distilbert_imdb_padding90model

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: 0.7443
• Accuracy: 0.9309

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
0.2395	1.0	1563	0.2329	0.9140
0.1704	2.0	3126	0.2702	0.9175
0.1157	3.0	4689	0.3082	0.9254
0.0725	4.0	6252	0.3575	0.9204
0.049	5.0	7815	0.4781	0.9152
0.0379	6.0	9378	0.4916	0.9257
0.0236	7.0	10941	0.5292	0.9244
0.0248	8.0	12504	0.5522	0.9249
0.0198	9.0	14067	0.5522	0.9273
0.018	10.0	15630	0.5759	0.9286
0.0126	11.0	17193	0.6480	0.9268
0.0061	12.0	18756	0.6711	0.9295
0.0091	13.0	20319	0.6219	0.9293
0.0049	14.0	21882	0.7301	0.9261
0.0085	15.0	23445	0.6748	0.9299
0.0039	16.0	25008	0.6808	0.9306
0.003	17.0	26571	0.7055	0.9289
0.0037	18.0	28134	0.7126	0.9292
0.0	19.0	29697	0.7484	0.9295
0.001	20.0	31260	0.7443	0.9309

Framework versions

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