bert-finetuned-ner

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

• Loss: 0.1128
• Precision: 0.9428
• Recall: 0.9534
• F1: 0.9480
• Accuracy: 0.9867

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: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.0937	1.0	1756	0.0660	0.9179	0.9332	0.9255	0.9825
0.0378	2.0	3512	0.0766	0.9246	0.9451	0.9348	0.9843
0.0245	3.0	5268	0.0667	0.9241	0.9409	0.9325	0.9843
0.017	4.0	7024	0.0712	0.9343	0.9505	0.9424	0.9863
0.0143	5.0	8780	0.0898	0.9366	0.9492	0.9428	0.9855
0.0049	6.0	10536	0.0964	0.9294	0.9482	0.9387	0.9853
0.0039	7.0	12292	0.1001	0.9353	0.9512	0.9432	0.9860
0.0036	8.0	14048	0.1002	0.9388	0.9522	0.9454	0.9862
0.0018	9.0	15804	0.1049	0.9363	0.9495	0.9428	0.9861
0.0019	10.0	17560	0.1191	0.9375	0.9497	0.9436	0.9849
0.0008	11.0	19316	0.1083	0.9396	0.9530	0.9463	0.9864
0.0003	12.0	21072	0.1064	0.9419	0.9530	0.9475	0.9864
0.0004	13.0	22828	0.1091	0.9448	0.9527	0.9487	0.9865
0.0006	14.0	24584	0.1132	0.9464	0.9542	0.9503	0.9867
0.0004	15.0	26340	0.1128	0.9428	0.9534	0.9480	0.9867

Framework versions

• Transformers 4.23.1
• Pytorch 1.12.1+cu113
• Datasets 2.6.1
• Tokenizers 0.13.1