N_bert_agnews_padding50model

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

• Loss: 0.5754
• Accuracy: 0.9467

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: 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	Accuracy
0.1754	1.0	7500	0.1909	0.9428
0.1362	2.0	15000	0.1928	0.9461
0.1139	3.0	22500	0.2106	0.9461
0.0825	4.0	30000	0.2544	0.9466
0.0565	5.0	37500	0.3046	0.9367
0.0372	6.0	45000	0.3764	0.9436
0.0347	7.0	52500	0.3646	0.9425
0.0346	8.0	60000	0.3826	0.9461
0.0247	9.0	67500	0.4244	0.9455
0.0113	10.0	75000	0.4418	0.9446
0.0166	11.0	82500	0.4917	0.9462
0.0157	12.0	90000	0.4662	0.9442
0.0124	13.0	97500	0.4864	0.9438
0.0055	14.0	105000	0.4912	0.9457
0.0102	15.0	112500	0.5040	0.9446
0.0045	16.0	120000	0.5200	0.9441
0.0038	17.0	127500	0.5374	0.9467
0.0012	18.0	135000	0.5605	0.9459
0.0005	19.0	142500	0.5809	0.9455
0.0003	20.0	150000	0.5754	0.9467

Framework versions

• Transformers 4.33.2
• Pytorch 2.0.1+cu117
• Datasets 2.14.5
• Tokenizers 0.13.3