bert-finetuned-ner4-new

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

• Loss: 0.1962
• Precision: 0.6905
• Recall: 0.8666
• F1: 0.7686
• Accuracy: 0.9402

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: 1e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use 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
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
0.852	1.0	2844	0.5164	0.3140	0.3681	0.3389	0.8628
0.467	2.0	5688	0.3406	0.3758	0.6604	0.4790	0.8733
0.3478	3.0	8532	0.2850	0.4874	0.7650	0.5955	0.8950
0.2887	4.0	11376	0.2695	0.5318	0.7932	0.6368	0.9030
0.2528	5.0	14220	0.2874	0.5381	0.8053	0.6451	0.9014
0.2294	6.0	17064	0.2320	0.5962	0.8273	0.6930	0.9198
0.2125	7.0	19908	0.2283	0.6103	0.8386	0.7065	0.9223
0.1997	8.0	22752	0.2070	0.6422	0.8384	0.7273	0.9302
0.1865	9.0	25596	0.2347	0.6287	0.8523	0.7236	0.9244
0.1788	10.0	28440	0.2379	0.6296	0.8547	0.7251	0.9248
0.1712	11.0	31284	0.2078	0.6580	0.8519	0.7425	0.9313
0.1632	12.0	34128	0.2111	0.6610	0.8585	0.7469	0.9331
0.1579	13.0	36972	0.2250	0.6515	0.8606	0.7416	0.9304
0.1531	14.0	39816	0.2027	0.6765	0.8615	0.7579	0.9375
0.1493	15.0	42660	0.2102	0.6766	0.8632	0.7586	0.9372
0.1451	16.0	45504	0.2098	0.6786	0.8653	0.7607	0.9376
0.143	17.0	48348	0.1962	0.6905	0.8666	0.7686	0.9402
0.1407	18.0	51192	0.2042	0.6900	0.8664	0.7682	0.9395
0.1385	19.0	54036	0.2058	0.6874	0.8685	0.7674	0.9388
0.1378	20.0	56880	0.2043	0.6890	0.8680	0.7682	0.9392

Framework versions

• Transformers 4.47.1
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0