roberta_sst5_padding0model

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

• Loss: 3.7599
• Accuracy: 0.5538

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
1.2226	1.0	534	1.1734	0.4774
0.9839	2.0	1068	1.0370	0.5561
0.8147	3.0	1602	1.0445	0.5701
0.693	4.0	2136	1.2026	0.5484
0.54	5.0	2670	1.2741	0.5597
0.441	6.0	3204	1.5593	0.5367
0.3333	7.0	3738	1.7048	0.5407
0.2661	8.0	4272	1.9049	0.5548
0.2292	9.0	4806	2.2051	0.5548
0.1968	10.0	5340	2.3851	0.5507
0.1714	11.0	5874	2.8158	0.5489
0.1524	12.0	6408	2.9160	0.5471
0.1367	13.0	6942	3.2566	0.5357
0.1087	14.0	7476	3.3613	0.5525
0.0874	15.0	8010	3.3934	0.5548
0.0561	16.0	8544	3.5486	0.5584
0.0637	17.0	9078	3.5989	0.5525
0.0402	18.0	9612	3.7290	0.5548
0.036	19.0	10146	3.7058	0.5552
0.0296	20.0	10680	3.7599	0.5538

Framework versions

• Transformers 4.32.1
• Pytorch 2.1.1
• Datasets 2.12.0
• Tokenizers 0.13.3