<!-- 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-id-jakarta
This model was trained from scratch on the inglish dataset. It achieves the following results on the evaluation set:
- Loss: 0.3122
- Bleu: 81.3279
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: 1e-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
- lr_scheduler_warmup_steps: 4000
- num_epochs: 25
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu |
|---|---|---|---|---|
| 0.9934 | 1.0 | 272 | 0.7998 | 57.6098 |
| 0.8752 | 2.0 | 544 | 0.6945 | 60.2893 |
| 0.762 | 3.0 | 816 | 0.6118 | 63.4407 |
| 0.6825 | 4.0 | 1088 | 0.5591 | 65.9376 |
| 0.6178 | 5.0 | 1360 | 0.5200 | 67.9008 |
| 0.5655 | 6.0 | 1632 | 0.4866 | 69.3854 |
| 0.516 | 7.0 | 1904 | 0.4580 | 70.8301 |
| 0.4685 | 8.0 | 2176 | 0.4334 | 72.0389 |
| 0.428 | 9.0 | 2448 | 0.4102 | 73.1174 |
| 0.3871 | 10.0 | 2720 | 0.3908 | 74.4526 |
| 0.3507 | 11.0 | 2992 | 0.3750 | 75.4508 |
| 0.3154 | 12.0 | 3264 | 0.3619 | 76.2748 |
| 0.2845 | 13.0 | 3536 | 0.3491 | 77.0737 |
| 0.2549 | 14.0 | 3808 | 0.3390 | 77.958 |
| 0.2269 | 15.0 | 4080 | 0.3312 | 78.6107 |
| 0.2045 | 16.0 | 4352 | 0.3240 | 79.3878 |
| 0.183 | 17.0 | 4624 | 0.3200 | 79.7554 |
| 0.1662 | 18.0 | 4896 | 0.3176 | 80.1064 |
| 0.1539 | 19.0 | 5168 | 0.3154 | 80.3286 |
| 0.1439 | 20.0 | 5440 | 0.3127 | 80.7742 |
| 0.133 | 21.0 | 5712 | 0.3126 | 80.8623 |
| 0.1262 | 22.0 | 5984 | 0.3133 | 81.0054 |
| 0.1213 | 23.0 | 6256 | 0.3132 | 81.2101 |
| 0.1169 | 24.0 | 6528 | 0.3123 | 81.2647 |
| 0.1149 | 25.0 | 6800 | 0.3122 | 81.3279 |
Framework versions
- Transformers 4.26.1
- Pytorch 2.0.0
- Datasets 2.10.1
- Tokenizers 0.11.0
