whisper-medium-azerbaijani-v1
This model is a fine-tuned version of openai/whisper-medium on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0521
- Wer: 6.3303
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: 1500
- num_epochs: 12
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0408 | 2.1038 | 1500 | 0.0797 | 10.5163 |
| 0.0097 | 4.2076 | 3000 | 0.0527 | 8.2184 |
| 0.0019 | 6.3114 | 4500 | 0.0494 | 7.3314 |
| 0.001 | 8.4151 | 6000 | 0.0507 | 6.1806 |
| 0.0002 | 10.5189 | 7500 | 0.0521 | 6.3303 |
Framework versions
- Transformers 4.57.0
- Pytorch 2.10.0.dev20251007+cu128
- Datasets 3.6.0
- Tokenizers 0.22.1