_nougat_CharJawi_AHR
This model is a fine-tuned version of bustamiyusoef/_base_nougat_JawiChar on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 2.0337
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: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 6
- total_train_batch_size: 48
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 30
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 1.0 | 9 | 2.4993 |
| 24.9223 | 2.0 | 18 | 2.3432 |
| 13.5598 | 3.0 | 27 | 2.2779 |
| 13.0195 | 4.0 | 36 | 2.2595 |
| 12.7676 | 5.0 | 45 | 2.2114 |
| 12.4741 | 6.0 | 54 | 2.1787 |
| 12.2735 | 7.0 | 63 | 2.1623 |
| 12.0191 | 8.0 | 72 | 2.1403 |
| 11.8536 | 9.0 | 81 | 2.1117 |
| 10.9289 | 10.0 | 90 | 2.0991 |
| 10.9289 | 11.0 | 99 | 2.0867 |
| 11.5685 | 12.0 | 108 | 2.0819 |
| 11.4487 | 13.0 | 117 | 2.0693 |
| 11.2871 | 14.0 | 126 | 2.0658 |
| 11.1921 | 15.0 | 135 | 2.0634 |
| 11.0656 | 16.0 | 144 | 2.0563 |
| 11.029 | 17.0 | 153 | 2.0515 |
| 10.957 | 18.0 | 162 | 2.0413 |
| 10.9147 | 19.0 | 171 | 2.0462 |
| 9.9879 | 20.0 | 180 | 2.0382 |
| 9.9879 | 21.0 | 189 | 2.0379 |
| 10.7999 | 22.0 | 198 | 2.0354 |
| 10.6849 | 23.0 | 207 | 2.0379 |
| 10.8111 | 24.0 | 216 | 2.0358 |
| 10.7872 | 25.0 | 225 | 2.0336 |
| 10.665 | 26.0 | 234 | 2.0340 |
| 10.743 | 26.72 | 240 | 2.0337 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0