<!-- 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. -->
PSST_spell_correction_V3
This model is a fine-tuned version of facebook/bart-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5390
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: 8
- eval_batch_size: 8
- 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 |
|---|---|---|---|
| 0.3418 | 1.0 | 3099 | 0.3206 |
| 0.2821 | 2.0 | 6198 | 0.3216 |
| 0.2467 | 3.0 | 9297 | 0.3501 |
| 0.2301 | 4.0 | 12396 | 0.3747 |
| 0.2152 | 5.0 | 15495 | 0.3614 |
| 0.2071 | 6.0 | 18594 | 0.3836 |
| 0.2002 | 7.0 | 21693 | 0.3923 |
| 0.1964 | 8.0 | 24792 | 0.4053 |
| 0.1953 | 9.0 | 27891 | 0.4154 |
| 0.1935 | 10.0 | 30990 | 0.4269 |
| 0.1918 | 11.0 | 34089 | 0.4381 |
| 0.1889 | 12.0 | 37188 | 0.4623 |
| 0.1886 | 13.0 | 40287 | 0.4586 |
| 0.1869 | 14.0 | 43386 | 0.4838 |
| 0.1867 | 15.0 | 46485 | 0.4984 |
| 0.1857 | 16.0 | 49584 | 0.4623 |
| 0.1843 | 17.0 | 52683 | 0.5020 |
| 0.1834 | 18.0 | 55782 | 0.5129 |
| 0.1832 | 19.0 | 58881 | 0.4972 |
| 0.1835 | 20.0 | 61980 | 0.4895 |
| 0.1822 | 21.0 | 65079 | 0.4992 |
| 0.1834 | 22.0 | 68178 | 0.5291 |
| 0.1816 | 23.0 | 71277 | 0.5351 |
| 0.1802 | 24.0 | 74376 | 0.5294 |
| 0.1796 | 25.0 | 77475 | 0.5443 |
| 0.1814 | 26.0 | 80574 | 0.5368 |
| 0.1804 | 27.0 | 83673 | 0.5375 |
| 0.1811 | 28.0 | 86772 | 0.5435 |
| 0.1811 | 29.0 | 89871 | 0.5391 |
| 0.1802 | 30.0 | 92970 | 0.5390 |
Framework versions
- Transformers 4.28.0
- Pytorch 1.12.1+cu102
- Datasets 2.13.1
- Tokenizers 0.13.3
