mt-en-vi
This model is a fine-tuned version of facebook/mbart-large-50-many-to-many-mmt on a dataset from iwslt2015. It achieves the following results on the evaluation set:
- Loss: 1.2564
- Bleu: 34.9585
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.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: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu |
|---|---|---|---|---|
| 1.1887 | 0.2399 | 500 | 1.2961 | 34.2118 |
| 1.1422 | 0.4798 | 1000 | 1.2807 | 34.6929 |
| 1.127 | 0.7198 | 1500 | 1.2689 | 34.7625 |
| 1.1215 | 0.9597 | 2000 | 1.2577 | 34.5726 |
| 1.031 | 1.1996 | 2500 | 1.2728 | 35.0683 |
| 1.0128 | 1.4395 | 3000 | 1.2632 | 34.8100 |
| 1.0184 | 1.6795 | 3500 | 1.2565 | 34.7721 |
| 1.0131 | 1.9194 | 4000 | 1.2564 | 34.9585 |
Framework versions
- Transformers 4.49.0
- Pytorch 2.6.0+cu124
- Datasets 2.20.0
- Tokenizers 0.21.1
- Downloads last month
- 10
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for hai2131/mt-en-vi
Base model
facebook/mbart-large-50-many-to-many-mmt