recipe-roberta-s

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

• Loss: 0.8870

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
1.3387	1.0	820	1.1529
1.187	2.0	1640	1.0643
1.1213	3.0	2460	1.0371
1.0859	4.0	3280	1.0000
1.0566	5.0	4100	0.9798
1.0338	6.0	4920	0.9637
1.0162	7.0	5740	0.9538
1.0003	8.0	6560	0.9332
0.9878	9.0	7380	0.9252
0.9767	10.0	8200	0.9189
0.9664	11.0	9020	0.9145
0.9627	12.0	9840	0.9065
0.9539	13.0	10660	0.9027
0.9461	14.0	11480	0.9029
0.9435	15.0	12300	0.8949
0.9404	16.0	13120	0.8924
0.9359	17.0	13940	0.8874
0.9307	18.0	14760	0.8842
0.9295	19.0	15580	0.8853
0.9263	20.0	16400	0.8870

Framework versions

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