recipe-distilbert-i

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

• Loss: 1.0288

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
2.3931	1.0	152	1.7738
1.7533	2.0	304	1.5109
1.5584	3.0	456	1.4003
1.443	4.0	608	1.3296
1.3551	5.0	760	1.2270
1.2981	6.0	912	1.1870
1.2577	7.0	1064	1.1511
1.2216	8.0	1216	1.1298
1.1958	9.0	1368	1.1087
1.1685	10.0	1520	1.0858
1.1533	11.0	1672	1.0820
1.1358	12.0	1824	1.0659
1.1286	13.0	1976	1.0382
1.1128	14.0	2128	1.0468
1.11	15.0	2280	1.0399
1.094	16.0	2432	1.0382
1.0969	17.0	2584	1.0096
1.0868	18.0	2736	1.0235
1.0845	19.0	2888	1.0227
1.0855	20.0	3040	1.0288

Framework versions

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