metadata
library_name: transformers
license: mit
base_model: facebook/m2m100_418M
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: m2m100-legal-translation-en-ur
results: []
m2m100-legal-translation-en-ur
This model is a fine-tuned version of facebook/m2m100_418M on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2290
- Bleu: 30.9872
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: 3e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- 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
- lr_scheduler_warmup_steps: 20
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu |
|---|---|---|---|---|
| 8.278 | 0.2688 | 25 | 6.3522 | 3.5831 |
| 5.3323 | 0.5376 | 50 | 4.9369 | 9.8940 |
| 4.3888 | 0.8065 | 75 | 3.7497 | 12.2129 |
| 3.0058 | 1.0753 | 100 | 2.6821 | 15.0216 |
| 2.2164 | 1.3441 | 125 | 1.7722 | 16.6899 |
| 1.2595 | 1.6129 | 150 | 1.0978 | 18.5312 |
| 0.8531 | 1.8817 | 175 | 0.7009 | 19.6160 |
| 0.579 | 2.1505 | 200 | 0.4962 | 21.4995 |
| 0.4599 | 2.4194 | 225 | 0.4073 | 22.5162 |
| 0.3362 | 2.6882 | 250 | 0.3502 | 23.5743 |
| 0.353 | 2.9570 | 275 | 0.3122 | 25.6336 |
| 0.313 | 3.2258 | 300 | 0.2887 | 27.3754 |
| 0.3446 | 3.4946 | 325 | 0.2702 | 28.0800 |
| 0.3345 | 3.7634 | 350 | 0.2541 | 29.5058 |
| 0.2018 | 4.0323 | 375 | 0.2439 | 30.0331 |
| 0.2381 | 4.3011 | 400 | 0.2360 | 30.5859 |
| 0.1609 | 4.5699 | 425 | 0.2315 | 30.8617 |
| 0.2169 | 4.8387 | 450 | 0.2290 | 30.9872 |
Framework versions
- Transformers 4.51.1
- Pytorch 2.5.1+cu124
- Datasets 3.5.0
- Tokenizers 0.21.0