distilbert-Nepali-NER

This model is a fine-tuned version of distilbert/distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.2917
• Precision: 0.0843
• Recall: 0.0538
• F1: 0.0657
• Accuracy: 0.9259

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: 64
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Accuracy	F1	Validation Loss	Precision	Recall
No log	0.92	200	0.8685	0.0	0.4700	0.0	0.0
No log	1.84	400	0.8984	0.0135	0.3581	0.0556	0.0077
0.4549	2.76	600	0.9087	0.0361	0.3188	0.0833	0.0231
0.4549	3.69	800	0.9111	0.0460	0.3040	0.0909	0.0308
0.2088	4.61	1000	0.9173	0.0396	0.2972	0.0556	0.0308
0.2088	5.53	1200	0.3065	0.0721	0.0615	0.0664	0.9100
0.2088	6.45	1400	0.2924	0.1724	0.0769	0.1064	0.9212
0.1601	7.37	1600	0.2929	0.0745	0.0538	0.0625	0.9234
0.1601	8.29	1800	0.2903	0.0893	0.0385	0.0538	0.9257
0.1114	9.22	2000	0.2917	0.0843	0.0538	0.0657	0.9259

Framework versions

• Transformers 4.39.1
• Pytorch 2.2.2+cu121
• Datasets 2.18.0
• Tokenizers 0.15.2