<!-- 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-id-en-ccmatrix-no-warmup
This model was trained from scratch on the ccmatrix dataset. It achieves the following results on the evaluation set:
- Loss: 0.9795
- Bleu: 56.6595
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: 0.0001
- 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: 25
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu |
|---|---|---|---|---|
| 0.8307 | 1.0 | 28125 | 0.8213 | 50.8248 |
| 0.7031 | 2.0 | 56250 | 0.7851 | 52.0712 |
| 0.6426 | 3.0 | 84375 | 0.7716 | 52.6974 |
| 0.5976 | 4.0 | 112500 | 0.7645 | 53.0742 |
| 0.5599 | 5.0 | 140625 | 0.7580 | 53.6526 |
| 0.5265 | 6.0 | 168750 | 0.7610 | 53.9336 |
| 0.4962 | 7.0 | 196875 | 0.7592 | 54.3884 |
| 0.4677 | 8.0 | 225000 | 0.7611 | 54.4904 |
| 0.4404 | 9.0 | 253125 | 0.7680 | 54.6679 |
| 0.4148 | 10.0 | 281250 | 0.7760 | 54.8934 |
| 0.3901 | 11.0 | 309375 | 0.7853 | 55.1644 |
| 0.3662 | 12.0 | 337500 | 0.7945 | 55.3496 |
| 0.3432 | 13.0 | 365625 | 0.8071 | 55.4056 |
| 0.3211 | 14.0 | 393750 | 0.8149 | 55.5914 |
| 0.2995 | 15.0 | 421875 | 0.8314 | 55.8594 |
| 0.2789 | 16.0 | 450000 | 0.8504 | 55.5482 |
| 0.2591 | 17.0 | 478125 | 0.8637 | 55.8965 |
| 0.24 | 18.0 | 506250 | 0.8789 | 55.9817 |
| 0.222 | 19.0 | 534375 | 0.8956 | 56.046 |
| 0.2047 | 20.0 | 562500 | 0.9096 | 56.1539 |
| 0.1888 | 21.0 | 590625 | 0.9279 | 56.2939 |
| 0.1739 | 22.0 | 618750 | 0.9409 | 56.4546 |
| 0.1606 | 23.0 | 646875 | 0.9580 | 56.5525 |
| 0.1488 | 24.0 | 675000 | 0.9713 | 56.6069 |
| 0.1392 | 25.0 | 703125 | 0.9795 | 56.6595 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.0
- Datasets 2.10.1
- Tokenizers 0.11.0
