N_distilbert_agnews_padding90model

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: 0.6235
• Accuracy: 0.9474

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
0.1806	1.0	7500	0.1897	0.9411
0.1426	2.0	15000	0.2027	0.9421
0.1169	3.0	22500	0.2151	0.9447
0.0978	4.0	30000	0.2648	0.9421
0.0634	5.0	37500	0.3140	0.9393
0.0385	6.0	45000	0.3887	0.9413
0.0335	7.0	52500	0.4336	0.9378
0.0317	8.0	60000	0.4547	0.9405
0.0192	9.0	67500	0.4665	0.9389
0.0091	10.0	75000	0.4814	0.9422
0.0142	11.0	82500	0.5405	0.9422
0.0131	12.0	90000	0.5560	0.9416
0.0131	13.0	97500	0.5343	0.9413
0.005	14.0	105000	0.5242	0.9430
0.0013	15.0	112500	0.5975	0.9425
0.003	16.0	120000	0.5905	0.9430
0.0024	17.0	127500	0.5740	0.9453
0.0009	18.0	135000	0.6001	0.9457
0.0012	19.0	142500	0.6177	0.9467
0.0002	20.0	150000	0.6235	0.9474

Framework versions

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