<!-- 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_F04_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: 0.0242
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 | 269 | 0.3354 |
| 1.9104 | 2.0 | 538 | 0.0604 |
| 1.9104 | 3.0 | 807 | 0.0405 |
| 0.0433 | 4.0 | 1076 | 0.0337 |
| 0.0433 | 5.0 | 1345 | 0.0285 |
| 0.0164 | 6.0 | 1614 | 0.0263 |
| 0.0164 | 7.0 | 1883 | 0.0268 |
| 0.0099 | 8.0 | 2152 | 0.0244 |
| 0.0099 | 9.0 | 2421 | 0.0262 |
| 0.0065 | 10.0 | 2690 | 0.0264 |
| 0.0065 | 11.0 | 2959 | 0.0255 |
| 0.0044 | 12.0 | 3228 | 0.0243 |
| 0.0044 | 13.0 | 3497 | 0.0262 |
| 0.0037 | 14.0 | 3766 | 0.0247 |
| 0.0035 | 15.0 | 4035 | 0.0233 |
| 0.0035 | 16.0 | 4304 | 0.0230 |
| 0.0031 | 17.0 | 4573 | 0.0242 |
| 0.0031 | 18.0 | 4842 | 0.0223 |
| 0.0029 | 19.0 | 5111 | 0.0226 |
| 0.0029 | 20.0 | 5380 | 0.0245 |
| 0.0023 | 21.0 | 5649 | 0.0249 |
| 0.0023 | 22.0 | 5918 | 0.0244 |
| 0.0023 | 23.0 | 6187 | 0.0238 |
| 0.0023 | 24.0 | 6456 | 0.0237 |
| 0.0025 | 25.0 | 6725 | 0.0242 |
| 0.0025 | 26.0 | 6994 | 0.0239 |
| 0.0017 | 27.0 | 7263 | 0.0236 |
| 0.0018 | 28.0 | 7532 | 0.0242 |
| 0.0018 | 29.0 | 7801 | 0.0240 |
| 0.0016 | 30.0 | 8070 | 0.0242 |
Framework versions
- Transformers 4.28.0
- Pytorch 1.12.1+cu102
- Datasets 2.13.1
- Tokenizers 0.13.3
