Regression_BERT_NOaug_MSEloss

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

  • Loss: 0.4928
  • Mse: 0.4928
  • Mae: 0.6337
  • R2: 0.0926
  • Accuracy: 0.4737

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: 4
  • eval_batch_size: 4
  • 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 Mse Mae R2 Accuracy
No log 1.0 33 0.3184 0.3184 0.5205 0.0487 0.5946
No log 2.0 66 0.2439 0.2439 0.3571 0.2712 0.7027
No log 3.0 99 0.2950 0.2950 0.3792 0.1185 0.6757
No log 4.0 132 0.3179 0.3179 0.4267 0.0503 0.6757
No log 5.0 165 0.2869 0.2869 0.3984 0.1426 0.6757
No log 6.0 198 0.2967 0.2967 0.3688 0.1134 0.7027
No log 7.0 231 0.2797 0.2797 0.3599 0.1643 0.7027
No log 8.0 264 0.2730 0.2730 0.3438 0.1844 0.7027
No log 9.0 297 0.2813 0.2813 0.3623 0.1596 0.7027
No log 10.0 330 0.2733 0.2733 0.3296 0.1835 0.7027
No log 11.0 363 0.2770 0.2770 0.3432 0.1725 0.7027
No log 12.0 396 0.3009 0.3009 0.3574 0.1010 0.6757
No log 13.0 429 0.2735 0.2735 0.3318 0.1827 0.7027
No log 14.0 462 0.2787 0.2787 0.3341 0.1672 0.7027
No log 15.0 495 0.2790 0.2790 0.3312 0.1663 0.7297
0.0804 16.0 528 0.2683 0.2683 0.3229 0.1984 0.7027
0.0804 17.0 561 0.2749 0.2749 0.3273 0.1785 0.7297
0.0804 18.0 594 0.2709 0.2709 0.3202 0.1906 0.7297
0.0804 19.0 627 0.2711 0.2711 0.3205 0.1901 0.7297
0.0804 20.0 660 0.2694 0.2694 0.3197 0.1950 0.7297

Framework versions

  • Transformers 4.28.1
  • Pytorch 2.0.0+cu118
  • Datasets 2.12.0
  • Tokenizers 0.13.3
Downloads last month
18
Inference Examples
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.