<!-- 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. -->
tst-translation-output4
This model is a fine-tuned version of facebook/mbart-large-cc25 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.3569
- Bleu: 11.508
- Gen Len: 14.3854
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: 5e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 16
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1500
- num_epochs: 40
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 1.5428 | 0.86 | 2000 | 1.4469 | 8.6101 | 16.9959 |
| 1.1195 | 1.71 | 4000 | 1.3471 | 10.3699 | 14.5055 |
| 0.8028 | 2.57 | 6000 | 1.3569 | 11.508 | 14.3955 |
| 0.5856 | 3.43 | 8000 | 1.4776 | 10.7663 | 15.0208 |
| 0.378 | 4.29 | 10000 | 1.6134 | 10.5237 | 14.3284 |
| 0.3093 | 5.14 | 12000 | 1.7701 | 10.7522 | 14.0819 |
Framework versions
- Transformers 4.34.0
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.14.0
