distilbert_imdb_padding20model

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: 0.7052
• Accuracy: 0.9330

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
0.2371	1.0	1563	0.2108	0.9213
0.1697	2.0	3126	0.2468	0.9285
0.1016	3.0	4689	0.3128	0.9250
0.0707	4.0	6252	0.3839	0.9192
0.0425	5.0	7815	0.4262	0.9238
0.033	6.0	9378	0.4711	0.9281
0.0226	7.0	10941	0.5034	0.9261
0.0274	8.0	12504	0.5279	0.9283
0.0092	9.0	14067	0.6002	0.9260
0.0099	10.0	15630	0.5944	0.9295
0.0035	11.0	17193	0.7042	0.9279
0.0119	12.0	18756	0.5989	0.9282
0.0068	13.0	20319	0.6468	0.9283
0.006	14.0	21882	0.6569	0.9307
0.0045	15.0	23445	0.7417	0.9299
0.0051	16.0	25008	0.6578	0.9322
0.0039	17.0	26571	0.6388	0.9325
0.0011	18.0	28134	0.6771	0.9324
0.0	19.0	29697	0.6996	0.9329
0.0017	20.0	31260	0.7052	0.9330

Framework versions

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