distilbert_imdb_padding80model

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: 0.7659
• Accuracy: 0.9289

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
0.2429	1.0	1563	0.2099	0.923
0.1777	2.0	3126	0.2439	0.9231
0.116	3.0	4689	0.3167	0.9248
0.0668	4.0	6252	0.3296	0.9229
0.0448	5.0	7815	0.4632	0.9216
0.0326	6.0	9378	0.5330	0.9122
0.0275	7.0	10941	0.5065	0.9242
0.0187	8.0	12504	0.5384	0.9238
0.0225	9.0	14067	0.4589	0.9260
0.0069	10.0	15630	0.6072	0.9269
0.0176	11.0	17193	0.5474	0.9269
0.0106	12.0	18756	0.6218	0.9272
0.0068	13.0	20319	0.6779	0.9263
0.0068	14.0	21882	0.6249	0.9263
0.0005	15.0	23445	0.6835	0.929
0.0014	16.0	25008	0.7223	0.929
0.002	17.0	26571	0.7401	0.9281
0.0011	18.0	28134	0.7324	0.9293
0.0	19.0	29697	0.7678	0.9289
0.0002	20.0	31260	0.7659	0.9289

Framework versions

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