whisper-small-cv-1
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: 0.2809
- Wer: 0.2588
- Cer: 0.2931
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 |
|---|---|---|---|---|---|
| 0.292 | 1.0 | 500 | 0.2837 | 0.2614 | 0.2950 |
| 0.1042 | 2.0 | 1000 | 0.2868 | 0.2584 | 0.2941 |
| 0.04 | 3.0 | 1500 | 0.2966 | 0.2625 | 0.2965 |
| 0.0102 | 4.0 | 2000 | 0.3444 | 0.2703 | 0.2966 |
| 0.0106 | 5.0 | 2500 | 0.4016 | 0.2648 | 0.2960 |
| 0.0018 | 6.0 | 3000 | 0.4035 | 0.2633 | 0.2959 |
| 0.001 | 7.0 | 3500 | 0.4156 | 0.2636 | 0.2956 |
| 0.0007 | 8.0 | 4000 | 0.4224 | 0.2610 | 0.2955 |
Framework versions
- Transformers 4.55.2
- Pytorch 2.7.0+cu126
- Datasets 2.19.1
- Tokenizers 0.21.4