N_roberta_imdb_padding50model

This model is a fine-tuned version of roberta-base on the imdb dataset. It achieves the following results on the evaluation set:

• Loss: 0.5385
• Accuracy: 0.9530

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
0.2002	1.0	1563	0.2254	0.9357
0.1628	2.0	3126	0.1732	0.9478
0.115	3.0	4689	0.2905	0.9365
0.0737	4.0	6252	0.2347	0.9474
0.062	5.0	7815	0.3516	0.9472
0.0466	6.0	9378	0.3532	0.9452
0.0295	7.0	10941	0.3115	0.9481
0.0213	8.0	12504	0.4286	0.9479
0.0196	9.0	14067	0.4348	0.9483
0.019	10.0	15630	0.5160	0.9376
0.0177	11.0	17193	0.4682	0.9467
0.004	12.0	18756	0.4670	0.9503
0.0076	13.0	20319	0.4573	0.9501
0.0054	14.0	21882	0.5279	0.9504
0.0055	15.0	23445	0.4883	0.9504
0.0051	16.0	25008	0.4782	0.9525
0.0021	17.0	26571	0.4732	0.9527
0.0007	18.0	28134	0.5154	0.9519
0.0029	19.0	29697	0.5317	0.9524
0.002	20.0	31260	0.5385	0.9530

Framework versions

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