N_distilbert_agnews_padding0model

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

• Loss: 0.6408
• Accuracy: 0.9464

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
0.1774	1.0	7500	0.1995	0.9392
0.1403	2.0	15000	0.1939	0.9447
0.1114	3.0	22500	0.2186	0.9459
0.0741	4.0	30000	0.2832	0.9446
0.0499	5.0	37500	0.3070	0.9408
0.0376	6.0	45000	0.3704	0.9434
0.0341	7.0	52500	0.3999	0.9426
0.0319	8.0	60000	0.4505	0.9425
0.0191	9.0	67500	0.4649	0.9399
0.013	10.0	75000	0.5064	0.9403
0.0184	11.0	82500	0.4858	0.9405
0.0081	12.0	90000	0.5358	0.9432
0.0065	13.0	97500	0.5440	0.9436
0.0053	14.0	105000	0.5755	0.9436
0.0017	15.0	112500	0.5907	0.9457
0.0042	16.0	120000	0.5916	0.9455
0.0031	17.0	127500	0.5976	0.9468
0.0017	18.0	135000	0.6063	0.9474
0.0003	19.0	142500	0.6248	0.9467
0.0003	20.0	150000	0.6408	0.9464

Framework versions

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