whisper-small-ptbr-cv-final1
This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4421
- Wer: 0.3407
- Cer: 0.4024
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: 8
- eval_batch_size: 2
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH 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: 100
- num_epochs: 12
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.393 | 1.0 | 250 | 0.4150 | 0.3464 | 0.4042 |
| 0.1372 | 2.0 | 500 | 0.4316 | 0.3415 | 0.4011 |
| 0.0399 | 3.0 | 750 | 0.4489 | 0.3505 | 0.4063 |
| 0.0201 | 4.0 | 1000 | 0.5175 | 0.3595 | 0.4095 |
| 0.007 | 5.0 | 1250 | 0.5595 | 0.3578 | 0.4064 |
| 0.0023 | 6.0 | 1500 | 0.5547 | 0.3603 | 0.4120 |
| 0.0012 | 7.0 | 1750 | 0.6038 | 0.3587 | 0.4073 |
| 0.0009 | 8.0 | 2000 | 0.6134 | 0.3619 | 0.4109 |
Framework versions
- Transformers 4.55.0
- Pytorch 2.7.0+cu126
- Datasets 2.19.1
- Tokenizers 0.21.4