1.6281114584920276e-05,
+      "loss": 1.0659,
+      "step": 15422
+    },
+    {
+      "epoch": 2.46768,
+      "grad_norm": 1.8630756796717325,
+      "learning_rate": 1.627166853254196e-05,
+      "loss": 1.1256,
+      "step": 15423
+    },
+    {
+      "epoch": 2.46784,
+      "grad_norm": 1.8966917541712196,
+      "learning_rate": 1.626222497849511e-05,
+      "loss": 1.0296,
+      "step": 15424
+    },
+    {
+      "epoch": 2.468,
+      "grad_norm": 2.0856508005092764,
+      "learning_rate": 1.6252783923061386e-05,
+      "loss": 1.0428,
+      "step": 15425
+    },
+    {
+      "epoch": 2.46816,
+      "grad_norm": 1.7292437772315759,
+      "learning_rate": 1.624334536652259e-05,
+      "loss": 1.11,
+      "step": 15426
+    },
+    {
+      "epoch": 2.46832,
+      "grad_norm": 1.7021452983362453,
+      "learning_rate": 1.6233909309160322e-05,
+      "loss": 1.109,
+      "step": 15427
+    },
+    {
+      "epoch": 2.46848,
+      "grad_norm": 2.2569028017918056,
+      "learning_rate": 1.6224475751256142e-05,
+      "loss": 1.0645,
+      "step": 15428
+    },
+    {
+      "epoch": 2.46864,
+      "grad_norm": 1.9232878668176434,
+      "learning_rate": 1.621504469309152e-05,
+      "loss": 1.1442,
+      "step": 15429
+    },
+    {
+      "epoch": 2.4688,
+      "grad_norm": 1.8123664565419184,
+      "learning_rate": 1.6205616134947886e-05,
+      "loss": 1.1074,
+      "step": 15430
+    },
+    {
+      "epoch": 2.46896,
+      "grad_norm": 2.039896463432742,
+      "learning_rate": 1.6196190077106566e-05,
+      "loss": 1.0353,
+      "step": 15431
+    },
+    {
+      "epoch": 2.46912,
+      "grad_norm": 1.923656979895947,
+      "learning_rate": 1.6186766519848805e-05,
+      "loss": 1.0633,
+      "step": 15432
+    },
+    {
+      "epoch": 2.46928,
+      "grad_norm": 1.9247059829056574,
+      "learning_rate": 1.6177345463455817e-05,
+      "loss": 1.0642,
+      "step": 15433
+    },
+    {
+      "epoch": 2.46944,
+      "grad_norm": 1.7440248715752287,
+      "learning_rate": 1.61679269082087e-05,
+      "loss": 1.085,
+      "step": 15434
+    },
+    {
+      "epoch": 2.4696,
+      "grad_norm": 1.90154704001976,
+      "learning_rate": 1.6158510854388455e-05,
+      "loss": 1.1017,
+      "step": 15435
+    },
+    {
+      "epoch": 2.46976,
+      "grad_norm": 2.0085802568016553,
+      "learning_rate": 1.6149097302276128e-05,
+      "loss": 1.0714,
+      "step": 15436
+    },
+    {
+      "epoch": 2.46992,
+      "grad_norm": 1.891563653762236,
+      "learning_rate": 1.61396862521525e-05,
+      "loss": 1.0449,
+      "step": 15437
+    },
+    {
+      "epoch": 2.47008,
+      "grad_norm": 1.7843363673662611,
+      "learning_rate": 1.613027770429849e-05,
+      "loss": 1.0832,
+      "step": 15438
+    },
+    {
+      "epoch": 2.47024,
+      "grad_norm": 1.9365325218367297,
+      "learning_rate": 1.612087165899473e-05,
+      "loss": 1.0601,
+      "step": 15439
+    },
+    {
+      "epoch": 2.4704,
+      "grad_norm": 1.5502164766587592,
+      "learning_rate": 1.6111468116521968e-05,
+      "loss": 1.1151,
+      "step": 15440
+    },
+    {
+      "epoch": 2.47056,
+      "grad_norm": 1.8560360999882295,
+      "learning_rate": 1.6102067077160754e-05,
+      "loss": 1.1113,
+      "step": 15441
+    },
+    {
+      "epoch": 2.47072,
+      "grad_norm": 1.8323242641458475,
+      "learning_rate": 1.60926685411916e-05,
+      "loss": 1.0761,
+      "step": 15442
+    },
+    {
+      "epoch": 2.47088,
+      "grad_norm": 1.6262668936839886,
+      "learning_rate": 1.6083272508894964e-05,
+      "loss": 1.0671,
+      "step": 15443
+    },
+    {
+      "epoch": 2.47104,
+      "grad_norm": 1.8565330736363446,
+      "learning_rate": 1.607387898055116e-05,
+      "loss": 1.0913,
+      "step": 15444
+    },
+    {
+      "epoch": 2.4712,
+      "grad_norm": 1.8693044099632286,
+      "learning_rate": 1.6064487956440567e-05,
+      "loss": 1.0822,
+      "step": 15445
+    },
+    {
+      "epoch": 2.47136,
+      "grad_norm": 1.671985936147801,
+      "learning_rate": 1.6055099436843303e-05,
+      "loss": 1.0578,
+      "step": 15446
+    },
+    {
+      "epoch": 2.47152,
+      "grad_norm": 1.6873806334154933,
+      "learning_rate": 1.6045713422039587e-05,
+      "loss": 1.0476,
+      "step": 15447
+    },
+    {
+      "epoch": 2.47168,
+      "grad_norm": 1.6998778851903684,
+      "learning_rate": 1.6036329912309413e-05,
+      "loss": 1.1193,
+      "step": 15448
+    },
+    {
+      "epoch": 2.47184,
+      "grad_norm": 1.6833301590697607,
+      "learning_rate": 1.602694890793285e-05,
+      "loss": 1.0803,
+      "step": 15449
+    },
+    {
+      "epoch": 2.472,
+      "grad_norm": 2.2090957266606592,
+      "learning_rate": 1.601757040918972e-05,
+      "loss": 1.1839,
+      "step": 15450
+    },
+    {
+      "epoch": 2.47216,
+      "grad_norm": 1.717615755258501,
+      "learning_rate": 1.6008194416359933e-05,
+      "loss": 1.0683,
+      "step": 15451
+    },
+    {
+      "epoch": 2.47232,
+      "grad_norm": 1.9887496943030936,
+      "learning_rate": 1.5998820929723246e-05,
+      "loss": 1.0789,
+      "step": 15452
+    },
+    {
+      "epoch": 2.47248,
+      "grad_norm": 1.7189549869452678,
+      "learning_rate": 1.5989449949559334e-05,
+      "loss": 1.1,
+      "step": 15453
+    },
+    {
+      "epoch": 2.47264,
+      "grad_norm": 1.7239152161726181,
+      "learning_rate": 1.5980081476147822e-05,
+      "loss": 1.0477,
+      "step": 15454
+    },
+    {
+      "epoch": 2.4728,
+      "grad_norm": 1.840247227092148,
+      "learning_rate": 1.5970715509768253e-05,
+      "loss": 1.1381,
+      "step": 15455
+    },
+    {
+      "epoch": 2.47296,
+      "grad_norm": 1.7729706064082706,
+      "learning_rate": 1.5961352050700083e-05,
+      "loss": 1.0874,
+      "step": 15456
+    },
+    {
+      "epoch": 2.4731199999999998,
+      "grad_norm": 1.6921195070371362,
+      "learning_rate": 1.5951991099222708e-05,
+      "loss": 1.1168,
+      "step": 15457
+    },
+    {
+      "epoch": 2.47328,
+      "grad_norm": 1.9680935586866712,
+      "learning_rate": 1.5942632655615454e-05,
+      "loss": 1.089,
+      "step": 15458
+    },
+    {
+      "epoch": 2.47344,
+      "grad_norm": 1.671463393673972,
+      "learning_rate": 1.5933276720157552e-05,
+      "loss": 1.079,
+      "step": 15459
+    },
+    {
+      "epoch": 2.4736000000000002,
+      "grad_norm": 1.8558498411736881,
+      "learning_rate": 1.592392329312815e-05,
+      "loss": 1.1521,
+      "step": 15460
+    },
+    {
+      "epoch": 2.47376,
+      "grad_norm": 2.093833445145665,
+      "learning_rate": 1.5914572374806413e-05,
+      "loss": 1.1116,
+      "step": 15461
+    },
+    {
+      "epoch": 2.47392,
+      "grad_norm": 1.892390278259856,
+      "learning_rate": 1.590522396547127e-05,
+      "loss": 1.1271,
+      "step": 15462
+    },
+    {
+      "epoch": 2.47408,
+      "grad_norm": 1.7735564571906373,
+      "learning_rate": 1.589587806540175e-05,
+      "loss": 1.0832,
+      "step": 15463
+    },
+    {
+      "epoch": 2.47424,
+      "grad_norm": 1.7208480401324624,
+      "learning_rate": 1.588653467487663e-05,
+      "loss": 1.1307,
+      "step": 15464
+    },
+    {
+      "epoch": 2.4744,
+      "grad_norm": 1.5019814752282947,
+      "learning_rate": 1.5877193794174793e-05,
+      "loss": 1.1075,
+      "step": 15465
+    },
+    {
+      "epoch": 2.47456,
+      "grad_norm": 1.6785783119090336,
+      "learning_rate": 1.5867855423574863e-05,
+      "loss": 1.0426,
+      "step": 15466
+    },
+    {
+      "epoch": 2.47472,
+      "grad_norm": 1.7065210816476488,
+      "learning_rate": 1.5858519563355555e-05,
+      "loss": 1.1179,
+      "step": 15467
+    },
+    {
+      "epoch": 2.47488,
+      "grad_norm": 1.8501912698742586,
+      "learning_rate": 1.584918621379543e-05,
+      "loss": 1.102,
+      "step": 15468
+    },
+    {
+      "epoch": 2.47504,
+      "grad_norm": 1.6041729182913214,
+      "learning_rate": 1.5839855375172953e-05,
+      "loss": 1.0795,
+      "step": 15469
+    },
+    {
+      "epoch": 2.4752,
+      "grad_norm": 1.6960343291604765,
+      "learning_rate": 1.5830527047766564e-05,
+      "loss": 1.1219,
+      "step": 15470
+    },
+    {
+      "epoch": 2.4753600000000002,
+      "grad_norm": 1.900258842210827,
+      "learning_rate": 1.5821201231854598e-05,
+      "loss": 1.0591,
+      "step": 15471
+    },
+    {
+      "epoch": 2.47552,
+      "grad_norm": 1.8445637911578312,
+      "learning_rate": 1.5811877927715337e-05,
+      "loss": 1.0517,
+      "step": 15472
+    },
+    {
+      "epoch": 2.47568,
+      "grad_norm": 1.8752312089953795,
+      "learning_rate": 1.580255713562695e-05,
+      "loss": 1.1026,
+      "step": 15473
+    },
+    {
+      "epoch": 2.47584,
+      "grad_norm": 1.868083671607334,
+      "learning_rate": 1.5793238855867576e-05,
+      "loss": 1.0809,
+      "step": 15474
+    },
+    {
+      "epoch": 2.476,
+      "grad_norm": 2.0062928015079984,
+      "learning_rate": 1.5783923088715245e-05,
+      "loss": 1.1125,
+      "step": 15475
+    },
+    {
+      "epoch": 2.47616,
+      "grad_norm": 1.5909795810999767,
+      "learning_rate": 1.577460983444792e-05,
+      "loss": 1.0892,
+      "step": 15476
+    },
+    {
+      "epoch": 2.47632,
+      "grad_norm": 1.8941427840239893,
+      "learning_rate": 1.5765299093343543e-05,
+      "loss": 1.0662,
+      "step": 15477
+    },
+    {
+      "epoch": 2.47648,
+      "grad_norm": 1.7649122002088062,
+      "learning_rate": 1.5755990865679858e-05,
+      "loss": 1.0375,
+      "step": 15478
+    },
+    {
+      "epoch": 2.47664,
+      "grad_norm": 1.737397696131996,
+      "learning_rate": 1.5746685151734696e-05,
+      "loss": 1.0622,
+      "step": 15479
+    },
+    {
+      "epoch": 2.4768,
+      "grad_norm": 1.7818989922585742,
+      "learning_rate": 1.5737381951785635e-05,
+      "loss": 1.0834,
+      "step": 15480
+    },
+    {
+      "epoch": 2.47696,
+      "grad_norm": 1.8536098855981307,
+      "learning_rate": 1.572808126611035e-05,
+      "loss": 1.1018,
+      "step": 15481
+    },
+    {
+      "epoch": 2.47712,
+      "grad_norm": 1.7053435412017044,
+      "learning_rate": 1.5718783094986288e-05,
+      "loss": 1.0548,
+      "step": 15482
+    },
+    {
+      "epoch": 2.47728,
+      "grad_norm": 1.8040096059353947,
+      "learning_rate": 1.570948743869094e-05,
+      "loss": 1.1222,
+      "step": 15483
+    },
+    {
+      "epoch": 2.47744,
+      "grad_norm": 1.698243703570369,
+      "learning_rate": 1.570019429750167e-05,
+      "loss": 1.1026,
+      "step": 15484
+    },
+    {
+      "epoch": 2.4776,
+      "grad_norm": 1.9391954039610584,
+      "learning_rate": 1.5690903671695766e-05,
+      "loss": 1.1474,
+      "step": 15485
+    },
+    {
+      "epoch": 2.47776,
+      "grad_norm": 1.772784195660152,
+      "learning_rate": 1.5681615561550443e-05,
+      "loss": 1.057,
+      "step": 15486
+    },
+    {
+      "epoch": 2.47792,
+      "grad_norm": 1.5003875178439714,
+      "learning_rate": 1.5672329967342846e-05,
+      "loss": 1.1022,
+      "step": 15487
+    },
+    {
+      "epoch": 2.47808,
+      "grad_norm": 1.8486240809341965,
+      "learning_rate": 1.566304688935004e-05,
+      "loss": 1.0958,
+      "step": 15488
+    },
+    {
+      "epoch": 2.47824,
+      "grad_norm": 1.6303179398658554,
+      "learning_rate": 1.5653766327849028e-05,
+      "loss": 1.1073,
+      "step": 15489
+    },
+    {
+      "epoch": 2.4784,
+      "grad_norm": 2.0593190024731065,
+      "learning_rate": 1.564448828311672e-05,
+      "loss": 1.083,
+      "step": 15490
+    },
+    {
+      "epoch": 2.47856,
+      "grad_norm": 1.863367986281599,
+      "learning_rate": 1.563521275542996e-05,
+      "loss": 1.073,
+      "step": 15491
+    },
+    {
+      "epoch": 2.47872,
+      "grad_norm": 1.5645461217596153,
+      "learning_rate": 1.562593974506551e-05,
+      "loss": 1.0747,
+      "step": 15492
+    },
+    {
+      "epoch": 2.47888,
+      "grad_norm": 1.8064729406291669,
+      "learning_rate": 1.5616669252300108e-05,
+      "loss": 1.0526,
+      "step": 15493
+    },
+    {
+      "epoch": 2.47904,
+      "grad_norm": 1.9468179657079947,
+      "learning_rate": 1.5607401277410295e-05,
+      "loss": 1.0389,
+      "step": 15494
+    },
+    {
+      "epoch": 2.4792,
+      "grad_norm": 2.8741904224941606,
+      "learning_rate": 1.5598135820672687e-05,
+      "loss": 1.1033,
+      "step": 15495
+    },
+    {
+      "epoch": 2.47936,
+      "grad_norm": 1.8297755718530522,
+      "learning_rate": 1.5588872882363694e-05,
+      "loss": 1.0837,
+      "step": 15496
+    },
+    {
+      "epoch": 2.47952,
+      "grad_norm": 1.6742313705816467,
+      "learning_rate": 1.5579612462759763e-05,
+      "loss": 1.0483,
+      "step": 15497
+    },
+    {
+      "epoch": 2.47968,
+      "grad_norm": 1.6558980743162512,
+      "learning_rate": 1.5570354562137146e-05,
+      "loss": 1.0848,
+      "step": 15498
+    },
+    {
+      "epoch": 2.47984,
+      "grad_norm": 1.7929702830171732,
+      "learning_rate": 1.556109918077214e-05,
+      "loss": 1.0923,
+      "step": 15499
+    },
+    {
+      "epoch": 2.48,
+      "grad_norm": 1.766466282239659,
+      "learning_rate": 1.55518463189409e-05,
+      "loss": 1.0379,
+      "step": 15500
+    },
+    {
+      "epoch": 2.48016,
+      "grad_norm": 1.6658302894079853,
+      "learning_rate": 1.5542595976919506e-05,
+      "loss": 1.0834,
+      "step": 15501
+    },
+    {
+      "epoch": 2.48032,
+      "grad_norm": 1.7328032937680835,
+      "learning_rate": 1.5533348154983974e-05,
+      "loss": 1.0607,
+      "step": 15502
+    },
+    {
+      "epoch": 2.48048,
+      "grad_norm": 1.849320257402432,
+      "learning_rate": 1.552410285341024e-05,
+      "loss": 1.0959,
+      "step": 15503
+    },
+    {
+      "epoch": 2.48064,
+      "grad_norm": 1.8791689619838152,
+      "learning_rate": 1.5514860072474223e-05,
+      "loss": 1.1363,
+      "step": 15504
+    },
+    {
+      "epoch": 2.4808,
+      "grad_norm": 1.7112635632200195,
+      "learning_rate": 1.5505619812451633e-05,
+      "loss": 1.0747,
+      "step": 15505
+    },
+    {
+      "epoch": 2.48096,
+      "grad_norm": 1.812567511480263,
+      "learning_rate": 1.5496382073618277e-05,
+      "loss": 1.0846,
+      "step": 15506
+    },
+    {
+      "epoch": 2.4811199999999998,
+      "grad_norm": 1.7110695166093612,
+      "learning_rate": 1.548714685624969e-05,
+      "loss": 1.0885,
+      "step": 15507
+    },
+    {
+      "epoch": 2.48128,
+      "grad_norm": 1.7314384844601496,
+      "learning_rate": 1.547791416062151e-05,
+      "loss": 1.1035,
+      "step": 15508
+    },
+    {
+      "epoch": 2.48144,
+      "grad_norm": 2.2900302085433335,
+      "learning_rate": 1.5468683987009224e-05,
+      "loss": 1.0795,
+      "step": 15509
+    },
+    {
+      "epoch": 2.4816,
+      "grad_norm": 1.808826694176807,
+      "learning_rate": 1.5459456335688227e-05,
+      "loss": 1.0736,
+      "step": 15510
+    },
+    {
+      "epoch": 2.48176,
+      "grad_norm": 1.6998452214670061,
+      "learning_rate": 1.5450231206933864e-05,
+      "loss": 1.0767,
+      "step": 15511
+    },
+    {
+      "epoch": 2.48192,
+      "grad_norm": 1.8510916620355002,
+      "learning_rate": 1.54410086010214e-05,
+      "loss": 1.1219,
+      "step": 15512
+    },
+    {
+      "epoch": 2.48208,
+      "grad_norm": 1.57589671192044,
+      "learning_rate": 1.543178851822602e-05,
+      "loss": 1.0868,
+      "step": 15513
+    },
+    {
+      "epoch": 2.48224,
+      "grad_norm": 1.8690131773761378,
+      "learning_rate": 1.5422570958822846e-05,
+      "loss": 1.0275,
+      "step": 15514
+    },
+    {
+      "epoch": 2.4824,
+      "grad_norm": 2.0489296775403623,
+      "learning_rate": 1.5413355923086915e-05,
+      "loss": 1.1075,
+      "step": 15515
+    },
+    {
+      "epoch": 2.48256,
+      "grad_norm": 1.9195092521486061,
+      "learning_rate": 1.5404143411293183e-05,
+      "loss": 1.1276,
+      "step": 15516
+    },
+    {
+      "epoch": 2.48272,
+      "grad_norm": 1.6980821571696378,
+      "learning_rate": 1.539493342371654e-05,
+      "loss": 1.0863,
+      "step": 15517
+    },
+    {
+      "epoch": 2.4828799999999998,
+      "grad_norm": 1.5885480673447607,
+      "learning_rate": 1.53857259606318e-05,
+      "loss": 1.1015,
+      "step": 15518
+    },
+    {
+      "epoch": 2.48304,
+      "grad_norm": 1.7025505575772883,
+      "learning_rate": 1.5376521022313682e-05,
+      "loss": 1.1052,
+      "step": 15519
+    },
+    {
+      "epoch": 2.4832,
+      "grad_norm": 1.7670228541803805,
+      "learning_rate": 1.5367318609036917e-05,
+      "loss": 1.1002,
+      "step": 15520
+    },
+    {
+      "epoch": 2.4833600000000002,
+      "grad_norm": 1.7298845454192169,
+      "learning_rate": 1.5358118721075987e-05,
+      "loss": 1.083,
+      "step": 15521
+    },
+    {
+      "epoch": 2.48352,
+      "grad_norm": 1.6694684504926847,
+      "learning_rate": 1.5348921358705516e-05,
+      "loss": 1.0706,
+      "step": 15522
+    },
+    {
+      "epoch": 2.48368,
+      "grad_norm": 1.7614553639833717,
+      "learning_rate": 1.5339726522199826e-05,
+      "loss": 1.1022,
+      "step": 15523
+    },
+    {
+      "epoch": 2.48384,
+      "grad_norm": 1.7348350157439514,
+      "learning_rate": 1.533053421183336e-05,
+      "loss": 1.0762,
+      "step": 15524
+    },
+    {
+      "epoch": 2.484,
+      "grad_norm": 1.8687029642635715,
+      "learning_rate": 1.532134442788038e-05,
+      "loss": 1.0922,
+      "step": 15525
+    },
+    {
+      "epoch": 2.48416,
+      "grad_norm": 1.7108834220871947,
+      "learning_rate": 1.5312157170615083e-05,
+      "loss": 1.0704,
+      "step": 15526
+    },
+    {
+      "epoch": 2.48432,
+      "grad_norm": 1.591412070587379,
+      "learning_rate": 1.5302972440311625e-05,
+      "loss": 1.059,
+      "step": 15527
+    },
+    {
+      "epoch": 2.48448,
+      "grad_norm": 1.927441113493386,
+      "learning_rate": 1.5293790237244044e-05,
+      "loss": 1.0772,
+      "step": 15528
+    },
+    {
+      "epoch": 2.48464,
+      "grad_norm": 1.632708893542301,
+      "learning_rate": 1.528461056168633e-05,
+      "loss": 1.104,
+      "step": 15529
+    },
+    {
+      "epoch": 2.4848,
+      "grad_norm": 1.8690491141372914,
+      "learning_rate": 1.52754334139124e-05,
+      "loss": 1.0986,
+      "step": 15530
+    },
+    {
+      "epoch": 2.48496,
+      "grad_norm": 1.8073393410157181,
+      "learning_rate": 1.5266258794196087e-05,
+      "loss": 1.0956,
+      "step": 15531
+    },
+    {
+      "epoch": 2.48512,
+      "grad_norm": 1.8107471699672986,
+      "learning_rate": 1.525708670281113e-05,
+      "loss": 1.0882,
+      "step": 15532
+    },
+    {
+      "epoch": 2.48528,
+      "grad_norm": 1.610339100255538,
+      "learning_rate": 1.5247917140031199e-05,
+      "loss": 1.0885,
+      "step": 15533
+    },
+    {
+      "epoch": 2.48544,
+      "grad_norm": 1.7798844803919305,
+      "learning_rate": 1.5238750106129973e-05,
+      "loss": 1.1061,
+      "step": 15534
+    },
+    {
+      "epoch": 2.4856,
+      "grad_norm": 2.0424760010613667,
+      "learning_rate": 1.5229585601380892e-05,
+      "loss": 1.0975,
+      "step": 15535
+    },
+    {
+      "epoch": 2.48576,
+      "grad_norm": 1.70860510469752,
+      "learning_rate": 1.522042362605749e-05,
+      "loss": 1.0856,
+      "step": 15536
+    },
+    {
+      "epoch": 2.48592,
+      "grad_norm": 1.8014289774270837,
+      "learning_rate": 1.521126418043307e-05,
+      "loss": 1.0672,
+      "step": 15537
+    },
+    {
+      "epoch": 2.48608,
+      "grad_norm": 2.050008471323017,
+      "learning_rate": 1.5202107264781018e-05,
+      "loss": 1.1064,
+      "step": 15538
+    },
+    {
+      "epoch": 2.48624,
+      "grad_norm": 1.7098409430806547,
+      "learning_rate": 1.5192952879374478e-05,
+      "loss": 1.0832,
+      "step": 15539
+    },
+    {
+      "epoch": 2.4864,
+      "grad_norm": 1.6147668602748333,
+      "learning_rate": 1.5183801024486677e-05,
+      "loss": 1.0106,
+      "step": 15540
+    },
+    {
+      "epoch": 2.48656,
+      "grad_norm": 1.6675277538318392,
+      "learning_rate": 1.5174651700390651e-05,
+      "loss": 1.052,
+      "step": 15541
+    },
+    {
+      "epoch": 2.48672,
+      "grad_norm": 1.8689616056868932,
+      "learning_rate": 1.5165504907359428e-05,
+      "loss": 1.1025,
+      "step": 15542
+    },
+    {
+      "epoch": 2.48688,
+      "grad_norm": 1.6013286584592572,
+      "learning_rate": 1.5156360645665912e-05,
+      "loss": 1.0875,
+      "step": 15543
+    },
+    {
+      "epoch": 2.48704,
+      "grad_norm": 1.7157666970638386,
+      "learning_rate": 1.5147218915582972e-05,
+      "loss": 1.1075,
+      "step": 15544
+    },
+    {
+      "epoch": 2.4872,
+      "grad_norm": 1.7195316890669576,
+      "learning_rate": 1.513807971738337e-05,
+      "loss": 1.1049,
+      "step": 15545
+    },
+    {
+      "epoch": 2.48736,
+      "grad_norm": 1.6092636446255015,
+      "learning_rate": 1.5128943051339817e-05,
+      "loss": 1.0911,
+      "step": 15546
+    },
+    {
+      "epoch": 2.48752,
+      "grad_norm": 1.8861033272562406,
+      "learning_rate": 1.511980891772492e-05,
+      "loss": 1.0988,
+      "step": 15547
+    },
+    {
+      "epoch": 2.48768,
+      "grad_norm": 1.7344781652051593,
+      "learning_rate": 1.511067731681125e-05,
+      "loss": 1.1132,
+      "step": 15548
+    },
+    {
+      "epoch": 2.48784,
+      "grad_norm": 1.81180889701364,
+      "learning_rate": 1.5101548248871255e-05,
+      "loss": 1.0661,
+      "step": 15549
+    },
+    {
+      "epoch": 2.488,
+      "grad_norm": 2.021447305854864,
+      "learning_rate": 1.5092421714177395e-05,
+      "loss": 1.094,
+      "step": 15550
+    },
+    {
+      "epoch": 2.48816,
+      "grad_norm": 2.29722136918485,
+      "learning_rate": 1.5083297713001898e-05,
+      "loss": 1.0873,
+      "step": 15551
+    },
+    {
+      "epoch": 2.48832,
+      "grad_norm": 1.71928272745789,
+      "learning_rate": 1.50741762456171e-05,
+      "loss": 1.0521,
+      "step": 15552
+    },
+    {
+      "epoch": 2.48848,
+      "grad_norm": 1.6713599145718372,
+      "learning_rate": 1.5065057312295095e-05,
+      "loss": 1.0842,
+      "step": 15553
+    },
+    {
+      "epoch": 2.48864,
+      "grad_norm": 2.030735276100062,
+      "learning_rate": 1.5055940913308065e-05,
+      "loss": 1.1133,
+      "step": 15554
+    },
+    {
+      "epoch": 2.4888,
+      "grad_norm": 1.9353351370030238,
+      "learning_rate": 1.5046827048927925e-05,
+      "loss": 1.0845,
+      "step": 15555
+    },
+    {
+      "epoch": 2.48896,
+      "grad_norm": 1.8773475820363406,
+      "learning_rate": 1.5037715719426703e-05,
+      "loss": 1.0792,
+      "step": 15556
+    },
+    {
+      "epoch": 2.4891199999999998,
+      "grad_norm": 1.741174668723876,
+      "learning_rate": 1.5028606925076249e-05,
+      "loss": 1.0817,
+      "step": 15557
+    },
+    {
+      "epoch": 2.48928,
+      "grad_norm": 2.264664757380566,
+      "learning_rate": 1.5019500666148334e-05,
+      "loss": 1.1746,
+      "step": 15558
+    },
+    {
+      "epoch": 2.48944,
+      "grad_norm": 1.636842002056724,
+      "learning_rate": 1.5010396942914705e-05,
+      "loss": 1.0628,
+      "step": 15559
+    },
+    {
+      "epoch": 2.4896,
+      "grad_norm": 1.7654336602308285,
+      "learning_rate": 1.5001295755646948e-05,
+      "loss": 1.0915,
+      "step": 15560
+    },
+    {
+      "epoch": 2.48976,
+      "grad_norm": 2.1580178516992676,
+      "learning_rate": 1.4992197104616735e-05,
+      "loss": 1.0563,
+      "step": 15561
+    },
+    {
+      "epoch": 2.48992,
+      "grad_norm": 1.7965591436502455,
+      "learning_rate": 1.4983100990095434e-05,
+      "loss": 1.0474,
+      "step": 15562
+    },
+    {
+      "epoch": 2.49008,
+      "grad_norm": 1.9822683517066828,
+      "learning_rate": 1.4974007412354563e-05,
+      "loss": 1.0329,
+      "step": 15563
+    },
+    {
+      "epoch": 2.49024,
+      "grad_norm": 1.9950792115985636,
+      "learning_rate": 1.4964916371665372e-05,
+      "loss": 1.1484,
+      "step": 15564
+    },
+    {
+      "epoch": 2.4904,
+      "grad_norm": 1.8058222648003797,
+      "learning_rate": 1.4955827868299189e-05,
+      "loss": 1.1049,
+      "step": 15565
+    },
+    {
+      "epoch": 2.49056,
+      "grad_norm": 1.6982916355851685,
+      "learning_rate": 1.4946741902527172e-05,
+      "loss": 1.1314,
+      "step": 15566
+    },
+    {
+      "epoch": 2.49072,
+      "grad_norm": 1.9177589428530855,
+      "learning_rate": 1.4937658474620453e-05,
+      "loss": 1.1112,
+      "step": 15567
+    },
+    {
+      "epoch": 2.4908799999999998,
+      "grad_norm": 1.5977361614170058,
+      "learning_rate": 1.4928577584850045e-05,
+      "loss": 1.0393,
+      "step": 15568
+    },
+    {
+      "epoch": 2.49104,
+      "grad_norm": 1.8717142968328355,
+      "learning_rate": 1.4919499233486923e-05,
+      "loss": 1.0839,
+      "step": 15569
+    },
+    {
+      "epoch": 2.4912,
+      "grad_norm": 1.8722091743920117,
+      "learning_rate": 1.4910423420801977e-05,
+      "loss": 1.0728,
+      "step": 15570
+    },
+    {
+      "epoch": 2.4913600000000002,
+      "grad_norm": 1.7692068127023748,
+      "learning_rate": 1.4901350147065995e-05,
+      "loss": 1.1024,
+      "step": 15571
+    },
+    {
+      "epoch": 2.49152,
+      "grad_norm": 1.7736822852493213,
+      "learning_rate": 1.4892279412549737e-05,
+      "loss": 1.1111,
+      "step": 15572
+    },
+    {
+      "epoch": 2.49168,
+      "grad_norm": 1.8127833625454888,
+      "learning_rate": 1.488321121752384e-05,
+      "loss": 1.0482,
+      "step": 15573
+    },
+    {
+      "epoch": 2.49184,
+      "grad_norm": 1.7404394961594587,
+      "learning_rate": 1.487414556225889e-05,
+      "loss": 1.1189,
+      "step": 15574
+    },
+    {
+      "epoch": 2.492,
+      "grad_norm": 1.6912572479564527,
+      "learning_rate": 1.4865082447025392e-05,
+      "loss": 1.0163,
+      "step": 15575
+    },
+    {
+      "epoch": 2.49216,
+      "grad_norm": 1.6529213403136107,
+      "learning_rate": 1.4856021872093761e-05,
+      "loss": 1.1277,
+      "step": 15576
+    },
+    {
+      "epoch": 2.49232,
+      "grad_norm": 1.6239209923473796,
+      "learning_rate": 1.4846963837734419e-05,
+      "loss": 1.0742,
+      "step": 15577
+    },
+    {
+      "epoch": 2.49248,
+      "grad_norm": 1.8455145729120002,
+      "learning_rate": 1.4837908344217533e-05,
+      "loss": 1.0538,
+      "step": 15578
+    },
+    {
+      "epoch": 2.49264,
+      "grad_norm": 1.7622454163883303,
+      "learning_rate": 1.4828855391813435e-05,
+      "loss": 1.0421,
+      "step": 15579
+    },
+    {
+      "epoch": 2.4928,
+      "grad_norm": 1.5994823460355398,
+      "learning_rate": 1.4819804980792119e-05,
+      "loss": 1.0829,
+      "step": 15580
+    },
+    {
+      "epoch": 2.49296,
+      "grad_norm": 1.7157512194494242,
+      "learning_rate": 1.4810757111423735e-05,
+      "loss": 1.0583,
+      "step": 15581
+    },
+    {
+      "epoch": 2.4931200000000002,
+      "grad_norm": 1.7709026109740826,
+      "learning_rate": 1.4801711783978223e-05,
+      "loss": 1.072,
+      "step": 15582
+    },
+    {
+      "epoch": 2.49328,
+      "grad_norm": 1.6825747055563416,
+      "learning_rate": 1.479266899872549e-05,
+      "loss": 1.0744,
+      "step": 15583
+    },
+    {
+      "epoch": 2.49344,
+      "grad_norm": 2.0167707063526863,
+      "learning_rate": 1.4783628755935352e-05,
+      "loss": 1.1272,
+      "step": 15584
+    },
+    {
+      "epoch": 2.4936,
+      "grad_norm": 1.7230238015959818,
+      "learning_rate": 1.4774591055877552e-05,
+      "loss": 1.0889,
+      "step": 15585
+    },
+    {
+      "epoch": 2.49376,
+      "grad_norm": 1.644022049245362,
+      "learning_rate": 1.476555589882177e-05,
+      "loss": 1.1166,
+      "step": 15586
+    },
+    {
+      "epoch": 2.49392,
+      "grad_norm": 1.8365300682870198,
+      "learning_rate": 1.4756523285037605e-05,
+      "loss": 1.0911,
+      "step": 15587
+    },
+    {
+      "epoch": 2.49408,
+      "grad_norm": 1.6642841601752598,
+      "learning_rate": 1.4747493214794572e-05,
+      "loss": 1.0734,
+      "step": 15588
+    },
+    {
+      "epoch": 2.49424,
+      "grad_norm": 1.7949140040807627,
+      "learning_rate": 1.4738465688362113e-05,
+      "loss": 1.0769,
+      "step": 15589
+    },
+    {
+      "epoch": 2.4944,
+      "grad_norm": 1.7900072569529892,
+      "learning_rate": 1.4729440706009601e-05,
+      "loss": 1.0829,
+      "step": 15590
+    },
+    {
+      "epoch": 2.49456,
+      "grad_norm": 2.086248930107601,
+      "learning_rate": 1.472041826800632e-05,
+      "loss": 1.1193,
+      "step": 15591
+    },
+    {
+      "epoch": 2.49472,
+      "grad_norm": 1.9477844906040036,
+      "learning_rate": 1.4711398374621477e-05,
+      "loss": 1.1002,
+      "step": 15592
+    },
+    {
+      "epoch": 2.49488,
+      "grad_norm": 1.9940048145927727,
+      "learning_rate": 1.4702381026124268e-05,
+      "loss": 1.1024,
+      "step": 15593
+    },
+    {
+      "epoch": 2.49504,
+      "grad_norm": 1.782533004776946,
+      "learning_rate": 1.4693366222783678e-05,
+      "loss": 1.075,
+      "step": 15594
+    },
+    {
+      "epoch": 2.4952,
+      "grad_norm": 1.7863457845685213,
+      "learning_rate": 1.4684353964868768e-05,
+      "loss": 1.0673,
+      "step": 15595
+    },
+    {
+      "epoch": 2.49536,
+      "grad_norm": 1.9466685501988328,
+      "learning_rate": 1.4675344252648382e-05,
+      "loss": 1.1018,
+      "step": 15596
+    },
+    {
+      "epoch": 2.49552,
+      "grad_norm": 1.752013439044761,
+      "learning_rate": 1.4666337086391413e-05,
+      "loss": 1.1021,
+      "step": 15597
+    },
+    {
+      "epoch": 2.49568,
+      "grad_norm": 1.8432996163632818,
+      "learning_rate": 1.4657332466366602e-05,
+      "loss": 1.073,
+      "step": 15598
+    },
+    {
+      "epoch": 2.49584,
+      "grad_norm": 1.9241691673514714,
+      "learning_rate": 1.4648330392842634e-05,
+      "loss": 1.0786,
+      "step": 15599
+    },
+    {
+      "epoch": 2.496,
+      "grad_norm": 1.9494191443259106,
+      "learning_rate": 1.4639330866088118e-05,
+      "loss": 1.1193,
+      "step": 15600
+    },
+    {
+      "epoch": 2.49616,
+      "grad_norm": 1.964353980663105,
+      "learning_rate": 1.4630333886371583e-05,
+      "loss": 1.0918,
+      "step": 15601
+    },
+    {
+      "epoch": 2.49632,
+      "grad_norm": 1.8305968417757241,
+      "learning_rate": 1.462133945396148e-05,
+      "loss": 1.1028,
+      "step": 15602
+    },
+    {
+      "epoch": 2.49648,
+      "grad_norm": 1.6814850735489832,
+      "learning_rate": 1.4612347569126205e-05,
+      "loss": 1.0374,
+      "step": 15603
+    },
+    {
+      "epoch": 2.49664,
+      "grad_norm": 1.8424581377392468,
+      "learning_rate": 1.4603358232134056e-05,
+      "loss": 1.1414,
+      "step": 15604
+    },
+    {
+      "epoch": 2.4968,
+      "grad_norm": 1.736657089602322,
+      "learning_rate": 1.4594371443253262e-05,
+      "loss": 1.0239,
+      "step": 15605
+    },
+    {
+      "epoch": 2.49696,
+      "grad_norm": 1.870959119810751,
+      "learning_rate": 1.4585387202751976e-05,
+      "loss": 1.0888,
+      "step": 15606
+    },
+    {
+      "epoch": 2.49712,
+      "grad_norm": 1.8998976239937693,
+      "learning_rate": 1.4576405510898271e-05,
+      "loss": 1.0566,
+      "step": 15607
+    },
+    {
+      "epoch": 2.49728,
+      "grad_norm": 1.9711504328870726,
+      "learning_rate": 1.4567426367960125e-05,
+      "loss": 1.1372,
+      "step": 15608
+    },
+    {
+      "epoch": 2.49744,
+      "grad_norm": 1.930914983851734,
+      "learning_rate": 1.455844977420554e-05,
+      "loss": 1.133,
+      "step": 15609
+    },
+    {
+      "epoch": 2.4976,
+      "grad_norm": 1.6627095995445413,
+      "learning_rate": 1.4549475729902262e-05,
+      "loss": 1.0582,
+      "step": 15610
+    },
+    {
+      "epoch": 2.49776,
+      "grad_norm": 1.9883127616808525,
+      "learning_rate": 1.4540504235318164e-05,
+      "loss": 1.0714,
+      "step": 15611
+    },
+    {
+      "epoch": 2.49792,
+      "grad_norm": 1.7043270850629817,
+      "learning_rate": 1.4531535290720844e-05,
+      "loss": 1.0674,
+      "step": 15612
+    },
+    {
+      "epoch": 2.49808,
+      "grad_norm": 2.046825869073315,
+      "learning_rate": 1.4522568896377986e-05,
+      "loss": 1.0762,
+      "step": 15613
+    },
+    {
+      "epoch": 2.49824,
+      "grad_norm": 2.155712658999197,
+      "learning_rate": 1.4513605052557122e-05,
+      "loss": 1.0619,
+      "step": 15614
+    },
+    {
+      "epoch": 2.4984,
+      "grad_norm": 1.6635186185837234,
+      "learning_rate": 1.4504643759525727e-05,
+      "loss": 1.0798,
+      "step": 15615
+    },
+    {
+      "epoch": 2.49856,
+      "grad_norm": 1.8392006136838968,
+      "learning_rate": 1.4495685017551175e-05,
+      "loss": 1.084,
+      "step": 15616
+    },
+    {
+      "epoch": 2.49872,
+      "grad_norm": 1.6205004552888198,
+      "learning_rate": 1.4486728826900764e-05,
+      "loss": 1.1207,
+      "step": 15617
+    },
+    {
+      "epoch": 2.4988799999999998,
+      "grad_norm": 1.7618403086945347,
+      "learning_rate": 1.4477775187841803e-05,
+      "loss": 1.0783,
+      "step": 15618
+    },
+    {
+      "epoch": 2.49904,
+      "grad_norm": 1.6875305266497325,
+      "learning_rate": 1.4468824100641365e-05,
+      "loss": 1.1112,
+      "step": 15619
+    },
+    {
+      "epoch": 2.4992,
+      "grad_norm": 2.065768562762801,
+      "learning_rate": 1.4459875565566638e-05,
+      "loss": 1.0865,
+      "step": 15620
+    },
+    {
+      "epoch": 2.4993600000000002,
+      "grad_norm": 1.7534041616763771,
+      "learning_rate": 1.445092958288452e-05,
+      "loss": 1.0879,
+      "step": 15621
+    },
+    {
+      "epoch": 2.49952,
+      "grad_norm": 1.744336090483246,
+      "learning_rate": 1.4441986152862042e-05,
+      "loss": 1.0734,
+      "step": 15622
+    },
+    {
+      "epoch": 2.49968,
+      "grad_norm": 1.711811478251657,
+      "learning_rate": 1.4433045275766022e-05,
+      "loss": 1.077,
+      "step": 15623
+    },
+    {
+      "epoch": 2.49984,
+      "grad_norm": 1.8799713780159095,
+      "learning_rate": 1.4424106951863237e-05,
+      "loss": 1.0851,
+      "step": 15624
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 1.8598867319346633,
+      "learning_rate": 1.4415171181420416e-05,
+      "loss": 1.0949,
+      "step": 15625
+    },
+    {
+      "epoch": 2.50016,
+      "grad_norm": 1.7168855744707219,
+      "learning_rate": 1.4406237964704173e-05,
+      "loss": 1.085,
+      "step": 15626
+    },
+    {
+      "epoch": 2.50032,
+      "grad_norm": 2.02956822774489,
+      "learning_rate": 1.4397307301981067e-05,
+      "loss": 1.1041,
+      "step": 15627
+    },
+    {
+      "epoch": 2.50048,
+      "grad_norm": 1.7562055701221988,
+      "learning_rate": 1.4388379193517565e-05,
+      "loss": 1.1163,
+      "step": 15628
+    },
+    {
+      "epoch": 2.5006399999999998,
+      "grad_norm": 1.6632818611791682,
+      "learning_rate": 1.4379453639580087e-05,
+      "loss": 1.0974,
+      "step": 15629
+    },
+    {
+      "epoch": 2.5008,
+      "grad_norm": 1.762920934782491,
+      "learning_rate": 1.4370530640434953e-05,
+      "loss": 1.0778,
+      "step": 15630
+    },
+    {
+      "epoch": 2.50096,
+      "grad_norm": 1.7608327438172373,
+      "learning_rate": 1.4361610196348396e-05,
+      "loss": 1.1077,
+      "step": 15631
+    },
+    {
+      "epoch": 2.5011200000000002,
+      "grad_norm": 1.7053072764970807,
+      "learning_rate": 1.4352692307586613e-05,
+      "loss": 1.073,
+      "step": 15632
+    },
+    {
+      "epoch": 2.50128,
+      "grad_norm": 1.6645011362093007,
+      "learning_rate": 1.4343776974415668e-05,
+      "loss": 1.0882,
+      "step": 15633
+    },
+    {
+      "epoch": 2.50144,
+      "grad_norm": 1.738513289835566,
+      "learning_rate": 1.433486419710165e-05,
+      "loss": 1.086,
+      "step": 15634
+    },
+    {
+      "epoch": 2.5016,
+      "grad_norm": 1.8738637634430877,
+      "learning_rate": 1.432595397591041e-05,
+      "loss": 1.0625,
+      "step": 15635
+    },
+    {
+      "epoch": 2.50176,
+      "grad_norm": 1.701121479518655,
+      "learning_rate": 1.4317046311107906e-05,
+      "loss": 1.06,
+      "step": 15636
+    },
+    {
+      "epoch": 2.50192,
+      "grad_norm": 1.5908623598287552,
+      "learning_rate": 1.4308141202959846e-05,
+      "loss": 1.0602,
+      "step": 15637
+    },
+    {
+      "epoch": 2.50208,
+      "grad_norm": 2.0323966644880613,
+      "learning_rate": 1.4299238651732006e-05,
+      "loss": 1.0684,
+      "step": 15638
+    },
+    {
+      "epoch": 2.50224,
+      "grad_norm": 2.0538582221264674,
+      "learning_rate": 1.4290338657690006e-05,
+      "loss": 1.1064,
+      "step": 15639
+    },
+    {
+      "epoch": 2.5023999999999997,
+      "grad_norm": 1.9329565966759281,
+      "learning_rate": 1.4281441221099402e-05,
+      "loss": 1.0593,
+      "step": 15640
+    },
+    {
+      "epoch": 2.50256,
+      "grad_norm": 1.79091831877344,
+      "learning_rate": 1.427254634222569e-05,
+      "loss": 1.0654,
+      "step": 15641
+    },
+    {
+      "epoch": 2.50272,
+      "grad_norm": 1.9438803019421316,
+      "learning_rate": 1.4263654021334283e-05,
+      "loss": 1.1113,
+      "step": 15642
+    },
+    {
+      "epoch": 2.50288,
+      "grad_norm": 1.9122851557894367,
+      "learning_rate": 1.42547642586905e-05,
+      "loss": 1.092,
+      "step": 15643
+    },
+    {
+      "epoch": 2.50304,
+      "grad_norm": 1.865173761973865,
+      "learning_rate": 1.424587705455962e-05,
+      "loss": 1.0873,
+      "step": 15644
+    },
+    {
+      "epoch": 2.5032,
+      "grad_norm": 1.8271459289264176,
+      "learning_rate": 1.4236992409206807e-05,
+      "loss": 1.0853,
+      "step": 15645
+    },
+    {
+      "epoch": 2.50336,
+      "grad_norm": 1.7854971694159119,
+      "learning_rate": 1.422811032289716e-05,
+      "loss": 1.1327,
+      "step": 15646
+    },
+    {
+      "epoch": 2.50352,
+      "grad_norm": 1.5987861602953388,
+      "learning_rate": 1.4219230795895733e-05,
+      "loss": 1.1188,
+      "step": 15647
+    },
+    {
+      "epoch": 2.50368,
+      "grad_norm": 1.8414557634211879,
+      "learning_rate": 1.4210353828467449e-05,
+      "loss": 1.1342,
+      "step": 15648
+    },
+    {
+      "epoch": 2.50384,
+      "grad_norm": 1.8389721924705007,
+      "learning_rate": 1.4201479420877184e-05,
+      "loss": 1.0969,
+      "step": 15649
+    },
+    {
+      "epoch": 2.504,
+      "grad_norm": 1.9228789887581048,
+      "learning_rate": 1.4192607573389805e-05,
+      "loss": 1.109,
+      "step": 15650
+    },
+    {
+      "epoch": 2.50416,
+      "grad_norm": 1.880865680772457,
+      "learning_rate": 1.4183738286269921e-05,
+      "loss": 1.0741,
+      "step": 15651
+    },
+    {
+      "epoch": 2.50432,
+      "grad_norm": 1.7049973067222741,
+      "learning_rate": 1.4174871559782289e-05,
+      "loss": 1.0728,
+      "step": 15652
+    },
+    {
+      "epoch": 2.50448,
+      "grad_norm": 1.7955464113056605,
+      "learning_rate": 1.4166007394191384e-05,
+      "loss": 1.0575,
+      "step": 15653
+    },
+    {
+      "epoch": 2.50464,
+      "grad_norm": 2.1196275955412593,
+      "learning_rate": 1.4157145789761772e-05,
+      "loss": 1.11,
+      "step": 15654
+    },
+    {
+      "epoch": 2.5048,
+      "grad_norm": 1.7983256594878956,
+      "learning_rate": 1.4148286746757844e-05,
+      "loss": 1.0786,
+      "step": 15655
+    },
+    {
+      "epoch": 2.50496,
+      "grad_norm": 1.652129893647557,
+      "learning_rate": 1.4139430265443931e-05,
+      "loss": 1.0851,
+      "step": 15656
+    },
+    {
+      "epoch": 2.50512,
+      "grad_norm": 1.877172847964934,
+      "learning_rate": 1.4130576346084312e-05,
+      "loss": 1.1853,
+      "step": 15657
+    },
+    {
+      "epoch": 2.50528,
+      "grad_norm": 1.8077468558532443,
+      "learning_rate": 1.4121724988943175e-05,
+      "loss": 1.0616,
+      "step": 15658
+    },
+    {
+      "epoch": 2.50544,
+      "grad_norm": 1.9299261149385236,
+      "learning_rate": 1.4112876194284619e-05,
+      "loss": 1.1099,
+      "step": 15659
+    },
+    {
+      "epoch": 2.5056000000000003,
+      "grad_norm": 2.3297585723360705,
+      "learning_rate": 1.4104029962372689e-05,
+      "loss": 1.0494,
+      "step": 15660
+    },
+    {
+      "epoch": 2.50576,
+      "grad_norm": 1.8256573612329658,
+      "learning_rate": 1.4095186293471341e-05,
+      "loss": 1.1067,
+      "step": 15661
+    },
+    {
+      "epoch": 2.50592,
+      "grad_norm": 1.875455534822485,
+      "learning_rate": 1.4086345187844453e-05,
+      "loss": 1.0627,
+      "step": 15662
+    },
+    {
+      "epoch": 2.50608,
+      "grad_norm": 1.7254149388435698,
+      "learning_rate": 1.4077506645755845e-05,
+      "loss": 1.0491,
+      "step": 15663
+    },
+    {
+      "epoch": 2.50624,
+      "grad_norm": 1.7057490101114767,
+      "learning_rate": 1.4068670667469231e-05,
+      "loss": 1.064,
+      "step": 15664
+    },
+    {
+      "epoch": 2.5064,
+      "grad_norm": 1.6972264632420144,
+      "learning_rate": 1.4059837253248243e-05,
+      "loss": 1.1021,
+      "step": 15665
+    },
+    {
+      "epoch": 2.50656,
+      "grad_norm": 1.6181875149591969,
+      "learning_rate": 1.4051006403356537e-05,
+      "loss": 1.1154,
+      "step": 15666
+    },
+    {
+      "epoch": 2.50672,
+      "grad_norm": 1.7908945631680664,
+      "learning_rate": 1.4042178118057503e-05,
+      "loss": 1.094,
+      "step": 15667
+    },
+    {
+      "epoch": 2.5068799999999998,
+      "grad_norm": 1.7986987227875209,
+      "learning_rate": 1.4033352397614674e-05,
+      "loss": 1.0821,
+      "step": 15668
+    },
+    {
+      "epoch": 2.50704,
+      "grad_norm": 2.0602566092997967,
+      "learning_rate": 1.4024529242291295e-05,
+      "loss": 1.0892,
+      "step": 15669
+    },
+    {
+      "epoch": 2.5072,
+      "grad_norm": 1.7932699655339992,
+      "learning_rate": 1.4015708652350712e-05,
+      "loss": 1.1001,
+      "step": 15670
+    },
+    {
+      "epoch": 2.5073600000000003,
+      "grad_norm": 1.7652578658559275,
+      "learning_rate": 1.400689062805608e-05,
+      "loss": 1.1036,
+      "step": 15671
+    },
+    {
+      "epoch": 2.50752,
+      "grad_norm": 1.8078125879973437,
+      "learning_rate": 1.3998075169670532e-05,
+      "loss": 1.0683,
+      "step": 15672
+    },
+    {
+      "epoch": 2.50768,
+      "grad_norm": 1.6155420893574928,
+      "learning_rate": 1.3989262277457105e-05,
+      "loss": 1.1025,
+      "step": 15673
+    },
+    {
+      "epoch": 2.50784,
+      "grad_norm": 1.8023789675849329,
+      "learning_rate": 1.398045195167873e-05,
+      "loss": 1.0552,
+      "step": 15674
+    },
+    {
+      "epoch": 2.508,
+      "grad_norm": 1.5850241703490624,
+      "learning_rate": 1.3971644192598376e-05,
+      "loss": 1.0437,
+      "step": 15675
+    },
+    {
+      "epoch": 2.50816,
+      "grad_norm": 1.901287521422136,
+      "learning_rate": 1.3962839000478766e-05,
+      "loss": 1.1383,
+      "step": 15676
+    },
+    {
+      "epoch": 2.50832,
+      "grad_norm": 1.6516762704303656,
+      "learning_rate": 1.3954036375582702e-05,
+      "loss": 1.1109,
+      "step": 15677
+    },
+    {
+      "epoch": 2.50848,
+      "grad_norm": 1.714685241398687,
+      "learning_rate": 1.3945236318172772e-05,
+      "loss": 1.1255,
+      "step": 15678
+    },
+    {
+      "epoch": 2.5086399999999998,
+      "grad_norm": 1.889862188959571,
+      "learning_rate": 1.3936438828511645e-05,
+      "loss": 1.0981,
+      "step": 15679
+    },
+    {
+      "epoch": 2.5088,
+      "grad_norm": 1.8814351745467917,
+      "learning_rate": 1.392764390686172e-05,
+      "loss": 1.1251,
+      "step": 15680
+    },
+    {
+      "epoch": 2.50896,
+      "grad_norm": 1.8004335974745447,
+      "learning_rate": 1.3918851553485501e-05,
+      "loss": 1.0592,
+      "step": 15681
+    },
+    {
+      "epoch": 2.5091200000000002,
+      "grad_norm": 1.7339609184156146,
+      "learning_rate": 1.3910061768645333e-05,
+      "loss": 1.066,
+      "step": 15682
+    },
+    {
+      "epoch": 2.50928,
+      "grad_norm": 1.841249893746043,
+      "learning_rate": 1.390127455260346e-05,
+      "loss": 1.0717,
+      "step": 15683
+    },
+    {
+      "epoch": 2.50944,
+      "grad_norm": 1.7588260123706945,
+      "learning_rate": 1.3892489905622107e-05,
+      "loss": 1.0666,
+      "step": 15684
+    },
+    {
+      "epoch": 2.5096,
+      "grad_norm": 1.7588260123706945,
+      "learning_rate": 1.3892489905622107e-05,
+      "loss": 1.2209,
+      "step": 15685
+    },
+    {
+      "epoch": 2.50976,
+      "grad_norm": 1.8112828034950461,
+      "learning_rate": 1.3883707827963377e-05,
+      "loss": 1.0463,
+      "step": 15686
+    },
+    {
+      "epoch": 2.50992,
+      "grad_norm": 1.7332057373465606,
+      "learning_rate": 1.3874928319889324e-05,
+      "loss": 1.1168,
+      "step": 15687
+    },
+    {
+      "epoch": 2.51008,
+      "grad_norm": 1.8505573021298156,
+      "learning_rate": 1.3866151381661906e-05,
+      "loss": 1.0766,
+      "step": 15688
+    },
+    {
+      "epoch": 2.51024,
+      "grad_norm": 1.836730252154148,
+      "learning_rate": 1.3857377013543028e-05,
+      "loss": 1.0931,
+      "step": 15689
+    },
+    {
+      "epoch": 2.5103999999999997,
+      "grad_norm": 1.8603468786797672,
+      "learning_rate": 1.3848605215794496e-05,
+      "loss": 1.0622,
+      "step": 15690
+    },
+    {
+      "epoch": 2.51056,
+      "grad_norm": 2.001390397723688,
+      "learning_rate": 1.3839835988678018e-05,
+      "loss": 1.0555,
+      "step": 15691
+    },
+    {
+      "epoch": 2.51072,
+      "grad_norm": 1.674458961941344,
+      "learning_rate": 1.3831069332455337e-05,
+      "loss": 1.0776,
+      "step": 15692
+    },
+    {
+      "epoch": 2.5108800000000002,
+      "grad_norm": 1.721734589365623,
+      "learning_rate": 1.3822305247387946e-05,
+      "loss": 1.0821,
+      "step": 15693
+    },
+    {
+      "epoch": 2.51104,
+      "grad_norm": 1.5904374192745117,
+      "learning_rate": 1.3813543733737422e-05,
+      "loss": 1.0972,
+      "step": 15694
+    },
+    {
+      "epoch": 2.5112,
+      "grad_norm": 1.5946882163233325,
+      "learning_rate": 1.3804784791765135e-05,
+      "loss": 1.0895,
+      "step": 15695
+    },
+    {
+      "epoch": 2.51136,
+      "grad_norm": 1.9412220167409533,
+      "learning_rate": 1.379602842173251e-05,
+      "loss": 1.0879,
+      "step": 15696
+    },
+    {
+      "epoch": 2.51152,
+      "grad_norm": 1.7715259408661321,
+      "learning_rate": 1.3787274623900737e-05,
+      "loss": 1.058,
+      "step": 15697
+    },
+    {
+      "epoch": 2.51168,
+      "grad_norm": 1.9785992073559875,
+      "learning_rate": 1.3778523398531095e-05,
+      "loss": 1.0516,
+      "step": 15698
+    },
+    {
+      "epoch": 2.51184,
+      "grad_norm": 1.5549937732156773,
+      "learning_rate": 1.3769774745884668e-05,
+      "loss": 1.1369,
+      "step": 15699
+    },
+    {
+      "epoch": 2.512,
+      "grad_norm": 1.7504098161240735,
+      "learning_rate": 1.376102866622252e-05,
+      "loss": 1.054,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5121599999999997,
+      "grad_norm": 2.1569290115795727,
+      "learning_rate": 1.37522851598056e-05,
+      "loss": 1.1009,
+      "step": 15701
+    },
+    {
+      "epoch": 2.51232,
+      "grad_norm": 2.102595461524389,
+      "learning_rate": 1.3743544226894822e-05,
+      "loss": 1.1113,
+      "step": 15702
+    },
+    {
+      "epoch": 2.51248,
+      "grad_norm": 1.7351113859693332,
+      "learning_rate": 1.3734805867751e-05,
+      "loss": 1.0583,
+      "step": 15703
+    },
+    {
+      "epoch": 2.51264,
+      "grad_norm": 1.8836833728601259,
+      "learning_rate": 1.3726070082634868e-05,
+      "loss": 1.0347,
+      "step": 15704
+    },
+    {
+      "epoch": 2.5128,
+      "grad_norm": 1.8369776203032235,
+      "learning_rate": 1.3717336871807086e-05,
+      "loss": 1.1277,
+      "step": 15705
+    },
+    {
+      "epoch": 2.51296,
+      "grad_norm": 1.8289410799302332,
+      "learning_rate": 1.3708606235528242e-05,
+      "loss": 1.0749,
+      "step": 15706
+    },
+    {
+      "epoch": 2.51312,
+      "grad_norm": 1.813519304464606,
+      "learning_rate": 1.3699878174058834e-05,
+      "loss": 1.0625,
+      "step": 15707
+    },
+    {
+      "epoch": 2.51328,
+      "grad_norm": 1.8215485446456692,
+      "learning_rate": 1.3691152687659347e-05,
+      "loss": 1.1575,
+      "step": 15708
+    },
+    {
+      "epoch": 2.51344,
+      "grad_norm": 1.9906376688263445,
+      "learning_rate": 1.3682429776590056e-05,
+      "loss": 1.1173,
+      "step": 15709
+    },
+    {
+      "epoch": 2.5136,
+      "grad_norm": 2.143136201120386,
+      "learning_rate": 1.3673709441111327e-05,
+      "loss": 1.1111,
+      "step": 15710
+    },
+    {
+      "epoch": 2.51376,
+      "grad_norm": 2.055904960076967,
+      "learning_rate": 1.3664991681483274e-05,
+      "loss": 1.1301,
+      "step": 15711
+    },
+    {
+      "epoch": 2.51392,
+      "grad_norm": 2.093346355488537,
+      "learning_rate": 1.3656276497966114e-05,
+      "loss": 1.0747,
+      "step": 15712
+    },
+    {
+      "epoch": 2.51408,
+      "grad_norm": 1.7300666452886742,
+      "learning_rate": 1.3647563890819792e-05,
+      "loss": 1.0646,
+      "step": 15713
+    },
+    {
+      "epoch": 2.51424,
+      "grad_norm": 1.7764780400870965,
+      "learning_rate": 1.3638853860304368e-05,
+      "loss": 1.1353,
+      "step": 15714
+    },
+    {
+      "epoch": 2.5144,
+      "grad_norm": 1.6494553468017543,
+      "learning_rate": 1.3630146406679701e-05,
+      "loss": 1.0543,
+      "step": 15715
+    },
+    {
+      "epoch": 2.51456,
+      "grad_norm": 1.82306201912164,
+      "learning_rate": 1.3621441530205603e-05,
+      "loss": 1.1207,
+      "step": 15716
+    },
+    {
+      "epoch": 2.51472,
+      "grad_norm": 1.8871388712982022,
+      "learning_rate": 1.3612739231141836e-05,
+      "loss": 1.1211,
+      "step": 15717
+    },
+    {
+      "epoch": 2.51488,
+      "grad_norm": 1.7066232216520636,
+      "learning_rate": 1.3604039509748035e-05,
+      "loss": 1.0962,
+      "step": 15718
+    },
+    {
+      "epoch": 2.51504,
+      "grad_norm": 1.98325977606049,
+      "learning_rate": 1.3595342366283814e-05,
+      "loss": 1.077,
+      "step": 15719
+    },
+    {
+      "epoch": 2.5152,
+      "grad_norm": 1.6236916270629496,
+      "learning_rate": 1.3586647801008667e-05,
+      "loss": 1.0761,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5153600000000003,
+      "grad_norm": 1.7439371634572882,
+      "learning_rate": 1.3577955814182041e-05,
+      "loss": 1.0504,
+      "step": 15721
+    },
+    {
+      "epoch": 2.51552,
+      "grad_norm": 1.6441298431055549,
+      "learning_rate": 1.3569266406063274e-05,
+      "loss": 1.1003,
+      "step": 15722
+    },
+    {
+      "epoch": 2.51568,
+      "grad_norm": 1.5755947878211707,
+      "learning_rate": 1.3560579576911636e-05,
+      "loss": 1.0326,
+      "step": 15723
+    },
+    {
+      "epoch": 2.51584,
+      "grad_norm": 1.5484102309593144,
+      "learning_rate": 1.3551895326986386e-05,
+      "loss": 1.078,
+      "step": 15724
+    },
+    {
+      "epoch": 2.516,
+      "grad_norm": 1.7383700162361875,
+      "learning_rate": 1.3543213656546572e-05,
+      "loss": 1.0848,
+      "step": 15725
+    },
+    {
+      "epoch": 2.51616,
+      "grad_norm": 1.8683086114756804,
+      "learning_rate": 1.3534534565851331e-05,
+      "loss": 1.0932,
+      "step": 15726
+    },
+    {
+      "epoch": 2.51632,
+      "grad_norm": 1.634777024130483,
+      "learning_rate": 1.3525858055159535e-05,
+      "loss": 1.0955,
+      "step": 15727
+    },
+    {
+      "epoch": 2.51648,
+      "grad_norm": 1.7360946903034782,
+      "learning_rate": 1.3517184124730153e-05,
+      "loss": 1.1318,
+      "step": 15728
+    },
+    {
+      "epoch": 2.5166399999999998,
+      "grad_norm": 1.6741486821075964,
+      "learning_rate": 1.3508512774821968e-05,
+      "loss": 1.1042,
+      "step": 15729
+    },
+    {
+      "epoch": 2.5168,
+      "grad_norm": 1.7096715335231498,
+      "learning_rate": 1.349984400569374e-05,
+      "loss": 1.0716,
+      "step": 15730
+    },
+    {
+      "epoch": 2.51696,
+      "grad_norm": 1.729181086643781,
+      "learning_rate": 1.3491177817604128e-05,
+      "loss": 1.093,
+      "step": 15731
+    },
+    {
+      "epoch": 2.5171200000000002,
+      "grad_norm": 1.8659228191522261,
+      "learning_rate": 1.3482514210811703e-05,
+      "loss": 1.138,
+      "step": 15732
+    },
+    {
+      "epoch": 2.51728,
+      "grad_norm": 1.6740435035116734,
+      "learning_rate": 1.3473853185574991e-05,
+      "loss": 1.085,
+      "step": 15733
+    },
+    {
+      "epoch": 2.51744,
+      "grad_norm": 1.7674804066418208,
+      "learning_rate": 1.3465194742152387e-05,
+      "loss": 1.1092,
+      "step": 15734
+    },
+    {
+      "epoch": 2.5176,
+      "grad_norm": 1.6951970091227366,
+      "learning_rate": 1.3456538880802328e-05,
+      "loss": 1.0717,
+      "step": 15735
+    },
+    {
+      "epoch": 2.51776,
+      "grad_norm": 2.002936160057017,
+      "learning_rate": 1.3447885601782995e-05,
+      "loss": 1.1194,
+      "step": 15736
+    },
+    {
+      "epoch": 2.51792,
+      "grad_norm": 1.9519106373898656,
+      "learning_rate": 1.3439234905352682e-05,
+      "loss": 1.1269,
+      "step": 15737
+    },
+    {
+      "epoch": 2.51808,
+      "grad_norm": 1.6692860550262627,
+      "learning_rate": 1.3430586791769428e-05,
+      "loss": 1.1058,
+      "step": 15738
+    },
+    {
+      "epoch": 2.51824,
+      "grad_norm": 1.539145490763042,
+      "learning_rate": 1.3421941261291338e-05,
+      "loss": 1.0797,
+      "step": 15739
+    },
+    {
+      "epoch": 2.5183999999999997,
+      "grad_norm": 1.7656127336437502,
+      "learning_rate": 1.3413298314176359e-05,
+      "loss": 1.1011,
+      "step": 15740
+    },
+    {
+      "epoch": 2.51856,
+      "grad_norm": 1.7149493255503525,
+      "learning_rate": 1.3404657950682398e-05,
+      "loss": 1.0336,
+      "step": 15741
+    },
+    {
+      "epoch": 2.51872,
+      "grad_norm": 1.534904261176185,
+      "learning_rate": 1.3396020171067247e-05,
+      "loss": 1.0931,
+      "step": 15742
+    },
+    {
+      "epoch": 2.5188800000000002,
+      "grad_norm": 2.0444797888563784,
+      "learning_rate": 1.3387384975588669e-05,
+      "loss": 1.139,
+      "step": 15743
+    },
+    {
+      "epoch": 2.51904,
+      "grad_norm": 2.1947730147925717,
+      "learning_rate": 1.337875236450431e-05,
+      "loss": 1.0989,
+      "step": 15744
+    },
+    {
+      "epoch": 2.5192,
+      "grad_norm": 1.6202072773798353,
+      "learning_rate": 1.3370122338071767e-05,
+      "loss": 1.0742,
+      "step": 15745
+    },
+    {
+      "epoch": 2.51936,
+      "grad_norm": 1.7436567033628008,
+      "learning_rate": 1.3361494896548533e-05,
+      "loss": 1.1075,
+      "step": 15746
+    },
+    {
+      "epoch": 2.51952,
+      "grad_norm": 1.9323386748110132,
+      "learning_rate": 1.3352870040192045e-05,
+      "loss": 1.0705,
+      "step": 15747
+    },
+    {
+      "epoch": 2.51968,
+      "grad_norm": 1.8017741010431658,
+      "learning_rate": 1.3344247769259633e-05,
+      "loss": 1.1003,
+      "step": 15748
+    },
+    {
+      "epoch": 2.51984,
+      "grad_norm": 1.6829421141165202,
+      "learning_rate": 1.3335628084008656e-05,
+      "loss": 1.0654,
+      "step": 15749
+    },
+    {
+      "epoch": 2.52,
+      "grad_norm": 1.5693779289853558,
+      "learning_rate": 1.3327010984696198e-05,
+      "loss": 1.0939,
+      "step": 15750
+    },
+    {
+      "epoch": 2.5201599999999997,
+      "grad_norm": 2.0664815387735076,
+      "learning_rate": 1.3318396471579475e-05,
+      "loss": 1.0982,
+      "step": 15751
+    },
+    {
+      "epoch": 2.52032,
+      "grad_norm": 1.8290102822962642,
+      "learning_rate": 1.330978454491546e-05,
+      "loss": 1.0788,
+      "step": 15752
+    },
+    {
+      "epoch": 2.52048,
+      "grad_norm": 1.880319712575036,
+      "learning_rate": 1.3301175204961191e-05,
+      "loss": 1.1058,
+      "step": 15753
+    },
+    {
+      "epoch": 2.52064,
+      "grad_norm": 2.0024509023185653,
+      "learning_rate": 1.3292568451973475e-05,
+      "loss": 1.0905,
+      "step": 15754
+    },
+    {
+      "epoch": 2.5208,
+      "grad_norm": 1.9379429444567517,
+      "learning_rate": 1.3283964286209194e-05,
+      "loss": 1.0865,
+      "step": 15755
+    },
+    {
+      "epoch": 2.52096,
+      "grad_norm": 1.9546282009615303,
+      "learning_rate": 1.3275362707925064e-05,
+      "loss": 1.0297,
+      "step": 15756
+    },
+    {
+      "epoch": 2.52112,
+      "grad_norm": 1.5765578396699016,
+      "learning_rate": 1.326676371737774e-05,
+      "loss": 1.0915,
+      "step": 15757
+    },
+    {
+      "epoch": 2.52128,
+      "grad_norm": 1.6567244220844553,
+      "learning_rate": 1.3258167314823799e-05,
+      "loss": 1.0852,
+      "step": 15758
+    },
+    {
+      "epoch": 2.52144,
+      "grad_norm": 1.6414342101709323,
+      "learning_rate": 1.3249573500519751e-05,
+      "loss": 1.0713,
+      "step": 15759
+    },
+    {
+      "epoch": 2.5216,
+      "grad_norm": 1.9251538800438006,
+      "learning_rate": 1.3240982274722014e-05,
+      "loss": 1.0818,
+      "step": 15760
+    },
+    {
+      "epoch": 2.52176,
+      "grad_norm": 1.8651421050787456,
+      "learning_rate": 1.3232393637686957e-05,
+      "loss": 1.1198,
+      "step": 15761
+    },
+    {
+      "epoch": 2.52192,
+      "grad_norm": 1.7532203416247842,
+      "learning_rate": 1.3223807589670834e-05,
+      "loss": 1.0282,
+      "step": 15762
+    },
+    {
+      "epoch": 2.52208,
+      "grad_norm": 1.677192220749273,
+      "learning_rate": 1.3215224130929837e-05,
+      "loss": 1.0599,
+      "step": 15763
+    },
+    {
+      "epoch": 2.52224,
+      "grad_norm": 1.7854873553800057,
+      "learning_rate": 1.3206643261720086e-05,
+      "loss": 1.0837,
+      "step": 15764
+    },
+    {
+      "epoch": 2.5224,
+      "grad_norm": 1.7437108339702847,
+      "learning_rate": 1.3198064982297664e-05,
+      "loss": 1.0932,
+      "step": 15765
+    },
+    {
+      "epoch": 2.52256,
+      "grad_norm": 1.7785572484020382,
+      "learning_rate": 1.3189489292918455e-05,
+      "loss": 1.0572,
+      "step": 15766
+    },
+    {
+      "epoch": 2.52272,
+      "grad_norm": 1.6688504887388633,
+      "learning_rate": 1.3180916193838443e-05,
+      "loss": 1.0519,
+      "step": 15767
+    },
+    {
+      "epoch": 2.52288,
+      "grad_norm": 1.587279236583126,
+      "learning_rate": 1.3172345685313325e-05,
+      "loss": 1.0672,
+      "step": 15768
+    },
+    {
+      "epoch": 2.52304,
+      "grad_norm": 1.9545448869619784,
+      "learning_rate": 1.3163777767598951e-05,
+      "loss": 1.0958,
+      "step": 15769
+    },
+    {
+      "epoch": 2.5232,
+      "grad_norm": 1.5673911919250987,
+      "learning_rate": 1.315521244095086e-05,
+      "loss": 1.0641,
+      "step": 15770
+    },
+    {
+      "epoch": 2.5233600000000003,
+      "grad_norm": 1.7759603453701434,
+      "learning_rate": 1.3146649705624703e-05,
+      "loss": 1.1099,
+      "step": 15771
+    },
+    {
+      "epoch": 2.52352,
+      "grad_norm": 1.7447469551605495,
+      "learning_rate": 1.3138089561875965e-05,
+      "loss": 1.0785,
+      "step": 15772
+    },
+    {
+      "epoch": 2.52368,
+      "grad_norm": 1.7595718266500204,
+      "learning_rate": 1.3129532009960067e-05,
+      "loss": 1.063,
+      "step": 15773
+    },
+    {
+      "epoch": 2.52384,
+      "grad_norm": 1.7251772779428123,
+      "learning_rate": 1.3120977050132344e-05,
+      "loss": 1.0631,
+      "step": 15774
+    },
+    {
+      "epoch": 2.524,
+      "grad_norm": 1.7684882080028883,
+      "learning_rate": 1.3112424682648073e-05,
+      "loss": 1.063,
+      "step": 15775
+    },
+    {
+      "epoch": 2.52416,
+      "grad_norm": 2.3181923422938313,
+      "learning_rate": 1.3103874907762437e-05,
+      "loss": 1.2848,
+      "step": 15776
+    },
+    {
+      "epoch": 2.52432,
+      "grad_norm": 1.65231196979327,
+      "learning_rate": 1.3095327725730555e-05,
+      "loss": 1.0725,
+      "step": 15777
+    },
+    {
+      "epoch": 2.52448,
+      "grad_norm": 1.6513391220435303,
+      "learning_rate": 1.3086783136807457e-05,
+      "loss": 1.0735,
+      "step": 15778
+    },
+    {
+      "epoch": 2.5246399999999998,
+      "grad_norm": 1.7503185422188563,
+      "learning_rate": 1.3078241141248104e-05,
+      "loss": 1.0974,
+      "step": 15779
+    },
+    {
+      "epoch": 2.5248,
+      "grad_norm": 1.7023361975057876,
+      "learning_rate": 1.306970173930736e-05,
+      "loss": 1.0977,
+      "step": 15780
+    },
+    {
+      "epoch": 2.52496,
+      "grad_norm": 1.9261329546935841,
+      "learning_rate": 1.3061164931240088e-05,
+      "loss": 1.0616,
+      "step": 15781
+    },
+    {
+      "epoch": 2.5251200000000003,
+      "grad_norm": 1.8248827575056719,
+      "learning_rate": 1.3052630717300928e-05,
+      "loss": 1.0265,
+      "step": 15782
+    },
+    {
+      "epoch": 2.52528,
+      "grad_norm": 1.923839180253186,
+      "learning_rate": 1.304409909774461e-05,
+      "loss": 1.0367,
+      "step": 15783
+    },
+    {
+      "epoch": 2.52544,
+      "grad_norm": 1.9970470366704225,
+      "learning_rate": 1.3035570072825632e-05,
+      "loss": 1.072,
+      "step": 15784
+    },
+    {
+      "epoch": 2.5256,
+      "grad_norm": 1.6256493142092014,
+      "learning_rate": 1.3027043642798564e-05,
+      "loss": 1.0864,
+      "step": 15785
+    },
+    {
+      "epoch": 2.52576,
+      "grad_norm": 1.7692670884094492,
+      "learning_rate": 1.301851980791773e-05,
+      "loss": 1.0762,
+      "step": 15786
+    },
+    {
+      "epoch": 2.52592,
+      "grad_norm": 2.020232500746277,
+      "learning_rate": 1.3009998568437543e-05,
+      "loss": 1.0997,
+      "step": 15787
+    },
+    {
+      "epoch": 2.52608,
+      "grad_norm": 1.927760026517816,
+      "learning_rate": 1.3001479924612248e-05,
+      "loss": 1.075,
+      "step": 15788
+    },
+    {
+      "epoch": 2.52624,
+      "grad_norm": 1.7169625761463265,
+      "learning_rate": 1.2992963876696019e-05,
+      "loss": 1.0405,
+      "step": 15789
+    },
+    {
+      "epoch": 2.5263999999999998,
+      "grad_norm": 1.8633648879579516,
+      "learning_rate": 1.2984450424942962e-05,
+      "loss": 1.0967,
+      "step": 15790
+    },
+    {
+      "epoch": 2.52656,
+      "grad_norm": 1.7964988034483993,
+      "learning_rate": 1.2975939569607088e-05,
+      "loss": 1.0634,
+      "step": 15791
+    },
+    {
+      "epoch": 2.52672,
+      "grad_norm": 1.8124748682627185,
+      "learning_rate": 1.2967431310942412e-05,
+      "loss": 1.1069,
+      "step": 15792
+    },
+    {
+      "epoch": 2.5268800000000002,
+      "grad_norm": 2.6988861589676945,
+      "learning_rate": 1.2958925649202725e-05,
+      "loss": 1.1106,
+      "step": 15793
+    },
+    {
+      "epoch": 2.52704,
+      "grad_norm": 1.8484460296345508,
+      "learning_rate": 1.2950422584641907e-05,
+      "loss": 1.0883,
+      "step": 15794
+    },
+    {
+      "epoch": 2.5272,
+      "grad_norm": 1.9833073063316262,
+      "learning_rate": 1.2941922117513594e-05,
+      "loss": 1.1166,
+      "step": 15795
+    },
+    {
+      "epoch": 2.52736,
+      "grad_norm": 1.7960597920981234,
+      "learning_rate": 1.293342424807149e-05,
+      "loss": 1.1226,
+      "step": 15796
+    },
+    {
+      "epoch": 2.52752,
+      "grad_norm": 1.6550846676526096,
+      "learning_rate": 1.2924928976569139e-05,
+      "loss": 1.0787,
+      "step": 15797
+    },
+    {
+      "epoch": 2.52768,
+      "grad_norm": 1.616295688905032,
+      "learning_rate": 1.2916436303260015e-05,
+      "loss": 1.1217,
+      "step": 15798
+    },
+    {
+      "epoch": 2.52784,
+      "grad_norm": 1.593194077390599,
+      "learning_rate": 1.290794622839755e-05,
+      "loss": 1.032,
+      "step": 15799
+    },
+    {
+      "epoch": 2.528,
+      "grad_norm": 2.077164198160908,
+      "learning_rate": 1.289945875223505e-05,
+      "loss": 1.0937,
+      "step": 15800
+    },
+    {
+      "epoch": 2.5281599999999997,
+      "grad_norm": 2.0949241348342014,
+      "learning_rate": 1.289097387502579e-05,
+      "loss": 1.1325,
+      "step": 15801
+    },
+    {
+      "epoch": 2.52832,
+      "grad_norm": 1.9704072735986506,
+      "learning_rate": 1.288249159702294e-05,
+      "loss": 1.0983,
+      "step": 15802
+    },
+    {
+      "epoch": 2.52848,
+      "grad_norm": 1.7488220041341305,
+      "learning_rate": 1.287401191847959e-05,
+      "loss": 1.0972,
+      "step": 15803
+    },
+    {
+      "epoch": 2.52864,
+      "grad_norm": 1.7645102927538967,
+      "learning_rate": 1.2865534839648774e-05,
+      "loss": 1.0308,
+      "step": 15804
+    },
+    {
+      "epoch": 2.5288,
+      "grad_norm": 1.891350273497294,
+      "learning_rate": 1.2857060360783402e-05,
+      "loss": 1.0523,
+      "step": 15805
+    },
+    {
+      "epoch": 2.52896,
+      "grad_norm": 2.0130450054917337,
+      "learning_rate": 1.2848588482136415e-05,
+      "loss": 1.0811,
+      "step": 15806
+    },
+    {
+      "epoch": 2.52912,
+      "grad_norm": 1.8596228701332609,
+      "learning_rate": 1.284011920396052e-05,
+      "loss": 1.044,
+      "step": 15807
+    },
+    {
+      "epoch": 2.52928,
+      "grad_norm": 1.7802983553285505,
+      "learning_rate": 1.2831652526508497e-05,
+      "loss": 1.0636,
+      "step": 15808
+    },
+    {
+      "epoch": 2.52944,
+      "grad_norm": 1.810614429013997,
+      "learning_rate": 1.2823188450032898e-05,
+      "loss": 1.0851,
+      "step": 15809
+    },
+    {
+      "epoch": 2.5296,
+      "grad_norm": 1.5814135763108512,
+      "learning_rate": 1.2814726974786374e-05,
+      "loss": 1.0903,
+      "step": 15810
+    },
+    {
+      "epoch": 2.52976,
+      "grad_norm": 1.8165510937582943,
+      "learning_rate": 1.2806268101021312e-05,
+      "loss": 1.0721,
+      "step": 15811
+    },
+    {
+      "epoch": 2.5299199999999997,
+      "grad_norm": 1.7370151678713674,
+      "learning_rate": 1.2797811828990181e-05,
+      "loss": 1.1069,
+      "step": 15812
+    },
+    {
+      "epoch": 2.53008,
+      "grad_norm": 1.7619232899672956,
+      "learning_rate": 1.2789358158945275e-05,
+      "loss": 1.0792,
+      "step": 15813
+    },
+    {
+      "epoch": 2.53024,
+      "grad_norm": 1.8856372662999994,
+      "learning_rate": 1.2780907091138838e-05,
+      "loss": 1.1474,
+      "step": 15814
+    },
+    {
+      "epoch": 2.5304,
+      "grad_norm": 2.070215853170538,
+      "learning_rate": 1.2772458625823048e-05,
+      "loss": 1.1438,
+      "step": 15815
+    },
+    {
+      "epoch": 2.53056,
+      "grad_norm": 1.7625818516918734,
+      "learning_rate": 1.2764012763249977e-05,
+      "loss": 1.0679,
+      "step": 15816
+    },
+    {
+      "epoch": 2.53072,
+      "grad_norm": 1.8691622460939308,
+      "learning_rate": 1.2755569503671661e-05,
+      "loss": 1.0941,
+      "step": 15817
+    },
+    {
+      "epoch": 2.53088,
+      "grad_norm": 1.6935775594721194,
+      "learning_rate": 1.2747128847340017e-05,
+      "loss": 1.0615,
+      "step": 15818
+    },
+    {
+      "epoch": 2.53104,
+      "grad_norm": 1.6909130731499242,
+      "learning_rate": 1.2738690794506903e-05,
+      "loss": 1.1079,
+      "step": 15819
+    },
+    {
+      "epoch": 2.5312,
+      "grad_norm": 1.7853294320085185,
+      "learning_rate": 1.2730255345424113e-05,
+      "loss": 1.1261,
+      "step": 15820
+    },
+    {
+      "epoch": 2.5313600000000003,
+      "grad_norm": 1.890027779191575,
+      "learning_rate": 1.2721822500343305e-05,
+      "loss": 1.0958,
+      "step": 15821
+    },
+    {
+      "epoch": 2.53152,
+      "grad_norm": 1.859115561891521,
+      "learning_rate": 1.2713392259516188e-05,
+      "loss": 1.1258,
+      "step": 15822
+    },
+    {
+      "epoch": 2.53168,
+      "grad_norm": 1.73115899437571,
+      "learning_rate": 1.2704964623194216e-05,
+      "loss": 1.1225,
+      "step": 15823
+    },
+    {
+      "epoch": 2.53184,
+      "grad_norm": 1.5481519748464823,
+      "learning_rate": 1.2696539591628932e-05,
+      "loss": 1.077,
+      "step": 15824
+    },
+    {
+      "epoch": 2.532,
+      "grad_norm": 1.8877547823134055,
+      "learning_rate": 1.2688117165071656e-05,
+      "loss": 1.0967,
+      "step": 15825
+    },
+    {
+      "epoch": 2.53216,
+      "grad_norm": 1.845799866259032,
+      "learning_rate": 1.2679697343773766e-05,
+      "loss": 1.1011,
+      "step": 15826
+    },
+    {
+      "epoch": 2.53232,
+      "grad_norm": 1.6691942741199024,
+      "learning_rate": 1.267128012798644e-05,
+      "loss": 1.087,
+      "step": 15827
+    },
+    {
+      "epoch": 2.53248,
+      "grad_norm": 1.8153833361931104,
+      "learning_rate": 1.2662865517960876e-05,
+      "loss": 1.0833,
+      "step": 15828
+    },
+    {
+      "epoch": 2.53264,
+      "grad_norm": 1.7048581304186141,
+      "learning_rate": 1.2654453513948151e-05,
+      "loss": 1.0924,
+      "step": 15829
+    },
+    {
+      "epoch": 2.5328,
+      "grad_norm": 1.7370896166966332,
+      "learning_rate": 1.2646044116199252e-05,
+      "loss": 1.1215,
+      "step": 15830
+    },
+    {
+      "epoch": 2.53296,
+      "grad_norm": 1.773026496480225,
+      "learning_rate": 1.2637637324965102e-05,
+      "loss": 1.0694,
+      "step": 15831
+    },
+    {
+      "epoch": 2.5331200000000003,
+      "grad_norm": 1.6890408171904783,
+      "learning_rate": 1.2629233140496565e-05,
+      "loss": 1.1025,
+      "step": 15832
+    },
+    {
+      "epoch": 2.53328,
+      "grad_norm": 1.93152040418224,
+      "learning_rate": 1.2620831563044389e-05,
+      "loss": 1.0844,
+      "step": 15833
+    },
+    {
+      "epoch": 2.53344,
+      "grad_norm": 1.534684310149666,
+      "learning_rate": 1.261243259285928e-05,
+      "loss": 1.0368,
+      "step": 15834
+    },
+    {
+      "epoch": 2.5336,
+      "grad_norm": 1.628084974976364,
+      "learning_rate": 1.260403623019184e-05,
+      "loss": 1.0739,
+      "step": 15835
+    },
+    {
+      "epoch": 2.53376,
+      "grad_norm": 2.0848506204546413,
+      "learning_rate": 1.2595642475292613e-05,
+      "loss": 1.0645,
+      "step": 15836
+    },
+    {
+      "epoch": 2.53392,
+      "grad_norm": 1.6981849740205468,
+      "learning_rate": 1.2587251328412031e-05,
+      "loss": 1.1463,
+      "step": 15837
+    },
+    {
+      "epoch": 2.53408,
+      "grad_norm": 1.7210371965542226,
+      "learning_rate": 1.2578862789800539e-05,
+      "loss": 1.0944,
+      "step": 15838
+    },
+    {
+      "epoch": 2.53424,
+      "grad_norm": 1.6502760160663927,
+      "learning_rate": 1.2570476859708357e-05,
+      "loss": 1.1028,
+      "step": 15839
+    },
+    {
+      "epoch": 2.5343999999999998,
+      "grad_norm": 1.5514259346726844,
+      "learning_rate": 1.2562093538385778e-05,
+      "loss": 1.0875,
+      "step": 15840
+    },
+    {
+      "epoch": 2.53456,
+      "grad_norm": 1.819330864251425,
+      "learning_rate": 1.2553712826082886e-05,
+      "loss": 1.1095,
+      "step": 15841
+    },
+    {
+      "epoch": 2.53472,
+      "grad_norm": 2.313202915380909,
+      "learning_rate": 1.2545334723049817e-05,
+      "loss": 1.0492,
+      "step": 15842
+    },
+    {
+      "epoch": 2.5348800000000002,
+      "grad_norm": 1.8036203373706567,
+      "learning_rate": 1.2536959229536483e-05,
+      "loss": 1.1241,
+      "step": 15843
+    },
+    {
+      "epoch": 2.53504,
+      "grad_norm": 1.6368096544144086,
+      "learning_rate": 1.252858634579287e-05,
+      "loss": 1.0314,
+      "step": 15844
+    },
+    {
+      "epoch": 2.5352,
+      "grad_norm": 1.7371309655332379,
+      "learning_rate": 1.252021607206877e-05,
+      "loss": 1.0445,
+      "step": 15845
+    },
+    {
+      "epoch": 2.53536,
+      "grad_norm": 1.8843626178567228,
+      "learning_rate": 1.2511848408613946e-05,
+      "loss": 1.0378,
+      "step": 15846
+    },
+    {
+      "epoch": 2.53552,
+      "grad_norm": 1.8866631666082305,
+      "learning_rate": 1.2503483355678091e-05,
+      "loss": 1.105,
+      "step": 15847
+    },
+    {
+      "epoch": 2.53568,
+      "grad_norm": 1.7604106236811374,
+      "learning_rate": 1.2495120913510793e-05,
+      "loss": 1.0706,
+      "step": 15848
+    },
+    {
+      "epoch": 2.53584,
+      "grad_norm": 1.9178140604524132,
+      "learning_rate": 1.2486761082361586e-05,
+      "loss": 1.0682,
+      "step": 15849
+    },
+    {
+      "epoch": 2.536,
+      "grad_norm": 1.5060194177610602,
+      "learning_rate": 1.2478403862479881e-05,
+      "loss": 1.0298,
+      "step": 15850
+    },
+    {
+      "epoch": 2.5361599999999997,
+      "grad_norm": 2.2448715729835658,
+      "learning_rate": 1.2470049254115112e-05,
+      "loss": 1.0933,
+      "step": 15851
+    },
+    {
+      "epoch": 2.53632,
+      "grad_norm": 1.8173315041711744,
+      "learning_rate": 1.2461697257516491e-05,
+      "loss": 1.0742,
+      "step": 15852
+    },
+    {
+      "epoch": 2.53648,
+      "grad_norm": 1.580837963374637,
+      "learning_rate": 1.2453347872933285e-05,
+      "loss": 1.0535,
+      "step": 15853
+    },
+    {
+      "epoch": 2.5366400000000002,
+      "grad_norm": 1.8686953132987894,
+      "learning_rate": 1.2445001100614617e-05,
+      "loss": 1.0851,
+      "step": 15854
+    },
+    {
+      "epoch": 2.5368,
+      "grad_norm": 1.9051368526240209,
+      "learning_rate": 1.2436656940809522e-05,
+      "loss": 1.0736,
+      "step": 15855
+    },
+    {
+      "epoch": 2.53696,
+      "grad_norm": 1.8048133740241508,
+      "learning_rate": 1.2428315393766998e-05,
+      "loss": 1.0803,
+      "step": 15856
+    },
+    {
+      "epoch": 2.53712,
+      "grad_norm": 1.6296325793402435,
+      "learning_rate": 1.2419976459735939e-05,
+      "loss": 1.1158,
+      "step": 15857
+    },
+    {
+      "epoch": 2.53728,
+      "grad_norm": 1.951053751306815,
+      "learning_rate": 1.2411640138965163e-05,
+      "loss": 1.0667,
+      "step": 15858
+    },
+    {
+      "epoch": 2.53744,
+      "grad_norm": 1.7164357049777246,
+      "learning_rate": 1.2403306431703422e-05,
+      "loss": 1.0769,
+      "step": 15859
+    },
+    {
+      "epoch": 2.5376,
+      "grad_norm": 1.7314107689945573,
+      "learning_rate": 1.2394975338199365e-05,
+      "loss": 1.0642,
+      "step": 15860
+    },
+    {
+      "epoch": 2.53776,
+      "grad_norm": 1.7983082578336314,
+      "learning_rate": 1.2386646858701589e-05,
+      "loss": 1.06,
+      "step": 15861
+    },
+    {
+      "epoch": 2.5379199999999997,
+      "grad_norm": 1.6619408249703607,
+      "learning_rate": 1.2378320993458592e-05,
+      "loss": 1.0503,
+      "step": 15862
+    },
+    {
+      "epoch": 2.53808,
+      "grad_norm": 1.9056067997219432,
+      "learning_rate": 1.2369997742718853e-05,
+      "loss": 1.126,
+      "step": 15863
+    },
+    {
+      "epoch": 2.53824,
+      "grad_norm": 1.7760883879802951,
+      "learning_rate": 1.2361677106730652e-05,
+      "loss": 1.0854,
+      "step": 15864
+    },
+    {
+      "epoch": 2.5384,
+      "grad_norm": 1.8480090407685468,
+      "learning_rate": 1.2353359085742344e-05,
+      "loss": 1.0973,
+      "step": 15865
+    },
+    {
+      "epoch": 2.53856,
+      "grad_norm": 1.8145349455476065,
+      "learning_rate": 1.2345043680002055e-05,
+      "loss": 1.1334,
+      "step": 15866
+    },
+    {
+      "epoch": 2.53872,
+      "grad_norm": 1.89924373836408,
+      "learning_rate": 1.2336730889757964e-05,
+      "loss": 1.0808,
+      "step": 15867
+    },
+    {
+      "epoch": 2.53888,
+      "grad_norm": 1.742920525853044,
+      "learning_rate": 1.232842071525805e-05,
+      "loss": 1.0307,
+      "step": 15868
+    },
+    {
+      "epoch": 2.53904,
+      "grad_norm": 2.0149444018807916,
+      "learning_rate": 1.2320113156750323e-05,
+      "loss": 1.162,
+      "step": 15869
+    },
+    {
+      "epoch": 2.5392,
+      "grad_norm": 1.688070689504765,
+      "learning_rate": 1.2311808214482667e-05,
+      "loss": 1.0587,
+      "step": 15870
+    },
+    {
+      "epoch": 2.53936,
+      "grad_norm": 1.7229594043952396,
+      "learning_rate": 1.230350588870287e-05,
+      "loss": 1.0589,
+      "step": 15871
+    },
+    {
+      "epoch": 2.53952,
+      "grad_norm": 1.713437775458759,
+      "learning_rate": 1.229520617965868e-05,
+      "loss": 1.0384,
+      "step": 15872
+    },
+    {
+      "epoch": 2.53968,
+      "grad_norm": 1.7037716794737832,
+      "learning_rate": 1.228690908759773e-05,
+      "loss": 1.1124,
+      "step": 15873
+    },
+    {
+      "epoch": 2.53984,
+      "grad_norm": 1.9851585945262649,
+      "learning_rate": 1.2278614612767603e-05,
+      "loss": 1.1011,
+      "step": 15874
+    },
+    {
+      "epoch": 2.54,
+      "grad_norm": 1.646863303348535,
+      "learning_rate": 1.2270322755415797e-05,
+      "loss": 1.0907,
+      "step": 15875
+    },
+    {
+      "epoch": 2.54016,
+      "grad_norm": 1.5998225758672442,
+      "learning_rate": 1.2262033515789718e-05,
+      "loss": 1.0876,
+      "step": 15876
+    },
+    {
+      "epoch": 2.54032,
+      "grad_norm": 1.747861970846681,
+      "learning_rate": 1.2253746894136709e-05,
+      "loss": 1.0614,
+      "step": 15877
+    },
+    {
+      "epoch": 2.54048,
+      "grad_norm": 1.767151910542062,
+      "learning_rate": 1.2245462890704019e-05,
+      "loss": 1.0924,
+      "step": 15878
+    },
+    {
+      "epoch": 2.54064,
+      "grad_norm": 1.727528250599725,
+      "learning_rate": 1.2237181505738882e-05,
+      "loss": 1.0546,
+      "step": 15879
+    },
+    {
+      "epoch": 2.5408,
+      "grad_norm": 1.7615752342675428,
+      "learning_rate": 1.2228902739488324e-05,
+      "loss": 1.0288,
+      "step": 15880
+    },
+    {
+      "epoch": 2.54096,
+      "grad_norm": 1.9240814019342523,
+      "learning_rate": 1.2220626592199457e-05,
+      "loss": 1.1201,
+      "step": 15881
+    },
+    {
+      "epoch": 2.5411200000000003,
+      "grad_norm": 1.737134994170106,
+      "learning_rate": 1.2212353064119142e-05,
+      "loss": 1.0921,
+      "step": 15882
+    },
+    {
+      "epoch": 2.54128,
+      "grad_norm": 1.5764710389090244,
+      "learning_rate": 1.2204082155494334e-05,
+      "loss": 1.0738,
+      "step": 15883
+    },
+    {
+      "epoch": 2.54144,
+      "grad_norm": 1.8940921321326971,
+      "learning_rate": 1.2195813866571737e-05,
+      "loss": 1.0672,
+      "step": 15884
+    },
+    {
+      "epoch": 2.5416,
+      "grad_norm": 1.6286169611301755,
+      "learning_rate": 1.218754819759813e-05,
+      "loss": 1.0684,
+      "step": 15885
+    },
+    {
+      "epoch": 2.54176,
+      "grad_norm": 1.6979132497073615,
+      "learning_rate": 1.2179285148820129e-05,
+      "loss": 1.1048,
+      "step": 15886
+    },
+    {
+      "epoch": 2.54192,
+      "grad_norm": 1.5963389605694922,
+      "learning_rate": 1.2171024720484292e-05,
+      "loss": 1.0804,
+      "step": 15887
+    },
+    {
+      "epoch": 2.54208,
+      "grad_norm": 1.8695340206885998,
+      "learning_rate": 1.2162766912837097e-05,
+      "loss": 1.1251,
+      "step": 15888
+    },
+    {
+      "epoch": 2.54224,
+      "grad_norm": 1.7725698567978299,
+      "learning_rate": 1.215451172612495e-05,
+      "loss": 1.092,
+      "step": 15889
+    },
+    {
+      "epoch": 2.5423999999999998,
+      "grad_norm": 1.6789358357944764,
+      "learning_rate": 1.2146259160594164e-05,
+      "loss": 1.091,
+      "step": 15890
+    },
+    {
+      "epoch": 2.54256,
+      "grad_norm": 1.6659990619101483,
+      "learning_rate": 1.2138009216490987e-05,
+      "loss": 1.0953,
+      "step": 15891
+    },
+    {
+      "epoch": 2.54272,
+      "grad_norm": 1.83765809494223,
+      "learning_rate": 1.2129761894061587e-05,
+      "loss": 1.0882,
+      "step": 15892
+    },
+    {
+      "epoch": 2.5428800000000003,
+      "grad_norm": 1.876945301470915,
+      "learning_rate": 1.2121517193552057e-05,
+      "loss": 1.1007,
+      "step": 15893
+    },
+    {
+      "epoch": 2.54304,
+      "grad_norm": 1.7814279450881183,
+      "learning_rate": 1.2113275115208378e-05,
+      "loss": 1.0648,
+      "step": 15894
+    },
+    {
+      "epoch": 2.5432,
+      "grad_norm": 1.6776667126971112,
+      "learning_rate": 1.2105035659276542e-05,
+      "loss": 1.0712,
+      "step": 15895
+    },
+    {
+      "epoch": 2.54336,
+      "grad_norm": 1.8738574267385735,
+      "learning_rate": 1.209679882600233e-05,
+      "loss": 1.1113,
+      "step": 15896
+    },
+    {
+      "epoch": 2.54352,
+      "grad_norm": 1.7294487034171462,
+      "learning_rate": 1.2088564615631592e-05,
+      "loss": 1.0345,
+      "step": 15897
+    },
+    {
+      "epoch": 2.54368,
+      "grad_norm": 1.7246483505928523,
+      "learning_rate": 1.208033302840994e-05,
+      "loss": 1.0899,
+      "step": 15898
+    },
+    {
+      "epoch": 2.54384,
+      "grad_norm": 1.7165474748356238,
+      "learning_rate": 1.207210406458308e-05,
+      "loss": 1.1479,
+      "step": 15899
+    },
+    {
+      "epoch": 2.544,
+      "grad_norm": 2.0104219157466527,
+      "learning_rate": 1.2063877724396466e-05,
+      "loss": 1.106,
+      "step": 15900
+    },
+    {
+      "epoch": 2.5441599999999998,
+      "grad_norm": 1.9425046962712658,
+      "learning_rate": 1.2055654008095629e-05,
+      "loss": 1.1081,
+      "step": 15901
+    },
+    {
+      "epoch": 2.54432,
+      "grad_norm": 1.813409934745918,
+      "learning_rate": 1.2047432915925938e-05,
+      "loss": 1.0914,
+      "step": 15902
+    },
+    {
+      "epoch": 2.54448,
+      "grad_norm": 1.776210759961322,
+      "learning_rate": 1.2039214448132674e-05,
+      "loss": 1.0868,
+      "step": 15903
+    },
+    {
+      "epoch": 2.5446400000000002,
+      "grad_norm": 2.0203614644658847,
+      "learning_rate": 1.2030998604961085e-05,
+      "loss": 1.0688,
+      "step": 15904
+    },
+    {
+      "epoch": 2.5448,
+      "grad_norm": 1.5386942620114628,
+      "learning_rate": 1.202278538665631e-05,
+      "loss": 1.1328,
+      "step": 15905
+    },
+    {
+      "epoch": 2.54496,
+      "grad_norm": 1.6222882622290273,
+      "learning_rate": 1.2014574793463428e-05,
+      "loss": 1.0775,
+      "step": 15906
+    },
+    {
+      "epoch": 2.54512,
+      "grad_norm": 1.6865097776874065,
+      "learning_rate": 1.2006366825627425e-05,
+      "loss": 1.0831,
+      "step": 15907
+    },
+    {
+      "epoch": 2.54528,
+      "grad_norm": 1.8517414141016921,
+      "learning_rate": 1.199816148339321e-05,
+      "loss": 1.1145,
+      "step": 15908
+    },
+    {
+      "epoch": 2.54544,
+      "grad_norm": 1.6840399477193246,
+      "learning_rate": 1.1989958767005605e-05,
+      "loss": 1.0417,
+      "step": 15909
+    },
+    {
+      "epoch": 2.5456,
+      "grad_norm": 1.848956355782471,
+      "learning_rate": 1.1981758676709409e-05,
+      "loss": 1.0909,
+      "step": 15910
+    },
+    {
+      "epoch": 2.54576,
+      "grad_norm": 1.777958559665139,
+      "learning_rate": 1.1973561212749274e-05,
+      "loss": 1.096,
+      "step": 15911
+    },
+    {
+      "epoch": 2.5459199999999997,
+      "grad_norm": 1.9807145350674635,
+      "learning_rate": 1.1965366375369812e-05,
+      "loss": 1.1109,
+      "step": 15912
+    },
+    {
+      "epoch": 2.54608,
+      "grad_norm": 1.8297632874630403,
+      "learning_rate": 1.1957174164815532e-05,
+      "loss": 1.0635,
+      "step": 15913
+    },
+    {
+      "epoch": 2.54624,
+      "grad_norm": 1.7977881091196624,
+      "learning_rate": 1.194898458133088e-05,
+      "loss": 1.0545,
+      "step": 15914
+    },
+    {
+      "epoch": 2.5464,
+      "grad_norm": 1.5893539039716547,
+      "learning_rate": 1.1940797625160227e-05,
+      "loss": 1.046,
+      "step": 15915
+    },
+    {
+      "epoch": 2.54656,
+      "grad_norm": 1.7241711469818604,
+      "learning_rate": 1.1932613296547868e-05,
+      "loss": 1.0613,
+      "step": 15916
+    },
+    {
+      "epoch": 2.54672,
+      "grad_norm": 1.7086160397988226,
+      "learning_rate": 1.1924431595737984e-05,
+      "loss": 1.0714,
+      "step": 15917
+    },
+    {
+      "epoch": 2.54688,
+      "grad_norm": 1.4679337164700221,
+      "learning_rate": 1.1916252522974735e-05,
+      "loss": 1.071,
+      "step": 15918
+    },
+    {
+      "epoch": 2.54704,
+      "grad_norm": 1.738473180943272,
+      "learning_rate": 1.1908076078502151e-05,
+      "loss": 1.0647,
+      "step": 15919
+    },
+    {
+      "epoch": 2.5472,
+      "grad_norm": 1.5735740928185573,
+      "learning_rate": 1.1899902262564211e-05,
+      "loss": 1.1387,
+      "step": 15920
+    },
+    {
+      "epoch": 2.54736,
+      "grad_norm": 1.8102887529040965,
+      "learning_rate": 1.1891731075404799e-05,
+      "loss": 1.092,
+      "step": 15921
+    },
+    {
+      "epoch": 2.54752,
+      "grad_norm": 1.7501572984523852,
+      "learning_rate": 1.1883562517267776e-05,
+      "loss": 1.0944,
+      "step": 15922
+    },
+    {
+      "epoch": 2.5476799999999997,
+      "grad_norm": 1.642458708870663,
+      "learning_rate": 1.1875396588396814e-05,
+      "loss": 1.075,
+      "step": 15923
+    },
+    {
+      "epoch": 2.54784,
+      "grad_norm": 1.9325910371254598,
+      "learning_rate": 1.186723328903565e-05,
+      "loss": 1.1391,
+      "step": 15924
+    },
+    {
+      "epoch": 2.548,
+      "grad_norm": 1.8299886767174027,
+      "learning_rate": 1.1859072619427769e-05,
+      "loss": 1.039,
+      "step": 15925
+    },
+    {
+      "epoch": 2.54816,
+      "grad_norm": 1.7404411438472565,
+      "learning_rate": 1.1850914579816763e-05,
+      "loss": 1.0637,
+      "step": 15926
+    },
+    {
+      "epoch": 2.54832,
+      "grad_norm": 1.8132607976009774,
+      "learning_rate": 1.1842759170446005e-05,
+      "loss": 1.0754,
+      "step": 15927
+    },
+    {
+      "epoch": 2.54848,
+      "grad_norm": 1.8433447206358642,
+      "learning_rate": 1.1834606391558867e-05,
+      "loss": 1.0725,
+      "step": 15928
+    },
+    {
+      "epoch": 2.54864,
+      "grad_norm": 1.8656214066445418,
+      "learning_rate": 1.1826456243398598e-05,
+      "loss": 1.0663,
+      "step": 15929
+    },
+    {
+      "epoch": 2.5488,
+      "grad_norm": 1.5182381719807927,
+      "learning_rate": 1.1818308726208393e-05,
+      "loss": 1.1008,
+      "step": 15930
+    },
+    {
+      "epoch": 2.54896,
+      "grad_norm": 1.8414154024443175,
+      "learning_rate": 1.1810163840231359e-05,
+      "loss": 1.0805,
+      "step": 15931
+    },
+    {
+      "epoch": 2.5491200000000003,
+      "grad_norm": 1.8222336735957496,
+      "learning_rate": 1.1802021585710532e-05,
+      "loss": 1.0954,
+      "step": 15932
+    },
+    {
+      "epoch": 2.54928,
+      "grad_norm": 1.9040157028008307,
+      "learning_rate": 1.1793881962888854e-05,
+      "loss": 1.103,
+      "step": 15933
+    },
+    {
+      "epoch": 2.54944,
+      "grad_norm": 1.838029013334733,
+      "learning_rate": 1.1785744972009216e-05,
+      "loss": 1.0812,
+      "step": 15934
+    },
+    {
+      "epoch": 2.5496,
+      "grad_norm": 1.5632712025421682,
+      "learning_rate": 1.1777610613314405e-05,
+      "loss": 1.0508,
+      "step": 15935
+    },
+    {
+      "epoch": 2.54976,
+      "grad_norm": 1.4129794603253603,
+      "learning_rate": 1.1769478887047147e-05,
+      "loss": 1.0405,
+      "step": 15936
+    },
+    {
+      "epoch": 2.54992,
+      "grad_norm": 1.6085180890119481,
+      "learning_rate": 1.1761349793450049e-05,
+      "loss": 1.0827,
+      "step": 15937
+    },
+    {
+      "epoch": 2.55008,
+      "grad_norm": 1.6884001342399397,
+      "learning_rate": 1.1753223332765729e-05,
+      "loss": 1.1578,
+      "step": 15938
+    },
+    {
+      "epoch": 2.55024,
+      "grad_norm": 1.6749470569871956,
+      "learning_rate": 1.17450995052366e-05,
+      "loss": 1.0827,
+      "step": 15939
+    },
+    {
+      "epoch": 2.5504,
+      "grad_norm": 1.7887070413452575,
+      "learning_rate": 1.173697831110514e-05,
+      "loss": 1.0928,
+      "step": 15940
+    },
+    {
+      "epoch": 2.55056,
+      "grad_norm": 1.647055074971043,
+      "learning_rate": 1.1728859750613596e-05,
+      "loss": 1.0875,
+      "step": 15941
+    },
+    {
+      "epoch": 2.55072,
+      "grad_norm": 1.620636453054706,
+      "learning_rate": 1.1720743824004277e-05,
+      "loss": 1.0951,
+      "step": 15942
+    },
+    {
+      "epoch": 2.5508800000000003,
+      "grad_norm": 1.5406927671726163,
+      "learning_rate": 1.171263053151932e-05,
+      "loss": 1.0941,
+      "step": 15943
+    },
+    {
+      "epoch": 2.55104,
+      "grad_norm": 1.7630162650329022,
+      "learning_rate": 1.1704519873400822e-05,
+      "loss": 1.0845,
+      "step": 15944
+    },
+    {
+      "epoch": 2.5512,
+      "grad_norm": 1.6704737657917912,
+      "learning_rate": 1.1696411849890788e-05,
+      "loss": 1.0926,
+      "step": 15945
+    },
+    {
+      "epoch": 2.55136,
+      "grad_norm": 1.6913387664740003,
+      "learning_rate": 1.1688306461231157e-05,
+      "loss": 1.0732,
+      "step": 15946
+    },
+    {
+      "epoch": 2.55152,
+      "grad_norm": 1.6685257156856212,
+      "learning_rate": 1.1680203707663773e-05,
+      "loss": 1.0678,
+      "step": 15947
+    },
+    {
+      "epoch": 2.55168,
+      "grad_norm": 1.6972063437760188,
+      "learning_rate": 1.1672103589430417e-05,
+      "loss": 1.102,
+      "step": 15948
+    },
+    {
+      "epoch": 2.55184,
+      "grad_norm": 1.4427791729663206,
+      "learning_rate": 1.1664006106772784e-05,
+      "loss": 1.0637,
+      "step": 15949
+    },
+    {
+      "epoch": 2.552,
+      "grad_norm": 1.785217425897302,
+      "learning_rate": 1.1655911259932483e-05,
+      "loss": 1.0891,
+      "step": 15950
+    },
+    {
+      "epoch": 2.5521599999999998,
+      "grad_norm": 1.658468424314127,
+      "learning_rate": 1.1647819049151043e-05,
+      "loss": 1.063,
+      "step": 15951
+    },
+    {
+      "epoch": 2.55232,
+      "grad_norm": 1.7194306899014693,
+      "learning_rate": 1.163972947466998e-05,
+      "loss": 1.155,
+      "step": 15952
+    },
+    {
+      "epoch": 2.55248,
+      "grad_norm": 1.6748563615404517,
+      "learning_rate": 1.1631642536730603e-05,
+      "loss": 1.111,
+      "step": 15953
+    },
+    {
+      "epoch": 2.5526400000000002,
+      "grad_norm": 1.9137004438252847,
+      "learning_rate": 1.1623558235574272e-05,
+      "loss": 1.0912,
+      "step": 15954
+    },
+    {
+      "epoch": 2.5528,
+      "grad_norm": 1.7697914552084473,
+      "learning_rate": 1.1615476571442152e-05,
+      "loss": 1.0962,
+      "step": 15955
+    },
+    {
+      "epoch": 2.55296,
+      "grad_norm": 1.8286863320711133,
+      "learning_rate": 1.1607397544575471e-05,
+      "loss": 1.0772,
+      "step": 15956
+    },
+    {
+      "epoch": 2.55312,
+      "grad_norm": 1.573451339883501,
+      "learning_rate": 1.1599321155215193e-05,
+      "loss": 1.0363,
+      "step": 15957
+    },
+    {
+      "epoch": 2.55328,
+      "grad_norm": 1.6297106058249828,
+      "learning_rate": 1.1591247403602379e-05,
+      "loss": 1.0661,
+      "step": 15958
+    },
+    {
+      "epoch": 2.55344,
+      "grad_norm": 1.6638066383230234,
+      "learning_rate": 1.1583176289977915e-05,
+      "loss": 1.0343,
+      "step": 15959
+    },
+    {
+      "epoch": 2.5536,
+      "grad_norm": 1.6640317097710757,
+      "learning_rate": 1.1575107814582642e-05,
+      "loss": 1.0501,
+      "step": 15960
+    },
+    {
+      "epoch": 2.55376,
+      "grad_norm": 1.5799004555391871,
+      "learning_rate": 1.15670419776573e-05,
+      "loss": 1.0272,
+      "step": 15961
+    },
+    {
+      "epoch": 2.5539199999999997,
+      "grad_norm": 1.9350677684066617,
+      "learning_rate": 1.1558978779442565e-05,
+      "loss": 1.113,
+      "step": 15962
+    },
+    {
+      "epoch": 2.55408,
+      "grad_norm": 1.7979591999404452,
+      "learning_rate": 1.155091822017903e-05,
+      "loss": 1.0539,
+      "step": 15963
+    },
+    {
+      "epoch": 2.55424,
+      "grad_norm": 1.8057437554919107,
+      "learning_rate": 1.1542860300107217e-05,
+      "loss": 1.058,
+      "step": 15964
+    },
+    {
+      "epoch": 2.5544000000000002,
+      "grad_norm": 1.823981112053837,
+      "learning_rate": 1.1534805019467543e-05,
+      "loss": 1.0258,
+      "step": 15965
+    },
+    {
+      "epoch": 2.55456,
+      "grad_norm": 1.6869553179083596,
+      "learning_rate": 1.1526752378500394e-05,
+      "loss": 1.078,
+      "step": 15966
+    },
+    {
+      "epoch": 2.55472,
+      "grad_norm": 1.5605039750769178,
+      "learning_rate": 1.1518702377446012e-05,
+      "loss": 1.071,
+      "step": 15967
+    },
+    {
+      "epoch": 2.55488,
+      "grad_norm": 1.8324969319293785,
+      "learning_rate": 1.1510655016544635e-05,
+      "loss": 1.0605,
+      "step": 15968
+    },
+    {
+      "epoch": 2.55504,
+      "grad_norm": 1.7275052689598018,
+      "learning_rate": 1.1502610296036377e-05,
+      "loss": 1.045,
+      "step": 15969
+    },
+    {
+      "epoch": 2.5552,
+      "grad_norm": 2.0468870702533355,
+      "learning_rate": 1.149456821616126e-05,
+      "loss": 1.0958,
+      "step": 15970
+    },
+    {
+      "epoch": 2.55536,
+      "grad_norm": 1.7152884625602751,
+      "learning_rate": 1.1486528777159267e-05,
+      "loss": 1.0908,
+      "step": 15971
+    },
+    {
+      "epoch": 2.55552,
+      "grad_norm": 1.7584351627232695,
+      "learning_rate": 1.147849197927029e-05,
+      "loss": 1.053,
+      "step": 15972
+    },
+    {
+      "epoch": 2.5556799999999997,
+      "grad_norm": 1.6589029564368476,
+      "learning_rate": 1.1470457822734103e-05,
+      "loss": 1.0795,
+      "step": 15973
+    },
+    {
+      "epoch": 2.55584,
+      "grad_norm": 1.537467245097855,
+      "learning_rate": 1.1462426307790475e-05,
+      "loss": 1.0838,
+      "step": 15974
+    },
+    {
+      "epoch": 2.556,
+      "grad_norm": 2.0571377893460263,
+      "learning_rate": 1.1454397434679021e-05,
+      "loss": 1.186,
+      "step": 15975
+    },
+    {
+      "epoch": 2.55616,
+      "grad_norm": 1.9255633523118736,
+      "learning_rate": 1.1446371203639316e-05,
+      "loss": 1.1178,
+      "step": 15976
+    },
+    {
+      "epoch": 2.55632,
+      "grad_norm": 1.6212038591811226,
+      "learning_rate": 1.1438347614910872e-05,
+      "loss": 1.0766,
+      "step": 15977
+    },
+    {
+      "epoch": 2.55648,
+      "grad_norm": 1.7943648595361574,
+      "learning_rate": 1.143032666873306e-05,
+      "loss": 1.1344,
+      "step": 15978
+    },
+    {
+      "epoch": 2.55664,
+      "grad_norm": 1.7720111318869602,
+      "learning_rate": 1.142230836534528e-05,
+      "loss": 1.0533,
+      "step": 15979
+    },
+    {
+      "epoch": 2.5568,
+      "grad_norm": 1.5104490108589887,
+      "learning_rate": 1.1414292704986706e-05,
+      "loss": 1.0576,
+      "step": 15980
+    },
+    {
+      "epoch": 2.55696,
+      "grad_norm": 1.5609974251822631,
+      "learning_rate": 1.1406279687896604e-05,
+      "loss": 1.0913,
+      "step": 15981
+    },
+    {
+      "epoch": 2.55712,
+      "grad_norm": 1.7695373884454049,
+      "learning_rate": 1.139826931431397e-05,
+      "loss": 1.1327,
+      "step": 15982
+    },
+    {
+      "epoch": 2.55728,
+      "grad_norm": 1.6936039099762266,
+      "learning_rate": 1.13902615844779e-05,
+      "loss": 1.0548,
+      "step": 15983
+    },
+    {
+      "epoch": 2.55744,
+      "grad_norm": 1.662559501347381,
+      "learning_rate": 1.1382256498627309e-05,
+      "loss": 1.0777,
+      "step": 15984
+    },
+    {
+      "epoch": 2.5576,
+      "grad_norm": 1.802846480763755,
+      "learning_rate": 1.1374254057001043e-05,
+      "loss": 1.0646,
+      "step": 15985
+    },
+    {
+      "epoch": 2.55776,
+      "grad_norm": 1.726473854745151,
+      "learning_rate": 1.1366254259837905e-05,
+      "loss": 1.0489,
+      "step": 15986
+    },
+    {
+      "epoch": 2.55792,
+      "grad_norm": 1.6898614342645604,
+      "learning_rate": 1.135825710737658e-05,
+      "loss": 1.0718,
+      "step": 15987
+    },
+    {
+      "epoch": 2.55808,
+      "grad_norm": 1.4304799003066562,
+      "learning_rate": 1.1350262599855699e-05,
+      "loss": 1.0487,
+      "step": 15988
+    },
+    {
+      "epoch": 2.55824,
+      "grad_norm": 1.9235405453523635,
+      "learning_rate": 1.1342270737513816e-05,
+      "loss": 1.1096,
+      "step": 15989
+    },
+    {
+      "epoch": 2.5584,
+      "grad_norm": 1.9278966839116394,
+      "learning_rate": 1.1334281520589374e-05,
+      "loss": 1.0779,
+      "step": 15990
+    },
+    {
+      "epoch": 2.55856,
+      "grad_norm": 1.7080248850814497,
+      "learning_rate": 1.1326294949320782e-05,
+      "loss": 1.0357,
+      "step": 15991
+    },
+    {
+      "epoch": 2.55872,
+      "grad_norm": 1.6775524426742943,
+      "learning_rate": 1.1318311023946326e-05,
+      "loss": 1.0267,
+      "step": 15992
+    },
+    {
+      "epoch": 2.5588800000000003,
+      "grad_norm": 1.6789357257538675,
+      "learning_rate": 1.131032974470425e-05,
+      "loss": 1.0576,
+      "step": 15993
+    },
+    {
+      "epoch": 2.55904,
+      "grad_norm": 2.0983993316272005,
+      "learning_rate": 1.1302351111832687e-05,
+      "loss": 1.0757,
+      "step": 15994
+    },
+    {
+      "epoch": 2.5592,
+      "grad_norm": 1.942835041960312,
+      "learning_rate": 1.1294375125569756e-05,
+      "loss": 1.0549,
+      "step": 15995
+    },
+    {
+      "epoch": 2.55936,
+      "grad_norm": 1.7688461354644602,
+      "learning_rate": 1.1286401786153366e-05,
+      "loss": 1.0679,
+      "step": 15996
+    },
+    {
+      "epoch": 2.55952,
+      "grad_norm": 1.9041004275205984,
+      "learning_rate": 1.1278431093821528e-05,
+      "loss": 1.1321,
+      "step": 15997
+    },
+    {
+      "epoch": 2.55968,
+      "grad_norm": 1.6290743707416682,
+      "learning_rate": 1.1270463048811985e-05,
+      "loss": 1.0437,
+      "step": 15998
+    },
+    {
+      "epoch": 2.55984,
+      "grad_norm": 1.5618797379634333,
+      "learning_rate": 1.1262497651362547e-05,
+      "loss": 1.059,
+      "step": 15999
+    },
+    {
+      "epoch": 2.56,
+      "grad_norm": 1.637134515109613,
+      "learning_rate": 1.1254534901710877e-05,
+      "loss": 1.075,
+      "step": 16000
+    },
+    {
+      "epoch": 2.5601599999999998,
+      "grad_norm": 1.8784225772664045,
+      "learning_rate": 1.1246574800094566e-05,
+      "loss": 1.0659,
+      "step": 16001
+    },
+    {
+      "epoch": 2.56032,
+      "grad_norm": 1.7388217284935266,
+      "learning_rate": 1.1238617346751134e-05,
+      "loss": 1.0802,
+      "step": 16002
+    },
+    {
+      "epoch": 2.56048,
+      "grad_norm": 1.8303186405268614,
+      "learning_rate": 1.1230662541918036e-05,
+      "loss": 1.0791,
+      "step": 16003
+    },
+    {
+      "epoch": 2.5606400000000002,
+      "grad_norm": 1.7050271600310152,
+      "learning_rate": 1.1222710385832602e-05,
+      "loss": 1.0529,
+      "step": 16004
+    },
+    {
+      "epoch": 2.5608,
+      "grad_norm": 1.8696543573457038,
+      "learning_rate": 1.1214760878732134e-05,
+      "loss": 1.0758,
+      "step": 16005
+    },
+    {
+      "epoch": 2.56096,
+      "grad_norm": 1.6496499977582393,
+      "learning_rate": 1.1206814020853828e-05,
+      "loss": 1.0641,
+      "step": 16006
+    },
+    {
+      "epoch": 2.56112,
+      "grad_norm": 2.2425964572902974,
+      "learning_rate": 1.119886981243481e-05,
+      "loss": 1.1231,
+      "step": 16007
+    },
+    {
+      "epoch": 2.56128,
+      "grad_norm": 1.6121433837387549,
+      "learning_rate": 1.1190928253712107e-05,
+      "loss": 1.0431,
+      "step": 16008
+    },
+    {
+      "epoch": 2.56144,
+      "grad_norm": 1.706562730389277,
+      "learning_rate": 1.1182989344922712e-05,
+      "loss": 1.0632,
+      "step": 16009
+    },
+    {
+      "epoch": 2.5616,
+      "grad_norm": 1.5401974566319467,
+      "learning_rate": 1.1175053086303466e-05,
+      "loss": 1.0841,
+      "step": 16010
+    },
+    {
+      "epoch": 2.56176,
+      "grad_norm": 1.7253349772028796,
+      "learning_rate": 1.1167119478091237e-05,
+      "loss": 1.0858,
+      "step": 16011
+    },
+    {
+      "epoch": 2.5619199999999998,
+      "grad_norm": 1.7240061991499105,
+      "learning_rate": 1.115918852052269e-05,
+      "loss": 1.0974,
+      "step": 16012
+    },
+    {
+      "epoch": 2.56208,
+      "grad_norm": 1.6698064038441667,
+      "learning_rate": 1.1151260213834547e-05,
+      "loss": 1.0696,
+      "step": 16013
+    },
+    {
+      "epoch": 2.56224,
+      "grad_norm": 2.0348884293058562,
+      "learning_rate": 1.1143334558263285e-05,
+      "loss": 1.0869,
+      "step": 16014
+    },
+    {
+      "epoch": 2.5624000000000002,
+      "grad_norm": 2.090299671190941,
+      "learning_rate": 1.113541155404546e-05,
+      "loss": 1.0862,
+      "step": 16015
+    },
+    {
+      "epoch": 2.56256,
+      "grad_norm": 1.6116069861620201,
+      "learning_rate": 1.1127491201417472e-05,
+      "loss": 1.0914,
+      "step": 16016
+    },
+    {
+      "epoch": 2.56272,
+      "grad_norm": 1.5311300785919164,
+      "learning_rate": 1.1119573500615642e-05,
+      "loss": 1.0791,
+      "step": 16017
+    },
+    {
+      "epoch": 2.56288,
+      "grad_norm": 1.6593230448523928,
+      "learning_rate": 1.1111658451876228e-05,
+      "loss": 1.0946,
+      "step": 16018
+    },
+    {
+      "epoch": 2.56304,
+      "grad_norm": 1.685935284130276,
+      "learning_rate": 1.1103746055435394e-05,
+      "loss": 1.0278,
+      "step": 16019
+    },
+    {
+      "epoch": 2.5632,
+      "grad_norm": 1.8155552630896257,
+      "learning_rate": 1.1095836311529251e-05,
+      "loss": 1.0532,
+      "step": 16020
+    },
+    {
+      "epoch": 2.56336,
+      "grad_norm": 1.7309503695438138,
+      "learning_rate": 1.10879292203938e-05,
+      "loss": 1.0697,
+      "step": 16021
+    },
+    {
+      "epoch": 2.56352,
+      "grad_norm": 1.7696625533838,
+      "learning_rate": 1.1080024782264986e-05,
+      "loss": 1.1429,
+      "step": 16022
+    },
+    {
+      "epoch": 2.5636799999999997,
+      "grad_norm": 1.765013758700164,
+      "learning_rate": 1.1072122997378665e-05,
+      "loss": 1.0478,
+      "step": 16023
+    },
+    {
+      "epoch": 2.56384,
+      "grad_norm": 1.6817615197503215,
+      "learning_rate": 1.1064223865970603e-05,
+      "loss": 1.0496,
+      "step": 16024
+    },
+    {
+      "epoch": 2.564,
+      "grad_norm": 1.862196142239124,
+      "learning_rate": 1.10563273882765e-05,
+      "loss": 1.1005,
+      "step": 16025
+    },
+    {
+      "epoch": 2.56416,
+      "grad_norm": 1.7588691000943293,
+      "learning_rate": 1.104843356453199e-05,
+      "loss": 1.0757,
+      "step": 16026
+    },
+    {
+      "epoch": 2.56432,
+      "grad_norm": 1.8716224134665693,
+      "learning_rate": 1.1040542394972619e-05,
+      "loss": 1.0458,
+      "step": 16027
+    },
+    {
+      "epoch": 2.56448,
+      "grad_norm": 1.9888146337949764,
+      "learning_rate": 1.1032653879833832e-05,
+      "loss": 1.0741,
+      "step": 16028
+    },
+    {
+      "epoch": 2.56464,
+      "grad_norm": 1.7287260695181816,
+      "learning_rate": 1.1024768019351005e-05,
+      "loss": 1.0905,
+      "step": 16029
+    },
+    {
+      "epoch": 2.5648,
+      "grad_norm": 1.8709624088731154,
+      "learning_rate": 1.1016884813759454e-05,
+      "loss": 1.0602,
+      "step": 16030
+    },
+    {
+      "epoch": 2.56496,
+      "grad_norm": 1.7030134693676418,
+      "learning_rate": 1.100900426329441e-05,
+      "loss": 1.0534,
+      "step": 16031
+    },
+    {
+      "epoch": 2.56512,
+      "grad_norm": 1.851016510283473,
+      "learning_rate": 1.1001126368190995e-05,
+      "loss": 1.0672,
+      "step": 16032
+    },
+    {
+      "epoch": 2.56528,
+      "grad_norm": 1.9203109400205653,
+      "learning_rate": 1.0993251128684279e-05,
+      "loss": 1.0684,
+      "step": 16033
+    },
+    {
+      "epoch": 2.56544,
+      "grad_norm": 1.605874004636067,
+      "learning_rate": 1.0985378545009262e-05,
+      "loss": 1.049,
+      "step": 16034
+    },
+    {
+      "epoch": 2.5656,
+      "grad_norm": 1.9812847355823024,
+      "learning_rate": 1.097750861740081e-05,
+      "loss": 1.0842,
+      "step": 16035
+    },
+    {
+      "epoch": 2.56576,
+      "grad_norm": 1.9085610163080051,
+      "learning_rate": 1.0969641346093828e-05,
+      "loss": 1.0971,
+      "step": 16036
+    },
+    {
+      "epoch": 2.56592,
+      "grad_norm": 1.8383262930114825,
+      "learning_rate": 1.0961776731322981e-05,
+      "loss": 1.0835,
+      "step": 16037
+    },
+    {
+      "epoch": 2.56608,
+      "grad_norm": 1.9917266817185642,
+      "learning_rate": 1.0953914773323016e-05,
+      "loss": 1.102,
+      "step": 16038
+    },
+    {
+      "epoch": 2.56624,
+      "grad_norm": 1.6853872211461591,
+      "learning_rate": 1.0946055472328431e-05,
+      "loss": 1.0676,
+      "step": 16039
+    },
+    {
+      "epoch": 2.5664,
+      "grad_norm": 1.8939885591308276,
+      "learning_rate": 1.093819882857381e-05,
+      "loss": 1.0922,
+      "step": 16040
+    },
+    {
+      "epoch": 2.56656,
+      "grad_norm": 1.7398784240771283,
+      "learning_rate": 1.093034484229356e-05,
+      "loss": 1.0917,
+      "step": 16041
+    },
+    {
+      "epoch": 2.56672,
+      "grad_norm": 1.5900633650692575,
+      "learning_rate": 1.0922493513722032e-05,
+      "loss": 1.1004,
+      "step": 16042
+    },
+    {
+      "epoch": 2.5668800000000003,
+      "grad_norm": 1.7450936791588572,
+      "learning_rate": 1.0914644843093491e-05,
+      "loss": 1.0631,
+      "step": 16043
+    },
+    {
+      "epoch": 2.56704,
+      "grad_norm": 1.947187976019258,
+      "learning_rate": 1.0906798830642151e-05,
+      "loss": 1.0983,
+      "step": 16044
+    },
+    {
+      "epoch": 2.5672,
+      "grad_norm": 1.6467418265122864,
+      "learning_rate": 1.0898955476602102e-05,
+      "loss": 1.1068,
+      "step": 16045
+    },
+    {
+      "epoch": 2.56736,
+      "grad_norm": 1.9672927506274926,
+      "learning_rate": 1.0891114781207379e-05,
+      "loss": 1.1011,
+      "step": 16046
+    },
+    {
+      "epoch": 2.56752,
+      "grad_norm": 1.60389615951522,
+      "learning_rate": 1.088327674469196e-05,
+      "loss": 1.0657,
+      "step": 16047
+    },
+    {
+      "epoch": 2.56768,
+      "grad_norm": 1.7193491181174883,
+      "learning_rate": 1.0875441367289696e-05,
+      "loss": 1.0412,
+      "step": 16048
+    },
+    {
+      "epoch": 2.56784,
+      "grad_norm": 2.073020650565538,
+      "learning_rate": 1.0867608649234406e-05,
+      "loss": 1.0385,
+      "step": 16049
+    },
+    {
+      "epoch": 2.568,
+      "grad_norm": 1.6952962204461501,
+      "learning_rate": 1.0859778590759784e-05,
+      "loss": 1.0645,
+      "step": 16050
+    },
+    {
+      "epoch": 2.5681599999999998,
+      "grad_norm": 1.734341323426314,
+      "learning_rate": 1.0851951192099462e-05,
+      "loss": 1.0687,
+      "step": 16051
+    },
+    {
+      "epoch": 2.56832,
+      "grad_norm": 1.5969634513268802,
+      "learning_rate": 1.0844126453487047e-05,
+      "loss": 1.0891,
+      "step": 16052
+    },
+    {
+      "epoch": 2.56848,
+      "grad_norm": 1.680187755669605,
+      "learning_rate": 1.0836304375155958e-05,
+      "loss": 1.0498,
+      "step": 16053
+    },
+    {
+      "epoch": 2.5686400000000003,
+      "grad_norm": 2.053981603536799,
+      "learning_rate": 1.0828484957339647e-05,
+      "loss": 1.119,
+      "step": 16054
+    },
+    {
+      "epoch": 2.5688,
+      "grad_norm": 1.6501381846388297,
+      "learning_rate": 1.0820668200271378e-05,
+      "loss": 1.0332,
+      "step": 16055
+    },
+    {
+      "epoch": 2.56896,
+      "grad_norm": 1.8099411420941058,
+      "learning_rate": 1.0812854104184433e-05,
+      "loss": 1.0655,
+      "step": 16056
+    },
+    {
+      "epoch": 2.56912,
+      "grad_norm": 1.9138611618137258,
+      "learning_rate": 1.0805042669311971e-05,
+      "loss": 1.0608,
+      "step": 16057
+    },
+    {
+      "epoch": 2.56928,
+      "grad_norm": 1.6271345762718894,
+      "learning_rate": 1.079723389588706e-05,
+      "loss": 1.0814,
+      "step": 16058
+    },
+    {
+      "epoch": 2.56944,
+      "grad_norm": 1.952430070905191,
+      "learning_rate": 1.0789427784142714e-05,
+      "loss": 1.0738,
+      "step": 16059
+    },
+    {
+      "epoch": 2.5696,
+      "grad_norm": 1.8412683386508442,
+      "learning_rate": 1.0781624334311847e-05,
+      "loss": 1.0911,
+      "step": 16060
+    },
+    {
+      "epoch": 2.56976,
+      "grad_norm": 1.9292023576991149,
+      "learning_rate": 1.0773823546627293e-05,
+      "loss": 1.0872,
+      "step": 16061
+    },
+    {
+      "epoch": 2.5699199999999998,
+      "grad_norm": 1.8374595039369899,
+      "learning_rate": 1.0766025421321845e-05,
+      "loss": 1.1127,
+      "step": 16062
+    },
+    {
+      "epoch": 2.57008,
+      "grad_norm": 1.665587698518949,
+      "learning_rate": 1.0758229958628163e-05,
+      "loss": 1.0597,
+      "step": 16063
+    },
+    {
+      "epoch": 2.57024,
+      "grad_norm": 1.8549243352382228,
+      "learning_rate": 1.075043715877886e-05,
+      "loss": 1.0805,
+      "step": 16064
+    },
+    {
+      "epoch": 2.5704000000000002,
+      "grad_norm": 1.6936296982643302,
+      "learning_rate": 1.0742647022006458e-05,
+      "loss": 1.0295,
+      "step": 16065
+    },
+    {
+      "epoch": 2.57056,
+      "grad_norm": 1.9137311170432405,
+      "learning_rate": 1.073485954854342e-05,
+      "loss": 1.1046,
+      "step": 16066
+    },
+    {
+      "epoch": 2.57072,
+      "grad_norm": 1.669770108020312,
+      "learning_rate": 1.0727074738622068e-05,
+      "loss": 1.1268,
+      "step": 16067
+    },
+    {
+      "epoch": 2.57088,
+      "grad_norm": 1.5465909767625705,
+      "learning_rate": 1.0719292592474761e-05,
+      "loss": 1.0772,
+      "step": 16068
+    },
+    {
+      "epoch": 2.57104,
+      "grad_norm": 1.9222169296490466,
+      "learning_rate": 1.0711513110333627e-05,
+      "loss": 1.0609,
+      "step": 16069
+    },
+    {
+      "epoch": 2.5712,
+      "grad_norm": 1.615622902717125,
+      "learning_rate": 1.0703736292430889e-05,
+      "loss": 1.0357,
+      "step": 16070
+    },
+    {
+      "epoch": 2.57136,
+      "grad_norm": 1.7874080869068674,
+      "learning_rate": 1.0695962138998483e-05,
+      "loss": 1.0725,
+      "step": 16071
+    },
+    {
+      "epoch": 2.57152,
+      "grad_norm": 1.5814915459803656,
+      "learning_rate": 1.068819065026847e-05,
+      "loss": 1.0547,
+      "step": 16072
+    },
+    {
+      "epoch": 2.5716799999999997,
+      "grad_norm": 1.6932200882897004,
+      "learning_rate": 1.0680421826472697e-05,
+      "loss": 1.0544,
+      "step": 16073
+    },
+    {
+      "epoch": 2.57184,
+      "grad_norm": 1.5962979407415763,
+      "learning_rate": 1.067265566784299e-05,
+      "loss": 1.0629,
+      "step": 16074
+    },
+    {
+      "epoch": 2.572,
+      "grad_norm": 1.966917604875235,
+      "learning_rate": 1.0664892174611075e-05,
+      "loss": 1.1084,
+      "step": 16075
+    },
+    {
+      "epoch": 2.5721600000000002,
+      "grad_norm": 1.8077522096956233,
+      "learning_rate": 1.0657131347008609e-05,
+      "loss": 1.068,
+      "step": 16076
+    },
+    {
+      "epoch": 2.57232,
+      "grad_norm": 1.796899600523046,
+      "learning_rate": 1.0649373185267142e-05,
+      "loss": 1.095,
+      "step": 16077
+    },
+    {
+      "epoch": 2.57248,
+      "grad_norm": 1.6087864867952306,
+      "learning_rate": 1.064161768961819e-05,
+      "loss": 1.0616,
+      "step": 16078
+    },
+    {
+      "epoch": 2.57264,
+      "grad_norm": 1.8051206700299556,
+      "learning_rate": 1.0633864860293153e-05,
+      "loss": 1.0741,
+      "step": 16079
+    },
+    {
+      "epoch": 2.5728,
+      "grad_norm": 1.6607031892462742,
+      "learning_rate": 1.0626114697523382e-05,
+      "loss": 1.1049,
+      "step": 16080
+    },
+    {
+      "epoch": 2.57296,
+      "grad_norm": 1.7233940226733495,
+      "learning_rate": 1.0618367201540102e-05,
+      "loss": 1.0767,
+      "step": 16081
+    },
+    {
+      "epoch": 2.57312,
+      "grad_norm": 1.8673484398046414,
+      "learning_rate": 1.0610622372574508e-05,
+      "loss": 1.0643,
+      "step": 16082
+    },
+    {
+      "epoch": 2.57328,
+      "grad_norm": 1.7201949678970592,
+      "learning_rate": 1.0602880210857668e-05,
+      "loss": 1.0436,
+      "step": 16083
+    },
+    {
+      "epoch": 2.5734399999999997,
+      "grad_norm": 1.6217406544426172,
+      "learning_rate": 1.0595140716620633e-05,
+      "loss": 1.0922,
+      "step": 16084
+    },
+    {
+      "epoch": 2.5736,
+      "grad_norm": 1.7936256983212853,
+      "learning_rate": 1.0587403890094328e-05,
+      "loss": 1.0804,
+      "step": 16085
+    },
+    {
+      "epoch": 2.57376,
+      "grad_norm": 1.7008421574222161,
+      "learning_rate": 1.05796697315096e-05,
+      "loss": 1.0516,
+      "step": 16086
+    },
+    {
+      "epoch": 2.57392,
+      "grad_norm": 1.7191085736453076,
+      "learning_rate": 1.0571938241097224e-05,
+      "loss": 1.0483,
+      "step": 16087
+    },
+    {
+      "epoch": 2.57408,
+      "grad_norm": 1.722868695240532,
+      "learning_rate": 1.0564209419087912e-05,
+      "loss": 1.1238,
+      "step": 16088
+    },
+    {
+      "epoch": 2.57424,
+      "grad_norm": 1.7074159691894908,
+      "learning_rate": 1.0556483265712258e-05,
+      "loss": 1.0849,
+      "step": 16089
+    },
+    {
+      "epoch": 2.5744,
+      "grad_norm": 1.3978907783855894,
+      "learning_rate": 1.0548759781200812e-05,
+      "loss": 1.0589,
+      "step": 16090
+    },
+    {
+      "epoch": 2.57456,
+      "grad_norm": 1.7341677918674223,
+      "learning_rate": 1.0541038965784034e-05,
+      "loss": 1.0828,
+      "step": 16091
+    },
+    {
+      "epoch": 2.57472,
+      "grad_norm": 1.806612213340676,
+      "learning_rate": 1.0533320819692271e-05,
+      "loss": 1.0463,
+      "step": 16092
+    },
+    {
+      "epoch": 2.57488,
+      "grad_norm": 1.912467462149946,
+      "learning_rate": 1.0525605343155898e-05,
+      "loss": 1.0816,
+      "step": 16093
+    },
+    {
+      "epoch": 2.57504,
+      "grad_norm": 1.6359848966344617,
+      "learning_rate": 1.0517892536405038e-05,
+      "loss": 1.1015,
+      "step": 16094
+    },
+    {
+      "epoch": 2.5752,
+      "grad_norm": 1.712793302581517,
+      "learning_rate": 1.051018239966991e-05,
+      "loss": 1.0625,
+      "step": 16095
+    },
+    {
+      "epoch": 2.57536,
+      "grad_norm": 1.7189587056243727,
+      "learning_rate": 1.0502474933180506e-05,
+      "loss": 1.1074,
+      "step": 16096
+    },
+    {
+      "epoch": 2.57552,
+      "grad_norm": 1.9857701718060634,
+      "learning_rate": 1.0494770137166877e-05,
+      "loss": 1.0774,
+      "step": 16097
+    },
+    {
+      "epoch": 2.57568,
+      "grad_norm": 1.7978029980005987,
+      "learning_rate": 1.0487068011858847e-05,
+      "loss": 1.0929,
+      "step": 16098
+    },
+    {
+      "epoch": 2.57584,
+      "grad_norm": 1.8481926227461005,
+      "learning_rate": 1.047936855748629e-05,
+      "loss": 1.0793,
+      "step": 16099
+    },
+    {
+      "epoch": 2.576,
+      "grad_norm": 1.6364191562683585,
+      "learning_rate": 1.0471671774278935e-05,
+      "loss": 1.0619,
+      "step": 16100
+    },
+    {
+      "epoch": 2.57616,
+      "grad_norm": 1.8292698746808513,
+      "learning_rate": 1.0463977662466429e-05,
+      "loss": 1.0703,
+      "step": 16101
+    },
+    {
+      "epoch": 2.57632,
+      "grad_norm": 1.751195798402254,
+      "learning_rate": 1.0456286222278366e-05,
+      "loss": 1.0836,
+      "step": 16102
+    },
+    {
+      "epoch": 2.57648,
+      "grad_norm": 1.7504386821632159,
+      "learning_rate": 1.0448597453944242e-05,
+      "loss": 1.1131,
+      "step": 16103
+    },
+    {
+      "epoch": 2.5766400000000003,
+      "grad_norm": 1.6780481772745135,
+      "learning_rate": 1.0440911357693484e-05,
+      "loss": 1.1347,
+      "step": 16104
+    },
+    {
+      "epoch": 2.5768,
+      "grad_norm": 1.745673491638137,
+      "learning_rate": 1.0433227933755429e-05,
+      "loss": 1.0364,
+      "step": 16105
+    },
+    {
+      "epoch": 2.57696,
+      "grad_norm": 1.7179890806386313,
+      "learning_rate": 1.0425547182359342e-05,
+      "loss": 1.0943,
+      "step": 16106
+    },
+    {
+      "epoch": 2.57712,
+      "grad_norm": 1.7139514635682571,
+      "learning_rate": 1.041786910373439e-05,
+      "loss": 1.1301,
+      "step": 16107
+    },
+    {
+      "epoch": 2.57728,
+      "grad_norm": 1.8031821604351306,
+      "learning_rate": 1.0410193698109683e-05,
+      "loss": 1.1133,
+      "step": 16108
+    },
+    {
+      "epoch": 2.57744,
+      "grad_norm": 1.7292606223490103,
+      "learning_rate": 1.040252096571428e-05,
+      "loss": 1.093,
+      "step": 16109
+    },
+    {
+      "epoch": 2.5776,
+      "grad_norm": 1.7292687136164941,
+      "learning_rate": 1.0394850906777065e-05,
+      "loss": 1.1006,
+      "step": 16110
+    },
+    {
+      "epoch": 2.57776,
+      "grad_norm": 1.6812650578858988,
+      "learning_rate": 1.0387183521526966e-05,
+      "loss": 1.0787,
+      "step": 16111
+    },
+    {
+      "epoch": 2.5779199999999998,
+      "grad_norm": 1.6194075572141278,
+      "learning_rate": 1.037951881019269e-05,
+      "loss": 1.0518,
+      "step": 16112
+    },
+    {
+      "epoch": 2.57808,
+      "grad_norm": 1.7297799241794807,
+      "learning_rate": 1.037185677300302e-05,
+      "loss": 1.0281,
+      "step": 16113
+    },
+    {
+      "epoch": 2.57824,
+      "grad_norm": 1.7471655403939172,
+      "learning_rate": 1.0364197410186504e-05,
+      "loss": 1.0585,
+      "step": 16114
+    },
+    {
+      "epoch": 2.5784000000000002,
+      "grad_norm": 1.7338347271204058,
+      "learning_rate": 1.0356540721971742e-05,
+      "loss": 1.0482,
+      "step": 16115
+    },
+    {
+      "epoch": 2.57856,
+      "grad_norm": 1.88904987288389,
+      "learning_rate": 1.0348886708587179e-05,
+      "loss": 1.091,
+      "step": 16116
+    },
+    {
+      "epoch": 2.57872,
+      "grad_norm": 1.7384864881984172,
+      "learning_rate": 1.0341235370261204e-05,
+      "loss": 1.0462,
+      "step": 16117
+    },
+    {
+      "epoch": 2.57888,
+      "grad_norm": 1.7345042485610693,
+      "learning_rate": 1.033358670722212e-05,
+      "loss": 1.0878,
+      "step": 16118
+    },
+    {
+      "epoch": 2.57904,
+      "grad_norm": 1.8318418765419178,
+      "learning_rate": 1.0325940719698145e-05,
+      "loss": 1.0829,
+      "step": 16119
+    },
+    {
+      "epoch": 2.5792,
+      "grad_norm": 1.8052772875487164,
+      "learning_rate": 1.0318297407917443e-05,
+      "loss": 1.0831,
+      "step": 16120
+    },
+    {
+      "epoch": 2.57936,
+      "grad_norm": 1.9819457346422336,
+      "learning_rate": 1.0310656772108052e-05,
+      "loss": 1.0719,
+      "step": 16121
+    },
+    {
+      "epoch": 2.57952,
+      "grad_norm": 1.9417329695641408,
+      "learning_rate": 1.0303018812497978e-05,
+      "loss": 1.0663,
+      "step": 16122
+    },
+    {
+      "epoch": 2.5796799999999998,
+      "grad_norm": 1.8058674493329538,
+      "learning_rate": 1.0295383529315117e-05,
+      "loss": 1.1165,
+      "step": 16123
+    },
+    {
+      "epoch": 2.57984,
+      "grad_norm": 1.8440117613049836,
+      "learning_rate": 1.0287750922787286e-05,
+      "loss": 1.0902,
+      "step": 16124
+    },
+    {
+      "epoch": 2.58,
+      "grad_norm": 1.6497932496197387,
+      "learning_rate": 1.0280120993142273e-05,
+      "loss": 1.0761,
+      "step": 16125
+    },
+    {
+      "epoch": 2.5801600000000002,
+      "grad_norm": 1.6954061404567178,
+      "learning_rate": 1.027249374060768e-05,
+      "loss": 1.1188,
+      "step": 16126
+    },
+    {
+      "epoch": 2.58032,
+      "grad_norm": 1.765256916253149,
+      "learning_rate": 1.0264869165411173e-05,
+      "loss": 1.0818,
+      "step": 16127
+    },
+    {
+      "epoch": 2.58048,
+      "grad_norm": 1.7504730950803102,
+      "learning_rate": 1.0257247267780157e-05,
+      "loss": 1.096,
+      "step": 16128
+    },
+    {
+      "epoch": 2.58064,
+      "grad_norm": 1.5335169721139976,
+      "learning_rate": 1.024962804794216e-05,
+      "loss": 1.0286,
+      "step": 16129
+    },
+    {
+      "epoch": 2.5808,
+      "grad_norm": 1.5977870837377555,
+      "learning_rate": 1.0242011506124449e-05,
+      "loss": 1.0655,
+      "step": 16130
+    },
+    {
+      "epoch": 2.58096,
+      "grad_norm": 1.8212954982961538,
+      "learning_rate": 1.0234397642554339e-05,
+      "loss": 1.0846,
+      "step": 16131
+    },
+    {
+      "epoch": 2.58112,
+      "grad_norm": 1.744805386293469,
+      "learning_rate": 1.0226786457459003e-05,
+      "loss": 1.0786,
+      "step": 16132
+    },
+    {
+      "epoch": 2.58128,
+      "grad_norm": 1.645802526080001,
+      "learning_rate": 1.021917795106554e-05,
+      "loss": 1.0877,
+      "step": 16133
+    },
+    {
+      "epoch": 2.5814399999999997,
+      "grad_norm": 1.8449695444819587,
+      "learning_rate": 1.0211572123600988e-05,
+      "loss": 1.0733,
+      "step": 16134
+    },
+    {
+      "epoch": 2.5816,
+      "grad_norm": 1.6150947199736996,
+      "learning_rate": 1.0203968975292289e-05,
+      "loss": 1.0238,
+      "step": 16135
+    },
+    {
+      "epoch": 2.58176,
+      "grad_norm": 1.60094439736983,
+      "learning_rate": 1.0196368506366305e-05,
+      "loss": 1.0599,
+      "step": 16136
+    },
+    {
+      "epoch": 2.58192,
+      "grad_norm": 1.886097177593462,
+      "learning_rate": 1.0188770717049822e-05,
+      "loss": 1.0251,
+      "step": 16137
+    },
+    {
+      "epoch": 2.58208,
+      "grad_norm": 1.6788105746378281,
+      "learning_rate": 1.018117560756956e-05,
+      "loss": 1.0931,
+      "step": 16138
+    },
+    {
+      "epoch": 2.58224,
+      "grad_norm": 1.8940979724743503,
+      "learning_rate": 1.0173583178152136e-05,
+      "loss": 1.0715,
+      "step": 16139
+    },
+    {
+      "epoch": 2.5824,
+      "grad_norm": 1.6960557453386869,
+      "learning_rate": 1.0165993429024078e-05,
+      "loss": 1.0702,
+      "step": 16140
+    },
+    {
+      "epoch": 2.58256,
+      "grad_norm": 2.0856644995384617,
+      "learning_rate": 1.015840636041191e-05,
+      "loss": 1.0907,
+      "step": 16141
+    },
+    {
+      "epoch": 2.58272,
+      "grad_norm": 1.7903490686873282,
+      "learning_rate": 1.0150821972541946e-05,
+      "loss": 1.1077,
+      "step": 16142
+    },
+    {
+      "epoch": 2.58288,
+      "grad_norm": 1.8033669073910714,
+      "learning_rate": 1.0143240265640563e-05,
+      "loss": 1.0517,
+      "step": 16143
+    },
+    {
+      "epoch": 2.58304,
+      "grad_norm": 1.8122207426246606,
+      "learning_rate": 1.0135661239933936e-05,
+      "loss": 1.0642,
+      "step": 16144
+    },
+    {
+      "epoch": 2.5832,
+      "grad_norm": 1.6483407834403165,
+      "learning_rate": 1.0128084895648249e-05,
+      "loss": 1.0641,
+      "step": 16145
+    },
+    {
+      "epoch": 2.58336,
+      "grad_norm": 1.8151167738372531,
+      "learning_rate": 1.0120511233009534e-05,
+      "loss": 1.1119,
+      "step": 16146
+    },
+    {
+      "epoch": 2.58352,
+      "grad_norm": 1.7680426715244337,
+      "learning_rate": 1.0112940252243808e-05,
+      "loss": 1.0662,
+      "step": 16147
+    },
+    {
+      "epoch": 2.58368,
+      "grad_norm": 1.8791415794395079,
+      "learning_rate": 1.0105371953576958e-05,
+      "loss": 1.0872,
+      "step": 16148
+    },
+    {
+      "epoch": 2.58384,
+      "grad_norm": 2.001592631745389,
+      "learning_rate": 1.0097806337234806e-05,
+      "loss": 1.1045,
+      "step": 16149
+    },
+    {
+      "epoch": 2.584,
+      "grad_norm": 1.90694374274218,
+      "learning_rate": 1.0090243403443144e-05,
+      "loss": 1.0955,
+      "step": 16150
+    },
+    {
+      "epoch": 2.58416,
+      "grad_norm": 1.6870497761883236,
+      "learning_rate": 1.0082683152427575e-05,
+      "loss": 1.0777,
+      "step": 16151
+    },
+    {
+      "epoch": 2.58432,
+      "grad_norm": 1.7866471676637237,
+      "learning_rate": 1.0075125584413748e-05,
+      "loss": 1.1244,
+      "step": 16152
+    },
+    {
+      "epoch": 2.58448,
+      "grad_norm": 1.6601702865890742,
+      "learning_rate": 1.0067570699627105e-05,
+      "loss": 1.0497,
+      "step": 16153
+    },
+    {
+      "epoch": 2.5846400000000003,
+      "grad_norm": 1.7921363328480708,
+      "learning_rate": 1.0060018498293145e-05,
+      "loss": 1.0647,
+      "step": 16154
+    },
+    {
+      "epoch": 2.5848,
+      "grad_norm": 1.6569153164706618,
+      "learning_rate": 1.0052468980637141e-05,
+      "loss": 1.1168,
+      "step": 16155
+    },
+    {
+      "epoch": 2.58496,
+      "grad_norm": 1.7218686183383864,
+      "learning_rate": 1.0044922146884406e-05,
+      "loss": 1.0672,
+      "step": 16156
+    },
+    {
+      "epoch": 2.58512,
+      "grad_norm": 1.6231141037737278,
+      "learning_rate": 1.0037377997260122e-05,
+      "loss": 1.0808,
+      "step": 16157
+    },
+    {
+      "epoch": 2.58528,
+      "grad_norm": 1.7340791659623658,
+      "learning_rate": 1.002983653198939e-05,
+      "loss": 1.094,
+      "step": 16158
+    },
+    {
+      "epoch": 2.58544,
+      "grad_norm": 1.6633414135985083,
+      "learning_rate": 1.0022297751297239e-05,
+      "loss": 1.0208,
+      "step": 16159
+    },
+    {
+      "epoch": 2.5856,
+      "grad_norm": 1.798745511631558,
+      "learning_rate": 1.00147616554086e-05,
+      "loss": 1.1135,
+      "step": 16160
+    },
+    {
+      "epoch": 2.58576,
+      "grad_norm": 1.816126022919094,
+      "learning_rate": 1.000722824454836e-05,
+      "loss": 1.0078,
+      "step": 16161
+    },
+    {
+      "epoch": 2.5859199999999998,
+      "grad_norm": 1.7376146854115446,
+      "learning_rate": 9.999697518941298e-06,
+      "loss": 1.021,
+      "step": 16162
+    },
+    {
+      "epoch": 2.58608,
+      "grad_norm": 1.9041828703530663,
+      "learning_rate": 9.992169478812108e-06,
+      "loss": 1.0843,
+      "step": 16163
+    },
+    {
+      "epoch": 2.58624,
+      "grad_norm": 1.8049571361886279,
+      "learning_rate": 9.984644124385434e-06,
+      "loss": 1.1172,
+      "step": 16164
+    },
+    {
+      "epoch": 2.5864000000000003,
+      "grad_norm": 1.909231231772935,
+      "learning_rate": 9.977121455885786e-06,
+      "loss": 1.0786,
+      "step": 16165
+    },
+    {
+      "epoch": 2.58656,
+      "grad_norm": 1.815797883457413,
+      "learning_rate": 9.969601473537704e-06,
+      "loss": 1.0725,
+      "step": 16166
+    },
+    {
+      "epoch": 2.58672,
+      "grad_norm": 1.7439767509338369,
+      "learning_rate": 9.962084177565479e-06,
+      "loss": 1.0966,
+      "step": 16167
+    },
+    {
+      "epoch": 2.58688,
+      "grad_norm": 1.833335768099524,
+      "learning_rate": 9.954569568193506e-06,
+      "loss": 1.1183,
+      "step": 16168
+    },
+    {
+      "epoch": 2.58704,
+      "grad_norm": 2.0364909887884197,
+      "learning_rate": 9.94705764564593e-06,
+      "loss": 1.0674,
+      "step": 16169
+    },
+    {
+      "epoch": 2.5872,
+      "grad_norm": 1.655796592645574,
+      "learning_rate": 9.93954841014697e-06,
+      "loss": 1.0452,
+      "step": 16170
+    },
+    {
+      "epoch": 2.58736,
+      "grad_norm": 1.740541734704766,
+      "learning_rate": 9.932041861920616e-06,
+      "loss": 1.0621,
+      "step": 16171
+    },
+    {
+      "epoch": 2.58752,
+      "grad_norm": 1.5237239569714287,
+      "learning_rate": 9.924538001190908e-06,
+      "loss": 1.0443,
+      "step": 16172
+    },
+    {
+      "epoch": 2.5876799999999998,
+      "grad_norm": 1.5405395919551694,
+      "learning_rate": 9.917036828181724e-06,
+      "loss": 1.0946,
+      "step": 16173
+    },
+    {
+      "epoch": 2.58784,
+      "grad_norm": 1.7250818147034512,
+      "learning_rate": 9.909538343116909e-06,
+      "loss": 1.1033,
+      "step": 16174
+    },
+    {
+      "epoch": 2.588,
+      "grad_norm": 1.686295779148556,
+      "learning_rate": 9.902042546220191e-06,
+      "loss": 1.0089,
+      "step": 16175
+    },
+    {
+      "epoch": 2.5881600000000002,
+      "grad_norm": 1.7202308493329639,
+      "learning_rate": 9.894549437715228e-06,
+      "loss": 1.093,
+      "step": 16176
+    },
+    {
+      "epoch": 2.58832,
+      "grad_norm": 1.709311090289206,
+      "learning_rate": 9.887059017825629e-06,
+      "loss": 1.0707,
+      "step": 16177
+    },
+    {
+      "epoch": 2.58848,
+      "grad_norm": 1.7136177083999007,
+      "learning_rate": 9.87957128677487e-06,
+      "loss": 1.0833,
+      "step": 16178
+    },
+    {
+      "epoch": 2.58864,
+      "grad_norm": 1.8144406768817223,
+      "learning_rate": 9.872086244786382e-06,
+      "loss": 1.0784,
+      "step": 16179
+    },
+    {
+      "epoch": 2.5888,
+      "grad_norm": 1.5280333270195297,
+      "learning_rate": 9.864603892083513e-06,
+      "loss": 1.0729,
+      "step": 16180
+    },
+    {
+      "epoch": 2.58896,
+      "grad_norm": 1.7560978623809254,
+      "learning_rate": 9.857124228889503e-06,
+      "loss": 1.1152,
+      "step": 16181
+    },
+    {
+      "epoch": 2.58912,
+      "grad_norm": 1.9106595210129698,
+      "learning_rate": 9.849647255427596e-06,
+      "loss": 1.0851,
+      "step": 16182
+    },
+    {
+      "epoch": 2.58928,
+      "grad_norm": 1.7690383163899528,
+      "learning_rate": 9.842172971920805e-06,
+      "loss": 1.0446,
+      "step": 16183
+    },
+    {
+      "epoch": 2.5894399999999997,
+      "grad_norm": 1.8058850728453182,
+      "learning_rate": 9.834701378592249e-06,
+      "loss": 1.0592,
+      "step": 16184
+    },
+    {
+      "epoch": 2.5896,
+      "grad_norm": 1.7452671140545992,
+      "learning_rate": 9.82723247566477e-06,
+      "loss": 1.0413,
+      "step": 16185
+    },
+    {
+      "epoch": 2.58976,
+      "grad_norm": 1.6441181674860834,
+      "learning_rate": 9.81976626336132e-06,
+      "loss": 1.0387,
+      "step": 16186
+    },
+    {
+      "epoch": 2.58992,
+      "grad_norm": 1.7849438177371315,
+      "learning_rate": 9.8123027419046e-06,
+      "loss": 1.0396,
+      "step": 16187
+    },
+    {
+      "epoch": 2.59008,
+      "grad_norm": 1.96146067594249,
+      "learning_rate": 9.804841911517371e-06,
+      "loss": 1.0906,
+      "step": 16188
+    },
+    {
+      "epoch": 2.59024,
+      "grad_norm": 1.746075950405558,
+      "learning_rate": 9.79738377242223e-06,
+      "loss": 1.0804,
+      "step": 16189
+    },
+    {
+      "epoch": 2.5904,
+      "grad_norm": 1.7620198531290232,
+      "learning_rate": 9.789928324841713e-06,
+      "loss": 1.0858,
+      "step": 16190
+    },
+    {
+      "epoch": 2.59056,
+      "grad_norm": 1.611866923658511,
+      "learning_rate": 9.782475568998295e-06,
+      "loss": 1.0802,
+      "step": 16191
+    },
+    {
+      "epoch": 2.59072,
+      "grad_norm": 1.680166874948286,
+      "learning_rate": 9.775025505114343e-06,
+      "loss": 1.0951,
+      "step": 16192
+    },
+    {
+      "epoch": 2.59088,
+      "grad_norm": 1.7641948243826648,
+      "learning_rate": 9.767578133412147e-06,
+      "loss": 1.0935,
+      "step": 16193
+    },
+    {
+      "epoch": 2.59104,
+      "grad_norm": 1.6282728132640079,
+      "learning_rate": 9.760133454113952e-06,
+      "loss": 1.0588,
+      "step": 16194
+    },
+    {
+      "epoch": 2.5911999999999997,
+      "grad_norm": 1.9163154328929868,
+      "learning_rate": 9.752691467441865e-06,
+      "loss": 1.0872,
+      "step": 16195
+    },
+    {
+      "epoch": 2.59136,
+      "grad_norm": 1.7664827789948665,
+      "learning_rate": 9.74525217361797e-06,
+      "loss": 1.0645,
+      "step": 16196
+    },
+    {
+      "epoch": 2.59152,
+      "grad_norm": 1.7127049559242347,
+      "learning_rate": 9.737815572864218e-06,
+      "loss": 1.1127,
+      "step": 16197
+    },
+    {
+      "epoch": 2.59168,
+      "grad_norm": 1.6257029966018899,
+      "learning_rate": 9.730381665402555e-06,
+      "loss": 1.0661,
+      "step": 16198
+    },
+    {
+      "epoch": 2.59184,
+      "grad_norm": 2.0232251508128676,
+      "learning_rate": 9.722950451454737e-06,
+      "loss": 1.078,
+      "step": 16199
+    },
+    {
+      "epoch": 2.592,
+      "grad_norm": 1.7833895196735507,
+      "learning_rate": 9.715521931242543e-06,
+      "loss": 1.0836,
+      "step": 16200
+    },
+    {
+      "epoch": 2.59216,
+      "grad_norm": 1.786753821782625,
+      "learning_rate": 9.708096104987618e-06,
+      "loss": 1.0203,
+      "step": 16201
+    },
+    {
+      "epoch": 2.59232,
+      "grad_norm": 1.7693683933565454,
+      "learning_rate": 9.70067297291154e-06,
+      "loss": 1.0758,
+      "step": 16202
+    },
+    {
+      "epoch": 2.59248,
+      "grad_norm": 1.6179275054352409,
+      "learning_rate": 9.693252535235797e-06,
+      "loss": 1.0425,
+      "step": 16203
+    },
+    {
+      "epoch": 2.59264,
+      "grad_norm": 1.6905675506196722,
+      "learning_rate": 9.685834792181803e-06,
+      "loss": 1.0626,
+      "step": 16204
+    },
+    {
+      "epoch": 2.5928,
+      "grad_norm": 1.542550097804661,
+      "learning_rate": 9.678419743970902e-06,
+      "loss": 1.0816,
+      "step": 16205
+    },
+    {
+      "epoch": 2.59296,
+      "grad_norm": 1.6050232637205388,
+      "learning_rate": 9.671007390824316e-06,
+      "loss": 1.1144,
+      "step": 16206
+    },
+    {
+      "epoch": 2.59312,
+      "grad_norm": 1.6735977353013491,
+      "learning_rate": 9.663597732963292e-06,
+      "loss": 1.0621,
+      "step": 16207
+    },
+    {
+      "epoch": 2.59328,
+      "grad_norm": 1.5664014933149844,
+      "learning_rate": 9.656190770608842e-06,
+      "loss": 1.0186,
+      "step": 16208
+    },
+    {
+      "epoch": 2.59344,
+      "grad_norm": 1.8191837361601189,
+      "learning_rate": 9.648786503982044e-06,
+      "loss": 1.0448,
+      "step": 16209
+    },
+    {
+      "epoch": 2.5936,
+      "grad_norm": 1.6338599708311115,
+      "learning_rate": 9.641384933303777e-06,
+      "loss": 1.0859,
+      "step": 16210
+    },
+    {
+      "epoch": 2.59376,
+      "grad_norm": 1.8243672361580006,
+      "learning_rate": 9.633986058794942e-06,
+      "loss": 1.1056,
+      "step": 16211
+    },
+    {
+      "epoch": 2.59392,
+      "grad_norm": 1.6897946695875188,
+      "learning_rate": 9.62658988067625e-06,
+      "loss": 1.0621,
+      "step": 16212
+    },
+    {
+      "epoch": 2.59408,
+      "grad_norm": 1.6409388431627359,
+      "learning_rate": 9.61919639916845e-06,
+      "loss": 1.0793,
+      "step": 16213
+    },
+    {
+      "epoch": 2.59424,
+      "grad_norm": 1.888535654011773,
+      "learning_rate": 9.61180561449213e-06,
+      "loss": 1.109,
+      "step": 16214
+    },
+    {
+      "epoch": 2.5944000000000003,
+      "grad_norm": 1.782114013047677,
+      "learning_rate": 9.604417526867826e-06,
+      "loss": 1.1445,
+      "step": 16215
+    },
+    {
+      "epoch": 2.59456,
+      "grad_norm": 1.736256828507259,
+      "learning_rate": 9.59703213651597e-06,
+      "loss": 1.1404,
+      "step": 16216
+    },
+    {
+      "epoch": 2.59472,
+      "grad_norm": 1.6129279075064606,
+      "learning_rate": 9.589649443656957e-06,
+      "loss": 1.0634,
+      "step": 16217
+    },
+    {
+      "epoch": 2.59488,
+      "grad_norm": 1.6522573262978422,
+      "learning_rate": 9.58226944851106e-06,
+      "loss": 1.07,
+      "step": 16218
+    },
+    {
+      "epoch": 2.59504,
+      "grad_norm": 1.9628207390150123,
+      "learning_rate": 9.574892151298487e-06,
+      "loss": 1.0676,
+      "step": 16219
+    },
+    {
+      "epoch": 2.5952,
+      "grad_norm": 1.9344633670467637,
+      "learning_rate": 9.567517552239369e-06,
+      "loss": 1.0843,
+      "step": 16220
+    },
+    {
+      "epoch": 2.59536,
+      "grad_norm": 1.8046100767517645,
+      "learning_rate": 9.560145651553754e-06,
+      "loss": 1.0702,
+      "step": 16221
+    },
+    {
+      "epoch": 2.59552,
+      "grad_norm": 1.791899807168691,
+      "learning_rate": 9.552776449461586e-06,
+      "loss": 1.0788,
+      "step": 16222
+    },
+    {
+      "epoch": 2.5956799999999998,
+      "grad_norm": 1.6780863429294295,
+      "learning_rate": 9.545409946182816e-06,
+      "loss": 1.0544,
+      "step": 16223
+    },
+    {
+      "epoch": 2.59584,
+      "grad_norm": 1.794790647277803,
+      "learning_rate": 9.538046141937162e-06,
+      "loss": 1.1166,
+      "step": 16224
+    },
+    {
+      "epoch": 2.596,
+      "grad_norm": 1.5228668244798704,
+      "learning_rate": 9.530685036944442e-06,
+      "loss": 1.0818,
+      "step": 16225
+    },
+    {
+      "epoch": 2.5961600000000002,
+      "grad_norm": 1.7624952001342622,
+      "learning_rate": 9.523326631424212e-06,
+      "loss": 1.0847,
+      "step": 16226
+    },
+    {
+      "epoch": 2.59632,
+      "grad_norm": 1.633818602480162,
+      "learning_rate": 9.515970925596107e-06,
+      "loss": 1.0731,
+      "step": 16227
+    },
+    {
+      "epoch": 2.59648,
+      "grad_norm": 1.4243210061242675,
+      "learning_rate": 9.508617919679552e-06,
+      "loss": 1.0744,
+      "step": 16228
+    },
+    {
+      "epoch": 2.59664,
+      "grad_norm": 1.7107877738486867,
+      "learning_rate": 9.501267613893994e-06,
+      "loss": 1.0785,
+      "step": 16229
+    },
+    {
+      "epoch": 2.5968,
+      "grad_norm": 1.7255833256921462,
+      "learning_rate": 9.493920008458734e-06,
+      "loss": 1.041,
+      "step": 16230
+    },
+    {
+      "epoch": 2.59696,
+      "grad_norm": 1.7062964671901244,
+      "learning_rate": 9.48657510359302e-06,
+      "loss": 1.0506,
+      "step": 16231
+    },
+    {
+      "epoch": 2.59712,
+      "grad_norm": 1.781880310688934,
+      "learning_rate": 9.479232899516022e-06,
+      "loss": 1.078,
+      "step": 16232
+    },
+    {
+      "epoch": 2.59728,
+      "grad_norm": 1.7433461738899514,
+      "learning_rate": 9.471893396446807e-06,
+      "loss": 1.0983,
+      "step": 16233
+    },
+    {
+      "epoch": 2.5974399999999997,
+      "grad_norm": 1.5995102419078688,
+      "learning_rate": 9.464556594604378e-06,
+      "loss": 1.0639,
+      "step": 16234
+    },
+    {
+      "epoch": 2.5976,
+      "grad_norm": 1.7164094844571087,
+      "learning_rate": 9.457222494207663e-06,
+      "loss": 1.0794,
+      "step": 16235
+    },
+    {
+      "epoch": 2.59776,
+      "grad_norm": 1.6474241173968507,
+      "learning_rate": 9.449891095475483e-06,
+      "loss": 1.0865,
+      "step": 16236
+    },
+    {
+      "epoch": 2.5979200000000002,
+      "grad_norm": 1.7610133951398088,
+      "learning_rate": 9.44256239862662e-06,
+      "loss": 1.0091,
+      "step": 16237
+    },
+    {
+      "epoch": 2.59808,
+      "grad_norm": 1.6566829751831307,
+      "learning_rate": 9.43523640387971e-06,
+      "loss": 1.0914,
+      "step": 16238
+    },
+    {
+      "epoch": 2.59824,
+      "grad_norm": 1.5354296371865337,
+      "learning_rate": 9.427913111453424e-06,
+      "loss": 1.0538,
+      "step": 16239
+    },
+    {
+      "epoch": 2.5984,
+      "grad_norm": 1.8014751787978938,
+      "learning_rate": 9.420592521566184e-06,
+      "loss": 1.0999,
+      "step": 16240
+    },
+    {
+      "epoch": 2.59856,
+      "grad_norm": 1.6777494593134816,
+      "learning_rate": 9.41327463443652e-06,
+      "loss": 1.1241,
+      "step": 16241
+    },
+    {
+      "epoch": 2.59872,
+      "grad_norm": 1.6917942518320932,
+      "learning_rate": 9.405959450282709e-06,
+      "loss": 1.0649,
+      "step": 16242
+    },
+    {
+      "epoch": 2.59888,
+      "grad_norm": 1.7668360482086236,
+      "learning_rate": 9.398646969323088e-06,
+      "loss": 1.1287,
+      "step": 16243
+    },
+    {
+      "epoch": 2.59904,
+      "grad_norm": 1.9720725320037622,
+      "learning_rate": 9.391337191775795e-06,
+      "loss": 1.121,
+      "step": 16244
+    },
+    {
+      "epoch": 2.5991999999999997,
+      "grad_norm": 1.7591841079852097,
+      "learning_rate": 9.384030117858977e-06,
+      "loss": 1.0691,
+      "step": 16245
+    },
+    {
+      "epoch": 2.59936,
+      "grad_norm": 1.9536859186843292,
+      "learning_rate": 9.37672574779066e-06,
+      "loss": 1.0672,
+      "step": 16246
+    },
+    {
+      "epoch": 2.59952,
+      "grad_norm": 1.8084010545733231,
+      "learning_rate": 9.369424081788803e-06,
+      "loss": 1.0348,
+      "step": 16247
+    },
+    {
+      "epoch": 2.59968,
+      "grad_norm": 1.8527669077215192,
+      "learning_rate": 9.362125120071264e-06,
+      "loss": 1.0609,
+      "step": 16248
+    },
+    {
+      "epoch": 2.59984,
+      "grad_norm": 1.6313305436038428,
+      "learning_rate": 9.354828862855835e-06,
+      "loss": 1.0518,
+      "step": 16249
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.7025601850164225,
+      "learning_rate": 9.347535310360233e-06,
+      "loss": 1.07,
+      "step": 16250
+    },
+    {
+      "epoch": 2.60016,
+      "grad_norm": 1.598225071767803,
+      "learning_rate": 9.340244462802073e-06,
+      "loss": 1.1312,
+      "step": 16251
+    },
+    {
+      "epoch": 2.60032,
+      "grad_norm": 1.8226148448978239,
+      "learning_rate": 9.332956320398923e-06,
+      "loss": 1.1125,
+      "step": 16252
+    },
+    {
+      "epoch": 2.60048,
+      "grad_norm": 1.6523391502288307,
+      "learning_rate": 9.325670883368242e-06,
+      "loss": 1.0702,
+      "step": 16253
+    },
+    {
+      "epoch": 2.60064,
+      "grad_norm": 1.5633267423302852,
+      "learning_rate": 9.318388151927393e-06,
+      "loss": 0.9938,
+      "step": 16254
+    },
+    {
+      "epoch": 2.6008,
+      "grad_norm": 1.834531679501991,
+      "learning_rate": 9.311108126293732e-06,
+      "loss": 1.1068,
+      "step": 16255
+    },
+    {
+      "epoch": 2.60096,
+      "grad_norm": 1.6528482438891867,
+      "learning_rate": 9.303830806684432e-06,
+      "loss": 1.0332,
+      "step": 16256
+    },
+    {
+      "epoch": 2.60112,
+      "grad_norm": 1.8690062008550912,
+      "learning_rate": 9.296556193316696e-06,
+      "loss": 1.0725,
+      "step": 16257
+    },
+    {
+      "epoch": 2.60128,
+      "grad_norm": 1.656364169649453,
+      "learning_rate": 9.289284286407519e-06,
+      "loss": 1.1082,
+      "step": 16258
+    },
+    {
+      "epoch": 2.60144,
+      "grad_norm": 2.0304611450935752,
+      "learning_rate": 9.282015086173935e-06,
+      "loss": 1.055,
+      "step": 16259
+    },
+    {
+      "epoch": 2.6016,
+      "grad_norm": 1.6731333353288809,
+      "learning_rate": 9.274748592832838e-06,
+      "loss": 1.053,
+      "step": 16260
+    },
+    {
+      "epoch": 2.60176,
+      "grad_norm": 2.0602513777189504,
+      "learning_rate": 9.267484806601034e-06,
+      "loss": 1.1152,
+      "step": 16261
+    },
+    {
+      "epoch": 2.60192,
+      "grad_norm": 1.6504827378105267,
+      "learning_rate": 9.260223727695273e-06,
+      "loss": 1.0843,
+      "step": 16262
+    },
+    {
+      "epoch": 2.60208,
+      "grad_norm": 1.9156384728472386,
+      "learning_rate": 9.252965356332199e-06,
+      "loss": 1.0524,
+      "step": 16263
+    },
+    {
+      "epoch": 2.60224,
+      "grad_norm": 1.803598371099873,
+      "learning_rate": 9.245709692728444e-06,
+      "loss": 1.1088,
+      "step": 16264
+    },
+    {
+      "epoch": 2.6024000000000003,
+      "grad_norm": 1.9996310743840124,
+      "learning_rate": 9.238456737100432e-06,
+      "loss": 1.1117,
+      "step": 16265
+    },
+    {
+      "epoch": 2.60256,
+      "grad_norm": 1.7360347794022162,
+      "learning_rate": 9.231206489664657e-06,
+      "loss": 1.0806,
+      "step": 16266
+    },
+    {
+      "epoch": 2.60272,
+      "grad_norm": 1.6700402106357728,
+      "learning_rate": 9.223958950637367e-06,
+      "loss": 1.0608,
+      "step": 16267
+    },
+    {
+      "epoch": 2.60288,
+      "grad_norm": 1.8065912842657998,
+      "learning_rate": 9.216714120234915e-06,
+      "loss": 1.0919,
+      "step": 16268
+    },
+    {
+      "epoch": 2.60304,
+      "grad_norm": 1.8284562840090621,
+      "learning_rate": 9.209471998673391e-06,
+      "loss": 1.0566,
+      "step": 16269
+    },
+    {
+      "epoch": 2.6032,
+      "grad_norm": 1.7551075036667865,
+      "learning_rate": 9.202232586168935e-06,
+      "loss": 1.1174,
+      "step": 16270
+    },
+    {
+      "epoch": 2.60336,
+      "grad_norm": 1.6626602080342685,
+      "learning_rate": 9.194995882937563e-06,
+      "loss": 1.0817,
+      "step": 16271
+    },
+    {
+      "epoch": 2.60352,
+      "grad_norm": 1.8130496251384187,
+      "learning_rate": 9.18776188919519e-06,
+      "loss": 1.041,
+      "step": 16272
+    },
+    {
+      "epoch": 2.6036799999999998,
+      "grad_norm": 1.6766939746017377,
+      "learning_rate": 9.180530605157678e-06,
+      "loss": 1.0475,
+      "step": 16273
+    },
+    {
+      "epoch": 2.60384,
+      "grad_norm": 1.7796865989212232,
+      "learning_rate": 9.173302031040798e-06,
+      "loss": 1.0743,
+      "step": 16274
+    },
+    {
+      "epoch": 2.604,
+      "grad_norm": 1.7009624236379255,
+      "learning_rate": 9.166076167060233e-06,
+      "loss": 1.0638,
+      "step": 16275
+    },
+    {
+      "epoch": 2.6041600000000003,
+      "grad_norm": 1.8980891895880703,
+      "learning_rate": 9.1588530134316e-06,
+      "loss": 1.0975,
+      "step": 16276
+    },
+    {
+      "epoch": 2.60432,
+      "grad_norm": 1.8418454483288338,
+      "learning_rate": 9.151632570370416e-06,
+      "loss": 1.1234,
+      "step": 16277
+    },
+    {
+      "epoch": 2.60448,
+      "grad_norm": 1.7306248020259942,
+      "learning_rate": 9.144414838092142e-06,
+      "loss": 1.0025,
+      "step": 16278
+    },
+    {
+      "epoch": 2.60464,
+      "grad_norm": 1.7603082475156016,
+      "learning_rate": 9.137199816812125e-06,
+      "loss": 1.0198,
+      "step": 16279
+    },
+    {
+      "epoch": 2.6048,
+      "grad_norm": 1.6022614584853334,
+      "learning_rate": 9.129987506745696e-06,
+      "loss": 1.052,
+      "step": 16280
+    },
+    {
+      "epoch": 2.60496,
+      "grad_norm": 1.863603077128101,
+      "learning_rate": 9.122777908107994e-06,
+      "loss": 1.0553,
+      "step": 16281
+    },
+    {
+      "epoch": 2.60512,
+      "grad_norm": 1.7866029943756225,
+      "learning_rate": 9.11557102111421e-06,
+      "loss": 1.0688,
+      "step": 16282
+    },
+    {
+      "epoch": 2.60528,
+      "grad_norm": 1.8032130967323101,
+      "learning_rate": 9.10836684597931e-06,
+      "loss": 1.0627,
+      "step": 16283
+    },
+    {
+      "epoch": 2.6054399999999998,
+      "grad_norm": 2.0557540722217946,
+      "learning_rate": 9.101165382918353e-06,
+      "loss": 1.0703,
+      "step": 16284
+    },
+    {
+      "epoch": 2.6056,
+      "grad_norm": 1.7326134744720108,
+      "learning_rate": 9.093966632146123e-06,
+      "loss": 1.0574,
+      "step": 16285
+    },
+    {
+      "epoch": 2.60576,
+      "grad_norm": 1.72981852390065,
+      "learning_rate": 9.08677059387748e-06,
+      "loss": 1.0496,
+      "step": 16286
+    },
+    {
+      "epoch": 2.6059200000000002,
+      "grad_norm": 1.773812536750979,
+      "learning_rate": 9.079577268327133e-06,
+      "loss": 1.1067,
+      "step": 16287
+    },
+    {
+      "epoch": 2.60608,
+      "grad_norm": 2.0378432536499873,
+      "learning_rate": 9.072386655709709e-06,
+      "loss": 1.106,
+      "step": 16288
+    },
+    {
+      "epoch": 2.60624,
+      "grad_norm": 1.732578591403689,
+      "learning_rate": 9.065198756239779e-06,
+      "loss": 1.0408,
+      "step": 16289
+    },
+    {
+      "epoch": 2.6064,
+      "grad_norm": 1.7437512941875675,
+      "learning_rate": 9.058013570131806e-06,
+      "loss": 1.0852,
+      "step": 16290
+    },
+    {
+      "epoch": 2.60656,
+      "grad_norm": 1.896505139243941,
+      "learning_rate": 9.050831097600187e-06,
+      "loss": 1.0204,
+      "step": 16291
+    },
+    {
+      "epoch": 2.60672,
+      "grad_norm": 1.8747586900152184,
+      "learning_rate": 9.04365133885925e-06,
+      "loss": 1.0614,
+      "step": 16292
+    },
+    {
+      "epoch": 2.60688,
+      "grad_norm": 1.738376037351955,
+      "learning_rate": 9.03647429412322e-06,
+      "loss": 1.051,
+      "step": 16293
+    },
+    {
+      "epoch": 2.60704,
+      "grad_norm": 1.715239961396869,
+      "learning_rate": 9.029299963606252e-06,
+      "loss": 1.1209,
+      "step": 16294
+    },
+    {
+      "epoch": 2.6071999999999997,
+      "grad_norm": 2.072180972323282,
+      "learning_rate": 9.022128347522385e-06,
+      "loss": 1.1221,
+      "step": 16295
+    },
+    {
+      "epoch": 2.60736,
+      "grad_norm": 1.7554244665270637,
+      "learning_rate": 9.01495944608568e-06,
+      "loss": 1.0348,
+      "step": 16296
+    },
+    {
+      "epoch": 2.60752,
+      "grad_norm": 1.7398267849872557,
+      "learning_rate": 9.007793259509978e-06,
+      "loss": 1.0725,
+      "step": 16297
+    },
+    {
+      "epoch": 2.60768,
+      "grad_norm": 1.6889011865818617,
+      "learning_rate": 9.000629788009163e-06,
+      "loss": 1.0872,
+      "step": 16298
+    },
+    {
+      "epoch": 2.60784,
+      "grad_norm": 1.6771146133131647,
+      "learning_rate": 8.99346903179692e-06,
+      "loss": 1.0543,
+      "step": 16299
+    },
+    {
+      "epoch": 2.608,
+      "grad_norm": 1.7359464751879434,
+      "learning_rate": 8.986310991086988e-06,
+      "loss": 1.0663,
+      "step": 16300
+    },
+    {
+      "epoch": 2.60816,
+      "grad_norm": 1.8256707991034757,
+      "learning_rate": 8.979155666092887e-06,
+      "loss": 1.0553,
+      "step": 16301
+    },
+    {
+      "epoch": 2.60832,
+      "grad_norm": 1.7958189596080878,
+      "learning_rate": 8.972003057028155e-06,
+      "loss": 1.0822,
+      "step": 16302
+    },
+    {
+      "epoch": 2.60848,
+      "grad_norm": 2.0042598278495074,
+      "learning_rate": 8.964853164106212e-06,
+      "loss": 1.1116,
+      "step": 16303
+    },
+    {
+      "epoch": 2.60864,
+      "grad_norm": 1.6370529500369486,
+      "learning_rate": 8.9577059875404e-06,
+      "loss": 1.0958,
+      "step": 16304
+    },
+    {
+      "epoch": 2.6088,
+      "grad_norm": 2.025049430461371,
+      "learning_rate": 8.950561527543988e-06,
+      "loss": 1.0981,
+      "step": 16305
+    },
+    {
+      "epoch": 2.6089599999999997,
+      "grad_norm": 1.960796875601097,
+      "learning_rate": 8.943419784330132e-06,
+      "loss": 1.0491,
+      "step": 16306
+    },
+    {
+      "epoch": 2.60912,
+      "grad_norm": 1.6579368920747357,
+      "learning_rate": 8.936280758111958e-06,
+      "loss": 1.0119,
+      "step": 16307
+    },
+    {
+      "epoch": 2.60928,
+      "grad_norm": 1.9327980599926693,
+      "learning_rate": 8.929144449102467e-06,
+      "loss": 1.0617,
+      "step": 16308
+    },
+    {
+      "epoch": 2.60944,
+      "grad_norm": 1.6989753668394618,
+      "learning_rate": 8.922010857514595e-06,
+      "loss": 1.0444,
+      "step": 16309
+    },
+    {
+      "epoch": 2.6096,
+      "grad_norm": 2.1617168495135743,
+      "learning_rate": 8.91487998356122e-06,
+      "loss": 1.0893,
+      "step": 16310
+    },
+    {
+      "epoch": 2.60976,
+      "grad_norm": 1.6575905400110211,
+      "learning_rate": 8.907751827455069e-06,
+      "loss": 1.1176,
+      "step": 16311
+    },
+    {
+      "epoch": 2.60992,
+      "grad_norm": 1.9935838290882328,
+      "learning_rate": 8.900626389408907e-06,
+      "loss": 1.0836,
+      "step": 16312
+    },
+    {
+      "epoch": 2.61008,
+      "grad_norm": 1.891080925108111,
+      "learning_rate": 8.893503669635273e-06,
+      "loss": 1.1303,
+      "step": 16313
+    },
+    {
+      "epoch": 2.61024,
+      "grad_norm": 1.824635385148132,
+      "learning_rate": 8.886383668346777e-06,
+      "loss": 1.0846,
+      "step": 16314
+    },
+    {
+      "epoch": 2.6104000000000003,
+      "grad_norm": 1.8557574162033486,
+      "learning_rate": 8.87926638575578e-06,
+      "loss": 1.1183,
+      "step": 16315
+    },
+    {
+      "epoch": 2.61056,
+      "grad_norm": 1.913528326729693,
+      "learning_rate": 8.872151822074737e-06,
+      "loss": 1.0321,
+      "step": 16316
+    },
+    {
+      "epoch": 2.61072,
+      "grad_norm": 1.6222068845994664,
+      "learning_rate": 8.865039977515854e-06,
+      "loss": 1.0827,
+      "step": 16317
+    },
+    {
+      "epoch": 2.61088,
+      "grad_norm": 1.612940092901964,
+      "learning_rate": 8.857930852291406e-06,
+      "loss": 1.0523,
+      "step": 16318
+    },
+    {
+      "epoch": 2.61104,
+      "grad_norm": 1.7478818006581518,
+      "learning_rate": 8.850824446613492e-06,
+      "loss": 1.0796,
+      "step": 16319
+    },
+    {
+      "epoch": 2.6112,
+      "grad_norm": 2.0751397388919837,
+      "learning_rate": 8.843720760694163e-06,
+      "loss": 1.0302,
+      "step": 16320
+    },
+    {
+      "epoch": 2.61136,
+      "grad_norm": 1.870421301853416,
+      "learning_rate": 8.836619794745371e-06,
+      "loss": 1.0534,
+      "step": 16321
+    },
+    {
+      "epoch": 2.61152,
+      "grad_norm": 1.6451465720259908,
+      "learning_rate": 8.82952154897898e-06,
+      "loss": 1.0469,
+      "step": 16322
+    },
+    {
+      "epoch": 2.61168,
+      "grad_norm": 1.6887043049260737,
+      "learning_rate": 8.822426023606867e-06,
+      "loss": 1.1023,
+      "step": 16323
+    },
+    {
+      "epoch": 2.61184,
+      "grad_norm": 1.6059008352469104,
+      "learning_rate": 8.81533321884066e-06,
+      "loss": 1.0608,
+      "step": 16324
+    },
+    {
+      "epoch": 2.612,
+      "grad_norm": 1.681195086099053,
+      "learning_rate": 8.80824313489207e-06,
+      "loss": 1.065,
+      "step": 16325
+    },
+    {
+      "epoch": 2.6121600000000003,
+      "grad_norm": 1.928698088204763,
+      "learning_rate": 8.801155771972602e-06,
+      "loss": 1.0731,
+      "step": 16326
+    },
+    {
+      "epoch": 2.61232,
+      "grad_norm": 1.8771094625577254,
+      "learning_rate": 8.794071130293757e-06,
+      "loss": 1.0785,
+      "step": 16327
+    },
+    {
+      "epoch": 2.61248,
+      "grad_norm": 1.7070967601168294,
+      "learning_rate": 8.786989210066943e-06,
+      "loss": 1.0655,
+      "step": 16328
+    },
+    {
+      "epoch": 2.61264,
+      "grad_norm": 1.8946743544026086,
+      "learning_rate": 8.779910011503445e-06,
+      "loss": 1.0453,
+      "step": 16329
+    },
+    {
+      "epoch": 2.6128,
+      "grad_norm": 1.7484111425744802,
+      "learning_rate": 8.772833534814518e-06,
+      "loss": 1.1024,
+      "step": 16330
+    },
+    {
+      "epoch": 2.61296,
+      "grad_norm": 1.7106935720828427,
+      "learning_rate": 8.765759780211314e-06,
+      "loss": 1.0664,
+      "step": 16331
+    },
+    {
+      "epoch": 2.61312,
+      "grad_norm": 1.6756238953810587,
+      "learning_rate": 8.758688747904887e-06,
+      "loss": 1.0463,
+      "step": 16332
+    },
+    {
+      "epoch": 2.61328,
+      "grad_norm": 1.9435899399964416,
+      "learning_rate": 8.751620438106234e-06,
+      "loss": 1.0763,
+      "step": 16333
+    },
+    {
+      "epoch": 2.6134399999999998,
+      "grad_norm": 1.7574884653859781,
+      "learning_rate": 8.744554851026266e-06,
+      "loss": 1.0644,
+      "step": 16334
+    },
+    {
+      "epoch": 2.6136,
+      "grad_norm": 1.643295463642763,
+      "learning_rate": 8.737491986875813e-06,
+      "loss": 1.0473,
+      "step": 16335
+    },
+    {
+      "epoch": 2.61376,
+      "grad_norm": 1.7298728499609055,
+      "learning_rate": 8.730431845865606e-06,
+      "loss": 1.0822,
+      "step": 16336
+    },
+    {
+      "epoch": 2.6139200000000002,
+      "grad_norm": 1.8777457527118204,
+      "learning_rate": 8.723374428206321e-06,
+      "loss": 1.1076,
+      "step": 16337
+    },
+    {
+      "epoch": 2.61408,
+      "grad_norm": 1.6614190504807405,
+      "learning_rate": 8.716319734108524e-06,
+      "loss": 1.0472,
+      "step": 16338
+    },
+    {
+      "epoch": 2.61424,
+      "grad_norm": 1.6360237442749528,
+      "learning_rate": 8.70926776378277e-06,
+      "loss": 1.0446,
+      "step": 16339
+    },
+    {
+      "epoch": 2.6144,
+      "grad_norm": 1.8085355755063783,
+      "learning_rate": 8.702218517439409e-06,
+      "loss": 1.0656,
+      "step": 16340
+    },
+    {
+      "epoch": 2.61456,
+      "grad_norm": 1.535918257389636,
+      "learning_rate": 8.695171995288843e-06,
+      "loss": 1.0973,
+      "step": 16341
+    },
+    {
+      "epoch": 2.61472,
+      "grad_norm": 1.7150902277915276,
+      "learning_rate": 8.688128197541257e-06,
+      "loss": 1.0274,
+      "step": 16342
+    },
+    {
+      "epoch": 2.61488,
+      "grad_norm": 1.826994458241076,
+      "learning_rate": 8.681087124406895e-06,
+      "loss": 1.0745,
+      "step": 16343
+    },
+    {
+      "epoch": 2.61504,
+      "grad_norm": 1.6589302644213935,
+      "learning_rate": 8.674048776095822e-06,
+      "loss": 1.1438,
+      "step": 16344
+    },
+    {
+      "epoch": 2.6151999999999997,
+      "grad_norm": 2.1343203022983412,
+      "learning_rate": 8.667013152818072e-06,
+      "loss": 1.0979,
+      "step": 16345
+    },
+    {
+      "epoch": 2.61536,
+      "grad_norm": 1.6890433816840693,
+      "learning_rate": 8.65998025478355e-06,
+      "loss": 1.0581,
+      "step": 16346
+    },
+    {
+      "epoch": 2.61552,
+      "grad_norm": 1.7295759455131223,
+      "learning_rate": 8.652950082202127e-06,
+      "loss": 1.0923,
+      "step": 16347
+    },
+    {
+      "epoch": 2.6156800000000002,
+      "grad_norm": 2.1586646076142584,
+      "learning_rate": 8.645922635283565e-06,
+      "loss": 1.0955,
+      "step": 16348
+    },
+    {
+      "epoch": 2.61584,
+      "grad_norm": 2.1169582458009417,
+      "learning_rate": 8.638897914237554e-06,
+      "loss": 1.0793,
+      "step": 16349
+    },
+    {
+      "epoch": 2.616,
+      "grad_norm": 1.8074956094661672,
+      "learning_rate": 8.631875919273691e-06,
+      "loss": 1.1214,
+      "step": 16350
+    },
+    {
+      "epoch": 2.61616,
+      "grad_norm": 1.8737472825973291,
+      "learning_rate": 8.624856650601532e-06,
+      "loss": 1.0916,
+      "step": 16351
+    },
+    {
+      "epoch": 2.61632,
+      "grad_norm": 1.8522936081094445,
+      "learning_rate": 8.617840108430497e-06,
+      "loss": 1.0438,
+      "step": 16352
+    },
+    {
+      "epoch": 2.61648,
+      "grad_norm": 1.6260180051037456,
+      "learning_rate": 8.610826292969942e-06,
+      "loss": 1.0443,
+      "step": 16353
+    },
+    {
+      "epoch": 2.61664,
+      "grad_norm": 1.7558169025303567,
+      "learning_rate": 8.603815204429167e-06,
+      "loss": 1.0425,
+      "step": 16354
+    },
+    {
+      "epoch": 2.6168,
+      "grad_norm": 1.9114171213381428,
+      "learning_rate": 8.596806843017391e-06,
+      "loss": 1.0961,
+      "step": 16355
+    },
+    {
+      "epoch": 2.6169599999999997,
+      "grad_norm": 1.9966227290586724,
+      "learning_rate": 8.589801208943671e-06,
+      "loss": 1.0809,
+      "step": 16356
+    },
+    {
+      "epoch": 2.61712,
+      "grad_norm": 1.8000006623034597,
+      "learning_rate": 8.582798302417128e-06,
+      "loss": 1.0854,
+      "step": 16357
+    },
+    {
+      "epoch": 2.61728,
+      "grad_norm": 1.847823240743816,
+      "learning_rate": 8.575798123646629e-06,
+      "loss": 1.0481,
+      "step": 16358
+    },
+    {
+      "epoch": 2.61744,
+      "grad_norm": 1.594326328379026,
+      "learning_rate": 8.568800672841127e-06,
+      "loss": 1.0696,
+      "step": 16359
+    },
+    {
+      "epoch": 2.6176,
+      "grad_norm": 1.577928371847913,
+      "learning_rate": 8.561805950209367e-06,
+      "loss": 1.0415,
+      "step": 16360
+    },
+    {
+      "epoch": 2.61776,
+      "grad_norm": 1.7587696150533116,
+      "learning_rate": 8.554813955960095e-06,
+      "loss": 1.0429,
+      "step": 16361
+    },
+    {
+      "epoch": 2.61792,
+      "grad_norm": 1.8221098163093963,
+      "learning_rate": 8.547824690301908e-06,
+      "loss": 1.1201,
+      "step": 16362
+    },
+    {
+      "epoch": 2.61808,
+      "grad_norm": 1.6879720038663202,
+      "learning_rate": 8.540838153443386e-06,
+      "loss": 1.0404,
+      "step": 16363
+    },
+    {
+      "epoch": 2.61824,
+      "grad_norm": 1.9058929742562334,
+      "learning_rate": 8.533854345592973e-06,
+      "loss": 1.0949,
+      "step": 16364
+    },
+    {
+      "epoch": 2.6184,
+      "grad_norm": 1.6441054335851117,
+      "learning_rate": 8.526873266959078e-06,
+      "loss": 1.1148,
+      "step": 16365
+    },
+    {
+      "epoch": 2.61856,
+      "grad_norm": 1.8019277633373476,
+      "learning_rate": 8.519894917749981e-06,
+      "loss": 1.0812,
+      "step": 16366
+    },
+    {
+      "epoch": 2.61872,
+      "grad_norm": 1.9739831197661326,
+      "learning_rate": 8.512919298173927e-06,
+      "loss": 1.069,
+      "step": 16367
+    },
+    {
+      "epoch": 2.61888,
+      "grad_norm": 1.6815579048587375,
+      "learning_rate": 8.505946408439037e-06,
+      "loss": 1.1071,
+      "step": 16368
+    },
+    {
+      "epoch": 2.61904,
+      "grad_norm": 1.9380957543874624,
+      "learning_rate": 8.49897624875341e-06,
+      "loss": 1.0548,
+      "step": 16369
+    },
+    {
+      "epoch": 2.6192,
+      "grad_norm": 1.7617443467338616,
+      "learning_rate": 8.492008819324971e-06,
+      "loss": 1.0947,
+      "step": 16370
+    },
+    {
+      "epoch": 2.61936,
+      "grad_norm": 1.8078778754979978,
+      "learning_rate": 8.485044120361686e-06,
+      "loss": 1.1089,
+      "step": 16371
+    },
+    {
+      "epoch": 2.61952,
+      "grad_norm": 1.536427152562346,
+      "learning_rate": 8.478082152071298e-06,
+      "loss": 1.0646,
+      "step": 16372
+    },
+    {
+      "epoch": 2.61968,
+      "grad_norm": 1.719588910091009,
+      "learning_rate": 8.4711229146616e-06,
+      "loss": 1.0837,
+      "step": 16373
+    },
+    {
+      "epoch": 2.61984,
+      "grad_norm": 1.9854849824645502,
+      "learning_rate": 8.464166408340201e-06,
+      "loss": 1.0981,
+      "step": 16374
+    },
+    {
+      "epoch": 2.62,
+      "grad_norm": 1.916672334862941,
+      "learning_rate": 8.457212633314704e-06,
+      "loss": 1.0596,
+      "step": 16375
+    },
+    {
+      "epoch": 2.6201600000000003,
+      "grad_norm": 1.7640769043149611,
+      "learning_rate": 8.450261589792596e-06,
+      "loss": 1.0555,
+      "step": 16376
+    },
+    {
+      "epoch": 2.62032,
+      "grad_norm": 1.7100856282449453,
+      "learning_rate": 8.44331327798128e-06,
+      "loss": 1.0748,
+      "step": 16377
+    },
+    {
+      "epoch": 2.62048,
+      "grad_norm": 1.8170383299744761,
+      "learning_rate": 8.436367698088078e-06,
+      "loss": 1.1179,
+      "step": 16378
+    },
+    {
+      "epoch": 2.62064,
+      "grad_norm": 1.5751076950830951,
+      "learning_rate": 8.429424850320223e-06,
+      "loss": 1.0575,
+      "step": 16379
+    },
+    {
+      "epoch": 2.6208,
+      "grad_norm": 1.7474790004388534,
+      "learning_rate": 8.422484734884939e-06,
+      "loss": 1.0769,
+      "step": 16380
+    },
+    {
+      "epoch": 2.62096,
+      "grad_norm": 1.6645875082425574,
+      "learning_rate": 8.415547351989228e-06,
+      "loss": 1.0968,
+      "step": 16381
+    },
+    {
+      "epoch": 2.62112,
+      "grad_norm": 1.7307323835630966,
+      "learning_rate": 8.408612701840169e-06,
+      "loss": 1.0381,
+      "step": 16382
+    },
+    {
+      "epoch": 2.62128,
+      "grad_norm": 1.7201203045734086,
+      "learning_rate": 8.401680784644617e-06,
+      "loss": 1.039,
+      "step": 16383
+    },
+    {
+      "epoch": 2.6214399999999998,
+      "grad_norm": 1.6874746851862867,
+      "learning_rate": 8.39475160060944e-06,
+      "loss": 1.0717,
+      "step": 16384
+    },
+    {
+      "epoch": 2.6216,
+      "grad_norm": 1.7210207799281905,
+      "learning_rate": 8.387825149941398e-06,
+      "loss": 1.0787,
+      "step": 16385
+    },
+    {
+      "epoch": 2.62176,
+      "grad_norm": 1.6025732914456303,
+      "learning_rate": 8.380901432847166e-06,
+      "loss": 1.0694,
+      "step": 16386
+    },
+    {
+      "epoch": 2.6219200000000003,
+      "grad_norm": 1.818630241850469,
+      "learning_rate": 8.373980449533324e-06,
+      "loss": 1.1017,
+      "step": 16387
+    },
+    {
+      "epoch": 2.62208,
+      "grad_norm": 2.0012654901952676,
+      "learning_rate": 8.367062200206399e-06,
+      "loss": 1.146,
+      "step": 16388
+    },
+    {
+      "epoch": 2.62224,
+      "grad_norm": 1.7514986045829906,
+      "learning_rate": 8.360146685072811e-06,
+      "loss": 1.0975,
+      "step": 16389
+    },
+    {
+      "epoch": 2.6224,
+      "grad_norm": 1.6200105300866383,
+      "learning_rate": 8.353233904338919e-06,
+      "loss": 1.0969,
+      "step": 16390
+    },
+    {
+      "epoch": 2.62256,
+      "grad_norm": 1.790874301930291,
+      "learning_rate": 8.34632385821098e-06,
+      "loss": 1.0883,
+      "step": 16391
+    },
+    {
+      "epoch": 2.62272,
+      "grad_norm": 1.689628771446555,
+      "learning_rate": 8.339416546895173e-06,
+      "loss": 1.1036,
+      "step": 16392
+    },
+    {
+      "epoch": 2.62288,
+      "grad_norm": 2.050790087852565,
+      "learning_rate": 8.332511970597623e-06,
+      "loss": 1.0837,
+      "step": 16393
+    },
+    {
+      "epoch": 2.62304,
+      "grad_norm": 2.0603077522197064,
+      "learning_rate": 8.325610129524342e-06,
+      "loss": 1.0544,
+      "step": 16394
+    },
+    {
+      "epoch": 2.6231999999999998,
+      "grad_norm": 1.7385794736958549,
+      "learning_rate": 8.318711023881254e-06,
+      "loss": 1.0206,
+      "step": 16395
+    },
+    {
+      "epoch": 2.62336,
+      "grad_norm": 2.0808657471666434,
+      "learning_rate": 8.311814653874272e-06,
+      "loss": 1.0772,
+      "step": 16396
+    },
+    {
+      "epoch": 2.62352,
+      "grad_norm": 2.092072022414496,
+      "learning_rate": 8.3049210197091e-06,
+      "loss": 1.099,
+      "step": 16397
+    },
+    {
+      "epoch": 2.6236800000000002,
+      "grad_norm": 1.6549201446768202,
+      "learning_rate": 8.298030121591493e-06,
+      "loss": 1.0631,
+      "step": 16398
+    },
+    {
+      "epoch": 2.62384,
+      "grad_norm": 1.9727440039813804,
+      "learning_rate": 8.291141959727023e-06,
+      "loss": 1.0685,
+      "step": 16399
+    },
+    {
+      "epoch": 2.624,
+      "grad_norm": 1.6299436040862758,
+      "learning_rate": 8.284256534321256e-06,
+      "loss": 1.0809,
+      "step": 16400
+    },
+    {
+      "epoch": 2.62416,
+      "grad_norm": 1.8220781010233729,
+      "learning_rate": 8.27737384557964e-06,
+      "loss": 1.0744,
+      "step": 16401
+    },
+    {
+      "epoch": 2.62432,
+      "grad_norm": 1.6774514181598152,
+      "learning_rate": 8.270493893707521e-06,
+      "loss": 1.0386,
+      "step": 16402
+    },
+    {
+      "epoch": 2.62448,
+      "grad_norm": 1.6744270336830536,
+      "learning_rate": 8.263616678910203e-06,
+      "loss": 1.0577,
+      "step": 16403
+    },
+    {
+      "epoch": 2.62464,
+      "grad_norm": 1.7113517148391684,
+      "learning_rate": 8.256742201392898e-06,
+      "loss": 1.0715,
+      "step": 16404
+    },
+    {
+      "epoch": 2.6248,
+      "grad_norm": 1.7007805743047606,
+      "learning_rate": 8.24987046136071e-06,
+      "loss": 1.0641,
+      "step": 16405
+    },
+    {
+      "epoch": 2.6249599999999997,
+      "grad_norm": 1.8846507820080323,
+      "learning_rate": 8.243001459018696e-06,
+      "loss": 1.0909,
+      "step": 16406
+    },
+    {
+      "epoch": 2.62512,
+      "grad_norm": 2.012803657503045,
+      "learning_rate": 8.236135194571814e-06,
+      "loss": 1.0941,
+      "step": 16407
+    },
+    {
+      "epoch": 2.62528,
+      "grad_norm": 1.6826961001755079,
+      "learning_rate": 8.229271668224948e-06,
+      "loss": 1.044,
+      "step": 16408
+    },
+    {
+      "epoch": 2.62544,
+      "grad_norm": 1.7104623331338462,
+      "learning_rate": 8.222410880182896e-06,
+      "loss": 1.0861,
+      "step": 16409
+    },
+    {
+      "epoch": 2.6256,
+      "grad_norm": 2.0596473446364802,
+      "learning_rate": 8.215552830650365e-06,
+      "loss": 1.0765,
+      "step": 16410
+    },
+    {
+      "epoch": 2.62576,
+      "grad_norm": 1.8589561878239975,
+      "learning_rate": 8.208697519831965e-06,
+      "loss": 1.0953,
+      "step": 16411
+    },
+    {
+      "epoch": 2.62592,
+      "grad_norm": 1.7964101727572872,
+      "learning_rate": 8.201844947932324e-06,
+      "loss": 1.0889,
+      "step": 16412
+    },
+    {
+      "epoch": 2.62608,
+      "grad_norm": 1.7348459912454344,
+      "learning_rate": 8.194995115155823e-06,
+      "loss": 1.0879,
+      "step": 16413
+    },
+    {
+      "epoch": 2.62624,
+      "grad_norm": 1.7060020510612919,
+      "learning_rate": 8.188148021706942e-06,
+      "loss": 1.0624,
+      "step": 16414
+    },
+    {
+      "epoch": 2.6264,
+      "grad_norm": 1.830734081452871,
+      "learning_rate": 8.181303667789897e-06,
+      "loss": 1.0667,
+      "step": 16415
+    },
+    {
+      "epoch": 2.62656,
+      "grad_norm": 2.1351811213797474,
+      "learning_rate": 8.174462053608966e-06,
+      "loss": 1.0526,
+      "step": 16416
+    },
+    {
+      "epoch": 2.62672,
+      "grad_norm": 1.8402648667270474,
+      "learning_rate": 8.167623179368289e-06,
+      "loss": 1.1188,
+      "step": 16417
+    },
+    {
+      "epoch": 2.62688,
+      "grad_norm": 1.8180355793663507,
+      "learning_rate": 8.160787045271923e-06,
+      "loss": 1.0965,
+      "step": 16418
+    },
+    {
+      "epoch": 2.62704,
+      "grad_norm": 1.8471747570810564,
+      "learning_rate": 8.153953651523848e-06,
+      "loss": 1.1124,
+      "step": 16419
+    },
+    {
+      "epoch": 2.6272,
+      "grad_norm": 1.7381254062307299,
+      "learning_rate": 8.147122998327972e-06,
+      "loss": 1.0373,
+      "step": 16420
+    },
+    {
+      "epoch": 2.62736,
+      "grad_norm": 2.3341899369183516,
+      "learning_rate": 8.140295085888094e-06,
+      "loss": 1.2501,
+      "step": 16421
+    },
+    {
+      "epoch": 2.62752,
+      "grad_norm": 1.9799922718896907,
+      "learning_rate": 8.133469914407955e-06,
+      "loss": 1.1552,
+      "step": 16422
+    },
+    {
+      "epoch": 2.62768,
+      "grad_norm": 2.009222869620577,
+      "learning_rate": 8.12664748409121e-06,
+      "loss": 1.0884,
+      "step": 16423
+    },
+    {
+      "epoch": 2.62784,
+      "grad_norm": 1.9799718862100906,
+      "learning_rate": 8.119827795141432e-06,
+      "loss": 1.1018,
+      "step": 16424
+    },
+    {
+      "epoch": 2.628,
+      "grad_norm": 1.7630892341012616,
+      "learning_rate": 8.113010847762103e-06,
+      "loss": 1.0639,
+      "step": 16425
+    },
+    {
+      "epoch": 2.6281600000000003,
+      "grad_norm": 1.9508701947782443,
+      "learning_rate": 8.10619664215665e-06,
+      "loss": 1.1009,
+      "step": 16426
+    },
+    {
+      "epoch": 2.62832,
+      "grad_norm": 1.6736216582557435,
+      "learning_rate": 8.099385178528351e-06,
+      "loss": 1.0423,
+      "step": 16427
+    },
+    {
+      "epoch": 2.62848,
+      "grad_norm": 1.7185163959264707,
+      "learning_rate": 8.092576457080536e-06,
+      "loss": 1.0476,
+      "step": 16428
+    },
+    {
+      "epoch": 2.62864,
+      "grad_norm": 1.6736305040798656,
+      "learning_rate": 8.085770478016275e-06,
+      "loss": 1.0598,
+      "step": 16429
+    },
+    {
+      "epoch": 2.6288,
+      "grad_norm": 1.89752050026815,
+      "learning_rate": 8.078967241538726e-06,
+      "loss": 1.0491,
+      "step": 16430
+    },
+    {
+      "epoch": 2.62896,
+      "grad_norm": 1.7769556634254093,
+      "learning_rate": 8.072166747850807e-06,
+      "loss": 1.0567,
+      "step": 16431
+    },
+    {
+      "epoch": 2.62912,
+      "grad_norm": 1.8325840597254406,
+      "learning_rate": 8.065368997155508e-06,
+      "loss": 1.0798,
+      "step": 16432
+    },
+    {
+      "epoch": 2.62928,
+      "grad_norm": 1.7135287244464534,
+      "learning_rate": 8.058573989655627e-06,
+      "loss": 1.1502,
+      "step": 16433
+    },
+    {
+      "epoch": 2.6294399999999998,
+      "grad_norm": 1.7222273887731883,
+      "learning_rate": 8.051781725553931e-06,
+      "loss": 1.0639,
+      "step": 16434
+    },
+    {
+      "epoch": 2.6296,
+      "grad_norm": 2.0224771307810303,
+      "learning_rate": 8.044992205053092e-06,
+      "loss": 1.0584,
+      "step": 16435
+    },
+    {
+      "epoch": 2.62976,
+      "grad_norm": 1.8236614185258493,
+      "learning_rate": 8.03820542835566e-06,
+      "loss": 1.0937,
+      "step": 16436
+    },
+    {
+      "epoch": 2.6299200000000003,
+      "grad_norm": 2.045176854213892,
+      "learning_rate": 8.031421395664217e-06,
+      "loss": 1.0405,
+      "step": 16437
+    },
+    {
+      "epoch": 2.63008,
+      "grad_norm": 1.705101479525581,
+      "learning_rate": 8.024640107181125e-06,
+      "loss": 1.0889,
+      "step": 16438
+    },
+    {
+      "epoch": 2.63024,
+      "grad_norm": 1.564015361443489,
+      "learning_rate": 8.017861563108764e-06,
+      "loss": 1.0499,
+      "step": 16439
+    },
+    {
+      "epoch": 2.6304,
+      "grad_norm": 1.562701091762301,
+      "learning_rate": 8.011085763649361e-06,
+      "loss": 1.0804,
+      "step": 16440
+    },
+    {
+      "epoch": 2.63056,
+      "grad_norm": 1.7857524410374839,
+      "learning_rate": 8.004312709005146e-06,
+      "loss": 1.0076,
+      "step": 16441
+    },
+    {
+      "epoch": 2.63072,
+      "grad_norm": 1.642408678566906,
+      "learning_rate": 7.997542399378155e-06,
+      "loss": 1.082,
+      "step": 16442
+    },
+    {
+      "epoch": 2.63088,
+      "grad_norm": 1.7376441335685313,
+      "learning_rate": 7.99077483497045e-06,
+      "loss": 1.0525,
+      "step": 16443
+    },
+    {
+      "epoch": 2.63104,
+      "grad_norm": 1.817772181577762,
+      "learning_rate": 7.984010015983956e-06,
+      "loss": 1.1105,
+      "step": 16444
+    },
+    {
+      "epoch": 2.6311999999999998,
+      "grad_norm": 1.8384334580693176,
+      "learning_rate": 7.977247942620525e-06,
+      "loss": 1.0733,
+      "step": 16445
+    },
+    {
+      "epoch": 2.63136,
+      "grad_norm": 1.714543259505355,
+      "learning_rate": 7.970488615081917e-06,
+      "loss": 1.087,
+      "step": 16446
+    },
+    {
+      "epoch": 2.63152,
+      "grad_norm": 1.7925601222759493,
+      "learning_rate": 7.963732033569837e-06,
+      "loss": 1.0592,
+      "step": 16447
+    },
+    {
+      "epoch": 2.6316800000000002,
+      "grad_norm": 1.7904437550537953,
+      "learning_rate": 7.95697819828588e-06,
+      "loss": 1.0449,
+      "step": 16448
+    },
+    {
+      "epoch": 2.63184,
+      "grad_norm": 1.6853578633056514,
+      "learning_rate": 7.950227109431562e-06,
+      "loss": 1.0675,
+      "step": 16449
+    },
+    {
+      "epoch": 2.632,
+      "grad_norm": 1.6872350703439245,
+      "learning_rate": 7.943478767208345e-06,
+      "loss": 1.0545,
+      "step": 16450
+    },
+    {
+      "epoch": 2.63216,
+      "grad_norm": 1.7409189157243457,
+      "learning_rate": 7.936733171817579e-06,
+      "loss": 1.0398,
+      "step": 16451
+    },
+    {
+      "epoch": 2.63232,
+      "grad_norm": 1.6535995102200807,
+      "learning_rate": 7.929990323460534e-06,
+      "loss": 1.0458,
+      "step": 16452
+    },
+    {
+      "epoch": 2.63248,
+      "grad_norm": 1.7009057376891457,
+      "learning_rate": 7.923250222338441e-06,
+      "loss": 1.0347,
+      "step": 16453
+    },
+    {
+      "epoch": 2.63264,
+      "grad_norm": 1.8709892852386323,
+      "learning_rate": 7.91651286865236e-06,
+      "loss": 1.0604,
+      "step": 16454
+    },
+    {
+      "epoch": 2.6328,
+      "grad_norm": 1.7673929040580776,
+      "learning_rate": 7.909778262603395e-06,
+      "loss": 1.1151,
+      "step": 16455
+    },
+    {
+      "epoch": 2.6329599999999997,
+      "grad_norm": 1.6487926558429107,
+      "learning_rate": 7.903046404392422e-06,
+      "loss": 1.0833,
+      "step": 16456
+    },
+    {
+      "epoch": 2.63312,
+      "grad_norm": 1.9003394042941364,
+      "learning_rate": 7.89631729422038e-06,
+      "loss": 1.1136,
+      "step": 16457
+    },
+    {
+      "epoch": 2.63328,
+      "grad_norm": 1.565205407879413,
+      "learning_rate": 7.889590932287983e-06,
+      "loss": 1.0718,
+      "step": 16458
+    },
+    {
+      "epoch": 2.6334400000000002,
+      "grad_norm": 1.909118028818733,
+      "learning_rate": 7.882867318795983e-06,
+      "loss": 1.0983,
+      "step": 16459
+    },
+    {
+      "epoch": 2.6336,
+      "grad_norm": 1.711941093452018,
+      "learning_rate": 7.876146453944999e-06,
+      "loss": 1.0481,
+      "step": 16460
+    },
+    {
+      "epoch": 2.63376,
+      "grad_norm": 1.8977321783990313,
+      "learning_rate": 7.86942833793557e-06,
+      "loss": 1.0896,
+      "step": 16461
+    },
+    {
+      "epoch": 2.63392,
+      "grad_norm": 1.6125986293649262,
+      "learning_rate": 7.862712970968156e-06,
+      "loss": 1.059,
+      "step": 16462
+    },
+    {
+      "epoch": 2.63408,
+      "grad_norm": 1.540939990402193,
+      "learning_rate": 7.856000353243121e-06,
+      "loss": 1.0497,
+      "step": 16463
+    },
+    {
+      "epoch": 2.63424,
+      "grad_norm": 1.788123362367553,
+      "learning_rate": 7.84929048496077e-06,
+      "loss": 1.0996,
+      "step": 16464
+    },
+    {
+      "epoch": 2.6344,
+      "grad_norm": 1.765263804736153,
+      "learning_rate": 7.84258336632131e-06,
+      "loss": 1.101,
+      "step": 16465
+    },
+    {
+      "epoch": 2.63456,
+      "grad_norm": 1.7939101404017705,
+      "learning_rate": 7.83587899752487e-06,
+      "loss": 1.0821,
+      "step": 16466
+    },
+    {
+      "epoch": 2.6347199999999997,
+      "grad_norm": 1.6436831907634448,
+      "learning_rate": 7.829177378771513e-06,
+      "loss": 1.1361,
+      "step": 16467
+    },
+    {
+      "epoch": 2.63488,
+      "grad_norm": 1.6659695699140216,
+      "learning_rate": 7.822478510261177e-06,
+      "loss": 1.0768,
+      "step": 16468
+    },
+    {
+      "epoch": 2.63504,
+      "grad_norm": 1.574667241024412,
+      "learning_rate": 7.815782392193793e-06,
+      "loss": 1.1238,
+      "step": 16469
+    },
+    {
+      "epoch": 2.6352,
+      "grad_norm": 1.6877472240438318,
+      "learning_rate": 7.809089024769111e-06,
+      "loss": 1.0972,
+      "step": 16470
+    },
+    {
+      "epoch": 2.63536,
+      "grad_norm": 1.7428445427158543,
+      "learning_rate": 7.802398408186906e-06,
+      "loss": 1.0636,
+      "step": 16471
+    },
+    {
+      "epoch": 2.63552,
+      "grad_norm": 1.7914142053496784,
+      "learning_rate": 7.79571054264675e-06,
+      "loss": 1.046,
+      "step": 16472
+    },
+    {
+      "epoch": 2.63568,
+      "grad_norm": 1.9133979200515085,
+      "learning_rate": 7.78902542834825e-06,
+      "loss": 1.0919,
+      "step": 16473
+    },
+    {
+      "epoch": 2.63584,
+      "grad_norm": 1.8098299291439308,
+      "learning_rate": 7.78234306549087e-06,
+      "loss": 1.0678,
+      "step": 16474
+    },
+    {
+      "epoch": 2.636,
+      "grad_norm": 1.7978240463442359,
+      "learning_rate": 7.775663454274006e-06,
+      "loss": 1.0834,
+      "step": 16475
+    },
+    {
+      "epoch": 2.63616,
+      "grad_norm": 1.7227084340391015,
+      "learning_rate": 7.768986594896954e-06,
+      "loss": 1.0378,
+      "step": 16476
+    },
+    {
+      "epoch": 2.63632,
+      "grad_norm": 1.8193119965314197,
+      "learning_rate": 7.762312487558942e-06,
+      "loss": 1.1053,
+      "step": 16477
+    },
+    {
+      "epoch": 2.63648,
+      "grad_norm": 1.6928340322180249,
+      "learning_rate": 7.755641132459124e-06,
+      "loss": 1.0606,
+      "step": 16478
+    },
+    {
+      "epoch": 2.63664,
+      "grad_norm": 1.6368801527667682,
+      "learning_rate": 7.74897252979656e-06,
+      "loss": 1.0652,
+      "step": 16479
+    },
+    {
+      "epoch": 2.6368,
+      "grad_norm": 2.132698718227588,
+      "learning_rate": 7.742306679770229e-06,
+      "loss": 1.0658,
+      "step": 16480
+    },
+    {
+      "epoch": 2.63696,
+      "grad_norm": 2.0716862457195098,
+      "learning_rate": 7.735643582579033e-06,
+      "loss": 1.0918,
+      "step": 16481
+    },
+    {
+      "epoch": 2.63712,
+      "grad_norm": 1.7329460573478068,
+      "learning_rate": 7.728983238421794e-06,
+      "loss": 1.1329,
+      "step": 16482
+    },
+    {
+      "epoch": 2.63728,
+      "grad_norm": 1.5744750163756998,
+      "learning_rate": 7.722325647497242e-06,
+      "loss": 1.0589,
+      "step": 16483
+    },
+    {
+      "epoch": 2.63744,
+      "grad_norm": 1.754141402163469,
+      "learning_rate": 7.715670810004017e-06,
+      "loss": 1.0695,
+      "step": 16484
+    },
+    {
+      "epoch": 2.6376,
+      "grad_norm": 1.8607628860559413,
+      "learning_rate": 7.709018726140727e-06,
+      "loss": 1.0753,
+      "step": 16485
+    },
+    {
+      "epoch": 2.63776,
+      "grad_norm": 1.761252708032898,
+      "learning_rate": 7.702369396105803e-06,
+      "loss": 1.1022,
+      "step": 16486
+    },
+    {
+      "epoch": 2.6379200000000003,
+      "grad_norm": 1.691505246821626,
+      "learning_rate": 7.695722820097718e-06,
+      "loss": 1.0378,
+      "step": 16487
+    },
+    {
+      "epoch": 2.63808,
+      "grad_norm": 1.6799061390636278,
+      "learning_rate": 7.689078998314736e-06,
+      "loss": 1.0337,
+      "step": 16488
+    },
+    {
+      "epoch": 2.63824,
+      "grad_norm": 1.7008652470267946,
+      "learning_rate": 7.682437930955133e-06,
+      "loss": 1.1193,
+      "step": 16489
+    },
+    {
+      "epoch": 2.6384,
+      "grad_norm": 1.930714045163956,
+      "learning_rate": 7.675799618217072e-06,
+      "loss": 1.0637,
+      "step": 16490
+    },
+    {
+      "epoch": 2.63856,
+      "grad_norm": 1.5635227564038723,
+      "learning_rate": 7.669164060298606e-06,
+      "loss": 1.048,
+      "step": 16491
+    },
+    {
+      "epoch": 2.63872,
+      "grad_norm": 1.8495861632431783,
+      "learning_rate": 7.662531257397754e-06,
+      "loss": 1.0776,
+      "step": 16492
+    },
+    {
+      "epoch": 2.63888,
+      "grad_norm": 1.6796010711800808,
+      "learning_rate": 7.655901209712402e-06,
+      "loss": 1.0656,
+      "step": 16493
+    },
+    {
+      "epoch": 2.63904,
+      "grad_norm": 1.7825353187442787,
+      "learning_rate": 7.649273917440436e-06,
+      "loss": 1.108,
+      "step": 16494
+    },
+    {
+      "epoch": 2.6391999999999998,
+      "grad_norm": 1.692647193934923,
+      "learning_rate": 7.64264938077952e-06,
+      "loss": 1.1135,
+      "step": 16495
+    },
+    {
+      "epoch": 2.63936,
+      "grad_norm": 1.699797034778703,
+      "learning_rate": 7.636027599927409e-06,
+      "loss": 1.0758,
+      "step": 16496
+    },
+    {
+      "epoch": 2.63952,
+      "grad_norm": 2.1591151346508255,
+      "learning_rate": 7.62940857508162e-06,
+      "loss": 1.0439,
+      "step": 16497
+    },
+    {
+      "epoch": 2.6396800000000002,
+      "grad_norm": 1.918991762277325,
+      "learning_rate": 7.6227923064397074e-06,
+      "loss": 1.0668,
+      "step": 16498
+    },
+    {
+      "epoch": 2.63984,
+      "grad_norm": 1.5738836405448282,
+      "learning_rate": 7.616178794199025e-06,
+      "loss": 1.0251,
+      "step": 16499
+    },
+    {
+      "epoch": 2.64,
+      "grad_norm": 1.9679699650319677,
+      "learning_rate": 7.60956803855698e-06,
+      "loss": 1.0496,
+      "step": 16500
+    },
+    {
+      "epoch": 2.64016,
+      "grad_norm": 1.660074325102113,
+      "learning_rate": 7.602960039710783e-06,
+      "loss": 1.0871,
+      "step": 16501
+    },
+    {
+      "epoch": 2.64032,
+      "grad_norm": 2.021985840506867,
+      "learning_rate": 7.59635479785763e-06,
+      "loss": 1.0461,
+      "step": 16502
+    },
+    {
+      "epoch": 2.64048,
+      "grad_norm": 1.5544317785809458,
+      "learning_rate": 7.589752313194598e-06,
+      "loss": 1.0374,
+      "step": 16503
+    },
+    {
+      "epoch": 2.64064,
+      "grad_norm": 1.834644649661309,
+      "learning_rate": 7.583152585918707e-06,
+      "loss": 1.1092,
+      "step": 16504
+    },
+    {
+      "epoch": 2.6408,
+      "grad_norm": 1.6121330984388325,
+      "learning_rate": 7.576555616226877e-06,
+      "loss": 1.0873,
+      "step": 16505
+    },
+    {
+      "epoch": 2.6409599999999998,
+      "grad_norm": 1.822685906668008,
+      "learning_rate": 7.569961404315951e-06,
+      "loss": 1.1277,
+      "step": 16506
+    },
+    {
+      "epoch": 2.64112,
+      "grad_norm": 2.1544014264132105,
+      "learning_rate": 7.563369950382693e-06,
+      "loss": 1.0732,
+      "step": 16507
+    },
+    {
+      "epoch": 2.64128,
+      "grad_norm": 1.958514018802683,
+      "learning_rate": 7.556781254623779e-06,
+      "loss": 1.0876,
+      "step": 16508
+    },
+    {
+      "epoch": 2.6414400000000002,
+      "grad_norm": 1.9605101354132295,
+      "learning_rate": 7.550195317235798e-06,
+      "loss": 1.0455,
+      "step": 16509
+    },
+    {
+      "epoch": 2.6416,
+      "grad_norm": 1.9009851790499102,
+      "learning_rate": 7.543612138415301e-06,
+      "loss": 1.0314,
+      "step": 16510
+    },
+    {
+      "epoch": 2.64176,
+      "grad_norm": 2.3832981639989637,
+      "learning_rate": 7.537031718358656e-06,
+      "loss": 1.0958,
+      "step": 16511
+    },
+    {
+      "epoch": 2.64192,
+      "grad_norm": 1.8926480459079524,
+      "learning_rate": 7.530454057262293e-06,
+      "loss": 1.063,
+      "step": 16512
+    },
+    {
+      "epoch": 2.64208,
+      "grad_norm": 1.8716743686962545,
+      "learning_rate": 7.523879155322411e-06,
+      "loss": 1.0652,
+      "step": 16513
+    },
+    {
+      "epoch": 2.64224,
+      "grad_norm": 1.6964699306754634,
+      "learning_rate": 7.517307012735242e-06,
+      "loss": 1.0523,
+      "step": 16514
+    },
+    {
+      "epoch": 2.6424,
+      "grad_norm": 1.7548328153155544,
+      "learning_rate": 7.510737629696851e-06,
+      "loss": 1.0541,
+      "step": 16515
+    },
+    {
+      "epoch": 2.64256,
+      "grad_norm": 1.5386529373056326,
+      "learning_rate": 7.504171006403282e-06,
+      "loss": 1.0589,
+      "step": 16516
+    },
+    {
+      "epoch": 2.6427199999999997,
+      "grad_norm": 1.6961078788869022,
+      "learning_rate": 7.497607143050478e-06,
+      "loss": 1.0546,
+      "step": 16517
+    },
+    {
+      "epoch": 2.64288,
+      "grad_norm": 1.8257891944481774,
+      "learning_rate": 7.491046039834293e-06,
+      "loss": 1.0697,
+      "step": 16518
+    },
+    {
+      "epoch": 2.64304,
+      "grad_norm": 1.7991969405441461,
+      "learning_rate": 7.4844876969504814e-06,
+      "loss": 1.0397,
+      "step": 16519
+    },
+    {
+      "epoch": 2.6432,
+      "grad_norm": 1.8103076405781866,
+      "learning_rate": 7.477932114594755e-06,
+      "loss": 1.0187,
+      "step": 16520
+    },
+    {
+      "epoch": 2.64336,
+      "grad_norm": 2.0690353673368675,
+      "learning_rate": 7.4713792929627105e-06,
+      "loss": 1.1064,
+      "step": 16521
+    },
+    {
+      "epoch": 2.64352,
+      "grad_norm": 1.8510517921582237,
+      "learning_rate": 7.464829232249882e-06,
+      "loss": 1.1196,
+      "step": 16522
+    },
+    {
+      "epoch": 2.64368,
+      "grad_norm": 1.6414253202857363,
+      "learning_rate": 7.458281932651712e-06,
+      "loss": 1.0909,
+      "step": 16523
+    },
+    {
+      "epoch": 2.64384,
+      "grad_norm": 1.8028730037046197,
+      "learning_rate": 7.451737394363556e-06,
+      "loss": 1.0505,
+      "step": 16524
+    },
+    {
+      "epoch": 2.644,
+      "grad_norm": 1.6901413455053997,
+      "learning_rate": 7.445195617580691e-06,
+      "loss": 1.0682,
+      "step": 16525
+    },
+    {
+      "epoch": 2.64416,
+      "grad_norm": 1.9541869128937424,
+      "learning_rate": 7.4386566024983505e-06,
+      "loss": 1.1097,
+      "step": 16526
+    },
+    {
+      "epoch": 2.64432,
+      "grad_norm": 1.6069799144580772,
+      "learning_rate": 7.432120349311589e-06,
+      "loss": 1.0591,
+      "step": 16527
+    },
+    {
+      "epoch": 2.64448,
+      "grad_norm": 1.6374404899182062,
+      "learning_rate": 7.425586858215506e-06,
+      "loss": 1.1048,
+      "step": 16528
+    },
+    {
+      "epoch": 2.64464,
+      "grad_norm": 1.679469999303298,
+      "learning_rate": 7.419056129404978e-06,
+      "loss": 1.0767,
+      "step": 16529
+    },
+    {
+      "epoch": 2.6448,
+      "grad_norm": 1.6746965482978209,
+      "learning_rate": 7.412528163074939e-06,
+      "loss": 1.0686,
+      "step": 16530
+    },
+    {
+      "epoch": 2.64496,
+      "grad_norm": 1.7434816803230326,
+      "learning_rate": 7.406002959420122e-06,
+      "loss": 1.0681,
+      "step": 16531
+    },
+    {
+      "epoch": 2.64512,
+      "grad_norm": 1.7491532606254159,
+      "learning_rate": 7.399480518635271e-06,
+      "loss": 1.081,
+      "step": 16532
+    },
+    {
+      "epoch": 2.64528,
+      "grad_norm": 1.703167376017038,
+      "learning_rate": 7.392960840914976e-06,
+      "loss": 1.062,
+      "step": 16533
+    },
+    {
+      "epoch": 2.64544,
+      "grad_norm": 1.8782379496964348,
+      "learning_rate": 7.386443926453801e-06,
+      "loss": 1.0908,
+      "step": 16534
+    },
+    {
+      "epoch": 2.6456,
+      "grad_norm": 1.7486805136321697,
+      "learning_rate": 7.379929775446171e-06,
+      "loss": 1.0632,
+      "step": 16535
+    },
+    {
+      "epoch": 2.64576,
+      "grad_norm": 1.7069756213027636,
+      "learning_rate": 7.3734183880864835e-06,
+      "loss": 1.0611,
+      "step": 16536
+    },
+    {
+      "epoch": 2.6459200000000003,
+      "grad_norm": 2.048956401636294,
+      "learning_rate": 7.36690976456903e-06,
+      "loss": 1.1082,
+      "step": 16537
+    },
+    {
+      "epoch": 2.64608,
+      "grad_norm": 1.8207821234700166,
+      "learning_rate": 7.360403905087998e-06,
+      "loss": 1.0504,
+      "step": 16538
+    },
+    {
+      "epoch": 2.64624,
+      "grad_norm": 1.7730410389445401,
+      "learning_rate": 7.3539008098375215e-06,
+      "loss": 1.0987,
+      "step": 16539
+    },
+    {
+      "epoch": 2.6464,
+      "grad_norm": 1.6900303247323416,
+      "learning_rate": 7.347400479011646e-06,
+      "loss": 1.0273,
+      "step": 16540
+    },
+    {
+      "epoch": 2.64656,
+      "grad_norm": 1.7341859460140008,
+      "learning_rate": 7.340902912804315e-06,
+      "loss": 1.0494,
+      "step": 16541
+    },
+    {
+      "epoch": 2.64672,
+      "grad_norm": 1.9463097024053841,
+      "learning_rate": 7.334408111409463e-06,
+      "loss": 1.0866,
+      "step": 16542
+    },
+    {
+      "epoch": 2.64688,
+      "grad_norm": 1.795708945836317,
+      "learning_rate": 7.327916075020802e-06,
+      "loss": 1.0804,
+      "step": 16543
+    },
+    {
+      "epoch": 2.64704,
+      "grad_norm": 1.6799750165158456,
+      "learning_rate": 7.321426803832132e-06,
+      "loss": 1.0507,
+      "step": 16544
+    },
+    {
+      "epoch": 2.6471999999999998,
+      "grad_norm": 2.00607588733994,
+      "learning_rate": 7.314940298036999e-06,
+      "loss": 1.0621,
+      "step": 16545
+    },
+    {
+      "epoch": 2.64736,
+      "grad_norm": 1.613529725497514,
+      "learning_rate": 7.3084565578290244e-06,
+      "loss": 1.0614,
+      "step": 16546
+    },
+    {
+      "epoch": 2.64752,
+      "grad_norm": 1.7991048694746874,
+      "learning_rate": 7.301975583401599e-06,
+      "loss": 1.0502,
+      "step": 16547
+    },
+    {
+      "epoch": 2.6476800000000003,
+      "grad_norm": 1.7127547613244687,
+      "learning_rate": 7.2954973749481796e-06,
+      "loss": 1.0483,
+      "step": 16548
+    },
+    {
+      "epoch": 2.64784,
+      "grad_norm": 1.5231738060312818,
+      "learning_rate": 7.289021932662021e-06,
+      "loss": 0.9974,
+      "step": 16549
+    },
+    {
+      "epoch": 2.648,
+      "grad_norm": 1.791452196630528,
+      "learning_rate": 7.28254925673636e-06,
+      "loss": 1.0805,
+      "step": 16550
+    },
+    {
+      "epoch": 2.64816,
+      "grad_norm": 1.674442224572413,
+      "learning_rate": 7.276079347364318e-06,
+      "loss": 1.0497,
+      "step": 16551
+    },
+    {
+      "epoch": 2.64832,
+      "grad_norm": 1.89552062811553,
+      "learning_rate": 7.269612204738951e-06,
+      "loss": 1.0459,
+      "step": 16552
+    },
+    {
+      "epoch": 2.64848,
+      "grad_norm": 1.701883539155119,
+      "learning_rate": 7.263147829053263e-06,
+      "loss": 1.0785,
+      "step": 16553
+    },
+    {
+      "epoch": 2.64864,
+      "grad_norm": 1.7549202340164949,
+      "learning_rate": 7.2566862205000754e-06,
+      "loss": 1.1118,
+      "step": 16554
+    },
+    {
+      "epoch": 2.6488,
+      "grad_norm": 1.6916981848318877,
+      "learning_rate": 7.250227379272257e-06,
+      "loss": 1.0688,
+      "step": 16555
+    },
+    {
+      "epoch": 2.6489599999999998,
+      "grad_norm": 1.8702462848255006,
+      "learning_rate": 7.243771305562474e-06,
+      "loss": 1.0786,
+      "step": 16556
+    },
+    {
+      "epoch": 2.64912,
+      "grad_norm": 1.8309203280426827,
+      "learning_rate": 7.23731799956342e-06,
+      "loss": 1.0664,
+      "step": 16557
+    },
+    {
+      "epoch": 2.64928,
+      "grad_norm": 1.754036157913778,
+      "learning_rate": 7.2308674614676165e-06,
+      "loss": 1.0519,
+      "step": 16558
+    },
+    {
+      "epoch": 2.6494400000000002,
+      "grad_norm": 1.707776567851464,
+      "learning_rate": 7.224419691467555e-06,
+      "loss": 1.0231,
+      "step": 16559
+    },
+    {
+      "epoch": 2.6496,
+      "grad_norm": 1.8969199506783336,
+      "learning_rate": 7.217974689755625e-06,
+      "loss": 1.0785,
+      "step": 16560
+    },
+    {
+      "epoch": 2.64976,
+      "grad_norm": 1.568499335519783,
+      "learning_rate": 7.211532456524128e-06,
+      "loss": 1.0387,
+      "step": 16561
+    },
+    {
+      "epoch": 2.64992,
+      "grad_norm": 2.036372935720816,
+      "learning_rate": 7.20509299196529e-06,
+      "loss": 1.0896,
+      "step": 16562
+    },
+    {
+      "epoch": 2.65008,
+      "grad_norm": 1.8043569648136664,
+      "learning_rate": 7.198656296271267e-06,
+      "loss": 1.1225,
+      "step": 16563
+    },
+    {
+      "epoch": 2.65024,
+      "grad_norm": 1.7587229657137553,
+      "learning_rate": 7.1922223696341165e-06,
+      "loss": 1.0894,
+      "step": 16564
+    },
+    {
+      "epoch": 2.6504,
+      "grad_norm": 1.6624789656123955,
+      "learning_rate": 7.185791212245807e-06,
+      "loss": 1.0905,
+      "step": 16565
+    },
+    {
+      "epoch": 2.65056,
+      "grad_norm": 1.7129267975011908,
+      "learning_rate": 7.17936282429823e-06,
+      "loss": 1.041,
+      "step": 16566
+    },
+    {
+      "epoch": 2.6507199999999997,
+      "grad_norm": 1.6833599428479704,
+      "learning_rate": 7.172937205983254e-06,
+      "loss": 1.0722,
+      "step": 16567
+    },
+    {
+      "epoch": 2.65088,
+      "grad_norm": 1.972003372933563,
+      "learning_rate": 7.166514357492526e-06,
+      "loss": 1.103,
+      "step": 16568
+    },
+    {
+      "epoch": 2.65104,
+      "grad_norm": 1.9599781297933845,
+      "learning_rate": 7.16009427901777e-06,
+      "loss": 1.1,
+      "step": 16569
+    },
+    {
+      "epoch": 2.6512000000000002,
+      "grad_norm": 1.84452308693034,
+      "learning_rate": 7.153676970750489e-06,
+      "loss": 1.0569,
+      "step": 16570
+    },
+    {
+      "epoch": 2.65136,
+      "grad_norm": 1.7453614293920297,
+      "learning_rate": 7.147262432882229e-06,
+      "loss": 1.0845,
+      "step": 16571
+    },
+    {
+      "epoch": 2.65152,
+      "grad_norm": 1.7196975832331793,
+      "learning_rate": 7.1408506656043264e-06,
+      "loss": 1.0779,
+      "step": 16572
+    },
+    {
+      "epoch": 2.65168,
+      "grad_norm": 1.9020162259893107,
+      "learning_rate": 7.134441669108138e-06,
+      "loss": 1.0582,
+      "step": 16573
+    },
+    {
+      "epoch": 2.65184,
+      "grad_norm": 1.6748701437116116,
+      "learning_rate": 7.128035443584902e-06,
+      "loss": 1.0774,
+      "step": 16574
+    },
+    {
+      "epoch": 2.652,
+      "grad_norm": 1.7838560128715444,
+      "learning_rate": 7.1216319892257525e-06,
+      "loss": 1.131,
+      "step": 16575
+    },
+    {
+      "epoch": 2.65216,
+      "grad_norm": 1.9100795290233263,
+      "learning_rate": 7.115231306221759e-06,
+      "loss": 1.0902,
+      "step": 16576
+    },
+    {
+      "epoch": 2.65232,
+      "grad_norm": 1.7548516437661723,
+      "learning_rate": 7.108833394763925e-06,
+      "loss": 1.0764,
+      "step": 16577
+    },
+    {
+      "epoch": 2.6524799999999997,
+      "grad_norm": 1.7190525255321645,
+      "learning_rate": 7.102438255043153e-06,
+      "loss": 1.0286,
+      "step": 16578
+    },
+    {
+      "epoch": 2.65264,
+      "grad_norm": 1.6614503467758894,
+      "learning_rate": 7.096045887250247e-06,
+      "loss": 1.0862,
+      "step": 16579
+    },
+    {
+      "epoch": 2.6528,
+      "grad_norm": 2.0569113416539615,
+      "learning_rate": 7.089656291575952e-06,
+      "loss": 1.1269,
+      "step": 16580
+    },
+    {
+      "epoch": 2.65296,
+      "grad_norm": 1.798250709041674,
+      "learning_rate": 7.08326946821094e-06,
+      "loss": 1.0989,
+      "step": 16581
+    },
+    {
+      "epoch": 2.65312,
+      "grad_norm": 1.801722449884284,
+      "learning_rate": 7.076885417345758e-06,
+      "loss": 1.0918,
+      "step": 16582
+    },
+    {
+      "epoch": 2.65328,
+      "grad_norm": 1.7332054585127157,
+      "learning_rate": 7.070504139170931e-06,
+      "loss": 1.0059,
+      "step": 16583
+    },
+    {
+      "epoch": 2.65344,
+      "grad_norm": 1.7610023893343538,
+      "learning_rate": 7.064125633876828e-06,
+      "loss": 1.054,
+      "step": 16584
+    },
+    {
+      "epoch": 2.6536,
+      "grad_norm": 1.5146536483867545,
+      "learning_rate": 7.05774990165382e-06,
+      "loss": 1.0156,
+      "step": 16585
+    },
+    {
+      "epoch": 2.65376,
+      "grad_norm": 1.6984783193306363,
+      "learning_rate": 7.05137694269209e-06,
+      "loss": 1.0571,
+      "step": 16586
+    },
+    {
+      "epoch": 2.65392,
+      "grad_norm": 1.901097234853623,
+      "learning_rate": 7.045006757181871e-06,
+      "loss": 1.0999,
+      "step": 16587
+    },
+    {
+      "epoch": 2.65408,
+      "grad_norm": 1.7713473786948999,
+      "learning_rate": 7.038639345313158e-06,
+      "loss": 1.0565,
+      "step": 16588
+    },
+    {
+      "epoch": 2.65424,
+      "grad_norm": 1.8673667121387785,
+      "learning_rate": 7.03227470727601e-06,
+      "loss": 1.0865,
+      "step": 16589
+    },
+    {
+      "epoch": 2.6544,
+      "grad_norm": 1.7807076661513193,
+      "learning_rate": 7.025912843260319e-06,
+      "loss": 1.0593,
+      "step": 16590
+    },
+    {
+      "epoch": 2.65456,
+      "grad_norm": 1.7558433185151614,
+      "learning_rate": 7.0195537534559095e-06,
+      "loss": 1.0601,
+      "step": 16591
+    },
+    {
+      "epoch": 2.65472,
+      "grad_norm": 1.7818175919749721,
+      "learning_rate": 7.013197438052532e-06,
+      "loss": 1.1071,
+      "step": 16592
+    },
+    {
+      "epoch": 2.65488,
+      "grad_norm": 1.8493288125489813,
+      "learning_rate": 7.006843897239846e-06,
+      "loss": 1.1166,
+      "step": 16593
+    },
+    {
+      "epoch": 2.65504,
+      "grad_norm": 1.806796920469894,
+      "learning_rate": 7.000493131207431e-06,
+      "loss": 1.0786,
+      "step": 16594
+    },
+    {
+      "epoch": 2.6552,
+      "grad_norm": 1.70237146163608,
+      "learning_rate": 6.994145140144781e-06,
+      "loss": 1.1011,
+      "step": 16595
+    },
+    {
+      "epoch": 2.65536,
+      "grad_norm": 1.6579244033804337,
+      "learning_rate": 6.987799924241323e-06,
+      "loss": 1.0731,
+      "step": 16596
+    },
+    {
+      "epoch": 2.65552,
+      "grad_norm": 1.7246192670644562,
+      "learning_rate": 6.981457483686371e-06,
+      "loss": 1.0778,
+      "step": 16597
+    },
+    {
+      "epoch": 2.6556800000000003,
+      "grad_norm": 1.765361757091676,
+      "learning_rate": 6.975117818669186e-06,
+      "loss": 1.0829,
+      "step": 16598
+    },
+    {
+      "epoch": 2.65584,
+      "grad_norm": 1.8987901215816514,
+      "learning_rate": 6.96878092937896e-06,
+      "loss": 1.09,
+      "step": 16599
+    },
+    {
+      "epoch": 2.656,
+      "grad_norm": 1.6904631749326535,
+      "learning_rate": 6.962446816004709e-06,
+      "loss": 1.0408,
+      "step": 16600
+    },
+    {
+      "epoch": 2.65616,
+      "grad_norm": 1.9590247838284929,
+      "learning_rate": 6.956115478735514e-06,
+      "loss": 1.1189,
+      "step": 16601
+    },
+    {
+      "epoch": 2.65632,
+      "grad_norm": 1.9014311671727848,
+      "learning_rate": 6.949786917760226e-06,
+      "loss": 1.079,
+      "step": 16602
+    },
+    {
+      "epoch": 2.65648,
+      "grad_norm": 1.7462265737323133,
+      "learning_rate": 6.943461133267737e-06,
+      "loss": 1.0613,
+      "step": 16603
+    },
+    {
+      "epoch": 2.65664,
+      "grad_norm": 1.764942987013084,
+      "learning_rate": 6.93713812544673e-06,
+      "loss": 1.0547,
+      "step": 16604
+    },
+    {
+      "epoch": 2.6568,
+      "grad_norm": 1.996755541060262,
+      "learning_rate": 6.930817894485942e-06,
+      "loss": 1.0693,
+      "step": 16605
+    },
+    {
+      "epoch": 2.6569599999999998,
+      "grad_norm": 1.8333740740175113,
+      "learning_rate": 6.924500440573922e-06,
+      "loss": 1.0866,
+      "step": 16606
+    },
+    {
+      "epoch": 2.65712,
+      "grad_norm": 1.6238121892454347,
+      "learning_rate": 6.9181857638991986e-06,
+      "loss": 1.0666,
+      "step": 16607
+    },
+    {
+      "epoch": 2.65728,
+      "grad_norm": 1.7376012262915383,
+      "learning_rate": 6.911873864650165e-06,
+      "loss": 1.0718,
+      "step": 16608
+    },
+    {
+      "epoch": 2.6574400000000002,
+      "grad_norm": 1.7427670615566733,
+      "learning_rate": 6.90556474301517e-06,
+      "loss": 1.0431,
+      "step": 16609
+    },
+    {
+      "epoch": 2.6576,
+      "grad_norm": 1.4382841090550726,
+      "learning_rate": 6.8992583991824735e-06,
+      "loss": 1.0989,
+      "step": 16610
+    },
+    {
+      "epoch": 2.65776,
+      "grad_norm": 1.6581626591129943,
+      "learning_rate": 6.892954833340226e-06,
+      "loss": 1.0834,
+      "step": 16611
+    },
+    {
+      "epoch": 2.65792,
+      "grad_norm": 1.676180303005088,
+      "learning_rate": 6.8866540456765774e-06,
+      "loss": 1.1003,
+      "step": 16612
+    },
+    {
+      "epoch": 2.65808,
+      "grad_norm": 1.8160878053249983,
+      "learning_rate": 6.880356036379454e-06,
+      "loss": 1.0961,
+      "step": 16613
+    },
+    {
+      "epoch": 2.65824,
+      "grad_norm": 1.557361344376133,
+      "learning_rate": 6.874060805636828e-06,
+      "loss": 1.0713,
+      "step": 16614
+    },
+    {
+      "epoch": 2.6584,
+      "grad_norm": 1.7120235157737995,
+      "learning_rate": 6.867768353636528e-06,
+      "loss": 1.0578,
+      "step": 16615
+    },
+    {
+      "epoch": 2.65856,
+      "grad_norm": 1.5697486156865357,
+      "learning_rate": 6.8614786805663246e-06,
+      "loss": 1.0247,
+      "step": 16616
+    },
+    {
+      "epoch": 2.6587199999999998,
+      "grad_norm": 1.5494596316293816,
+      "learning_rate": 6.855191786613868e-06,
+      "loss": 1.0444,
+      "step": 16617
+    },
+    {
+      "epoch": 2.65888,
+      "grad_norm": 1.851436772873892,
+      "learning_rate": 6.848907671966765e-06,
+      "loss": 1.0678,
+      "step": 16618
+    },
+    {
+      "epoch": 2.65904,
+      "grad_norm": 1.692573862994979,
+      "learning_rate": 6.842626336812519e-06,
+      "loss": 1.0728,
+      "step": 16619
+    },
+    {
+      "epoch": 2.6592000000000002,
+      "grad_norm": 1.69404316598082,
+      "learning_rate": 6.836347781338559e-06,
+      "loss": 1.0689,
+      "step": 16620
+    },
+    {
+      "epoch": 2.65936,
+      "grad_norm": 1.8806249478951724,
+      "learning_rate": 6.8300720057322355e-06,
+      "loss": 1.0696,
+      "step": 16621
+    },
+    {
+      "epoch": 2.65952,
+      "grad_norm": 1.8115433895991326,
+      "learning_rate": 6.823799010180798e-06,
+      "loss": 1.0599,
+      "step": 16622
+    },
+    {
+      "epoch": 2.65968,
+      "grad_norm": 1.74846694777749,
+      "learning_rate": 6.8175287948714084e-06,
+      "loss": 1.0777,
+      "step": 16623
+    },
+    {
+      "epoch": 2.65984,
+      "grad_norm": 1.6206850740822838,
+      "learning_rate": 6.811261359991206e-06,
+      "loss": 1.1005,
+      "step": 16624
+    },
+    {
+      "epoch": 2.66,
+      "grad_norm": 1.7709633542527918,
+      "learning_rate": 6.80499670572714e-06,
+      "loss": 1.0437,
+      "step": 16625
+    },
+    {
+      "epoch": 2.66016,
+      "grad_norm": 1.9335544509300246,
+      "learning_rate": 6.798734832266218e-06,
+      "loss": 1.0848,
+      "step": 16626
+    },
+    {
+      "epoch": 2.66032,
+      "grad_norm": 1.6560021475911035,
+      "learning_rate": 6.792475739795201e-06,
+      "loss": 1.0593,
+      "step": 16627
+    },
+    {
+      "epoch": 2.6604799999999997,
+      "grad_norm": 1.7805899071579605,
+      "learning_rate": 6.786219428500918e-06,
+      "loss": 1.0874,
+      "step": 16628
+    },
+    {
+      "epoch": 2.66064,
+      "grad_norm": 1.7578509704367027,
+      "learning_rate": 6.779965898569996e-06,
+      "loss": 1.1013,
+      "step": 16629
+    },
+    {
+      "epoch": 2.6608,
+      "grad_norm": 1.5943213679049433,
+      "learning_rate": 6.773715150189064e-06,
+      "loss": 1.0568,
+      "step": 16630
+    },
+    {
+      "epoch": 2.66096,
+      "grad_norm": 1.92725803617634,
+      "learning_rate": 6.767467183544629e-06,
+      "loss": 1.0927,
+      "step": 16631
+    },
+    {
+      "epoch": 2.66112,
+      "grad_norm": 1.6967810607987794,
+      "learning_rate": 6.76122199882312e-06,
+      "loss": 1.039,
+      "step": 16632
+    },
+    {
+      "epoch": 2.66128,
+      "grad_norm": 1.5875869610233566,
+      "learning_rate": 6.7549795962108864e-06,
+      "loss": 1.0647,
+      "step": 16633
+    },
+    {
+      "epoch": 2.66144,
+      "grad_norm": 1.7493160580494331,
+      "learning_rate": 6.748739975894181e-06,
+      "loss": 1.0799,
+      "step": 16634
+    },
+    {
+      "epoch": 2.6616,
+      "grad_norm": 1.7260958807988322,
+      "learning_rate": 6.742503138059198e-06,
+      "loss": 1.0481,
+      "step": 16635
+    },
+    {
+      "epoch": 2.66176,
+      "grad_norm": 1.75821222237887,
+      "learning_rate": 6.736269082892033e-06,
+      "loss": 1.0409,
+      "step": 16636
+    },
+    {
+      "epoch": 2.66192,
+      "grad_norm": 1.9705796148340045,
+      "learning_rate": 6.7300378105787044e-06,
+      "loss": 1.0833,
+      "step": 16637
+    },
+    {
+      "epoch": 2.66208,
+      "grad_norm": 1.6115561206314761,
+      "learning_rate": 6.72380932130513e-06,
+      "loss": 1.0409,
+      "step": 16638
+    },
+    {
+      "epoch": 2.66224,
+      "grad_norm": 1.7289588320443123,
+      "learning_rate": 6.71758361525715e-06,
+      "loss": 1.1047,
+      "step": 16639
+    },
+    {
+      "epoch": 2.6624,
+      "grad_norm": 1.7998518135153383,
+      "learning_rate": 6.711360692620583e-06,
+      "loss": 1.0835,
+      "step": 16640
+    },
+    {
+      "epoch": 2.66256,
+      "grad_norm": 1.9819514906723192,
+      "learning_rate": 6.705140553581046e-06,
+      "loss": 1.0592,
+      "step": 16641
+    },
+    {
+      "epoch": 2.66272,
+      "grad_norm": 1.6648922392289363,
+      "learning_rate": 6.698923198324203e-06,
+      "loss": 1.0804,
+      "step": 16642
+    },
+    {
+      "epoch": 2.66288,
+      "grad_norm": 1.6346776096786784,
+      "learning_rate": 6.692708627035505e-06,
+      "loss": 1.0455,
+      "step": 16643
+    },
+    {
+      "epoch": 2.66304,
+      "grad_norm": 1.8559716002161064,
+      "learning_rate": 6.686496839900447e-06,
+      "loss": 1.0336,
+      "step": 16644
+    },
+    {
+      "epoch": 2.6632,
+      "grad_norm": 1.7347388134794102,
+      "learning_rate": 6.6802878371043266e-06,
+      "loss": 1.0277,
+      "step": 16645
+    },
+    {
+      "epoch": 2.66336,
+      "grad_norm": 1.721352841976973,
+      "learning_rate": 6.674081618832451e-06,
+      "loss": 1.0424,
+      "step": 16646
+    },
+    {
+      "epoch": 2.66352,
+      "grad_norm": 1.754118059083687,
+      "learning_rate": 6.6678781852699825e-06,
+      "loss": 1.0683,
+      "step": 16647
+    },
+    {
+      "epoch": 2.6636800000000003,
+      "grad_norm": 1.6837787275725329,
+      "learning_rate": 6.661677536602029e-06,
+      "loss": 1.0796,
+      "step": 16648
+    },
+    {
+      "epoch": 2.66384,
+      "grad_norm": 1.7549477691536592,
+      "learning_rate": 6.655479673013621e-06,
+      "loss": 1.105,
+      "step": 16649
+    },
+    {
+      "epoch": 2.664,
+      "grad_norm": 1.6758727954491819,
+      "learning_rate": 6.649284594689675e-06,
+      "loss": 1.0699,
+      "step": 16650
+    },
+    {
+      "epoch": 2.66416,
+      "grad_norm": 1.7113458878319139,
+      "learning_rate": 6.643092301815046e-06,
+      "loss": 1.0264,
+      "step": 16651
+    },
+    {
+      "epoch": 2.66432,
+      "grad_norm": 1.886675784341746,
+      "learning_rate": 6.6369027945745176e-06,
+      "loss": 1.1267,
+      "step": 16652
+    },
+    {
+      "epoch": 2.66448,
+      "grad_norm": 1.5887157577061684,
+      "learning_rate": 6.630716073152754e-06,
+      "loss": 1.0543,
+      "step": 16653
+    },
+    {
+      "epoch": 2.66464,
+      "grad_norm": 1.8327737512015343,
+      "learning_rate": 6.624532137734374e-06,
+      "loss": 1.1177,
+      "step": 16654
+    },
+    {
+      "epoch": 2.6648,
+      "grad_norm": 1.6227336899955056,
+      "learning_rate": 6.618350988503885e-06,
+      "loss": 1.0993,
+      "step": 16655
+    },
+    {
+      "epoch": 2.6649599999999998,
+      "grad_norm": 1.7117719109703287,
+      "learning_rate": 6.612172625645752e-06,
+      "loss": 1.0499,
+      "step": 16656
+    },
+    {
+      "epoch": 2.66512,
+      "grad_norm": 1.6086231332427408,
+      "learning_rate": 6.605997049344281e-06,
+      "loss": 1.0767,
+      "step": 16657
+    },
+    {
+      "epoch": 2.66528,
+      "grad_norm": 1.8214826582521642,
+      "learning_rate": 6.599824259783793e-06,
+      "loss": 1.0382,
+      "step": 16658
+    },
+    {
+      "epoch": 2.6654400000000003,
+      "grad_norm": 1.6517210448914728,
+      "learning_rate": 6.5936542571484295e-06,
+      "loss": 1.0643,
+      "step": 16659
+    },
+    {
+      "epoch": 2.6656,
+      "grad_norm": 1.651819139936185,
+      "learning_rate": 6.587487041622342e-06,
+      "loss": 1.0556,
+      "step": 16660
+    },
+    {
+      "epoch": 2.66576,
+      "grad_norm": 1.697431302410647,
+      "learning_rate": 6.581322613389484e-06,
+      "loss": 1.1024,
+      "step": 16661
+    },
+    {
+      "epoch": 2.66592,
+      "grad_norm": 1.805774169178222,
+      "learning_rate": 6.575160972633865e-06,
+      "loss": 1.0555,
+      "step": 16662
+    },
+    {
+      "epoch": 2.66608,
+      "grad_norm": 1.823683214660601,
+      "learning_rate": 6.569002119539302e-06,
+      "loss": 1.0623,
+      "step": 16663
+    },
+    {
+      "epoch": 2.66624,
+      "grad_norm": 1.645860932679413,
+      "learning_rate": 6.562846054289573e-06,
+      "loss": 1.0549,
+      "step": 16664
+    },
+    {
+      "epoch": 2.6664,
+      "grad_norm": 2.25208923376021,
+      "learning_rate": 6.556692777068363e-06,
+      "loss": 1.1203,
+      "step": 16665
+    },
+    {
+      "epoch": 2.66656,
+      "grad_norm": 1.6343928248935198,
+      "learning_rate": 6.550542288059281e-06,
+      "loss": 1.0953,
+      "step": 16666
+    },
+    {
+      "epoch": 2.6667199999999998,
+      "grad_norm": 1.803225236840368,
+      "learning_rate": 6.5443945874458455e-06,
+      "loss": 1.1204,
+      "step": 16667
+    },
+    {
+      "epoch": 2.66688,
+      "grad_norm": 2.0317469638347516,
+      "learning_rate": 6.538249675411479e-06,
+      "loss": 1.0895,
+      "step": 16668
+    },
+    {
+      "epoch": 2.66704,
+      "grad_norm": 1.648981122130869,
+      "learning_rate": 6.532107552139588e-06,
+      "loss": 1.0323,
+      "step": 16669
+    },
+    {
+      "epoch": 2.6672000000000002,
+      "grad_norm": 1.9381475339850402,
+      "learning_rate": 6.525968217813383e-06,
+      "loss": 1.114,
+      "step": 16670
+    },
+    {
+      "epoch": 2.66736,
+      "grad_norm": 1.731435813853549,
+      "learning_rate": 6.519831672616095e-06,
+      "loss": 1.0824,
+      "step": 16671
+    },
+    {
+      "epoch": 2.66752,
+      "grad_norm": 1.8384804863256812,
+      "learning_rate": 6.5136979167308095e-06,
+      "loss": 1.097,
+      "step": 16672
+    },
+    {
+      "epoch": 2.66768,
+      "grad_norm": 1.5965542036455775,
+      "learning_rate": 6.50756695034056e-06,
+      "loss": 1.0516,
+      "step": 16673
+    },
+    {
+      "epoch": 2.66784,
+      "grad_norm": 1.9124368893751615,
+      "learning_rate": 6.501438773628277e-06,
+      "loss": 1.0582,
+      "step": 16674
+    },
+    {
+      "epoch": 2.668,
+      "grad_norm": 1.6834222807457329,
+      "learning_rate": 6.4953133867768134e-06,
+      "loss": 1.0504,
+      "step": 16675
+    },
+    {
+      "epoch": 2.66816,
+      "grad_norm": 1.7990512285647473,
+      "learning_rate": 6.489190789968957e-06,
+      "loss": 1.0454,
+      "step": 16676
+    },
+    {
+      "epoch": 2.66832,
+      "grad_norm": 1.9107469431770578,
+      "learning_rate": 6.483070983387385e-06,
+      "loss": 1.0749,
+      "step": 16677
+    },
+    {
+      "epoch": 2.6684799999999997,
+      "grad_norm": 2.098595980320015,
+      "learning_rate": 6.4769539672147055e-06,
+      "loss": 1.1184,
+      "step": 16678
+    },
+    {
+      "epoch": 2.66864,
+      "grad_norm": 1.6764347186599529,
+      "learning_rate": 6.470839741633438e-06,
+      "loss": 1.1025,
+      "step": 16679
+    },
+    {
+      "epoch": 2.6688,
+      "grad_norm": 1.886951739719016,
+      "learning_rate": 6.464728306826018e-06,
+      "loss": 1.0738,
+      "step": 16680
+    },
+    {
+      "epoch": 2.66896,
+      "grad_norm": 1.6334205095049703,
+      "learning_rate": 6.458619662974841e-06,
+      "loss": 1.0607,
+      "step": 16681
+    },
+    {
+      "epoch": 2.66912,
+      "grad_norm": 1.7788788784105338,
+      "learning_rate": 6.452513810262117e-06,
+      "loss": 1.0375,
+      "step": 16682
+    },
+    {
+      "epoch": 2.66928,
+      "grad_norm": 1.606391595388745,
+      "learning_rate": 6.446410748870091e-06,
+      "loss": 1.0671,
+      "step": 16683
+    },
+    {
+      "epoch": 2.66944,
+      "grad_norm": 1.7318748280178442,
+      "learning_rate": 6.440310478980827e-06,
+      "loss": 1.0918,
+      "step": 16684
+    },
+    {
+      "epoch": 2.6696,
+      "grad_norm": 2.0363668564797512,
+      "learning_rate": 6.43421300077639e-06,
+      "loss": 1.046,
+      "step": 16685
+    },
+    {
+      "epoch": 2.66976,
+      "grad_norm": 1.6719901716557244,
+      "learning_rate": 6.4281183144386695e-06,
+      "loss": 1.02,
+      "step": 16686
+    },
+    {
+      "epoch": 2.66992,
+      "grad_norm": 1.6181891560137613,
+      "learning_rate": 6.4220264201495625e-06,
+      "loss": 1.0894,
+      "step": 16687
+    },
+    {
+      "epoch": 2.67008,
+      "grad_norm": 2.0718253953321524,
+      "learning_rate": 6.4159373180908256e-06,
+      "loss": 1.0876,
+      "step": 16688
+    },
+    {
+      "epoch": 2.6702399999999997,
+      "grad_norm": 1.5962587738578815,
+      "learning_rate": 6.4098510084441675e-06,
+      "loss": 1.0654,
+      "step": 16689
+    },
+    {
+      "epoch": 2.6704,
+      "grad_norm": 1.810232120215129,
+      "learning_rate": 6.403767491391166e-06,
+      "loss": 1.0674,
+      "step": 16690
+    },
+    {
+      "epoch": 2.67056,
+      "grad_norm": 1.7740554164187614,
+      "learning_rate": 6.397686767113364e-06,
+      "loss": 1.0842,
+      "step": 16691
+    },
+    {
+      "epoch": 2.67072,
+      "grad_norm": 1.7510908538094443,
+      "learning_rate": 6.391608835792195e-06,
+      "loss": 1.086,
+      "step": 16692
+    },
+    {
+      "epoch": 2.67088,
+      "grad_norm": 1.9454729498198557,
+      "learning_rate": 6.385533697609014e-06,
+      "loss": 1.0618,
+      "step": 16693
+    },
+    {
+      "epoch": 2.67104,
+      "grad_norm": 1.853452104477374,
+      "learning_rate": 6.379461352745109e-06,
+      "loss": 1.066,
+      "step": 16694
+    },
+    {
+      "epoch": 2.6712,
+      "grad_norm": 1.7587029411698585,
+      "learning_rate": 6.373391801381645e-06,
+      "loss": 1.1299,
+      "step": 16695
+    },
+    {
+      "epoch": 2.67136,
+      "grad_norm": 1.8259530534938517,
+      "learning_rate": 6.367325043699734e-06,
+      "loss": 1.0471,
+      "step": 16696
+    },
+    {
+      "epoch": 2.67152,
+      "grad_norm": 1.7915078571490262,
+      "learning_rate": 6.361261079880432e-06,
+      "loss": 1.0457,
+      "step": 16697
+    },
+    {
+      "epoch": 2.6716800000000003,
+      "grad_norm": 1.7973489861744394,
+      "learning_rate": 6.355199910104614e-06,
+      "loss": 1.0503,
+      "step": 16698
+    },
+    {
+      "epoch": 2.67184,
+      "grad_norm": 1.9269866216966838,
+      "learning_rate": 6.349141534553227e-06,
+      "loss": 1.0718,
+      "step": 16699
+    },
+    {
+      "epoch": 2.672,
+      "grad_norm": 1.7874382386722942,
+      "learning_rate": 6.3430859534069465e-06,
+      "loss": 1.1326,
+      "step": 16700
+    },
+    {
+      "epoch": 2.67216,
+      "grad_norm": 1.6207438411718447,
+      "learning_rate": 6.33703316684654e-06,
+      "loss": 1.0936,
+      "step": 16701
+    },
+    {
+      "epoch": 2.67232,
+      "grad_norm": 1.8271572006599446,
+      "learning_rate": 6.330983175052563e-06,
+      "loss": 1.0392,
+      "step": 16702
+    },
+    {
+      "epoch": 2.67248,
+      "grad_norm": 1.9633068255902626,
+      "learning_rate": 6.32493597820556e-06,
+      "loss": 1.0552,
+      "step": 16703
+    },
+    {
+      "epoch": 2.67264,
+      "grad_norm": 1.5635324116689455,
+      "learning_rate": 6.318891576485975e-06,
+      "loss": 1.0533,
+      "step": 16704
+    },
+    {
+      "epoch": 2.6728,
+      "grad_norm": 1.9840793388919642,
+      "learning_rate": 6.312849970074152e-06,
+      "loss": 1.0929,
+      "step": 16705
+    },
+    {
+      "epoch": 2.67296,
+      "grad_norm": 1.6749524679685766,
+      "learning_rate": 6.306811159150383e-06,
+      "loss": 1.0682,
+      "step": 16706
+    },
+    {
+      "epoch": 2.67312,
+      "grad_norm": 1.858985529132933,
+      "learning_rate": 6.300775143894833e-06,
+      "loss": 1.1405,
+      "step": 16707
+    },
+    {
+      "epoch": 2.67328,
+      "grad_norm": 1.499576020593364,
+      "learning_rate": 6.294741924487612e-06,
+      "loss": 1.035,
+      "step": 16708
+    },
+    {
+      "epoch": 2.6734400000000003,
+      "grad_norm": 1.8826487126868843,
+      "learning_rate": 6.288711501108757e-06,
+      "loss": 1.0587,
+      "step": 16709
+    },
+    {
+      "epoch": 2.6736,
+      "grad_norm": 1.802794375811451,
+      "learning_rate": 6.2826838739382e-06,
+      "loss": 1.0607,
+      "step": 16710
+    },
+    {
+      "epoch": 2.67376,
+      "grad_norm": 1.82225697106221,
+      "learning_rate": 6.276659043155797e-06,
+      "loss": 1.0346,
+      "step": 16711
+    },
+    {
+      "epoch": 2.67392,
+      "grad_norm": 1.7672683276470253,
+      "learning_rate": 6.2706370089413045e-06,
+      "loss": 1.0782,
+      "step": 16712
+    },
+    {
+      "epoch": 2.67408,
+      "grad_norm": 1.7861990376704537,
+      "learning_rate": 6.264617771474457e-06,
+      "loss": 1.0464,
+      "step": 16713
+    },
+    {
+      "epoch": 2.67424,
+      "grad_norm": 1.5859805615438696,
+      "learning_rate": 6.258601330934788e-06,
+      "loss": 1.0613,
+      "step": 16714
+    },
+    {
+      "epoch": 2.6744,
+      "grad_norm": 1.9732441317113454,
+      "learning_rate": 6.252587687501898e-06,
+      "loss": 1.1073,
+      "step": 16715
+    },
+    {
+      "epoch": 2.67456,
+      "grad_norm": 1.8087091323991238,
+      "learning_rate": 6.2465768413551565e-06,
+      "loss": 1.0933,
+      "step": 16716
+    },
+    {
+      "epoch": 2.6747199999999998,
+      "grad_norm": 1.8967829135621899,
+      "learning_rate": 6.240568792673973e-06,
+      "loss": 1.0694,
+      "step": 16717
+    },
+    {
+      "epoch": 2.67488,
+      "grad_norm": 1.656163292547558,
+      "learning_rate": 6.234563541637573e-06,
+      "loss": 1.0553,
+      "step": 16718
+    },
+    {
+      "epoch": 2.67504,
+      "grad_norm": 1.625148579894448,
+      "learning_rate": 6.228561088425178e-06,
+      "loss": 1.0431,
+      "step": 16719
+    },
+    {
+      "epoch": 2.6752000000000002,
+      "grad_norm": 1.8580377980044325,
+      "learning_rate": 6.222561433215879e-06,
+      "loss": 1.0703,
+      "step": 16720
+    },
+    {
+      "epoch": 2.67536,
+      "grad_norm": 1.744670099487896,
+      "learning_rate": 6.2165645761887e-06,
+      "loss": 1.0717,
+      "step": 16721
+    },
+    {
+      "epoch": 2.67552,
+      "grad_norm": 1.77787612093104,
+      "learning_rate": 6.2105705175225736e-06,
+      "loss": 1.0577,
+      "step": 16722
+    },
+    {
+      "epoch": 2.67568,
+      "grad_norm": 1.6957112521267022,
+      "learning_rate": 6.204579257396359e-06,
+      "loss": 1.0591,
+      "step": 16723
+    },
+    {
+      "epoch": 2.67584,
+      "grad_norm": 1.8848261143821425,
+      "learning_rate": 6.198590795988824e-06,
+      "loss": 1.0985,
+      "step": 16724
+    },
+    {
+      "epoch": 2.676,
+      "grad_norm": 1.5930230560336438,
+      "learning_rate": 6.192605133478657e-06,
+      "loss": 1.092,
+      "step": 16725
+    },
+    {
+      "epoch": 2.67616,
+      "grad_norm": 2.187145793345799,
+      "learning_rate": 6.186622270044462e-06,
+      "loss": 1.0684,
+      "step": 16726
+    },
+    {
+      "epoch": 2.67632,
+      "grad_norm": 1.7910089018174489,
+      "learning_rate": 6.180642205864751e-06,
+      "loss": 1.0618,
+      "step": 16727
+    },
+    {
+      "epoch": 2.6764799999999997,
+      "grad_norm": 1.7841396101443996,
+      "learning_rate": 6.174664941117969e-06,
+      "loss": 1.0765,
+      "step": 16728
+    },
+    {
+      "epoch": 2.67664,
+      "grad_norm": 1.8420269105603795,
+      "learning_rate": 6.1686904759824746e-06,
+      "loss": 1.1155,
+      "step": 16729
+    },
+    {
+      "epoch": 2.6768,
+      "grad_norm": 1.701362549910364,
+      "learning_rate": 6.1627188106365255e-06,
+      "loss": 1.09,
+      "step": 16730
+    },
+    {
+      "epoch": 2.6769600000000002,
+      "grad_norm": 1.6724834727687485,
+      "learning_rate": 6.156749945258322e-06,
+      "loss": 1.0685,
+      "step": 16731
+    },
+    {
+      "epoch": 2.67712,
+      "grad_norm": 1.7754813751333045,
+      "learning_rate": 6.150783880025945e-06,
+      "loss": 1.0877,
+      "step": 16732
+    },
+    {
+      "epoch": 2.67728,
+      "grad_norm": 1.621342193826028,
+      "learning_rate": 6.14482061511743e-06,
+      "loss": 1.0538,
+      "step": 16733
+    },
+    {
+      "epoch": 2.67744,
+      "grad_norm": 1.576352708913261,
+      "learning_rate": 6.138860150710702e-06,
+      "loss": 1.0611,
+      "step": 16734
+    },
+    {
+      "epoch": 2.6776,
+      "grad_norm": 2.15105518095439,
+      "learning_rate": 6.132902486983616e-06,
+      "loss": 1.1305,
+      "step": 16735
+    },
+    {
+      "epoch": 2.67776,
+      "grad_norm": 1.8548447275411097,
+      "learning_rate": 6.126947624113943e-06,
+      "loss": 1.0785,
+      "step": 16736
+    },
+    {
+      "epoch": 2.67792,
+      "grad_norm": 1.4677889799137775,
+      "learning_rate": 6.1209955622793745e-06,
+      "loss": 1.0885,
+      "step": 16737
+    },
+    {
+      "epoch": 2.67808,
+      "grad_norm": 1.6559154259434372,
+      "learning_rate": 6.11504630165749e-06,
+      "loss": 1.0865,
+      "step": 16738
+    },
+    {
+      "epoch": 2.6782399999999997,
+      "grad_norm": 1.8480372645238563,
+      "learning_rate": 6.109099842425803e-06,
+      "loss": 1.0532,
+      "step": 16739
+    },
+    {
+      "epoch": 2.6784,
+      "grad_norm": 1.6650372823871051,
+      "learning_rate": 6.103156184761804e-06,
+      "loss": 1.0627,
+      "step": 16740
+    },
+    {
+      "epoch": 2.67856,
+      "grad_norm": 1.7837994214602368,
+      "learning_rate": 6.097215328842764e-06,
+      "loss": 1.0557,
+      "step": 16741
+    },
+    {
+      "epoch": 2.67872,
+      "grad_norm": 1.728374849563502,
+      "learning_rate": 6.091277274846019e-06,
+      "loss": 1.031,
+      "step": 16742
+    },
+    {
+      "epoch": 2.67888,
+      "grad_norm": 1.829786888830899,
+      "learning_rate": 6.085342022948692e-06,
+      "loss": 1.106,
+      "step": 16743
+    },
+    {
+      "epoch": 2.67904,
+      "grad_norm": 1.9847870937980112,
+      "learning_rate": 6.079409573327932e-06,
+      "loss": 1.1748,
+      "step": 16744
+    },
+    {
+      "epoch": 2.6792,
+      "grad_norm": 1.742291984769699,
+      "learning_rate": 6.07347992616073e-06,
+      "loss": 1.0397,
+      "step": 16745
+    },
+    {
+      "epoch": 2.67936,
+      "grad_norm": 1.862968326821259,
+      "learning_rate": 6.067553081624011e-06,
+      "loss": 1.0904,
+      "step": 16746
+    },
+    {
+      "epoch": 2.67952,
+      "grad_norm": 1.7475644518700288,
+      "learning_rate": 6.061629039894645e-06,
+      "loss": 1.0264,
+      "step": 16747
+    },
+    {
+      "epoch": 2.67968,
+      "grad_norm": 1.6317480951001013,
+      "learning_rate": 6.0557078011493795e-06,
+      "loss": 1.0692,
+      "step": 16748
+    },
+    {
+      "epoch": 2.67984,
+      "grad_norm": 1.7040707881903183,
+      "learning_rate": 6.049789365564906e-06,
+      "loss": 1.0604,
+      "step": 16749
+    },
+    {
+      "epoch": 2.68,
+      "grad_norm": 1.6207146329063131,
+      "learning_rate": 6.043873733317817e-06,
+      "loss": 1.1035,
+      "step": 16750
+    },
+    {
+      "epoch": 2.68016,
+      "grad_norm": 1.856916998335297,
+      "learning_rate": 6.037960904584627e-06,
+      "loss": 1.1222,
+      "step": 16751
+    },
+    {
+      "epoch": 2.68032,
+      "grad_norm": 1.893113427254502,
+      "learning_rate": 6.032050879541773e-06,
+      "loss": 1.0491,
+      "step": 16752
+    },
+    {
+      "epoch": 2.68048,
+      "grad_norm": 1.7818983624032945,
+      "learning_rate": 6.026143658365591e-06,
+      "loss": 1.0654,
+      "step": 16753
+    },
+    {
+      "epoch": 2.68064,
+      "grad_norm": 1.7565900291428342,
+      "learning_rate": 6.020239241232339e-06,
+      "loss": 1.0934,
+      "step": 16754
+    },
+    {
+      "epoch": 2.6808,
+      "grad_norm": 1.85576559730876,
+      "learning_rate": 6.0143376283182005e-06,
+      "loss": 1.0704,
+      "step": 16755
+    },
+    {
+      "epoch": 2.68096,
+      "grad_norm": 1.9697203479819758,
+      "learning_rate": 6.0084388197993e-06,
+      "loss": 1.0363,
+      "step": 16756
+    },
+    {
+      "epoch": 2.68112,
+      "grad_norm": 1.7431569422598483,
+      "learning_rate": 6.002542815851597e-06,
+      "loss": 1.0554,
+      "step": 16757
+    },
+    {
+      "epoch": 2.68128,
+      "grad_norm": 1.7672359125732184,
+      "learning_rate": 5.9966496166510845e-06,
+      "loss": 1.0729,
+      "step": 16758
+    },
+    {
+      "epoch": 2.6814400000000003,
+      "grad_norm": 2.063827487840779,
+      "learning_rate": 5.990759222373532e-06,
+      "loss": 1.137,
+      "step": 16759
+    },
+    {
+      "epoch": 2.6816,
+      "grad_norm": 1.7468965239656407,
+      "learning_rate": 5.984871633194755e-06,
+      "loss": 1.09,
+      "step": 16760
+    },
+    {
+      "epoch": 2.68176,
+      "grad_norm": 1.9164051189823965,
+      "learning_rate": 5.978986849290414e-06,
+      "loss": 1.0479,
+      "step": 16761
+    },
+    {
+      "epoch": 2.68192,
+      "grad_norm": 1.6321417706284493,
+      "learning_rate": 5.973104870836099e-06,
+      "loss": 1.0325,
+      "step": 16762
+    },
+    {
+      "epoch": 2.68208,
+      "grad_norm": 1.7502128781585817,
+      "learning_rate": 5.967225698007329e-06,
+      "loss": 1.0653,
+      "step": 16763
+    },
+    {
+      "epoch": 2.68224,
+      "grad_norm": 1.8101206352209738,
+      "learning_rate": 5.961349330979516e-06,
+      "loss": 1.0185,
+      "step": 16764
+    },
+    {
+      "epoch": 2.6824,
+      "grad_norm": 1.936618937168233,
+      "learning_rate": 5.95547576992801e-06,
+      "loss": 1.0794,
+      "step": 16765
+    },
+    {
+      "epoch": 2.68256,
+      "grad_norm": 1.8553329004887524,
+      "learning_rate": 5.9496050150280815e-06,
+      "loss": 1.1197,
+      "step": 16766
+    },
+    {
+      "epoch": 2.6827199999999998,
+      "grad_norm": 1.7855097144833387,
+      "learning_rate": 5.9437370664548795e-06,
+      "loss": 1.0845,
+      "step": 16767
+    },
+    {
+      "epoch": 2.68288,
+      "grad_norm": 1.7480288008719147,
+      "learning_rate": 5.93787192438352e-06,
+      "loss": 1.0591,
+      "step": 16768
+    },
+    {
+      "epoch": 2.68304,
+      "grad_norm": 1.8138798055588459,
+      "learning_rate": 5.932009588988996e-06,
+      "loss": 1.0784,
+      "step": 16769
+    },
+    {
+      "epoch": 2.6832000000000003,
+      "grad_norm": 1.6503285569916397,
+      "learning_rate": 5.926150060446234e-06,
+      "loss": 1.0629,
+      "step": 16770
+    },
+    {
+      "epoch": 2.68336,
+      "grad_norm": 1.7158233898881032,
+      "learning_rate": 5.920293338930061e-06,
+      "loss": 1.0784,
+      "step": 16771
+    },
+    {
+      "epoch": 2.68352,
+      "grad_norm": 1.6624143723327638,
+      "learning_rate": 5.91443942461527e-06,
+      "loss": 1.0182,
+      "step": 16772
+    },
+    {
+      "epoch": 2.68368,
+      "grad_norm": 1.8126371405631598,
+      "learning_rate": 5.908588317676489e-06,
+      "loss": 1.0269,
+      "step": 16773
+    },
+    {
+      "epoch": 2.68384,
+      "grad_norm": 1.8054257301508925,
+      "learning_rate": 5.902740018288344e-06,
+      "loss": 1.0877,
+      "step": 16774
+    },
+    {
+      "epoch": 2.684,
+      "grad_norm": 1.7917519245107045,
+      "learning_rate": 5.896894526625307e-06,
+      "loss": 1.0856,
+      "step": 16775
+    },
+    {
+      "epoch": 2.68416,
+      "grad_norm": 1.7155446432219752,
+      "learning_rate": 5.891051842861817e-06,
+      "loss": 1.0511,
+      "step": 16776
+    },
+    {
+      "epoch": 2.68432,
+      "grad_norm": 1.6634131234274243,
+      "learning_rate": 5.88521196717221e-06,
+      "loss": 1.0631,
+      "step": 16777
+    },
+    {
+      "epoch": 2.6844799999999998,
+      "grad_norm": 1.6888599714419226,
+      "learning_rate": 5.879374899730749e-06,
+      "loss": 1.0982,
+      "step": 16778
+    },
+    {
+      "epoch": 2.68464,
+      "grad_norm": 1.721523044954552,
+      "learning_rate": 5.873540640711584e-06,
+      "loss": 1.0878,
+      "step": 16779
+    },
+    {
+      "epoch": 2.6848,
+      "grad_norm": 1.8313470026097682,
+      "learning_rate": 5.867709190288817e-06,
+      "loss": 1.0389,
+      "step": 16780
+    },
+    {
+      "epoch": 2.6849600000000002,
+      "grad_norm": 1.6891266711117312,
+      "learning_rate": 5.861880548636433e-06,
+      "loss": 1.0555,
+      "step": 16781
+    },
+    {
+      "epoch": 2.68512,
+      "grad_norm": 1.6922775312660743,
+      "learning_rate": 5.856054715928361e-06,
+      "loss": 1.1248,
+      "step": 16782
+    },
+    {
+      "epoch": 2.68528,
+      "grad_norm": 1.6369974499755302,
+      "learning_rate": 5.850231692338437e-06,
+      "loss": 1.035,
+      "step": 16783
+    },
+    {
+      "epoch": 2.68544,
+      "grad_norm": 1.6909235448904376,
+      "learning_rate": 5.844411478040413e-06,
+      "loss": 1.0697,
+      "step": 16784
+    },
+    {
+      "epoch": 2.6856,
+      "grad_norm": 1.6190603065129723,
+      "learning_rate": 5.838594073207926e-06,
+      "loss": 1.0517,
+      "step": 16785
+    },
+    {
+      "epoch": 2.68576,
+      "grad_norm": 2.064961833135708,
+      "learning_rate": 5.832779478014605e-06,
+      "loss": 1.0727,
+      "step": 16786
+    },
+    {
+      "epoch": 2.68592,
+      "grad_norm": 1.8828394636126846,
+      "learning_rate": 5.826967692633933e-06,
+      "loss": 1.082,
+      "step": 16787
+    },
+    {
+      "epoch": 2.68608,
+      "grad_norm": 1.6129671712896998,
+      "learning_rate": 5.821158717239317e-06,
+      "loss": 1.0511,
+      "step": 16788
+    },
+    {
+      "epoch": 2.6862399999999997,
+      "grad_norm": 1.9012506414030101,
+      "learning_rate": 5.815352552004083e-06,
+      "loss": 1.0891,
+      "step": 16789
+    },
+    {
+      "epoch": 2.6864,
+      "grad_norm": 1.767335056089078,
+      "learning_rate": 5.809549197101505e-06,
+      "loss": 1.0578,
+      "step": 16790
+    },
+    {
+      "epoch": 2.68656,
+      "grad_norm": 1.946680524278586,
+      "learning_rate": 5.80374865270471e-06,
+      "loss": 1.0916,
+      "step": 16791
+    },
+    {
+      "epoch": 2.68672,
+      "grad_norm": 1.6447210060127015,
+      "learning_rate": 5.7979509189868055e-06,
+      "loss": 0.9998,
+      "step": 16792
+    },
+    {
+      "epoch": 2.68688,
+      "grad_norm": 1.8685007503115043,
+      "learning_rate": 5.792155996120785e-06,
+      "loss": 1.0754,
+      "step": 16793
+    },
+    {
+      "epoch": 2.68704,
+      "grad_norm": 1.78826671783885,
+      "learning_rate": 5.786363884279545e-06,
+      "loss": 1.0837,
+      "step": 16794
+    },
+    {
+      "epoch": 2.6872,
+      "grad_norm": 1.931020548628291,
+      "learning_rate": 5.780574583635923e-06,
+      "loss": 1.0537,
+      "step": 16795
+    },
+    {
+      "epoch": 2.68736,
+      "grad_norm": 1.8679903237691167,
+      "learning_rate": 5.774788094362649e-06,
+      "loss": 1.1121,
+      "step": 16796
+    },
+    {
+      "epoch": 2.68752,
+      "grad_norm": 1.9667834106967474,
+      "learning_rate": 5.76900441663244e-06,
+      "loss": 1.0673,
+      "step": 16797
+    },
+    {
+      "epoch": 2.68768,
+      "grad_norm": 1.9320379386990991,
+      "learning_rate": 5.76322355061778e-06,
+      "loss": 1.1701,
+      "step": 16798
+    },
+    {
+      "epoch": 2.68784,
+      "grad_norm": 1.6204931977841412,
+      "learning_rate": 5.757445496491254e-06,
+      "loss": 1.0793,
+      "step": 16799
+    },
+    {
+      "epoch": 2.6879999999999997,
+      "grad_norm": 1.7157169836174881,
+      "learning_rate": 5.751670254425201e-06,
+      "loss": 1.0933,
+      "step": 16800
+    },
+    {
+      "epoch": 2.68816,
+      "grad_norm": 1.712447081346326,
+      "learning_rate": 5.745897824591984e-06,
+      "loss": 1.1393,
+      "step": 16801
+    },
+    {
+      "epoch": 2.68832,
+      "grad_norm": 1.7606161488848862,
+      "learning_rate": 5.74012820716383e-06,
+      "loss": 1.1378,
+      "step": 16802
+    },
+    {
+      "epoch": 2.68848,
+      "grad_norm": 1.9531472010058506,
+      "learning_rate": 5.734361402312904e-06,
+      "loss": 1.0633,
+      "step": 16803
+    },
+    {
+      "epoch": 2.68864,
+      "grad_norm": 1.7276504242087478,
+      "learning_rate": 5.728597410211267e-06,
+      "loss": 1.0748,
+      "step": 16804
+    },
+    {
+      "epoch": 2.6888,
+      "grad_norm": 1.7467756330861939,
+      "learning_rate": 5.7228362310309255e-06,
+      "loss": 1.0766,
+      "step": 16805
+    },
+    {
+      "epoch": 2.68896,
+      "grad_norm": 1.6940336548705455,
+      "learning_rate": 5.717077864943765e-06,
+      "loss": 1.0936,
+      "step": 16806
+    },
+    {
+      "epoch": 2.68912,
+      "grad_norm": 1.5843336714552136,
+      "learning_rate": 5.711322312121625e-06,
+      "loss": 1.0681,
+      "step": 16807
+    },
+    {
+      "epoch": 2.68928,
+      "grad_norm": 1.686010645167752,
+      "learning_rate": 5.705569572736224e-06,
+      "loss": 1.1197,
+      "step": 16808
+    },
+    {
+      "epoch": 2.6894400000000003,
+      "grad_norm": 1.7177201664204964,
+      "learning_rate": 5.699819646959237e-06,
+      "loss": 1.0547,
+      "step": 16809
+    },
+    {
+      "epoch": 2.6896,
+      "grad_norm": 1.742737093969776,
+      "learning_rate": 5.694072534962214e-06,
+      "loss": 1.0121,
+      "step": 16810
+    },
+    {
+      "epoch": 2.68976,
+      "grad_norm": 1.7007150440733392,
+      "learning_rate": 5.688328236916651e-06,
+      "loss": 1.0816,
+      "step": 16811
+    },
+    {
+      "epoch": 2.68992,
+      "grad_norm": 1.7497060811206995,
+      "learning_rate": 5.6825867529939235e-06,
+      "loss": 1.0604,
+      "step": 16812
+    },
+    {
+      "epoch": 2.69008,
+      "grad_norm": 1.5229037549200235,
+      "learning_rate": 5.6768480833654156e-06,
+      "loss": 1.1166,
+      "step": 16813
+    },
+    {
+      "epoch": 2.69024,
+      "grad_norm": 1.938672337206763,
+      "learning_rate": 5.67111222820228e-06,
+      "loss": 1.0851,
+      "step": 16814
+    },
+    {
+      "epoch": 2.6904,
+      "grad_norm": 1.7289847637950668,
+      "learning_rate": 5.665379187675734e-06,
+      "loss": 1.0928,
+      "step": 16815
+    },
+    {
+      "epoch": 2.69056,
+      "grad_norm": 1.6170808202291607,
+      "learning_rate": 5.659648961956787e-06,
+      "loss": 1.057,
+      "step": 16816
+    },
+    {
+      "epoch": 2.69072,
+      "grad_norm": 1.6274607026207364,
+      "learning_rate": 5.653921551216457e-06,
+      "loss": 1.0583,
+      "step": 16817
+    },
+    {
+      "epoch": 2.69088,
+      "grad_norm": 1.700991378047855,
+      "learning_rate": 5.648196955625628e-06,
+      "loss": 1.0501,
+      "step": 16818
+    },
+    {
+      "epoch": 2.69104,
+      "grad_norm": 1.7039786594174515,
+      "learning_rate": 5.642475175355122e-06,
+      "loss": 1.1041,
+      "step": 16819
+    },
+    {
+      "epoch": 2.6912000000000003,
+      "grad_norm": 1.7325330312909633,
+      "learning_rate": 5.636756210575667e-06,
+      "loss": 1.0904,
+      "step": 16820
+    },
+    {
+      "epoch": 2.69136,
+      "grad_norm": 1.7278344283677263,
+      "learning_rate": 5.631040061457904e-06,
+      "loss": 1.1006,
+      "step": 16821
+    },
+    {
+      "epoch": 2.69152,
+      "grad_norm": 1.540466349298439,
+      "learning_rate": 5.625326728172386e-06,
+      "loss": 1.0314,
+      "step": 16822
+    },
+    {
+      "epoch": 2.69168,
+      "grad_norm": 1.8274379904032931,
+      "learning_rate": 5.619616210889611e-06,
+      "loss": 1.1126,
+      "step": 16823
+    },
+    {
+      "epoch": 2.69184,
+      "grad_norm": 1.8545932502438096,
+      "learning_rate": 5.613908509779953e-06,
+      "loss": 1.0381,
+      "step": 16824
+    },
+    {
+      "epoch": 2.692,
+      "grad_norm": 1.7628598934750586,
+      "learning_rate": 5.608203625013741e-06,
+      "loss": 1.0768,
+      "step": 16825
+    },
+    {
+      "epoch": 2.69216,
+      "grad_norm": 1.8337850746287805,
+      "learning_rate": 5.602501556761175e-06,
+      "loss": 1.1092,
+      "step": 16826
+    },
+    {
+      "epoch": 2.69232,
+      "grad_norm": 1.5882885659843828,
+      "learning_rate": 5.596802305192428e-06,
+      "loss": 1.0257,
+      "step": 16827
+    },
+    {
+      "epoch": 2.6924799999999998,
+      "grad_norm": 1.6208734851328896,
+      "learning_rate": 5.591105870477509e-06,
+      "loss": 1.0678,
+      "step": 16828
+    },
+    {
+      "epoch": 2.69264,
+      "grad_norm": 1.8159792658637668,
+      "learning_rate": 5.58541225278646e-06,
+      "loss": 1.0451,
+      "step": 16829
+    },
+    {
+      "epoch": 2.6928,
+      "grad_norm": 1.7426396736588925,
+      "learning_rate": 5.5797214522890996e-06,
+      "loss": 1.0554,
+      "step": 16830
+    },
+    {
+      "epoch": 2.6929600000000002,
+      "grad_norm": 1.8154998479852023,
+      "learning_rate": 5.574033469155304e-06,
+      "loss": 1.0316,
+      "step": 16831
+    },
+    {
+      "epoch": 2.69312,
+      "grad_norm": 1.9181996368639997,
+      "learning_rate": 5.568348303554716e-06,
+      "loss": 1.0948,
+      "step": 16832
+    },
+    {
+      "epoch": 2.69328,
+      "grad_norm": 1.7248829252973086,
+      "learning_rate": 5.562665955657031e-06,
+      "loss": 1.0765,
+      "step": 16833
+    },
+    {
+      "epoch": 2.69344,
+      "grad_norm": 1.8030519401475162,
+      "learning_rate": 5.556986425631794e-06,
+      "loss": 1.0846,
+      "step": 16834
+    },
+    {
+      "epoch": 2.6936,
+      "grad_norm": 1.5787949816366134,
+      "learning_rate": 5.551309713648456e-06,
+      "loss": 1.0996,
+      "step": 16835
+    },
+    {
+      "epoch": 2.69376,
+      "grad_norm": 1.707346198341028,
+      "learning_rate": 5.545635819876405e-06,
+      "loss": 1.0448,
+      "step": 16836
+    },
+    {
+      "epoch": 2.69392,
+      "grad_norm": 1.6306599829542552,
+      "learning_rate": 5.539964744484949e-06,
+      "loss": 1.0768,
+      "step": 16837
+    },
+    {
+      "epoch": 2.69408,
+      "grad_norm": 1.566063443418684,
+      "learning_rate": 5.534296487643309e-06,
+      "loss": 1.0441,
+      "step": 16838
+    },
+    {
+      "epoch": 2.6942399999999997,
+      "grad_norm": 1.772869032019509,
+      "learning_rate": 5.528631049520605e-06,
+      "loss": 1.0772,
+      "step": 16839
+    },
+    {
+      "epoch": 2.6944,
+      "grad_norm": 1.9488899875076549,
+      "learning_rate": 5.522968430285891e-06,
+      "loss": 1.0698,
+      "step": 16840
+    },
+    {
+      "epoch": 2.69456,
+      "grad_norm": 1.777829989462336,
+      "learning_rate": 5.517308630108142e-06,
+      "loss": 1.0962,
+      "step": 16841
+    },
+    {
+      "epoch": 2.6947200000000002,
+      "grad_norm": 1.850558512822286,
+      "learning_rate": 5.511651649156213e-06,
+      "loss": 1.0949,
+      "step": 16842
+    },
+    {
+      "epoch": 2.69488,
+      "grad_norm": 1.7319403353633294,
+      "learning_rate": 5.505997487598924e-06,
+      "loss": 1.0655,
+      "step": 16843
+    },
+    {
+      "epoch": 2.69504,
+      "grad_norm": 1.6484897416458042,
+      "learning_rate": 5.5003461456049625e-06,
+      "loss": 1.0387,
+      "step": 16844
+    },
+    {
+      "epoch": 2.6952,
+      "grad_norm": 1.742433548966416,
+      "learning_rate": 5.494697623342981e-06,
+      "loss": 1.0611,
+      "step": 16845
+    },
+    {
+      "epoch": 2.69536,
+      "grad_norm": 1.9182342206244358,
+      "learning_rate": 5.489051920981514e-06,
+      "loss": 1.0345,
+      "step": 16846
+    },
+    {
+      "epoch": 2.69552,
+      "grad_norm": 2.1608555300443695,
+      "learning_rate": 5.483409038689025e-06,
+      "loss": 1.1177,
+      "step": 16847
+    },
+    {
+      "epoch": 2.69568,
+      "grad_norm": 1.7460801570628466,
+      "learning_rate": 5.477768976633879e-06,
+      "loss": 1.018,
+      "step": 16848
+    },
+    {
+      "epoch": 2.69584,
+      "grad_norm": 1.806380215302962,
+      "learning_rate": 5.472131734984376e-06,
+      "loss": 1.0382,
+      "step": 16849
+    },
+    {
+      "epoch": 2.6959999999999997,
+      "grad_norm": 1.908383481462503,
+      "learning_rate": 5.466497313908725e-06,
+      "loss": 1.0601,
+      "step": 16850
+    },
+    {
+      "epoch": 2.69616,
+      "grad_norm": 1.9217253451126497,
+      "learning_rate": 5.460865713575036e-06,
+      "loss": 1.0888,
+      "step": 16851
+    },
+    {
+      "epoch": 2.69632,
+      "grad_norm": 1.6572729542276408,
+      "learning_rate": 5.455236934151364e-06,
+      "loss": 1.0258,
+      "step": 16852
+    },
+    {
+      "epoch": 2.69648,
+      "grad_norm": 1.8946381128527388,
+      "learning_rate": 5.449610975805641e-06,
+      "loss": 1.0913,
+      "step": 16853
+    },
+    {
+      "epoch": 2.69664,
+      "grad_norm": 2.0781149818054403,
+      "learning_rate": 5.4439878387057775e-06,
+      "loss": 1.1214,
+      "step": 16854
+    },
+    {
+      "epoch": 2.6968,
+      "grad_norm": 1.628294236102029,
+      "learning_rate": 5.438367523019505e-06,
+      "loss": 1.0631,
+      "step": 16855
+    },
+    {
+      "epoch": 2.69696,
+      "grad_norm": 1.8473011096989898,
+      "learning_rate": 5.432750028914579e-06,
+      "loss": 1.0657,
+      "step": 16856
+    },
+    {
+      "epoch": 2.69712,
+      "grad_norm": 1.4964272977073179,
+      "learning_rate": 5.427135356558566e-06,
+      "loss": 1.0796,
+      "step": 16857
+    },
+    {
+      "epoch": 2.69728,
+      "grad_norm": 1.8851868182571536,
+      "learning_rate": 5.421523506119064e-06,
+      "loss": 1.1556,
+      "step": 16858
+    },
+    {
+      "epoch": 2.69744,
+      "grad_norm": 1.8123638317551667,
+      "learning_rate": 5.415914477763451e-06,
+      "loss": 1.0453,
+      "step": 16859
+    },
+    {
+      "epoch": 2.6976,
+      "grad_norm": 1.6682925884443809,
+      "learning_rate": 5.410308271659137e-06,
+      "loss": 1.0585,
+      "step": 16860
+    },
+    {
+      "epoch": 2.69776,
+      "grad_norm": 2.0643298726323493,
+      "learning_rate": 5.4047048879734e-06,
+      "loss": 1.0872,
+      "step": 16861
+    },
+    {
+      "epoch": 2.69792,
+      "grad_norm": 1.9574780585708584,
+      "learning_rate": 5.399104326873428e-06,
+      "loss": 1.0851,
+      "step": 16862
+    },
+    {
+      "epoch": 2.69808,
+      "grad_norm": 1.7553196421339428,
+      "learning_rate": 5.393506588526332e-06,
+      "loss": 1.0141,
+      "step": 16863
+    },
+    {
+      "epoch": 2.69824,
+      "grad_norm": 2.194284532056391,
+      "learning_rate": 5.387911673099144e-06,
+      "loss": 1.0536,
+      "step": 16864
+    },
+    {
+      "epoch": 2.6984,
+      "grad_norm": 1.6732486057269023,
+      "learning_rate": 5.38231958075881e-06,
+      "loss": 1.0736,
+      "step": 16865
+    },
+    {
+      "epoch": 2.69856,
+      "grad_norm": 1.8991018383352622,
+      "learning_rate": 5.376730311672184e-06,
+      "loss": 1.0907,
+      "step": 16866
+    },
+    {
+      "epoch": 2.69872,
+      "grad_norm": 1.7069183079334895,
+      "learning_rate": 5.371143866006045e-06,
+      "loss": 1.0922,
+      "step": 16867
+    },
+    {
+      "epoch": 2.69888,
+      "grad_norm": 2.0201537821449924,
+      "learning_rate": 5.365560243927082e-06,
+      "loss": 1.081,
+      "step": 16868
+    },
+    {
+      "epoch": 2.69904,
+      "grad_norm": 1.8202750584768714,
+      "learning_rate": 5.3599794456018925e-06,
+      "loss": 1.0407,
+      "step": 16869
+    },
+    {
+      "epoch": 2.6992000000000003,
+      "grad_norm": 1.9313827089096423,
+      "learning_rate": 5.354401471197035e-06,
+      "loss": 1.0212,
+      "step": 16870
+    },
+    {
+      "epoch": 2.69936,
+      "grad_norm": 1.7908443195746944,
+      "learning_rate": 5.348826320878897e-06,
+      "loss": 1.0839,
+      "step": 16871
+    },
+    {
+      "epoch": 2.69952,
+      "grad_norm": 1.615757258087005,
+      "learning_rate": 5.343253994813901e-06,
+      "loss": 1.0327,
+      "step": 16872
+    },
+    {
+      "epoch": 2.69968,
+      "grad_norm": 1.746987213910968,
+      "learning_rate": 5.337684493168238e-06,
+      "loss": 1.0777,
+      "step": 16873
+    },
+    {
+      "epoch": 2.69984,
+      "grad_norm": 1.717795590080769,
+      "learning_rate": 5.332117816108162e-06,
+      "loss": 1.0997,
+      "step": 16874
+    },
+    {
+      "epoch": 2.7,
+      "grad_norm": 1.5788950060640883,
+      "learning_rate": 5.3265539637997075e-06,
+      "loss": 1.0514,
+      "step": 16875
+    },
+    {
+      "epoch": 2.70016,
+      "grad_norm": 1.6158518267852913,
+      "learning_rate": 5.320992936408953e-06,
+      "loss": 1.1097,
+      "step": 16876
+    },
+    {
+      "epoch": 2.70032,
+      "grad_norm": 1.7463853278700774,
+      "learning_rate": 5.31543473410181e-06,
+      "loss": 1.0496,
+      "step": 16877
+    },
+    {
+      "epoch": 2.7004799999999998,
+      "grad_norm": 1.8899453430835633,
+      "learning_rate": 5.309879357044123e-06,
+      "loss": 1.0642,
+      "step": 16878
+    },
+    {
+      "epoch": 2.70064,
+      "grad_norm": 1.6979301622637162,
+      "learning_rate": 5.304326805401649e-06,
+      "loss": 1.0359,
+      "step": 16879
+    },
+    {
+      "epoch": 2.7008,
+      "grad_norm": 1.7817269854397917,
+      "learning_rate": 5.298777079340089e-06,
+      "loss": 1.0735,
+      "step": 16880
+    },
+    {
+      "epoch": 2.7009600000000002,
+      "grad_norm": 1.5721637861250144,
+      "learning_rate": 5.293230179025021e-06,
+      "loss": 1.0935,
+      "step": 16881
+    },
+    {
+      "epoch": 2.70112,
+      "grad_norm": 1.9225157410167009,
+      "learning_rate": 5.287686104621958e-06,
+      "loss": 1.0759,
+      "step": 16882
+    },
+    {
+      "epoch": 2.70128,
+      "grad_norm": 1.738857237682843,
+      "learning_rate": 5.2821448562963335e-06,
+      "loss": 1.0762,
+      "step": 16883
+    },
+    {
+      "epoch": 2.70144,
+      "grad_norm": 1.9400326386901343,
+      "learning_rate": 5.276606434213494e-06,
+      "loss": 1.1689,
+      "step": 16884
+    },
+    {
+      "epoch": 2.7016,
+      "grad_norm": 1.7747960060668475,
+      "learning_rate": 5.271070838538672e-06,
+      "loss": 1.0981,
+      "step": 16885
+    },
+    {
+      "epoch": 2.70176,
+      "grad_norm": 1.545157018113336,
+      "learning_rate": 5.2655380694370925e-06,
+      "loss": 1.0423,
+      "step": 16886
+    },
+    {
+      "epoch": 2.70192,
+      "grad_norm": 1.893770923244631,
+      "learning_rate": 5.26000812707379e-06,
+      "loss": 1.1181,
+      "step": 16887
+    },
+    {
+      "epoch": 2.70208,
+      "grad_norm": 1.6462956426380986,
+      "learning_rate": 5.25448101161381e-06,
+      "loss": 1.0546,
+      "step": 16888
+    },
+    {
+      "epoch": 2.7022399999999998,
+      "grad_norm": 1.9661189095220657,
+      "learning_rate": 5.248956723222043e-06,
+      "loss": 1.1083,
+      "step": 16889
+    },
+    {
+      "epoch": 2.7024,
+      "grad_norm": 2.0673078807691367,
+      "learning_rate": 5.243435262063345e-06,
+      "loss": 1.0782,
+      "step": 16890
+    },
+    {
+      "epoch": 2.70256,
+      "grad_norm": 1.9112316772894484,
+      "learning_rate": 5.237916628302464e-06,
+      "loss": 1.1665,
+      "step": 16891
+    },
+    {
+      "epoch": 2.7027200000000002,
+      "grad_norm": 1.738453799603318,
+      "learning_rate": 5.232400822104077e-06,
+      "loss": 1.0303,
+      "step": 16892
+    },
+    {
+      "epoch": 2.70288,
+      "grad_norm": 1.8031548724666229,
+      "learning_rate": 5.226887843632755e-06,
+      "loss": 1.1119,
+      "step": 16893
+    },
+    {
+      "epoch": 2.70304,
+      "grad_norm": 1.8070242315073972,
+      "learning_rate": 5.221377693052998e-06,
+      "loss": 1.0406,
+      "step": 16894
+    },
+    {
+      "epoch": 2.7032,
+      "grad_norm": 1.8073967708976864,
+      "learning_rate": 5.215870370529219e-06,
+      "loss": 1.0993,
+      "step": 16895
+    },
+    {
+      "epoch": 2.70336,
+      "grad_norm": 1.6944136915553165,
+      "learning_rate": 5.210365876225764e-06,
+      "loss": 1.0692,
+      "step": 16896
+    },
+    {
+      "epoch": 2.70352,
+      "grad_norm": 1.5319420507134618,
+      "learning_rate": 5.204864210306859e-06,
+      "loss": 1.0633,
+      "step": 16897
+    },
+    {
+      "epoch": 2.70368,
+      "grad_norm": 2.0623392502391176,
+      "learning_rate": 5.199365372936682e-06,
+      "loss": 1.0698,
+      "step": 16898
+    },
+    {
+      "epoch": 2.70384,
+      "grad_norm": 2.030376549583921,
+      "learning_rate": 5.193869364279291e-06,
+      "loss": 1.0788,
+      "step": 16899
+    },
+    {
+      "epoch": 2.7039999999999997,
+      "grad_norm": 1.7785807289149231,
+      "learning_rate": 5.1883761844987e-06,
+      "loss": 1.0658,
+      "step": 16900
+    },
+    {
+      "epoch": 2.70416,
+      "grad_norm": 1.7618923351067828,
+      "learning_rate": 5.182885833758799e-06,
+      "loss": 1.0728,
+      "step": 16901
+    },
+    {
+      "epoch": 2.70432,
+      "grad_norm": 1.5913001942006881,
+      "learning_rate": 5.177398312223425e-06,
+      "loss": 1.0476,
+      "step": 16902
+    },
+    {
+      "epoch": 2.70448,
+      "grad_norm": 1.822941163072659,
+      "learning_rate": 5.171913620056324e-06,
+      "loss": 1.0576,
+      "step": 16903
+    },
+    {
+      "epoch": 2.70464,
+      "grad_norm": 1.6791120074762313,
+      "learning_rate": 5.1664317574211415e-06,
+      "loss": 1.0362,
+      "step": 16904
+    },
+    {
+      "epoch": 2.7048,
+      "grad_norm": 1.6582846736098988,
+      "learning_rate": 5.1609527244814385e-06,
+      "loss": 1.0605,
+      "step": 16905
+    },
+    {
+      "epoch": 2.70496,
+      "grad_norm": 1.7328304410053021,
+      "learning_rate": 5.155476521400715e-06,
+      "loss": 1.1307,
+      "step": 16906
+    },
+    {
+      "epoch": 2.70512,
+      "grad_norm": 1.881979621058653,
+      "learning_rate": 5.150003148342375e-06,
+      "loss": 1.1267,
+      "step": 16907
+    },
+    {
+      "epoch": 2.70528,
+      "grad_norm": 1.8236798591641965,
+      "learning_rate": 5.1445326054697205e-06,
+      "loss": 1.0945,
+      "step": 16908
+    },
+    {
+      "epoch": 2.70544,
+      "grad_norm": 1.6339299052242695,
+      "learning_rate": 5.139064892945999e-06,
+      "loss": 1.083,
+      "step": 16909
+    },
+    {
+      "epoch": 2.7056,
+      "grad_norm": 1.4820869696686805,
+      "learning_rate": 5.133600010934348e-06,
+      "loss": 1.0546,
+      "step": 16910
+    },
+    {
+      "epoch": 2.70576,
+      "grad_norm": 1.7400315592304894,
+      "learning_rate": 5.128137959597856e-06,
+      "loss": 1.1078,
+      "step": 16911
+    },
+    {
+      "epoch": 2.70592,
+      "grad_norm": 1.822985066522462,
+      "learning_rate": 5.122678739099451e-06,
+      "loss": 1.0418,
+      "step": 16912
+    },
+    {
+      "epoch": 2.70608,
+      "grad_norm": 1.7328903787315462,
+      "learning_rate": 5.117222349602102e-06,
+      "loss": 1.0791,
+      "step": 16913
+    },
+    {
+      "epoch": 2.70624,
+      "grad_norm": 2.10154998857529,
+      "learning_rate": 5.111768791268545e-06,
+      "loss": 1.0654,
+      "step": 16914
+    },
+    {
+      "epoch": 2.7064,
+      "grad_norm": 1.6270099535525038,
+      "learning_rate": 5.106318064261572e-06,
+      "loss": 1.0715,
+      "step": 16915
+    },
+    {
+      "epoch": 2.70656,
+      "grad_norm": 1.9806485125584097,
+      "learning_rate": 5.100870168743765e-06,
+      "loss": 1.0854,
+      "step": 16916
+    },
+    {
+      "epoch": 2.70672,
+      "grad_norm": 1.6717407498224128,
+      "learning_rate": 5.095425104877716e-06,
+      "loss": 1.0795,
+      "step": 16917
+    },
+    {
+      "epoch": 2.70688,
+      "grad_norm": 1.755625560507427,
+      "learning_rate": 5.089982872825894e-06,
+      "loss": 1.0505,
+      "step": 16918
+    },
+    {
+      "epoch": 2.70704,
+      "grad_norm": 1.7129885812178838,
+      "learning_rate": 5.084543472750691e-06,
+      "loss": 1.0813,
+      "step": 16919
+    },
+    {
+      "epoch": 2.7072000000000003,
+      "grad_norm": 1.8347447347433297,
+      "learning_rate": 5.079106904814402e-06,
+      "loss": 1.1004,
+      "step": 16920
+    },
+    {
+      "epoch": 2.70736,
+      "grad_norm": 1.8092875186783839,
+      "learning_rate": 5.07367316917925e-06,
+      "loss": 1.0795,
+      "step": 16921
+    },
+    {
+      "epoch": 2.70752,
+      "grad_norm": 1.7702600569823599,
+      "learning_rate": 5.068242266007361e-06,
+      "loss": 1.0993,
+      "step": 16922
+    },
+    {
+      "epoch": 2.70768,
+      "grad_norm": 1.873116242307311,
+      "learning_rate": 5.062814195460797e-06,
+      "loss": 1.0818,
+      "step": 16923
+    },
+    {
+      "epoch": 2.70784,
+      "grad_norm": 1.6190061396345927,
+      "learning_rate": 5.057388957701525e-06,
+      "loss": 1.0171,
+      "step": 16924
+    },
+    {
+      "epoch": 2.708,
+      "grad_norm": 1.7037580456844734,
+      "learning_rate": 5.051966552891407e-06,
+      "loss": 1.0584,
+      "step": 16925
+    },
+    {
+      "epoch": 2.70816,
+      "grad_norm": 1.8918520615776737,
+      "learning_rate": 5.046546981192246e-06,
+      "loss": 1.1333,
+      "step": 16926
+    },
+    {
+      "epoch": 2.70832,
+      "grad_norm": 1.8864958088083426,
+      "learning_rate": 5.04113024276579e-06,
+      "loss": 1.0939,
+      "step": 16927
+    },
+    {
+      "epoch": 2.7084799999999998,
+      "grad_norm": 1.7930289244282083,
+      "learning_rate": 5.035716337773611e-06,
+      "loss": 1.0416,
+      "step": 16928
+    },
+    {
+      "epoch": 2.70864,
+      "grad_norm": 1.629762945740843,
+      "learning_rate": 5.030305266377311e-06,
+      "loss": 1.0392,
+      "step": 16929
+    },
+    {
+      "epoch": 2.7088,
+      "grad_norm": 1.691159667183272,
+      "learning_rate": 5.024897028738285e-06,
+      "loss": 1.0619,
+      "step": 16930
+    },
+    {
+      "epoch": 2.7089600000000003,
+      "grad_norm": 1.9314227329791571,
+      "learning_rate": 5.019491625017958e-06,
+      "loss": 1.0849,
+      "step": 16931
+    },
+    {
+      "epoch": 2.70912,
+      "grad_norm": 1.871241374623343,
+      "learning_rate": 5.01408905537758e-06,
+      "loss": 1.0818,
+      "step": 16932
+    },
+    {
+      "epoch": 2.70928,
+      "grad_norm": 1.6537958684305472,
+      "learning_rate": 5.008689319978388e-06,
+      "loss": 1.0475,
+      "step": 16933
+    },
+    {
+      "epoch": 2.70944,
+      "grad_norm": 1.7075263836781862,
+      "learning_rate": 5.003292418981498e-06,
+      "loss": 1.0733,
+      "step": 16934
+    },
+    {
+      "epoch": 2.7096,
+      "grad_norm": 1.5315163206135949,
+      "learning_rate": 4.997898352547925e-06,
+      "loss": 1.0606,
+      "step": 16935
+    },
+    {
+      "epoch": 2.70976,
+      "grad_norm": 1.6838430517899357,
+      "learning_rate": 4.992507120838641e-06,
+      "loss": 1.0596,
+      "step": 16936
+    },
+    {
+      "epoch": 2.70992,
+      "grad_norm": 1.660731447365455,
+      "learning_rate": 4.987118724014494e-06,
+      "loss": 1.0041,
+      "step": 16937
+    },
+    {
+      "epoch": 2.71008,
+      "grad_norm": 1.7089431092780845,
+      "learning_rate": 4.98173316223628e-06,
+      "loss": 1.0754,
+      "step": 16938
+    },
+    {
+      "epoch": 2.7102399999999998,
+      "grad_norm": 1.851081965580602,
+      "learning_rate": 4.976350435664701e-06,
+      "loss": 1.0346,
+      "step": 16939
+    },
+    {
+      "epoch": 2.7104,
+      "grad_norm": 1.599074401692495,
+      "learning_rate": 4.970970544460352e-06,
+      "loss": 1.0667,
+      "step": 16940
+    },
+    {
+      "epoch": 2.71056,
+      "grad_norm": 1.5360903008503257,
+      "learning_rate": 4.965593488783771e-06,
+      "loss": 1.0779,
+      "step": 16941
+    },
+    {
+      "epoch": 2.7107200000000002,
+      "grad_norm": 1.8954034062777534,
+      "learning_rate": 4.9602192687953855e-06,
+      "loss": 1.0742,
+      "step": 16942
+    },
+    {
+      "epoch": 2.71088,
+      "grad_norm": 1.7475868790809699,
+      "learning_rate": 4.9548478846556e-06,
+      "loss": 1.0375,
+      "step": 16943
+    },
+    {
+      "epoch": 2.71104,
+      "grad_norm": 1.6790947397408442,
+      "learning_rate": 4.949479336524632e-06,
+      "loss": 1.0437,
+      "step": 16944
+    },
+    {
+      "epoch": 2.7112,
+      "grad_norm": 1.523528934936308,
+      "learning_rate": 4.944113624562719e-06,
+      "loss": 1.0884,
+      "step": 16945
+    },
+    {
+      "epoch": 2.71136,
+      "grad_norm": 1.565318618252114,
+      "learning_rate": 4.938750748929922e-06,
+      "loss": 1.0412,
+      "step": 16946
+    },
+    {
+      "epoch": 2.71152,
+      "grad_norm": 1.57539270155636,
+      "learning_rate": 4.933390709786301e-06,
+      "loss": 1.037,
+      "step": 16947
+    },
+    {
+      "epoch": 2.71168,
+      "grad_norm": 1.698604929567853,
+      "learning_rate": 4.928033507291763e-06,
+      "loss": 1.0892,
+      "step": 16948
+    },
+    {
+      "epoch": 2.71184,
+      "grad_norm": 1.6704095574767197,
+      "learning_rate": 4.922679141606179e-06,
+      "loss": 1.0293,
+      "step": 16949
+    },
+    {
+      "epoch": 2.7119999999999997,
+      "grad_norm": 1.7726193382054258,
+      "learning_rate": 4.9173276128893e-06,
+      "loss": 1.11,
+      "step": 16950
+    },
+    {
+      "epoch": 2.71216,
+      "grad_norm": 1.6531782840482996,
+      "learning_rate": 4.911978921300819e-06,
+      "loss": 1.0707,
+      "step": 16951
+    },
+    {
+      "epoch": 2.71232,
+      "grad_norm": 1.9117915992542651,
+      "learning_rate": 4.906633067000332e-06,
+      "loss": 1.054,
+      "step": 16952
+    },
+    {
+      "epoch": 2.7124800000000002,
+      "grad_norm": 1.7702971330892876,
+      "learning_rate": 4.901290050147345e-06,
+      "loss": 1.0478,
+      "step": 16953
+    },
+    {
+      "epoch": 2.71264,
+      "grad_norm": 1.8347083967291626,
+      "learning_rate": 4.895949870901295e-06,
+      "loss": 1.0284,
+      "step": 16954
+    },
+    {
+      "epoch": 2.7128,
+      "grad_norm": 1.9232331284331035,
+      "learning_rate": 4.890612529421512e-06,
+      "loss": 1.0483,
+      "step": 16955
+    },
+    {
+      "epoch": 2.71296,
+      "grad_norm": 1.6179276751462282,
+      "learning_rate": 4.885278025867268e-06,
+      "loss": 1.0441,
+      "step": 16956
+    },
+    {
+      "epoch": 2.71312,
+      "grad_norm": 1.7790059638718538,
+      "learning_rate": 4.879946360397724e-06,
+      "loss": 1.0711,
+      "step": 16957
+    },
+    {
+      "epoch": 2.71328,
+      "grad_norm": 1.6961962704741402,
+      "learning_rate": 4.874617533171965e-06,
+      "loss": 1.0589,
+      "step": 16958
+    },
+    {
+      "epoch": 2.71344,
+      "grad_norm": 1.9354177502365533,
+      "learning_rate": 4.869291544349042e-06,
+      "loss": 1.0741,
+      "step": 16959
+    },
+    {
+      "epoch": 2.7136,
+      "grad_norm": 1.6574578189589646,
+      "learning_rate": 4.863968394087803e-06,
+      "loss": 1.1276,
+      "step": 16960
+    },
+    {
+      "epoch": 2.7137599999999997,
+      "grad_norm": 1.7277462869261384,
+      "learning_rate": 4.858648082547135e-06,
+      "loss": 1.0252,
+      "step": 16961
+    },
+    {
+      "epoch": 2.71392,
+      "grad_norm": 1.7799778060680416,
+      "learning_rate": 4.8533306098857755e-06,
+      "loss": 1.0424,
+      "step": 16962
+    },
+    {
+      "epoch": 2.71408,
+      "grad_norm": 1.9098057531443025,
+      "learning_rate": 4.8480159762623874e-06,
+      "loss": 1.0687,
+      "step": 16963
+    },
+    {
+      "epoch": 2.71424,
+      "grad_norm": 1.9079036017686886,
+      "learning_rate": 4.842704181835545e-06,
+      "loss": 1.034,
+      "step": 16964
+    },
+    {
+      "epoch": 2.7144,
+      "grad_norm": 1.8441337492101193,
+      "learning_rate": 4.8373952267637526e-06,
+      "loss": 1.0948,
+      "step": 16965
+    },
+    {
+      "epoch": 2.71456,
+      "grad_norm": 1.919624883687896,
+      "learning_rate": 4.832089111205429e-06,
+      "loss": 1.0678,
+      "step": 16966
+    },
+    {
+      "epoch": 2.71472,
+      "grad_norm": 1.827524009069975,
+      "learning_rate": 4.82678583531887e-06,
+      "loss": 1.0731,
+      "step": 16967
+    },
+    {
+      "epoch": 2.71488,
+      "grad_norm": 1.903435647582755,
+      "learning_rate": 4.8214853992623706e-06,
+      "loss": 1.0769,
+      "step": 16968
+    },
+    {
+      "epoch": 2.71504,
+      "grad_norm": 1.7904856906473792,
+      "learning_rate": 4.816187803194028e-06,
+      "loss": 1.0685,
+      "step": 16969
+    },
+    {
+      "epoch": 2.7152,
+      "grad_norm": 1.4183982635255694,
+      "learning_rate": 4.810893047271969e-06,
+      "loss": 1.0629,
+      "step": 16970
+    },
+    {
+      "epoch": 2.71536,
+      "grad_norm": 1.731577538548586,
+      "learning_rate": 4.805601131654136e-06,
+      "loss": 1.0486,
+      "step": 16971
+    },
+    {
+      "epoch": 2.71552,
+      "grad_norm": 1.7976791609193934,
+      "learning_rate": 4.800312056498468e-06,
+      "loss": 1.0801,
+      "step": 16972
+    },
+    {
+      "epoch": 2.71568,
+      "grad_norm": 2.195651581334908,
+      "learning_rate": 4.7950258219627506e-06,
+      "loss": 1.0503,
+      "step": 16973
+    },
+    {
+      "epoch": 2.71584,
+      "grad_norm": 1.8098582819327012,
+      "learning_rate": 4.789742428204747e-06,
+      "loss": 1.102,
+      "step": 16974
+    },
+    {
+      "epoch": 2.716,
+      "grad_norm": 1.819258306512835,
+      "learning_rate": 4.784461875382085e-06,
+      "loss": 1.0616,
+      "step": 16975
+    },
+    {
+      "epoch": 2.71616,
+      "grad_norm": 1.7084908644735388,
+      "learning_rate": 4.779184163652339e-06,
+      "loss": 1.0462,
+      "step": 16976
+    },
+    {
+      "epoch": 2.71632,
+      "grad_norm": 1.6419308015464908,
+      "learning_rate": 4.773909293172995e-06,
+      "loss": 0.9993,
+      "step": 16977
+    },
+    {
+      "epoch": 2.71648,
+      "grad_norm": 1.8533211295580168,
+      "learning_rate": 4.768637264101439e-06,
+      "loss": 1.2121,
+      "step": 16978
+    },
+    {
+      "epoch": 2.71664,
+      "grad_norm": 1.7396659773574574,
+      "learning_rate": 4.763368076594965e-06,
+      "loss": 1.0397,
+      "step": 16979
+    },
+    {
+      "epoch": 2.7168,
+      "grad_norm": 1.6966332558131563,
+      "learning_rate": 4.758101730810827e-06,
+      "loss": 1.0665,
+      "step": 16980
+    },
+    {
+      "epoch": 2.7169600000000003,
+      "grad_norm": 1.7305215696225194,
+      "learning_rate": 4.752838226906153e-06,
+      "loss": 1.0871,
+      "step": 16981
+    },
+    {
+      "epoch": 2.71712,
+      "grad_norm": 1.7000451254421554,
+      "learning_rate": 4.747577565037986e-06,
+      "loss": 1.0195,
+      "step": 16982
+    },
+    {
+      "epoch": 2.71728,
+      "grad_norm": 1.5917948436778953,
+      "learning_rate": 4.7423197453632995e-06,
+      "loss": 1.0397,
+      "step": 16983
+    },
+    {
+      "epoch": 2.71744,
+      "grad_norm": 1.661764355211005,
+      "learning_rate": 4.737064768039013e-06,
+      "loss": 1.0659,
+      "step": 16984
+    },
+    {
+      "epoch": 2.7176,
+      "grad_norm": 1.6643612565744221,
+      "learning_rate": 4.7318126332218775e-06,
+      "loss": 1.0444,
+      "step": 16985
+    },
+    {
+      "epoch": 2.71776,
+      "grad_norm": 1.5245869493709006,
+      "learning_rate": 4.726563341068646e-06,
+      "loss": 1.0496,
+      "step": 16986
+    },
+    {
+      "epoch": 2.71792,
+      "grad_norm": 1.9998168984606268,
+      "learning_rate": 4.721316891735927e-06,
+      "loss": 1.0725,
+      "step": 16987
+    },
+    {
+      "epoch": 2.71808,
+      "grad_norm": 1.6874080298638174,
+      "learning_rate": 4.716073285380296e-06,
+      "loss": 1.0451,
+      "step": 16988
+    },
+    {
+      "epoch": 2.7182399999999998,
+      "grad_norm": 1.6375100097380597,
+      "learning_rate": 4.7108325221581596e-06,
+      "loss": 1.0148,
+      "step": 16989
+    },
+    {
+      "epoch": 2.7184,
+      "grad_norm": 1.7616782296299005,
+      "learning_rate": 4.7055946022259604e-06,
+      "loss": 1.0824,
+      "step": 16990
+    },
+    {
+      "epoch": 2.71856,
+      "grad_norm": 1.9035574773620823,
+      "learning_rate": 4.700359525739939e-06,
+      "loss": 1.1006,
+      "step": 16991
+    },
+    {
+      "epoch": 2.7187200000000002,
+      "grad_norm": 1.7952511791175405,
+      "learning_rate": 4.695127292856338e-06,
+      "loss": 1.0366,
+      "step": 16992
+    },
+    {
+      "epoch": 2.71888,
+      "grad_norm": 2.0983533106066306,
+      "learning_rate": 4.689897903731266e-06,
+      "loss": 1.0472,
+      "step": 16993
+    },
+    {
+      "epoch": 2.71904,
+      "grad_norm": 1.9505097405221108,
+      "learning_rate": 4.684671358520753e-06,
+      "loss": 1.1031,
+      "step": 16994
+    },
+    {
+      "epoch": 2.7192,
+      "grad_norm": 1.7072396546199131,
+      "learning_rate": 4.679447657380753e-06,
+      "loss": 1.066,
+      "step": 16995
+    },
+    {
+      "epoch": 2.71936,
+      "grad_norm": 1.544540669509309,
+      "learning_rate": 4.674226800467141e-06,
+      "loss": 1.0432,
+      "step": 16996
+    },
+    {
+      "epoch": 2.71952,
+      "grad_norm": 1.7105750405695392,
+      "learning_rate": 4.669008787935702e-06,
+      "loss": 1.0539,
+      "step": 16997
+    },
+    {
+      "epoch": 2.71968,
+      "grad_norm": 1.6820984531308385,
+      "learning_rate": 4.663793619942125e-06,
+      "loss": 1.0476,
+      "step": 16998
+    },
+    {
+      "epoch": 2.71984,
+      "grad_norm": 1.7595527046586346,
+      "learning_rate": 4.658581296642018e-06,
+      "loss": 1.0627,
+      "step": 16999
+    },
+    {
+      "epoch": 2.7199999999999998,
+      "grad_norm": 1.6394858760731685,
+      "learning_rate": 4.6533718181909435e-06,
+      "loss": 1.0475,
+      "step": 17000
+    },
+    {
+      "epoch": 2.72016,
+      "grad_norm": 1.8956080137579219,
+      "learning_rate": 4.648165184744291e-06,
+      "loss": 1.0964,
+      "step": 17001
+    },
+    {
+      "epoch": 2.72032,
+      "grad_norm": 1.7037597907008704,
+      "learning_rate": 4.642961396457468e-06,
+      "loss": 1.0676,
+      "step": 17002
+    },
+    {
+      "epoch": 2.7204800000000002,
+      "grad_norm": 1.707836165420738,
+      "learning_rate": 4.637760453485707e-06,
+      "loss": 1.0561,
+      "step": 17003
+    },
+    {
+      "epoch": 2.72064,
+      "grad_norm": 1.8200674155545937,
+      "learning_rate": 4.632562355984249e-06,
+      "loss": 1.071,
+      "step": 17004
+    },
+    {
+      "epoch": 2.7208,
+      "grad_norm": 2.009479468483343,
+      "learning_rate": 4.627367104108127e-06,
+      "loss": 1.1225,
+      "step": 17005
+    },
+    {
+      "epoch": 2.72096,
+      "grad_norm": 1.810671867938889,
+      "learning_rate": 4.622174698012416e-06,
+      "loss": 1.0985,
+      "step": 17006
+    },
+    {
+      "epoch": 2.72112,
+      "grad_norm": 1.9079765077206943,
+      "learning_rate": 4.616985137852026e-06,
+      "loss": 1.0811,
+      "step": 17007
+    },
+    {
+      "epoch": 2.72128,
+      "grad_norm": 1.698991515400214,
+      "learning_rate": 4.611798423781821e-06,
+      "loss": 1.051,
+      "step": 17008
+    },
+    {
+      "epoch": 2.72144,
+      "grad_norm": 1.975347050415953,
+      "learning_rate": 4.6066145559565565e-06,
+      "loss": 1.0768,
+      "step": 17009
+    },
+    {
+      "epoch": 2.7216,
+      "grad_norm": 1.7317124887678408,
+      "learning_rate": 4.601433534530908e-06,
+      "loss": 1.0989,
+      "step": 17010
+    },
+    {
+      "epoch": 2.7217599999999997,
+      "grad_norm": 1.8926058348058166,
+      "learning_rate": 4.5962553596594625e-06,
+      "loss": 1.1131,
+      "step": 17011
+    },
+    {
+      "epoch": 2.72192,
+      "grad_norm": 1.7273541998885453,
+      "learning_rate": 4.591080031496753e-06,
+      "loss": 1.1249,
+      "step": 17012
+    },
+    {
+      "epoch": 2.72208,
+      "grad_norm": 1.7582862480431485,
+      "learning_rate": 4.5859075501971885e-06,
+      "loss": 1.055,
+      "step": 17013
+    },
+    {
+      "epoch": 2.72224,
+      "grad_norm": 1.6839716657540782,
+      "learning_rate": 4.580737915915101e-06,
+      "loss": 1.0535,
+      "step": 17014
+    },
+    {
+      "epoch": 2.7224,
+      "grad_norm": 1.777799843730147,
+      "learning_rate": 4.575571128804745e-06,
+      "loss": 1.0651,
+      "step": 17015
+    },
+    {
+      "epoch": 2.72256,
+      "grad_norm": 1.7509930501209212,
+      "learning_rate": 4.57040718902032e-06,
+      "loss": 1.0508,
+      "step": 17016
+    },
+    {
+      "epoch": 2.72272,
+      "grad_norm": 1.6306812494931668,
+      "learning_rate": 4.565246096715869e-06,
+      "loss": 1.0721,
+      "step": 17017
+    },
+    {
+      "epoch": 2.72288,
+      "grad_norm": 1.7260985018655561,
+      "learning_rate": 4.560087852045436e-06,
+      "loss": 1.0306,
+      "step": 17018
+    },
+    {
+      "epoch": 2.72304,
+      "grad_norm": 1.9585956616808946,
+      "learning_rate": 4.554932455162886e-06,
+      "loss": 1.1204,
+      "step": 17019
+    },
+    {
+      "epoch": 2.7232,
+      "grad_norm": 1.638196665218523,
+      "learning_rate": 4.549779906222085e-06,
+      "loss": 1.0688,
+      "step": 17020
+    },
+    {
+      "epoch": 2.72336,
+      "grad_norm": 1.6266789230139356,
+      "learning_rate": 4.544630205376765e-06,
+      "loss": 1.0597,
+      "step": 17021
+    },
+    {
+      "epoch": 2.72352,
+      "grad_norm": 1.8578035869891023,
+      "learning_rate": 4.5394833527805935e-06,
+      "loss": 1.1032,
+      "step": 17022
+    },
+    {
+      "epoch": 2.72368,
+      "grad_norm": 1.660132880090215,
+      "learning_rate": 4.534339348587124e-06,
+      "loss": 1.0757,
+      "step": 17023
+    },
+    {
+      "epoch": 2.72384,
+      "grad_norm": 1.612054450733661,
+      "learning_rate": 4.529198192949858e-06,
+      "loss": 1.059,
+      "step": 17024
+    },
+    {
+      "epoch": 2.724,
+      "grad_norm": 1.5361838355352522,
+      "learning_rate": 4.524059886022225e-06,
+      "loss": 1.0381,
+      "step": 17025
+    },
+    {
+      "epoch": 2.72416,
+      "grad_norm": 1.791549777825415,
+      "learning_rate": 4.5189244279574935e-06,
+      "loss": 1.0773,
+      "step": 17026
+    },
+    {
+      "epoch": 2.72432,
+      "grad_norm": 1.855843251661214,
+      "learning_rate": 4.513791818908963e-06,
+      "loss": 1.0498,
+      "step": 17027
+    },
+    {
+      "epoch": 2.72448,
+      "grad_norm": 1.7960431579370344,
+      "learning_rate": 4.508662059029712e-06,
+      "loss": 1.0647,
+      "step": 17028
+    },
+    {
+      "epoch": 2.72464,
+      "grad_norm": 2.038864140374273,
+      "learning_rate": 4.50353514847286e-06,
+      "loss": 1.0854,
+      "step": 17029
+    },
+    {
+      "epoch": 2.7248,
+      "grad_norm": 1.7974600691628357,
+      "learning_rate": 4.498411087391352e-06,
+      "loss": 1.0542,
+      "step": 17030
+    },
+    {
+      "epoch": 2.7249600000000003,
+      "grad_norm": 1.756098436130211,
+      "learning_rate": 4.4932898759381005e-06,
+      "loss": 1.0933,
+      "step": 17031
+    },
+    {
+      "epoch": 2.72512,
+      "grad_norm": 1.698010357122067,
+      "learning_rate": 4.488171514265915e-06,
+      "loss": 1.0411,
+      "step": 17032
+    },
+    {
+      "epoch": 2.72528,
+      "grad_norm": 1.5225765945055139,
+      "learning_rate": 4.483056002527519e-06,
+      "loss": 1.0323,
+      "step": 17033
+    },
+    {
+      "epoch": 2.72544,
+      "grad_norm": 1.8331974967776896,
+      "learning_rate": 4.477943340875546e-06,
+      "loss": 1.0713,
+      "step": 17034
+    },
+    {
+      "epoch": 2.7256,
+      "grad_norm": 1.8225677100599267,
+      "learning_rate": 4.47283352946255e-06,
+      "loss": 1.0372,
+      "step": 17035
+    },
+    {
+      "epoch": 2.72576,
+      "grad_norm": 1.9759568942238015,
+      "learning_rate": 4.4677265684410105e-06,
+      "loss": 1.0646,
+      "step": 17036
+    },
+    {
+      "epoch": 2.72592,
+      "grad_norm": 1.9744514209207602,
+      "learning_rate": 4.462622457963295e-06,
+      "loss": 1.0687,
+      "step": 17037
+    },
+    {
+      "epoch": 2.72608,
+      "grad_norm": 1.7645366689512574,
+      "learning_rate": 4.457521198181724e-06,
+      "loss": 1.1165,
+      "step": 17038
+    },
+    {
+      "epoch": 2.7262399999999998,
+      "grad_norm": 1.6297397480650992,
+      "learning_rate": 4.452422789248489e-06,
+      "loss": 1.0778,
+      "step": 17039
+    },
+    {
+      "epoch": 2.7264,
+      "grad_norm": 1.6289636904979738,
+      "learning_rate": 4.447327231315723e-06,
+      "loss": 1.0284,
+      "step": 17040
+    },
+    {
+      "epoch": 2.72656,
+      "grad_norm": 1.7491065100041903,
+      "learning_rate": 4.442234524535504e-06,
+      "loss": 1.0742,
+      "step": 17041
+    },
+    {
+      "epoch": 2.7267200000000003,
+      "grad_norm": 1.7031873856489415,
+      "learning_rate": 4.437144669059745e-06,
+      "loss": 1.0685,
+      "step": 17042
+    },
+    {
+      "epoch": 2.72688,
+      "grad_norm": 1.533591617603067,
+      "learning_rate": 4.432057665040356e-06,
+      "loss": 1.0334,
+      "step": 17043
+    },
+    {
+      "epoch": 2.72704,
+      "grad_norm": 1.527961146980031,
+      "learning_rate": 4.426973512629084e-06,
+      "loss": 1.0382,
+      "step": 17044
+    },
+    {
+      "epoch": 2.7272,
+      "grad_norm": 1.6454762071917444,
+      "learning_rate": 4.421892211977685e-06,
+      "loss": 1.0954,
+      "step": 17045
+    },
+    {
+      "epoch": 2.72736,
+      "grad_norm": 1.5854066194521614,
+      "learning_rate": 4.416813763237726e-06,
+      "loss": 1.0599,
+      "step": 17046
+    },
+    {
+      "epoch": 2.72752,
+      "grad_norm": 1.6871758490489588,
+      "learning_rate": 4.411738166560775e-06,
+      "loss": 1.0591,
+      "step": 17047
+    },
+    {
+      "epoch": 2.72768,
+      "grad_norm": 1.70068082747103,
+      "learning_rate": 4.406665422098278e-06,
+      "loss": 1.0326,
+      "step": 17048
+    },
+    {
+      "epoch": 2.72784,
+      "grad_norm": 1.8477986320359985,
+      "learning_rate": 4.401595530001579e-06,
+      "loss": 1.1053,
+      "step": 17049
+    },
+    {
+      "epoch": 2.7279999999999998,
+      "grad_norm": 1.7139761431825453,
+      "learning_rate": 4.396528490421981e-06,
+      "loss": 1.0549,
+      "step": 17050
+    },
+    {
+      "epoch": 2.72816,
+      "grad_norm": 1.688440353440603,
+      "learning_rate": 4.391464303510651e-06,
+      "loss": 1.0492,
+      "step": 17051
+    },
+    {
+      "epoch": 2.72832,
+      "grad_norm": 1.7105517456882229,
+      "learning_rate": 4.386402969418712e-06,
+      "loss": 1.0774,
+      "step": 17052
+    },
+    {
+      "epoch": 2.7284800000000002,
+      "grad_norm": 1.7938188654287108,
+      "learning_rate": 4.38134448829719e-06,
+      "loss": 1.0842,
+      "step": 17053
+    },
+    {
+      "epoch": 2.72864,
+      "grad_norm": 1.6567303894743475,
+      "learning_rate": 4.376288860297018e-06,
+      "loss": 1.0536,
+      "step": 17054
+    },
+    {
+      "epoch": 2.7288,
+      "grad_norm": 1.6309993677476469,
+      "learning_rate": 4.371236085569042e-06,
+      "loss": 1.0257,
+      "step": 17055
+    },
+    {
+      "epoch": 2.72896,
+      "grad_norm": 1.7445221682693157,
+      "learning_rate": 4.36618616426403e-06,
+      "loss": 1.0792,
+      "step": 17056
+    },
+    {
+      "epoch": 2.72912,
+      "grad_norm": 1.9171477045884007,
+      "learning_rate": 4.361139096532685e-06,
+      "loss": 1.0845,
+      "step": 17057
+    },
+    {
+      "epoch": 2.72928,
+      "grad_norm": 1.581452757896525,
+      "learning_rate": 4.356094882525574e-06,
+      "loss": 1.0406,
+      "step": 17058
+    },
+    {
+      "epoch": 2.72944,
+      "grad_norm": 1.695317140507946,
+      "learning_rate": 4.351053522393245e-06,
+      "loss": 1.0549,
+      "step": 17059
+    },
+    {
+      "epoch": 2.7296,
+      "grad_norm": 1.8383362196149693,
+      "learning_rate": 4.346015016286076e-06,
+      "loss": 1.0647,
+      "step": 17060
+    },
+    {
+      "epoch": 2.7297599999999997,
+      "grad_norm": 1.8605391791723898,
+      "learning_rate": 4.3409793643544694e-06,
+      "loss": 1.0871,
+      "step": 17061
+    },
+    {
+      "epoch": 2.72992,
+      "grad_norm": 2.102741037146089,
+      "learning_rate": 4.335946566748616e-06,
+      "loss": 1.0799,
+      "step": 17062
+    },
+    {
+      "epoch": 2.73008,
+      "grad_norm": 1.682640706686746,
+      "learning_rate": 4.330916623618731e-06,
+      "loss": 1.022,
+      "step": 17063
+    },
+    {
+      "epoch": 2.7302400000000002,
+      "grad_norm": 1.833369627983564,
+      "learning_rate": 4.325889535114891e-06,
+      "loss": 1.0878,
+      "step": 17064
+    },
+    {
+      "epoch": 2.7304,
+      "grad_norm": 1.8167051064643933,
+      "learning_rate": 4.320865301387101e-06,
+      "loss": 1.0338,
+      "step": 17065
+    },
+    {
+      "epoch": 2.73056,
+      "grad_norm": 1.7141684977294467,
+      "learning_rate": 4.315843922585261e-06,
+      "loss": 1.0658,
+      "step": 17066
+    },
+    {
+      "epoch": 2.73072,
+      "grad_norm": 1.8804268114578835,
+      "learning_rate": 4.3108253988592195e-06,
+      "loss": 1.0665,
+      "step": 17067
+    },
+    {
+      "epoch": 2.73088,
+      "grad_norm": 1.7580393858026961,
+      "learning_rate": 4.3058097303587115e-06,
+      "loss": 1.1152,
+      "step": 17068
+    },
+    {
+      "epoch": 2.73104,
+      "grad_norm": 1.5060797575808034,
+      "learning_rate": 4.300796917233396e-06,
+      "loss": 1.0966,
+      "step": 17069
+    },
+    {
+      "epoch": 2.7312,
+      "grad_norm": 1.703919065969119,
+      "learning_rate": 4.2957869596328635e-06,
+      "loss": 1.0453,
+      "step": 17070
+    },
+    {
+      "epoch": 2.73136,
+      "grad_norm": 1.7316118871358561,
+      "learning_rate": 4.290779857706584e-06,
+      "loss": 1.044,
+      "step": 17071
+    },
+    {
+      "epoch": 2.7315199999999997,
+      "grad_norm": 1.526871777214231,
+      "learning_rate": 4.285775611603959e-06,
+      "loss": 1.0651,
+      "step": 17072
+    },
+    {
+      "epoch": 2.73168,
+      "grad_norm": 1.93777679776247,
+      "learning_rate": 4.280774221474349e-06,
+      "loss": 1.0477,
+      "step": 17073
+    },
+    {
+      "epoch": 2.73184,
+      "grad_norm": 2.1364660646428257,
+      "learning_rate": 4.275775687466921e-06,
+      "loss": 1.0915,
+      "step": 17074
+    },
+    {
+      "epoch": 2.732,
+      "grad_norm": 1.5815123893853789,
+      "learning_rate": 4.270780009730902e-06,
+      "loss": 1.0598,
+      "step": 17075
+    },
+    {
+      "epoch": 2.73216,
+      "grad_norm": 1.9972289067337694,
+      "learning_rate": 4.265787188415282e-06,
+      "loss": 1.0965,
+      "step": 17076
+    },
+    {
+      "epoch": 2.73232,
+      "grad_norm": 1.8318836597137604,
+      "learning_rate": 4.2607972236691e-06,
+      "loss": 1.038,
+      "step": 17077
+    },
+    {
+      "epoch": 2.73248,
+      "grad_norm": 1.7611321896853909,
+      "learning_rate": 4.25581011564119e-06,
+      "loss": 1.0151,
+      "step": 17078
+    },
+    {
+      "epoch": 2.73264,
+      "grad_norm": 1.597226732924166,
+      "learning_rate": 4.250825864480412e-06,
+      "loss": 1.0994,
+      "step": 17079
+    },
+    {
+      "epoch": 2.7328,
+      "grad_norm": 2.0167853336617707,
+      "learning_rate": 4.24584447033547e-06,
+      "loss": 1.0788,
+      "step": 17080
+    },
+    {
+      "epoch": 2.73296,
+      "grad_norm": 2.096521610356161,
+      "learning_rate": 4.240865933354999e-06,
+      "loss": 1.0642,
+      "step": 17081
+    },
+    {
+      "epoch": 2.73312,
+      "grad_norm": 1.853568182927154,
+      "learning_rate": 4.2358902536875575e-06,
+      "loss": 1.0587,
+      "step": 17082
+    },
+    {
+      "epoch": 2.73328,
+      "grad_norm": 1.840702054461689,
+      "learning_rate": 4.230917431481585e-06,
+      "loss": 1.0886,
+      "step": 17083
+    },
+    {
+      "epoch": 2.73344,
+      "grad_norm": 1.873060219882373,
+      "learning_rate": 4.225947466885516e-06,
+      "loss": 1.0664,
+      "step": 17084
+    },
+    {
+      "epoch": 2.7336,
+      "grad_norm": 1.7561369988716973,
+      "learning_rate": 4.220980360047589e-06,
+      "loss": 1.008,
+      "step": 17085
+    },
+    {
+      "epoch": 2.73376,
+      "grad_norm": 1.8955535916231832,
+      "learning_rate": 4.216016111116072e-06,
+      "loss": 1.0734,
+      "step": 17086
+    },
+    {
+      "epoch": 2.73392,
+      "grad_norm": 1.7956759123470423,
+      "learning_rate": 4.211054720239038e-06,
+      "loss": 1.0847,
+      "step": 17087
+    },
+    {
+      "epoch": 2.73408,
+      "grad_norm": 1.7110163998377947,
+      "learning_rate": 4.206096187564568e-06,
+      "loss": 1.0581,
+      "step": 17088
+    },
+    {
+      "epoch": 2.73424,
+      "grad_norm": 1.761027795810302,
+      "learning_rate": 4.201140513240586e-06,
+      "loss": 1.0867,
+      "step": 17089
+    },
+    {
+      "epoch": 2.7344,
+      "grad_norm": 1.8874192882786216,
+      "learning_rate": 4.196187697414988e-06,
+      "loss": 1.0748,
+      "step": 17090
+    },
+    {
+      "epoch": 2.73456,
+      "grad_norm": 1.830011921219371,
+      "learning_rate": 4.191237740235554e-06,
+      "loss": 1.1182,
+      "step": 17091
+    },
+    {
+      "epoch": 2.7347200000000003,
+      "grad_norm": 1.7354757943169772,
+      "learning_rate": 4.186290641849966e-06,
+      "loss": 1.0812,
+      "step": 17092
+    },
+    {
+      "epoch": 2.73488,
+      "grad_norm": 2.019886870143781,
+      "learning_rate": 4.181346402405861e-06,
+      "loss": 1.0642,
+      "step": 17093
+    },
+    {
+      "epoch": 2.73504,
+      "grad_norm": 1.9853253547393577,
+      "learning_rate": 4.176405022050744e-06,
+      "loss": 1.1206,
+      "step": 17094
+    },
+    {
+      "epoch": 2.7352,
+      "grad_norm": 1.5372772191806552,
+      "learning_rate": 4.171466500932087e-06,
+      "loss": 1.0399,
+      "step": 17095
+    },
+    {
+      "epoch": 2.73536,
+      "grad_norm": 1.9269935816408759,
+      "learning_rate": 4.166530839197224e-06,
+      "loss": 1.1011,
+      "step": 17096
+    },
+    {
+      "epoch": 2.73552,
+      "grad_norm": 1.76931571190935,
+      "learning_rate": 4.161598036993419e-06,
+      "loss": 1.082,
+      "step": 17097
+    },
+    {
+      "epoch": 2.73568,
+      "grad_norm": 1.9510032603893595,
+      "learning_rate": 4.156668094467908e-06,
+      "loss": 1.1266,
+      "step": 17098
+    },
+    {
+      "epoch": 2.73584,
+      "grad_norm": 1.9599196432521915,
+      "learning_rate": 4.15174101176774e-06,
+      "loss": 1.1039,
+      "step": 17099
+    },
+    {
+      "epoch": 2.7359999999999998,
+      "grad_norm": 1.7073403369903903,
+      "learning_rate": 4.146816789039975e-06,
+      "loss": 1.0931,
+      "step": 17100
+    },
+    {
+      "epoch": 2.73616,
+      "grad_norm": 1.9948318402652632,
+      "learning_rate": 4.141895426431497e-06,
+      "loss": 1.0828,
+      "step": 17101
+    },
+    {
+      "epoch": 2.73632,
+      "grad_norm": 1.9451859977713273,
+      "learning_rate": 4.136976924089209e-06,
+      "loss": 1.059,
+      "step": 17102
+    },
+    {
+      "epoch": 2.7364800000000002,
+      "grad_norm": 1.9522599674780037,
+      "learning_rate": 4.132061282159816e-06,
+      "loss": 1.0659,
+      "step": 17103
+    },
+    {
+      "epoch": 2.73664,
+      "grad_norm": 1.7125062662065544,
+      "learning_rate": 4.127148500790034e-06,
+      "loss": 1.0687,
+      "step": 17104
+    },
+    {
+      "epoch": 2.7368,
+      "grad_norm": 1.7659392119648374,
+      "learning_rate": 4.1222385801264365e-06,
+      "loss": 1.1055,
+      "step": 17105
+    },
+    {
+      "epoch": 2.73696,
+      "grad_norm": 1.7137903713029718,
+      "learning_rate": 4.117331520315537e-06,
+      "loss": 1.1188,
+      "step": 17106
+    },
+    {
+      "epoch": 2.73712,
+      "grad_norm": 1.6179125688379148,
+      "learning_rate": 4.1124273215037425e-06,
+      "loss": 1.039,
+      "step": 17107
+    },
+    {
+      "epoch": 2.73728,
+      "grad_norm": 1.825731220484016,
+      "learning_rate": 4.107525983837401e-06,
+      "loss": 1.0992,
+      "step": 17108
+    },
+    {
+      "epoch": 2.73744,
+      "grad_norm": 1.682645868847165,
+      "learning_rate": 4.102627507462753e-06,
+      "loss": 1.0676,
+      "step": 17109
+    },
+    {
+      "epoch": 2.7376,
+      "grad_norm": 1.5111860659305034,
+      "learning_rate": 4.097731892525958e-06,
+      "loss": 1.0492,
+      "step": 17110
+    },
+    {
+      "epoch": 2.7377599999999997,
+      "grad_norm": 1.7767959108694102,
+      "learning_rate": 4.0928391391730994e-06,
+      "loss": 1.0496,
+      "step": 17111
+    },
+    {
+      "epoch": 2.73792,
+      "grad_norm": 1.8022411131743175,
+      "learning_rate": 4.0879492475501715e-06,
+      "loss": 1.0747,
+      "step": 17112
+    },
+    {
+      "epoch": 2.73808,
+      "grad_norm": 1.8625100166511286,
+      "learning_rate": 4.083062217803058e-06,
+      "loss": 1.0692,
+      "step": 17113
+    },
+    {
+      "epoch": 2.7382400000000002,
+      "grad_norm": 1.6180348702734098,
+      "learning_rate": 4.0781780500776416e-06,
+      "loss": 1.0806,
+      "step": 17114
+    },
+    {
+      "epoch": 2.7384,
+      "grad_norm": 2.0878553137166276,
+      "learning_rate": 4.073296744519583e-06,
+      "loss": 1.0946,
+      "step": 17115
+    },
+    {
+      "epoch": 2.73856,
+      "grad_norm": 1.711572237948366,
+      "learning_rate": 4.068418301274601e-06,
+      "loss": 1.0551,
+      "step": 17116
+    },
+    {
+      "epoch": 2.73872,
+      "grad_norm": 1.686149840791658,
+      "learning_rate": 4.0635427204882e-06,
+      "loss": 1.023,
+      "step": 17117
+    },
+    {
+      "epoch": 2.73888,
+      "grad_norm": 1.9144882367903715,
+      "learning_rate": 4.0586700023059195e-06,
+      "loss": 1.0689,
+      "step": 17118
+    },
+    {
+      "epoch": 2.73904,
+      "grad_norm": 1.7999244239526173,
+      "learning_rate": 4.053800146873099e-06,
+      "loss": 1.0844,
+      "step": 17119
+    },
+    {
+      "epoch": 2.7392,
+      "grad_norm": 1.6275967289357314,
+      "learning_rate": 4.048933154335088e-06,
+      "loss": 1.0677,
+      "step": 17120
+    },
+    {
+      "epoch": 2.73936,
+      "grad_norm": 1.6147249043490037,
+      "learning_rate": 4.044069024837094e-06,
+      "loss": 1.0757,
+      "step": 17121
+    },
+    {
+      "epoch": 2.7395199999999997,
+      "grad_norm": 1.8969796901143505,
+      "learning_rate": 4.039207758524255e-06,
+      "loss": 1.0555,
+      "step": 17122
+    },
+    {
+      "epoch": 2.73968,
+      "grad_norm": 1.7573073852883974,
+      "learning_rate": 4.034349355541634e-06,
+      "loss": 1.1249,
+      "step": 17123
+    },
+    {
+      "epoch": 2.73984,
+      "grad_norm": 1.926665897377508,
+      "learning_rate": 4.02949381603418e-06,
+      "loss": 1.0611,
+      "step": 17124
+    },
+    {
+      "epoch": 2.74,
+      "grad_norm": 1.5846348146031128,
+      "learning_rate": 4.024641140146801e-06,
+      "loss": 1.0681,
+      "step": 17125
+    },
+    {
+      "epoch": 2.74016,
+      "grad_norm": 1.709121927217864,
+      "learning_rate": 4.019791328024269e-06,
+      "loss": 1.1299,
+      "step": 17126
+    },
+    {
+      "epoch": 2.74032,
+      "grad_norm": 1.6250537764181183,
+      "learning_rate": 4.014944379811303e-06,
+      "loss": 1.0379,
+      "step": 17127
+    },
+    {
+      "epoch": 2.74048,
+      "grad_norm": 1.6296780585054824,
+      "learning_rate": 4.010100295652541e-06,
+      "loss": 1.089,
+      "step": 17128
+    },
+    {
+      "epoch": 2.74064,
+      "grad_norm": 1.7236408643510825,
+      "learning_rate": 4.005259075692502e-06,
+      "loss": 1.0744,
+      "step": 17129
+    },
+    {
+      "epoch": 2.7408,
+      "grad_norm": 1.6807942663312205,
+      "learning_rate": 4.00042072007567e-06,
+      "loss": 1.0968,
+      "step": 17130
+    },
+    {
+      "epoch": 2.74096,
+      "grad_norm": 1.6527872908346228,
+      "learning_rate": 3.995585228946374e-06,
+      "loss": 1.087,
+      "step": 17131
+    },
+    {
+      "epoch": 2.74112,
+      "grad_norm": 1.7661578770321404,
+      "learning_rate": 3.990752602448955e-06,
+      "loss": 1.0851,
+      "step": 17132
+    },
+    {
+      "epoch": 2.74128,
+      "grad_norm": 1.8522138380133772,
+      "learning_rate": 3.98592284072754e-06,
+      "loss": 1.0514,
+      "step": 17133
+    },
+    {
+      "epoch": 2.74144,
+      "grad_norm": 2.309425934190024,
+      "learning_rate": 3.981095943926316e-06,
+      "loss": 1.1048,
+      "step": 17134
+    },
+    {
+      "epoch": 2.7416,
+      "grad_norm": 1.6555732711758544,
+      "learning_rate": 3.976271912189244e-06,
+      "loss": 1.0488,
+      "step": 17135
+    },
+    {
+      "epoch": 2.74176,
+      "grad_norm": 1.880545568231261,
+      "learning_rate": 3.971450745660321e-06,
+      "loss": 1.0526,
+      "step": 17136
+    },
+    {
+      "epoch": 2.74192,
+      "grad_norm": 1.6867279991412683,
+      "learning_rate": 3.966632444483365e-06,
+      "loss": 1.0295,
+      "step": 17137
+    },
+    {
+      "epoch": 2.74208,
+      "grad_norm": 1.9587548404079134,
+      "learning_rate": 3.961817008802171e-06,
+      "loss": 1.0511,
+      "step": 17138
+    },
+    {
+      "epoch": 2.74224,
+      "grad_norm": 1.6211859295215871,
+      "learning_rate": 3.957004438760403e-06,
+      "loss": 1.0547,
+      "step": 17139
+    },
+    {
+      "epoch": 2.7424,
+      "grad_norm": 1.6020777964050945,
+      "learning_rate": 3.952194734501669e-06,
+      "loss": 1.0619,
+      "step": 17140
+    },
+    {
+      "epoch": 2.74256,
+      "grad_norm": 1.7705758544232642,
+      "learning_rate": 3.947387896169519e-06,
+      "loss": 1.0833,
+      "step": 17141
+    },
+    {
+      "epoch": 2.7427200000000003,
+      "grad_norm": 1.7695230047250496,
+      "learning_rate": 3.942583923907317e-06,
+      "loss": 1.0557,
+      "step": 17142
+    },
+    {
+      "epoch": 2.74288,
+      "grad_norm": 1.6881577140961674,
+      "learning_rate": 3.93778281785846e-06,
+      "loss": 1.085,
+      "step": 17143
+    },
+    {
+      "epoch": 2.74304,
+      "grad_norm": 1.6992951432776424,
+      "learning_rate": 3.932984578166177e-06,
+      "loss": 1.0715,
+      "step": 17144
+    },
+    {
+      "epoch": 2.7432,
+      "grad_norm": 1.8642776352672834,
+      "learning_rate": 3.9281892049736425e-06,
+      "loss": 1.0662,
+      "step": 17145
+    },
+    {
+      "epoch": 2.74336,
+      "grad_norm": 1.885439437526476,
+      "learning_rate": 3.923396698423965e-06,
+      "loss": 1.1397,
+      "step": 17146
+    },
+    {
+      "epoch": 2.74352,
+      "grad_norm": 1.8795973255849956,
+      "learning_rate": 3.918607058660128e-06,
+      "loss": 1.0253,
+      "step": 17147
+    },
+    {
+      "epoch": 2.74368,
+      "grad_norm": 1.5469376158767363,
+      "learning_rate": 3.913820285825054e-06,
+      "loss": 1.0295,
+      "step": 17148
+    },
+    {
+      "epoch": 2.74384,
+      "grad_norm": 1.7310963160851438,
+      "learning_rate": 3.90903638006157e-06,
+      "loss": 1.0812,
+      "step": 17149
+    },
+    {
+      "epoch": 2.7439999999999998,
+      "grad_norm": 1.5515741262018032,
+      "learning_rate": 3.904255341512419e-06,
+      "loss": 1.0128,
+      "step": 17150
+    },
+    {
+      "epoch": 2.74416,
+      "grad_norm": 1.6906869161440086,
+      "learning_rate": 3.8994771703202535e-06,
+      "loss": 1.0631,
+      "step": 17151
+    },
+    {
+      "epoch": 2.74432,
+      "grad_norm": 1.8531355598276213,
+      "learning_rate": 3.89470186662767e-06,
+      "loss": 1.0784,
+      "step": 17152
+    },
+    {
+      "epoch": 2.7444800000000003,
+      "grad_norm": 1.9367391821228104,
+      "learning_rate": 3.889929430577133e-06,
+      "loss": 1.0846,
+      "step": 17153
+    },
+    {
+      "epoch": 2.74464,
+      "grad_norm": 2.076165449940177,
+      "learning_rate": 3.885159862311061e-06,
+      "loss": 1.1137,
+      "step": 17154
+    },
+    {
+      "epoch": 2.7448,
+      "grad_norm": 1.7580760843910914,
+      "learning_rate": 3.880393161971763e-06,
+      "loss": 1.0422,
+      "step": 17155
+    },
+    {
+      "epoch": 2.74496,
+      "grad_norm": 1.6098423694564026,
+      "learning_rate": 3.875629329701458e-06,
+      "loss": 1.0337,
+      "step": 17156
+    },
+    {
+      "epoch": 2.74512,
+      "grad_norm": 1.8345954240981264,
+      "learning_rate": 3.870868365642333e-06,
+      "loss": 1.0381,
+      "step": 17157
+    },
+    {
+      "epoch": 2.74528,
+      "grad_norm": 1.806173722894291,
+      "learning_rate": 3.866110269936396e-06,
+      "loss": 1.0109,
+      "step": 17158
+    },
+    {
+      "epoch": 2.74544,
+      "grad_norm": 1.7598641993081436,
+      "learning_rate": 3.861355042725668e-06,
+      "loss": 1.0227,
+      "step": 17159
+    },
+    {
+      "epoch": 2.7456,
+      "grad_norm": 1.7208332500853079,
+      "learning_rate": 3.856602684151989e-06,
+      "loss": 1.0793,
+      "step": 17160
+    },
+    {
+      "epoch": 2.7457599999999998,
+      "grad_norm": 1.8593517759413203,
+      "learning_rate": 3.851853194357213e-06,
+      "loss": 1.0811,
+      "step": 17161
+    },
+    {
+      "epoch": 2.74592,
+      "grad_norm": 1.835912695248546,
+      "learning_rate": 3.847106573483028e-06,
+      "loss": 1.0952,
+      "step": 17162
+    },
+    {
+      "epoch": 2.74608,
+      "grad_norm": 2.0203238410215474,
+      "learning_rate": 3.842362821671075e-06,
+      "loss": 1.0944,
+      "step": 17163
+    },
+    {
+      "epoch": 2.7462400000000002,
+      "grad_norm": 1.8582829028176715,
+      "learning_rate": 3.837621939062897e-06,
+      "loss": 1.044,
+      "step": 17164
+    },
+    {
+      "epoch": 2.7464,
+      "grad_norm": 1.7190423038047478,
+      "learning_rate": 3.832883925799957e-06,
+      "loss": 1.0809,
+      "step": 17165
+    },
+    {
+      "epoch": 2.74656,
+      "grad_norm": 1.5363061587267821,
+      "learning_rate": 3.828148782023633e-06,
+      "loss": 1.0639,
+      "step": 17166
+    },
+    {
+      "epoch": 2.74672,
+      "grad_norm": 1.755835545561015,
+      "learning_rate": 3.823416507875211e-06,
+      "loss": 1.0569,
+      "step": 17167
+    },
+    {
+      "epoch": 2.74688,
+      "grad_norm": 1.9219463879491994,
+      "learning_rate": 3.8186871034958995e-06,
+      "loss": 1.0517,
+      "step": 17168
+    },
+    {
+      "epoch": 2.74704,
+      "grad_norm": 2.053162198354527,
+      "learning_rate": 3.813960569026809e-06,
+      "loss": 1.081,
+      "step": 17169
+    },
+    {
+      "epoch": 2.7472,
+      "grad_norm": 1.5861959523126998,
+      "learning_rate": 3.8092369046089704e-06,
+      "loss": 1.0456,
+      "step": 17170
+    },
+    {
+      "epoch": 2.74736,
+      "grad_norm": 1.7846281996639035,
+      "learning_rate": 3.8045161103833493e-06,
+      "loss": 1.1019,
+      "step": 17171
+    },
+    {
+      "epoch": 2.7475199999999997,
+      "grad_norm": 1.8008593499225694,
+      "learning_rate": 3.7997981864907883e-06,
+      "loss": 1.0633,
+      "step": 17172
+    },
+    {
+      "epoch": 2.74768,
+      "grad_norm": 1.7466755729652972,
+      "learning_rate": 3.795083133072086e-06,
+      "loss": 1.0995,
+      "step": 17173
+    },
+    {
+      "epoch": 2.74784,
+      "grad_norm": 1.7388921242003827,
+      "learning_rate": 3.7903709502678962e-06,
+      "loss": 1.0426,
+      "step": 17174
+    },
+    {
+      "epoch": 2.748,
+      "grad_norm": 1.639649188947903,
+      "learning_rate": 3.7856616382188627e-06,
+      "loss": 1.0541,
+      "step": 17175
+    },
+    {
+      "epoch": 2.74816,
+      "grad_norm": 1.707356825401792,
+      "learning_rate": 3.780955197065472e-06,
+      "loss": 1.0766,
+      "step": 17176
+    },
+    {
+      "epoch": 2.74832,
+      "grad_norm": 1.6785200128528879,
+      "learning_rate": 3.7762516269481795e-06,
+      "loss": 1.0462,
+      "step": 17177
+    },
+    {
+      "epoch": 2.74848,
+      "grad_norm": 1.7359356366523828,
+      "learning_rate": 3.771550928007328e-06,
+      "loss": 1.0651,
+      "step": 17178
+    },
+    {
+      "epoch": 2.74864,
+      "grad_norm": 1.6607880425094317,
+      "learning_rate": 3.766853100383172e-06,
+      "loss": 1.0897,
+      "step": 17179
+    },
+    {
+      "epoch": 2.7488,
+      "grad_norm": 1.8329252416746291,
+      "learning_rate": 3.7621581442158993e-06,
+      "loss": 1.0324,
+      "step": 17180
+    },
+    {
+      "epoch": 2.74896,
+      "grad_norm": 1.8416265519482111,
+      "learning_rate": 3.7574660596455867e-06,
+      "loss": 1.0147,
+      "step": 17181
+    },
+    {
+      "epoch": 2.74912,
+      "grad_norm": 1.6619665836416468,
+      "learning_rate": 3.752776846812256e-06,
+      "loss": 1.0937,
+      "step": 17182
+    },
+    {
+      "epoch": 2.7492799999999997,
+      "grad_norm": 1.7778471896443198,
+      "learning_rate": 3.7480905058558056e-06,
+      "loss": 1.073,
+      "step": 17183
+    },
+    {
+      "epoch": 2.74944,
+      "grad_norm": 1.6739659586217572,
+      "learning_rate": 3.7434070369160913e-06,
+      "loss": 1.1076,
+      "step": 17184
+    },
+    {
+      "epoch": 2.7496,
+      "grad_norm": 1.6072106934589272,
+      "learning_rate": 3.738726440132845e-06,
+      "loss": 1.0509,
+      "step": 17185
+    },
+    {
+      "epoch": 2.74976,
+      "grad_norm": 2.0800334339086826,
+      "learning_rate": 3.7340487156457326e-06,
+      "loss": 1.0947,
+      "step": 17186
+    },
+    {
+      "epoch": 2.74992,
+      "grad_norm": 1.8743859886116934,
+      "learning_rate": 3.729373863594343e-06,
+      "loss": 1.0615,
+      "step": 17187
+    },
+    {
+      "epoch": 2.75008,
+      "grad_norm": 1.7175683944589377,
+      "learning_rate": 3.724701884118131e-06,
+      "loss": 1.0854,
+      "step": 17188
+    },
+    {
+      "epoch": 2.75024,
+      "grad_norm": 1.636902433988632,
+      "learning_rate": 3.7200327773565523e-06,
+      "loss": 1.0785,
+      "step": 17189
+    },
+    {
+      "epoch": 2.7504,
+      "grad_norm": 1.7084874426905352,
+      "learning_rate": 3.7153665434488837e-06,
+      "loss": 1.0417,
+      "step": 17190
+    },
+    {
+      "epoch": 2.75056,
+      "grad_norm": 1.6946981181317504,
+      "learning_rate": 3.710703182534403e-06,
+      "loss": 1.0599,
+      "step": 17191
+    },
+    {
+      "epoch": 2.7507200000000003,
+      "grad_norm": 1.8541803588639219,
+      "learning_rate": 3.7060426947521985e-06,
+      "loss": 1.0731,
+      "step": 17192
+    },
+    {
+      "epoch": 2.75088,
+      "grad_norm": 1.8337113055470964,
+      "learning_rate": 3.701385080241382e-06,
+      "loss": 1.1132,
+      "step": 17193
+    },
+    {
+      "epoch": 2.75104,
+      "grad_norm": 2.0650105634845923,
+      "learning_rate": 3.696730339140908e-06,
+      "loss": 1.0527,
+      "step": 17194
+    },
+    {
+      "epoch": 2.7512,
+      "grad_norm": 1.8215580449907984,
+      "learning_rate": 3.6920784715896774e-06,
+      "loss": 1.1,
+      "step": 17195
+    },
+    {
+      "epoch": 2.75136,
+      "grad_norm": 1.9312837126698623,
+      "learning_rate": 3.6874294777264783e-06,
+      "loss": 1.1053,
+      "step": 17196
+    },
+    {
+      "epoch": 2.75152,
+      "grad_norm": 1.7034136245402451,
+      "learning_rate": 3.682783357690045e-06,
+      "loss": 1.0653,
+      "step": 17197
+    },
+    {
+      "epoch": 2.75168,
+      "grad_norm": 1.8669877656868719,
+      "learning_rate": 3.6781401116190217e-06,
+      "loss": 1.0785,
+      "step": 17198
+    },
+    {
+      "epoch": 2.75184,
+      "grad_norm": 1.5547353917909899,
+      "learning_rate": 3.67349973965192e-06,
+      "loss": 1.0527,
+      "step": 17199
+    },
+    {
+      "epoch": 2.752,
+      "grad_norm": 1.6642932384268772,
+      "learning_rate": 3.668862241927251e-06,
+      "loss": 1.0445,
+      "step": 17200
+    },
+    {
+      "epoch": 2.75216,
+      "grad_norm": 1.692363112671913,
+      "learning_rate": 3.664227618583327e-06,
+      "loss": 1.0635,
+      "step": 17201
+    },
+    {
+      "epoch": 2.75232,
+      "grad_norm": 1.7891292896294682,
+      "learning_rate": 3.659595869758503e-06,
+      "loss": 1.0904,
+      "step": 17202
+    },
+    {
+      "epoch": 2.7524800000000003,
+      "grad_norm": 1.7849160531379311,
+      "learning_rate": 3.6549669955909473e-06,
+      "loss": 1.1077,
+      "step": 17203
+    },
+    {
+      "epoch": 2.75264,
+      "grad_norm": 1.9723701408805367,
+      "learning_rate": 3.650340996218782e-06,
+      "loss": 1.023,
+      "step": 17204
+    },
+    {
+      "epoch": 2.7528,
+      "grad_norm": 1.8187963184369689,
+      "learning_rate": 3.6457178717800524e-06,
+      "loss": 1.0568,
+      "step": 17205
+    },
+    {
+      "epoch": 2.75296,
+      "grad_norm": 1.748967940759884,
+      "learning_rate": 3.6410976224127035e-06,
+      "loss": 1.0572,
+      "step": 17206
+    },
+    {
+      "epoch": 2.75312,
+      "grad_norm": 1.7733200364062207,
+      "learning_rate": 3.6364802482545814e-06,
+      "loss": 1.0794,
+      "step": 17207
+    },
+    {
+      "epoch": 2.75328,
+      "grad_norm": 1.6814392560011788,
+      "learning_rate": 3.6318657494434748e-06,
+      "loss": 1.0571,
+      "step": 17208
+    },
+    {
+      "epoch": 2.75344,
+      "grad_norm": 1.763249070440137,
+      "learning_rate": 3.6272541261170746e-06,
+      "loss": 1.0772,
+      "step": 17209
+    },
+    {
+      "epoch": 2.7536,
+      "grad_norm": 1.7400900788996838,
+      "learning_rate": 3.622645378412981e-06,
+      "loss": 1.0898,
+      "step": 17210
+    },
+    {
+      "epoch": 2.7537599999999998,
+      "grad_norm": 1.998490220024469,
+      "learning_rate": 3.618039506468707e-06,
+      "loss": 1.0762,
+      "step": 17211
+    },
+    {
+      "epoch": 2.75392,
+      "grad_norm": 1.7176585055818263,
+      "learning_rate": 3.613436510421697e-06,
+      "loss": 1.0887,
+      "step": 17212
+    },
+    {
+      "epoch": 2.75408,
+      "grad_norm": 1.6755932679249705,
+      "learning_rate": 3.608836390409276e-06,
+      "loss": 1.0562,
+      "step": 17213
+    },
+    {
+      "epoch": 2.7542400000000002,
+      "grad_norm": 1.661311290479207,
+      "learning_rate": 3.604239146568744e-06,
+      "loss": 1.06,
+      "step": 17214
+    },
+    {
+      "epoch": 2.7544,
+      "grad_norm": 1.7823603498690175,
+      "learning_rate": 3.5996447790372255e-06,
+      "loss": 1.0654,
+      "step": 17215
+    },
+    {
+      "epoch": 2.75456,
+      "grad_norm": 1.7049129740380375,
+      "learning_rate": 3.595053287951866e-06,
+      "loss": 1.0491,
+      "step": 17216
+    },
+    {
+      "epoch": 2.75472,
+      "grad_norm": 1.8317169715891217,
+      "learning_rate": 3.5904646734496117e-06,
+      "loss": 1.0393,
+      "step": 17217
+    },
+    {
+      "epoch": 2.75488,
+      "grad_norm": 1.9342636269419688,
+      "learning_rate": 3.5858789356674304e-06,
+      "loss": 1.0794,
+      "step": 17218
+    },
+    {
+      "epoch": 2.75504,
+      "grad_norm": 1.8229969214907455,
+      "learning_rate": 3.5812960747421244e-06,
+      "loss": 1.0652,
+      "step": 17219
+    },
+    {
+      "epoch": 2.7552,
+      "grad_norm": 1.8636625666617388,
+      "learning_rate": 3.576716090810439e-06,
+      "loss": 1.0691,
+      "step": 17220
+    },
+    {
+      "epoch": 2.75536,
+      "grad_norm": 1.678210239053867,
+      "learning_rate": 3.572138984009055e-06,
+      "loss": 1.06,
+      "step": 17221
+    },
+    {
+      "epoch": 2.7555199999999997,
+      "grad_norm": 1.800926850636659,
+      "learning_rate": 3.5675647544745284e-06,
+      "loss": 1.0402,
+      "step": 17222
+    },
+    {
+      "epoch": 2.75568,
+      "grad_norm": 2.0587486267515356,
+      "learning_rate": 3.5629934023433508e-06,
+      "loss": 1.0441,
+      "step": 17223
+    },
+    {
+      "epoch": 2.75584,
+      "grad_norm": 1.7902655737513622,
+      "learning_rate": 3.5584249277519243e-06,
+      "loss": 1.1025,
+      "step": 17224
+    },
+    {
+      "epoch": 2.7560000000000002,
+      "grad_norm": 1.9499435441222708,
+      "learning_rate": 3.553859330836573e-06,
+      "loss": 1.0561,
+      "step": 17225
+    },
+    {
+      "epoch": 2.75616,
+      "grad_norm": 1.5317224352247707,
+      "learning_rate": 3.5492966117335103e-06,
+      "loss": 1.043,
+      "step": 17226
+    },
+    {
+      "epoch": 2.75632,
+      "grad_norm": 1.971405133516192,
+      "learning_rate": 3.5447367705789046e-06,
+      "loss": 1.087,
+      "step": 17227
+    },
+    {
+      "epoch": 2.75648,
+      "grad_norm": 1.6792846180406575,
+      "learning_rate": 3.5401798075088035e-06,
+      "loss": 1.117,
+      "step": 17228
+    },
+    {
+      "epoch": 2.75664,
+      "grad_norm": 1.9101142770663506,
+      "learning_rate": 3.535625722659164e-06,
+      "loss": 1.1054,
+      "step": 17229
+    },
+    {
+      "epoch": 2.7568,
+      "grad_norm": 1.8638934145827484,
+      "learning_rate": 3.531074516165922e-06,
+      "loss": 1.0725,
+      "step": 17230
+    },
+    {
+      "epoch": 2.75696,
+      "grad_norm": 1.8030696663611439,
+      "learning_rate": 3.5265261881648137e-06,
+      "loss": 1.1082,
+      "step": 17231
+    },
+    {
+      "epoch": 2.75712,
+      "grad_norm": 1.746420010678839,
+      "learning_rate": 3.5219807387916083e-06,
+      "loss": 1.096,
+      "step": 17232
+    },
+    {
+      "epoch": 2.7572799999999997,
+      "grad_norm": 1.7991911062840797,
+      "learning_rate": 3.517438168181897e-06,
+      "loss": 1.0619,
+      "step": 17233
+    },
+    {
+      "epoch": 2.75744,
+      "grad_norm": 1.4446314812654641,
+      "learning_rate": 3.51289847647126e-06,
+      "loss": 1.0296,
+      "step": 17234
+    },
+    {
+      "epoch": 2.7576,
+      "grad_norm": 1.7027928747478023,
+      "learning_rate": 3.5083616637951235e-06,
+      "loss": 1.0771,
+      "step": 17235
+    },
+    {
+      "epoch": 2.75776,
+      "grad_norm": 1.639039213173833,
+      "learning_rate": 3.503827730288878e-06,
+      "loss": 1.0308,
+      "step": 17236
+    },
+    {
+      "epoch": 2.75792,
+      "grad_norm": 1.7471607899538009,
+      "learning_rate": 3.499296676087793e-06,
+      "loss": 1.0396,
+      "step": 17237
+    },
+    {
+      "epoch": 2.75808,
+      "grad_norm": 1.7485279425584348,
+      "learning_rate": 3.494768501327095e-06,
+      "loss": 1.0652,
+      "step": 17238
+    },
+    {
+      "epoch": 2.75824,
+      "grad_norm": 1.630752059570085,
+      "learning_rate": 3.4902432061418745e-06,
+      "loss": 1.1021,
+      "step": 17239
+    },
+    {
+      "epoch": 2.7584,
+      "grad_norm": 1.815229622042669,
+      "learning_rate": 3.485720790667157e-06,
+      "loss": 1.1118,
+      "step": 17240
+    },
+    {
+      "epoch": 2.75856,
+      "grad_norm": 1.4417722158170996,
+      "learning_rate": 3.4812012550379134e-06,
+      "loss": 1.0266,
+      "step": 17241
+    },
+    {
+      "epoch": 2.75872,
+      "grad_norm": 1.5406926607174907,
+      "learning_rate": 3.4766845993889684e-06,
+      "loss": 1.0496,
+      "step": 17242
+    },
+    {
+      "epoch": 2.75888,
+      "grad_norm": 1.7525463000944634,
+      "learning_rate": 3.472170823855114e-06,
+      "loss": 1.0518,
+      "step": 17243
+    },
+    {
+      "epoch": 2.75904,
+      "grad_norm": 1.5543470242326578,
+      "learning_rate": 3.467659928571021e-06,
+      "loss": 1.0639,
+      "step": 17244
+    },
+    {
+      "epoch": 2.7592,
+      "grad_norm": 1.7932738135384438,
+      "learning_rate": 3.463151913671281e-06,
+      "loss": 1.0722,
+      "step": 17245
+    },
+    {
+      "epoch": 2.75936,
+      "grad_norm": 1.7110351744596763,
+      "learning_rate": 3.458646779290442e-06,
+      "loss": 1.0416,
+      "step": 17246
+    },
+    {
+      "epoch": 2.75952,
+      "grad_norm": 1.6991982130141794,
+      "learning_rate": 3.4541445255628857e-06,
+      "loss": 1.0856,
+      "step": 17247
+    },
+    {
+      "epoch": 2.75968,
+      "grad_norm": 1.7236011730679377,
+      "learning_rate": 3.4496451526229934e-06,
+      "loss": 1.0517,
+      "step": 17248
+    },
+    {
+      "epoch": 2.75984,
+      "grad_norm": 1.9214040272777109,
+      "learning_rate": 3.445148660604991e-06,
+      "loss": 1.0515,
+      "step": 17249
+    },
+    {
+      "epoch": 2.76,
+      "grad_norm": 1.818826049549941,
+      "learning_rate": 3.44065504964306e-06,
+      "loss": 1.0341,
+      "step": 17250
+    },
+    {
+      "epoch": 2.76016,
+      "grad_norm": 1.805118403773661,
+      "learning_rate": 3.4361643198712713e-06,
+      "loss": 1.0682,
+      "step": 17251
+    },
+    {
+      "epoch": 2.76032,
+      "grad_norm": 1.5698925271252662,
+      "learning_rate": 3.4316764714236503e-06,
+      "loss": 1.0341,
+      "step": 17252
+    },
+    {
+      "epoch": 2.7604800000000003,
+      "grad_norm": 1.7455542853494326,
+      "learning_rate": 3.4271915044340685e-06,
+      "loss": 1.1073,
+      "step": 17253
+    },
+    {
+      "epoch": 2.76064,
+      "grad_norm": 1.8912505949013383,
+      "learning_rate": 3.422709419036374e-06,
+      "loss": 1.101,
+      "step": 17254
+    },
+    {
+      "epoch": 2.7608,
+      "grad_norm": 1.7343362657567123,
+      "learning_rate": 3.4182302153643264e-06,
+      "loss": 1.041,
+      "step": 17255
+    },
+    {
+      "epoch": 2.76096,
+      "grad_norm": 1.7244145346525694,
+      "learning_rate": 3.413753893551519e-06,
+      "loss": 1.0197,
+      "step": 17256
+    },
+    {
+      "epoch": 2.76112,
+      "grad_norm": 1.6425531506588822,
+      "learning_rate": 3.4092804537315894e-06,
+      "loss": 1.0499,
+      "step": 17257
+    },
+    {
+      "epoch": 2.76128,
+      "grad_norm": 1.8786598891064796,
+      "learning_rate": 3.4048098960379525e-06,
+      "loss": 1.0517,
+      "step": 17258
+    },
+    {
+      "epoch": 2.76144,
+      "grad_norm": 1.8253035045677968,
+      "learning_rate": 3.400342220604058e-06,
+      "loss": 1.065,
+      "step": 17259
+    },
+    {
+      "epoch": 2.7616,
+      "grad_norm": 1.6153624502317008,
+      "learning_rate": 3.3958774275631765e-06,
+      "loss": 1.0494,
+      "step": 17260
+    },
+    {
+      "epoch": 2.7617599999999998,
+      "grad_norm": 1.695786749215098,
+      "learning_rate": 3.391415517048546e-06,
+      "loss": 1.1167,
+      "step": 17261
+    },
+    {
+      "epoch": 2.76192,
+      "grad_norm": 1.7886131702988632,
+      "learning_rate": 3.3869564891933047e-06,
+      "loss": 1.0428,
+      "step": 17262
+    },
+    {
+      "epoch": 2.76208,
+      "grad_norm": 1.5619485691099175,
+      "learning_rate": 3.3825003441305127e-06,
+      "loss": 1.1123,
+      "step": 17263
+    },
+    {
+      "epoch": 2.7622400000000003,
+      "grad_norm": 1.8823472176376286,
+      "learning_rate": 3.3780470819931077e-06,
+      "loss": 1.0309,
+      "step": 17264
+    },
+    {
+      "epoch": 2.7624,
+      "grad_norm": 1.6215620182253858,
+      "learning_rate": 3.3735967029139946e-06,
+      "loss": 1.0604,
+      "step": 17265
+    },
+    {
+      "epoch": 2.76256,
+      "grad_norm": 1.7403016210816864,
+      "learning_rate": 3.369149207025957e-06,
+      "loss": 1.0599,
+      "step": 17266
+    },
+    {
+      "epoch": 2.76272,
+      "grad_norm": 1.8714030212015886,
+      "learning_rate": 3.3647045944616983e-06,
+      "loss": 1.0621,
+      "step": 17267
+    },
+    {
+      "epoch": 2.76288,
+      "grad_norm": 1.6337199072590252,
+      "learning_rate": 3.360262865353847e-06,
+      "loss": 1.0517,
+      "step": 17268
+    },
+    {
+      "epoch": 2.76304,
+      "grad_norm": 1.7556676723343334,
+      "learning_rate": 3.355824019834919e-06,
+      "loss": 1.0967,
+      "step": 17269
+    },
+    {
+      "epoch": 2.7632,
+      "grad_norm": 1.684383854023124,
+      "learning_rate": 3.351388058037375e-06,
+      "loss": 1.0762,
+      "step": 17270
+    },
+    {
+      "epoch": 2.76336,
+      "grad_norm": 1.6671806222679295,
+      "learning_rate": 3.346954980093608e-06,
+      "loss": 1.0716,
+      "step": 17271
+    },
+    {
+      "epoch": 2.7635199999999998,
+      "grad_norm": 1.9113391621123543,
+      "learning_rate": 3.3425247861358367e-06,
+      "loss": 1.0837,
+      "step": 17272
+    },
+    {
+      "epoch": 2.76368,
+      "grad_norm": 1.6406811635375165,
+      "learning_rate": 3.338097476296298e-06,
+      "loss": 1.006,
+      "step": 17273
+    },
+    {
+      "epoch": 2.76384,
+      "grad_norm": 1.7412886304796327,
+      "learning_rate": 3.333673050707065e-06,
+      "loss": 1.1119,
+      "step": 17274
+    },
+    {
+      "epoch": 2.7640000000000002,
+      "grad_norm": 1.9025359883255488,
+      "learning_rate": 3.3292515095001863e-06,
+      "loss": 1.123,
+      "step": 17275
+    },
+    {
+      "epoch": 2.76416,
+      "grad_norm": 1.964899416742802,
+      "learning_rate": 3.324832852807569e-06,
+      "loss": 1.0862,
+      "step": 17276
+    },
+    {
+      "epoch": 2.76432,
+      "grad_norm": 1.7606623582772156,
+      "learning_rate": 3.320417080761062e-06,
+      "loss": 1.0809,
+      "step": 17277
+    },
+    {
+      "epoch": 2.76448,
+      "grad_norm": 1.9336029078857941,
+      "learning_rate": 3.31600419349245e-06,
+      "loss": 1.0803,
+      "step": 17278
+    },
+    {
+      "epoch": 2.76464,
+      "grad_norm": 1.675094047224311,
+      "learning_rate": 3.3115941911333827e-06,
+      "loss": 1.1249,
+      "step": 17279
+    },
+    {
+      "epoch": 2.7648,
+      "grad_norm": 1.689332707999544,
+      "learning_rate": 3.307187073815443e-06,
+      "loss": 1.0928,
+      "step": 17280
+    },
+    {
+      "epoch": 2.76496,
+      "grad_norm": 1.7011591064703977,
+      "learning_rate": 3.30278284167016e-06,
+      "loss": 1.1082,
+      "step": 17281
+    },
+    {
+      "epoch": 2.76512,
+      "grad_norm": 1.8053516211963685,
+      "learning_rate": 3.298381494828928e-06,
+      "loss": 1.0193,
+      "step": 17282
+    },
+    {
+      "epoch": 2.7652799999999997,
+      "grad_norm": 1.723959744415416,
+      "learning_rate": 3.293983033423087e-06,
+      "loss": 1.0107,
+      "step": 17283
+    },
+    {
+      "epoch": 2.76544,
+      "grad_norm": 1.9738221631337185,
+      "learning_rate": 3.289587457583876e-06,
+      "loss": 1.1179,
+      "step": 17284
+    },
+    {
+      "epoch": 2.7656,
+      "grad_norm": 1.581212944996635,
+      "learning_rate": 3.2851947674424456e-06,
+      "loss": 1.0295,
+      "step": 17285
+    },
+    {
+      "epoch": 2.76576,
+      "grad_norm": 1.7905201397656745,
+      "learning_rate": 3.2808049631298687e-06,
+      "loss": 1.0561,
+      "step": 17286
+    },
+    {
+      "epoch": 2.76592,
+      "grad_norm": 1.9337498004484697,
+      "learning_rate": 3.2764180447771632e-06,
+      "loss": 1.059,
+      "step": 17287
+    },
+    {
+      "epoch": 2.76608,
+      "grad_norm": 1.6362762312614074,
+      "learning_rate": 3.2720340125151793e-06,
+      "loss": 1.0557,
+      "step": 17288
+    },
+    {
+      "epoch": 2.76624,
+      "grad_norm": 1.8193189437060897,
+      "learning_rate": 3.2676528664747687e-06,
+      "loss": 1.0161,
+      "step": 17289
+    },
+    {
+      "epoch": 2.7664,
+      "grad_norm": 1.6809714828477056,
+      "learning_rate": 3.263274606786626e-06,
+      "loss": 1.0057,
+      "step": 17290
+    },
+    {
+      "epoch": 2.76656,
+      "grad_norm": 1.7990112426748106,
+      "learning_rate": 3.258899233581425e-06,
+      "loss": 1.0609,
+      "step": 17291
+    },
+    {
+      "epoch": 2.76672,
+      "grad_norm": 1.7513708831645458,
+      "learning_rate": 3.254526746989683e-06,
+      "loss": 1.0738,
+      "step": 17292
+    },
+    {
+      "epoch": 2.76688,
+      "grad_norm": 1.7115649415202192,
+      "learning_rate": 3.250157147141908e-06,
+      "loss": 1.0399,
+      "step": 17293
+    },
+    {
+      "epoch": 2.76704,
+      "grad_norm": 2.0357910011853875,
+      "learning_rate": 3.245790434168461e-06,
+      "loss": 1.0609,
+      "step": 17294
+    },
+    {
+      "epoch": 2.7672,
+      "grad_norm": 1.6653248354768204,
+      "learning_rate": 3.241426608199638e-06,
+      "loss": 1.0701,
+      "step": 17295
+    },
+    {
+      "epoch": 2.76736,
+      "grad_norm": 1.6470517286735822,
+      "learning_rate": 3.2370656693656575e-06,
+      "loss": 1.072,
+      "step": 17296
+    },
+    {
+      "epoch": 2.76752,
+      "grad_norm": 1.6222363651245801,
+      "learning_rate": 3.2327076177966374e-06,
+      "loss": 1.0573,
+      "step": 17297
+    },
+    {
+      "epoch": 2.76768,
+      "grad_norm": 1.8800925371299069,
+      "learning_rate": 3.2283524536226184e-06,
+      "loss": 1.0748,
+      "step": 17298
+    },
+    {
+      "epoch": 2.76784,
+      "grad_norm": 1.7524787274113585,
+      "learning_rate": 3.224000176973552e-06,
+      "loss": 1.0936,
+      "step": 17299
+    },
+    {
+      "epoch": 2.768,
+      "grad_norm": 1.650837207329334,
+      "learning_rate": 3.2196507879793003e-06,
+      "loss": 1.0363,
+      "step": 17300
+    },
+    {
+      "epoch": 2.76816,
+      "grad_norm": 1.9044721938623443,
+      "learning_rate": 3.2153042867696383e-06,
+      "loss": 1.0939,
+      "step": 17301
+    },
+    {
+      "epoch": 2.76832,
+      "grad_norm": 1.9839876701476036,
+      "learning_rate": 3.2109606734742616e-06,
+      "loss": 1.0409,
+      "step": 17302
+    },
+    {
+      "epoch": 2.7684800000000003,
+      "grad_norm": 1.7690411537687158,
+      "learning_rate": 3.206619948222811e-06,
+      "loss": 1.0872,
+      "step": 17303
+    },
+    {
+      "epoch": 2.76864,
+      "grad_norm": 1.7705033397953402,
+      "learning_rate": 3.2022821111447497e-06,
+      "loss": 1.087,
+      "step": 17304
+    },
+    {
+      "epoch": 2.7688,
+      "grad_norm": 1.949425055483647,
+      "learning_rate": 3.1979471623695635e-06,
+      "loss": 1.0782,
+      "step": 17305
+    },
+    {
+      "epoch": 2.76896,
+      "grad_norm": 1.915793914363276,
+      "learning_rate": 3.1936151020265483e-06,
+      "loss": 1.0761,
+      "step": 17306
+    },
+    {
+      "epoch": 2.76912,
+      "grad_norm": 1.8174659275150193,
+      "learning_rate": 3.1892859302450116e-06,
+      "loss": 1.0476,
+      "step": 17307
+    },
+    {
+      "epoch": 2.76928,
+      "grad_norm": 1.5769580556708132,
+      "learning_rate": 3.184959647154129e-06,
+      "loss": 1.0263,
+      "step": 17308
+    },
+    {
+      "epoch": 2.76944,
+      "grad_norm": 1.861551372027792,
+      "learning_rate": 3.1806362528829627e-06,
+      "loss": 1.0519,
+      "step": 17309
+    },
+    {
+      "epoch": 2.7696,
+      "grad_norm": 1.6931981995669065,
+      "learning_rate": 3.176315747560532e-06,
+      "loss": 1.056,
+      "step": 17310
+    },
+    {
+      "epoch": 2.7697599999999998,
+      "grad_norm": 1.8430078905248042,
+      "learning_rate": 3.1719981313157455e-06,
+      "loss": 1.0773,
+      "step": 17311
+    },
+    {
+      "epoch": 2.76992,
+      "grad_norm": 2.087460078154182,
+      "learning_rate": 3.167683404277455e-06,
+      "loss": 1.0875,
+      "step": 17312
+    },
+    {
+      "epoch": 2.77008,
+      "grad_norm": 1.9369042983149911,
+      "learning_rate": 3.163371566574369e-06,
+      "loss": 1.0261,
+      "step": 17313
+    },
+    {
+      "epoch": 2.7702400000000003,
+      "grad_norm": 1.9990539777680176,
+      "learning_rate": 3.1590626183352066e-06,
+      "loss": 1.1192,
+      "step": 17314
+    },
+    {
+      "epoch": 2.7704,
+      "grad_norm": 1.618521986657911,
+      "learning_rate": 3.1547565596884653e-06,
+      "loss": 1.0366,
+      "step": 17315
+    },
+    {
+      "epoch": 2.77056,
+      "grad_norm": 1.6765639083681465,
+      "learning_rate": 3.1504533907626975e-06,
+      "loss": 1.0401,
+      "step": 17316
+    },
+    {
+      "epoch": 2.77072,
+      "grad_norm": 1.6224566124776718,
+      "learning_rate": 3.1461531116862562e-06,
+      "loss": 1.0227,
+      "step": 17317
+    },
+    {
+      "epoch": 2.77088,
+      "grad_norm": 1.6225861210149695,
+      "learning_rate": 3.1418557225874835e-06,
+      "loss": 1.0229,
+      "step": 17318
+    },
+    {
+      "epoch": 2.77104,
+      "grad_norm": 1.8110603454934746,
+      "learning_rate": 3.1375612235946094e-06,
+      "loss": 1.0216,
+      "step": 17319
+    },
+    {
+      "epoch": 2.7712,
+      "grad_norm": 1.7552284424576892,
+      "learning_rate": 3.1332696148357545e-06,
+      "loss": 1.0394,
+      "step": 17320
+    },
+    {
+      "epoch": 2.77136,
+      "grad_norm": 1.9006939262610454,
+      "learning_rate": 3.128980896438982e-06,
+      "loss": 1.0492,
+      "step": 17321
+    },
+    {
+      "epoch": 2.7715199999999998,
+      "grad_norm": 1.7296596388912924,
+      "learning_rate": 3.1246950685322685e-06,
+      "loss": 1.0093,
+      "step": 17322
+    },
+    {
+      "epoch": 2.77168,
+      "grad_norm": 1.819640482962972,
+      "learning_rate": 3.120412131243489e-06,
+      "loss": 1.0801,
+      "step": 17323
+    },
+    {
+      "epoch": 2.77184,
+      "grad_norm": 1.8453962341576302,
+      "learning_rate": 3.1161320847004514e-06,
+      "loss": 1.0439,
+      "step": 17324
+    },
+    {
+      "epoch": 2.7720000000000002,
+      "grad_norm": 1.8751429964121817,
+      "learning_rate": 3.111854929030844e-06,
+      "loss": 1.0354,
+      "step": 17325
+    },
+    {
+      "epoch": 2.77216,
+      "grad_norm": 1.5436491306660274,
+      "learning_rate": 3.1075806643623195e-06,
+      "loss": 1.0563,
+      "step": 17326
+    },
+    {
+      "epoch": 2.77232,
+      "grad_norm": 1.7072280369632118,
+      "learning_rate": 3.103309290822376e-06,
+      "loss": 1.1021,
+      "step": 17327
+    },
+    {
+      "epoch": 2.77248,
+      "grad_norm": 1.8552583705418564,
+      "learning_rate": 3.0990408085385225e-06,
+      "loss": 1.0723,
+      "step": 17328
+    },
+    {
+      "epoch": 2.77264,
+      "grad_norm": 1.8338986838706375,
+      "learning_rate": 3.0947752176380574e-06,
+      "loss": 1.0754,
+      "step": 17329
+    },
+    {
+      "epoch": 2.7728,
+      "grad_norm": 1.7235927884154734,
+      "learning_rate": 3.090512518248334e-06,
+      "loss": 1.1084,
+      "step": 17330
+    },
+    {
+      "epoch": 2.77296,
+      "grad_norm": 1.7059794086443625,
+      "learning_rate": 3.0862527104964623e-06,
+      "loss": 1.0245,
+      "step": 17331
+    },
+    {
+      "epoch": 2.77312,
+      "grad_norm": 2.1851684708447143,
+      "learning_rate": 3.081995794509629e-06,
+      "loss": 1.0293,
+      "step": 17332
+    },
+    {
+      "epoch": 2.7732799999999997,
+      "grad_norm": 1.7076277684823273,
+      "learning_rate": 3.077741770414788e-06,
+      "loss": 1.0395,
+      "step": 17333
+    },
+    {
+      "epoch": 2.77344,
+      "grad_norm": 1.6442535090748125,
+      "learning_rate": 3.073490638338916e-06,
+      "loss": 1.0845,
+      "step": 17334
+    },
+    {
+      "epoch": 2.7736,
+      "grad_norm": 1.719756995770863,
+      "learning_rate": 3.0692423984088446e-06,
+      "loss": 1.051,
+      "step": 17335
+    },
+    {
+      "epoch": 2.7737600000000002,
+      "grad_norm": 1.5410367568216523,
+      "learning_rate": 3.0649970507513394e-06,
+      "loss": 1.0674,
+      "step": 17336
+    },
+    {
+      "epoch": 2.77392,
+      "grad_norm": 1.8025429225770093,
+      "learning_rate": 3.0607545954930648e-06,
+      "loss": 1.0767,
+      "step": 17337
+    },
+    {
+      "epoch": 2.77408,
+      "grad_norm": 1.6684377233936638,
+      "learning_rate": 3.056515032760632e-06,
+      "loss": 1.034,
+      "step": 17338
+    },
+    {
+      "epoch": 2.77424,
+      "grad_norm": 1.6631573662310088,
+      "learning_rate": 3.052278362680516e-06,
+      "loss": 1.0815,
+      "step": 17339
+    },
+    {
+      "epoch": 2.7744,
+      "grad_norm": 1.7530802356519843,
+      "learning_rate": 3.0480445853791507e-06,
+      "loss": 1.0522,
+      "step": 17340
+    },
+    {
+      "epoch": 2.77456,
+      "grad_norm": 1.731256734019679,
+      "learning_rate": 3.043813700982867e-06,
+      "loss": 1.0569,
+      "step": 17341
+    },
+    {
+      "epoch": 2.77472,
+      "grad_norm": 2.167589387747978,
+      "learning_rate": 3.039585709617887e-06,
+      "loss": 1.0489,
+      "step": 17342
+    },
+    {
+      "epoch": 2.77488,
+      "grad_norm": 1.7602055590930321,
+      "learning_rate": 3.035360611410387e-06,
+      "loss": 1.0717,
+      "step": 17343
+    },
+    {
+      "epoch": 2.7750399999999997,
+      "grad_norm": 1.6351584674897386,
+      "learning_rate": 3.0311384064864555e-06,
+      "loss": 1.0836,
+      "step": 17344
+    },
+    {
+      "epoch": 2.7752,
+      "grad_norm": 1.722534362188963,
+      "learning_rate": 3.026919094972025e-06,
+      "loss": 1.071,
+      "step": 17345
+    },
+    {
+      "epoch": 2.77536,
+      "grad_norm": 1.9245084099111558,
+      "learning_rate": 3.02270267699305e-06,
+      "loss": 1.0891,
+      "step": 17346
+    },
+    {
+      "epoch": 2.77552,
+      "grad_norm": 1.778531658756194,
+      "learning_rate": 3.0184891526752966e-06,
+      "loss": 1.0578,
+      "step": 17347
+    },
+    {
+      "epoch": 2.77568,
+      "grad_norm": 1.665290989644023,
+      "learning_rate": 3.0142785221445202e-06,
+      "loss": 1.0304,
+      "step": 17348
+    },
+    {
+      "epoch": 2.77584,
+      "grad_norm": 1.7798064617945761,
+      "learning_rate": 3.010070785526342e-06,
+      "loss": 1.1087,
+      "step": 17349
+    },
+    {
+      "epoch": 2.776,
+      "grad_norm": 1.6001167805460894,
+      "learning_rate": 3.0058659429463287e-06,
+      "loss": 1.0295,
+      "step": 17350
+    },
+    {
+      "epoch": 2.77616,
+      "grad_norm": 1.8451401978364217,
+      "learning_rate": 3.001663994529935e-06,
+      "loss": 1.0715,
+      "step": 17351
+    },
+    {
+      "epoch": 2.77632,
+      "grad_norm": 1.8127767523126384,
+      "learning_rate": 2.9974649404025503e-06,
+      "loss": 1.0976,
+      "step": 17352
+    },
+    {
+      "epoch": 2.77648,
+      "grad_norm": 2.02967835619787,
+      "learning_rate": 2.993268780689462e-06,
+      "loss": 1.0841,
+      "step": 17353
+    },
+    {
+      "epoch": 2.77664,
+      "grad_norm": 2.052823684754729,
+      "learning_rate": 2.989075515515882e-06,
+      "loss": 1.0935,
+      "step": 17354
+    },
+    {
+      "epoch": 2.7768,
+      "grad_norm": 1.9105418354754773,
+      "learning_rate": 2.9848851450069328e-06,
+      "loss": 1.0286,
+      "step": 17355
+    },
+    {
+      "epoch": 2.77696,
+      "grad_norm": 1.770201612331569,
+      "learning_rate": 2.9806976692876465e-06,
+      "loss": 1.0826,
+      "step": 17356
+    },
+    {
+      "epoch": 2.77712,
+      "grad_norm": 1.6992630051445161,
+      "learning_rate": 2.9765130884829573e-06,
+      "loss": 1.07,
+      "step": 17357
+    },
+    {
+      "epoch": 2.77728,
+      "grad_norm": 1.9396011917590583,
+      "learning_rate": 2.9723314027177428e-06,
+      "loss": 1.0877,
+      "step": 17358
+    },
+    {
+      "epoch": 2.77744,
+      "grad_norm": 1.8452579676639829,
+      "learning_rate": 2.96815261211677e-06,
+      "loss": 1.0767,
+      "step": 17359
+    },
+    {
+      "epoch": 2.7776,
+      "grad_norm": 1.5677423680493987,
+      "learning_rate": 2.9639767168047505e-06,
+      "loss": 1.0955,
+      "step": 17360
+    },
+    {
+      "epoch": 2.77776,
+      "grad_norm": 1.7671430341904635,
+      "learning_rate": 2.959803716906251e-06,
+      "loss": 1.0939,
+      "step": 17361
+    },
+    {
+      "epoch": 2.77792,
+      "grad_norm": 1.7419438021869136,
+      "learning_rate": 2.9556336125458273e-06,
+      "loss": 1.1323,
+      "step": 17362
+    },
+    {
+      "epoch": 2.77808,
+      "grad_norm": 1.6746590400934318,
+      "learning_rate": 2.9514664038478688e-06,
+      "loss": 1.0765,
+      "step": 17363
+    },
+    {
+      "epoch": 2.7782400000000003,
+      "grad_norm": 1.7486349762338143,
+      "learning_rate": 2.9473020909367544e-06,
+      "loss": 1.0833,
+      "step": 17364
+    },
+    {
+      "epoch": 2.7784,
+      "grad_norm": 1.904591757397066,
+      "learning_rate": 2.943140673936706e-06,
+      "loss": 1.1146,
+      "step": 17365
+    },
+    {
+      "epoch": 2.77856,
+      "grad_norm": 1.69327134166997,
+      "learning_rate": 2.938982152971914e-06,
+      "loss": 1.0624,
+      "step": 17366
+    },
+    {
+      "epoch": 2.77872,
+      "grad_norm": 2.113788866955142,
+      "learning_rate": 2.9348265281664787e-06,
+      "loss": 1.0629,
+      "step": 17367
+    },
+    {
+      "epoch": 2.77888,
+      "grad_norm": 1.706116137152951,
+      "learning_rate": 2.9306737996443677e-06,
+      "loss": 1.0163,
+      "step": 17368
+    },
+    {
+      "epoch": 2.77904,
+      "grad_norm": 1.7407614230972674,
+      "learning_rate": 2.926523967529515e-06,
+      "loss": 1.0515,
+      "step": 17369
+    },
+    {
+      "epoch": 2.7792,
+      "grad_norm": 1.602350284833157,
+      "learning_rate": 2.9223770319457224e-06,
+      "loss": 1.038,
+      "step": 17370
+    },
+    {
+      "epoch": 2.77936,
+      "grad_norm": 1.6912672573903251,
+      "learning_rate": 2.9182329930167564e-06,
+      "loss": 1.0282,
+      "step": 17371
+    },
+    {
+      "epoch": 2.7795199999999998,
+      "grad_norm": 1.5220811641484917,
+      "learning_rate": 2.9140918508662407e-06,
+      "loss": 1.0503,
+      "step": 17372
+    },
+    {
+      "epoch": 2.77968,
+      "grad_norm": 1.7807349344202212,
+      "learning_rate": 2.909953605617777e-06,
+      "loss": 1.0794,
+      "step": 17373
+    },
+    {
+      "epoch": 2.77984,
+      "grad_norm": 1.7720728601863063,
+      "learning_rate": 2.905818257394799e-06,
+      "loss": 1.0739,
+      "step": 17374
+    },
+    {
+      "epoch": 2.7800000000000002,
+      "grad_norm": 1.767617174485821,
+      "learning_rate": 2.901685806320731e-06,
+      "loss": 1.088,
+      "step": 17375
+    },
+    {
+      "epoch": 2.78016,
+      "grad_norm": 1.5436538021454802,
+      "learning_rate": 2.897556252518874e-06,
+      "loss": 1.0626,
+      "step": 17376
+    },
+    {
+      "epoch": 2.78032,
+      "grad_norm": 2.0815396072927994,
+      "learning_rate": 2.89342959611244e-06,
+      "loss": 1.0948,
+      "step": 17377
+    },
+    {
+      "epoch": 2.78048,
+      "grad_norm": 1.765970540006468,
+      "learning_rate": 2.8893058372245763e-06,
+      "loss": 1.0928,
+      "step": 17378
+    },
+    {
+      "epoch": 2.78064,
+      "grad_norm": 1.9661299437127593,
+      "learning_rate": 2.885184975978317e-06,
+      "loss": 1.0078,
+      "step": 17379
+    },
+    {
+      "epoch": 2.7808,
+      "grad_norm": 1.7575271514248452,
+      "learning_rate": 2.88106701249663e-06,
+      "loss": 1.0791,
+      "step": 17380
+    },
+    {
+      "epoch": 2.78096,
+      "grad_norm": 1.55429293546934,
+      "learning_rate": 2.876951946902384e-06,
+      "loss": 1.0276,
+      "step": 17381
+    },
+    {
+      "epoch": 2.78112,
+      "grad_norm": 1.6553986052468743,
+      "learning_rate": 2.8728397793183703e-06,
+      "loss": 1.089,
+      "step": 17382
+    },
+    {
+      "epoch": 2.7812799999999998,
+      "grad_norm": 1.7378382870409534,
+      "learning_rate": 2.868730509867279e-06,
+      "loss": 1.1144,
+      "step": 17383
+    },
+    {
+      "epoch": 2.78144,
+      "grad_norm": 1.7538116276958156,
+      "learning_rate": 2.864624138671734e-06,
+      "loss": 1.0881,
+      "step": 17384
+    },
+    {
+      "epoch": 2.7816,
+      "grad_norm": 1.6296463022339167,
+      "learning_rate": 2.8605206658542826e-06,
+      "loss": 1.098,
+      "step": 17385
+    },
+    {
+      "epoch": 2.7817600000000002,
+      "grad_norm": 1.9525565330758017,
+      "learning_rate": 2.8564200915373264e-06,
+      "loss": 1.0321,
+      "step": 17386
+    },
+    {
+      "epoch": 2.78192,
+      "grad_norm": 1.7384059676256067,
+      "learning_rate": 2.852322415843256e-06,
+      "loss": 1.0807,
+      "step": 17387
+    },
+    {
+      "epoch": 2.78208,
+      "grad_norm": 1.9502488534280351,
+      "learning_rate": 2.8482276388943073e-06,
+      "loss": 1.0734,
+      "step": 17388
+    },
+    {
+      "epoch": 2.78224,
+      "grad_norm": 1.563787002864677,
+      "learning_rate": 2.844135760812705e-06,
+      "loss": 1.0294,
+      "step": 17389
+    },
+    {
+      "epoch": 2.7824,
+      "grad_norm": 1.8178222064532719,
+      "learning_rate": 2.8400467817205067e-06,
+      "loss": 1.0216,
+      "step": 17390
+    },
+    {
+      "epoch": 2.78256,
+      "grad_norm": 1.7302256523928912,
+      "learning_rate": 2.8359607017397373e-06,
+      "loss": 1.0451,
+      "step": 17391
+    },
+    {
+      "epoch": 2.78272,
+      "grad_norm": 1.628234918258641,
+      "learning_rate": 2.83187752099231e-06,
+      "loss": 1.0298,
+      "step": 17392
+    },
+    {
+      "epoch": 2.78288,
+      "grad_norm": 1.9669408043631205,
+      "learning_rate": 2.8277972396000827e-06,
+      "loss": 1.0816,
+      "step": 17393
+    },
+    {
+      "epoch": 2.7830399999999997,
+      "grad_norm": 1.7303460796378707,
+      "learning_rate": 2.8237198576847813e-06,
+      "loss": 1.0932,
+      "step": 17394
+    },
+    {
+      "epoch": 2.7832,
+      "grad_norm": 1.7744128766517464,
+      "learning_rate": 2.819645375368085e-06,
+      "loss": 1.0446,
+      "step": 17395
+    },
+    {
+      "epoch": 2.78336,
+      "grad_norm": 1.8264791891098253,
+      "learning_rate": 2.815573792771564e-06,
+      "loss": 1.0318,
+      "step": 17396
+    },
+    {
+      "epoch": 2.78352,
+      "grad_norm": 2.0416652283081342,
+      "learning_rate": 2.811505110016699e-06,
+      "loss": 1.0718,
+      "step": 17397
+    },
+    {
+      "epoch": 2.78368,
+      "grad_norm": 1.9739018586267802,
+      "learning_rate": 2.8074393272249144e-06,
+      "loss": 1.0686,
+      "step": 17398
+    },
+    {
+      "epoch": 2.78384,
+      "grad_norm": 1.7661135627282571,
+      "learning_rate": 2.8033764445175136e-06,
+      "loss": 1.0717,
+      "step": 17399
+    },
+    {
+      "epoch": 2.784,
+      "grad_norm": 1.6865754187655695,
+      "learning_rate": 2.799316462015711e-06,
+      "loss": 1.015,
+      "step": 17400
+    },
+    {
+      "epoch": 2.78416,
+      "grad_norm": 1.772947035299686,
+      "learning_rate": 2.795259379840698e-06,
+      "loss": 1.1129,
+      "step": 17401
+    },
+    {
+      "epoch": 2.78432,
+      "grad_norm": 1.8186709686280895,
+      "learning_rate": 2.791205198113478e-06,
+      "loss": 1.1025,
+      "step": 17402
+    },
+    {
+      "epoch": 2.78448,
+      "grad_norm": 1.7132608715430453,
+      "learning_rate": 2.7871539169550655e-06,
+      "loss": 1.0225,
+      "step": 17403
+    },
+    {
+      "epoch": 2.78464,
+      "grad_norm": 1.841369972660779,
+      "learning_rate": 2.7831055364863077e-06,
+      "loss": 1.0229,
+      "step": 17404
+    },
+    {
+      "epoch": 2.7848,
+      "grad_norm": 1.9359459424832697,
+      "learning_rate": 2.779060056828042e-06,
+      "loss": 1.0654,
+      "step": 17405
+    },
+    {
+      "epoch": 2.78496,
+      "grad_norm": 1.7899542986153918,
+      "learning_rate": 2.7750174781009386e-06,
+      "loss": 1.0633,
+      "step": 17406
+    },
+    {
+      "epoch": 2.78512,
+      "grad_norm": 1.7878535527638904,
+      "learning_rate": 2.770977800425645e-06,
+      "loss": 1.0864,
+      "step": 17407
+    },
+    {
+      "epoch": 2.78528,
+      "grad_norm": 1.9373237331836697,
+      "learning_rate": 2.766941023922709e-06,
+      "loss": 1.058,
+      "step": 17408
+    },
+    {
+      "epoch": 2.78544,
+      "grad_norm": 1.7661911793916318,
+      "learning_rate": 2.7629071487125567e-06,
+      "loss": 1.067,
+      "step": 17409
+    },
+    {
+      "epoch": 2.7856,
+      "grad_norm": 1.5865928162987888,
+      "learning_rate": 2.75887617491557e-06,
+      "loss": 1.0638,
+      "step": 17410
+    },
+    {
+      "epoch": 2.78576,
+      "grad_norm": 1.8612574628057785,
+      "learning_rate": 2.75484810265203e-06,
+      "loss": 1.0671,
+      "step": 17411
+    },
+    {
+      "epoch": 2.78592,
+      "grad_norm": 1.5102966215420046,
+      "learning_rate": 2.7508229320421185e-06,
+      "loss": 1.031,
+      "step": 17412
+    },
+    {
+      "epoch": 2.78608,
+      "grad_norm": 1.8131514289505692,
+      "learning_rate": 2.7468006632059396e-06,
+      "loss": 1.0824,
+      "step": 17413
+    },
+    {
+      "epoch": 2.7862400000000003,
+      "grad_norm": 1.6541235519747133,
+      "learning_rate": 2.7427812962635193e-06,
+      "loss": 1.1308,
+      "step": 17414
+    },
+    {
+      "epoch": 2.7864,
+      "grad_norm": 1.5611126364268582,
+      "learning_rate": 2.7387648313347834e-06,
+      "loss": 1.043,
+      "step": 17415
+    },
+    {
+      "epoch": 2.78656,
+      "grad_norm": 1.7527023003537396,
+      "learning_rate": 2.7347512685395705e-06,
+      "loss": 1.058,
+      "step": 17416
+    },
+    {
+      "epoch": 2.78672,
+      "grad_norm": 1.8444780652318533,
+      "learning_rate": 2.7307406079976727e-06,
+      "loss": 1.1032,
+      "step": 17417
+    },
+    {
+      "epoch": 2.78688,
+      "grad_norm": 1.7942661458099272,
+      "learning_rate": 2.726732849828717e-06,
+      "loss": 1.1043,
+      "step": 17418
+    },
+    {
+      "epoch": 2.78704,
+      "grad_norm": 1.9026820016528323,
+      "learning_rate": 2.72272799415233e-06,
+      "loss": 1.0269,
+      "step": 17419
+    },
+    {
+      "epoch": 2.7872,
+      "grad_norm": 1.7278970128793036,
+      "learning_rate": 2.718726041087971e-06,
+      "loss": 1.1042,
+      "step": 17420
+    },
+    {
+      "epoch": 2.78736,
+      "grad_norm": 1.737972452354174,
+      "learning_rate": 2.71472699075509e-06,
+      "loss": 1.1054,
+      "step": 17421
+    },
+    {
+      "epoch": 2.7875199999999998,
+      "grad_norm": 2.0715535744102063,
+      "learning_rate": 2.7107308432729796e-06,
+      "loss": 1.1225,
+      "step": 17422
+    },
+    {
+      "epoch": 2.78768,
+      "grad_norm": 1.7356917182613483,
+      "learning_rate": 2.706737598760911e-06,
+      "loss": 1.0662,
+      "step": 17423
+    },
+    {
+      "epoch": 2.78784,
+      "grad_norm": 1.666410284684658,
+      "learning_rate": 2.702747257338012e-06,
+      "loss": 1.073,
+      "step": 17424
+    },
+    {
+      "epoch": 2.7880000000000003,
+      "grad_norm": 1.5965430692858233,
+      "learning_rate": 2.6987598191233644e-06,
+      "loss": 1.0489,
+      "step": 17425
+    },
+    {
+      "epoch": 2.78816,
+      "grad_norm": 1.7497263796066316,
+      "learning_rate": 2.69477528423594e-06,
+      "loss": 1.095,
+      "step": 17426
+    },
+    {
+      "epoch": 2.78832,
+      "grad_norm": 1.6564995354031848,
+      "learning_rate": 2.690793652794621e-06,
+      "loss": 1.0716,
+      "step": 17427
+    },
+    {
+      "epoch": 2.78848,
+      "grad_norm": 1.7736896952247985,
+      "learning_rate": 2.686814924918235e-06,
+      "loss": 1.0572,
+      "step": 17428
+    },
+    {
+      "epoch": 2.78864,
+      "grad_norm": 1.6545467042264053,
+      "learning_rate": 2.682839100725487e-06,
+      "loss": 1.0593,
+      "step": 17429
+    },
+    {
+      "epoch": 2.7888,
+      "grad_norm": 1.7409659525383498,
+      "learning_rate": 2.678866180335027e-06,
+      "loss": 1.06,
+      "step": 17430
+    },
+    {
+      "epoch": 2.78896,
+      "grad_norm": 1.670635036711542,
+      "learning_rate": 2.67489616386537e-06,
+      "loss": 1.0361,
+      "step": 17431
+    },
+    {
+      "epoch": 2.78912,
+      "grad_norm": 1.6360353251711486,
+      "learning_rate": 2.6709290514350007e-06,
+      "loss": 1.0606,
+      "step": 17432
+    },
+    {
+      "epoch": 2.7892799999999998,
+      "grad_norm": 1.6625658346623478,
+      "learning_rate": 2.6669648431622895e-06,
+      "loss": 1.0084,
+      "step": 17433
+    },
+    {
+      "epoch": 2.78944,
+      "grad_norm": 1.7505246543862822,
+      "learning_rate": 2.6630035391655204e-06,
+      "loss": 1.0972,
+      "step": 17434
+    },
+    {
+      "epoch": 2.7896,
+      "grad_norm": 1.7554552864542925,
+      "learning_rate": 2.659045139562877e-06,
+      "loss": 1.0688,
+      "step": 17435
+    },
+    {
+      "epoch": 2.7897600000000002,
+      "grad_norm": 1.5906350264891649,
+      "learning_rate": 2.655089644472497e-06,
+      "loss": 1.1092,
+      "step": 17436
+    },
+    {
+      "epoch": 2.78992,
+      "grad_norm": 1.7446347432915694,
+      "learning_rate": 2.6511370540123983e-06,
+      "loss": 1.1055,
+      "step": 17437
+    },
+    {
+      "epoch": 2.79008,
+      "grad_norm": 1.6233819327078443,
+      "learning_rate": 2.6471873683005076e-06,
+      "loss": 1.1043,
+      "step": 17438
+    },
+    {
+      "epoch": 2.79024,
+      "grad_norm": 1.7357485620004915,
+      "learning_rate": 2.643240587454698e-06,
+      "loss": 1.0557,
+      "step": 17439
+    },
+    {
+      "epoch": 2.7904,
+      "grad_norm": 1.8381339069426643,
+      "learning_rate": 2.63929671159272e-06,
+      "loss": 1.0873,
+      "step": 17440
+    },
+    {
+      "epoch": 2.79056,
+      "grad_norm": 1.6689694702681752,
+      "learning_rate": 2.635355740832246e-06,
+      "loss": 1.0423,
+      "step": 17441
+    },
+    {
+      "epoch": 2.79072,
+      "grad_norm": 1.8858734293666575,
+      "learning_rate": 2.6314176752909035e-06,
+      "loss": 1.044,
+      "step": 17442
+    },
+    {
+      "epoch": 2.79088,
+      "grad_norm": 1.6252368083507387,
+      "learning_rate": 2.6274825150861657e-06,
+      "loss": 1.0531,
+      "step": 17443
+    },
+    {
+      "epoch": 2.7910399999999997,
+      "grad_norm": 1.765069967441556,
+      "learning_rate": 2.6235502603354722e-06,
+      "loss": 1.0238,
+      "step": 17444
+    },
+    {
+      "epoch": 2.7912,
+      "grad_norm": 1.7359662754606122,
+      "learning_rate": 2.6196209111561285e-06,
+      "loss": 1.0289,
+      "step": 17445
+    },
+    {
+      "epoch": 2.79136,
+      "grad_norm": 1.6712362520053867,
+      "learning_rate": 2.6156944676654195e-06,
+      "loss": 1.0774,
+      "step": 17446
+    },
+    {
+      "epoch": 2.7915200000000002,
+      "grad_norm": 1.6314319039376421,
+      "learning_rate": 2.6117709299804614e-06,
+      "loss": 1.1144,
+      "step": 17447
+    },
+    {
+      "epoch": 2.79168,
+      "grad_norm": 1.9627856874320209,
+      "learning_rate": 2.6078502982183617e-06,
+      "loss": 1.0979,
+      "step": 17448
+    },
+    {
+      "epoch": 2.79184,
+      "grad_norm": 1.752691125998695,
+      "learning_rate": 2.603932572496093e-06,
+      "loss": 1.0382,
+      "step": 17449
+    },
+    {
+      "epoch": 2.792,
+      "grad_norm": 1.839823314652347,
+      "learning_rate": 2.6000177529305615e-06,
+      "loss": 1.0298,
+      "step": 17450
+    },
+    {
+      "epoch": 2.79216,
+      "grad_norm": 1.6636631671458633,
+      "learning_rate": 2.596105839638563e-06,
+      "loss": 1.0492,
+      "step": 17451
+    },
+    {
+      "epoch": 2.79232,
+      "grad_norm": 2.1447185362812538,
+      "learning_rate": 2.5921968327368372e-06,
+      "loss": 1.0555,
+      "step": 17452
+    },
+    {
+      "epoch": 2.79248,
+      "grad_norm": 1.7023288602126718,
+      "learning_rate": 2.588290732342025e-06,
+      "loss": 1.067,
+      "step": 17453
+    },
+    {
+      "epoch": 2.79264,
+      "grad_norm": 1.7860549764583094,
+      "learning_rate": 2.5843875385706652e-06,
+      "loss": 1.0492,
+      "step": 17454
+    },
+    {
+      "epoch": 2.7927999999999997,
+      "grad_norm": 1.893251638907195,
+      "learning_rate": 2.580487251539243e-06,
+      "loss": 1.1175,
+      "step": 17455
+    },
+    {
+      "epoch": 2.79296,
+      "grad_norm": 1.8860386018082604,
+      "learning_rate": 2.57658987136411e-06,
+      "loss": 1.0692,
+      "step": 17456
+    },
+    {
+      "epoch": 2.79312,
+      "grad_norm": 1.88159845952871,
+      "learning_rate": 2.5726953981615842e-06,
+      "loss": 1.0647,
+      "step": 17457
+    },
+    {
+      "epoch": 2.79328,
+      "grad_norm": 1.7669954065194675,
+      "learning_rate": 2.5688038320478614e-06,
+      "loss": 1.0618,
+      "step": 17458
+    },
+    {
+      "epoch": 2.79344,
+      "grad_norm": 1.7328524896525073,
+      "learning_rate": 2.5649151731390486e-06,
+      "loss": 1.0383,
+      "step": 17459
+    },
+    {
+      "epoch": 2.7936,
+      "grad_norm": 1.989913176062622,
+      "learning_rate": 2.561029421551209e-06,
+      "loss": 1.039,
+      "step": 17460
+    },
+    {
+      "epoch": 2.79376,
+      "grad_norm": 1.8148413740507892,
+      "learning_rate": 2.557146577400249e-06,
+      "loss": 1.0659,
+      "step": 17461
+    },
+    {
+      "epoch": 2.79392,
+      "grad_norm": 1.675928737452294,
+      "learning_rate": 2.5532666408020655e-06,
+      "loss": 1.0775,
+      "step": 17462
+    },
+    {
+      "epoch": 2.79408,
+      "grad_norm": 1.7889228424768258,
+      "learning_rate": 2.5493896118723882e-06,
+      "loss": 1.0904,
+      "step": 17463
+    },
+    {
+      "epoch": 2.79424,
+      "grad_norm": 1.8906621483068673,
+      "learning_rate": 2.545515490726935e-06,
+      "loss": 1.064,
+      "step": 17464
+    },
+    {
+      "epoch": 2.7944,
+      "grad_norm": 1.7624188225463453,
+      "learning_rate": 2.541644277481303e-06,
+      "loss": 1.0738,
+      "step": 17465
+    },
+    {
+      "epoch": 2.79456,
+      "grad_norm": 1.722445637903292,
+      "learning_rate": 2.5377759722509777e-06,
+      "loss": 1.0591,
+      "step": 17466
+    },
+    {
+      "epoch": 2.79472,
+      "grad_norm": 1.8191040044668547,
+      "learning_rate": 2.5339105751514105e-06,
+      "loss": 1.1972,
+      "step": 17467
+    },
+    {
+      "epoch": 2.79488,
+      "grad_norm": 1.7368745132546541,
+      "learning_rate": 2.5300480862979316e-06,
+      "loss": 1.0406,
+      "step": 17468
+    },
+    {
+      "epoch": 2.79504,
+      "grad_norm": 1.9217327752799922,
+      "learning_rate": 2.526188505805782e-06,
+      "loss": 1.0596,
+      "step": 17469
+    },
+    {
+      "epoch": 2.7952,
+      "grad_norm": 1.5630926645137349,
+      "learning_rate": 2.522331833790137e-06,
+      "loss": 1.0469,
+      "step": 17470
+    },
+    {
+      "epoch": 2.79536,
+      "grad_norm": 1.6644351723875672,
+      "learning_rate": 2.5184780703660592e-06,
+      "loss": 1.0522,
+      "step": 17471
+    },
+    {
+      "epoch": 2.79552,
+      "grad_norm": 1.6460395412671711,
+      "learning_rate": 2.5146272156485572e-06,
+      "loss": 1.0315,
+      "step": 17472
+    },
+    {
+      "epoch": 2.79568,
+      "grad_norm": 1.639703275918034,
+      "learning_rate": 2.510779269752528e-06,
+      "loss": 1.0541,
+      "step": 17473
+    },
+    {
+      "epoch": 2.79584,
+      "grad_norm": 1.715258338882066,
+      "learning_rate": 2.5069342327927904e-06,
+      "loss": 1.0584,
+      "step": 17474
+    },
+    {
+      "epoch": 2.7960000000000003,
+      "grad_norm": 1.5821016436901556,
+      "learning_rate": 2.503092104884064e-06,
+      "loss": 1.0651,
+      "step": 17475
+    },
+    {
+      "epoch": 2.79616,
+      "grad_norm": 1.718292464300394,
+      "learning_rate": 2.4992528861410237e-06,
+      "loss": 1.0797,
+      "step": 17476
+    },
+    {
+      "epoch": 2.79632,
+      "grad_norm": 1.8149420410706254,
+      "learning_rate": 2.495416576678178e-06,
+      "loss": 1.076,
+      "step": 17477
+    },
+    {
+      "epoch": 2.79648,
+      "grad_norm": 1.8671320771140512,
+      "learning_rate": 2.4915831766100462e-06,
+      "loss": 1.0871,
+      "step": 17478
+    },
+    {
+      "epoch": 2.79664,
+      "grad_norm": 1.8185428943383901,
+      "learning_rate": 2.4877526860509594e-06,
+      "loss": 1.081,
+      "step": 17479
+    },
+    {
+      "epoch": 2.7968,
+      "grad_norm": 1.7141343064614292,
+      "learning_rate": 2.483925105115259e-06,
+      "loss": 1.0342,
+      "step": 17480
+    },
+    {
+      "epoch": 2.79696,
+      "grad_norm": 1.759414078050816,
+      "learning_rate": 2.480100433917143e-06,
+      "loss": 1.1206,
+      "step": 17481
+    },
+    {
+      "epoch": 2.79712,
+      "grad_norm": 1.814633291673303,
+      "learning_rate": 2.47627867257072e-06,
+      "loss": 1.0751,
+      "step": 17482
+    },
+    {
+      "epoch": 2.7972799999999998,
+      "grad_norm": 1.8057125519279125,
+      "learning_rate": 2.4724598211900317e-06,
+      "loss": 1.0338,
+      "step": 17483
+    },
+    {
+      "epoch": 2.79744,
+      "grad_norm": 1.8092425350544632,
+      "learning_rate": 2.4686438798890322e-06,
+      "loss": 1.0415,
+      "step": 17484
+    },
+    {
+      "epoch": 2.7976,
+      "grad_norm": 1.7345038568683886,
+      "learning_rate": 2.464830848781585e-06,
+      "loss": 1.0357,
+      "step": 17485
+    },
+    {
+      "epoch": 2.7977600000000002,
+      "grad_norm": 1.701638151073405,
+      "learning_rate": 2.461020727981467e-06,
+      "loss": 1.0372,
+      "step": 17486
+    },
+    {
+      "epoch": 2.79792,
+      "grad_norm": 1.792322019968945,
+      "learning_rate": 2.4572135176023524e-06,
+      "loss": 1.0975,
+      "step": 17487
+    },
+    {
+      "epoch": 2.79808,
+      "grad_norm": 1.8226352533790393,
+      "learning_rate": 2.4534092177578403e-06,
+      "loss": 1.1277,
+      "step": 17488
+    },
+    {
+      "epoch": 2.79824,
+      "grad_norm": 1.619491995476564,
+      "learning_rate": 2.4496078285614734e-06,
+      "loss": 1.0888,
+      "step": 17489
+    },
+    {
+      "epoch": 2.7984,
+      "grad_norm": 1.7722049000714903,
+      "learning_rate": 2.44580935012666e-06,
+      "loss": 1.0474,
+      "step": 17490
+    },
+    {
+      "epoch": 2.79856,
+      "grad_norm": 1.6303684292202696,
+      "learning_rate": 2.4420137825667433e-06,
+      "loss": 1.0521,
+      "step": 17491
+    },
+    {
+      "epoch": 2.79872,
+      "grad_norm": 1.939173059284266,
+      "learning_rate": 2.438221125994977e-06,
+      "loss": 1.0285,
+      "step": 17492
+    },
+    {
+      "epoch": 2.79888,
+      "grad_norm": 1.7181623119262073,
+      "learning_rate": 2.4344313805245267e-06,
+      "loss": 1.0437,
+      "step": 17493
+    },
+    {
+      "epoch": 2.7990399999999998,
+      "grad_norm": 1.8515504439467956,
+      "learning_rate": 2.430644546268479e-06,
+      "loss": 1.0785,
+      "step": 17494
+    },
+    {
+      "epoch": 2.7992,
+      "grad_norm": 1.9363298944623326,
+      "learning_rate": 2.4268606233398216e-06,
+      "loss": 1.0713,
+      "step": 17495
+    },
+    {
+      "epoch": 2.79936,
+      "grad_norm": 1.6738932427379996,
+      "learning_rate": 2.423079611851475e-06,
+      "loss": 1.0718,
+      "step": 17496
+    },
+    {
+      "epoch": 2.7995200000000002,
+      "grad_norm": 1.7171301944346156,
+      "learning_rate": 2.419301511916239e-06,
+      "loss": 1.036,
+      "step": 17497
+    },
+    {
+      "epoch": 2.79968,
+      "grad_norm": 1.6121757346528642,
+      "learning_rate": 2.4155263236468552e-06,
+      "loss": 1.0391,
+      "step": 17498
+    },
+    {
+      "epoch": 2.79984,
+      "grad_norm": 1.6530794818573131,
+      "learning_rate": 2.4117540471559784e-06,
+      "loss": 1.0467,
+      "step": 17499
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.6874333516364666,
+      "learning_rate": 2.4079846825561413e-06,
+      "loss": 1.0605,
+      "step": 17500
+    },
+    {
+      "epoch": 2.80016,
+      "grad_norm": 1.847601302285175,
+      "learning_rate": 2.4042182299598647e-06,
+      "loss": 1.0877,
+      "step": 17501
+    },
+    {
+      "epoch": 2.80032,
+      "grad_norm": 1.7257083218193698,
+      "learning_rate": 2.4004546894794698e-06,
+      "loss": 1.074,
+      "step": 17502
+    },
+    {
+      "epoch": 2.80048,
+      "grad_norm": 1.6978707187681774,
+      "learning_rate": 2.3966940612273227e-06,
+      "loss": 1.0848,
+      "step": 17503
+    },
+    {
+      "epoch": 2.80064,
+      "grad_norm": 1.594491707258665,
+      "learning_rate": 2.3929363453155774e-06,
+      "loss": 1.0568,
+      "step": 17504
+    },
+    {
+      "epoch": 2.8007999999999997,
+      "grad_norm": 1.5460717196563025,
+      "learning_rate": 2.389181541856389e-06,
+      "loss": 0.9748,
+      "step": 17505
+    },
+    {
+      "epoch": 2.80096,
+      "grad_norm": 1.6900029873045432,
+      "learning_rate": 2.38542965096179e-06,
+      "loss": 1.0855,
+      "step": 17506
+    },
+    {
+      "epoch": 2.80112,
+      "grad_norm": 1.6587504218565712,
+      "learning_rate": 2.3816806727437358e-06,
+      "loss": 1.0749,
+      "step": 17507
+    },
+    {
+      "epoch": 2.80128,
+      "grad_norm": 1.6422854929547146,
+      "learning_rate": 2.3779346073140807e-06,
+      "loss": 1.0397,
+      "step": 17508
+    },
+    {
+      "epoch": 2.80144,
+      "grad_norm": 1.8620498405660224,
+      "learning_rate": 2.374191454784602e-06,
+      "loss": 1.0905,
+      "step": 17509
+    },
+    {
+      "epoch": 2.8016,
+      "grad_norm": 1.6845344252841052,
+      "learning_rate": 2.3704512152669998e-06,
+      "loss": 1.068,
+      "step": 17510
+    },
+    {
+      "epoch": 2.80176,
+      "grad_norm": 1.700462488722197,
+      "learning_rate": 2.3667138888728735e-06,
+      "loss": 0.9967,
+      "step": 17511
+    },
+    {
+      "epoch": 2.80192,
+      "grad_norm": 1.8962682561785678,
+      "learning_rate": 2.3629794757137445e-06,
+      "loss": 1.0373,
+      "step": 17512
+    },
+    {
+      "epoch": 2.80208,
+      "grad_norm": 2.1191042439146663,
+      "learning_rate": 2.3592479759010244e-06,
+      "loss": 1.0832,
+      "step": 17513
+    },
+    {
+      "epoch": 2.80224,
+      "grad_norm": 1.6324874123754058,
+      "learning_rate": 2.3555193895460682e-06,
+      "loss": 1.0644,
+      "step": 17514
+    },
+    {
+      "epoch": 2.8024,
+      "grad_norm": 1.639081598437492,
+      "learning_rate": 2.3517937167601423e-06,
+      "loss": 1.0546,
+      "step": 17515
+    },
+    {
+      "epoch": 2.80256,
+      "grad_norm": 1.6621922901587904,
+      "learning_rate": 2.3480709576543804e-06,
+      "loss": 1.0788,
+      "step": 17516
+    },
+    {
+      "epoch": 2.80272,
+      "grad_norm": 1.566442937632009,
+      "learning_rate": 2.3443511123399152e-06,
+      "loss": 1.0526,
+      "step": 17517
+    },
+    {
+      "epoch": 2.80288,
+      "grad_norm": 1.8830420157768974,
+      "learning_rate": 2.340634180927692e-06,
+      "loss": 1.019,
+      "step": 17518
+    },
+    {
+      "epoch": 2.80304,
+      "grad_norm": 1.6560972803238703,
+      "learning_rate": 2.3369201635286553e-06,
+      "loss": 1.0377,
+      "step": 17519
+    },
+    {
+      "epoch": 2.8032,
+      "grad_norm": 1.697533490731012,
+      "learning_rate": 2.333209060253594e-06,
+      "loss": 1.0199,
+      "step": 17520
+    },
+    {
+      "epoch": 2.80336,
+      "grad_norm": 1.8255728880130924,
+      "learning_rate": 2.3295008712132636e-06,
+      "loss": 1.1146,
+      "step": 17521
+    },
+    {
+      "epoch": 2.80352,
+      "grad_norm": 1.866368330900619,
+      "learning_rate": 2.32579559651831e-06,
+      "loss": 1.0337,
+      "step": 17522
+    },
+    {
+      "epoch": 2.80368,
+      "grad_norm": 1.8018191994570798,
+      "learning_rate": 2.322093236279288e-06,
+      "loss": 1.0596,
+      "step": 17523
+    },
+    {
+      "epoch": 2.80384,
+      "grad_norm": 1.8953707902494483,
+      "learning_rate": 2.3183937906066766e-06,
+      "loss": 1.1188,
+      "step": 17524
+    },
+    {
+      "epoch": 2.8040000000000003,
+      "grad_norm": 1.6367129885394585,
+      "learning_rate": 2.314697259610854e-06,
+      "loss": 1.0478,
+      "step": 17525
+    },
+    {
+      "epoch": 2.80416,
+      "grad_norm": 1.7565890809804268,
+      "learning_rate": 2.311003643402121e-06,
+      "loss": 1.0544,
+      "step": 17526
+    },
+    {
+      "epoch": 2.80432,
+      "grad_norm": 1.8206674241366088,
+      "learning_rate": 2.3073129420907004e-06,
+      "loss": 1.0681,
+      "step": 17527
+    },
+    {
+      "epoch": 2.80448,
+      "grad_norm": 1.693671028306695,
+      "learning_rate": 2.303625155786704e-06,
+      "loss": 1.0581,
+      "step": 17528
+    },
+    {
+      "epoch": 2.80464,
+      "grad_norm": 1.7931400597967415,
+      "learning_rate": 2.2999402846001663e-06,
+      "loss": 1.0329,
+      "step": 17529
+    },
+    {
+      "epoch": 2.8048,
+      "grad_norm": 1.758111176655071,
+      "learning_rate": 2.2962583286410434e-06,
+      "loss": 1.0976,
+      "step": 17530
+    },
+    {
+      "epoch": 2.80496,
+      "grad_norm": 1.6073681377442939,
+      "learning_rate": 2.2925792880192252e-06,
+      "loss": 1.056,
+      "step": 17531
+    },
+    {
+      "epoch": 2.80512,
+      "grad_norm": 1.6926796745509054,
+      "learning_rate": 2.2889031628444356e-06,
+      "loss": 1.1112,
+      "step": 17532
+    },
+    {
+      "epoch": 2.8052799999999998,
+      "grad_norm": 1.8361121265908726,
+      "learning_rate": 2.285229953226431e-06,
+      "loss": 1.0711,
+      "step": 17533
+    },
+    {
+      "epoch": 2.80544,
+      "grad_norm": 1.6280835932282411,
+      "learning_rate": 2.2815596592747455e-06,
+      "loss": 1.0737,
+      "step": 17534
+    },
+    {
+      "epoch": 2.8056,
+      "grad_norm": 1.7288709453774938,
+      "learning_rate": 2.277892281098959e-06,
+      "loss": 1.098,
+      "step": 17535
+    },
+    {
+      "epoch": 2.8057600000000003,
+      "grad_norm": 1.7912259930890524,
+      "learning_rate": 2.274227818808439e-06,
+      "loss": 1.0919,
+      "step": 17536
+    },
+    {
+      "epoch": 2.80592,
+      "grad_norm": 1.845392026315309,
+      "learning_rate": 2.270566272512575e-06,
+      "loss": 1.0841,
+      "step": 17537
+    },
+    {
+      "epoch": 2.80608,
+      "grad_norm": 1.9034954322865523,
+      "learning_rate": 2.266907642320615e-06,
+      "loss": 1.0845,
+      "step": 17538
+    },
+    {
+      "epoch": 2.80624,
+      "grad_norm": 1.8296117349320455,
+      "learning_rate": 2.2632519283417144e-06,
+      "loss": 1.0804,
+      "step": 17539
+    },
+    {
+      "epoch": 2.8064,
+      "grad_norm": 1.663390690685949,
+      "learning_rate": 2.2595991306849527e-06,
+      "loss": 1.03,
+      "step": 17540
+    },
+    {
+      "epoch": 2.80656,
+      "grad_norm": 1.5589773814142283,
+      "learning_rate": 2.2559492494593325e-06,
+      "loss": 1.0792,
+      "step": 17541
+    },
+    {
+      "epoch": 2.80672,
+      "grad_norm": 1.5373257496716886,
+      "learning_rate": 2.252302284773755e-06,
+      "loss": 1.0756,
+      "step": 17542
+    },
+    {
+      "epoch": 2.80688,
+      "grad_norm": 1.7348314994368406,
+      "learning_rate": 2.2486582367370558e-06,
+      "loss": 1.0442,
+      "step": 17543
+    },
+    {
+      "epoch": 2.8070399999999998,
+      "grad_norm": 1.7620632326704362,
+      "learning_rate": 2.2450171054579473e-06,
+      "loss": 1.0749,
+      "step": 17544
+    },
+    {
+      "epoch": 2.8072,
+      "grad_norm": 2.0385881918868,
+      "learning_rate": 2.241378891045076e-06,
+      "loss": 1.0737,
+      "step": 17545
+    },
+    {
+      "epoch": 2.80736,
+      "grad_norm": 1.695087455932865,
+      "learning_rate": 2.237743593607011e-06,
+      "loss": 1.0458,
+      "step": 17546
+    },
+    {
+      "epoch": 2.8075200000000002,
+      "grad_norm": 1.8647772318679257,
+      "learning_rate": 2.2341112132522323e-06,
+      "loss": 1.0475,
+      "step": 17547
+    },
+    {
+      "epoch": 2.80768,
+      "grad_norm": 1.84245177203884,
+      "learning_rate": 2.2304817500891195e-06,
+      "loss": 1.0519,
+      "step": 17548
+    },
+    {
+      "epoch": 2.80784,
+      "grad_norm": 1.797130730800278,
+      "learning_rate": 2.2268552042259528e-06,
+      "loss": 1.0549,
+      "step": 17549
+    },
+    {
+      "epoch": 2.808,
+      "grad_norm": 1.526941123368533,
+      "learning_rate": 2.2232315757709677e-06,
+      "loss": 1.0515,
+      "step": 17550
+    },
+    {
+      "epoch": 2.80816,
+      "grad_norm": 1.465903529293568,
+      "learning_rate": 2.2196108648322665e-06,
+      "loss": 1.031,
+      "step": 17551
+    },
+    {
+      "epoch": 2.80832,
+      "grad_norm": 1.8393010508149419,
+      "learning_rate": 2.2159930715179077e-06,
+      "loss": 1.0503,
+      "step": 17552
+    },
+    {
+      "epoch": 2.80848,
+      "grad_norm": 1.7250924926208688,
+      "learning_rate": 2.212378195935816e-06,
+      "loss": 1.0629,
+      "step": 17553
+    },
+    {
+      "epoch": 2.80864,
+      "grad_norm": 1.664814104709552,
+      "learning_rate": 2.208766238193882e-06,
+      "loss": 1.076,
+      "step": 17554
+    },
+    {
+      "epoch": 2.8087999999999997,
+      "grad_norm": 1.8810602189935182,
+      "learning_rate": 2.2051571983998542e-06,
+      "loss": 1.051,
+      "step": 17555
+    },
+    {
+      "epoch": 2.80896,
+      "grad_norm": 1.8013748575798763,
+      "learning_rate": 2.2015510766614343e-06,
+      "loss": 1.0806,
+      "step": 17556
+    },
+    {
+      "epoch": 2.80912,
+      "grad_norm": 1.6836247116617868,
+      "learning_rate": 2.197947873086226e-06,
+      "loss": 1.0383,
+      "step": 17557
+    },
+    {
+      "epoch": 2.80928,
+      "grad_norm": 1.7779629901908685,
+      "learning_rate": 2.194347587781742e-06,
+      "loss": 1.0612,
+      "step": 17558
+    },
+    {
+      "epoch": 2.80944,
+      "grad_norm": 1.6519684241733377,
+      "learning_rate": 2.1907502208554085e-06,
+      "loss": 1.0703,
+      "step": 17559
+    },
+    {
+      "epoch": 2.8096,
+      "grad_norm": 1.982294941560627,
+      "learning_rate": 2.187155772414573e-06,
+      "loss": 1.07,
+      "step": 17560
+    },
+    {
+      "epoch": 2.80976,
+      "grad_norm": 1.8001112361297142,
+      "learning_rate": 2.1835642425664603e-06,
+      "loss": 1.0626,
+      "step": 17561
+    },
+    {
+      "epoch": 2.80992,
+      "grad_norm": 1.7634303962365467,
+      "learning_rate": 2.179975631418263e-06,
+      "loss": 1.0882,
+      "step": 17562
+    },
+    {
+      "epoch": 2.81008,
+      "grad_norm": 1.5849032337181475,
+      "learning_rate": 2.1763899390770614e-06,
+      "loss": 1.073,
+      "step": 17563
+    },
+    {
+      "epoch": 2.81024,
+      "grad_norm": 1.748586364695047,
+      "learning_rate": 2.172807165649837e-06,
+      "loss": 1.0746,
+      "step": 17564
+    },
+    {
+      "epoch": 2.8104,
+      "grad_norm": 1.6787301186436072,
+      "learning_rate": 2.1692273112435046e-06,
+      "loss": 1.064,
+      "step": 17565
+    },
+    {
+      "epoch": 2.8105599999999997,
+      "grad_norm": 1.6126813726644116,
+      "learning_rate": 2.165650375964867e-06,
+      "loss": 1.0399,
+      "step": 17566
+    },
+    {
+      "epoch": 2.81072,
+      "grad_norm": 1.706649675771459,
+      "learning_rate": 2.1620763599206616e-06,
+      "loss": 1.0738,
+      "step": 17567
+    },
+    {
+      "epoch": 2.81088,
+      "grad_norm": 1.6266789893648004,
+      "learning_rate": 2.158505263217536e-06,
+      "loss": 1.0709,
+      "step": 17568
+    },
+    {
+      "epoch": 2.81104,
+      "grad_norm": 1.7481642329889873,
+      "learning_rate": 2.154937085962039e-06,
+      "loss": 1.1064,
+      "step": 17569
+    },
+    {
+      "epoch": 2.8112,
+      "grad_norm": 1.9615722178522614,
+      "learning_rate": 2.151371828260651e-06,
+      "loss": 1.099,
+      "step": 17570
+    },
+    {
+      "epoch": 2.81136,
+      "grad_norm": 1.693243066540237,
+      "learning_rate": 2.147809490219743e-06,
+      "loss": 1.0709,
+      "step": 17571
+    },
+    {
+      "epoch": 2.81152,
+      "grad_norm": 1.7339096770826838,
+      "learning_rate": 2.1442500719456193e-06,
+      "loss": 1.0593,
+      "step": 17572
+    },
+    {
+      "epoch": 2.81168,
+      "grad_norm": 1.880346889337437,
+      "learning_rate": 2.140693573544461e-06,
+      "loss": 1.0551,
+      "step": 17573
+    },
+    {
+      "epoch": 2.81184,
+      "grad_norm": 1.6904487136965496,
+      "learning_rate": 2.1371399951224392e-06,
+      "loss": 1.0526,
+      "step": 17574
+    },
+    {
+      "epoch": 2.8120000000000003,
+      "grad_norm": 2.077509123791589,
+      "learning_rate": 2.1335893367855355e-06,
+      "loss": 1.0818,
+      "step": 17575
+    },
+    {
+      "epoch": 2.81216,
+      "grad_norm": 1.7966798884711188,
+      "learning_rate": 2.1300415986397427e-06,
+      "loss": 1.0785,
+      "step": 17576
+    },
+    {
+      "epoch": 2.81232,
+      "grad_norm": 1.6996390958794507,
+      "learning_rate": 2.1264967807908764e-06,
+      "loss": 1.0559,
+      "step": 17577
+    },
+    {
+      "epoch": 2.81248,
+      "grad_norm": 1.652111629765719,
+      "learning_rate": 2.122954883344741e-06,
+      "loss": 1.082,
+      "step": 17578
+    },
+    {
+      "epoch": 2.81264,
+      "grad_norm": 1.479818970494364,
+      "learning_rate": 2.1194159064070073e-06,
+      "loss": 1.0687,
+      "step": 17579
+    },
+    {
+      "epoch": 2.8128,
+      "grad_norm": 1.8272665463311677,
+      "learning_rate": 2.1158798500832798e-06,
+      "loss": 1.0707,
+      "step": 17580
+    },
+    {
+      "epoch": 2.81296,
+      "grad_norm": 1.7183875081986177,
+      "learning_rate": 2.112346714479052e-06,
+      "loss": 1.0428,
+      "step": 17581
+    },
+    {
+      "epoch": 2.81312,
+      "grad_norm": 1.9109613976272823,
+      "learning_rate": 2.1088164996997727e-06,
+      "loss": 1.058,
+      "step": 17582
+    },
+    {
+      "epoch": 2.81328,
+      "grad_norm": 1.6680398100258558,
+      "learning_rate": 2.105289205850769e-06,
+      "loss": 1.0633,
+      "step": 17583
+    },
+    {
+      "epoch": 2.81344,
+      "grad_norm": 1.7520742208159459,
+      "learning_rate": 2.1017648330372785e-06,
+      "loss": 1.0562,
+      "step": 17584
+    },
+    {
+      "epoch": 2.8136,
+      "grad_norm": 1.5370179419672265,
+      "learning_rate": 2.098243381364473e-06,
+      "loss": 1.0446,
+      "step": 17585
+    },
+    {
+      "epoch": 2.8137600000000003,
+      "grad_norm": 1.6515651228038832,
+      "learning_rate": 2.0947248509374353e-06,
+      "loss": 1.098,
+      "step": 17586
+    },
+    {
+      "epoch": 2.81392,
+      "grad_norm": 1.691876653232597,
+      "learning_rate": 2.0912092418611364e-06,
+      "loss": 1.0536,
+      "step": 17587
+    },
+    {
+      "epoch": 2.81408,
+      "grad_norm": 1.656494045582602,
+      "learning_rate": 2.0876965542404925e-06,
+      "loss": 1.0329,
+      "step": 17588
+    },
+    {
+      "epoch": 2.81424,
+      "grad_norm": 1.7417781618341075,
+      "learning_rate": 2.084186788180298e-06,
+      "loss": 1.1186,
+      "step": 17589
+    },
+    {
+      "epoch": 2.8144,
+      "grad_norm": 1.8095535805536804,
+      "learning_rate": 2.080679943785313e-06,
+      "loss": 1.0244,
+      "step": 17590
+    },
+    {
+      "epoch": 2.81456,
+      "grad_norm": 1.7147015602798228,
+      "learning_rate": 2.077176021160132e-06,
+      "loss": 1.0478,
+      "step": 17591
+    },
+    {
+      "epoch": 2.81472,
+      "grad_norm": 2.0602234128952204,
+      "learning_rate": 2.073675020409349e-06,
+      "loss": 1.0636,
+      "step": 17592
+    },
+    {
+      "epoch": 2.81488,
+      "grad_norm": 1.8223556641919985,
+      "learning_rate": 2.0701769416373917e-06,
+      "loss": 1.1383,
+      "step": 17593
+    },
+    {
+      "epoch": 2.8150399999999998,
+      "grad_norm": 1.461735307056323,
+      "learning_rate": 2.066681784948665e-06,
+      "loss": 1.0303,
+      "step": 17594
+    },
+    {
+      "epoch": 2.8152,
+      "grad_norm": 1.8551893691098684,
+      "learning_rate": 2.063189550447453e-06,
+      "loss": 1.0734,
+      "step": 17595
+    },
+    {
+      "epoch": 2.81536,
+      "grad_norm": 1.848748757066335,
+      "learning_rate": 2.0597002382379495e-06,
+      "loss": 1.0668,
+      "step": 17596
+    },
+    {
+      "epoch": 2.8155200000000002,
+      "grad_norm": 1.6246217494708095,
+      "learning_rate": 2.056213848424282e-06,
+      "loss": 1.0472,
+      "step": 17597
+    },
+    {
+      "epoch": 2.81568,
+      "grad_norm": 1.6855794612453359,
+      "learning_rate": 2.052730381110468e-06,
+      "loss": 1.0197,
+      "step": 17598
+    },
+    {
+      "epoch": 2.81584,
+      "grad_norm": 1.688304951558564,
+      "learning_rate": 2.0492498364004684e-06,
+      "loss": 1.0888,
+      "step": 17599
+    },
+    {
+      "epoch": 2.816,
+      "grad_norm": 1.8030676716302287,
+      "learning_rate": 2.0457722143981116e-06,
+      "loss": 1.0573,
+      "step": 17600
+    },
+    {
+      "epoch": 2.81616,
+      "grad_norm": 2.006403160374799,
+      "learning_rate": 2.042297515207181e-06,
+      "loss": 1.0655,
+      "step": 17601
+    },
+    {
+      "epoch": 2.81632,
+      "grad_norm": 1.775757670877627,
+      "learning_rate": 2.0388257389313494e-06,
+      "loss": 1.0773,
+      "step": 17602
+    },
+    {
+      "epoch": 2.81648,
+      "grad_norm": 2.0052585170231074,
+      "learning_rate": 2.0353568856742224e-06,
+      "loss": 1.0633,
+      "step": 17603
+    },
+    {
+      "epoch": 2.81664,
+      "grad_norm": 1.8190891892759162,
+      "learning_rate": 2.031890955539284e-06,
+      "loss": 1.064,
+      "step": 17604
+    },
+    {
+      "epoch": 2.8167999999999997,
+      "grad_norm": 2.2953547801410217,
+      "learning_rate": 2.028427948629963e-06,
+      "loss": 1.0813,
+      "step": 17605
+    },
+    {
+      "epoch": 2.81696,
+      "grad_norm": 1.5350388206806138,
+      "learning_rate": 2.0249678650496098e-06,
+      "loss": 1.0139,
+      "step": 17606
+    },
+    {
+      "epoch": 2.81712,
+      "grad_norm": 1.6005017565220057,
+      "learning_rate": 2.0215107049014416e-06,
+      "loss": 1.0277,
+      "step": 17607
+    },
+    {
+      "epoch": 2.8172800000000002,
+      "grad_norm": 1.7631413748669382,
+      "learning_rate": 2.0180564682886204e-06,
+      "loss": 1.0208,
+      "step": 17608
+    },
+    {
+      "epoch": 2.81744,
+      "grad_norm": 1.7598701809646111,
+      "learning_rate": 2.014605155314231e-06,
+      "loss": 1.0654,
+      "step": 17609
+    },
+    {
+      "epoch": 2.8176,
+      "grad_norm": 1.914332671604063,
+      "learning_rate": 2.0111567660812347e-06,
+      "loss": 1.0666,
+      "step": 17610
+    },
+    {
+      "epoch": 2.81776,
+      "grad_norm": 1.6591780547783508,
+      "learning_rate": 2.0077113006925386e-06,
+      "loss": 1.0523,
+      "step": 17611
+    },
+    {
+      "epoch": 2.81792,
+      "grad_norm": 1.689801831883471,
+      "learning_rate": 2.004268759250938e-06,
+      "loss": 1.0735,
+      "step": 17612
+    },
+    {
+      "epoch": 2.81808,
+      "grad_norm": 1.7232712583775516,
+      "learning_rate": 2.0008291418591727e-06,
+      "loss": 1.1035,
+      "step": 17613
+    },
+    {
+      "epoch": 2.81824,
+      "grad_norm": 1.755182401963464,
+      "learning_rate": 1.9973924486198615e-06,
+      "loss": 1.0367,
+      "step": 17614
+    },
+    {
+      "epoch": 2.8184,
+      "grad_norm": 1.7536336031866822,
+      "learning_rate": 1.9939586796355657e-06,
+      "loss": 1.0515,
+      "step": 17615
+    },
+    {
+      "epoch": 2.8185599999999997,
+      "grad_norm": 1.7187787552625713,
+      "learning_rate": 1.9905278350087152e-06,
+      "loss": 1.0538,
+      "step": 17616
+    },
+    {
+      "epoch": 2.81872,
+      "grad_norm": 1.8140514041585896,
+      "learning_rate": 1.987099914841717e-06,
+      "loss": 1.0403,
+      "step": 17617
+    },
+    {
+      "epoch": 2.81888,
+      "grad_norm": 1.6968502587493641,
+      "learning_rate": 1.983674919236822e-06,
+      "loss": 1.0555,
+      "step": 17618
+    },
+    {
+      "epoch": 2.81904,
+      "grad_norm": 1.716061288776023,
+      "learning_rate": 1.980252848296249e-06,
+      "loss": 1.075,
+      "step": 17619
+    },
+    {
+      "epoch": 2.8192,
+      "grad_norm": 1.6947788589911563,
+      "learning_rate": 1.9768337021221052e-06,
+      "loss": 1.0829,
+      "step": 17620
+    },
+    {
+      "epoch": 2.81936,
+      "grad_norm": 2.734120503251936,
+      "learning_rate": 1.9734174808164084e-06,
+      "loss": 1.1654,
+      "step": 17621
+    },
+    {
+      "epoch": 2.81952,
+      "grad_norm": 1.6288966968268248,
+      "learning_rate": 1.970004184481089e-06,
+      "loss": 1.0039,
+      "step": 17622
+    },
+    {
+      "epoch": 2.81968,
+      "grad_norm": 1.4902964390164775,
+      "learning_rate": 1.9665938132179986e-06,
+      "loss": 1.0786,
+      "step": 17623
+    },
+    {
+      "epoch": 2.81984,
+      "grad_norm": 1.8132220680884803,
+      "learning_rate": 1.9631863671289107e-06,
+      "loss": 1.0479,
+      "step": 17624
+    },
+    {
+      "epoch": 2.82,
+      "grad_norm": 1.782807517837788,
+      "learning_rate": 1.959781846315478e-06,
+      "loss": 1.0365,
+      "step": 17625
+    },
+    {
+      "epoch": 2.82016,
+      "grad_norm": 1.8310440665493335,
+      "learning_rate": 1.956380250879297e-06,
+      "loss": 1.0545,
+      "step": 17626
+    },
+    {
+      "epoch": 2.82032,
+      "grad_norm": 1.6799123680485704,
+      "learning_rate": 1.9529815809218644e-06,
+      "loss": 1.0916,
+      "step": 17627
+    },
+    {
+      "epoch": 2.82048,
+      "grad_norm": 1.8728254773677682,
+      "learning_rate": 1.9495858365445874e-06,
+      "loss": 1.0748,
+      "step": 17628
+    },
+    {
+      "epoch": 2.82064,
+      "grad_norm": 1.5788914311521887,
+      "learning_rate": 1.9461930178487965e-06,
+      "loss": 1.0608,
+      "step": 17629
+    },
+    {
+      "epoch": 2.8208,
+      "grad_norm": 1.6672134577165598,
+      "learning_rate": 1.9428031249357213e-06,
+      "loss": 1.1079,
+      "step": 17630
+    },
+    {
+      "epoch": 2.82096,
+      "grad_norm": 1.8185545114289527,
+      "learning_rate": 1.9394161579065375e-06,
+      "loss": 1.0732,
+      "step": 17631
+    },
+    {
+      "epoch": 2.82112,
+      "grad_norm": 1.8977988888204178,
+      "learning_rate": 1.9360321168622632e-06,
+      "loss": 1.0523,
+      "step": 17632
+    },
+    {
+      "epoch": 2.82128,
+      "grad_norm": 1.6052899634180355,
+      "learning_rate": 1.9326510019039067e-06,
+      "loss": 1.0605,
+      "step": 17633
+    },
+    {
+      "epoch": 2.82144,
+      "grad_norm": 2.0163202361218766,
+      "learning_rate": 1.9292728131323432e-06,
+      "loss": 1.0756,
+      "step": 17634
+    },
+    {
+      "epoch": 2.8216,
+      "grad_norm": 1.7280094695904942,
+      "learning_rate": 1.92589755064837e-06,
+      "loss": 1.0337,
+      "step": 17635
+    },
+    {
+      "epoch": 2.8217600000000003,
+      "grad_norm": 1.5233118379003978,
+      "learning_rate": 1.9225252145527174e-06,
+      "loss": 1.0168,
+      "step": 17636
+    },
+    {
+      "epoch": 2.82192,
+      "grad_norm": 1.8358092361048688,
+      "learning_rate": 1.9191558049459824e-06,
+      "loss": 1.0861,
+      "step": 17637
+    },
+    {
+      "epoch": 2.82208,
+      "grad_norm": 1.6511332972468684,
+      "learning_rate": 1.9157893219287293e-06,
+      "loss": 1.0563,
+      "step": 17638
+    },
+    {
+      "epoch": 2.82224,
+      "grad_norm": 1.871611950451005,
+      "learning_rate": 1.9124257656014e-06,
+      "loss": 1.0911,
+      "step": 17639
+    },
+    {
+      "epoch": 2.8224,
+      "grad_norm": 1.753091574499871,
+      "learning_rate": 1.9090651360643586e-06,
+      "loss": 1.0895,
+      "step": 17640
+    },
+    {
+      "epoch": 2.82256,
+      "grad_norm": 1.5798902969573458,
+      "learning_rate": 1.9057074334178693e-06,
+      "loss": 1.0893,
+      "step": 17641
+    },
+    {
+      "epoch": 2.82272,
+      "grad_norm": 1.7864602646803538,
+      "learning_rate": 1.9023526577621409e-06,
+      "loss": 1.0706,
+      "step": 17642
+    },
+    {
+      "epoch": 2.82288,
+      "grad_norm": 1.793721541920493,
+      "learning_rate": 1.899000809197271e-06,
+      "loss": 1.0336,
+      "step": 17643
+    },
+    {
+      "epoch": 2.8230399999999998,
+      "grad_norm": 1.7427164739338938,
+      "learning_rate": 1.8956518878232576e-06,
+      "loss": 1.038,
+      "step": 17644
+    },
+    {
+      "epoch": 2.8232,
+      "grad_norm": 1.8545887484340597,
+      "learning_rate": 1.892305893740043e-06,
+      "loss": 1.0896,
+      "step": 17645
+    },
+    {
+      "epoch": 2.82336,
+      "grad_norm": 1.7458351096910238,
+      "learning_rate": 1.888962827047458e-06,
+      "loss": 1.0674,
+      "step": 17646
+    },
+    {
+      "epoch": 2.8235200000000003,
+      "grad_norm": 1.6614077545316144,
+      "learning_rate": 1.8856226878452675e-06,
+      "loss": 1.0706,
+      "step": 17647
+    },
+    {
+      "epoch": 2.82368,
+      "grad_norm": 1.7091199288593446,
+      "learning_rate": 1.8822854762331143e-06,
+      "loss": 1.0565,
+      "step": 17648
+    },
+    {
+      "epoch": 2.82384,
+      "grad_norm": 1.883251495035687,
+      "learning_rate": 1.8789511923106073e-06,
+      "loss": 1.0807,
+      "step": 17649
+    },
+    {
+      "epoch": 2.824,
+      "grad_norm": 1.784070960433627,
+      "learning_rate": 1.8756198361772004e-06,
+      "loss": 1.0726,
+      "step": 17650
+    },
+    {
+      "epoch": 2.82416,
+      "grad_norm": 1.8188807817777295,
+      "learning_rate": 1.872291407932336e-06,
+      "loss": 1.0385,
+      "step": 17651
+    },
+    {
+      "epoch": 2.82432,
+      "grad_norm": 1.5511586840664036,
+      "learning_rate": 1.8689659076752908e-06,
+      "loss": 1.0419,
+      "step": 17652
+    },
+    {
+      "epoch": 2.82448,
+      "grad_norm": 1.698397556779928,
+      "learning_rate": 1.8656433355053182e-06,
+      "loss": 1.0923,
+      "step": 17653
+    },
+    {
+      "epoch": 2.82464,
+      "grad_norm": 1.9712422707831507,
+      "learning_rate": 1.8623236915215614e-06,
+      "loss": 1.0363,
+      "step": 17654
+    },
+    {
+      "epoch": 2.8247999999999998,
+      "grad_norm": 1.743545651092612,
+      "learning_rate": 1.859006975823052e-06,
+      "loss": 1.0814,
+      "step": 17655
+    },
+    {
+      "epoch": 2.82496,
+      "grad_norm": 1.7147399125779923,
+      "learning_rate": 1.8556931885087669e-06,
+      "loss": 1.0681,
+      "step": 17656
+    },
+    {
+      "epoch": 2.82512,
+      "grad_norm": 1.642125251682061,
+      "learning_rate": 1.8523823296775933e-06,
+      "loss": 1.0155,
+      "step": 17657
+    },
+    {
+      "epoch": 2.8252800000000002,
+      "grad_norm": 1.9466439985697603,
+      "learning_rate": 1.8490743994283077e-06,
+      "loss": 1.078,
+      "step": 17658
+    },
+    {
+      "epoch": 2.82544,
+      "grad_norm": 1.4153783957593058,
+      "learning_rate": 1.8457693978596313e-06,
+      "loss": 1.0703,
+      "step": 17659
+    },
+    {
+      "epoch": 2.8256,
+      "grad_norm": 1.7619298558470329,
+      "learning_rate": 1.842467325070163e-06,
+      "loss": 1.0953,
+      "step": 17660
+    },
+    {
+      "epoch": 2.82576,
+      "grad_norm": 1.592547378498794,
+      "learning_rate": 1.839168181158435e-06,
+      "loss": 1.106,
+      "step": 17661
+    },
+    {
+      "epoch": 2.82592,
+      "grad_norm": 1.872378812424889,
+      "learning_rate": 1.8358719662228906e-06,
+      "loss": 1.0626,
+      "step": 17662
+    },
+    {
+      "epoch": 2.82608,
+      "grad_norm": 1.703102079082968,
+      "learning_rate": 1.832578680361896e-06,
+      "loss": 1.0396,
+      "step": 17663
+    },
+    {
+      "epoch": 2.82624,
+      "grad_norm": 1.795975213204685,
+      "learning_rate": 1.8292883236737058e-06,
+      "loss": 1.0877,
+      "step": 17664
+    },
+    {
+      "epoch": 2.8264,
+      "grad_norm": 1.9091875073923394,
+      "learning_rate": 1.8260008962565079e-06,
+      "loss": 1.0647,
+      "step": 17665
+    },
+    {
+      "epoch": 2.8265599999999997,
+      "grad_norm": 2.0329952524363453,
+      "learning_rate": 1.8227163982083795e-06,
+      "loss": 1.0882,
+      "step": 17666
+    },
+    {
+      "epoch": 2.82672,
+      "grad_norm": 1.6505135107731375,
+      "learning_rate": 1.8194348296273422e-06,
+      "loss": 1.0681,
+      "step": 17667
+    },
+    {
+      "epoch": 2.82688,
+      "grad_norm": 1.919885081994042,
+      "learning_rate": 1.8161561906112956e-06,
+      "loss": 1.0974,
+      "step": 17668
+    },
+    {
+      "epoch": 2.82704,
+      "grad_norm": 1.6047385050668752,
+      "learning_rate": 1.8128804812580836e-06,
+      "loss": 1.0695,
+      "step": 17669
+    },
+    {
+      "epoch": 2.8272,
+      "grad_norm": 1.542270293401668,
+      "learning_rate": 1.80960770166545e-06,
+      "loss": 1.0598,
+      "step": 17670
+    },
+    {
+      "epoch": 2.82736,
+      "grad_norm": 1.6800386194050076,
+      "learning_rate": 1.806337851931028e-06,
+      "loss": 1.0889,
+      "step": 17671
+    },
+    {
+      "epoch": 2.82752,
+      "grad_norm": 1.7798482182614401,
+      "learning_rate": 1.8030709321524286e-06,
+      "loss": 1.0835,
+      "step": 17672
+    },
+    {
+      "epoch": 2.82768,
+      "grad_norm": 1.8510062655741237,
+      "learning_rate": 1.7998069424270737e-06,
+      "loss": 1.0891,
+      "step": 17673
+    },
+    {
+      "epoch": 2.82784,
+      "grad_norm": 1.763670459768661,
+      "learning_rate": 1.7965458828524072e-06,
+      "loss": 1.0322,
+      "step": 17674
+    },
+    {
+      "epoch": 2.828,
+      "grad_norm": 1.9081031549520509,
+      "learning_rate": 1.7932877535257076e-06,
+      "loss": 1.0201,
+      "step": 17675
+    },
+    {
+      "epoch": 2.82816,
+      "grad_norm": 1.765355082288134,
+      "learning_rate": 1.7900325545442076e-06,
+      "loss": 1.0307,
+      "step": 17676
+    },
+    {
+      "epoch": 2.8283199999999997,
+      "grad_norm": 1.8539298430522724,
+      "learning_rate": 1.7867802860050075e-06,
+      "loss": 1.0561,
+      "step": 17677
+    },
+    {
+      "epoch": 2.82848,
+      "grad_norm": 1.7906412039863784,
+      "learning_rate": 1.7835309480051965e-06,
+      "loss": 1.0488,
+      "step": 17678
+    },
+    {
+      "epoch": 2.82864,
+      "grad_norm": 1.5023708958068382,
+      "learning_rate": 1.7802845406416856e-06,
+      "loss": 1.0163,
+      "step": 17679
+    },
+    {
+      "epoch": 2.8288,
+      "grad_norm": 1.7687540616278463,
+      "learning_rate": 1.7770410640113754e-06,
+      "loss": 1.0522,
+      "step": 17680
+    },
+    {
+      "epoch": 2.82896,
+      "grad_norm": 1.7781550100943606,
+      "learning_rate": 1.7738005182110328e-06,
+      "loss": 1.0827,
+      "step": 17681
+    },
+    {
+      "epoch": 2.82912,
+      "grad_norm": 1.7086432734620192,
+      "learning_rate": 1.7705629033373472e-06,
+      "loss": 1.0745,
+      "step": 17682
+    },
+    {
+      "epoch": 2.82928,
+      "grad_norm": 1.7181875732743963,
+      "learning_rate": 1.7673282194869412e-06,
+      "loss": 1.0857,
+      "step": 17683
+    },
+    {
+      "epoch": 2.82944,
+      "grad_norm": 1.700791963905602,
+      "learning_rate": 1.7640964667563041e-06,
+      "loss": 1.0307,
+      "step": 17684
+    },
+    {
+      "epoch": 2.8296,
+      "grad_norm": 1.7139909845082253,
+      "learning_rate": 1.7608676452418927e-06,
+      "loss": 1.0661,
+      "step": 17685
+    },
+    {
+      "epoch": 2.8297600000000003,
+      "grad_norm": 1.6656074554111084,
+      "learning_rate": 1.7576417550400404e-06,
+      "loss": 1.0603,
+      "step": 17686
+    },
+    {
+      "epoch": 2.82992,
+      "grad_norm": 1.7494410957060982,
+      "learning_rate": 1.7544187962469928e-06,
+      "loss": 1.0471,
+      "step": 17687
+    },
+    {
+      "epoch": 2.83008,
+      "grad_norm": 1.7568598842400234,
+      "learning_rate": 1.7511987689589504e-06,
+      "loss": 1.0645,
+      "step": 17688
+    },
+    {
+      "epoch": 2.83024,
+      "grad_norm": 1.7991456667744965,
+      "learning_rate": 1.747981673271959e-06,
+      "loss": 1.0614,
+      "step": 17689
+    },
+    {
+      "epoch": 2.8304,
+      "grad_norm": 1.7640513513630667,
+      "learning_rate": 1.7447675092820304e-06,
+      "loss": 1.0708,
+      "step": 17690
+    },
+    {
+      "epoch": 2.83056,
+      "grad_norm": 1.7486338131603723,
+      "learning_rate": 1.7415562770850657e-06,
+      "loss": 1.0413,
+      "step": 17691
+    },
+    {
+      "epoch": 2.83072,
+      "grad_norm": 1.7503764243811941,
+      "learning_rate": 1.7383479767768884e-06,
+      "loss": 1.1022,
+      "step": 17692
+    },
+    {
+      "epoch": 2.83088,
+      "grad_norm": 1.7293881351566032,
+      "learning_rate": 1.7351426084532108e-06,
+      "loss": 1.1118,
+      "step": 17693
+    },
+    {
+      "epoch": 2.83104,
+      "grad_norm": 1.9528718761392947,
+      "learning_rate": 1.7319401722097007e-06,
+      "loss": 1.1064,
+      "step": 17694
+    },
+    {
+      "epoch": 2.8312,
+      "grad_norm": 1.6696123427639442,
+      "learning_rate": 1.7287406681419038e-06,
+      "loss": 1.1019,
+      "step": 17695
+    },
+    {
+      "epoch": 2.83136,
+      "grad_norm": 1.7055386949672031,
+      "learning_rate": 1.7255440963452884e-06,
+      "loss": 1.0036,
+      "step": 17696
+    },
+    {
+      "epoch": 2.8315200000000003,
+      "grad_norm": 1.5892400399349191,
+      "learning_rate": 1.7223504569152448e-06,
+      "loss": 1.0792,
+      "step": 17697
+    },
+    {
+      "epoch": 2.83168,
+      "grad_norm": 1.5664761516361734,
+      "learning_rate": 1.7191597499470413e-06,
+      "loss": 1.0532,
+      "step": 17698
+    },
+    {
+      "epoch": 2.83184,
+      "grad_norm": 1.7289291568663587,
+      "learning_rate": 1.7159719755359127e-06,
+      "loss": 1.0738,
+      "step": 17699
+    },
+    {
+      "epoch": 2.832,
+      "grad_norm": 1.9472989722577492,
+      "learning_rate": 1.7127871337769608e-06,
+      "loss": 1.0706,
+      "step": 17700
+    },
+    {
+      "epoch": 2.83216,
+      "grad_norm": 1.6804318558810087,
+      "learning_rate": 1.7096052247652206e-06,
+      "loss": 1.0838,
+      "step": 17701
+    },
+    {
+      "epoch": 2.83232,
+      "grad_norm": 1.8366522378254302,
+      "learning_rate": 1.7064262485956384e-06,
+      "loss": 1.0845,
+      "step": 17702
+    },
+    {
+      "epoch": 2.83248,
+      "grad_norm": 1.6019494057018207,
+      "learning_rate": 1.7032502053630605e-06,
+      "loss": 1.0735,
+      "step": 17703
+    },
+    {
+      "epoch": 2.83264,
+      "grad_norm": 1.6672621558028138,
+      "learning_rate": 1.700077095162278e-06,
+      "loss": 1.1016,
+      "step": 17704
+    },
+    {
+      "epoch": 2.8327999999999998,
+      "grad_norm": 1.6918880744462708,
+      "learning_rate": 1.6969069180879482e-06,
+      "loss": 1.0658,
+      "step": 17705
+    },
+    {
+      "epoch": 2.83296,
+      "grad_norm": 1.7642725401109778,
+      "learning_rate": 1.6937396742346846e-06,
+      "loss": 1.0965,
+      "step": 17706
+    },
+    {
+      "epoch": 2.83312,
+      "grad_norm": 2.0284089739922937,
+      "learning_rate": 1.6905753636969557e-06,
+      "loss": 1.1101,
+      "step": 17707
+    },
+    {
+      "epoch": 2.8332800000000002,
+      "grad_norm": 1.8893496840981998,
+      "learning_rate": 1.6874139865692308e-06,
+      "loss": 1.0693,
+      "step": 17708
+    },
+    {
+      "epoch": 2.83344,
+      "grad_norm": 1.7351210498234373,
+      "learning_rate": 1.6842555429458006e-06,
+      "loss": 1.0529,
+      "step": 17709
+    },
+    {
+      "epoch": 2.8336,
+      "grad_norm": 1.6506403487661478,
+      "learning_rate": 1.6811000329209347e-06,
+      "loss": 1.0238,
+      "step": 17710
+    },
+    {
+      "epoch": 2.83376,
+      "grad_norm": 1.5643381321684056,
+      "learning_rate": 1.6779474565887686e-06,
+      "loss": 1.0468,
+      "step": 17711
+    },
+    {
+      "epoch": 2.83392,
+      "grad_norm": 1.7058957675816389,
+      "learning_rate": 1.6747978140433829e-06,
+      "loss": 1.122,
+      "step": 17712
+    },
+    {
+      "epoch": 2.83408,
+      "grad_norm": 1.7401828447255405,
+      "learning_rate": 1.6716511053787576e-06,
+      "loss": 1.0732,
+      "step": 17713
+    },
+    {
+      "epoch": 2.83424,
+      "grad_norm": 1.8226781185022436,
+      "learning_rate": 1.6685073306887844e-06,
+      "loss": 1.0752,
+      "step": 17714
+    },
+    {
+      "epoch": 2.8344,
+      "grad_norm": 1.693948243798983,
+      "learning_rate": 1.665366490067266e-06,
+      "loss": 1.0749,
+      "step": 17715
+    },
+    {
+      "epoch": 2.8345599999999997,
+      "grad_norm": 1.8130392306743486,
+      "learning_rate": 1.6622285836079276e-06,
+      "loss": 1.0457,
+      "step": 17716
+    },
+    {
+      "epoch": 2.83472,
+      "grad_norm": 1.6694589866871232,
+      "learning_rate": 1.6590936114043942e-06,
+      "loss": 1.0346,
+      "step": 17717
+    },
+    {
+      "epoch": 2.83488,
+      "grad_norm": 1.6162219392419688,
+      "learning_rate": 1.655961573550202e-06,
+      "loss": 1.0651,
+      "step": 17718
+    },
+    {
+      "epoch": 2.8350400000000002,
+      "grad_norm": 1.788699913328752,
+      "learning_rate": 1.6528324701388098e-06,
+      "loss": 1.0513,
+      "step": 17719
+    },
+    {
+      "epoch": 2.8352,
+      "grad_norm": 1.5796660401308111,
+      "learning_rate": 1.6497063012636093e-06,
+      "loss": 1.0195,
+      "step": 17720
+    },
+    {
+      "epoch": 2.83536,
+      "grad_norm": 1.772973845909865,
+      "learning_rate": 1.6465830670178484e-06,
+      "loss": 1.0402,
+      "step": 17721
+    },
+    {
+      "epoch": 2.83552,
+      "grad_norm": 1.8453935789343876,
+      "learning_rate": 1.64346276749473e-06,
+      "loss": 1.0392,
+      "step": 17722
+    },
+    {
+      "epoch": 2.83568,
+      "grad_norm": 1.7966791384723904,
+      "learning_rate": 1.6403454027873687e-06,
+      "loss": 1.0821,
+      "step": 17723
+    },
+    {
+      "epoch": 2.83584,
+      "grad_norm": 1.7122399753027622,
+      "learning_rate": 1.6372309729887791e-06,
+      "loss": 1.0597,
+      "step": 17724
+    },
+    {
+      "epoch": 2.836,
+      "grad_norm": 1.846292755712143,
+      "learning_rate": 1.6341194781918757e-06,
+      "loss": 1.1007,
+      "step": 17725
+    },
+    {
+      "epoch": 2.83616,
+      "grad_norm": 2.14753665471828,
+      "learning_rate": 1.6310109184895172e-06,
+      "loss": 1.0988,
+      "step": 17726
+    },
+    {
+      "epoch": 2.8363199999999997,
+      "grad_norm": 1.6578280222510142,
+      "learning_rate": 1.6279052939744633e-06,
+      "loss": 0.9812,
+      "step": 17727
+    },
+    {
+      "epoch": 2.83648,
+      "grad_norm": 1.9089994924401565,
+      "learning_rate": 1.6248026047393505e-06,
+      "loss": 1.1138,
+      "step": 17728
+    },
+    {
+      "epoch": 2.83664,
+      "grad_norm": 1.5539984931297464,
+      "learning_rate": 1.6217028508768051e-06,
+      "loss": 1.0696,
+      "step": 17729
+    },
+    {
+      "epoch": 2.8368,
+      "grad_norm": 1.7337357538642089,
+      "learning_rate": 1.618606032479275e-06,
+      "loss": 1.0744,
+      "step": 17730
+    },
+    {
+      "epoch": 2.83696,
+      "grad_norm": 2.017003433008788,
+      "learning_rate": 1.615512149639209e-06,
+      "loss": 1.0608,
+      "step": 17731
+    },
+    {
+      "epoch": 2.83712,
+      "grad_norm": 1.8384755439383085,
+      "learning_rate": 1.6124212024488771e-06,
+      "loss": 1.0873,
+      "step": 17732
+    },
+    {
+      "epoch": 2.83728,
+      "grad_norm": 1.751318160547042,
+      "learning_rate": 1.6093331910005394e-06,
+      "loss": 1.0666,
+      "step": 17733
+    },
+    {
+      "epoch": 2.83744,
+      "grad_norm": 1.6361112503369364,
+      "learning_rate": 1.606248115386322e-06,
+      "loss": 1.0709,
+      "step": 17734
+    },
+    {
+      "epoch": 2.8376,
+      "grad_norm": 1.5902235132748133,
+      "learning_rate": 1.603165975698284e-06,
+      "loss": 1.0429,
+      "step": 17735
+    },
+    {
+      "epoch": 2.83776,
+      "grad_norm": 1.703934061561271,
+      "learning_rate": 1.6000867720284084e-06,
+      "loss": 1.0852,
+      "step": 17736
+    },
+    {
+      "epoch": 2.83792,
+      "grad_norm": 1.6424586334070057,
+      "learning_rate": 1.5970105044685436e-06,
+      "loss": 1.065,
+      "step": 17737
+    },
+    {
+      "epoch": 2.83808,
+      "grad_norm": 1.6830916048721314,
+      "learning_rate": 1.5939371731105045e-06,
+      "loss": 1.042,
+      "step": 17738
+    },
+    {
+      "epoch": 2.83824,
+      "grad_norm": 1.8735919720442866,
+      "learning_rate": 1.5908667780459963e-06,
+      "loss": 1.0553,
+      "step": 17739
+    },
+    {
+      "epoch": 2.8384,
+      "grad_norm": 1.8112438155591488,
+      "learning_rate": 1.5877993193666118e-06,
+      "loss": 1.0676,
+      "step": 17740
+    },
+    {
+      "epoch": 2.83856,
+      "grad_norm": 1.5777943958160479,
+      "learning_rate": 1.5847347971639004e-06,
+      "loss": 1.0505,
+      "step": 17741
+    },
+    {
+      "epoch": 2.83872,
+      "grad_norm": 1.671344506035117,
+      "learning_rate": 1.5816732115292998e-06,
+      "loss": 1.0814,
+      "step": 17742
+    },
+    {
+      "epoch": 2.83888,
+      "grad_norm": 1.7313973491580155,
+      "learning_rate": 1.5786145625541593e-06,
+      "loss": 1.0642,
+      "step": 17743
+    },
+    {
+      "epoch": 2.83904,
+      "grad_norm": 1.581438440656593,
+      "learning_rate": 1.5755588503297281e-06,
+      "loss": 1.0421,
+      "step": 17744
+    },
+    {
+      "epoch": 2.8392,
+      "grad_norm": 1.9909902757553066,
+      "learning_rate": 1.5725060749472221e-06,
+      "loss": 1.0563,
+      "step": 17745
+    },
+    {
+      "epoch": 2.83936,
+      "grad_norm": 1.7064470981681286,
+      "learning_rate": 1.5694562364977018e-06,
+      "loss": 1.0618,
+      "step": 17746
+    },
+    {
+      "epoch": 2.8395200000000003,
+      "grad_norm": 1.8685435155941148,
+      "learning_rate": 1.5664093350721831e-06,
+      "loss": 1.0625,
+      "step": 17747
+    },
+    {
+      "epoch": 2.83968,
+      "grad_norm": 1.787969354163663,
+      "learning_rate": 1.5633653707615714e-06,
+      "loss": 1.0516,
+      "step": 17748
+    },
+    {
+      "epoch": 2.83984,
+      "grad_norm": 1.7307389410261365,
+      "learning_rate": 1.560324343656705e-06,
+      "loss": 1.0839,
+      "step": 17749
+    },
+    {
+      "epoch": 2.84,
+      "grad_norm": 1.8815998642616905,
+      "learning_rate": 1.5572862538483112e-06,
+      "loss": 1.0553,
+      "step": 17750
+    },
+    {
+      "epoch": 2.84016,
+      "grad_norm": 1.5483710249687315,
+      "learning_rate": 1.5542511014270512e-06,
+      "loss": 1.0669,
+      "step": 17751
+    },
+    {
+      "epoch": 2.84032,
+      "grad_norm": 1.881239140544155,
+      "learning_rate": 1.5512188864834965e-06,
+      "loss": 1.0157,
+      "step": 17752
+    },
+    {
+      "epoch": 2.84048,
+      "grad_norm": 1.8570856772602597,
+      "learning_rate": 1.5481896091081083e-06,
+      "loss": 1.0945,
+      "step": 17753
+    },
+    {
+      "epoch": 2.84064,
+      "grad_norm": 1.7311869763456844,
+      "learning_rate": 1.545163269391281e-06,
+      "loss": 1.0737,
+      "step": 17754
+    },
+    {
+      "epoch": 2.8407999999999998,
+      "grad_norm": 1.7225202125447083,
+      "learning_rate": 1.54213986742332e-06,
+      "loss": 1.0534,
+      "step": 17755
+    },
+    {
+      "epoch": 2.84096,
+      "grad_norm": 1.6972780457325851,
+      "learning_rate": 1.5391194032944422e-06,
+      "loss": 1.0677,
+      "step": 17756
+    },
+    {
+      "epoch": 2.84112,
+      "grad_norm": 1.7316130493436444,
+      "learning_rate": 1.5361018770947645e-06,
+      "loss": 1.0415,
+      "step": 17757
+    },
+    {
+      "epoch": 2.8412800000000002,
+      "grad_norm": 1.7485516682939792,
+      "learning_rate": 1.5330872889143365e-06,
+      "loss": 1.0605,
+      "step": 17758
+    },
+    {
+      "epoch": 2.84144,
+      "grad_norm": 1.8181747335599838,
+      "learning_rate": 1.530075638843087e-06,
+      "loss": 1.0083,
+      "step": 17759
+    },
+    {
+      "epoch": 2.8416,
+      "grad_norm": 1.661824513636878,
+      "learning_rate": 1.5270669269708993e-06,
+      "loss": 1.0086,
+      "step": 17760
+    },
+    {
+      "epoch": 2.84176,
+      "grad_norm": 1.676812054079814,
+      "learning_rate": 1.5240611533875571e-06,
+      "loss": 1.0699,
+      "step": 17761
+    },
+    {
+      "epoch": 2.84192,
+      "grad_norm": 1.696400411482675,
+      "learning_rate": 1.521058318182722e-06,
+      "loss": 1.0654,
+      "step": 17762
+    },
+    {
+      "epoch": 2.84208,
+      "grad_norm": 1.5458706045301593,
+      "learning_rate": 1.5180584214460225e-06,
+      "loss": 1.0367,
+      "step": 17763
+    },
+    {
+      "epoch": 2.84224,
+      "grad_norm": 1.845563368317059,
+      "learning_rate": 1.5150614632669314e-06,
+      "loss": 1.0478,
+      "step": 17764
+    },
+    {
+      "epoch": 2.8424,
+      "grad_norm": 1.617120115527207,
+      "learning_rate": 1.5120674437349214e-06,
+      "loss": 1.0859,
+      "step": 17765
+    },
+    {
+      "epoch": 2.8425599999999998,
+      "grad_norm": 1.6280956480219304,
+      "learning_rate": 1.509076362939299e-06,
+      "loss": 1.0776,
+      "step": 17766
+    },
+    {
+      "epoch": 2.84272,
+      "grad_norm": 1.6392001228920583,
+      "learning_rate": 1.506088220969315e-06,
+      "loss": 0.9902,
+      "step": 17767
+    },
+    {
+      "epoch": 2.84288,
+      "grad_norm": 1.614702784807406,
+      "learning_rate": 1.5031030179141426e-06,
+      "loss": 1.059,
+      "step": 17768
+    },
+    {
+      "epoch": 2.8430400000000002,
+      "grad_norm": 1.707678126986815,
+      "learning_rate": 1.5001207538628547e-06,
+      "loss": 1.1103,
+      "step": 17769
+    },
+    {
+      "epoch": 2.8432,
+      "grad_norm": 1.5774082372465208,
+      "learning_rate": 1.4971414289044361e-06,
+      "loss": 1.0484,
+      "step": 17770
+    },
+    {
+      "epoch": 2.84336,
+      "grad_norm": 1.7766420505849292,
+      "learning_rate": 1.4941650431277709e-06,
+      "loss": 1.1381,
+      "step": 17771
+    },
+    {
+      "epoch": 2.84352,
+      "grad_norm": 1.864730063512103,
+      "learning_rate": 1.4911915966216993e-06,
+      "loss": 1.0267,
+      "step": 17772
+    },
+    {
+      "epoch": 2.84368,
+      "grad_norm": 1.8604924021856564,
+      "learning_rate": 1.4882210894749171e-06,
+      "loss": 1.0809,
+      "step": 17773
+    },
+    {
+      "epoch": 2.84384,
+      "grad_norm": 1.8624686905539,
+      "learning_rate": 1.4852535217760754e-06,
+      "loss": 1.0888,
+      "step": 17774
+    },
+    {
+      "epoch": 2.844,
+      "grad_norm": 1.821290049555846,
+      "learning_rate": 1.4822888936137148e-06,
+      "loss": 1.0374,
+      "step": 17775
+    },
+    {
+      "epoch": 2.84416,
+      "grad_norm": 1.6818099654448933,
+      "learning_rate": 1.4793272050762864e-06,
+      "loss": 1.0502,
+      "step": 17776
+    },
+    {
+      "epoch": 2.8443199999999997,
+      "grad_norm": 1.4738015706089524,
+      "learning_rate": 1.4763684562521974e-06,
+      "loss": 1.0911,
+      "step": 17777
+    },
+    {
+      "epoch": 2.84448,
+      "grad_norm": 1.7970004655997567,
+      "learning_rate": 1.4734126472296882e-06,
+      "loss": 1.0837,
+      "step": 17778
+    },
+    {
+      "epoch": 2.84464,
+      "grad_norm": 1.6547267514612762,
+      "learning_rate": 1.4704597780969886e-06,
+      "loss": 1.0046,
+      "step": 17779
+    },
+    {
+      "epoch": 2.8448,
+      "grad_norm": 1.766567253853582,
+      "learning_rate": 1.4675098489421946e-06,
+      "loss": 1.0996,
+      "step": 17780
+    },
+    {
+      "epoch": 2.84496,
+      "grad_norm": 1.4727880042001198,
+      "learning_rate": 1.4645628598533357e-06,
+      "loss": 1.037,
+      "step": 17781
+    },
+    {
+      "epoch": 2.84512,
+      "grad_norm": 1.654773283327991,
+      "learning_rate": 1.4616188109183304e-06,
+      "loss": 1.0488,
+      "step": 17782
+    },
+    {
+      "epoch": 2.84528,
+      "grad_norm": 1.7357325510339903,
+      "learning_rate": 1.458677702225042e-06,
+      "loss": 1.0809,
+      "step": 17783
+    },
+    {
+      "epoch": 2.84544,
+      "grad_norm": 1.898196092447233,
+      "learning_rate": 1.4557395338612113e-06,
+      "loss": 1.0433,
+      "step": 17784
+    },
+    {
+      "epoch": 2.8456,
+      "grad_norm": 1.7581577170158877,
+      "learning_rate": 1.4528043059145236e-06,
+      "loss": 1.0648,
+      "step": 17785
+    },
+    {
+      "epoch": 2.84576,
+      "grad_norm": 1.7864374150998903,
+      "learning_rate": 1.4498720184725646e-06,
+      "loss": 1.0334,
+      "step": 17786
+    },
+    {
+      "epoch": 2.84592,
+      "grad_norm": 1.58782675931826,
+      "learning_rate": 1.4469426716228085e-06,
+      "loss": 1.0233,
+      "step": 17787
+    },
+    {
+      "epoch": 2.84608,
+      "grad_norm": 1.5965551442422388,
+      "learning_rate": 1.4440162654526856e-06,
+      "loss": 1.0702,
+      "step": 17788
+    },
+    {
+      "epoch": 2.84624,
+      "grad_norm": 1.7461979958785538,
+      "learning_rate": 1.4410928000495038e-06,
+      "loss": 1.0266,
+      "step": 17789
+    },
+    {
+      "epoch": 2.8464,
+      "grad_norm": 2.0656245748411974,
+      "learning_rate": 1.4381722755005045e-06,
+      "loss": 1.0849,
+      "step": 17790
+    },
+    {
+      "epoch": 2.84656,
+      "grad_norm": 2.2392331921371453,
+      "learning_rate": 1.4352546918928067e-06,
+      "loss": 1.0977,
+      "step": 17791
+    },
+    {
+      "epoch": 2.84672,
+      "grad_norm": 1.6535405074874068,
+      "learning_rate": 1.4323400493134964e-06,
+      "loss": 1.0369,
+      "step": 17792
+    },
+    {
+      "epoch": 2.84688,
+      "grad_norm": 1.779318294865006,
+      "learning_rate": 1.4294283478495262e-06,
+      "loss": 1.0925,
+      "step": 17793
+    },
+    {
+      "epoch": 2.84704,
+      "grad_norm": 1.7356237354770199,
+      "learning_rate": 1.4265195875877935e-06,
+      "loss": 1.0823,
+      "step": 17794
+    },
+    {
+      "epoch": 2.8472,
+      "grad_norm": 1.7904433821294037,
+      "learning_rate": 1.423613768615062e-06,
+      "loss": 1.0732,
+      "step": 17795
+    },
+    {
+      "epoch": 2.84736,
+      "grad_norm": 1.9157014272795623,
+      "learning_rate": 1.4207108910180734e-06,
+      "loss": 1.0716,
+      "step": 17796
+    },
+    {
+      "epoch": 2.8475200000000003,
+      "grad_norm": 1.9307460988927623,
+      "learning_rate": 1.417810954883414e-06,
+      "loss": 1.1103,
+      "step": 17797
+    },
+    {
+      "epoch": 2.84768,
+      "grad_norm": 1.9926569704333663,
+      "learning_rate": 1.4149139602976258e-06,
+      "loss": 1.0756,
+      "step": 17798
+    },
+    {
+      "epoch": 2.84784,
+      "grad_norm": 1.959763992730222,
+      "learning_rate": 1.4120199073471618e-06,
+      "loss": 1.0535,
+      "step": 17799
+    },
+    {
+      "epoch": 2.848,
+      "grad_norm": 1.7708476141879785,
+      "learning_rate": 1.409128796118353e-06,
+      "loss": 1.095,
+      "step": 17800
+    },
+    {
+      "epoch": 2.84816,
+      "grad_norm": 1.6703463586001104,
+      "learning_rate": 1.4062406266974859e-06,
+      "loss": 1.0357,
+      "step": 17801
+    },
+    {
+      "epoch": 2.84832,
+      "grad_norm": 1.5808350352697402,
+      "learning_rate": 1.403355399170736e-06,
+      "loss": 1.0874,
+      "step": 17802
+    },
+    {
+      "epoch": 2.84848,
+      "grad_norm": 1.678321955326813,
+      "learning_rate": 1.400473113624179e-06,
+      "loss": 1.0588,
+      "step": 17803
+    },
+    {
+      "epoch": 2.84864,
+      "grad_norm": 2.1582154246261354,
+      "learning_rate": 1.3975937701438457e-06,
+      "loss": 1.1258,
+      "step": 17804
+    },
+    {
+      "epoch": 2.8487999999999998,
+      "grad_norm": 1.6653670469144328,
+      "learning_rate": 1.3947173688156123e-06,
+      "loss": 1.1045,
+      "step": 17805
+    },
+    {
+      "epoch": 2.84896,
+      "grad_norm": 1.8170813415687928,
+      "learning_rate": 1.3918439097253544e-06,
+      "loss": 1.0666,
+      "step": 17806
+    },
+    {
+      "epoch": 2.84912,
+      "grad_norm": 1.7219389198399888,
+      "learning_rate": 1.3889733929587701e-06,
+      "loss": 1.0691,
+      "step": 17807
+    },
+    {
+      "epoch": 2.8492800000000003,
+      "grad_norm": 1.7816662233637715,
+      "learning_rate": 1.3861058186015352e-06,
+      "loss": 1.0757,
+      "step": 17808
+    },
+    {
+      "epoch": 2.84944,
+      "grad_norm": 1.9327464467976678,
+      "learning_rate": 1.3832411867392036e-06,
+      "loss": 1.1043,
+      "step": 17809
+    },
+    {
+      "epoch": 2.8496,
+      "grad_norm": 1.731638848201838,
+      "learning_rate": 1.3803794974572625e-06,
+      "loss": 1.019,
+      "step": 17810
+    },
+    {
+      "epoch": 2.84976,
+      "grad_norm": 1.7804443653779947,
+      "learning_rate": 1.377520750841088e-06,
+      "loss": 1.0759,
+      "step": 17811
+    },
+    {
+      "epoch": 2.84992,
+      "grad_norm": 1.862349068731881,
+      "learning_rate": 1.3746649469759787e-06,
+      "loss": 1.0402,
+      "step": 17812
+    },
+    {
+      "epoch": 2.85008,
+      "grad_norm": 1.8293671889976657,
+      "learning_rate": 1.371812085947166e-06,
+      "loss": 1.0867,
+      "step": 17813
+    },
+    {
+      "epoch": 2.85024,
+      "grad_norm": 1.5592125693888843,
+      "learning_rate": 1.3689621678397601e-06,
+      "loss": 1.0964,
+      "step": 17814
+    },
+    {
+      "epoch": 2.8504,
+      "grad_norm": 1.548701589599262,
+      "learning_rate": 1.366115192738804e-06,
+      "loss": 1.0435,
+      "step": 17815
+    },
+    {
+      "epoch": 2.8505599999999998,
+      "grad_norm": 1.6223165584637378,
+      "learning_rate": 1.3632711607292515e-06,
+      "loss": 1.1253,
+      "step": 17816
+    },
+    {
+      "epoch": 2.85072,
+      "grad_norm": 1.8782247725071806,
+      "learning_rate": 1.3604300718959462e-06,
+      "loss": 1.0549,
+      "step": 17817
+    },
+    {
+      "epoch": 2.85088,
+      "grad_norm": 1.803329625071633,
+      "learning_rate": 1.357591926323687e-06,
+      "loss": 1.0674,
+      "step": 17818
+    },
+    {
+      "epoch": 2.8510400000000002,
+      "grad_norm": 1.9137853204154243,
+      "learning_rate": 1.354756724097128e-06,
+      "loss": 1.0704,
+      "step": 17819
+    },
+    {
+      "epoch": 2.8512,
+      "grad_norm": 1.845056378837364,
+      "learning_rate": 1.3519244653009134e-06,
+      "loss": 1.0365,
+      "step": 17820
+    },
+    {
+      "epoch": 2.85136,
+      "grad_norm": 1.703702823327098,
+      "learning_rate": 1.3490951500195081e-06,
+      "loss": 1.1097,
+      "step": 17821
+    },
+    {
+      "epoch": 2.85152,
+      "grad_norm": 1.681810347787278,
+      "learning_rate": 1.3462687783373673e-06,
+      "loss": 1.1019,
+      "step": 17822
+    },
+    {
+      "epoch": 2.85168,
+      "grad_norm": 1.7750151989170122,
+      "learning_rate": 1.34344535033879e-06,
+      "loss": 1.0882,
+      "step": 17823
+    },
+    {
+      "epoch": 2.85184,
+      "grad_norm": 1.7599276851593355,
+      "learning_rate": 1.3406248661080644e-06,
+      "loss": 1.0627,
+      "step": 17824
+    },
+    {
+      "epoch": 2.852,
+      "grad_norm": 1.6415762103275473,
+      "learning_rate": 1.3378073257293232e-06,
+      "loss": 1.0669,
+      "step": 17825
+    },
+    {
+      "epoch": 2.85216,
+      "grad_norm": 1.9276234487874915,
+      "learning_rate": 1.334992729286655e-06,
+      "loss": 1.0497,
+      "step": 17826
+    },
+    {
+      "epoch": 2.8523199999999997,
+      "grad_norm": 1.7208788592388202,
+      "learning_rate": 1.3321810768640252e-06,
+      "loss": 1.0698,
+      "step": 17827
+    },
+    {
+      "epoch": 2.85248,
+      "grad_norm": 2.0077478665355715,
+      "learning_rate": 1.3293723685453453e-06,
+      "loss": 1.0597,
+      "step": 17828
+    },
+    {
+      "epoch": 2.85264,
+      "grad_norm": 1.7390365938143104,
+      "learning_rate": 1.3265666044144142e-06,
+      "loss": 1.0492,
+      "step": 17829
+    },
+    {
+      "epoch": 2.8528000000000002,
+      "grad_norm": 1.7091837566290256,
+      "learning_rate": 1.3237637845549432e-06,
+      "loss": 1.0677,
+      "step": 17830
+    },
+    {
+      "epoch": 2.85296,
+      "grad_norm": 1.82102696022358,
+      "learning_rate": 1.3209639090505875e-06,
+      "loss": 1.0833,
+      "step": 17831
+    },
+    {
+      "epoch": 2.85312,
+      "grad_norm": 1.9595312152155375,
+      "learning_rate": 1.3181669779848692e-06,
+      "loss": 1.079,
+      "step": 17832
+    },
+    {
+      "epoch": 2.85328,
+      "grad_norm": 2.0696837189930157,
+      "learning_rate": 1.315372991441255e-06,
+      "loss": 1.1102,
+      "step": 17833
+    },
+    {
+      "epoch": 2.85344,
+      "grad_norm": 1.733696623292707,
+      "learning_rate": 1.3125819495031221e-06,
+      "loss": 1.0659,
+      "step": 17834
+    },
+    {
+      "epoch": 2.8536,
+      "grad_norm": 1.6949116877343848,
+      "learning_rate": 1.3097938522537157e-06,
+      "loss": 1.0795,
+      "step": 17835
+    },
+    {
+      "epoch": 2.85376,
+      "grad_norm": 1.6572241360017401,
+      "learning_rate": 1.30700869977628e-06,
+      "loss": 1.0651,
+      "step": 17836
+    },
+    {
+      "epoch": 2.85392,
+      "grad_norm": 1.6637634263377523,
+      "learning_rate": 1.3042264921538816e-06,
+      "loss": 1.0699,
+      "step": 17837
+    },
+    {
+      "epoch": 2.8540799999999997,
+      "grad_norm": 1.8221013209390688,
+      "learning_rate": 1.3014472294695545e-06,
+      "loss": 1.0559,
+      "step": 17838
+    },
+    {
+      "epoch": 2.85424,
+      "grad_norm": 1.6392552507536775,
+      "learning_rate": 1.2986709118062212e-06,
+      "loss": 1.0898,
+      "step": 17839
+    },
+    {
+      "epoch": 2.8544,
+      "grad_norm": 1.6794639233194448,
+      "learning_rate": 1.2958975392467265e-06,
+      "loss": 1.0469,
+      "step": 17840
+    },
+    {
+      "epoch": 2.85456,
+      "grad_norm": 1.8433316701758846,
+      "learning_rate": 1.2931271118738152e-06,
+      "loss": 1.0522,
+      "step": 17841
+    },
+    {
+      "epoch": 2.85472,
+      "grad_norm": 1.7586827608902762,
+      "learning_rate": 1.2903596297701547e-06,
+      "loss": 1.0486,
+      "step": 17842
+    },
+    {
+      "epoch": 2.85488,
+      "grad_norm": 1.7489395120966917,
+      "learning_rate": 1.2875950930183455e-06,
+      "loss": 1.0689,
+      "step": 17843
+    },
+    {
+      "epoch": 2.85504,
+      "grad_norm": 1.6295039240709517,
+      "learning_rate": 1.284833501700855e-06,
+      "loss": 1.0384,
+      "step": 17844
+    },
+    {
+      "epoch": 2.8552,
+      "grad_norm": 1.6039792911754294,
+      "learning_rate": 1.282074855900095e-06,
+      "loss": 1.0728,
+      "step": 17845
+    },
+    {
+      "epoch": 2.85536,
+      "grad_norm": 1.8781317644244888,
+      "learning_rate": 1.2793191556983553e-06,
+      "loss": 1.0783,
+      "step": 17846
+    },
+    {
+      "epoch": 2.85552,
+      "grad_norm": 1.9074251165542881,
+      "learning_rate": 1.2765664011778921e-06,
+      "loss": 1.0787,
+      "step": 17847
+    },
+    {
+      "epoch": 2.85568,
+      "grad_norm": 1.873841428604183,
+      "learning_rate": 1.2738165924208289e-06,
+      "loss": 1.0676,
+      "step": 17848
+    },
+    {
+      "epoch": 2.85584,
+      "grad_norm": 1.7961784272307844,
+      "learning_rate": 1.2710697295092222e-06,
+      "loss": 1.0655,
+      "step": 17849
+    },
+    {
+      "epoch": 2.856,
+      "grad_norm": 1.848154812621062,
+      "learning_rate": 1.2683258125250285e-06,
+      "loss": 1.0654,
+      "step": 17850
+    },
+    {
+      "epoch": 2.85616,
+      "grad_norm": 1.871405009523617,
+      "learning_rate": 1.2655848415501271e-06,
+      "loss": 1.0416,
+      "step": 17851
+    },
+    {
+      "epoch": 2.85632,
+      "grad_norm": 1.7552031607140728,
+      "learning_rate": 1.2628468166663077e-06,
+      "loss": 1.1148,
+      "step": 17852
+    },
+    {
+      "epoch": 2.85648,
+      "grad_norm": 1.6648094653972456,
+      "learning_rate": 1.2601117379552608e-06,
+      "loss": 1.0659,
+      "step": 17853
+    },
+    {
+      "epoch": 2.85664,
+      "grad_norm": 1.7838691998555145,
+      "learning_rate": 1.2573796054985986e-06,
+      "loss": 1.0636,
+      "step": 17854
+    },
+    {
+      "epoch": 2.8568,
+      "grad_norm": 1.608133369599632,
+      "learning_rate": 1.2546504193778452e-06,
+      "loss": 1.0707,
+      "step": 17855
+    },
+    {
+      "epoch": 2.85696,
+      "grad_norm": 1.6736658923951413,
+      "learning_rate": 1.2519241796744464e-06,
+      "loss": 1.0754,
+      "step": 17856
+    },
+    {
+      "epoch": 2.85712,
+      "grad_norm": 1.6922306879499427,
+      "learning_rate": 1.2492008864697258e-06,
+      "loss": 1.0343,
+      "step": 17857
+    },
+    {
+      "epoch": 2.8572800000000003,
+      "grad_norm": 1.7038851078666437,
+      "learning_rate": 1.2464805398449519e-06,
+      "loss": 1.0551,
+      "step": 17858
+    },
+    {
+      "epoch": 2.85744,
+      "grad_norm": 1.675180626071444,
+      "learning_rate": 1.2437631398813154e-06,
+      "loss": 1.0414,
+      "step": 17859
+    },
+    {
+      "epoch": 2.8576,
+      "grad_norm": 1.6701026528603322,
+      "learning_rate": 1.2410486866598737e-06,
+      "loss": 1.0558,
+      "step": 17860
+    },
+    {
+      "epoch": 2.85776,
+      "grad_norm": 1.7456469455289687,
+      "learning_rate": 1.2383371802616394e-06,
+      "loss": 1.07,
+      "step": 17861
+    },
+    {
+      "epoch": 2.85792,
+      "grad_norm": 1.770698871668978,
+      "learning_rate": 1.2356286207675039e-06,
+      "loss": 1.0978,
+      "step": 17862
+    },
+    {
+      "epoch": 2.85808,
+      "grad_norm": 1.759136956720233,
+      "learning_rate": 1.2329230082583022e-06,
+      "loss": 1.055,
+      "step": 17863
+    },
+    {
+      "epoch": 2.85824,
+      "grad_norm": 1.7706986232828048,
+      "learning_rate": 1.2302203428147474e-06,
+      "loss": 1.0297,
+      "step": 17864
+    },
+    {
+      "epoch": 2.8584,
+      "grad_norm": 1.890365668758866,
+      "learning_rate": 1.2275206245175085e-06,
+      "loss": 1.0553,
+      "step": 17865
+    },
+    {
+      "epoch": 2.8585599999999998,
+      "grad_norm": 1.7195283046780299,
+      "learning_rate": 1.2248238534471213e-06,
+      "loss": 1.0551,
+      "step": 17866
+    },
+    {
+      "epoch": 2.85872,
+      "grad_norm": 1.947821678510605,
+      "learning_rate": 1.2221300296840654e-06,
+      "loss": 1.0414,
+      "step": 17867
+    },
+    {
+      "epoch": 2.85888,
+      "grad_norm": 1.58891184216314,
+      "learning_rate": 1.2194391533087102e-06,
+      "loss": 1.0509,
+      "step": 17868
+    },
+    {
+      "epoch": 2.8590400000000002,
+      "grad_norm": 1.63955153016183,
+      "learning_rate": 1.216751224401347e-06,
+      "loss": 1.0271,
+      "step": 17869
+    },
+    {
+      "epoch": 2.8592,
+      "grad_norm": 1.8092974738567544,
+      "learning_rate": 1.214066243042189e-06,
+      "loss": 1.0746,
+      "step": 17870
+    },
+    {
+      "epoch": 2.85936,
+      "grad_norm": 1.7656281137712602,
+      "learning_rate": 1.2113842093113504e-06,
+      "loss": 1.0688,
+      "step": 17871
+    },
+    {
+      "epoch": 2.85952,
+      "grad_norm": 1.8126518855798484,
+      "learning_rate": 1.2087051232888558e-06,
+      "loss": 1.1457,
+      "step": 17872
+    },
+    {
+      "epoch": 2.85968,
+      "grad_norm": 1.999864449191034,
+      "learning_rate": 1.2060289850546524e-06,
+      "loss": 1.0927,
+      "step": 17873
+    },
+    {
+      "epoch": 2.85984,
+      "grad_norm": 1.662935062182916,
+      "learning_rate": 1.2033557946885765e-06,
+      "loss": 1.0311,
+      "step": 17874
+    },
+    {
+      "epoch": 2.86,
+      "grad_norm": 1.7179334428170467,
+      "learning_rate": 1.2006855522704086e-06,
+      "loss": 1.0799,
+      "step": 17875
+    },
+    {
+      "epoch": 2.86016,
+      "grad_norm": 1.7520999557616208,
+      "learning_rate": 1.1980182578798073e-06,
+      "loss": 1.097,
+      "step": 17876
+    },
+    {
+      "epoch": 2.8603199999999998,
+      "grad_norm": 1.7404483751872262,
+      "learning_rate": 1.1953539115963864e-06,
+      "loss": 1.0613,
+      "step": 17877
+    },
+    {
+      "epoch": 2.86048,
+      "grad_norm": 1.7064083276483881,
+      "learning_rate": 1.192692513499616e-06,
+      "loss": 1.023,
+      "step": 17878
+    },
+    {
+      "epoch": 2.86064,
+      "grad_norm": 1.940577323638622,
+      "learning_rate": 1.1900340636689433e-06,
+      "loss": 1.082,
+      "step": 17879
+    },
+    {
+      "epoch": 2.8608000000000002,
+      "grad_norm": 1.6811529926493518,
+      "learning_rate": 1.1873785621836497e-06,
+      "loss": 1.0827,
+      "step": 17880
+    },
+    {
+      "epoch": 2.86096,
+      "grad_norm": 1.6876959552983088,
+      "learning_rate": 1.184726009123005e-06,
+      "loss": 1.0888,
+      "step": 17881
+    },
+    {
+      "epoch": 2.86112,
+      "grad_norm": 1.9223365537252979,
+      "learning_rate": 1.1820764045661458e-06,
+      "loss": 1.1352,
+      "step": 17882
+    },
+    {
+      "epoch": 2.86128,
+      "grad_norm": 1.8847485294687787,
+      "learning_rate": 1.1794297485921423e-06,
+      "loss": 1.1059,
+      "step": 17883
+    },
+    {
+      "epoch": 2.86144,
+      "grad_norm": 1.6176438203138406,
+      "learning_rate": 1.1767860412799425e-06,
+      "loss": 1.0988,
+      "step": 17884
+    },
+    {
+      "epoch": 2.8616,
+      "grad_norm": 1.7520490376315452,
+      "learning_rate": 1.174145282708461e-06,
+      "loss": 1.0449,
+      "step": 17885
+    },
+    {
+      "epoch": 2.86176,
+      "grad_norm": 1.7800128842258889,
+      "learning_rate": 1.1715074729564678e-06,
+      "loss": 1.0723,
+      "step": 17886
+    },
+    {
+      "epoch": 2.86192,
+      "grad_norm": 1.8357206989555264,
+      "learning_rate": 1.1688726121027005e-06,
+      "loss": 1.0949,
+      "step": 17887
+    },
+    {
+      "epoch": 2.8620799999999997,
+      "grad_norm": 1.5641050500137603,
+      "learning_rate": 1.1662407002257514e-06,
+      "loss": 1.0634,
+      "step": 17888
+    },
+    {
+      "epoch": 2.86224,
+      "grad_norm": 1.8153872829010054,
+      "learning_rate": 1.1636117374041578e-06,
+      "loss": 1.02,
+      "step": 17889
+    },
+    {
+      "epoch": 2.8624,
+      "grad_norm": 1.7494536370562004,
+      "learning_rate": 1.160985723716368e-06,
+      "loss": 1.065,
+      "step": 17890
+    },
+    {
+      "epoch": 2.86256,
+      "grad_norm": 1.683919408257388,
+      "learning_rate": 1.1583626592407638e-06,
+      "loss": 1.0435,
+      "step": 17891
+    },
+    {
+      "epoch": 2.86272,
+      "grad_norm": 1.7372946292247702,
+      "learning_rate": 1.1557425440555602e-06,
+      "loss": 1.0836,
+      "step": 17892
+    },
+    {
+      "epoch": 2.86288,
+      "grad_norm": 1.6822336755733749,
+      "learning_rate": 1.153125378238995e-06,
+      "loss": 1.0839,
+      "step": 17893
+    },
+    {
+      "epoch": 2.86304,
+      "grad_norm": 1.7057728447631872,
+      "learning_rate": 1.1505111618691168e-06,
+      "loss": 1.0601,
+      "step": 17894
+    },
+    {
+      "epoch": 2.8632,
+      "grad_norm": 1.8240812804276612,
+      "learning_rate": 1.1478998950239518e-06,
+      "loss": 1.0552,
+      "step": 17895
+    },
+    {
+      "epoch": 2.86336,
+      "grad_norm": 1.7103877418603801,
+      "learning_rate": 1.1452915777814043e-06,
+      "loss": 1.0455,
+      "step": 17896
+    },
+    {
+      "epoch": 2.86352,
+      "grad_norm": 1.7449012686913163,
+      "learning_rate": 1.1426862102193236e-06,
+      "loss": 1.0579,
+      "step": 17897
+    },
+    {
+      "epoch": 2.86368,
+      "grad_norm": 1.7915125128866023,
+      "learning_rate": 1.1400837924154252e-06,
+      "loss": 1.0525,
+      "step": 17898
+    },
+    {
+      "epoch": 2.86384,
+      "grad_norm": 1.856531983641352,
+      "learning_rate": 1.13748432444738e-06,
+      "loss": 1.056,
+      "step": 17899
+    },
+    {
+      "epoch": 2.864,
+      "grad_norm": 1.8801580899899515,
+      "learning_rate": 1.1348878063927372e-06,
+      "loss": 1.0367,
+      "step": 17900
+    },
+    {
+      "epoch": 2.86416,
+      "grad_norm": 1.8457877918107584,
+      "learning_rate": 1.1322942383289791e-06,
+      "loss": 1.0735,
+      "step": 17901
+    },
+    {
+      "epoch": 2.86432,
+      "grad_norm": 1.7653220383290895,
+      "learning_rate": 1.1297036203335e-06,
+      "loss": 1.0887,
+      "step": 17902
+    },
+    {
+      "epoch": 2.86448,
+      "grad_norm": 2.006893083582193,
+      "learning_rate": 1.1271159524835927e-06,
+      "loss": 1.1654,
+      "step": 17903
+    },
+    {
+      "epoch": 2.86464,
+      "grad_norm": 1.6730753870285413,
+      "learning_rate": 1.1245312348564852e-06,
+      "loss": 1.1013,
+      "step": 17904
+    },
+    {
+      "epoch": 2.8648,
+      "grad_norm": 1.8533911794037854,
+      "learning_rate": 1.1219494675292708e-06,
+      "loss": 1.0653,
+      "step": 17905
+    },
+    {
+      "epoch": 2.86496,
+      "grad_norm": 1.7677903582241021,
+      "learning_rate": 1.1193706505790102e-06,
+      "loss": 1.0915,
+      "step": 17906
+    },
+    {
+      "epoch": 2.86512,
+      "grad_norm": 1.7252030914007752,
+      "learning_rate": 1.116794784082642e-06,
+      "loss": 1.069,
+      "step": 17907
+    },
+    {
+      "epoch": 2.8652800000000003,
+      "grad_norm": 1.9863355377937133,
+      "learning_rate": 1.114221868117038e-06,
+      "loss": 1.109,
+      "step": 17908
+    },
+    {
+      "epoch": 2.86544,
+      "grad_norm": 1.7273251491688661,
+      "learning_rate": 1.111651902758959e-06,
+      "loss": 1.1159,
+      "step": 17909
+    },
+    {
+      "epoch": 2.8656,
+      "grad_norm": 1.7158687933409384,
+      "learning_rate": 1.109084888085088e-06,
+      "loss": 1.0856,
+      "step": 17910
+    },
+    {
+      "epoch": 2.86576,
+      "grad_norm": 1.7182895873794348,
+      "learning_rate": 1.1065208241720303e-06,
+      "loss": 1.0671,
+      "step": 17911
+    },
+    {
+      "epoch": 2.86592,
+      "grad_norm": 1.8077834652066591,
+      "learning_rate": 1.1039597110962808e-06,
+      "loss": 1.0628,
+      "step": 17912
+    },
+    {
+      "epoch": 2.86608,
+      "grad_norm": 1.5331471460766184,
+      "learning_rate": 1.1014015489342776e-06,
+      "loss": 1.115,
+      "step": 17913
+    },
+    {
+      "epoch": 2.86624,
+      "grad_norm": 1.7703049024701543,
+      "learning_rate": 1.0988463377623271e-06,
+      "loss": 1.0668,
+      "step": 17914
+    },
+    {
+      "epoch": 2.8664,
+      "grad_norm": 1.7114810717510678,
+      "learning_rate": 1.096294077656701e-06,
+      "loss": 1.0803,
+      "step": 17915
+    },
+    {
+      "epoch": 2.8665599999999998,
+      "grad_norm": 2.0134828027955503,
+      "learning_rate": 1.0937447686935387e-06,
+      "loss": 1.0696,
+      "step": 17916
+    },
+    {
+      "epoch": 2.86672,
+      "grad_norm": 1.8925380996964145,
+      "learning_rate": 1.0911984109489016e-06,
+      "loss": 1.0541,
+      "step": 17917
+    },
+    {
+      "epoch": 2.86688,
+      "grad_norm": 1.692039049853282,
+      "learning_rate": 1.0886550044987841e-06,
+      "loss": 1.0729,
+      "step": 17918
+    },
+    {
+      "epoch": 2.8670400000000003,
+      "grad_norm": 1.9854540297100163,
+      "learning_rate": 1.0861145494190706e-06,
+      "loss": 1.0499,
+      "step": 17919
+    },
+    {
+      "epoch": 2.8672,
+      "grad_norm": 1.6637061010190755,
+      "learning_rate": 1.0835770457855777e-06,
+      "loss": 1.0967,
+      "step": 17920
+    },
+    {
+      "epoch": 2.86736,
+      "grad_norm": 1.5884469427069898,
+      "learning_rate": 1.0810424936739894e-06,
+      "loss": 1.0501,
+      "step": 17921
+    },
+    {
+      "epoch": 2.86752,
+      "grad_norm": 1.7617713773935821,
+      "learning_rate": 1.0785108931599674e-06,
+      "loss": 1.0455,
+      "step": 17922
+    },
+    {
+      "epoch": 2.86768,
+      "grad_norm": 1.691783059254402,
+      "learning_rate": 1.0759822443190294e-06,
+      "loss": 1.101,
+      "step": 17923
+    },
+    {
+      "epoch": 2.86784,
+      "grad_norm": 1.5946146802692474,
+      "learning_rate": 1.0734565472266368e-06,
+      "loss": 1.0932,
+      "step": 17924
+    },
+    {
+      "epoch": 2.868,
+      "grad_norm": 1.665323789421903,
+      "learning_rate": 1.0709338019581517e-06,
+      "loss": 1.0781,
+      "step": 17925
+    },
+    {
+      "epoch": 2.86816,
+      "grad_norm": 1.6170534519667117,
+      "learning_rate": 1.0684140085888471e-06,
+      "loss": 1.0365,
+      "step": 17926
+    },
+    {
+      "epoch": 2.8683199999999998,
+      "grad_norm": 1.7617885576165557,
+      "learning_rate": 1.0658971671939077e-06,
+      "loss": 1.0621,
+      "step": 17927
+    },
+    {
+      "epoch": 2.86848,
+      "grad_norm": 1.7020629903102351,
+      "learning_rate": 1.06338327784844e-06,
+      "loss": 1.0612,
+      "step": 17928
+    },
+    {
+      "epoch": 2.86864,
+      "grad_norm": 1.811297981653757,
+      "learning_rate": 1.060872340627439e-06,
+      "loss": 1.0488,
+      "step": 17929
+    },
+    {
+      "epoch": 2.8688000000000002,
+      "grad_norm": 1.924780356244853,
+      "learning_rate": 1.0583643556058453e-06,
+      "loss": 1.0341,
+      "step": 17930
+    },
+    {
+      "epoch": 2.86896,
+      "grad_norm": 1.7939221880687284,
+      "learning_rate": 1.055859322858488e-06,
+      "loss": 1.0969,
+      "step": 17931
+    },
+    {
+      "epoch": 2.86912,
+      "grad_norm": 1.7820842423140373,
+      "learning_rate": 1.0533572424601069e-06,
+      "loss": 1.1201,
+      "step": 17932
+    },
+    {
+      "epoch": 2.86928,
+      "grad_norm": 1.8420063852773565,
+      "learning_rate": 1.0508581144853646e-06,
+      "loss": 1.0898,
+      "step": 17933
+    },
+    {
+      "epoch": 2.86944,
+      "grad_norm": 1.5496339976743292,
+      "learning_rate": 1.048361939008835e-06,
+      "loss": 1.0609,
+      "step": 17934
+    },
+    {
+      "epoch": 2.8696,
+      "grad_norm": 1.6765965513760903,
+      "learning_rate": 1.0458687161049919e-06,
+      "loss": 1.0292,
+      "step": 17935
+    },
+    {
+      "epoch": 2.86976,
+      "grad_norm": 1.67561978336355,
+      "learning_rate": 1.043378445848242e-06,
+      "loss": 1.0837,
+      "step": 17936
+    },
+    {
+      "epoch": 2.86992,
+      "grad_norm": 1.664138774051775,
+      "learning_rate": 1.040891128312882e-06,
+      "loss": 1.0412,
+      "step": 17937
+    },
+    {
+      "epoch": 2.8700799999999997,
+      "grad_norm": 1.6932189086544942,
+      "learning_rate": 1.0384067635731298e-06,
+      "loss": 1.0853,
+      "step": 17938
+    },
+    {
+      "epoch": 2.87024,
+      "grad_norm": 1.784651865104559,
+      "learning_rate": 1.0359253517031153e-06,
+      "loss": 1.0891,
+      "step": 17939
+    },
+    {
+      "epoch": 2.8704,
+      "grad_norm": 1.7212789716918384,
+      "learning_rate": 1.0334468927768903e-06,
+      "loss": 1.0399,
+      "step": 17940
+    },
+    {
+      "epoch": 2.8705600000000002,
+      "grad_norm": 1.903261542727953,
+      "learning_rate": 1.0309713868683957e-06,
+      "loss": 1.0535,
+      "step": 17941
+    },
+    {
+      "epoch": 2.87072,
+      "grad_norm": 1.777534925759284,
+      "learning_rate": 1.0284988340515055e-06,
+      "loss": 1.0757,
+      "step": 17942
+    },
+    {
+      "epoch": 2.87088,
+      "grad_norm": 1.8267797897684583,
+      "learning_rate": 1.0260292343999833e-06,
+      "loss": 1.0604,
+      "step": 17943
+    },
+    {
+      "epoch": 2.87104,
+      "grad_norm": 1.7960391058886114,
+      "learning_rate": 1.0235625879875366e-06,
+      "loss": 1.0565,
+      "step": 17944
+    },
+    {
+      "epoch": 2.8712,
+      "grad_norm": 1.7078284873644833,
+      "learning_rate": 1.021098894887762e-06,
+      "loss": 1.0597,
+      "step": 17945
+    },
+    {
+      "epoch": 2.87136,
+      "grad_norm": 2.092553314835943,
+      "learning_rate": 1.0186381551741674e-06,
+      "loss": 1.094,
+      "step": 17946
+    },
+    {
+      "epoch": 2.87152,
+      "grad_norm": 1.767031576066266,
+      "learning_rate": 1.016180368920161e-06,
+      "loss": 1.0256,
+      "step": 17947
+    },
+    {
+      "epoch": 2.87168,
+      "grad_norm": 1.5453765900806162,
+      "learning_rate": 1.0137255361991171e-06,
+      "loss": 1.0766,
+      "step": 17948
+    },
+    {
+      "epoch": 2.8718399999999997,
+      "grad_norm": 1.8873309159469076,
+      "learning_rate": 1.0112736570842552e-06,
+      "loss": 1.059,
+      "step": 17949
+    },
+    {
+      "epoch": 2.872,
+      "grad_norm": 1.6144723197229451,
+      "learning_rate": 1.008824731648761e-06,
+      "loss": 1.1288,
+      "step": 17950
+    },
+    {
+      "epoch": 2.87216,
+      "grad_norm": 1.721272020954657,
+      "learning_rate": 1.006378759965676e-06,
+      "loss": 1.0965,
+      "step": 17951
+    },
+    {
+      "epoch": 2.87232,
+      "grad_norm": 1.5519953614861244,
+      "learning_rate": 1.0039357421080087e-06,
+      "loss": 1.0342,
+      "step": 17952
+    },
+    {
+      "epoch": 2.87248,
+      "grad_norm": 1.6278716581975567,
+      "learning_rate": 1.001495678148634e-06,
+      "loss": 1.0856,
+      "step": 17953
+    },
+    {
+      "epoch": 2.87264,
+      "grad_norm": 1.8494311592104387,
+      "learning_rate": 9.990585681603827e-07,
+      "loss": 1.0661,
+      "step": 17954
+    },
+    {
+      "epoch": 2.8728,
+      "grad_norm": 1.5377966607254738,
+      "learning_rate": 9.966244122159629e-07,
+      "loss": 0.9884,
+      "step": 17955
+    },
+    {
+      "epoch": 2.87296,
+      "grad_norm": 1.8102905656218453,
+      "learning_rate": 9.941932103880058e-07,
+      "loss": 1.0733,
+      "step": 17956
+    },
+    {
+      "epoch": 2.87312,
+      "grad_norm": 1.4762385929675659,
+      "learning_rate": 9.917649627490644e-07,
+      "loss": 1.0833,
+      "step": 17957
+    },
+    {
+      "epoch": 2.87328,
+      "grad_norm": 1.7559549896838194,
+      "learning_rate": 9.893396693715694e-07,
+      "loss": 1.034,
+      "step": 17958
+    },
+    {
+      "epoch": 2.87344,
+      "grad_norm": 1.6219800414046035,
+      "learning_rate": 9.869173303279189e-07,
+      "loss": 1.0707,
+      "step": 17959
+    },
+    {
+      "epoch": 2.8736,
+      "grad_norm": 1.693723913582844,
+      "learning_rate": 9.844979456903769e-07,
+      "loss": 1.0262,
+      "step": 17960
+    },
+    {
+      "epoch": 2.87376,
+      "grad_norm": 1.5813967161674294,
+      "learning_rate": 9.820815155311302e-07,
+      "loss": 1.0269,
+      "step": 17961
+    },
+    {
+      "epoch": 2.87392,
+      "grad_norm": 1.771336511966746,
+      "learning_rate": 9.79668039922288e-07,
+      "loss": 1.084,
+      "step": 17962
+    },
+    {
+      "epoch": 2.87408,
+      "grad_norm": 1.7419418513141807,
+      "learning_rate": 9.772575189358702e-07,
+      "loss": 1.1027,
+      "step": 17963
+    },
+    {
+      "epoch": 2.87424,
+      "grad_norm": 1.9546056624430446,
+      "learning_rate": 9.74849952643786e-07,
+      "loss": 1.1246,
+      "step": 17964
+    },
+    {
+      "epoch": 2.8744,
+      "grad_norm": 1.9221383626066486,
+      "learning_rate": 9.72445341117889e-07,
+      "loss": 1.0558,
+      "step": 17965
+    },
+    {
+      "epoch": 2.87456,
+      "grad_norm": 1.5354111358860565,
+      "learning_rate": 9.70043684429922e-07,
+      "loss": 1.0416,
+      "step": 17966
+    },
+    {
+      "epoch": 2.87472,
+      "grad_norm": 1.835462904711899,
+      "learning_rate": 9.67644982651561e-07,
+      "loss": 1.0446,
+      "step": 17967
+    },
+    {
+      "epoch": 2.87488,
+      "grad_norm": 1.7568165665214481,
+      "learning_rate": 9.652492358543596e-07,
+      "loss": 1.067,
+      "step": 17968
+    },
+    {
+      "epoch": 2.8750400000000003,
+      "grad_norm": 1.7407134791293808,
+      "learning_rate": 9.628564441098054e-07,
+      "loss": 1.1159,
+      "step": 17969
+    },
+    {
+      "epoch": 2.8752,
+      "grad_norm": 1.516206456456417,
+      "learning_rate": 9.604666074893187e-07,
+      "loss": 1.0823,
+      "step": 17970
+    },
+    {
+      "epoch": 2.87536,
+      "grad_norm": 1.7651357326184944,
+      "learning_rate": 9.58079726064176e-07,
+      "loss": 1.0555,
+      "step": 17971
+    },
+    {
+      "epoch": 2.87552,
+      "grad_norm": 1.704913218761663,
+      "learning_rate": 9.556957999056204e-07,
+      "loss": 1.1007,
+      "step": 17972
+    },
+    {
+      "epoch": 2.87568,
+      "grad_norm": 1.7194967668454713,
+      "learning_rate": 9.533148290847837e-07,
+      "loss": 1.0889,
+      "step": 17973
+    },
+    {
+      "epoch": 2.87584,
+      "grad_norm": 1.8479179087951412,
+      "learning_rate": 9.509368136726981e-07,
+      "loss": 1.0964,
+      "step": 17974
+    },
+    {
+      "epoch": 2.876,
+      "grad_norm": 1.7985877821585632,
+      "learning_rate": 9.485617537403402e-07,
+      "loss": 1.0518,
+      "step": 17975
+    },
+    {
+      "epoch": 2.87616,
+      "grad_norm": 1.6775505073116068,
+      "learning_rate": 9.461896493585532e-07,
+      "loss": 1.093,
+      "step": 17976
+    },
+    {
+      "epoch": 2.8763199999999998,
+      "grad_norm": 1.6906409905529167,
+      "learning_rate": 9.43820500598147e-07,
+      "loss": 1.0571,
+      "step": 17977
+    },
+    {
+      "epoch": 2.87648,
+      "grad_norm": 1.7163119199954244,
+      "learning_rate": 9.414543075297766e-07,
+      "loss": 1.0958,
+      "step": 17978
+    },
+    {
+      "epoch": 2.87664,
+      "grad_norm": 1.6272345645679747,
+      "learning_rate": 9.390910702240852e-07,
+      "loss": 1.0606,
+      "step": 17979
+    },
+    {
+      "epoch": 2.8768000000000002,
+      "grad_norm": 1.7544111299038112,
+      "learning_rate": 9.367307887515608e-07,
+      "loss": 1.1066,
+      "step": 17980
+    },
+    {
+      "epoch": 2.87696,
+      "grad_norm": 1.7323081949733223,
+      "learning_rate": 9.343734631826362e-07,
+      "loss": 1.0915,
+      "step": 17981
+    },
+    {
+      "epoch": 2.87712,
+      "grad_norm": 1.7801716250981843,
+      "learning_rate": 9.320190935876549e-07,
+      "loss": 1.0665,
+      "step": 17982
+    },
+    {
+      "epoch": 2.87728,
+      "grad_norm": 1.64626348599775,
+      "learning_rate": 9.29667680036872e-07,
+      "loss": 1.0157,
+      "step": 17983
+    },
+    {
+      "epoch": 2.87744,
+      "grad_norm": 1.8463673856447087,
+      "learning_rate": 9.273192226004423e-07,
+      "loss": 1.0776,
+      "step": 17984
+    },
+    {
+      "epoch": 2.8776,
+      "grad_norm": 1.6812077459949681,
+      "learning_rate": 9.249737213484433e-07,
+      "loss": 1.126,
+      "step": 17985
+    },
+    {
+      "epoch": 2.87776,
+      "grad_norm": 1.6408804862026474,
+      "learning_rate": 9.226311763508521e-07,
+      "loss": 1.0933,
+      "step": 17986
+    },
+    {
+      "epoch": 2.87792,
+      "grad_norm": 1.7931872966420381,
+      "learning_rate": 9.202915876775909e-07,
+      "loss": 1.044,
+      "step": 17987
+    },
+    {
+      "epoch": 2.8780799999999997,
+      "grad_norm": 1.9964323379495792,
+      "learning_rate": 9.179549553984479e-07,
+      "loss": 1.0759,
+      "step": 17988
+    },
+    {
+      "epoch": 2.87824,
+      "grad_norm": 1.682058821380474,
+      "learning_rate": 9.156212795831454e-07,
+      "loss": 1.0681,
+      "step": 17989
+    },
+    {
+      "epoch": 2.8784,
+      "grad_norm": 1.6396293737408085,
+      "learning_rate": 9.132905603013165e-07,
+      "loss": 1.0485,
+      "step": 17990
+    },
+    {
+      "epoch": 2.8785600000000002,
+      "grad_norm": 1.668745675728345,
+      "learning_rate": 9.109627976225388e-07,
+      "loss": 1.1048,
+      "step": 17991
+    },
+    {
+      "epoch": 2.87872,
+      "grad_norm": 1.5935932393427492,
+      "learning_rate": 9.086379916162125e-07,
+      "loss": 1.072,
+      "step": 17992
+    },
+    {
+      "epoch": 2.87888,
+      "grad_norm": 1.5919589111671524,
+      "learning_rate": 9.063161423517597e-07,
+      "loss": 1.0663,
+      "step": 17993
+    },
+    {
+      "epoch": 2.87904,
+      "grad_norm": 1.6030136005493758,
+      "learning_rate": 9.03997249898425e-07,
+      "loss": 1.0594,
+      "step": 17994
+    },
+    {
+      "epoch": 2.8792,
+      "grad_norm": 1.898855556771452,
+      "learning_rate": 9.0168131432542e-07,
+      "loss": 1.0604,
+      "step": 17995
+    },
+    {
+      "epoch": 2.87936,
+      "grad_norm": 1.7391561096200583,
+      "learning_rate": 8.993683357018445e-07,
+      "loss": 1.068,
+      "step": 17996
+    },
+    {
+      "epoch": 2.87952,
+      "grad_norm": 1.9021943378463286,
+      "learning_rate": 8.970583140967103e-07,
+      "loss": 1.0562,
+      "step": 17997
+    },
+    {
+      "epoch": 2.87968,
+      "grad_norm": 1.6315770846288933,
+      "learning_rate": 8.947512495789512e-07,
+      "loss": 1.0669,
+      "step": 17998
+    },
+    {
+      "epoch": 2.8798399999999997,
+      "grad_norm": 1.8653498270320734,
+      "learning_rate": 8.924471422174008e-07,
+      "loss": 1.1257,
+      "step": 17999
+    },
+    {
+      "epoch": 2.88,
+      "grad_norm": 1.599996876845988,
+      "learning_rate": 8.901459920808152e-07,
+      "loss": 1.0674,
+      "step": 18000
+    },
+    {
+      "epoch": 2.88016,
+      "grad_norm": 2.029986412962527,
+      "learning_rate": 8.878477992378509e-07,
+      "loss": 1.0663,
+      "step": 18001
+    },
+    {
+      "epoch": 2.88032,
+      "grad_norm": 1.6562305184798092,
+      "learning_rate": 8.855525637570861e-07,
+      "loss": 1.0469,
+      "step": 18002
+    },
+    {
+      "epoch": 2.88048,
+      "grad_norm": 1.8438063345183122,
+      "learning_rate": 8.832602857070105e-07,
+      "loss": 1.0731,
+      "step": 18003
+    },
+    {
+      "epoch": 2.88064,
+      "grad_norm": 1.6774051545157995,
+      "learning_rate": 8.80970965156025e-07,
+      "loss": 1.0698,
+      "step": 18004
+    },
+    {
+      "epoch": 2.8808,
+      "grad_norm": 1.6501776191503814,
+      "learning_rate": 8.786846021724304e-07,
+      "loss": 1.0199,
+      "step": 18005
+    },
+    {
+      "epoch": 2.88096,
+      "grad_norm": 1.7298684773933168,
+      "learning_rate": 8.76401196824439e-07,
+      "loss": 1.0503,
+      "step": 18006
+    },
+    {
+      "epoch": 2.88112,
+      "grad_norm": 1.7315033748801814,
+      "learning_rate": 8.741207491802183e-07,
+      "loss": 1.0373,
+      "step": 18007
+    },
+    {
+      "epoch": 2.88128,
+      "grad_norm": 1.8488165708314475,
+      "learning_rate": 8.718432593077807e-07,
+      "loss": 1.0383,
+      "step": 18008
+    },
+    {
+      "epoch": 2.88144,
+      "grad_norm": 1.8179967323333008,
+      "learning_rate": 8.695687272751051e-07,
+      "loss": 1.0644,
+      "step": 18009
+    },
+    {
+      "epoch": 2.8816,
+      "grad_norm": 1.6849185493478935,
+      "learning_rate": 8.672971531500374e-07,
+      "loss": 1.1071,
+      "step": 18010
+    },
+    {
+      "epoch": 2.88176,
+      "grad_norm": 1.6799075512277362,
+      "learning_rate": 8.650285370003675e-07,
+      "loss": 1.1001,
+      "step": 18011
+    },
+    {
+      "epoch": 2.88192,
+      "grad_norm": 1.5960240934608056,
+      "learning_rate": 8.627628788938081e-07,
+      "loss": 1.0773,
+      "step": 18012
+    },
+    {
+      "epoch": 2.88208,
+      "grad_norm": 1.6432669010842633,
+      "learning_rate": 8.605001788979383e-07,
+      "loss": 1.1011,
+      "step": 18013
+    },
+    {
+      "epoch": 2.88224,
+      "grad_norm": 1.5239442549302022,
+      "learning_rate": 8.582404370802821e-07,
+      "loss": 1.034,
+      "step": 18014
+    },
+    {
+      "epoch": 2.8824,
+      "grad_norm": 1.9008374999274915,
+      "learning_rate": 8.559836535082632e-07,
+      "loss": 1.0454,
+      "step": 18015
+    },
+    {
+      "epoch": 2.88256,
+      "grad_norm": 1.6559626466036594,
+      "learning_rate": 8.537298282492279e-07,
+      "loss": 1.0565,
+      "step": 18016
+    },
+    {
+      "epoch": 2.88272,
+      "grad_norm": 1.9956381113597774,
+      "learning_rate": 8.514789613704222e-07,
+      "loss": 1.0732,
+      "step": 18017
+    },
+    {
+      "epoch": 2.88288,
+      "grad_norm": 1.7564920902783703,
+      "learning_rate": 8.492310529390146e-07,
+      "loss": 1.1532,
+      "step": 18018
+    },
+    {
+      "epoch": 2.8830400000000003,
+      "grad_norm": 1.8191507376662142,
+      "learning_rate": 8.469861030220628e-07,
+      "loss": 1.0676,
+      "step": 18019
+    },
+    {
+      "epoch": 2.8832,
+      "grad_norm": 1.9056059338530438,
+      "learning_rate": 8.447441116865684e-07,
+      "loss": 1.0391,
+      "step": 18020
+    },
+    {
+      "epoch": 2.88336,
+      "grad_norm": 1.7560438351392087,
+      "learning_rate": 8.425050789994226e-07,
+      "loss": 1.1067,
+      "step": 18021
+    },
+    {
+      "epoch": 2.88352,
+      "grad_norm": 1.659621324540273,
+      "learning_rate": 8.402690050274497e-07,
+      "loss": 1.0713,
+      "step": 18022
+    },
+    {
+      "epoch": 2.88368,
+      "grad_norm": 1.5636093751063473,
+      "learning_rate": 8.380358898373519e-07,
+      "loss": 1.0196,
+      "step": 18023
+    },
+    {
+      "epoch": 2.88384,
+      "grad_norm": 1.6333403961993274,
+      "learning_rate": 8.358057334957647e-07,
+      "loss": 1.0766,
+      "step": 18024
+    },
+    {
+      "epoch": 2.884,
+      "grad_norm": 1.8626419820268483,
+      "learning_rate": 8.335785360692461e-07,
+      "loss": 1.0895,
+      "step": 18025
+    },
+    {
+      "epoch": 2.88416,
+      "grad_norm": 1.6643643149318517,
+      "learning_rate": 8.313542976242316e-07,
+      "loss": 1.0859,
+      "step": 18026
+    },
+    {
+      "epoch": 2.8843199999999998,
+      "grad_norm": 1.8609117917185636,
+      "learning_rate": 8.291330182271129e-07,
+      "loss": 1.0658,
+      "step": 18027
+    },
+    {
+      "epoch": 2.88448,
+      "grad_norm": 1.7645388543431986,
+      "learning_rate": 8.269146979441477e-07,
+      "loss": 1.0501,
+      "step": 18028
+    },
+    {
+      "epoch": 2.88464,
+      "grad_norm": 1.6171746344518179,
+      "learning_rate": 8.246993368415501e-07,
+      "loss": 1.0504,
+      "step": 18029
+    },
+    {
+      "epoch": 2.8848000000000003,
+      "grad_norm": 1.6954377672023984,
+      "learning_rate": 8.224869349854003e-07,
+      "loss": 1.015,
+      "step": 18030
+    },
+    {
+      "epoch": 2.88496,
+      "grad_norm": 1.656367674568442,
+      "learning_rate": 8.202774924417233e-07,
+      "loss": 1.0768,
+      "step": 18031
+    },
+    {
+      "epoch": 2.88512,
+      "grad_norm": 1.7408774369214546,
+      "learning_rate": 8.180710092764554e-07,
+      "loss": 1.0798,
+      "step": 18032
+    },
+    {
+      "epoch": 2.88528,
+      "grad_norm": 1.9323159529902039,
+      "learning_rate": 8.158674855554105e-07,
+      "loss": 1.0547,
+      "step": 18033
+    },
+    {
+      "epoch": 2.88544,
+      "grad_norm": 1.7219908359550293,
+      "learning_rate": 8.136669213443582e-07,
+      "loss": 1.0643,
+      "step": 18034
+    },
+    {
+      "epoch": 2.8856,
+      "grad_norm": 1.8453622091114996,
+      "learning_rate": 8.114693167089571e-07,
+      "loss": 1.0557,
+      "step": 18035
+    },
+    {
+      "epoch": 2.88576,
+      "grad_norm": 1.6176259629594703,
+      "learning_rate": 8.09274671714777e-07,
+      "loss": 1.0836,
+      "step": 18036
+    },
+    {
+      "epoch": 2.88592,
+      "grad_norm": 1.8686905792011164,
+      "learning_rate": 8.0708298642731e-07,
+      "loss": 1.0584,
+      "step": 18037
+    },
+    {
+      "epoch": 2.8860799999999998,
+      "grad_norm": 1.8722586374668655,
+      "learning_rate": 8.048942609119481e-07,
+      "loss": 1.0948,
+      "step": 18038
+    },
+    {
+      "epoch": 2.88624,
+      "grad_norm": 1.670821884496907,
+      "learning_rate": 8.027084952339947e-07,
+      "loss": 1.0754,
+      "step": 18039
+    },
+    {
+      "epoch": 2.8864,
+      "grad_norm": 1.7153398177062211,
+      "learning_rate": 8.005256894586754e-07,
+      "loss": 1.0409,
+      "step": 18040
+    },
+    {
+      "epoch": 2.8865600000000002,
+      "grad_norm": 1.692963405856512,
+      "learning_rate": 7.983458436511159e-07,
+      "loss": 1.0588,
+      "step": 18041
+    },
+    {
+      "epoch": 2.88672,
+      "grad_norm": 1.6833143790996934,
+      "learning_rate": 7.961689578763643e-07,
+      "loss": 1.0337,
+      "step": 18042
+    },
+    {
+      "epoch": 2.88688,
+      "grad_norm": 1.700546297645002,
+      "learning_rate": 7.939950321993794e-07,
+      "loss": 1.0328,
+      "step": 18043
+    },
+    {
+      "epoch": 2.88704,
+      "grad_norm": 1.603410838162496,
+      "learning_rate": 7.918240666850318e-07,
+      "loss": 1.0455,
+      "step": 18044
+    },
+    {
+      "epoch": 2.8872,
+      "grad_norm": 1.6019716189940134,
+      "learning_rate": 7.896560613980808e-07,
+      "loss": 1.0514,
+      "step": 18045
+    },
+    {
+      "epoch": 2.88736,
+      "grad_norm": 1.6572794612609212,
+      "learning_rate": 7.8749101640323e-07,
+      "loss": 1.0177,
+      "step": 18046
+    },
+    {
+      "epoch": 2.88752,
+      "grad_norm": 1.7699436641416189,
+      "learning_rate": 7.853289317650836e-07,
+      "loss": 1.0782,
+      "step": 18047
+    },
+    {
+      "epoch": 2.88768,
+      "grad_norm": 1.789278155908635,
+      "learning_rate": 7.831698075481563e-07,
+      "loss": 1.0449,
+      "step": 18048
+    },
+    {
+      "epoch": 2.8878399999999997,
+      "grad_norm": 1.6987558547550394,
+      "learning_rate": 7.810136438168525e-07,
+      "loss": 1.0586,
+      "step": 18049
+    },
+    {
+      "epoch": 2.888,
+      "grad_norm": 1.4370780128262077,
+      "learning_rate": 7.788604406355315e-07,
+      "loss": 1.0899,
+      "step": 18050
+    },
+    {
+      "epoch": 2.88816,
+      "grad_norm": 1.750613375819182,
+      "learning_rate": 7.76710198068431e-07,
+      "loss": 1.1167,
+      "step": 18051
+    },
+    {
+      "epoch": 2.88832,
+      "grad_norm": 1.7527213288825363,
+      "learning_rate": 7.745629161797219e-07,
+      "loss": 1.0772,
+      "step": 18052
+    },
+    {
+      "epoch": 2.88848,
+      "grad_norm": 1.4114904444767096,
+      "learning_rate": 7.72418595033464e-07,
+      "loss": 1.0469,
+      "step": 18053
+    },
+    {
+      "epoch": 2.88864,
+      "grad_norm": 1.6529845722800283,
+      "learning_rate": 7.702772346936504e-07,
+      "loss": 1.0178,
+      "step": 18054
+    },
+    {
+      "epoch": 2.8888,
+      "grad_norm": 1.8300883324895936,
+      "learning_rate": 7.681388352241636e-07,
+      "loss": 1.1042,
+      "step": 18055
+    },
+    {
+      "epoch": 2.88896,
+      "grad_norm": 1.7323840111640176,
+      "learning_rate": 7.660033966888191e-07,
+      "loss": 1.0634,
+      "step": 18056
+    },
+    {
+      "epoch": 2.88912,
+      "grad_norm": 1.8840442185826323,
+      "learning_rate": 7.638709191513438e-07,
+      "loss": 1.0723,
+      "step": 18057
+    },
+    {
+      "epoch": 2.88928,
+      "grad_norm": 1.6305127558045565,
+      "learning_rate": 7.617414026753533e-07,
+      "loss": 1.1006,
+      "step": 18058
+    },
+    {
+      "epoch": 2.88944,
+      "grad_norm": 1.6688011836536598,
+      "learning_rate": 7.596148473243858e-07,
+      "loss": 1.0727,
+      "step": 18059
+    },
+    {
+      "epoch": 2.8895999999999997,
+      "grad_norm": 1.6581005729457694,
+      "learning_rate": 7.574912531619128e-07,
+      "loss": 1.0814,
+      "step": 18060
+    },
+    {
+      "epoch": 2.88976,
+      "grad_norm": 1.5581544958436633,
+      "learning_rate": 7.553706202512834e-07,
+      "loss": 1.0913,
+      "step": 18061
+    },
+    {
+      "epoch": 2.88992,
+      "grad_norm": 1.7874137400032122,
+      "learning_rate": 7.532529486557805e-07,
+      "loss": 1.0882,
+      "step": 18062
+    },
+    {
+      "epoch": 2.89008,
+      "grad_norm": 1.5240004240269387,
+      "learning_rate": 7.511382384385868e-07,
+      "loss": 1.02,
+      "step": 18063
+    },
+    {
+      "epoch": 2.89024,
+      "grad_norm": 1.8097034549213604,
+      "learning_rate": 7.490264896628185e-07,
+      "loss": 1.0612,
+      "step": 18064
+    },
+    {
+      "epoch": 2.8904,
+      "grad_norm": 1.6120235727410999,
+      "learning_rate": 7.469177023914698e-07,
+      "loss": 1.0628,
+      "step": 18065
+    },
+    {
+      "epoch": 2.89056,
+      "grad_norm": 1.6328119123264375,
+      "learning_rate": 7.448118766874679e-07,
+      "loss": 1.0452,
+      "step": 18066
+    },
+    {
+      "epoch": 2.89072,
+      "grad_norm": 1.6678013421329425,
+      "learning_rate": 7.427090126136515e-07,
+      "loss": 1.1044,
+      "step": 18067
+    },
+    {
+      "epoch": 2.89088,
+      "grad_norm": 1.600878838581474,
+      "learning_rate": 7.406091102327705e-07,
+      "loss": 1.0608,
+      "step": 18068
+    },
+    {
+      "epoch": 2.8910400000000003,
+      "grad_norm": 2.072320315486386,
+      "learning_rate": 7.385121696074638e-07,
+      "loss": 1.3287,
+      "step": 18069
+    },
+    {
+      "epoch": 2.8912,
+      "grad_norm": 1.8021635420020565,
+      "learning_rate": 7.364181908003254e-07,
+      "loss": 1.0512,
+      "step": 18070
+    },
+    {
+      "epoch": 2.89136,
+      "grad_norm": 1.6706445191151027,
+      "learning_rate": 7.343271738738278e-07,
+      "loss": 1.0481,
+      "step": 18071
+    },
+    {
+      "epoch": 2.89152,
+      "grad_norm": 1.6156043861468403,
+      "learning_rate": 7.322391188903543e-07,
+      "loss": 1.0884,
+      "step": 18072
+    },
+    {
+      "epoch": 2.89168,
+      "grad_norm": 1.7917735689732994,
+      "learning_rate": 7.30154025912222e-07,
+      "loss": 1.0347,
+      "step": 18073
+    },
+    {
+      "epoch": 2.89184,
+      "grad_norm": 1.6402448492880148,
+      "learning_rate": 7.280718950016363e-07,
+      "loss": 1.064,
+      "step": 18074
+    },
+    {
+      "epoch": 2.892,
+      "grad_norm": 1.6940372219178983,
+      "learning_rate": 7.259927262207367e-07,
+      "loss": 1.0866,
+      "step": 18075
+    },
+    {
+      "epoch": 2.89216,
+      "grad_norm": 1.8568818373541232,
+      "learning_rate": 7.239165196315511e-07,
+      "loss": 1.0531,
+      "step": 18076
+    },
+    {
+      "epoch": 2.89232,
+      "grad_norm": 1.8085621980607802,
+      "learning_rate": 7.218432752960524e-07,
+      "loss": 1.0477,
+      "step": 18077
+    },
+    {
+      "epoch": 2.89248,
+      "grad_norm": 1.5995343385520235,
+      "learning_rate": 7.197729932760688e-07,
+      "loss": 1.0109,
+      "step": 18078
+    },
+    {
+      "epoch": 2.89264,
+      "grad_norm": 1.7443833492737462,
+      "learning_rate": 7.177056736333953e-07,
+      "loss": 1.0565,
+      "step": 18079
+    },
+    {
+      "epoch": 2.8928000000000003,
+      "grad_norm": 1.6914543950698453,
+      "learning_rate": 7.15641316429716e-07,
+      "loss": 1.0815,
+      "step": 18080
+    },
+    {
+      "epoch": 2.89296,
+      "grad_norm": 1.6398627879507728,
+      "learning_rate": 7.13579921726637e-07,
+      "loss": 1.0554,
+      "step": 18081
+    },
+    {
+      "epoch": 2.89312,
+      "grad_norm": 1.7449917701677085,
+      "learning_rate": 7.115214895856426e-07,
+      "loss": 1.0508,
+      "step": 18082
+    },
+    {
+      "epoch": 2.89328,
+      "grad_norm": 1.6069032195354438,
+      "learning_rate": 7.094660200681724e-07,
+      "loss": 1.053,
+      "step": 18083
+    },
+    {
+      "epoch": 2.89344,
+      "grad_norm": 1.6764480346965536,
+      "learning_rate": 7.074135132355664e-07,
+      "loss": 1.1176,
+      "step": 18084
+    },
+    {
+      "epoch": 2.8936,
+      "grad_norm": 1.8330842484450016,
+      "learning_rate": 7.053639691490421e-07,
+      "loss": 1.0938,
+      "step": 18085
+    },
+    {
+      "epoch": 2.89376,
+      "grad_norm": 2.087022030313569,
+      "learning_rate": 7.03317387869773e-07,
+      "loss": 1.0746,
+      "step": 18086
+    },
+    {
+      "epoch": 2.89392,
+      "grad_norm": 1.6875955417127657,
+      "learning_rate": 7.012737694588322e-07,
+      "loss": 1.0842,
+      "step": 18087
+    },
+    {
+      "epoch": 2.8940799999999998,
+      "grad_norm": 1.719438776334062,
+      "learning_rate": 6.992331139771713e-07,
+      "loss": 1.1569,
+      "step": 18088
+    },
+    {
+      "epoch": 2.89424,
+      "grad_norm": 1.4965482309314333,
+      "learning_rate": 6.971954214857191e-07,
+      "loss": 0.9916,
+      "step": 18089
+    },
+    {
+      "epoch": 2.8944,
+      "grad_norm": 1.840177549014273,
+      "learning_rate": 6.951606920452381e-07,
+      "loss": 1.0614,
+      "step": 18090
+    },
+    {
+      "epoch": 2.8945600000000002,
+      "grad_norm": 1.69990136105105,
+      "learning_rate": 6.931289257164686e-07,
+      "loss": 1.0584,
+      "step": 18091
+    },
+    {
+      "epoch": 2.89472,
+      "grad_norm": 1.7127644904114059,
+      "learning_rate": 6.911001225600178e-07,
+      "loss": 1.0624,
+      "step": 18092
+    },
+    {
+      "epoch": 2.89488,
+      "grad_norm": 1.4347009846046783,
+      "learning_rate": 6.890742826364371e-07,
+      "loss": 1.0592,
+      "step": 18093
+    },
+    {
+      "epoch": 2.89504,
+      "grad_norm": 1.7851006485813317,
+      "learning_rate": 6.870514060061672e-07,
+      "loss": 1.0498,
+      "step": 18094
+    },
+    {
+      "epoch": 2.8952,
+      "grad_norm": 1.6465401365253887,
+      "learning_rate": 6.850314927295709e-07,
+      "loss": 1.0392,
+      "step": 18095
+    },
+    {
+      "epoch": 2.89536,
+      "grad_norm": 1.6227295121921412,
+      "learning_rate": 6.830145428669221e-07,
+      "loss": 1.0563,
+      "step": 18096
+    },
+    {
+      "epoch": 2.89552,
+      "grad_norm": 2.725732156798697,
+      "learning_rate": 6.81000556478395e-07,
+      "loss": 1.264,
+      "step": 18097
+    },
+    {
+      "epoch": 2.89568,
+      "grad_norm": 1.8872977222492289,
+      "learning_rate": 6.78989533624086e-07,
+      "loss": 1.0338,
+      "step": 18098
+    },
+    {
+      "epoch": 2.8958399999999997,
+      "grad_norm": 1.7917682855629056,
+      "learning_rate": 6.769814743640024e-07,
+      "loss": 1.091,
+      "step": 18099
+    },
+    {
+      "epoch": 2.896,
+      "grad_norm": 1.6647500109394293,
+      "learning_rate": 6.749763787580632e-07,
+      "loss": 1.0886,
+      "step": 18100
+    },
+    {
+      "epoch": 2.89616,
+      "grad_norm": 1.4784611533136256,
+      "learning_rate": 6.729742468660982e-07,
+      "loss": 1.1001,
+      "step": 18101
+    },
+    {
+      "epoch": 2.8963200000000002,
+      "grad_norm": 1.5236061721271774,
+      "learning_rate": 6.709750787478486e-07,
+      "loss": 1.0677,
+      "step": 18102
+    },
+    {
+      "epoch": 2.89648,
+      "grad_norm": 1.8642247158607645,
+      "learning_rate": 6.689788744629666e-07,
+      "loss": 1.0606,
+      "step": 18103
+    },
+    {
+      "epoch": 2.89664,
+      "grad_norm": 1.8392588194944686,
+      "learning_rate": 6.669856340710046e-07,
+      "loss": 1.0284,
+      "step": 18104
+    },
+    {
+      "epoch": 2.8968,
+      "grad_norm": 1.9038959582280879,
+      "learning_rate": 6.649953576314593e-07,
+      "loss": 1.0848,
+      "step": 18105
+    },
+    {
+      "epoch": 2.89696,
+      "grad_norm": 1.84276932118979,
+      "learning_rate": 6.630080452036946e-07,
+      "loss": 1.0332,
+      "step": 18106
+    },
+    {
+      "epoch": 2.89712,
+      "grad_norm": 1.8024090220525368,
+      "learning_rate": 6.610236968470295e-07,
+      "loss": 1.0476,
+      "step": 18107
+    },
+    {
+      "epoch": 2.89728,
+      "grad_norm": 1.7812445805154058,
+      "learning_rate": 6.590423126206613e-07,
+      "loss": 1.0474,
+      "step": 18108
+    },
+    {
+      "epoch": 2.89744,
+      "grad_norm": 1.6718824603471911,
+      "learning_rate": 6.570638925837203e-07,
+      "loss": 1.084,
+      "step": 18109
+    },
+    {
+      "epoch": 2.8975999999999997,
+      "grad_norm": 1.909065639630347,
+      "learning_rate": 6.550884367952259e-07,
+      "loss": 1.0257,
+      "step": 18110
+    },
+    {
+      "epoch": 2.89776,
+      "grad_norm": 1.6227483510897507,
+      "learning_rate": 6.531159453141423e-07,
+      "loss": 1.0598,
+      "step": 18111
+    },
+    {
+      "epoch": 2.89792,
+      "grad_norm": 1.7022983350895113,
+      "learning_rate": 6.51146418199311e-07,
+      "loss": 1.0911,
+      "step": 18112
+    },
+    {
+      "epoch": 2.89808,
+      "grad_norm": 1.80243590667553,
+      "learning_rate": 6.491798555095074e-07,
+      "loss": 1.1072,
+      "step": 18113
+    },
+    {
+      "epoch": 2.89824,
+      "grad_norm": 1.6396560002700529,
+      "learning_rate": 6.472162573033958e-07,
+      "loss": 1.0305,
+      "step": 18114
+    },
+    {
+      "epoch": 2.8984,
+      "grad_norm": 1.5613025580411461,
+      "learning_rate": 6.452556236395957e-07,
+      "loss": 1.054,
+      "step": 18115
+    },
+    {
+      "epoch": 2.89856,
+      "grad_norm": 1.8816722591590342,
+      "learning_rate": 6.432979545765827e-07,
+      "loss": 1.0142,
+      "step": 18116
+    },
+    {
+      "epoch": 2.89872,
+      "grad_norm": 1.6964177878594053,
+      "learning_rate": 6.413432501727767e-07,
+      "loss": 1.054,
+      "step": 18117
+    },
+    {
+      "epoch": 2.89888,
+      "grad_norm": 1.9055199150303888,
+      "learning_rate": 6.393915104865089e-07,
+      "loss": 1.0713,
+      "step": 18118
+    },
+    {
+      "epoch": 2.89904,
+      "grad_norm": 1.7066525006991573,
+      "learning_rate": 6.374427355760215e-07,
+      "loss": 1.072,
+      "step": 18119
+    },
+    {
+      "epoch": 2.8992,
+      "grad_norm": 1.8432261361467632,
+      "learning_rate": 6.354969254994458e-07,
+      "loss": 1.0512,
+      "step": 18120
+    },
+    {
+      "epoch": 2.89936,
+      "grad_norm": 1.9053178890967286,
+      "learning_rate": 6.335540803148688e-07,
+      "loss": 1.0858,
+      "step": 18121
+    },
+    {
+      "epoch": 2.89952,
+      "grad_norm": 1.5830638844294256,
+      "learning_rate": 6.316142000802327e-07,
+      "loss": 1.068,
+      "step": 18122
+    },
+    {
+      "epoch": 2.89968,
+      "grad_norm": 1.6890379483675309,
+      "learning_rate": 6.29677284853436e-07,
+      "loss": 1.0716,
+      "step": 18123
+    },
+    {
+      "epoch": 2.89984,
+      "grad_norm": 1.7923912817378709,
+      "learning_rate": 6.277433346922656e-07,
+      "loss": 1.0736,
+      "step": 18124
+    },
+    {
+      "epoch": 2.9,
+      "grad_norm": 1.9308112772998403,
+      "learning_rate": 6.258123496544532e-07,
+      "loss": 1.0775,
+      "step": 18125
+    },
+    {
+      "epoch": 2.90016,
+      "grad_norm": 1.5574145553319074,
+      "learning_rate": 6.23884329797575e-07,
+      "loss": 1.0384,
+      "step": 18126
+    },
+    {
+      "epoch": 2.90032,
+      "grad_norm": 1.836933341427705,
+      "learning_rate": 6.219592751791959e-07,
+      "loss": 1.0087,
+      "step": 18127
+    },
+    {
+      "epoch": 2.90048,
+      "grad_norm": 1.527183674281191,
+      "learning_rate": 6.200371858567367e-07,
+      "loss": 1.0161,
+      "step": 18128
+    },
+    {
+      "epoch": 2.90064,
+      "grad_norm": 1.6366992034629047,
+      "learning_rate": 6.181180618875515e-07,
+      "loss": 1.0605,
+      "step": 18129
+    },
+    {
+      "epoch": 2.9008000000000003,
+      "grad_norm": 1.7069892341372797,
+      "learning_rate": 6.162019033289168e-07,
+      "loss": 1.0309,
+      "step": 18130
+    },
+    {
+      "epoch": 2.90096,
+      "grad_norm": 1.765122247936031,
+      "learning_rate": 6.14288710237998e-07,
+      "loss": 1.0235,
+      "step": 18131
+    },
+    {
+      "epoch": 2.90112,
+      "grad_norm": 1.8150120838996457,
+      "learning_rate": 6.123784826718826e-07,
+      "loss": 1.1122,
+      "step": 18132
+    },
+    {
+      "epoch": 2.90128,
+      "grad_norm": 1.7627480397633426,
+      "learning_rate": 6.104712206875696e-07,
+      "loss": 1.0438,
+      "step": 18133
+    },
+    {
+      "epoch": 2.90144,
+      "grad_norm": 1.936539301513309,
+      "learning_rate": 6.085669243419801e-07,
+      "loss": 1.0617,
+      "step": 18134
+    },
+    {
+      "epoch": 2.9016,
+      "grad_norm": 1.811205952489237,
+      "learning_rate": 6.066655936919019e-07,
+      "loss": 1.0648,
+      "step": 18135
+    },
+    {
+      "epoch": 2.90176,
+      "grad_norm": 1.6771451941500655,
+      "learning_rate": 6.04767228794112e-07,
+      "loss": 1.0507,
+      "step": 18136
+    },
+    {
+      "epoch": 2.90192,
+      "grad_norm": 1.8667634460932672,
+      "learning_rate": 6.028718297052205e-07,
+      "loss": 1.075,
+      "step": 18137
+    },
+    {
+      "epoch": 2.9020799999999998,
+      "grad_norm": 1.7900784250819302,
+      "learning_rate": 6.009793964818045e-07,
+      "loss": 1.0839,
+      "step": 18138
+    },
+    {
+      "epoch": 2.90224,
+      "grad_norm": 1.8008697571022227,
+      "learning_rate": 5.990899291803187e-07,
+      "loss": 1.0371,
+      "step": 18139
+    },
+    {
+      "epoch": 2.9024,
+      "grad_norm": 1.7612239973914323,
+      "learning_rate": 5.972034278571514e-07,
+      "loss": 1.0347,
+      "step": 18140
+    },
+    {
+      "epoch": 2.9025600000000003,
+      "grad_norm": 1.8441474448483264,
+      "learning_rate": 5.95319892568591e-07,
+      "loss": 1.0853,
+      "step": 18141
+    },
+    {
+      "epoch": 2.90272,
+      "grad_norm": 1.4865412798313984,
+      "learning_rate": 5.934393233708369e-07,
+      "loss": 1.0747,
+      "step": 18142
+    },
+    {
+      "epoch": 2.90288,
+      "grad_norm": 1.5872936958626904,
+      "learning_rate": 5.915617203199997e-07,
+      "loss": 1.0857,
+      "step": 18143
+    },
+    {
+      "epoch": 2.90304,
+      "grad_norm": 1.8305991717500676,
+      "learning_rate": 5.896870834721013e-07,
+      "loss": 1.0663,
+      "step": 18144
+    },
+    {
+      "epoch": 2.9032,
+      "grad_norm": 1.6790488756906299,
+      "learning_rate": 5.87815412883086e-07,
+      "loss": 1.0638,
+      "step": 18145
+    },
+    {
+      "epoch": 2.90336,
+      "grad_norm": 1.5471763652741197,
+      "learning_rate": 5.859467086088088e-07,
+      "loss": 1.06,
+      "step": 18146
+    },
+    {
+      "epoch": 2.90352,
+      "grad_norm": 1.743319785255101,
+      "learning_rate": 5.840809707050144e-07,
+      "loss": 1.0597,
+      "step": 18147
+    },
+    {
+      "epoch": 2.90368,
+      "grad_norm": 1.74141123034735,
+      "learning_rate": 5.822181992273801e-07,
+      "loss": 1.0362,
+      "step": 18148
+    },
+    {
+      "epoch": 2.9038399999999998,
+      "grad_norm": 1.7022870979717963,
+      "learning_rate": 5.803583942314839e-07,
+      "loss": 1.0753,
+      "step": 18149
+    },
+    {
+      "epoch": 2.904,
+      "grad_norm": 1.830137615058492,
+      "learning_rate": 5.785015557728369e-07,
+      "loss": 1.0543,
+      "step": 18150
+    },
+    {
+      "epoch": 2.90416,
+      "grad_norm": 1.5584505888798548,
+      "learning_rate": 5.766476839068169e-07,
+      "loss": 1.0195,
+      "step": 18151
+    },
+    {
+      "epoch": 2.9043200000000002,
+      "grad_norm": 1.6568047393155687,
+      "learning_rate": 5.747967786887576e-07,
+      "loss": 1.0925,
+      "step": 18152
+    },
+    {
+      "epoch": 2.90448,
+      "grad_norm": 1.7305905680953242,
+      "learning_rate": 5.729488401738925e-07,
+      "loss": 1.0986,
+      "step": 18153
+    },
+    {
+      "epoch": 2.90464,
+      "grad_norm": 1.497028524560315,
+      "learning_rate": 5.711038684173442e-07,
+      "loss": 1.0371,
+      "step": 18154
+    },
+    {
+      "epoch": 2.9048,
+      "grad_norm": 1.9853421475458244,
+      "learning_rate": 5.692618634741798e-07,
+      "loss": 1.0675,
+      "step": 18155
+    },
+    {
+      "epoch": 2.90496,
+      "grad_norm": 1.6700397448287894,
+      "learning_rate": 5.674228253993441e-07,
+      "loss": 1.0542,
+      "step": 18156
+    },
+    {
+      "epoch": 2.90512,
+      "grad_norm": 1.78880297144279,
+      "learning_rate": 5.655867542477378e-07,
+      "loss": 1.076,
+      "step": 18157
+    },
+    {
+      "epoch": 2.90528,
+      "grad_norm": 1.64641475717678,
+      "learning_rate": 5.637536500741281e-07,
+      "loss": 1.0324,
+      "step": 18158
+    },
+    {
+      "epoch": 2.90544,
+      "grad_norm": 1.7134021576620242,
+      "learning_rate": 5.619235129332046e-07,
+      "loss": 1.0366,
+      "step": 18159
+    },
+    {
+      "epoch": 2.9055999999999997,
+      "grad_norm": 1.4802059206792635,
+      "learning_rate": 5.600963428796014e-07,
+      "loss": 1.0511,
+      "step": 18160
+    },
+    {
+      "epoch": 2.90576,
+      "grad_norm": 1.6897421959923715,
+      "learning_rate": 5.582721399678082e-07,
+      "loss": 1.0824,
+      "step": 18161
+    },
+    {
+      "epoch": 2.90592,
+      "grad_norm": 1.8281620504703364,
+      "learning_rate": 5.564509042522814e-07,
+      "loss": 1.0576,
+      "step": 18162
+    },
+    {
+      "epoch": 2.90608,
+      "grad_norm": 1.9876344602939477,
+      "learning_rate": 5.546326357873555e-07,
+      "loss": 1.0864,
+      "step": 18163
+    },
+    {
+      "epoch": 2.90624,
+      "grad_norm": 1.7731117967411907,
+      "learning_rate": 5.528173346272869e-07,
+      "loss": 1.038,
+      "step": 18164
+    },
+    {
+      "epoch": 2.9064,
+      "grad_norm": 1.5549685948644276,
+      "learning_rate": 5.510050008262324e-07,
+      "loss": 1.0287,
+      "step": 18165
+    },
+    {
+      "epoch": 2.90656,
+      "grad_norm": 1.9928455606448292,
+      "learning_rate": 5.491956344382821e-07,
+      "loss": 1.0561,
+      "step": 18166
+    },
+    {
+      "epoch": 2.90672,
+      "grad_norm": 1.6182900504249846,
+      "learning_rate": 5.473892355174148e-07,
+      "loss": 1.0914,
+      "step": 18167
+    },
+    {
+      "epoch": 2.90688,
+      "grad_norm": 1.6027753990103915,
+      "learning_rate": 5.455858041175433e-07,
+      "loss": 1.0004,
+      "step": 18168
+    },
+    {
+      "epoch": 2.90704,
+      "grad_norm": 1.5876084336354952,
+      "learning_rate": 5.437853402924576e-07,
+      "loss": 1.0441,
+      "step": 18169
+    },
+    {
+      "epoch": 2.9072,
+      "grad_norm": 1.8008623551898937,
+      "learning_rate": 5.419878440959036e-07,
+      "loss": 1.0598,
+      "step": 18170
+    },
+    {
+      "epoch": 2.90736,
+      "grad_norm": 1.6372159142689453,
+      "learning_rate": 5.401933155815053e-07,
+      "loss": 1.0636,
+      "step": 18171
+    },
+    {
+      "epoch": 2.90752,
+      "grad_norm": 1.6659818691082364,
+      "learning_rate": 5.384017548028086e-07,
+      "loss": 1.1116,
+      "step": 18172
+    },
+    {
+      "epoch": 2.90768,
+      "grad_norm": 1.7526448966461046,
+      "learning_rate": 5.366131618132708e-07,
+      "loss": 1.0333,
+      "step": 18173
+    },
+    {
+      "epoch": 2.90784,
+      "grad_norm": 1.9220833280976868,
+      "learning_rate": 5.348275366662713e-07,
+      "loss": 1.0688,
+      "step": 18174
+    },
+    {
+      "epoch": 2.908,
+      "grad_norm": 1.803134896652972,
+      "learning_rate": 5.330448794150788e-07,
+      "loss": 1.0821,
+      "step": 18175
+    },
+    {
+      "epoch": 2.90816,
+      "grad_norm": 1.6953456162884322,
+      "learning_rate": 5.31265190112884e-07,
+      "loss": 1.0559,
+      "step": 18176
+    },
+    {
+      "epoch": 2.90832,
+      "grad_norm": 1.9365372459127437,
+      "learning_rate": 5.294884688127888e-07,
+      "loss": 1.1232,
+      "step": 18177
+    },
+    {
+      "epoch": 2.90848,
+      "grad_norm": 1.6980588499371192,
+      "learning_rate": 5.277147155678286e-07,
+      "loss": 1.0818,
+      "step": 18178
+    },
+    {
+      "epoch": 2.90864,
+      "grad_norm": 1.5139424662406697,
+      "learning_rate": 5.259439304309055e-07,
+      "loss": 1.0555,
+      "step": 18179
+    },
+    {
+      "epoch": 2.9088000000000003,
+      "grad_norm": 1.8004004072731117,
+      "learning_rate": 5.241761134548662e-07,
+      "loss": 1.0584,
+      "step": 18180
+    },
+    {
+      "epoch": 2.90896,
+      "grad_norm": 1.855937815952237,
+      "learning_rate": 5.224112646924683e-07,
+      "loss": 1.0582,
+      "step": 18181
+    },
+    {
+      "epoch": 2.90912,
+      "grad_norm": 1.7734426483929524,
+      "learning_rate": 5.206493841963589e-07,
+      "loss": 1.0764,
+      "step": 18182
+    },
+    {
+      "epoch": 2.90928,
+      "grad_norm": 1.6711926453723922,
+      "learning_rate": 5.188904720191068e-07,
+      "loss": 1.0524,
+      "step": 18183
+    },
+    {
+      "epoch": 2.90944,
+      "grad_norm": 1.4830620584730256,
+      "learning_rate": 5.171345282132145e-07,
+      "loss": 1.0269,
+      "step": 18184
+    },
+    {
+      "epoch": 2.9096,
+      "grad_norm": 1.8270135708852557,
+      "learning_rate": 5.153815528310624e-07,
+      "loss": 1.0647,
+      "step": 18185
+    },
+    {
+      "epoch": 2.90976,
+      "grad_norm": 1.897972418508606,
+      "learning_rate": 5.136315459249641e-07,
+      "loss": 1.0389,
+      "step": 18186
+    },
+    {
+      "epoch": 2.90992,
+      "grad_norm": 1.6495836395753825,
+      "learning_rate": 5.118845075471335e-07,
+      "loss": 1.0815,
+      "step": 18187
+    },
+    {
+      "epoch": 2.91008,
+      "grad_norm": 1.813454510517023,
+      "learning_rate": 5.101404377496955e-07,
+      "loss": 1.0737,
+      "step": 18188
+    },
+    {
+      "epoch": 2.91024,
+      "grad_norm": 1.6746486146842827,
+      "learning_rate": 5.083993365846973e-07,
+      "loss": 1.029,
+      "step": 18189
+    },
+    {
+      "epoch": 2.9104,
+      "grad_norm": 1.7645017376671521,
+      "learning_rate": 5.066612041040753e-07,
+      "loss": 1.0783,
+      "step": 18190
+    },
+    {
+      "epoch": 2.9105600000000003,
+      "grad_norm": 1.7225826873856989,
+      "learning_rate": 5.049260403597212e-07,
+      "loss": 1.0583,
+      "step": 18191
+    },
+    {
+      "epoch": 2.91072,
+      "grad_norm": 1.718918098840264,
+      "learning_rate": 5.031938454033935e-07,
+      "loss": 1.0398,
+      "step": 18192
+    },
+    {
+      "epoch": 2.91088,
+      "grad_norm": 1.6651657208213861,
+      "learning_rate": 5.014646192867733e-07,
+      "loss": 1.042,
+      "step": 18193
+    },
+    {
+      "epoch": 2.91104,
+      "grad_norm": 1.99550327158524,
+      "learning_rate": 4.997383620614638e-07,
+      "loss": 1.1109,
+      "step": 18194
+    },
+    {
+      "epoch": 2.9112,
+      "grad_norm": 1.5410036728247978,
+      "learning_rate": 4.98015073778979e-07,
+      "loss": 1.0322,
+      "step": 18195
+    },
+    {
+      "epoch": 2.91136,
+      "grad_norm": 1.9992022957870983,
+      "learning_rate": 4.962947544907337e-07,
+      "loss": 1.142,
+      "step": 18196
+    },
+    {
+      "epoch": 2.91152,
+      "grad_norm": 1.623229331973671,
+      "learning_rate": 4.945774042480644e-07,
+      "loss": 1.0427,
+      "step": 18197
+    },
+    {
+      "epoch": 2.91168,
+      "grad_norm": 1.6258200063527233,
+      "learning_rate": 4.92863023102208e-07,
+      "loss": 1.0426,
+      "step": 18198
+    },
+    {
+      "epoch": 2.9118399999999998,
+      "grad_norm": 1.7544458465237256,
+      "learning_rate": 4.911516111043235e-07,
+      "loss": 1.1051,
+      "step": 18199
+    },
+    {
+      "epoch": 2.912,
+      "grad_norm": 1.687353354189666,
+      "learning_rate": 4.894431683054701e-07,
+      "loss": 1.0775,
+      "step": 18200
+    },
+    {
+      "epoch": 2.91216,
+      "grad_norm": 1.7400558432581452,
+      "learning_rate": 4.877376947566293e-07,
+      "loss": 1.0512,
+      "step": 18201
+    },
+    {
+      "epoch": 2.9123200000000002,
+      "grad_norm": 1.786288489430517,
+      "learning_rate": 4.860351905086935e-07,
+      "loss": 1.0832,
+      "step": 18202
+    },
+    {
+      "epoch": 2.91248,
+      "grad_norm": 1.6297165679129073,
+      "learning_rate": 4.843356556124668e-07,
+      "loss": 1.0629,
+      "step": 18203
+    },
+    {
+      "epoch": 2.91264,
+      "grad_norm": 1.697671954292518,
+      "learning_rate": 4.826390901186418e-07,
+      "loss": 1.0724,
+      "step": 18204
+    },
+    {
+      "epoch": 2.9128,
+      "grad_norm": 1.614954874332069,
+      "learning_rate": 4.809454940778668e-07,
+      "loss": 1.126,
+      "step": 18205
+    },
+    {
+      "epoch": 2.91296,
+      "grad_norm": 1.619612151278821,
+      "learning_rate": 4.792548675406572e-07,
+      "loss": 1.0172,
+      "step": 18206
+    },
+    {
+      "epoch": 2.91312,
+      "grad_norm": 1.592345156159891,
+      "learning_rate": 4.775672105574725e-07,
+      "loss": 1.0527,
+      "step": 18207
+    },
+    {
+      "epoch": 2.91328,
+      "grad_norm": 1.8003639463789076,
+      "learning_rate": 4.7588252317865014e-07,
+      "loss": 1.0611,
+      "step": 18208
+    },
+    {
+      "epoch": 2.91344,
+      "grad_norm": 1.7408756371707559,
+      "learning_rate": 4.742008054544722e-07,
+      "loss": 1.1468,
+      "step": 18209
+    },
+    {
+      "epoch": 2.9135999999999997,
+      "grad_norm": 1.5505278656768764,
+      "learning_rate": 4.725220574351319e-07,
+      "loss": 1.0981,
+      "step": 18210
+    },
+    {
+      "epoch": 2.91376,
+      "grad_norm": 1.6918727664180608,
+      "learning_rate": 4.7084627917070023e-07,
+      "loss": 1.0783,
+      "step": 18211
+    },
+    {
+      "epoch": 2.91392,
+      "grad_norm": 1.7237040482874313,
+      "learning_rate": 4.6917347071118167e-07,
+      "loss": 1.0486,
+      "step": 18212
+    },
+    {
+      "epoch": 2.9140800000000002,
+      "grad_norm": 1.8568018351946574,
+      "learning_rate": 4.6750363210649186e-07,
+      "loss": 1.0962,
+      "step": 18213
+    },
+    {
+      "epoch": 2.91424,
+      "grad_norm": 1.8169578967543518,
+      "learning_rate": 4.6583676340646867e-07,
+      "loss": 1.0682,
+      "step": 18214
+    },
+    {
+      "epoch": 2.9144,
+      "grad_norm": 1.8134531814323458,
+      "learning_rate": 4.64172864660839e-07,
+      "loss": 1.0722,
+      "step": 18215
+    },
+    {
+      "epoch": 2.91456,
+      "grad_norm": 1.7850872161136069,
+      "learning_rate": 4.62511935919252e-07,
+      "loss": 1.0898,
+      "step": 18216
+    },
+    {
+      "epoch": 2.91472,
+      "grad_norm": 1.8270019629586434,
+      "learning_rate": 4.60853977231257e-07,
+      "loss": 1.0555,
+      "step": 18217
+    },
+    {
+      "epoch": 2.91488,
+      "grad_norm": 1.9144266296157524,
+      "learning_rate": 4.591989886463366e-07,
+      "loss": 1.1081,
+      "step": 18218
+    },
+    {
+      "epoch": 2.91504,
+      "grad_norm": 1.8050867017403889,
+      "learning_rate": 4.575469702138846e-07,
+      "loss": 1.0288,
+      "step": 18219
+    },
+    {
+      "epoch": 2.9152,
+      "grad_norm": 1.6060286925792517,
+      "learning_rate": 4.558979219831727e-07,
+      "loss": 1.0136,
+      "step": 18220
+    },
+    {
+      "epoch": 2.9153599999999997,
+      "grad_norm": 1.7926155779948505,
+      "learning_rate": 4.5425184400342826e-07,
+      "loss": 1.0681,
+      "step": 18221
+    },
+    {
+      "epoch": 2.91552,
+      "grad_norm": 1.7649507417766614,
+      "learning_rate": 4.5260873632374525e-07,
+      "loss": 1.0638,
+      "step": 18222
+    },
+    {
+      "epoch": 2.91568,
+      "grad_norm": 1.638543855748606,
+      "learning_rate": 4.5096859899316224e-07,
+      "loss": 1.0583,
+      "step": 18223
+    },
+    {
+      "epoch": 2.91584,
+      "grad_norm": 1.8020308217142185,
+      "learning_rate": 4.493314320606179e-07,
+      "loss": 1.053,
+      "step": 18224
+    },
+    {
+      "epoch": 2.916,
+      "grad_norm": 1.5893999318013732,
+      "learning_rate": 4.4769723557496204e-07,
+      "loss": 1.0326,
+      "step": 18225
+    },
+    {
+      "epoch": 2.91616,
+      "grad_norm": 1.4498501507420758,
+      "learning_rate": 4.460660095849556e-07,
+      "loss": 1.0222,
+      "step": 18226
+    },
+    {
+      "epoch": 2.91632,
+      "grad_norm": 1.6887238365926427,
+      "learning_rate": 4.444377541392819e-07,
+      "loss": 1.08,
+      "step": 18227
+    },
+    {
+      "epoch": 2.91648,
+      "grad_norm": 1.836278223017296,
+      "learning_rate": 4.428124692865132e-07,
+      "loss": 1.1236,
+      "step": 18228
+    },
+    {
+      "epoch": 2.91664,
+      "grad_norm": 1.6434155053013113,
+      "learning_rate": 4.4119015507515514e-07,
+      "loss": 1.0887,
+      "step": 18229
+    },
+    {
+      "epoch": 2.9168,
+      "grad_norm": 1.5400355339775245,
+      "learning_rate": 4.3957081155360235e-07,
+      "loss": 1.021,
+      "step": 18230
+    },
+    {
+      "epoch": 2.91696,
+      "grad_norm": 1.710530374189153,
+      "learning_rate": 4.3795443877018285e-07,
+      "loss": 1.0407,
+      "step": 18231
+    },
+    {
+      "epoch": 2.91712,
+      "grad_norm": 1.6036988078697179,
+      "learning_rate": 4.3634103677312466e-07,
+      "loss": 1.0943,
+      "step": 18232
+    },
+    {
+      "epoch": 2.91728,
+      "grad_norm": 1.6885184952861734,
+      "learning_rate": 4.347306056105782e-07,
+      "loss": 1.0593,
+      "step": 18233
+    },
+    {
+      "epoch": 2.91744,
+      "grad_norm": 1.7480833240798388,
+      "learning_rate": 4.3312314533058285e-07,
+      "loss": 1.0658,
+      "step": 18234
+    },
+    {
+      "epoch": 2.9176,
+      "grad_norm": 1.83550524998713,
+      "learning_rate": 4.315186559811113e-07,
+      "loss": 1.0845,
+      "step": 18235
+    },
+    {
+      "epoch": 2.91776,
+      "grad_norm": 1.5371792648651876,
+      "learning_rate": 4.299171376100364e-07,
+      "loss": 1.0432,
+      "step": 18236
+    },
+    {
+      "epoch": 2.91792,
+      "grad_norm": 1.7702894576051424,
+      "learning_rate": 4.283185902651421e-07,
+      "loss": 1.1047,
+      "step": 18237
+    },
+    {
+      "epoch": 2.91808,
+      "grad_norm": 1.7622512456157013,
+      "learning_rate": 4.2672301399412363e-07,
+      "loss": 1.0877,
+      "step": 18238
+    },
+    {
+      "epoch": 2.91824,
+      "grad_norm": 1.4181285204392833,
+      "learning_rate": 4.2513040884460953e-07,
+      "loss": 1.0611,
+      "step": 18239
+    },
+    {
+      "epoch": 2.9184,
+      "grad_norm": 1.6110225963826994,
+      "learning_rate": 4.2354077486410627e-07,
+      "loss": 1.0869,
+      "step": 18240
+    },
+    {
+      "epoch": 2.9185600000000003,
+      "grad_norm": 1.4178493207532512,
+      "learning_rate": 4.2195411210004256e-07,
+      "loss": 1.0245,
+      "step": 18241
+    },
+    {
+      "epoch": 2.91872,
+      "grad_norm": 1.6283382349246496,
+      "learning_rate": 4.203704205997694e-07,
+      "loss": 1.0548,
+      "step": 18242
+    },
+    {
+      "epoch": 2.91888,
+      "grad_norm": 1.6869211623554319,
+      "learning_rate": 4.187897004105379e-07,
+      "loss": 1.056,
+      "step": 18243
+    },
+    {
+      "epoch": 2.91904,
+      "grad_norm": 1.8883526998835272,
+      "learning_rate": 4.172119515795214e-07,
+      "loss": 1.1087,
+      "step": 18244
+    },
+    {
+      "epoch": 2.9192,
+      "grad_norm": 1.9106933520017604,
+      "learning_rate": 4.1563717415379344e-07,
+      "loss": 1.0883,
+      "step": 18245
+    },
+    {
+      "epoch": 2.91936,
+      "grad_norm": 1.7628611299491685,
+      "learning_rate": 4.140653681803386e-07,
+      "loss": 1.0355,
+      "step": 18246
+    },
+    {
+      "epoch": 2.91952,
+      "grad_norm": 1.6499523191328656,
+      "learning_rate": 4.1249653370605267e-07,
+      "loss": 1.0433,
+      "step": 18247
+    },
+    {
+      "epoch": 2.91968,
+      "grad_norm": 1.5504705956123526,
+      "learning_rate": 4.10930670777776e-07,
+      "loss": 1.0856,
+      "step": 18248
+    },
+    {
+      "epoch": 2.9198399999999998,
+      "grad_norm": 1.9150499696554089,
+      "learning_rate": 4.093677794421935e-07,
+      "loss": 1.0004,
+      "step": 18249
+    },
+    {
+      "epoch": 2.92,
+      "grad_norm": 1.9101795678133089,
+      "learning_rate": 4.078078597459678e-07,
+      "loss": 1.057,
+      "step": 18250
+    },
+    {
+      "epoch": 2.92016,
+      "grad_norm": 1.5432670819883427,
+      "learning_rate": 4.0625091173562837e-07,
+      "loss": 1.0663,
+      "step": 18251
+    },
+    {
+      "epoch": 2.9203200000000002,
+      "grad_norm": 1.6034830121748187,
+      "learning_rate": 4.0469693545764906e-07,
+      "loss": 1.024,
+      "step": 18252
+    },
+    {
+      "epoch": 2.92048,
+      "grad_norm": 1.6980566379969604,
+      "learning_rate": 4.0314593095837073e-07,
+      "loss": 1.0405,
+      "step": 18253
+    },
+    {
+      "epoch": 2.92064,
+      "grad_norm": 1.844752476178814,
+      "learning_rate": 4.0159789828410067e-07,
+      "loss": 1.1222,
+      "step": 18254
+    },
+    {
+      "epoch": 2.9208,
+      "grad_norm": 1.7618647595028496,
+      "learning_rate": 4.000528374810242e-07,
+      "loss": 1.0777,
+      "step": 18255
+    },
+    {
+      "epoch": 2.92096,
+      "grad_norm": 1.7282716552531472,
+      "learning_rate": 3.985107485952266e-07,
+      "loss": 1.1092,
+      "step": 18256
+    },
+    {
+      "epoch": 2.92112,
+      "grad_norm": 1.877337610324912,
+      "learning_rate": 3.969716316727379e-07,
+      "loss": 1.0719,
+      "step": 18257
+    },
+    {
+      "epoch": 2.92128,
+      "grad_norm": 1.6903197271016281,
+      "learning_rate": 3.9543548675948784e-07,
+      "loss": 1.0997,
+      "step": 18258
+    },
+    {
+      "epoch": 2.92144,
+      "grad_norm": 1.6246976366670436,
+      "learning_rate": 3.9390231390128427e-07,
+      "loss": 1.0161,
+      "step": 18259
+    },
+    {
+      "epoch": 2.9215999999999998,
+      "grad_norm": 1.6450358637704245,
+      "learning_rate": 3.9237211314391286e-07,
+      "loss": 1.0383,
+      "step": 18260
+    },
+    {
+      "epoch": 2.92176,
+      "grad_norm": 1.6742667084320098,
+      "learning_rate": 3.908448845330037e-07,
+      "loss": 1.0684,
+      "step": 18261
+    },
+    {
+      "epoch": 2.92192,
+      "grad_norm": 1.5309683151170552,
+      "learning_rate": 3.893206281141426e-07,
+      "loss": 1.0639,
+      "step": 18262
+    },
+    {
+      "epoch": 2.9220800000000002,
+      "grad_norm": 1.8604082664676989,
+      "learning_rate": 3.8779934393280426e-07,
+      "loss": 1.0614,
+      "step": 18263
+    },
+    {
+      "epoch": 2.92224,
+      "grad_norm": 1.5802370189783221,
+      "learning_rate": 3.862810320343857e-07,
+      "loss": 1.0158,
+      "step": 18264
+    },
+    {
+      "epoch": 2.9224,
+      "grad_norm": 1.9156480720782467,
+      "learning_rate": 3.8476569246417296e-07,
+      "loss": 1.0343,
+      "step": 18265
+    },
+    {
+      "epoch": 2.92256,
+      "grad_norm": 1.8945903804795656,
+      "learning_rate": 3.8325332526741866e-07,
+      "loss": 1.0809,
+      "step": 18266
+    },
+    {
+      "epoch": 2.92272,
+      "grad_norm": 1.8384831013113512,
+      "learning_rate": 3.8174393048922007e-07,
+      "loss": 1.0503,
+      "step": 18267
+    },
+    {
+      "epoch": 2.92288,
+      "grad_norm": 1.8208685437728613,
+      "learning_rate": 3.80237508174619e-07,
+      "loss": 1.079,
+      "step": 18268
+    },
+    {
+      "epoch": 2.92304,
+      "grad_norm": 1.9813775211917397,
+      "learning_rate": 3.7873405836856834e-07,
+      "loss": 1.0595,
+      "step": 18269
+    },
+    {
+      "epoch": 2.9232,
+      "grad_norm": 1.6728781293675221,
+      "learning_rate": 3.772335811159322e-07,
+      "loss": 1.0371,
+      "step": 18270
+    },
+    {
+      "epoch": 2.9233599999999997,
+      "grad_norm": 1.6131323443662147,
+      "learning_rate": 3.757360764614859e-07,
+      "loss": 1.0207,
+      "step": 18271
+    },
+    {
+      "epoch": 2.92352,
+      "grad_norm": 1.7573990017063346,
+      "learning_rate": 3.7424154444990476e-07,
+      "loss": 1.0708,
+      "step": 18272
+    },
+    {
+      "epoch": 2.92368,
+      "grad_norm": 1.9753373299545123,
+      "learning_rate": 3.727499851257865e-07,
+      "loss": 1.067,
+      "step": 18273
+    },
+    {
+      "epoch": 2.92384,
+      "grad_norm": 1.7350734014461493,
+      "learning_rate": 3.7126139853362884e-07,
+      "loss": 1.0355,
+      "step": 18274
+    },
+    {
+      "epoch": 2.924,
+      "grad_norm": 1.8242443440724716,
+      "learning_rate": 3.697757847178629e-07,
+      "loss": 1.0943,
+      "step": 18275
+    },
+    {
+      "epoch": 2.92416,
+      "grad_norm": 1.7933246364793562,
+      "learning_rate": 3.6829314372281984e-07,
+      "loss": 1.0834,
+      "step": 18276
+    },
+    {
+      "epoch": 2.92432,
+      "grad_norm": 1.9432973955370123,
+      "learning_rate": 3.6681347559270886e-07,
+      "loss": 1.0981,
+      "step": 18277
+    },
+    {
+      "epoch": 2.92448,
+      "grad_norm": 1.5442139917429303,
+      "learning_rate": 3.653367803717278e-07,
+      "loss": 1.0594,
+      "step": 18278
+    },
+    {
+      "epoch": 2.92464,
+      "grad_norm": 1.7086162071366033,
+      "learning_rate": 3.6386305810389707e-07,
+      "loss": 1.0904,
+      "step": 18279
+    },
+    {
+      "epoch": 2.9248,
+      "grad_norm": 1.6757084799279056,
+      "learning_rate": 3.6239230883321483e-07,
+      "loss": 1.0821,
+      "step": 18280
+    },
+    {
+      "epoch": 2.92496,
+      "grad_norm": 1.8926390623061378,
+      "learning_rate": 3.60924532603546e-07,
+      "loss": 1.1051,
+      "step": 18281
+    },
+    {
+      "epoch": 2.92512,
+      "grad_norm": 1.8191525479047554,
+      "learning_rate": 3.59459729458711e-07,
+      "loss": 1.051,
+      "step": 18282
+    },
+    {
+      "epoch": 2.92528,
+      "grad_norm": 1.693370324194335,
+      "learning_rate": 3.579978994424082e-07,
+      "loss": 1.0328,
+      "step": 18283
+    },
+    {
+      "epoch": 2.92544,
+      "grad_norm": 1.7329746940473136,
+      "learning_rate": 3.565390425982473e-07,
+      "loss": 1.0555,
+      "step": 18284
+    },
+    {
+      "epoch": 2.9256,
+      "grad_norm": 1.6172871327128635,
+      "learning_rate": 3.550831589697712e-07,
+      "loss": 1.0124,
+      "step": 18285
+    },
+    {
+      "epoch": 2.92576,
+      "grad_norm": 1.9964247302690705,
+      "learning_rate": 3.536302486004117e-07,
+      "loss": 1.1356,
+      "step": 18286
+    },
+    {
+      "epoch": 2.92592,
+      "grad_norm": 1.7274055593040283,
+      "learning_rate": 3.5218031153353423e-07,
+      "loss": 1.0439,
+      "step": 18287
+    },
+    {
+      "epoch": 2.92608,
+      "grad_norm": 1.7124806855926071,
+      "learning_rate": 3.50733347812382e-07,
+      "loss": 1.0459,
+      "step": 18288
+    },
+    {
+      "epoch": 2.92624,
+      "grad_norm": 2.0007241837906222,
+      "learning_rate": 3.4928935748015367e-07,
+      "loss": 1.0314,
+      "step": 18289
+    },
+    {
+      "epoch": 2.9264,
+      "grad_norm": 1.7276800300979287,
+      "learning_rate": 3.478483405799149e-07,
+      "loss": 1.0611,
+      "step": 18290
+    },
+    {
+      "epoch": 2.9265600000000003,
+      "grad_norm": 1.8581241563135666,
+      "learning_rate": 3.4641029715467564e-07,
+      "loss": 1.1063,
+      "step": 18291
+    },
+    {
+      "epoch": 2.92672,
+      "grad_norm": 1.890473872371131,
+      "learning_rate": 3.449752272473461e-07,
+      "loss": 1.0586,
+      "step": 18292
+    },
+    {
+      "epoch": 2.92688,
+      "grad_norm": 1.598612855146468,
+      "learning_rate": 3.435431309007364e-07,
+      "loss": 1.1074,
+      "step": 18293
+    },
+    {
+      "epoch": 2.92704,
+      "grad_norm": 1.5035046470658002,
+      "learning_rate": 3.421140081575902e-07,
+      "loss": 0.9944,
+      "step": 18294
+    },
+    {
+      "epoch": 2.9272,
+      "grad_norm": 1.899186416712617,
+      "learning_rate": 3.4068785906054e-07,
+      "loss": 1.1015,
+      "step": 18295
+    },
+    {
+      "epoch": 2.92736,
+      "grad_norm": 1.8327445151342396,
+      "learning_rate": 3.392646836521518e-07,
+      "loss": 1.1614,
+      "step": 18296
+    },
+    {
+      "epoch": 2.92752,
+      "grad_norm": 1.581134986272302,
+      "learning_rate": 3.378444819748805e-07,
+      "loss": 1.1066,
+      "step": 18297
+    },
+    {
+      "epoch": 2.92768,
+      "grad_norm": 1.7041477799680795,
+      "learning_rate": 3.3642725407110333e-07,
+      "loss": 1.0381,
+      "step": 18298
+    },
+    {
+      "epoch": 2.9278399999999998,
+      "grad_norm": 1.6634519541570523,
+      "learning_rate": 3.3501299998311976e-07,
+      "loss": 1.0982,
+      "step": 18299
+    },
+    {
+      "epoch": 2.928,
+      "grad_norm": 1.7708583660658594,
+      "learning_rate": 3.336017197531183e-07,
+      "loss": 1.0826,
+      "step": 18300
+    },
+    {
+      "epoch": 2.92816,
+      "grad_norm": 1.791812028772664,
+      "learning_rate": 3.3219341342320964e-07,
+      "loss": 1.0491,
+      "step": 18301
+    },
+    {
+      "epoch": 2.9283200000000003,
+      "grad_norm": 1.8011669995047366,
+      "learning_rate": 3.3078808103542693e-07,
+      "loss": 1.1084,
+      "step": 18302
+    },
+    {
+      "epoch": 2.92848,
+      "grad_norm": 1.6880817016953977,
+      "learning_rate": 3.2938572263168097e-07,
+      "loss": 1.1351,
+      "step": 18303
+    },
+    {
+      "epoch": 2.92864,
+      "grad_norm": 1.5670921253743602,
+      "learning_rate": 3.279863382538384e-07,
+      "loss": 1.1134,
+      "step": 18304
+    },
+    {
+      "epoch": 2.9288,
+      "grad_norm": 1.8151762436684153,
+      "learning_rate": 3.2658992794364353e-07,
+      "loss": 1.049,
+      "step": 18305
+    },
+    {
+      "epoch": 2.92896,
+      "grad_norm": 1.721024733058849,
+      "learning_rate": 3.2519649174276304e-07,
+      "loss": 1.0877,
+      "step": 18306
+    },
+    {
+      "epoch": 2.92912,
+      "grad_norm": 1.6548573630943928,
+      "learning_rate": 3.2380602969277473e-07,
+      "loss": 1.0748,
+      "step": 18307
+    },
+    {
+      "epoch": 2.92928,
+      "grad_norm": 2.0798683787203363,
+      "learning_rate": 3.224185418351788e-07,
+      "loss": 1.0955,
+      "step": 18308
+    },
+    {
+      "epoch": 2.92944,
+      "grad_norm": 1.7556976953815677,
+      "learning_rate": 3.2103402821136443e-07,
+      "loss": 1.0669,
+      "step": 18309
+    },
+    {
+      "epoch": 2.9295999999999998,
+      "grad_norm": 1.6879011518988785,
+      "learning_rate": 3.19652488862654e-07,
+      "loss": 1.0367,
+      "step": 18310
+    },
+    {
+      "epoch": 2.92976,
+      "grad_norm": 1.6062532336817184,
+      "learning_rate": 3.18273923830259e-07,
+      "loss": 1.0314,
+      "step": 18311
+    },
+    {
+      "epoch": 2.92992,
+      "grad_norm": 1.833205761348502,
+      "learning_rate": 3.1689833315531326e-07,
+      "loss": 1.0724,
+      "step": 18312
+    },
+    {
+      "epoch": 2.9300800000000002,
+      "grad_norm": 1.8966165436866045,
+      "learning_rate": 3.155257168788728e-07,
+      "loss": 1.0471,
+      "step": 18313
+    },
+    {
+      "epoch": 2.93024,
+      "grad_norm": 1.7986252958175566,
+      "learning_rate": 3.141560750418937e-07,
+      "loss": 1.0099,
+      "step": 18314
+    },
+    {
+      "epoch": 2.9304,
+      "grad_norm": 1.7145502938594879,
+      "learning_rate": 3.1278940768523224e-07,
+      "loss": 1.0415,
+      "step": 18315
+    },
+    {
+      "epoch": 2.93056,
+      "grad_norm": 1.5365661786539013,
+      "learning_rate": 3.1142571484967796e-07,
+      "loss": 1.0207,
+      "step": 18316
+    },
+    {
+      "epoch": 2.93072,
+      "grad_norm": 1.6991929223617925,
+      "learning_rate": 3.100649965759206e-07,
+      "loss": 1.0496,
+      "step": 18317
+    },
+    {
+      "epoch": 2.93088,
+      "grad_norm": 1.680853771888096,
+      "learning_rate": 3.087072529045498e-07,
+      "loss": 1.0233,
+      "step": 18318
+    },
+    {
+      "epoch": 2.93104,
+      "grad_norm": 1.8534223815354758,
+      "learning_rate": 3.0735248387609996e-07,
+      "loss": 1.0524,
+      "step": 18319
+    },
+    {
+      "epoch": 2.9312,
+      "grad_norm": 1.5958280062391748,
+      "learning_rate": 3.0600068953098306e-07,
+      "loss": 1.0459,
+      "step": 18320
+    },
+    {
+      "epoch": 2.9313599999999997,
+      "grad_norm": 1.8620420828634443,
+      "learning_rate": 3.046518699095335e-07,
+      "loss": 1.0723,
+      "step": 18321
+    },
+    {
+      "epoch": 2.93152,
+      "grad_norm": 1.7311881967668776,
+      "learning_rate": 3.0330602505199704e-07,
+      "loss": 1.075,
+      "step": 18322
+    },
+    {
+      "epoch": 2.93168,
+      "grad_norm": 1.6546812510557454,
+      "learning_rate": 3.0196315499854135e-07,
+      "loss": 1.0607,
+      "step": 18323
+    },
+    {
+      "epoch": 2.9318400000000002,
+      "grad_norm": 1.9935495541165003,
+      "learning_rate": 3.0062325978923445e-07,
+      "loss": 1.1248,
+      "step": 18324
+    },
+    {
+      "epoch": 2.932,
+      "grad_norm": 1.5319559686968152,
+      "learning_rate": 2.992863394640444e-07,
+      "loss": 1.0291,
+      "step": 18325
+    },
+    {
+      "epoch": 2.93216,
+      "grad_norm": 1.8136090657176172,
+      "learning_rate": 2.9795239406287255e-07,
+      "loss": 1.0984,
+      "step": 18326
+    },
+    {
+      "epoch": 2.93232,
+      "grad_norm": 1.6847360713061428,
+      "learning_rate": 2.966214236255205e-07,
+      "loss": 1.0515,
+      "step": 18327
+    },
+    {
+      "epoch": 2.93248,
+      "grad_norm": 1.8064177077411987,
+      "learning_rate": 2.952934281917008e-07,
+      "loss": 1.0612,
+      "step": 18328
+    },
+    {
+      "epoch": 2.93264,
+      "grad_norm": 1.7123803055258537,
+      "learning_rate": 2.939684078010374e-07,
+      "loss": 1.0614,
+      "step": 18329
+    },
+    {
+      "epoch": 2.9328,
+      "grad_norm": 1.8948408917295947,
+      "learning_rate": 2.926463624930764e-07,
+      "loss": 1.075,
+      "step": 18330
+    },
+    {
+      "epoch": 2.93296,
+      "grad_norm": 1.818433317852076,
+      "learning_rate": 2.9132729230725297e-07,
+      "loss": 1.0562,
+      "step": 18331
+    },
+    {
+      "epoch": 2.9331199999999997,
+      "grad_norm": 2.5117511603495024,
+      "learning_rate": 2.9001119728293557e-07,
+      "loss": 1.1695,
+      "step": 18332
+    },
+    {
+      "epoch": 2.93328,
+      "grad_norm": 1.7950706623028923,
+      "learning_rate": 2.886980774593928e-07,
+      "loss": 1.0597,
+      "step": 18333
+    },
+    {
+      "epoch": 2.93344,
+      "grad_norm": 2.103511010076261,
+      "learning_rate": 2.8738793287580443e-07,
+      "loss": 1.0684,
+      "step": 18334
+    },
+    {
+      "epoch": 2.9336,
+      "grad_norm": 1.6135410841477194,
+      "learning_rate": 2.8608076357126146e-07,
+      "loss": 1.0485,
+      "step": 18335
+    },
+    {
+      "epoch": 2.93376,
+      "grad_norm": 1.7579558294774444,
+      "learning_rate": 2.8477656958476596e-07,
+      "loss": 1.0713,
+      "step": 18336
+    },
+    {
+      "epoch": 2.93392,
+      "grad_norm": 1.627428490135553,
+      "learning_rate": 2.834753509552424e-07,
+      "loss": 1.084,
+      "step": 18337
+    },
+    {
+      "epoch": 2.93408,
+      "grad_norm": 1.666152097180518,
+      "learning_rate": 2.821771077215041e-07,
+      "loss": 1.0586,
+      "step": 18338
+    },
+    {
+      "epoch": 2.93424,
+      "grad_norm": 1.6561522491661531,
+      "learning_rate": 2.80881839922309e-07,
+      "loss": 1.0909,
+      "step": 18339
+    },
+    {
+      "epoch": 2.9344,
+      "grad_norm": 1.7230987242436233,
+      "learning_rate": 2.7958954759628174e-07,
+      "loss": 1.0239,
+      "step": 18340
+    },
+    {
+      "epoch": 2.93456,
+      "grad_norm": 1.3895173231389537,
+      "learning_rate": 2.783002307820026e-07,
+      "loss": 1.0613,
+      "step": 18341
+    },
+    {
+      "epoch": 2.93472,
+      "grad_norm": 2.128173222143219,
+      "learning_rate": 2.7701388951794084e-07,
+      "loss": 1.0952,
+      "step": 18342
+    },
+    {
+      "epoch": 2.93488,
+      "grad_norm": 1.8220983650495872,
+      "learning_rate": 2.757305238424657e-07,
+      "loss": 1.0053,
+      "step": 18343
+    },
+    {
+      "epoch": 2.93504,
+      "grad_norm": 1.6216484535453652,
+      "learning_rate": 2.7445013379386874e-07,
+      "loss": 1.0485,
+      "step": 18344
+    },
+    {
+      "epoch": 2.9352,
+      "grad_norm": 1.8543643036086954,
+      "learning_rate": 2.7317271941037504e-07,
+      "loss": 1.0556,
+      "step": 18345
+    },
+    {
+      "epoch": 2.93536,
+      "grad_norm": 1.8793573373977919,
+      "learning_rate": 2.718982807300874e-07,
+      "loss": 1.0487,
+      "step": 18346
+    },
+    {
+      "epoch": 2.93552,
+      "grad_norm": 1.6057543529434062,
+      "learning_rate": 2.7062681779104203e-07,
+      "loss": 1.0482,
+      "step": 18347
+    },
+    {
+      "epoch": 2.93568,
+      "grad_norm": 1.6073449916793434,
+      "learning_rate": 2.693583306311642e-07,
+      "loss": 1.0635,
+      "step": 18348
+    },
+    {
+      "epoch": 2.93584,
+      "grad_norm": 1.715799417832246,
+      "learning_rate": 2.680928192883125e-07,
+      "loss": 1.0736,
+      "step": 18349
+    },
+    {
+      "epoch": 2.936,
+      "grad_norm": 1.6581774788852786,
+      "learning_rate": 2.668302838002457e-07,
+      "loss": 1.0337,
+      "step": 18350
+    },
+    {
+      "epoch": 2.93616,
+      "grad_norm": 1.6475281585879231,
+      "learning_rate": 2.6557072420463347e-07,
+      "loss": 1.0455,
+      "step": 18351
+    },
+    {
+      "epoch": 2.9363200000000003,
+      "grad_norm": 1.8144255282085233,
+      "learning_rate": 2.643141405390681e-07,
+      "loss": 1.048,
+      "step": 18352
+    },
+    {
+      "epoch": 2.93648,
+      "grad_norm": 1.7406046083413738,
+      "learning_rate": 2.6306053284104183e-07,
+      "loss": 1.0528,
+      "step": 18353
+    },
+    {
+      "epoch": 2.93664,
+      "grad_norm": 1.8082739469500104,
+      "learning_rate": 2.618099011479469e-07,
+      "loss": 1.0549,
+      "step": 18354
+    },
+    {
+      "epoch": 2.9368,
+      "grad_norm": 1.6556741066433793,
+      "learning_rate": 2.6056224549710907e-07,
+      "loss": 1.0451,
+      "step": 18355
+    },
+    {
+      "epoch": 2.93696,
+      "grad_norm": 1.685682290018379,
+      "learning_rate": 2.5931756592576517e-07,
+      "loss": 1.0727,
+      "step": 18356
+    },
+    {
+      "epoch": 2.93712,
+      "grad_norm": 1.9810861704866716,
+      "learning_rate": 2.580758624710522e-07,
+      "loss": 1.0817,
+      "step": 18357
+    },
+    {
+      "epoch": 2.93728,
+      "grad_norm": 1.7059342181141939,
+      "learning_rate": 2.568371351700072e-07,
+      "loss": 1.0871,
+      "step": 18358
+    },
+    {
+      "epoch": 2.93744,
+      "grad_norm": 1.5825827314815961,
+      "learning_rate": 2.5560138405961167e-07,
+      "loss": 1.075,
+      "step": 18359
+    },
+    {
+      "epoch": 2.9375999999999998,
+      "grad_norm": 1.758064829068191,
+      "learning_rate": 2.543686091767139e-07,
+      "loss": 1.0982,
+      "step": 18360
+    },
+    {
+      "epoch": 2.93776,
+      "grad_norm": 1.682270181241566,
+      "learning_rate": 2.531388105581178e-07,
+      "loss": 1.0092,
+      "step": 18361
+    },
+    {
+      "epoch": 2.93792,
+      "grad_norm": 1.9922652378704475,
+      "learning_rate": 2.519119882405163e-07,
+      "loss": 1.0789,
+      "step": 18362
+    },
+    {
+      "epoch": 2.9380800000000002,
+      "grad_norm": 2.463035887640298,
+      "learning_rate": 2.506881422605134e-07,
+      "loss": 1.1181,
+      "step": 18363
+    },
+    {
+      "epoch": 2.93824,
+      "grad_norm": 1.9976604017568387,
+      "learning_rate": 2.494672726546243e-07,
+      "loss": 1.0388,
+      "step": 18364
+    },
+    {
+      "epoch": 2.9384,
+      "grad_norm": 1.812404423824692,
+      "learning_rate": 2.482493794592755e-07,
+      "loss": 1.0828,
+      "step": 18365
+    },
+    {
+      "epoch": 2.93856,
+      "grad_norm": 1.7033548498111213,
+      "learning_rate": 2.4703446271081565e-07,
+      "loss": 1.0503,
+      "step": 18366
+    },
+    {
+      "epoch": 2.93872,
+      "grad_norm": 1.8033045923755457,
+      "learning_rate": 2.4582252244549354e-07,
+      "loss": 1.0992,
+      "step": 18367
+    },
+    {
+      "epoch": 2.93888,
+      "grad_norm": 2.0803752793323484,
+      "learning_rate": 2.446135586994691e-07,
+      "loss": 1.1218,
+      "step": 18368
+    },
+    {
+      "epoch": 2.93904,
+      "grad_norm": 1.6580315124868887,
+      "learning_rate": 2.4340757150881355e-07,
+      "loss": 1.1311,
+      "step": 18369
+    },
+    {
+      "epoch": 2.9392,
+      "grad_norm": 1.7569412286086807,
+      "learning_rate": 2.4220456090952027e-07,
+      "loss": 1.0761,
+      "step": 18370
+    },
+    {
+      "epoch": 2.9393599999999998,
+      "grad_norm": 1.7187497620766987,
+      "learning_rate": 2.4100452693747167e-07,
+      "loss": 1.062,
+      "step": 18371
+    },
+    {
+      "epoch": 2.93952,
+      "grad_norm": 1.6884192297226117,
+      "learning_rate": 2.3980746962849464e-07,
+      "loss": 1.0626,
+      "step": 18372
+    },
+    {
+      "epoch": 2.93968,
+      "grad_norm": 1.9277859303505693,
+      "learning_rate": 2.38613389018294e-07,
+      "loss": 1.098,
+      "step": 18373
+    },
+    {
+      "epoch": 2.9398400000000002,
+      "grad_norm": 1.6062349433721737,
+      "learning_rate": 2.3742228514249677e-07,
+      "loss": 1.0347,
+      "step": 18374
+    },
+    {
+      "epoch": 2.94,
+      "grad_norm": 1.6398500638444704,
+      "learning_rate": 2.3623415803665227e-07,
+      "loss": 1.0383,
+      "step": 18375
+    },
+    {
+      "epoch": 2.94016,
+      "grad_norm": 1.6499555097514438,
+      "learning_rate": 2.3504900773620996e-07,
+      "loss": 1.0539,
+      "step": 18376
+    },
+    {
+      "epoch": 2.94032,
+      "grad_norm": 1.6554953056945285,
+      "learning_rate": 2.3386683427653044e-07,
+      "loss": 1.0521,
+      "step": 18377
+    },
+    {
+      "epoch": 2.94048,
+      "grad_norm": 1.838787087669512,
+      "learning_rate": 2.326876376928966e-07,
+      "loss": 1.0805,
+      "step": 18378
+    },
+    {
+      "epoch": 2.94064,
+      "grad_norm": 1.6707791400575513,
+      "learning_rate": 2.3151141802048027e-07,
+      "loss": 1.0311,
+      "step": 18379
+    },
+    {
+      "epoch": 2.9408,
+      "grad_norm": 1.715508454114667,
+      "learning_rate": 2.3033817529438673e-07,
+      "loss": 1.0765,
+      "step": 18380
+    },
+    {
+      "epoch": 2.94096,
+      "grad_norm": 1.541914728153115,
+      "learning_rate": 2.291679095496213e-07,
+      "loss": 1.0697,
+      "step": 18381
+    },
+    {
+      "epoch": 2.9411199999999997,
+      "grad_norm": 1.873360761284708,
+      "learning_rate": 2.2800062082111162e-07,
+      "loss": 1.0539,
+      "step": 18382
+    },
+    {
+      "epoch": 2.94128,
+      "grad_norm": 1.7148708454006492,
+      "learning_rate": 2.2683630914367427e-07,
+      "loss": 1.0041,
+      "step": 18383
+    },
+    {
+      "epoch": 2.94144,
+      "grad_norm": 1.830186806318246,
+      "learning_rate": 2.256749745520481e-07,
+      "loss": 1.0991,
+      "step": 18384
+    },
+    {
+      "epoch": 2.9416,
+      "grad_norm": 1.6558961189568848,
+      "learning_rate": 2.2451661708090542e-07,
+      "loss": 1.0697,
+      "step": 18385
+    },
+    {
+      "epoch": 2.94176,
+      "grad_norm": 1.5002333323191792,
+      "learning_rate": 2.2336123676479636e-07,
+      "loss": 1.0569,
+      "step": 18386
+    },
+    {
+      "epoch": 2.94192,
+      "grad_norm": 1.716540018489498,
+      "learning_rate": 2.2220883363819333e-07,
+      "loss": 1.0014,
+      "step": 18387
+    },
+    {
+      "epoch": 2.94208,
+      "grad_norm": 1.791612759138079,
+      "learning_rate": 2.2105940773549106e-07,
+      "loss": 1.094,
+      "step": 18388
+    },
+    {
+      "epoch": 2.94224,
+      "grad_norm": 1.6223642638991667,
+      "learning_rate": 2.199129590909843e-07,
+      "loss": 1.0323,
+      "step": 18389
+    },
+    {
+      "epoch": 2.9424,
+      "grad_norm": 1.8118386177582229,
+      "learning_rate": 2.1876948773887905e-07,
+      "loss": 1.0927,
+      "step": 18390
+    },
+    {
+      "epoch": 2.94256,
+      "grad_norm": 1.672505269792672,
+      "learning_rate": 2.1762899371329248e-07,
+      "loss": 1.0648,
+      "step": 18391
+    },
+    {
+      "epoch": 2.94272,
+      "grad_norm": 2.0676530202662846,
+      "learning_rate": 2.16491477048264e-07,
+      "loss": 1.1122,
+      "step": 18392
+    },
+    {
+      "epoch": 2.94288,
+      "grad_norm": 1.8727426368042592,
+      "learning_rate": 2.1535693777773313e-07,
+      "loss": 1.06,
+      "step": 18393
+    },
+    {
+      "epoch": 2.94304,
+      "grad_norm": 1.7970261379115275,
+      "learning_rate": 2.1422537593553948e-07,
+      "loss": 1.0468,
+      "step": 18394
+    },
+    {
+      "epoch": 2.9432,
+      "grad_norm": 1.8552491359555292,
+      "learning_rate": 2.1309679155545604e-07,
+      "loss": 1.0642,
+      "step": 18395
+    },
+    {
+      "epoch": 2.94336,
+      "grad_norm": 1.5714069943009423,
+      "learning_rate": 2.1197118467116696e-07,
+      "loss": 1.0347,
+      "step": 18396
+    },
+    {
+      "epoch": 2.94352,
+      "grad_norm": 1.8600210812406497,
+      "learning_rate": 2.1084855531625646e-07,
+      "loss": 1.0348,
+      "step": 18397
+    },
+    {
+      "epoch": 2.94368,
+      "grad_norm": 1.4447281349525896,
+      "learning_rate": 2.0972890352420892e-07,
+      "loss": 1.0711,
+      "step": 18398
+    },
+    {
+      "epoch": 2.94384,
+      "grad_norm": 1.7536564425515364,
+      "learning_rate": 2.0861222932844205e-07,
+      "loss": 1.0213,
+      "step": 18399
+    },
+    {
+      "epoch": 2.944,
+      "grad_norm": 1.7427856688766008,
+      "learning_rate": 2.074985327622736e-07,
+      "loss": 1.0395,
+      "step": 18400
+    },
+    {
+      "epoch": 2.94416,
+      "grad_norm": 1.922825092891001,
+      "learning_rate": 2.063878138589326e-07,
+      "loss": 1.0123,
+      "step": 18401
+    },
+    {
+      "epoch": 2.9443200000000003,
+      "grad_norm": 1.8062136293613527,
+      "learning_rate": 2.0528007265157024e-07,
+      "loss": 1.0618,
+      "step": 18402
+    },
+    {
+      "epoch": 2.94448,
+      "grad_norm": 1.70792259802009,
+      "learning_rate": 2.041753091732379e-07,
+      "loss": 1.1121,
+      "step": 18403
+    },
+    {
+      "epoch": 2.94464,
+      "grad_norm": 1.6923743161899574,
+      "learning_rate": 2.03073523456887e-07,
+      "loss": 1.1153,
+      "step": 18404
+    },
+    {
+      "epoch": 2.9448,
+      "grad_norm": 1.6225740834846938,
+      "learning_rate": 2.0197471553540236e-07,
+      "loss": 1.0393,
+      "step": 18405
+    },
+    {
+      "epoch": 2.94496,
+      "grad_norm": 1.708553105322896,
+      "learning_rate": 2.0087888544157997e-07,
+      "loss": 1.1275,
+      "step": 18406
+    },
+    {
+      "epoch": 2.94512,
+      "grad_norm": 1.702623253320345,
+      "learning_rate": 1.9978603320810475e-07,
+      "loss": 1.0678,
+      "step": 18407
+    },
+    {
+      "epoch": 2.94528,
+      "grad_norm": 1.7804986686696578,
+      "learning_rate": 1.9869615886759508e-07,
+      "loss": 1.0659,
+      "step": 18408
+    },
+    {
+      "epoch": 2.94544,
+      "grad_norm": 1.7164678887158127,
+      "learning_rate": 1.976092624525583e-07,
+      "loss": 1.1082,
+      "step": 18409
+    },
+    {
+      "epoch": 2.9455999999999998,
+      "grad_norm": 1.8637002392344375,
+      "learning_rate": 1.965253439954351e-07,
+      "loss": 1.0494,
+      "step": 18410
+    },
+    {
+      "epoch": 2.94576,
+      "grad_norm": 1.870115644894585,
+      "learning_rate": 1.954444035285663e-07,
+      "loss": 1.0391,
+      "step": 18411
+    },
+    {
+      "epoch": 2.94592,
+      "grad_norm": 1.6198405194628056,
+      "learning_rate": 1.9436644108420388e-07,
+      "loss": 1.0396,
+      "step": 18412
+    },
+    {
+      "epoch": 2.9460800000000003,
+      "grad_norm": 1.7885176621033319,
+      "learning_rate": 1.932914566945221e-07,
+      "loss": 1.0769,
+      "step": 18413
+    },
+    {
+      "epoch": 2.94624,
+      "grad_norm": 1.6739050378020321,
+      "learning_rate": 1.9221945039158418e-07,
+      "loss": 1.0774,
+      "step": 18414
+    },
+    {
+      "epoch": 2.9464,
+      "grad_norm": 1.735795126109619,
+      "learning_rate": 1.9115042220738676e-07,
+      "loss": 1.0517,
+      "step": 18415
+    },
+    {
+      "epoch": 2.94656,
+      "grad_norm": 1.7535506917183126,
+      "learning_rate": 1.9008437217381547e-07,
+      "loss": 1.0129,
+      "step": 18416
+    },
+    {
+      "epoch": 2.94672,
+      "grad_norm": 1.7025092883359698,
+      "learning_rate": 1.8902130032270037e-07,
+      "loss": 1.0622,
+      "step": 18417
+    },
+    {
+      "epoch": 2.94688,
+      "grad_norm": 1.8746032758042974,
+      "learning_rate": 1.8796120668573834e-07,
+      "loss": 1.1045,
+      "step": 18418
+    },
+    {
+      "epoch": 2.94704,
+      "grad_norm": 1.7396127375451849,
+      "learning_rate": 1.8690409129458186e-07,
+      "loss": 1.0644,
+      "step": 18419
+    },
+    {
+      "epoch": 2.9472,
+      "grad_norm": 1.6161919017802324,
+      "learning_rate": 1.8584995418076123e-07,
+      "loss": 1.0721,
+      "step": 18420
+    },
+    {
+      "epoch": 2.9473599999999998,
+      "grad_norm": 1.7210463017820905,
+      "learning_rate": 1.8479879537572908e-07,
+      "loss": 1.009,
+      "step": 18421
+    },
+    {
+      "epoch": 2.94752,
+      "grad_norm": 1.6207175587980862,
+      "learning_rate": 1.837506149108603e-07,
+      "loss": 1.0659,
+      "step": 18422
+    },
+    {
+      "epoch": 2.94768,
+      "grad_norm": 1.7622081238074245,
+      "learning_rate": 1.8270541281741883e-07,
+      "loss": 1.0329,
+      "step": 18423
+    },
+    {
+      "epoch": 2.9478400000000002,
+      "grad_norm": 1.6131007889622528,
+      "learning_rate": 1.8166318912660186e-07,
+      "loss": 1.0566,
+      "step": 18424
+    },
+    {
+      "epoch": 2.948,
+      "grad_norm": 1.7664080012136976,
+      "learning_rate": 1.8062394386950676e-07,
+      "loss": 1.0621,
+      "step": 18425
+    },
+    {
+      "epoch": 2.94816,
+      "grad_norm": 1.8206684500858796,
+      "learning_rate": 1.7958767707714207e-07,
+      "loss": 1.0444,
+      "step": 18426
+    },
+    {
+      "epoch": 2.94832,
+      "grad_norm": 1.7660353640660271,
+      "learning_rate": 1.7855438878042752e-07,
+      "loss": 1.1163,
+      "step": 18427
+    },
+    {
+      "epoch": 2.94848,
+      "grad_norm": 1.7628500538210816,
+      "learning_rate": 1.7752407901018286e-07,
+      "loss": 1.0673,
+      "step": 18428
+    },
+    {
+      "epoch": 2.94864,
+      "grad_norm": 1.683641695704978,
+      "learning_rate": 1.764967477971724e-07,
+      "loss": 1.0625,
+      "step": 18429
+    },
+    {
+      "epoch": 2.9488,
+      "grad_norm": 1.763695572971801,
+      "learning_rate": 1.7547239517204939e-07,
+      "loss": 1.0832,
+      "step": 18430
+    },
+    {
+      "epoch": 2.94896,
+      "grad_norm": 1.7507606314555284,
+      "learning_rate": 1.7445102116535605e-07,
+      "loss": 1.0734,
+      "step": 18431
+    },
+    {
+      "epoch": 2.9491199999999997,
+      "grad_norm": 1.774082587183648,
+      "learning_rate": 1.7343262580759023e-07,
+      "loss": 1.0681,
+      "step": 18432
+    },
+    {
+      "epoch": 2.94928,
+      "grad_norm": 1.6798471303609066,
+      "learning_rate": 1.7241720912912762e-07,
+      "loss": 1.0664,
+      "step": 18433
+    },
+    {
+      "epoch": 2.94944,
+      "grad_norm": 1.7534968984990333,
+      "learning_rate": 1.714047711602773e-07,
+      "loss": 1.0755,
+      "step": 18434
+    },
+    {
+      "epoch": 2.9496,
+      "grad_norm": 2.149416969124901,
+      "learning_rate": 1.7039531193123737e-07,
+      "loss": 1.1163,
+      "step": 18435
+    },
+    {
+      "epoch": 2.94976,
+      "grad_norm": 1.6283556350479917,
+      "learning_rate": 1.6938883147213924e-07,
+      "loss": 1.0644,
+      "step": 18436
+    },
+    {
+      "epoch": 2.94992,
+      "grad_norm": 1.6563681697903203,
+      "learning_rate": 1.6838532981300336e-07,
+      "loss": 1.0481,
+      "step": 18437
+    },
+    {
+      "epoch": 2.95008,
+      "grad_norm": 1.7295656352868711,
+      "learning_rate": 1.6738480698378356e-07,
+      "loss": 1.1223,
+      "step": 18438
+    },
+    {
+      "epoch": 2.95024,
+      "grad_norm": 1.661993129086454,
+      "learning_rate": 1.6638726301433372e-07,
+      "loss": 1.0426,
+      "step": 18439
+    },
+    {
+      "epoch": 2.9504,
+      "grad_norm": 1.7751717439485841,
+      "learning_rate": 1.6539269793440783e-07,
+      "loss": 1.082,
+      "step": 18440
+    },
+    {
+      "epoch": 2.95056,
+      "grad_norm": 1.7575416376040083,
+      "learning_rate": 1.644011117736932e-07,
+      "loss": 1.1177,
+      "step": 18441
+    },
+    {
+      "epoch": 2.95072,
+      "grad_norm": 1.818182040369806,
+      "learning_rate": 1.6341250456176626e-07,
+      "loss": 1.0369,
+      "step": 18442
+    },
+    {
+      "epoch": 2.9508799999999997,
+      "grad_norm": 1.6814136451371244,
+      "learning_rate": 1.624268763281478e-07,
+      "loss": 1.0579,
+      "step": 18443
+    },
+    {
+      "epoch": 2.95104,
+      "grad_norm": 1.89059046327249,
+      "learning_rate": 1.614442271022254e-07,
+      "loss": 1.0585,
+      "step": 18444
+    },
+    {
+      "epoch": 2.9512,
+      "grad_norm": 1.559807908812348,
+      "learning_rate": 1.6046455691332008e-07,
+      "loss": 1.089,
+      "step": 18445
+    },
+    {
+      "epoch": 2.95136,
+      "grad_norm": 1.874273736221127,
+      "learning_rate": 1.594878657906751e-07,
+      "loss": 1.0678,
+      "step": 18446
+    },
+    {
+      "epoch": 2.95152,
+      "grad_norm": 1.7836149460148036,
+      "learning_rate": 1.5851415376343382e-07,
+      "loss": 1.0865,
+      "step": 18447
+    },
+    {
+      "epoch": 2.95168,
+      "grad_norm": 1.777127412677148,
+      "learning_rate": 1.5754342086063967e-07,
+      "loss": 1.048,
+      "step": 18448
+    },
+    {
+      "epoch": 2.95184,
+      "grad_norm": 1.9004496944596316,
+      "learning_rate": 1.5657566711126947e-07,
+      "loss": 1.0487,
+      "step": 18449
+    },
+    {
+      "epoch": 2.952,
+      "grad_norm": 1.6852933579502771,
+      "learning_rate": 1.5561089254418903e-07,
+      "loss": 1.1235,
+      "step": 18450
+    },
+    {
+      "epoch": 2.95216,
+      "grad_norm": 1.5920924075088516,
+      "learning_rate": 1.5464909718819754e-07,
+      "loss": 1.0649,
+      "step": 18451
+    },
+    {
+      "epoch": 2.9523200000000003,
+      "grad_norm": 1.7125058281295975,
+      "learning_rate": 1.5369028107198314e-07,
+      "loss": 1.0609,
+      "step": 18452
+    },
+    {
+      "epoch": 2.95248,
+      "grad_norm": 1.6755853388782662,
+      "learning_rate": 1.5273444422415628e-07,
+      "loss": 1.0588,
+      "step": 18453
+    },
+    {
+      "epoch": 2.95264,
+      "grad_norm": 1.7194912964784663,
+      "learning_rate": 1.517815866732386e-07,
+      "loss": 1.0514,
+      "step": 18454
+    },
+    {
+      "epoch": 2.9528,
+      "grad_norm": 1.7386244813098093,
+      "learning_rate": 1.5083170844767402e-07,
+      "loss": 1.0985,
+      "step": 18455
+    },
+    {
+      "epoch": 2.95296,
+      "grad_norm": 1.6736193764823701,
+      "learning_rate": 1.4988480957578432e-07,
+      "loss": 1.0522,
+      "step": 18456
+    },
+    {
+      "epoch": 2.95312,
+      "grad_norm": 1.6004793703664413,
+      "learning_rate": 1.4894089008584688e-07,
+      "loss": 1.0293,
+      "step": 18457
+    },
+    {
+      "epoch": 2.95328,
+      "grad_norm": 2.103999807503354,
+      "learning_rate": 1.4799995000600587e-07,
+      "loss": 1.1167,
+      "step": 18458
+    },
+    {
+      "epoch": 2.95344,
+      "grad_norm": 1.845542132909247,
+      "learning_rate": 1.4706198936434988e-07,
+      "loss": 1.1328,
+      "step": 18459
+    },
+    {
+      "epoch": 2.9536,
+      "grad_norm": 1.7475462512565423,
+      "learning_rate": 1.461270081888677e-07,
+      "loss": 1.0398,
+      "step": 18460
+    },
+    {
+      "epoch": 2.95376,
+      "grad_norm": 1.5524154877511764,
+      "learning_rate": 1.4519500650744811e-07,
+      "loss": 1.0123,
+      "step": 18461
+    },
+    {
+      "epoch": 2.95392,
+      "grad_norm": 1.828392716839928,
+      "learning_rate": 1.4426598434790218e-07,
+      "loss": 1.0642,
+      "step": 18462
+    },
+    {
+      "epoch": 2.9540800000000003,
+      "grad_norm": 1.565844490092529,
+      "learning_rate": 1.4333994173796328e-07,
+      "loss": 1.0608,
+      "step": 18463
+    },
+    {
+      "epoch": 2.95424,
+      "grad_norm": 1.682246862716771,
+      "learning_rate": 1.4241687870524268e-07,
+      "loss": 1.055,
+      "step": 18464
+    },
+    {
+      "epoch": 2.9544,
+      "grad_norm": 1.7106613949776468,
+      "learning_rate": 1.414967952773072e-07,
+      "loss": 1.0063,
+      "step": 18465
+    },
+    {
+      "epoch": 2.95456,
+      "grad_norm": 1.8056516927530875,
+      "learning_rate": 1.4057969148159045e-07,
+      "loss": 1.0312,
+      "step": 18466
+    },
+    {
+      "epoch": 2.95472,
+      "grad_norm": 1.7183345663362954,
+      "learning_rate": 1.3966556734545944e-07,
+      "loss": 1.0915,
+      "step": 18467
+    },
+    {
+      "epoch": 2.95488,
+      "grad_norm": 1.6503142905507528,
+      "learning_rate": 1.3875442289620344e-07,
+      "loss": 1.0698,
+      "step": 18468
+    },
+    {
+      "epoch": 2.95504,
+      "grad_norm": 1.8335187343908788,
+      "learning_rate": 1.3784625816100073e-07,
+      "loss": 1.0607,
+      "step": 18469
+    },
+    {
+      "epoch": 2.9552,
+      "grad_norm": 1.5562418192234233,
+      "learning_rate": 1.3694107316695182e-07,
+      "loss": 1.0904,
+      "step": 18470
+    },
+    {
+      "epoch": 2.9553599999999998,
+      "grad_norm": 1.952779619621553,
+      "learning_rate": 1.3603886794106847e-07,
+      "loss": 1.0443,
+      "step": 18471
+    },
+    {
+      "epoch": 2.95552,
+      "grad_norm": 1.651241727848978,
+      "learning_rate": 1.3513964251026244e-07,
+      "loss": 1.0481,
+      "step": 18472
+    },
+    {
+      "epoch": 2.95568,
+      "grad_norm": 1.7577271025000645,
+      "learning_rate": 1.34243396901379e-07,
+      "loss": 1.0242,
+      "step": 18473
+    },
+    {
+      "epoch": 2.9558400000000002,
+      "grad_norm": 1.690692048021979,
+      "learning_rate": 1.3335013114114114e-07,
+      "loss": 1.0597,
+      "step": 18474
+    },
+    {
+      "epoch": 2.956,
+      "grad_norm": 1.9298503221007954,
+      "learning_rate": 1.3245984525622758e-07,
+      "loss": 1.1106,
+      "step": 18475
+    },
+    {
+      "epoch": 2.95616,
+      "grad_norm": 1.759370533077804,
+      "learning_rate": 1.315725392731837e-07,
+      "loss": 1.0846,
+      "step": 18476
+    },
+    {
+      "epoch": 2.95632,
+      "grad_norm": 1.6677066747548916,
+      "learning_rate": 1.3068821321848835e-07,
+      "loss": 1.039,
+      "step": 18477
+    },
+    {
+      "epoch": 2.95648,
+      "grad_norm": 1.7174071984569335,
+      "learning_rate": 1.2980686711853153e-07,
+      "loss": 1.0278,
+      "step": 18478
+    },
+    {
+      "epoch": 2.95664,
+      "grad_norm": 2.0076055746864845,
+      "learning_rate": 1.289285009996144e-07,
+      "loss": 1.0645,
+      "step": 18479
+    },
+    {
+      "epoch": 2.9568,
+      "grad_norm": 1.7658520147710992,
+      "learning_rate": 1.2805311488794936e-07,
+      "loss": 1.0984,
+      "step": 18480
+    },
+    {
+      "epoch": 2.95696,
+      "grad_norm": 1.6561122534775046,
+      "learning_rate": 1.2718070880963772e-07,
+      "loss": 1.0261,
+      "step": 18481
+    },
+    {
+      "epoch": 2.9571199999999997,
+      "grad_norm": 1.652193329105997,
+      "learning_rate": 1.2631128279073645e-07,
+      "loss": 1.0349,
+      "step": 18482
+    },
+    {
+      "epoch": 2.95728,
+      "grad_norm": 1.7215163751650098,
+      "learning_rate": 1.2544483685716924e-07,
+      "loss": 1.0701,
+      "step": 18483
+    },
+    {
+      "epoch": 2.95744,
+      "grad_norm": 1.5361548560189355,
+      "learning_rate": 1.2458137103479317e-07,
+      "loss": 1.0535,
+      "step": 18484
+    },
+    {
+      "epoch": 2.9576000000000002,
+      "grad_norm": 1.7987311139640965,
+      "learning_rate": 1.2372088534937653e-07,
+      "loss": 1.0415,
+      "step": 18485
+    },
+    {
+      "epoch": 2.95776,
+      "grad_norm": 1.630708896167641,
+      "learning_rate": 1.228633798265877e-07,
+      "loss": 1.0192,
+      "step": 18486
+    },
+    {
+      "epoch": 2.95792,
+      "grad_norm": 1.6008292637128156,
+      "learning_rate": 1.2200885449202838e-07,
+      "loss": 1.0452,
+      "step": 18487
+    },
+    {
+      "epoch": 2.95808,
+      "grad_norm": 1.7740747600235756,
+      "learning_rate": 1.211573093711893e-07,
+      "loss": 1.0544,
+      "step": 18488
+    },
+    {
+      "epoch": 2.95824,
+      "grad_norm": 1.6125737934342188,
+      "learning_rate": 1.2030874448947238e-07,
+      "loss": 1.0287,
+      "step": 18489
+    },
+    {
+      "epoch": 2.9584,
+      "grad_norm": 1.735569996475905,
+      "learning_rate": 1.194631598722018e-07,
+      "loss": 1.0288,
+      "step": 18490
+    },
+    {
+      "epoch": 2.95856,
+      "grad_norm": 1.6620480409690357,
+      "learning_rate": 1.1862055554460183e-07,
+      "loss": 1.0701,
+      "step": 18491
+    },
+    {
+      "epoch": 2.95872,
+      "grad_norm": 1.6195845603127907,
+      "learning_rate": 1.1778093153184122e-07,
+      "loss": 1.0992,
+      "step": 18492
+    },
+    {
+      "epoch": 2.9588799999999997,
+      "grad_norm": 1.8317436749403104,
+      "learning_rate": 1.169442878589333e-07,
+      "loss": 1.0606,
+      "step": 18493
+    },
+    {
+      "epoch": 2.95904,
+      "grad_norm": 1.8310967640532516,
+      "learning_rate": 1.1611062455088029e-07,
+      "loss": 1.0552,
+      "step": 18494
+    },
+    {
+      "epoch": 2.9592,
+      "grad_norm": 1.5468130459069818,
+      "learning_rate": 1.1527994163252898e-07,
+      "loss": 1.0429,
+      "step": 18495
+    },
+    {
+      "epoch": 2.95936,
+      "grad_norm": 1.6761111601393646,
+      "learning_rate": 1.1445223912868175e-07,
+      "loss": 1.0556,
+      "step": 18496
+    },
+    {
+      "epoch": 2.95952,
+      "grad_norm": 1.7072261213863626,
+      "learning_rate": 1.1362751706402996e-07,
+      "loss": 1.0856,
+      "step": 18497
+    },
+    {
+      "epoch": 2.95968,
+      "grad_norm": 2.0104065019507456,
+      "learning_rate": 1.1280577546317617e-07,
+      "loss": 1.0974,
+      "step": 18498
+    },
+    {
+      "epoch": 2.95984,
+      "grad_norm": 1.6322771312483733,
+      "learning_rate": 1.1198701435065628e-07,
+      "loss": 1.0686,
+      "step": 18499
+    },
+    {
+      "epoch": 2.96,
+      "grad_norm": 1.6493166060266595,
+      "learning_rate": 1.1117123375088411e-07,
+      "loss": 1.0568,
+      "step": 18500
+    },
+    {
+      "epoch": 2.96016,
+      "grad_norm": 1.5729056842638525,
+      "learning_rate": 1.1035843368821797e-07,
+      "loss": 1.0447,
+      "step": 18501
+    },
+    {
+      "epoch": 2.96032,
+      "grad_norm": 1.7425158326112389,
+      "learning_rate": 1.09548614186894e-07,
+      "loss": 1.0404,
+      "step": 18502
+    },
+    {
+      "epoch": 2.96048,
+      "grad_norm": 1.8050342723743706,
+      "learning_rate": 1.0874177527109285e-07,
+      "loss": 1.1021,
+      "step": 18503
+    },
+    {
+      "epoch": 2.96064,
+      "grad_norm": 1.8749142157235856,
+      "learning_rate": 1.079379169648731e-07,
+      "loss": 1.0625,
+      "step": 18504
+    },
+    {
+      "epoch": 2.9608,
+      "grad_norm": 1.5924045291313407,
+      "learning_rate": 1.0713703929222663e-07,
+      "loss": 1.08,
+      "step": 18505
+    },
+    {
+      "epoch": 2.96096,
+      "grad_norm": 2.073216482065503,
+      "learning_rate": 1.0633914227705655e-07,
+      "loss": 1.0808,
+      "step": 18506
+    },
+    {
+      "epoch": 2.96112,
+      "grad_norm": 2.0448646134905117,
+      "learning_rate": 1.0554422594315494e-07,
+      "loss": 1.0775,
+      "step": 18507
+    },
+    {
+      "epoch": 2.96128,
+      "grad_norm": 1.6446251128814144,
+      "learning_rate": 1.0475229031426948e-07,
+      "loss": 1.1116,
+      "step": 18508
+    },
+    {
+      "epoch": 2.96144,
+      "grad_norm": 1.7624089192962513,
+      "learning_rate": 1.039633354140035e-07,
+      "loss": 1.0459,
+      "step": 18509
+    },
+    {
+      "epoch": 2.9616,
+      "grad_norm": 1.5374212331078405,
+      "learning_rate": 1.0317736126590483e-07,
+      "loss": 1.0986,
+      "step": 18510
+    },
+    {
+      "epoch": 2.96176,
+      "grad_norm": 2.005112262953835,
+      "learning_rate": 1.0239436789343249e-07,
+      "loss": 1.0993,
+      "step": 18511
+    },
+    {
+      "epoch": 2.96192,
+      "grad_norm": 1.6648778296092275,
+      "learning_rate": 1.0161435531994556e-07,
+      "loss": 1.0956,
+      "step": 18512
+    },
+    {
+      "epoch": 2.9620800000000003,
+      "grad_norm": 1.669274810120546,
+      "learning_rate": 1.0083732356871433e-07,
+      "loss": 1.0056,
+      "step": 18513
+    },
+    {
+      "epoch": 2.96224,
+      "grad_norm": 1.7043470278827537,
+      "learning_rate": 1.0006327266293136e-07,
+      "loss": 1.0316,
+      "step": 18514
+    },
+    {
+      "epoch": 2.9624,
+      "grad_norm": 1.5862117609687418,
+      "learning_rate": 9.929220262570038e-08,
+      "loss": 1.0514,
+      "step": 18515
+    },
+    {
+      "epoch": 2.96256,
+      "grad_norm": 1.5432661902599851,
+      "learning_rate": 9.852411348000301e-08,
+      "loss": 1.0752,
+      "step": 18516
+    },
+    {
+      "epoch": 2.96272,
+      "grad_norm": 1.6042002964852151,
+      "learning_rate": 9.775900524878757e-08,
+      "loss": 1.0685,
+      "step": 18517
+    },
+    {
+      "epoch": 2.96288,
+      "grad_norm": 1.8313075386226585,
+      "learning_rate": 9.699687795486912e-08,
+      "loss": 1.0719,
+      "step": 18518
+    },
+    {
+      "epoch": 2.96304,
+      "grad_norm": 1.6917882006364635,
+      "learning_rate": 9.623773162098503e-08,
+      "loss": 1.0288,
+      "step": 18519
+    },
+    {
+      "epoch": 2.9632,
+      "grad_norm": 1.7729941316188278,
+      "learning_rate": 9.548156626979498e-08,
+      "loss": 1.0954,
+      "step": 18520
+    },
+    {
+      "epoch": 2.9633599999999998,
+      "grad_norm": 1.6791618467114802,
+      "learning_rate": 9.472838192384759e-08,
+      "loss": 1.1078,
+      "step": 18521
+    },
+    {
+      "epoch": 2.96352,
+      "grad_norm": 1.869661571493202,
+      "learning_rate": 9.397817860564707e-08,
+      "loss": 1.01,
+      "step": 18522
+    },
+    {
+      "epoch": 2.96368,
+      "grad_norm": 1.655322727486473,
+      "learning_rate": 9.32309563375422e-08,
+      "loss": 1.0795,
+      "step": 18523
+    },
+    {
+      "epoch": 2.9638400000000003,
+      "grad_norm": 1.8089235140696462,
+      "learning_rate": 9.248671514185958e-08,
+      "loss": 1.1171,
+      "step": 18524
+    },
+    {
+      "epoch": 2.964,
+      "grad_norm": 1.5660154991933453,
+      "learning_rate": 9.174545504078147e-08,
+      "loss": 1.0309,
+      "step": 18525
+    },
+    {
+      "epoch": 2.96416,
+      "grad_norm": 1.7416143688363757,
+      "learning_rate": 9.100717605644571e-08,
+      "loss": 1.066,
+      "step": 18526
+    },
+    {
+      "epoch": 2.96432,
+      "grad_norm": 1.7937196666863606,
+      "learning_rate": 9.027187821086802e-08,
+      "loss": 1.0404,
+      "step": 18527
+    },
+    {
+      "epoch": 2.96448,
+      "grad_norm": 1.8150679841656803,
+      "learning_rate": 8.953956152599752e-08,
+      "loss": 1.0744,
+      "step": 18528
+    },
+    {
+      "epoch": 2.96464,
+      "grad_norm": 1.528187925768953,
+      "learning_rate": 8.881022602367228e-08,
+      "loss": 1.0713,
+      "step": 18529
+    },
+    {
+      "epoch": 2.9648,
+      "grad_norm": 1.736948912617879,
+      "learning_rate": 8.808387172567489e-08,
+      "loss": 1.0679,
+      "step": 18530
+    },
+    {
+      "epoch": 2.96496,
+      "grad_norm": 1.5761937070857006,
+      "learning_rate": 8.73604986536658e-08,
+      "loss": 1.0439,
+      "step": 18531
+    },
+    {
+      "epoch": 2.9651199999999998,
+      "grad_norm": 1.9060571153019392,
+      "learning_rate": 8.664010682922775e-08,
+      "loss": 1.0966,
+      "step": 18532
+    },
+    {
+      "epoch": 2.96528,
+      "grad_norm": 1.6782652940913008,
+      "learning_rate": 8.592269627385464e-08,
+      "loss": 1.0828,
+      "step": 18533
+    },
+    {
+      "epoch": 2.96544,
+      "grad_norm": 1.906925582527806,
+      "learning_rate": 8.520826700897378e-08,
+      "loss": 1.0653,
+      "step": 18534
+    },
+    {
+      "epoch": 2.9656000000000002,
+      "grad_norm": 1.855393793649198,
+      "learning_rate": 8.449681905587925e-08,
+      "loss": 1.1301,
+      "step": 18535
+    },
+    {
+      "epoch": 2.96576,
+      "grad_norm": 1.676252613932323,
+      "learning_rate": 8.378835243579852e-08,
+      "loss": 1.0518,
+      "step": 18536
+    },
+    {
+      "epoch": 2.96592,
+      "grad_norm": 1.7574093753282898,
+      "learning_rate": 8.308286716989244e-08,
+      "loss": 1.0973,
+      "step": 18537
+    },
+    {
+      "epoch": 2.96608,
+      "grad_norm": 1.540209017580875,
+      "learning_rate": 8.238036327918863e-08,
+      "loss": 1.076,
+      "step": 18538
+    },
+    {
+      "epoch": 2.96624,
+      "grad_norm": 1.6156814276919504,
+      "learning_rate": 8.168084078467031e-08,
+      "loss": 1.0508,
+      "step": 18539
+    },
+    {
+      "epoch": 2.9664,
+      "grad_norm": 1.6717671074549825,
+      "learning_rate": 8.098429970718746e-08,
+      "loss": 1.1137,
+      "step": 18540
+    },
+    {
+      "epoch": 2.96656,
+      "grad_norm": 1.6992355856510166,
+      "learning_rate": 8.029074006754566e-08,
+      "loss": 0.9973,
+      "step": 18541
+    },
+    {
+      "epoch": 2.96672,
+      "grad_norm": 1.7851677871908485,
+      "learning_rate": 7.960016188642838e-08,
+      "loss": 0.9937,
+      "step": 18542
+    },
+    {
+      "epoch": 2.9668799999999997,
+      "grad_norm": 1.7857349687074258,
+      "learning_rate": 7.891256518444134e-08,
+      "loss": 1.0454,
+      "step": 18543
+    },
+    {
+      "epoch": 2.96704,
+      "grad_norm": 1.6204881305050767,
+      "learning_rate": 7.822794998210147e-08,
+      "loss": 1.065,
+      "step": 18544
+    },
+    {
+      "epoch": 2.9672,
+      "grad_norm": 1.7135724913403192,
+      "learning_rate": 7.754631629983688e-08,
+      "loss": 1.0986,
+      "step": 18545
+    },
+    {
+      "epoch": 2.96736,
+      "grad_norm": 1.6181123662456582,
+      "learning_rate": 7.686766415798685e-08,
+      "loss": 1.0981,
+      "step": 18546
+    },
+    {
+      "epoch": 2.96752,
+      "grad_norm": 1.562881480015165,
+      "learning_rate": 7.619199357681295e-08,
+      "loss": 1.0857,
+      "step": 18547
+    },
+    {
+      "epoch": 2.96768,
+      "grad_norm": 1.8919008483940256,
+      "learning_rate": 7.551930457645462e-08,
+      "loss": 1.0387,
+      "step": 18548
+    },
+    {
+      "epoch": 2.96784,
+      "grad_norm": 1.8138006374102227,
+      "learning_rate": 7.48495971769958e-08,
+      "loss": 1.0731,
+      "step": 18549
+    },
+    {
+      "epoch": 2.968,
+      "grad_norm": 1.7835409798680915,
+      "learning_rate": 7.418287139842051e-08,
+      "loss": 1.0868,
+      "step": 18550
+    },
+    {
+      "epoch": 2.96816,
+      "grad_norm": 1.5937601490371809,
+      "learning_rate": 7.351912726062393e-08,
+      "loss": 1.039,
+      "step": 18551
+    },
+    {
+      "epoch": 2.96832,
+      "grad_norm": 1.6504534696095274,
+      "learning_rate": 7.285836478341246e-08,
+      "loss": 1.0402,
+      "step": 18552
+    },
+    {
+      "epoch": 2.96848,
+      "grad_norm": 1.7376012554941695,
+      "learning_rate": 7.220058398649254e-08,
+      "loss": 1.0899,
+      "step": 18553
+    },
+    {
+      "epoch": 2.9686399999999997,
+      "grad_norm": 1.9109615657924823,
+      "learning_rate": 7.154578488949293e-08,
+      "loss": 1.0884,
+      "step": 18554
+    },
+    {
+      "epoch": 2.9688,
+      "grad_norm": 1.7230334163672874,
+      "learning_rate": 7.089396751195354e-08,
+      "loss": 1.0706,
+      "step": 18555
+    },
+    {
+      "epoch": 2.96896,
+      "grad_norm": 1.6367141219559609,
+      "learning_rate": 7.024513187333659e-08,
+      "loss": 1.0504,
+      "step": 18556
+    },
+    {
+      "epoch": 2.96912,
+      "grad_norm": 1.8018613920048985,
+      "learning_rate": 6.959927799298216e-08,
+      "loss": 1.0553,
+      "step": 18557
+    },
+    {
+      "epoch": 2.96928,
+      "grad_norm": 1.7332112714214993,
+      "learning_rate": 6.895640589018593e-08,
+      "loss": 1.0626,
+      "step": 18558
+    },
+    {
+      "epoch": 2.96944,
+      "grad_norm": 1.8559757229435307,
+      "learning_rate": 6.831651558409923e-08,
+      "loss": 1.0769,
+      "step": 18559
+    },
+    {
+      "epoch": 2.9696,
+      "grad_norm": 1.7416067674452673,
+      "learning_rate": 6.767960709384013e-08,
+      "loss": 1.0686,
+      "step": 18560
+    },
+    {
+      "epoch": 2.96976,
+      "grad_norm": 1.8953131382781723,
+      "learning_rate": 6.704568043840454e-08,
+      "loss": 1.1293,
+      "step": 18561
+    },
+    {
+      "epoch": 2.96992,
+      "grad_norm": 1.9341828566275536,
+      "learning_rate": 6.641473563671064e-08,
+      "loss": 1.1142,
+      "step": 18562
+    },
+    {
+      "epoch": 2.9700800000000003,
+      "grad_norm": 1.567111605208282,
+      "learning_rate": 6.578677270757671e-08,
+      "loss": 1.0629,
+      "step": 18563
+    },
+    {
+      "epoch": 2.97024,
+      "grad_norm": 1.605710937114884,
+      "learning_rate": 6.516179166974335e-08,
+      "loss": 1.0574,
+      "step": 18564
+    },
+    {
+      "epoch": 2.9704,
+      "grad_norm": 1.6664712241858686,
+      "learning_rate": 6.453979254187336e-08,
+      "loss": 1.0397,
+      "step": 18565
+    },
+    {
+      "epoch": 2.97056,
+      "grad_norm": 1.9829644592126494,
+      "learning_rate": 6.392077534250751e-08,
+      "loss": 1.0299,
+      "step": 18566
+    },
+    {
+      "epoch": 2.97072,
+      "grad_norm": 1.629907302910119,
+      "learning_rate": 6.330474009011988e-08,
+      "loss": 1.0488,
+      "step": 18567
+    },
+    {
+      "epoch": 2.97088,
+      "grad_norm": 1.7778897608816946,
+      "learning_rate": 6.269168680309579e-08,
+      "loss": 1.0574,
+      "step": 18568
+    },
+    {
+      "epoch": 2.97104,
+      "grad_norm": 1.8092184579685147,
+      "learning_rate": 6.208161549973168e-08,
+      "loss": 1.0301,
+      "step": 18569
+    },
+    {
+      "epoch": 2.9712,
+      "grad_norm": 1.6796436106293495,
+      "learning_rate": 6.147452619822413e-08,
+      "loss": 1.0325,
+      "step": 18570
+    },
+    {
+      "epoch": 2.97136,
+      "grad_norm": 1.7991379225425617,
+      "learning_rate": 6.087041891670308e-08,
+      "loss": 1.0903,
+      "step": 18571
+    },
+    {
+      "epoch": 2.97152,
+      "grad_norm": 1.5186923126814338,
+      "learning_rate": 6.026929367316525e-08,
+      "loss": 1.0503,
+      "step": 18572
+    },
+    {
+      "epoch": 2.97168,
+      "grad_norm": 1.5919185071559125,
+      "learning_rate": 5.967115048557403e-08,
+      "loss": 1.0594,
+      "step": 18573
+    },
+    {
+      "epoch": 2.9718400000000003,
+      "grad_norm": 1.9432825433187357,
+      "learning_rate": 5.907598937175962e-08,
+      "loss": 1.0739,
+      "step": 18574
+    },
+    {
+      "epoch": 2.972,
+      "grad_norm": 1.8360454264931014,
+      "learning_rate": 5.848381034949668e-08,
+      "loss": 1.0486,
+      "step": 18575
+    },
+    {
+      "epoch": 2.97216,
+      "grad_norm": 1.9327151833642315,
+      "learning_rate": 5.789461343644886e-08,
+      "loss": 1.0803,
+      "step": 18576
+    },
+    {
+      "epoch": 2.97232,
+      "grad_norm": 1.7230591615904198,
+      "learning_rate": 5.7308398650190994e-08,
+      "loss": 1.065,
+      "step": 18577
+    },
+    {
+      "epoch": 2.97248,
+      "grad_norm": 1.7428047202651533,
+      "learning_rate": 5.67251660082091e-08,
+      "loss": 1.1235,
+      "step": 18578
+    },
+    {
+      "epoch": 2.97264,
+      "grad_norm": 1.8373334349425001,
+      "learning_rate": 5.614491552792256e-08,
+      "loss": 1.0872,
+      "step": 18579
+    },
+    {
+      "epoch": 2.9728,
+      "grad_norm": 1.6272402109158137,
+      "learning_rate": 5.556764722663976e-08,
+      "loss": 1.1381,
+      "step": 18580
+    },
+    {
+      "epoch": 2.97296,
+      "grad_norm": 1.768567851758852,
+      "learning_rate": 5.4993361121580265e-08,
+      "loss": 1.113,
+      "step": 18581
+    },
+    {
+      "epoch": 2.9731199999999998,
+      "grad_norm": 1.6127048697178825,
+      "learning_rate": 5.442205722988592e-08,
+      "loss": 1.0535,
+      "step": 18582
+    },
+    {
+      "epoch": 2.97328,
+      "grad_norm": 1.738663357925348,
+      "learning_rate": 5.385373556860973e-08,
+      "loss": 1.0679,
+      "step": 18583
+    },
+    {
+      "epoch": 2.97344,
+      "grad_norm": 1.7813874736578177,
+      "learning_rate": 5.328839615469372e-08,
+      "loss": 1.0219,
+      "step": 18584
+    },
+    {
+      "epoch": 2.9736000000000002,
+      "grad_norm": 1.6082807378012827,
+      "learning_rate": 5.272603900501327e-08,
+      "loss": 1.0613,
+      "step": 18585
+    },
+    {
+      "epoch": 2.97376,
+      "grad_norm": 1.8469081598923283,
+      "learning_rate": 5.2166664136343855e-08,
+      "loss": 1.0955,
+      "step": 18586
+    },
+    {
+      "epoch": 2.97392,
+      "grad_norm": 1.6061603628585082,
+      "learning_rate": 5.161027156539433e-08,
+      "loss": 1.0507,
+      "step": 18587
+    },
+    {
+      "epoch": 2.97408,
+      "grad_norm": 1.9182736607056514,
+      "learning_rate": 5.1056861308740324e-08,
+      "loss": 1.067,
+      "step": 18588
+    },
+    {
+      "epoch": 2.97424,
+      "grad_norm": 1.6471597119601082,
+      "learning_rate": 5.0506433382913054e-08,
+      "loss": 1.0495,
+      "step": 18589
+    },
+    {
+      "epoch": 2.9744,
+      "grad_norm": 1.5803986268375305,
+      "learning_rate": 4.995898780434383e-08,
+      "loss": 1.0593,
+      "step": 18590
+    },
+    {
+      "epoch": 2.97456,
+      "grad_norm": 1.7394865654089093,
+      "learning_rate": 4.941452458934182e-08,
+      "loss": 1.0489,
+      "step": 18591
+    },
+    {
+      "epoch": 2.97472,
+      "grad_norm": 1.6469597069044688,
+      "learning_rate": 4.8873043754171786e-08,
+      "loss": 1.0157,
+      "step": 18592
+    },
+    {
+      "epoch": 2.9748799999999997,
+      "grad_norm": 1.6334961380764774,
+      "learning_rate": 4.833454531498749e-08,
+      "loss": 1.077,
+      "step": 18593
+    },
+    {
+      "epoch": 2.97504,
+      "grad_norm": 1.5486782088591093,
+      "learning_rate": 4.7799029287853846e-08,
+      "loss": 1.0503,
+      "step": 18594
+    },
+    {
+      "epoch": 2.9752,
+      "grad_norm": 1.606124705698592,
+      "learning_rate": 4.726649568874697e-08,
+      "loss": 1.059,
+      "step": 18595
+    },
+    {
+      "epoch": 2.9753600000000002,
+      "grad_norm": 1.8356532924442273,
+      "learning_rate": 4.673694453356525e-08,
+      "loss": 1.0738,
+      "step": 18596
+    },
+    {
+      "epoch": 2.97552,
+      "grad_norm": 1.5374658090599347,
+      "learning_rate": 4.6210375838107164e-08,
+      "loss": 1.063,
+      "step": 18597
+    },
+    {
+      "epoch": 2.97568,
+      "grad_norm": 1.8346451055456547,
+      "learning_rate": 4.568678961808237e-08,
+      "loss": 1.026,
+      "step": 18598
+    },
+    {
+      "epoch": 2.97584,
+      "grad_norm": 1.7987716909823275,
+      "learning_rate": 4.51661858891117e-08,
+      "loss": 1.0876,
+      "step": 18599
+    },
+    {
+      "epoch": 2.976,
+      "grad_norm": 1.674397986833735,
+      "learning_rate": 4.464856466673828e-08,
+      "loss": 1.0612,
+      "step": 18600
+    },
+    {
+      "epoch": 2.97616,
+      "grad_norm": 1.921229537053706,
+      "learning_rate": 4.413392596640531e-08,
+      "loss": 1.061,
+      "step": 18601
+    },
+    {
+      "epoch": 2.97632,
+      "grad_norm": 1.6349041301882032,
+      "learning_rate": 4.3622269803456075e-08,
+      "loss": 1.0491,
+      "step": 18602
+    },
+    {
+      "epoch": 2.97648,
+      "grad_norm": 1.7258058688571298,
+      "learning_rate": 4.3113596193167235e-08,
+      "loss": 1.0357,
+      "step": 18603
+    },
+    {
+      "epoch": 2.9766399999999997,
+      "grad_norm": 1.5493814789775966,
+      "learning_rate": 4.2607905150715556e-08,
+      "loss": 1.029,
+      "step": 18604
+    },
+    {
+      "epoch": 2.9768,
+      "grad_norm": 1.811997527850029,
+      "learning_rate": 4.210519669118895e-08,
+      "loss": 1.0665,
+      "step": 18605
+    },
+    {
+      "epoch": 2.97696,
+      "grad_norm": 1.6226596056178606,
+      "learning_rate": 4.1605470829597646e-08,
+      "loss": 1.063,
+      "step": 18606
+    },
+    {
+      "epoch": 2.97712,
+      "grad_norm": 1.4738075647837392,
+      "learning_rate": 4.1108727580840836e-08,
+      "loss": 1.0541,
+      "step": 18607
+    },
+    {
+      "epoch": 2.97728,
+      "grad_norm": 2.0428416842779087,
+      "learning_rate": 4.0614966959739984e-08,
+      "loss": 1.0548,
+      "step": 18608
+    },
+    {
+      "epoch": 2.97744,
+      "grad_norm": 1.7697775846813857,
+      "learning_rate": 4.012418898103887e-08,
+      "loss": 1.118,
+      "step": 18609
+    },
+    {
+      "epoch": 2.9776,
+      "grad_norm": 1.659904134001889,
+      "learning_rate": 3.963639365937022e-08,
+      "loss": 1.0411,
+      "step": 18610
+    },
+    {
+      "epoch": 2.97776,
+      "grad_norm": 1.8159762936507995,
+      "learning_rate": 3.915158100928906e-08,
+      "loss": 1.0545,
+      "step": 18611
+    },
+    {
+      "epoch": 2.97792,
+      "grad_norm": 1.357040427963884,
+      "learning_rate": 3.8669751045272705e-08,
+      "loss": 1.0848,
+      "step": 18612
+    },
+    {
+      "epoch": 2.97808,
+      "grad_norm": 1.5992398300297022,
+      "learning_rate": 3.8190903781698536e-08,
+      "loss": 1.0352,
+      "step": 18613
+    },
+    {
+      "epoch": 2.97824,
+      "grad_norm": 1.6696302116318253,
+      "learning_rate": 3.771503923284403e-08,
+      "loss": 1.0625,
+      "step": 18614
+    },
+    {
+      "epoch": 2.9784,
+      "grad_norm": 1.6241083187608925,
+      "learning_rate": 3.724215741290893e-08,
+      "loss": 1.0815,
+      "step": 18615
+    },
+    {
+      "epoch": 2.97856,
+      "grad_norm": 1.7660087971906941,
+      "learning_rate": 3.6772258336004176e-08,
+      "loss": 1.0833,
+      "step": 18616
+    },
+    {
+      "epoch": 2.97872,
+      "grad_norm": 1.7820006847540006,
+      "learning_rate": 3.6305342016151876e-08,
+      "loss": 1.0586,
+      "step": 18617
+    },
+    {
+      "epoch": 2.97888,
+      "grad_norm": 1.6452958169499983,
+      "learning_rate": 3.5841408467296444e-08,
+      "loss": 1.0844,
+      "step": 18618
+    },
+    {
+      "epoch": 2.97904,
+      "grad_norm": 2.0108315164573107,
+      "learning_rate": 3.538045770326015e-08,
+      "loss": 1.1196,
+      "step": 18619
+    },
+    {
+      "epoch": 2.9792,
+      "grad_norm": 1.6862847917032169,
+      "learning_rate": 3.492248973780976e-08,
+      "loss": 1.0338,
+      "step": 18620
+    },
+    {
+      "epoch": 2.97936,
+      "grad_norm": 1.6960780789875918,
+      "learning_rate": 3.4467504584612123e-08,
+      "loss": 1.0493,
+      "step": 18621
+    },
+    {
+      "epoch": 2.97952,
+      "grad_norm": 1.7988541810852998,
+      "learning_rate": 3.401550225723416e-08,
+      "loss": 1.1052,
+      "step": 18622
+    },
+    {
+      "epoch": 2.97968,
+      "grad_norm": 1.7298074447757266,
+      "learning_rate": 3.356648276917618e-08,
+      "loss": 1.0349,
+      "step": 18623
+    },
+    {
+      "epoch": 2.9798400000000003,
+      "grad_norm": 1.5512823660731263,
+      "learning_rate": 3.3120446133816374e-08,
+      "loss": 1.0797,
+      "step": 18624
+    },
+    {
+      "epoch": 2.98,
+      "grad_norm": 1.6437935127861167,
+      "learning_rate": 3.2677392364477425e-08,
+      "loss": 1.05,
+      "step": 18625
+    },
+    {
+      "epoch": 2.98016,
+      "grad_norm": 2.0199660033942206,
+      "learning_rate": 3.223732147438208e-08,
+      "loss": 1.0289,
+      "step": 18626
+    },
+    {
+      "epoch": 2.98032,
+      "grad_norm": 1.655500810662828,
+      "learning_rate": 3.180023347665317e-08,
+      "loss": 1.0729,
+      "step": 18627
+    },
+    {
+      "epoch": 2.98048,
+      "grad_norm": 1.597911866668118,
+      "learning_rate": 3.136612838433584e-08,
+      "loss": 1.0641,
+      "step": 18628
+    },
+    {
+      "epoch": 2.98064,
+      "grad_norm": 1.770539086498168,
+      "learning_rate": 3.093500621038636e-08,
+      "loss": 1.0935,
+      "step": 18629
+    },
+    {
+      "epoch": 2.9808,
+      "grad_norm": 1.683524305362178,
+      "learning_rate": 3.0506866967672244e-08,
+      "loss": 1.0394,
+      "step": 18630
+    },
+    {
+      "epoch": 2.98096,
+      "grad_norm": 1.7124562395903296,
+      "learning_rate": 3.008171066894994e-08,
+      "loss": 1.0494,
+      "step": 18631
+    },
+    {
+      "epoch": 2.9811199999999998,
+      "grad_norm": 1.643926565675597,
+      "learning_rate": 2.9659537326920393e-08,
+      "loss": 1.0713,
+      "step": 18632
+    },
+    {
+      "epoch": 2.98128,
+      "grad_norm": 1.6664198095799134,
+      "learning_rate": 2.924034695418465e-08,
+      "loss": 1.1064,
+      "step": 18633
+    },
+    {
+      "epoch": 2.98144,
+      "grad_norm": 1.7951036367443456,
+      "learning_rate": 2.882413956324381e-08,
+      "loss": 1.0682,
+      "step": 18634
+    },
+    {
+      "epoch": 2.9816000000000003,
+      "grad_norm": 1.7248446449546404,
+      "learning_rate": 2.8410915166510178e-08,
+      "loss": 1.0225,
+      "step": 18635
+    },
+    {
+      "epoch": 2.98176,
+      "grad_norm": 1.8809986848819125,
+      "learning_rate": 2.8000673776329422e-08,
+      "loss": 1.0888,
+      "step": 18636
+    },
+    {
+      "epoch": 2.98192,
+      "grad_norm": 1.6758453255265722,
+      "learning_rate": 2.7593415404925104e-08,
+      "loss": 1.044,
+      "step": 18637
+    },
+    {
+      "epoch": 2.98208,
+      "grad_norm": 1.7912072102924437,
+      "learning_rate": 2.7189140064465267e-08,
+      "loss": 1.1225,
+      "step": 18638
+    },
+    {
+      "epoch": 2.98224,
+      "grad_norm": 1.5724757010257224,
+      "learning_rate": 2.6787847766995834e-08,
+      "loss": 1.0365,
+      "step": 18639
+    },
+    {
+      "epoch": 2.9824,
+      "grad_norm": 1.5080872428463739,
+      "learning_rate": 2.6389538524507208e-08,
+      "loss": 1.0881,
+      "step": 18640
+    },
+    {
+      "epoch": 2.98256,
+      "grad_norm": 1.7848675710645385,
+      "learning_rate": 2.5994212348878778e-08,
+      "loss": 1.079,
+      "step": 18641
+    },
+    {
+      "epoch": 2.98272,
+      "grad_norm": 1.5695180995902351,
+      "learning_rate": 2.560186925190111e-08,
+      "loss": 1.07,
+      "step": 18642
+    },
+    {
+      "epoch": 2.9828799999999998,
+      "grad_norm": 1.8114652390558887,
+      "learning_rate": 2.5212509245287064e-08,
+      "loss": 1.0752,
+      "step": 18643
+    },
+    {
+      "epoch": 2.98304,
+      "grad_norm": 1.8114652390558887,
+      "learning_rate": 2.5212509245287064e-08,
+      "loss": 1.4004,
+      "step": 18644
+    },
+    {
+      "epoch": 2.9832,
+      "grad_norm": 1.7635081414109424,
+      "learning_rate": 2.482613234066067e-08,
+      "loss": 1.0533,
+      "step": 18645
+    },
+    {
+      "epoch": 2.9833600000000002,
+      "grad_norm": 1.616669892800771,
+      "learning_rate": 2.4442738549534937e-08,
+      "loss": 0.9975,
+      "step": 18646
+    },
+    {
+      "epoch": 2.98352,
+      "grad_norm": 1.806444813283481,
+      "learning_rate": 2.4062327883356273e-08,
+      "loss": 1.1315,
+      "step": 18647
+    },
+    {
+      "epoch": 2.98368,
+      "grad_norm": 1.6795085646591112,
+      "learning_rate": 2.3684900353482252e-08,
+      "loss": 1.0723,
+      "step": 18648
+    },
+    {
+      "epoch": 2.98384,
+      "grad_norm": 1.6537385506190145,
+      "learning_rate": 2.331045597117054e-08,
+      "loss": 1.0674,
+      "step": 18649
+    },
+    {
+      "epoch": 2.984,
+      "grad_norm": 1.714608287485481,
+      "learning_rate": 2.2938994747601082e-08,
+      "loss": 1.0932,
+      "step": 18650
+    },
+    {
+      "epoch": 2.98416,
+      "grad_norm": 1.758422117024579,
+      "learning_rate": 2.25705166938317e-08,
+      "loss": 1.0788,
+      "step": 18651
+    },
+    {
+      "epoch": 2.98432,
+      "grad_norm": 1.5530714243196524,
+      "learning_rate": 2.2205021820886906e-08,
+      "loss": 1.0411,
+      "step": 18652
+    },
+    {
+      "epoch": 2.98448,
+      "grad_norm": 1.9155915386240134,
+      "learning_rate": 2.184251013965799e-08,
+      "loss": 1.0992,
+      "step": 18653
+    },
+    {
+      "epoch": 2.9846399999999997,
+      "grad_norm": 1.865683945166493,
+      "learning_rate": 2.1482981660969625e-08,
+      "loss": 1.073,
+      "step": 18654
+    },
+    {
+      "epoch": 2.9848,
+      "grad_norm": 1.6774103883358893,
+      "learning_rate": 2.1126436395546566e-08,
+      "loss": 1.1075,
+      "step": 18655
+    },
+    {
+      "epoch": 2.98496,
+      "grad_norm": 1.7411956252902654,
+      "learning_rate": 2.0772874354013648e-08,
+      "loss": 1.0836,
+      "step": 18656
+    },
+    {
+      "epoch": 2.98512,
+      "grad_norm": 1.9479340890700814,
+      "learning_rate": 2.0422295546940196e-08,
+      "loss": 1.0945,
+      "step": 18657
+    },
+    {
+      "epoch": 2.98528,
+      "grad_norm": 1.7217239307696444,
+      "learning_rate": 2.0074699984784505e-08,
+      "loss": 1.067,
+      "step": 18658
+    },
+    {
+      "epoch": 2.98544,
+      "grad_norm": 1.7522021295715304,
+      "learning_rate": 1.9730087677904962e-08,
+      "loss": 1.0907,
+      "step": 18659
+    },
+    {
+      "epoch": 2.9856,
+      "grad_norm": 1.5605005551346864,
+      "learning_rate": 1.9388458636604435e-08,
+      "loss": 1.0655,
+      "step": 18660
+    },
+    {
+      "epoch": 2.98576,
+      "grad_norm": 1.8184799988760032,
+      "learning_rate": 1.9049812871052563e-08,
+      "loss": 1.0901,
+      "step": 18661
+    },
+    {
+      "epoch": 2.98592,
+      "grad_norm": 1.694926012003406,
+      "learning_rate": 1.8714150391374584e-08,
+      "loss": 1.0657,
+      "step": 18662
+    },
+    {
+      "epoch": 2.98608,
+      "grad_norm": 1.6861604048208725,
+      "learning_rate": 1.838147120757361e-08,
+      "loss": 1.0062,
+      "step": 18663
+    },
+    {
+      "epoch": 2.98624,
+      "grad_norm": 1.834384723274792,
+      "learning_rate": 1.8051775329586128e-08,
+      "loss": 1.0737,
+      "step": 18664
+    },
+    {
+      "epoch": 2.9864,
+      "grad_norm": 1.6424700948770032,
+      "learning_rate": 1.772506276723762e-08,
+      "loss": 1.0473,
+      "step": 18665
+    },
+    {
+      "epoch": 2.98656,
+      "grad_norm": 1.7423459243357615,
+      "learning_rate": 1.7401333530286944e-08,
+      "loss": 1.0302,
+      "step": 18666
+    },
+    {
+      "epoch": 2.98672,
+      "grad_norm": 1.7397602403187644,
+      "learning_rate": 1.708058762839304e-08,
+      "loss": 1.0638,
+      "step": 18667
+    },
+    {
+      "epoch": 2.98688,
+      "grad_norm": 1.7979534166680828,
+      "learning_rate": 1.676282507112603e-08,
+      "loss": 1.091,
+      "step": 18668
+    },
+    {
+      "epoch": 2.98704,
+      "grad_norm": 1.8190040223501842,
+      "learning_rate": 1.6448045867956117e-08,
+      "loss": 1.0598,
+      "step": 18669
+    },
+    {
+      "epoch": 2.9872,
+      "grad_norm": 1.6425433106725296,
+      "learning_rate": 1.6136250028286894e-08,
+      "loss": 1.0726,
+      "step": 18670
+    },
+    {
+      "epoch": 2.98736,
+      "grad_norm": 2.276104224592726,
+      "learning_rate": 1.582743756143312e-08,
+      "loss": 1.0451,
+      "step": 18671
+    },
+    {
+      "epoch": 2.98752,
+      "grad_norm": 1.838110130489731,
+      "learning_rate": 1.5521608476587458e-08,
+      "loss": 1.0354,
+      "step": 18672
+    },
+    {
+      "epoch": 2.98768,
+      "grad_norm": 1.685570095657968,
+      "learning_rate": 1.5218762782875928e-08,
+      "loss": 1.0739,
+      "step": 18673
+    },
+    {
+      "epoch": 2.9878400000000003,
+      "grad_norm": 1.7171509544193915,
+      "learning_rate": 1.4918900489357957e-08,
+      "loss": 1.0558,
+      "step": 18674
+    },
+    {
+      "epoch": 2.988,
+      "grad_norm": 1.83597766071068,
+      "learning_rate": 1.4622021604948633e-08,
+      "loss": 1.0854,
+      "step": 18675
+    },
+    {
+      "epoch": 2.98816,
+      "grad_norm": 1.5705726274488798,
+      "learning_rate": 1.4328126138540843e-08,
+      "loss": 1.043,
+      "step": 18676
+    },
+    {
+      "epoch": 2.98832,
+      "grad_norm": 1.7481913771088158,
+      "learning_rate": 1.4037214098883145e-08,
+      "loss": 1.0879,
+      "step": 18677
+    },
+    {
+      "epoch": 2.98848,
+      "grad_norm": 1.706851385729219,
+      "learning_rate": 1.3749285494657482e-08,
+      "loss": 1.064,
+      "step": 18678
+    },
+    {
+      "epoch": 2.98864,
+      "grad_norm": 1.6039922952503898,
+      "learning_rate": 1.3464340334456983e-08,
+      "loss": 1.0221,
+      "step": 18679
+    },
+    {
+      "epoch": 2.9888,
+      "grad_norm": 1.5724679033034277,
+      "learning_rate": 1.3182378626785951e-08,
+      "loss": 1.0687,
+      "step": 18680
+    },
+    {
+      "epoch": 2.98896,
+      "grad_norm": 1.566393210547132,
+      "learning_rate": 1.2903400380048781e-08,
+      "loss": 1.0348,
+      "step": 18681
+    },
+    {
+      "epoch": 2.9891199999999998,
+      "grad_norm": 1.6716307267857813,
+      "learning_rate": 1.2627405602594344e-08,
+      "loss": 1.0356,
+      "step": 18682
+    },
+    {
+      "epoch": 2.98928,
+      "grad_norm": 1.9898080825657114,
+      "learning_rate": 1.2354394302627193e-08,
+      "loss": 1.1404,
+      "step": 18683
+    },
+    {
+      "epoch": 2.98944,
+      "grad_norm": 1.6124873410211322,
+      "learning_rate": 1.2084366488318566e-08,
+      "loss": 1.04,
+      "step": 18684
+    },
+    {
+      "epoch": 2.9896000000000003,
+      "grad_norm": 1.6128299388425473,
+      "learning_rate": 1.181732216770648e-08,
+      "loss": 1.1137,
+      "step": 18685
+    },
+    {
+      "epoch": 2.98976,
+      "grad_norm": 1.7726949013845887,
+      "learning_rate": 1.1553261348762335e-08,
+      "loss": 1.0339,
+      "step": 18686
+    },
+    {
+      "epoch": 2.98992,
+      "grad_norm": 1.6811828979333332,
+      "learning_rate": 1.1292184039379816e-08,
+      "loss": 1.0793,
+      "step": 18687
+    },
+    {
+      "epoch": 2.99008,
+      "grad_norm": 1.669335531422997,
+      "learning_rate": 1.1034090247341588e-08,
+      "loss": 1.0337,
+      "step": 18688
+    },
+    {
+      "epoch": 2.99024,
+      "grad_norm": 1.697880115786991,
+      "learning_rate": 1.0778979980341497e-08,
+      "loss": 1.0821,
+      "step": 18689
+    },
+    {
+      "epoch": 2.9904,
+      "grad_norm": 1.6459614713710822,
+      "learning_rate": 1.0526853245995671e-08,
+      "loss": 1.0483,
+      "step": 18690
+    },
+    {
+      "epoch": 2.99056,
+      "grad_norm": 1.779700173516101,
+      "learning_rate": 1.0277710051831425e-08,
+      "loss": 1.0559,
+      "step": 18691
+    },
+    {
+      "epoch": 2.99072,
+      "grad_norm": 1.870363689583133,
+      "learning_rate": 1.0031550405276147e-08,
+      "loss": 1.0374,
+      "step": 18692
+    },
+    {
+      "epoch": 2.9908799999999998,
+      "grad_norm": 1.4885323453915207,
+      "learning_rate": 9.788374313679516e-09,
+      "loss": 1.0743,
+      "step": 18693
+    },
+    {
+      "epoch": 2.99104,
+      "grad_norm": 1.5732058112036915,
+      "learning_rate": 9.54818178430239e-09,
+      "loss": 1.0457,
+      "step": 18694
+    },
+    {
+      "epoch": 2.9912,
+      "grad_norm": 1.6654573925280447,
+      "learning_rate": 9.310972824305707e-09,
+      "loss": 1.0566,
+      "step": 18695
+    },
+    {
+      "epoch": 2.9913600000000002,
+      "grad_norm": 1.5422295001746744,
+      "learning_rate": 9.076747440761591e-09,
+      "loss": 1.0262,
+      "step": 18696
+    },
+    {
+      "epoch": 2.99152,
+      "grad_norm": 1.8369243482650819,
+      "learning_rate": 8.845505640664442e-09,
+      "loss": 1.0919,
+      "step": 18697
+    },
+    {
+      "epoch": 2.99168,
+      "grad_norm": 1.8711197023270858,
+      "learning_rate": 8.61724743091985e-09,
+      "loss": 1.0508,
+      "step": 18698
+    },
+    {
+      "epoch": 2.99184,
+      "grad_norm": 1.6897635327670335,
+      "learning_rate": 8.391972818333482e-09,
+      "loss": 1.0556,
+      "step": 18699
+    },
+    {
+      "epoch": 2.992,
+      "grad_norm": 1.724850989579992,
+      "learning_rate": 8.169681809622188e-09,
+      "loss": 1.026,
+      "step": 18700
+    },
+    {
+      "epoch": 2.99216,
+      "grad_norm": 1.7702106255768415,
+      "learning_rate": 7.950374411425098e-09,
+      "loss": 1.051,
+      "step": 18701
+    },
+    {
+      "epoch": 2.99232,
+      "grad_norm": 1.8108248924762822,
+      "learning_rate": 7.734050630292533e-09,
+      "loss": 1.0746,
+      "step": 18702
+    },
+    {
+      "epoch": 2.99248,
+      "grad_norm": 1.7452088188845296,
+      "learning_rate": 7.520710472663784e-09,
+      "loss": 1.064,
+      "step": 18703
+    },
+    {
+      "epoch": 2.9926399999999997,
+      "grad_norm": 1.5061876956398692,
+      "learning_rate": 7.310353944922632e-09,
+      "loss": 1.0709,
+      "step": 18704
+    },
+    {
+      "epoch": 2.9928,
+      "grad_norm": 1.6328142767619818,
+      "learning_rate": 7.102981053330738e-09,
+      "loss": 1.0558,
+      "step": 18705
+    },
+    {
+      "epoch": 2.99296,
+      "grad_norm": 1.8149257443203581,
+      "learning_rate": 6.898591804072041e-09,
+      "loss": 1.0459,
+      "step": 18706
+    },
+    {
+      "epoch": 2.9931200000000002,
+      "grad_norm": 1.5487735965949119,
+      "learning_rate": 6.697186203263872e-09,
+      "loss": 1.0436,
+      "step": 18707
+    },
+    {
+      "epoch": 2.99328,
+      "grad_norm": 2.080393177813149,
+      "learning_rate": 6.498764256901435e-09,
+      "loss": 1.0901,
+      "step": 18708
+    },
+    {
+      "epoch": 2.99344,
+      "grad_norm": 1.6042207536297244,
+      "learning_rate": 6.303325970913321e-09,
+      "loss": 1.0163,
+      "step": 18709
+    },
+    {
+      "epoch": 2.9936,
+      "grad_norm": 1.6830774875863914,
+      "learning_rate": 6.1108713511282e-09,
+      "loss": 1.0609,
+      "step": 18710
+    },
+    {
+      "epoch": 2.99376,
+      "grad_norm": 2.1062419240268913,
+      "learning_rate": 5.921400403285926e-09,
+      "loss": 1.1678,
+      "step": 18711
+    },
+    {
+      "epoch": 2.99392,
+      "grad_norm": 1.7420863977828576,
+      "learning_rate": 5.734913133037534e-09,
+      "loss": 1.0459,
+      "step": 18712
+    },
+    {
+      "epoch": 2.99408,
+      "grad_norm": 1.6886735237911377,
+      "learning_rate": 5.551409545967445e-09,
+      "loss": 1.031,
+      "step": 18713
+    },
+    {
+      "epoch": 2.99424,
+      "grad_norm": 1.8123184096083012,
+      "learning_rate": 5.3708896475157534e-09,
+      "loss": 1.0466,
+      "step": 18714
+    },
+    {
+      "epoch": 2.9943999999999997,
+      "grad_norm": 1.8184477856415067,
+      "learning_rate": 5.193353443100346e-09,
+      "loss": 1.0534,
+      "step": 18715
+    },
+    {
+      "epoch": 2.99456,
+      "grad_norm": 2.006979096123033,
+      "learning_rate": 5.018800938005886e-09,
+      "loss": 1.0455,
+      "step": 18716
+    },
+    {
+      "epoch": 2.99472,
+      "grad_norm": 1.834412063986649,
+      "learning_rate": 4.847232137450419e-09,
+      "loss": 1.0432,
+      "step": 18717
+    },
+    {
+      "epoch": 2.99488,
+      "grad_norm": 1.7360765801606,
+      "learning_rate": 4.6786470465298715e-09,
+      "loss": 1.0277,
+      "step": 18718
+    },
+    {
+      "epoch": 2.99504,
+      "grad_norm": 1.629880706323518,
+      "learning_rate": 4.513045670306859e-09,
+      "loss": 1.0352,
+      "step": 18719
+    },
+    {
+      "epoch": 2.9952,
+      "grad_norm": 1.765858593051811,
+      "learning_rate": 4.350428013688568e-09,
+      "loss": 1.0422,
+      "step": 18720
+    },
+    {
+      "epoch": 2.99536,
+      "grad_norm": 1.6109020715858753,
+      "learning_rate": 4.190794081559979e-09,
+      "loss": 1.0756,
+      "step": 18721
+    },
+    {
+      "epoch": 2.99552,
+      "grad_norm": 1.6991224757515908,
+      "learning_rate": 4.034143878661745e-09,
+      "loss": 1.0498,
+      "step": 18722
+    },
+    {
+      "epoch": 2.99568,
+      "grad_norm": 1.7377172780454797,
+      "learning_rate": 3.880477409667904e-09,
+      "loss": 1.0614,
+      "step": 18723
+    },
+    {
+      "epoch": 2.99584,
+      "grad_norm": 1.720418543065921,
+      "learning_rate": 3.729794679174781e-09,
+      "loss": 1.1034,
+      "step": 18724
+    },
+    {
+      "epoch": 2.996,
+      "grad_norm": 1.8002952496002294,
+      "learning_rate": 3.5820956916787774e-09,
+      "loss": 1.0438,
+      "step": 18725
+    },
+    {
+      "epoch": 2.99616,
+      "grad_norm": 1.9103593514683979,
+      "learning_rate": 3.4373804515874797e-09,
+      "loss": 1.1331,
+      "step": 18726
+    },
+    {
+      "epoch": 2.99632,
+      "grad_norm": 1.6488085299044282,
+      "learning_rate": 3.295648963208553e-09,
+      "loss": 1.0633,
+      "step": 18727
+    },
+    {
+      "epoch": 2.99648,
+      "grad_norm": 1.9807442954117043,
+      "learning_rate": 3.156901230771947e-09,
+      "loss": 1.1617,
+      "step": 18728
+    },
+    {
+      "epoch": 2.99664,
+      "grad_norm": 1.8488937838952038,
+      "learning_rate": 3.0211372584298958e-09,
+      "loss": 1.053,
+      "step": 18729
+    },
+    {
+      "epoch": 2.9968,
+      "grad_norm": 1.6442603021644022,
+      "learning_rate": 2.8883570502236114e-09,
+      "loss": 1.0585,
+      "step": 18730
+    },
+    {
+      "epoch": 2.99696,
+      "grad_norm": 1.750674643615086,
+      "learning_rate": 2.7585606101165895e-09,
+      "loss": 0.9915,
+      "step": 18731
+    },
+    {
+      "epoch": 2.99712,
+      "grad_norm": 1.688166226658389,
+      "learning_rate": 2.631747941983509e-09,
+      "loss": 1.0589,
+      "step": 18732
+    },
+    {
+      "epoch": 2.99728,
+      "grad_norm": 1.6827127317680364,
+      "learning_rate": 2.50791904961023e-09,
+      "loss": 1.0093,
+      "step": 18733
+    },
+    {
+      "epoch": 2.99744,
+      "grad_norm": 1.8019253689894206,
+      "learning_rate": 2.3870739366826934e-09,
+      "loss": 1.0553,
+      "step": 18734
+    },
+    {
+      "epoch": 2.9976000000000003,
+      "grad_norm": 1.858633745765144,
+      "learning_rate": 2.2692126068202256e-09,
+      "loss": 1.0289,
+      "step": 18735
+    },
+    {
+      "epoch": 2.99776,
+      "grad_norm": 1.713755058151155,
+      "learning_rate": 2.154335063531132e-09,
+      "loss": 1.0524,
+      "step": 18736
+    },
+    {
+      "epoch": 2.99792,
+      "grad_norm": 1.5317078363503267,
+      "learning_rate": 2.0424413102348995e-09,
+      "loss": 1.0031,
+      "step": 18737
+    },
+    {
+      "epoch": 2.99808,
+      "grad_norm": 1.8862561579300896,
+      "learning_rate": 1.9335313502955033e-09,
+      "loss": 1.0735,
+      "step": 18738
+    },
+    {
+      "epoch": 2.99824,
+      "grad_norm": 1.6677541719173647,
+      "learning_rate": 1.8276051869325905e-09,
+      "loss": 1.0197,
+      "step": 18739
+    },
+    {
+      "epoch": 2.9984,
+      "grad_norm": 1.6991315809210767,
+      "learning_rate": 1.7246628233325013e-09,
+      "loss": 1.0676,
+      "step": 18740
+    },
+    {
+      "epoch": 2.99856,
+      "grad_norm": 1.8343288535118227,
+      "learning_rate": 1.6247042625483488e-09,
+      "loss": 1.0645,
+      "step": 18741
+    },
+    {
+      "epoch": 2.99872,
+      "grad_norm": 1.7095597857711538,
+      "learning_rate": 1.5277295075666331e-09,
+      "loss": 1.0766,
+      "step": 18742
+    },
+    {
+      "epoch": 2.9988799999999998,
+      "grad_norm": 1.6739544392769243,
+      "learning_rate": 1.4337385612961385e-09,
+      "loss": 1.0488,
+      "step": 18743
+    },
+    {
+      "epoch": 2.99904,
+      "grad_norm": 1.6839919206577105,
+      "learning_rate": 1.3427314265235246e-09,
+      "loss": 1.0578,
+      "step": 18744
+    },
+    {
+      "epoch": 2.9992,
+      "grad_norm": 1.758564056823037,
+      "learning_rate": 1.2547081059688382e-09,
+      "loss": 1.0416,
+      "step": 18745
+    },
+    {
+      "epoch": 2.9993600000000002,
+      "grad_norm": 1.8339125519327073,
+      "learning_rate": 1.1696686022522051e-09,
+      "loss": 1.0139,
+      "step": 18746
+    },
+    {
+      "epoch": 2.99952,
+      "grad_norm": 1.800480962108195,
+      "learning_rate": 1.0876129179271389e-09,
+      "loss": 1.0588,
+      "step": 18747
+    },
+    {
+      "epoch": 2.99968,
+      "grad_norm": 1.7035340366375726,
+      "learning_rate": 1.0085410554361297e-09,
+      "loss": 1.0737,
+      "step": 18748
+    },
+    {
+      "epoch": 2.99984,
+      "grad_norm": 1.6298709005628516,
+      "learning_rate": 9.324530171328506e-10,
+      "loss": 1.0905,
+      "step": 18749
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.7639782816545797,
+      "learning_rate": 8.593488052932586e-10,
+      "loss": 1.0648,
+      "step": 18750
+    },
+    {
+      "epoch": 3.0,
+      "step": 18750,
+      "total_flos": 3.9455981417927803e+18,
+      "train_loss": 1.597369184363683,
+      "train_runtime": 34447.4589,
+      "train_samples_per_second": 17.418,
+      "train_steps_per_second": 0.544
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 18750,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.9455981417927803e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}