N_distilbert_agnews_padding70model

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: 0.6396
• Accuracy: 0.945

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
0.1861	1.0	7500	0.1881	0.9424
0.1359	2.0	15000	0.1967	0.9451
0.1194	3.0	22500	0.2221	0.9429
0.0881	4.0	30000	0.2671	0.9451
0.0577	5.0	37500	0.3216	0.9367
0.0503	6.0	45000	0.3547	0.9420
0.0402	7.0	52500	0.4013	0.9433
0.0279	8.0	60000	0.4334	0.9438
0.0169	9.0	67500	0.5243	0.9371
0.0104	10.0	75000	0.4979	0.9425
0.0149	11.0	82500	0.5179	0.9436
0.0129	12.0	90000	0.5020	0.9413
0.0054	13.0	97500	0.5674	0.9420
0.01	14.0	105000	0.5720	0.9433
0.0087	15.0	112500	0.5746	0.9409
0.0022	16.0	120000	0.5759	0.9436
0.0012	17.0	127500	0.5933	0.9436
0.0005	18.0	135000	0.6308	0.9429
0.0021	19.0	142500	0.6259	0.9445
0.0017	20.0	150000	0.6396	0.945

Framework versions

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