<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
spell_correction_F03_LM
This model is a fine-tuned version of facebook/bart-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0490
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: 1e-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: 30
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 1.0 | 248 | 1.0015 |
| No log | 2.0 | 496 | 0.3499 |
| 2.4465 | 3.0 | 744 | 0.1456 |
| 2.4465 | 4.0 | 992 | 0.0854 |
| 0.1986 | 5.0 | 1240 | 0.0689 |
| 0.1986 | 6.0 | 1488 | 0.0619 |
| 0.0578 | 7.0 | 1736 | 0.0571 |
| 0.0578 | 8.0 | 1984 | 0.0525 |
| 0.0269 | 9.0 | 2232 | 0.0524 |
| 0.0269 | 10.0 | 2480 | 0.0517 |
| 0.0172 | 11.0 | 2728 | 0.0498 |
| 0.0172 | 12.0 | 2976 | 0.0488 |
| 0.0113 | 13.0 | 3224 | 0.0500 |
| 0.0113 | 14.0 | 3472 | 0.0493 |
| 0.0088 | 15.0 | 3720 | 0.0496 |
| 0.0088 | 16.0 | 3968 | 0.0495 |
| 0.0077 | 17.0 | 4216 | 0.0477 |
| 0.0077 | 18.0 | 4464 | 0.0489 |
| 0.0066 | 19.0 | 4712 | 0.0476 |
| 0.0066 | 20.0 | 4960 | 0.0489 |
| 0.0064 | 21.0 | 5208 | 0.0486 |
| 0.0064 | 22.0 | 5456 | 0.0494 |
| 0.0063 | 23.0 | 5704 | 0.0494 |
| 0.0063 | 24.0 | 5952 | 0.0497 |
| 0.004 | 25.0 | 6200 | 0.0495 |
| 0.004 | 26.0 | 6448 | 0.0493 |
| 0.0045 | 27.0 | 6696 | 0.0491 |
| 0.0045 | 28.0 | 6944 | 0.0490 |
| 0.0039 | 29.0 | 7192 | 0.0490 |
| 0.0039 | 30.0 | 7440 | 0.0490 |
Framework versions
- Transformers 4.28.0
- Pytorch 1.12.1+cu102
- Datasets 2.13.1
- Tokenizers 0.13.3
