N_distilbert_imdb_padding100model

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: 0.7393
• Accuracy: 0.9294

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
0.2387	1.0	1563	0.2354	0.919
0.1866	2.0	3126	0.2345	0.9248
0.1194	3.0	4689	0.3117	0.9212
0.0615	4.0	6252	0.3370	0.9219
0.0475	5.0	7815	0.5367	0.9131
0.0394	6.0	9378	0.5018	0.9236
0.0281	7.0	10941	0.5039	0.9243
0.0262	8.0	12504	0.5149	0.9238
0.0203	9.0	14067	0.5159	0.9275
0.0194	10.0	15630	0.5855	0.927
0.0092	11.0	17193	0.6452	0.9259
0.0097	12.0	18756	0.6318	0.9262
0.0024	13.0	20319	0.6537	0.9292
0.0056	14.0	21882	0.7551	0.9268
0.0037	15.0	23445	0.7516	0.9255
0.0073	16.0	25008	0.7335	0.9281
0.0025	17.0	26571	0.6959	0.9301
0.0008	18.0	28134	0.7439	0.9276
0.0005	19.0	29697	0.7300	0.9296
0.0004	20.0	31260	0.7393	0.9294

Framework versions

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