favs_token_classification_v2_updated_data

This model is a fine-tuned version of bert-base-cased on the token_classification_v2 dataset. It achieves the following results on the evaluation set:

• Loss: 0.5346
• Precision: 0.6923
• Recall: 0.8357
• F1: 0.7573
• Accuracy: 0.8493

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: 1.5e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
2.3096	1.0	13	1.9927	0.3011	0.2	0.2403	0.3726
2.038	2.0	26	1.7093	0.2569	0.2643	0.2606	0.4274
1.8391	3.0	39	1.4452	0.3057	0.4214	0.3544	0.5562
1.4912	4.0	52	1.2176	0.4130	0.5429	0.4691	0.6493
1.3296	5.0	65	1.0368	0.4973	0.6643	0.5688	0.7123
1.2036	6.0	78	0.9084	0.5053	0.6786	0.5793	0.7260
0.9244	7.0	91	0.8148	0.5543	0.7286	0.6296	0.7616
0.8293	8.0	104	0.7482	0.5698	0.7286	0.6395	0.7726
0.7422	9.0	117	0.6961	0.5833	0.75	0.6562	0.7836
0.6379	10.0	130	0.6613	0.6124	0.7786	0.6855	0.8027
0.6071	11.0	143	0.6357	0.6193	0.7786	0.6899	0.8082
0.5526	12.0	156	0.6033	0.6433	0.7857	0.7074	0.8164
0.537	13.0	169	0.5813	0.6512	0.8	0.7179	0.8301
0.4806	14.0	182	0.5706	0.6608	0.8071	0.7267	0.8329
0.4503	15.0	195	0.5594	0.6647	0.8071	0.7290	0.8356
0.4149	16.0	208	0.5503	0.6805	0.8214	0.7443	0.8438
0.4175	17.0	221	0.5430	0.6824	0.8286	0.7484	0.8438
0.4337	18.0	234	0.5396	0.6923	0.8357	0.7573	0.8493
0.3965	19.0	247	0.5361	0.6882	0.8357	0.7548	0.8493
0.3822	20.0	260	0.5346	0.6923	0.8357	0.7573	0.8493

Framework versions

• Transformers 4.21.1
• Pytorch 1.12.1
• Datasets 2.4.0
• Tokenizers 0.12.1