metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- bleu
- rouge
model-index:
- name: esp-to-lsm-model-split
results: []
esp-to-lsm-model-split
This model is a fine-tuned version of Helsinki-NLP/opus-mt-es-es on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.5258
- Bleu: 83.8782
- Ter: 10.6805
- Rouge1: 0.9238
- Rouge2: 0.8539
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: 0.00015
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Ter | Rouge1 | Rouge2 |
|---|---|---|---|---|---|---|---|
| 1.0312 | 1.0 | 75 | 0.6799 | 70.9487 | 17.7958 | 0.8736 | 0.7858 |
| 0.4186 | 2.0 | 150 | 0.4926 | 50.7542 | 12.2855 | 0.9190 | 0.8516 |
| 0.2945 | 3.0 | 225 | 0.4174 | 72.8458 | 17.0732 | 0.9290 | 0.8649 |
| 0.1752 | 4.0 | 300 | 0.4505 | 81.4327 | 10.3884 | 0.9254 | 0.8485 |
| 0.1483 | 5.0 | 375 | 0.4863 | 39.1763 | 10.7498 | 0.9139 | 0.8411 |
| 0.1049 | 6.0 | 450 | 0.4795 | 81.2675 | 10.0271 | 0.9250 | 0.8572 |
| 0.0808 | 7.0 | 525 | 0.4824 | 80.2681 | 11.9241 | 0.9145 | 0.8251 |
| 0.0723 | 8.0 | 600 | 0.4388 | 83.9041 | 11.0208 | 0.9308 | 0.8690 |
| 0.0652 | 9.0 | 675 | 0.4277 | 85.0942 | 9.3948 | 0.9306 | 0.8718 |
| 0.0469 | 10.0 | 750 | 0.4310 | 84.8003 | 9.2141 | 0.9337 | 0.8766 |
| 0.043 | 11.0 | 825 | 0.4471 | 84.1771 | 9.9368 | 0.9297 | 0.8669 |
| 0.0278 | 12.0 | 900 | 0.4326 | 86.4289 | 8.5818 | 0.9404 | 0.8829 |
| 0.0258 | 13.0 | 975 | 0.4369 | 85.7966 | 8.7624 | 0.9363 | 0.8772 |
| 0.0258 | 14.0 | 1050 | 0.4296 | 86.5295 | 8.4914 | 0.9393 | 0.8856 |
| 0.0216 | 15.0 | 1125 | 0.4379 | 78.3156 | 14.7245 | 0.9321 | 0.8710 |
| 0.0132 | 16.0 | 1200 | 0.4400 | 87.1342 | 8.0397 | 0.9392 | 0.8839 |
| 0.0129 | 17.0 | 1275 | 0.4438 | 86.9282 | 8.1301 | 0.9383 | 0.8831 |
| 0.0109 | 18.0 | 1350 | 0.4445 | 86.7807 | 8.4914 | 0.9372 | 0.8813 |
| 0.0109 | 19.0 | 1425 | 0.4455 | 86.7006 | 8.4914 | 0.9371 | 0.8808 |
| 0.014 | 20.0 | 1500 | 0.4472 | 86.8515 | 8.3107 | 0.9378 | 0.882 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.5.0+cu121
- Datasets 3.0.2
- Tokenizers 0.13.3