N_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: 3.5511
• Accuracy: 0.5647

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
1.4421	1.0	534	1.1883	0.4525
1.0818	2.0	1068	1.0264	0.5471
0.9047	3.0	1602	1.0223	0.5593
0.7803	4.0	2136	1.1087	0.5561
0.6307	5.0	2670	1.3012	0.5434
0.5203	6.0	3204	1.3949	0.5638
0.4113	7.0	3738	1.6258	0.5561
0.319	8.0	4272	1.6429	0.5706
0.2758	9.0	4806	1.8020	0.5606
0.225	10.0	5340	2.2183	0.5552
0.2068	11.0	5874	2.6069	0.5376
0.1818	12.0	6408	2.7364	0.5511
0.1518	13.0	6942	2.9808	0.5570
0.1545	14.0	7476	3.2431	0.5624
0.096	15.0	8010	3.2850	0.5606
0.0852	16.0	8544	3.4121	0.5597
0.0698	17.0	9078	3.4816	0.5652
0.0541	18.0	9612	3.5019	0.5692
0.0453	19.0	10146	3.5686	0.5670
0.038	20.0	10680	3.5511	0.5647

Framework versions

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