N_distilbert_agnews_padding40model

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: 0.6441
• Accuracy: 0.9449

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
0.181	1.0	7500	0.1943	0.9399
0.1378	2.0	15000	0.2044	0.9443
0.1183	3.0	22500	0.2246	0.9459
0.088	4.0	30000	0.2517	0.9445
0.0614	5.0	37500	0.3074	0.9382
0.0464	6.0	45000	0.3765	0.9407
0.0368	7.0	52500	0.4057	0.9416
0.0245	8.0	60000	0.4436	0.9430
0.0202	9.0	67500	0.4608	0.9420
0.0119	10.0	75000	0.4479	0.9425
0.0125	11.0	82500	0.5133	0.9436
0.0147	12.0	90000	0.5036	0.9451
0.0103	13.0	97500	0.5727	0.9437
0.0051	14.0	105000	0.5684	0.9430
0.0056	15.0	112500	0.5746	0.9424
0.0031	16.0	120000	0.6067	0.9436
0.0009	17.0	127500	0.5994	0.9455
0.0025	18.0	135000	0.6187	0.9433
0.0024	19.0	142500	0.6413	0.9449
0.0011	20.0	150000	0.6441	0.9449

Framework versions

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