<!-- 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. -->
opus-mt-en-es-finetuned-es-to-sja
This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-es on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.5834
- Bleu: 2.8027
- Gen Len: 85.512
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: 2e-05
- 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| No log | 1.0 | 199 | 2.2739 | 0.4646 | 138.5397 |
| No log | 2.0 | 398 | 1.9315 | 1.2125 | 95.1778 |
| 2.6483 | 3.0 | 597 | 1.8010 | 1.6411 | 88.4023 |
| 2.6483 | 4.0 | 796 | 1.7271 | 1.9492 | 89.2963 |
| 2.6483 | 5.0 | 995 | 1.6721 | 2.4431 | 86.1551 |
| 1.8418 | 6.0 | 1194 | 1.6353 | 2.4596 | 84.7491 |
| 1.8418 | 7.0 | 1393 | 1.6126 | 2.6671 | 85.6545 |
| 1.6984 | 8.0 | 1592 | 1.5978 | 2.663 | 83.6557 |
| 1.6984 | 9.0 | 1791 | 1.5861 | 2.8288 | 84.9042 |
| 1.6984 | 10.0 | 1990 | 1.5834 | 2.8027 | 85.512 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3
