metadata
base_model: drewcurran/translation_model
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: model-spanglish
results: []
model-spanglish
This model is a fine-tuned version of drewcurran/translation_model on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.5595
- Bleu: 5.861
- Gen Len: 17.8
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: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| 2.3242 | 1.0 | 10 | 1.8794 | 1.372 | 18.4 |
| 2.2034 | 2.0 | 20 | 1.7643 | 1.8888 | 18.125 |
| 2.0818 | 3.0 | 30 | 1.6884 | 2.1152 | 18.125 |
| 2.0447 | 4.0 | 40 | 1.6408 | 2.0966 | 17.925 |
| 1.9457 | 5.0 | 50 | 1.6089 | 2.286 | 17.925 |
| 1.9454 | 6.0 | 60 | 1.5881 | 2.3929 | 17.95 |
| 1.8906 | 7.0 | 70 | 1.5756 | 2.4497 | 17.95 |
| 1.8829 | 8.0 | 80 | 1.5669 | 4.2201 | 18.025 |
| 1.8483 | 9.0 | 90 | 1.5613 | 5.861 | 17.8 |
| 1.8454 | 10.0 | 100 | 1.5595 | 5.861 | 17.8 |
Framework versions
- Transformers 4.40.1
- Pytorch 2.2.1+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1