N_distilbert_imdb_padding60model

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: 0.7224
• Accuracy: 0.9327

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
0.2346	1.0	1563	0.2252	0.916
0.1742	2.0	3126	0.2406	0.9204
0.1246	3.0	4689	0.3171	0.9224
0.0738	4.0	6252	0.3747	0.9245
0.0507	5.0	7815	0.4165	0.9278
0.0327	6.0	9378	0.5113	0.9248
0.0218	7.0	10941	0.5063	0.9210
0.0221	8.0	12504	0.5326	0.9279
0.0231	9.0	14067	0.5171	0.9279
0.0111	10.0	15630	0.6266	0.9275
0.0096	11.0	17193	0.6049	0.9255
0.0092	12.0	18756	0.6766	0.9237
0.0079	13.0	20319	0.6736	0.9273
0.0082	14.0	21882	0.6786	0.9296
0.0047	15.0	23445	0.6562	0.9298
0.003	16.0	25008	0.6903	0.9301
0.0028	17.0	26571	0.7158	0.9291
0.0	18.0	28134	0.7324	0.9321
0.0	19.0	29697	0.7185	0.9325
0.0003	20.0	31260	0.7224	0.9327

Framework versions

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