Whisper Small ESPAÑOL - GRUPO 1
This model is a fine-tuned version of openai/whisper-small on the google-fleurs dataset. It achieves the following results on the evaluation set:
- Loss: 0.1817
- Wer: 7.7406
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 50
- training_steps: 400
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.0475 | 5.0 | 25 | 0.3134 | 9.4142 |
| 0.4441 | 10.0 | 50 | 0.1597 | 7.5314 |
| 0.0936 | 15.0 | 75 | 0.1676 | 7.7406 |
| 0.0147 | 20.0 | 100 | 0.1719 | 7.5314 |
| 0.0069 | 25.0 | 125 | 0.1720 | 7.7406 |
| 0.0039 | 30.0 | 150 | 0.1730 | 7.7406 |
| 0.0025 | 35.0 | 175 | 0.1756 | 7.7406 |
| 0.0012 | 40.0 | 200 | 0.1765 | 7.7406 |
| 0.0008 | 45.0 | 225 | 0.1780 | 7.7406 |
| 0.0006 | 50.0 | 250 | 0.1789 | 7.7406 |
| 0.0006 | 55.0 | 275 | 0.1798 | 7.7406 |
| 0.0005 | 60.0 | 300 | 0.1805 | 7.7406 |
| 0.0005 | 65.0 | 325 | 0.1810 | 7.7406 |
| 0.0005 | 70.0 | 350 | 0.1814 | 7.7406 |
| 0.0005 | 75.0 | 375 | 0.1817 | 7.7406 |
| 0.0005 | 80.0 | 400 | 0.1817 | 7.7406 |
Framework versions
- Transformers 5.12.1
- Pytorch 2.11.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2