N_roberta_imdb_padding70model

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

• Loss: 0.4584
• Accuracy: 0.9506

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
0.2111	1.0	1563	0.1864	0.9412
0.1628	2.0	3126	0.1973	0.9474
0.1178	3.0	4689	0.3140	0.9406
0.0745	4.0	6252	0.2698	0.9469
0.057	5.0	7815	0.3568	0.9424
0.0478	6.0	9378	0.3114	0.9477
0.0281	7.0	10941	0.3123	0.9491
0.0269	8.0	12504	0.3732	0.9464
0.0169	9.0	14067	0.4043	0.9466
0.0156	10.0	15630	0.3296	0.9480
0.0186	11.0	17193	0.4413	0.9466
0.015	12.0	18756	0.3944	0.9488
0.0091	13.0	20319	0.4310	0.9481
0.0061	14.0	21882	0.4363	0.9506
0.0073	15.0	23445	0.4315	0.9496
0.003	16.0	25008	0.4623	0.9499
0.0048	17.0	26571	0.4589	0.9494
0.0	18.0	28134	0.4556	0.9504
0.0028	19.0	29697	0.4608	0.9509
0.0013	20.0	31260	0.4584	0.9506

Framework versions

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