marianmt_sanlish_to_bangla_1934_lr2_duijonChara
This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-mul on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6735
- Bleu: 14.5937
- Chrf: 39.3821
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 25
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf |
|---|---|---|---|---|---|
| 2.8387 | 1.0 | 196 | 2.2046 | 0.6422 | 15.3476 |
| 2.1040 | 2.0 | 392 | 1.9528 | 1.3720 | 19.8281 |
| 1.8124 | 3.0 | 588 | 1.7823 | 2.9177 | 24.4750 |
| 1.5736 | 4.0 | 784 | 1.6974 | 6.5991 | 28.8474 |
| 1.3648 | 5.0 | 980 | 1.6412 | 5.6960 | 30.1347 |
| 1.2160 | 6.0 | 1176 | 1.6423 | 9.0882 | 32.2068 |
| 1.0999 | 7.0 | 1372 | 1.6123 | 9.3129 | 33.2494 |
| 1.0003 | 8.0 | 1568 | 1.6063 | 10.9129 | 33.4587 |
| 0.9010 | 9.0 | 1764 | 1.5883 | 10.5600 | 35.6147 |
| 0.8126 | 10.0 | 1960 | 1.5826 | 13.8346 | 36.8960 |
| 0.7450 | 11.0 | 2156 | 1.5904 | 10.9749 | 37.7168 |
| 0.6777 | 12.0 | 2352 | 1.6041 | 14.4225 | 38.6026 |
| 0.6260 | 13.0 | 2548 | 1.5994 | 14.0511 | 37.9777 |
| 0.5762 | 14.0 | 2744 | 1.6140 | 15.6413 | 40.1306 |
| 0.5417 | 15.0 | 2940 | 1.6293 | 14.7248 | 39.4384 |
| 0.4926 | 16.0 | 3136 | 1.6299 | 15.6485 | 39.7425 |
| 0.4622 | 17.0 | 3332 | 1.6423 | 15.1540 | 38.6979 |
| 0.4387 | 18.0 | 3528 | 1.6342 | 15.1172 | 38.8870 |
| 0.4134 | 19.0 | 3724 | 1.6588 | 14.4538 | 39.5993 |
| 0.3898 | 20.0 | 3920 | 1.6585 | 14.5759 | 39.0584 |
| 0.3818 | 21.0 | 4116 | 1.6620 | 14.5808 | 39.0646 |
| 0.3653 | 22.0 | 4312 | 1.6677 | 14.0915 | 39.0470 |
| 0.3502 | 23.0 | 4508 | 1.6737 | 14.2037 | 38.8062 |
| 0.3457 | 24.0 | 4704 | 1.6736 | 14.4161 | 39.4732 |
| 0.3424 | 25.0 | 4900 | 1.6735 | 14.5937 | 39.3821 |
Framework versions
- Transformers 5.15.0
- Pytorch 2.11.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2