bert-finetuned-ner

This model is a fine-tuned version of bert-base-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.0032
• Precision: 0.9927
• Recall: 0.9937
• F1: 0.9931
• Accuracy: 0.9815

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: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	136	0.0056	0.9826	0.9803	0.9814	0.9446
No log	2.0	272	0.0038	0.9878	0.9810	0.9843	0.9557
No log	3.0	408	0.0026	0.9920	0.9890	0.9905	0.9668
0.0296	4.0	544	0.0020	0.9927	0.9908	0.9917	0.9742
0.0296	5.0	680	0.0021	0.9927	0.9919	0.9923	0.9779
0.0296	6.0	816	0.0026	0.9900	0.9937	0.9918	0.9779
0.0296	7.0	952	0.0024	0.9921	0.9943	0.9931	0.9815
0.0008	8.0	1088	0.0026	0.9921	0.9924	0.9922	0.9742
0.0008	9.0	1224	0.0030	0.9921	0.9937	0.9929	0.9779
0.0008	10.0	1360	0.0032	0.9921	0.9943	0.9931	0.9815
0.0008	11.0	1496	0.0021	0.9954	0.9937	0.9945	0.9852
0.0003	12.0	1632	0.0021	0.9927	0.9943	0.9934	0.9852
0.0003	13.0	1768	0.0026	0.9927	0.9943	0.9934	0.9852
0.0003	14.0	1904	0.0026	0.9927	0.9943	0.9934	0.9852
0.0001	15.0	2040	0.0027	0.9927	0.9943	0.9934	0.9852
0.0001	16.0	2176	0.0026	0.9927	0.9943	0.9934	0.9852
0.0001	17.0	2312	0.0027	0.9927	0.9943	0.9934	0.9852
0.0001	18.0	2448	0.0029	0.9927	0.9937	0.9931	0.9815
0.0001	19.0	2584	0.0033	0.9927	0.9937	0.9931	0.9815
0.0001	20.0	2720	0.0032	0.9927	0.9937	0.9931	0.9815

Framework versions

• Transformers 4.53.3
• Pytorch 2.6.0+cu124
• Datasets 3.6.0
• Tokenizers 0.21.2