N_distilbert_imdb_padding40model

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: 0.7640
• Accuracy: 0.9305

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
0.2367	1.0	1563	0.3081	0.8873
0.18	2.0	3126	0.2079	0.9299
0.1146	3.0	4689	0.3326	0.9227
0.0688	4.0	6252	0.3477	0.9238
0.0389	5.0	7815	0.4432	0.9256
0.0338	6.0	9378	0.4389	0.9252
0.0269	7.0	10941	0.4876	0.9254
0.0146	8.0	12504	0.5673	0.9272
0.0178	9.0	14067	0.5712	0.9249
0.0108	10.0	15630	0.5723	0.9303
0.0137	11.0	17193	0.5582	0.9289
0.0104	12.0	18756	0.6285	0.9303
0.0071	13.0	20319	0.6775	0.9296
0.0057	14.0	21882	0.7206	0.9262
0.0067	15.0	23445	0.7085	0.929
0.0055	16.0	25008	0.7183	0.9296
0.0027	17.0	26571	0.7296	0.9299
0.0005	18.0	28134	0.7465	0.9313
0.0004	19.0	29697	0.7610	0.9309
0.0	20.0	31260	0.7640	0.9305

Framework versions

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