recipe-roberta-i

This model is a fine-tuned version of roberta-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9919

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: 256
• eval_batch_size: 256
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
1.871	1.0	149	1.4670
1.528	2.0	298	1.3426
1.41	3.0	447	1.2636
1.3332	4.0	596	1.2029
1.2804	5.0	745	1.1646
1.2441	6.0	894	1.1351
1.21	7.0	1043	1.0898
1.182	8.0	1192	1.0725
1.1604	9.0	1341	1.0718
1.1402	10.0	1490	1.0529
1.1308	11.0	1639	1.0512
1.1191	12.0	1788	1.0245
1.0986	13.0	1937	1.0203
1.0919	14.0	2086	1.0158
1.084	15.0	2235	0.9930
1.0797	16.0	2384	0.9855
1.0697	17.0	2533	1.0061
1.0652	18.0	2682	0.9725
1.0658	19.0	2831	0.9861
1.0642	20.0	2980	0.9919

Framework versions

• Transformers 4.19.0.dev0
• Pytorch 1.11.0+cu102
• Datasets 2.3.2
• Tokenizers 0.12.1