roberta_sst5_padding30model

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

• Loss: 3.7712
• Accuracy: 0.5674

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
1.2819	1.0	534	1.1781	0.4683
1.0082	2.0	1068	0.9981	0.5715
0.8514	3.0	1602	1.0001	0.5715
0.7126	4.0	2136	1.1284	0.5665
0.582	5.0	2670	1.3241	0.5462
0.4655	6.0	3204	1.4761	0.5466
0.3524	7.0	3738	1.6291	0.5457
0.2865	8.0	4272	1.9088	0.5335
0.2393	9.0	4806	2.0502	0.5538
0.2127	10.0	5340	2.3658	0.5516
0.1828	11.0	5874	2.9156	0.5493
0.1483	12.0	6408	2.9242	0.5566
0.1433	13.0	6942	3.2224	0.5362
0.1184	14.0	7476	3.3634	0.5502
0.1026	15.0	8010	3.4638	0.5602
0.0826	16.0	8544	3.5596	0.5511
0.0571	17.0	9078	3.5359	0.5710
0.0416	18.0	9612	3.7094	0.5579
0.0345	19.0	10146	3.7244	0.5674
0.0297	20.0	10680	3.7712	0.5674

Framework versions

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