opus-mt-mul-en-finetuned-dz-to-en-n1
This model was trained from scratch on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 2.2735
- Bleu: 12.0953
- Gen Len: 15.8819
Model description
More information needed
Intended uses & limitations
Gives inacurrate translation, use at your own risk
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: 16
- eval_batch_size: 16
- seed: 42
- 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: 9
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|
| 3.2761 | 1.0 | 562 | 2.9572 | 5.5363 | 15.5395 |
| 2.8321 | 2.0 | 1124 | 2.6795 | 7.9287 | 15.2723 |
| 2.5606 | 3.0 | 1686 | 2.5424 | 9.3415 | 15.5245 |
| 2.346 | 4.0 | 2248 | 2.4262 | 10.811 | 15.8619 |
| 2.1696 | 5.0 | 2810 | 2.3613 | 10.7336 | 16.0801 |
| 2.0639 | 6.0 | 3372 | 2.3282 | 11.2871 | 15.8388 |
| 1.9692 | 7.0 | 3934 | 2.2879 | 12.7783 | 15.4164 |
| 1.893 | 8.0 | 4496 | 2.2758 | 12.6661 | 15.3634 |
| 1.7954 | 9.0 | 5058 | 2.2735 | 12.0953 | 15.8819 |
Framework versions
- Transformers 4.47.1
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0
