<!-- 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-ro-finetuned-eng-to-para-finetuned-eng-to-para
This model is a fine-tuned version of Tri1/opus-mt-en-ro-finetuned-eng-to-para on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0565
- Bleu: 25.7815
- Gen Len: 21.92
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 0.0759 | 1.0 | 6250 | 0.0738 | 24.9996 | 23.28 |
| 0.0675 | 2.0 | 12500 | 0.0665 | 25.8205 | 23.0 |
| 0.063 | 3.0 | 18750 | 0.0620 | 26.0631 | 22.72 |
| 0.0593 | 4.0 | 25000 | 0.0590 | 24.7911 | 22.56 |
| 0.0569 | 5.0 | 31250 | 0.0571 | 25.548 | 21.92 |
| 0.0559 | 6.0 | 37500 | 0.0565 | 25.7815 | 21.92 |
Framework versions
- Transformers 4.30.1
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3
