whisper-es-pr-pe-lora-unsloth-small-v13
This model is a fine-tuned version of unsloth/whisper-small on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.1560
- Wer: 0.0817
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
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Use OptimizerNames.ADAMW_8BIT 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: 200
- training_steps: 1200
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3511 | 0.3333 | 400 | 0.3670 | 0.0933 |
| 0.1454 | 1.0983 | 800 | 0.1746 | 0.0857 |
| 0.1264 | 1.4317 | 1200 | 0.1560 | 0.0817 |
Framework versions
- PEFT 0.18.0
- Transformers 4.57.3
- Pytorch 2.9.1+cu128
- Datasets 4.3.0
- Tokenizers 0.22.1