flipped_results

This model is a fine-tuned version of google/vit-base-patch16-224-in21k on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 2.5433
• Accuracy: 0.4099

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 64
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
1.3063	1.0	37	1.2485	0.4592
1.2272	2.0	74	1.1915	0.4626
1.1518	3.0	111	1.1770	0.4575
1.0504	4.0	148	1.1724	0.4745
0.9525	5.0	185	1.1966	0.4898
0.7485	6.0	222	1.2403	0.4626
0.5645	7.0	259	1.3973	0.4235
0.4645	8.0	296	1.4260	0.4898
0.374	9.0	333	1.5838	0.4405
0.2721	10.0	370	1.6747	0.4354
0.2679	11.0	407	1.7427	0.4507
0.2284	12.0	444	1.8097	0.4269
0.2003	13.0	481	1.9738	0.3997
0.1844	14.0	518	1.9745	0.4524
0.1631	15.0	555	2.0326	0.4456
0.1135	16.0	592	2.1294	0.4184
0.1188	17.0	629	2.1613	0.4065
0.1204	18.0	666	2.1795	0.4252
0.1083	19.0	703	2.2433	0.4167
0.0794	20.0	740	2.2762	0.4150
0.0589	21.0	777	2.3736	0.4065
0.0646	22.0	814	2.3644	0.4252
0.0744	23.0	851	2.4478	0.4099
0.0728	24.0	888	2.4367	0.4099
0.0382	25.0	925	2.5123	0.3997
0.033	26.0	962	2.5202	0.4031
0.0326	27.0	999	2.5145	0.4099
0.023	28.0	1036	2.5309	0.4099
0.0377	29.0	1073	2.5409	0.4099
0.0243	30.0	1110	2.5433	0.4099

Framework versions

• Transformers 4.30.2
• Pytorch 2.0.1+cu117
• Datasets 2.13.1
• Tokenizers 0.13.3