Whisper Small TK - Abdyrahman Gudratullayew
This model is a fine-tuned version of openai/whisper-small on the Common Voice 17.0 dataset. It achieves the following results on the evaluation set:
- Loss: 1.3114
- Wer: 57.9337
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: 500
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0083 | 14.0845 | 1000 | 1.1117 | 60.3571 |
| 0.0003 | 28.1690 | 2000 | 1.2099 | 57.7041 |
| 0.0002 | 42.2535 | 3000 | 1.2640 | 58.0102 |
| 0.0001 | 56.3380 | 4000 | 1.2973 | 58.1378 |
| 0.0001 | 70.4225 | 5000 | 1.3114 | 57.9337 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.6.0+cu118
- Datasets 3.2.0
- Tokenizers 0.21.0