distilbert_imdb_padding10model

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

• Loss: 0.6898
• Accuracy: 0.9336

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
0.2352	1.0	1563	0.2181	0.9211
0.1717	2.0	3126	0.2141	0.9301
0.1043	3.0	4689	0.2965	0.9269
0.0652	4.0	6252	0.3615	0.9291
0.044	5.0	7815	0.4307	0.9279
0.0368	6.0	9378	0.5081	0.9253
0.0198	7.0	10941	0.4579	0.928
0.0258	8.0	12504	0.5307	0.9271
0.0154	9.0	14067	0.5795	0.9262
0.0154	10.0	15630	0.6031	0.9242
0.0074	11.0	17193	0.6427	0.9289
0.0085	12.0	18756	0.6715	0.9291
0.007	13.0	20319	0.6645	0.9291
0.0047	14.0	21882	0.6918	0.9294
0.0022	15.0	23445	0.7447	0.9304
0.0032	16.0	25008	0.7146	0.9329
0.004	17.0	26571	0.7050	0.9324
0.0027	18.0	28134	0.6860	0.9340
0.0031	19.0	29697	0.6935	0.9330
0.0004	20.0	31260	0.6898	0.9336

Framework versions

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