bert-finetuned-ner

This model is a fine-tuned version of bert-base-uncased on the conll2003 dataset. It achieves the following results on the evaluation set:

• Loss: 0.0589
• Overall Precision: 0.9362
• Overall Recall: 0.9500
• Overall F1: 0.9430
• Overall Accuracy: 0.9873
• Loc F1: 0.9616
• Misc F1: 0.8783
• Org F1: 0.9121
• Per F1: 0.9797

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: 2e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Overall Precision	Overall Recall	Overall F1	Overall Accuracy	Loc F1	Misc F1	Org F1	Per F1
0.0745	1.0	1756	0.0556	0.9183	0.9345	0.9263	0.9848	0.9501	0.8499	0.8775	0.9765
0.0321	2.0	3512	0.0542	0.9346	0.9475	0.9410	0.9872	0.9618	0.8761	0.9073	0.9773
0.0172	3.0	5268	0.0589	0.9362	0.9500	0.9430	0.9873	0.9616	0.8783	0.9121	0.9797

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu102
• Datasets 2.6.1
• Tokenizers 0.13.1