whisper-small-arabic-letters-pod
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: 0.0493
- Wer: 6.3830
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 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
- training_steps: 1000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0415 | 2.7778 | 200 | 0.0298 | 6.8493 |
| 0.0107 | 5.5556 | 400 | 0.0280 | 5.4795 |
| 0.0061 | 8.3333 | 600 | 0.0248 | 6.1644 |
| 0.0067 | 11.1111 | 800 | 0.0283 | 8.2192 |
| 0.0057 | 13.8889 | 1000 | 0.0321 | 8.9041 |
Framework versions
- Transformers 5.12.1
- Pytorch 2.4.1+cu124
- Datasets 5.0.0
- Tokenizers 0.22.2