favs_token_classification_v2

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: 0.5498
• Precision: 0.6610
• Recall: 0.8417
• F1: 0.7405
• Accuracy: 0.8575

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
2.2225	1.0	13	1.9093	0.3735	0.2230	0.2793	0.3808
1.9616	2.0	26	1.6124	0.3101	0.3525	0.3300	0.4877
1.7778	3.0	39	1.3562	0.3632	0.4964	0.4195	0.6219
1.4003	4.0	52	1.1595	0.4278	0.5755	0.4908	0.6685
1.2374	5.0	65	1.0260	0.4462	0.6259	0.5210	0.6904
1.1184	6.0	78	0.9223	0.4895	0.6691	0.5653	0.7205
0.8801	7.0	91	0.8179	0.5027	0.6763	0.5767	0.7397
0.8246	8.0	104	0.7591	0.5543	0.7338	0.6316	0.7699
0.7177	9.0	117	0.7037	0.5683	0.7482	0.6460	0.7890
0.6277	10.0	130	0.6652	0.5870	0.7770	0.6687	0.8
0.5744	11.0	143	0.6344	0.6011	0.7914	0.6832	0.8164
0.528	12.0	156	0.6117	0.6292	0.8058	0.7066	0.8329
0.4981	13.0	169	0.5919	0.6348	0.8129	0.7129	0.8384
0.4423	14.0	182	0.5841	0.6461	0.8273	0.7256	0.8438
0.4864	15.0	195	0.5781	0.6461	0.8273	0.7256	0.8521
0.3975	16.0	208	0.5677	0.6517	0.8345	0.7319	0.8548
0.3846	17.0	221	0.5563	0.6517	0.8345	0.7319	0.8548
0.3729	18.0	234	0.5503	0.6610	0.8417	0.7405	0.8575
0.3367	19.0	247	0.5504	0.6610	0.8417	0.7405	0.8575
0.3492	20.0	260	0.5498	0.6610	0.8417	0.7405	0.8575

Framework versions

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