Whisper Small N - Final
This model is a fine-tuned version of openai/whisper-small on the N Demo Final dataset. It achieves the following results on the evaluation set:
- Loss: 0.4388
- Wer: 54.2913
- Cer: 20.4380
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: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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: cosine
- lr_scheduler_warmup_steps: 0.1
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 1.0732 | 0.9965 | 285 | 0.5126 | 66.0236 | 22.9883 |
| 0.7468 | 1.9930 | 570 | 0.4514 | 59.8819 | 22.2823 |
| 0.4662 | 2.9895 | 855 | 0.4296 | 54.1732 | 20.3588 |
| 0.3399 | 3.9860 | 1140 | 0.4388 | 54.2913 | 20.4380 |
Framework versions
- Transformers 5.2.0
- Pytorch 2.9.0+cu128
- Datasets 4.5.0
- Tokenizers 0.22.2