Whisper tiny Spanish
This model is a fine-tuned version of openai/whisper-tiny on the Spanish English dataset. It achieves the following results on the evaluation set:
- Loss: 0.2429
- Wer: 15.2239
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: 3.75e-05
- train_batch_size: 16
- eval_batch_size: 16
- 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: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3514 | 1.0 | 129 | 0.2619 | 13.5323 |
| 0.2228 | 2.0 | 258 | 0.2443 | 14.6269 |
| 0.0890 | 3.0 | 387 | 0.2345 | 12.3632 |
| 0.0448 | 4.0 | 516 | 0.2576 | 13.9552 |
| 0.0143 | 5.0 | 645 | 0.2429 | 15.2239 |
Framework versions
- Transformers 5.21.0.dev0
- Pytorch 2.11.0+cu130
- Datasets 5.1.0
- Tokenizers 0.23.2