iva_mt_wslot-m2m100_418M-zh-en
This model is a fine-tuned version of facebook/m2m100_418M on the iva_mt_wslot dataset. It achieves the following results on the evaluation set:
- Loss: 0.0139
- Bleu: 67.9385
- Gen Len: 18.9988
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: 4
- eval_batch_size: 4
- 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 |
|---|---|---|---|---|---|
| 0.016 | 1.0 | 2437 | 0.0138 | 62.077 | 18.6077 |
| 0.0114 | 2.0 | 4874 | 0.0126 | 64.3834 | 18.9019 |
| 0.0084 | 3.0 | 7311 | 0.0123 | 66.0012 | 18.9206 |
| 0.0067 | 4.0 | 9748 | 0.0123 | 66.7838 | 19.0018 |
| 0.005 | 5.0 | 12185 | 0.0124 | 66.9053 | 18.9527 |
| 0.0039 | 6.0 | 14622 | 0.0128 | 67.5252 | 18.9918 |
| 0.003 | 7.0 | 17059 | 0.0131 | 67.3664 | 18.9609 |
| 0.0025 | 8.0 | 19496 | 0.0135 | 67.792 | 19.0198 |
| 0.0019 | 9.0 | 21933 | 0.0137 | 67.7256 | 18.9591 |
| 0.0015 | 10.0 | 24370 | 0.0139 | 67.9385 | 18.9988 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.1.0+cu121
- Datasets 2.15.0
- Tokenizers 0.13.3