N_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: 0.7863
• Accuracy: 0.9307

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
0.2481	1.0	1563	0.2298	0.9142
0.1762	2.0	3126	0.2394	0.9208
0.1152	3.0	4689	0.2973	0.9262
0.0741	4.0	6252	0.3401	0.9242
0.0442	5.0	7815	0.5043	0.9184
0.04	6.0	9378	0.6373	0.9107
0.0258	7.0	10941	0.4737	0.9246
0.0199	8.0	12504	0.5434	0.9255
0.0198	9.0	14067	0.5922	0.9231
0.0122	10.0	15630	0.6507	0.924
0.0102	11.0	17193	0.6104	0.9274
0.0141	12.0	18756	0.6091	0.9267
0.0071	13.0	20319	0.6356	0.9255
0.0076	14.0	21882	0.6566	0.9296
0.0053	15.0	23445	0.6492	0.9296
0.0046	16.0	25008	0.7113	0.9284
0.0002	17.0	26571	0.7791	0.9286
0.0023	18.0	28134	0.7519	0.9308
0.0002	19.0	29697	0.7739	0.9303
0.0005	20.0	31260	0.7863	0.9307

Framework versions

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