metadata
library_name: transformers
license: apache-2.0
base_model: Helsinki-NLP/opus-mt-en-ar
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: Motarjem-v0.1
results: []
Motarjem-v0.1
This model is a fine-tuned version of Helsinki-NLP/opus-mt-en-ar on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.1540
- Bleu: 29.991
- Gen Len: 11.75
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: 0.001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 128
- 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_ratio: 0.03
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 3.7908 | 0.5857 | 1000 | 3.7377 | 6.18 | 16.2781 |
| 3.5926 | 1.1716 | 2000 | 3.3270 | 8.0493 | 16.4099 |
| 3.2234 | 1.7572 | 3000 | 3.0900 | 9.7393 | 15.2034 |
| 2.9178 | 2.3432 | 4000 | 2.8567 | 9.8133 | 17.6295 |
| 2.7302 | 2.9288 | 5000 | 2.6926 | 15.7344 | 12.8374 |
| 2.4324 | 3.5148 | 6000 | 2.5841 | 16.7132 | 12.9787 |
| 2.3135 | 4.1007 | 7000 | 2.4542 | 17.5718 | 13.288 |
| 2.0808 | 4.6864 | 8000 | 2.3554 | 18.0186 | 13.6381 |
| 1.9353 | 5.2723 | 9000 | 2.3031 | 20.0612 | 12.8465 |
| 1.8244 | 5.8580 | 10000 | 2.1954 | 19.1501 | 14.0303 |
| 1.6091 | 6.4439 | 11000 | 2.2027 | 23.3063 | 12.3297 |
| 1.5548 | 7.0299 | 12000 | 2.1910 | 23.6853 | 12.4534 |
| 1.325 | 7.6155 | 13000 | 2.1362 | 27.8799 | 11.4823 |
| 1.2595 | 8.2015 | 14000 | 2.1696 | 25.782 | 12.1096 |
| 1.1253 | 8.7871 | 15000 | 2.1173 | 27.8543 | 11.8945 |
| 0.9858 | 9.3731 | 16000 | 2.1696 | 29.3317 | 11.7985 |
| 0.938 | 9.9587 | 17000 | 2.1540 | 29.991 | 11.75 |
Framework versions
- Transformers 4.48.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0