whisper-large-v3-turbo-l21l30
This model is a fine-tuned version of openai/whisper-large-v3-turbo on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9327
- Wer: 39.1887
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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: 2000
- training_steps: 20000
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0 | 0 | 5.7681 | 1174.3994 |
| 0.8929 | 4.0112 | 2000 | 1.1186 | 45.0571 |
| 0.6935 | 9.0002 | 4000 | 0.9811 | 50.2954 |
| 0.7716 | 13.0114 | 6000 | 0.9477 | 44.3482 |
| 0.6844 | 18.0004 | 8000 | 0.9464 | 43.5605 |
| 0.5635 | 22.0116 | 10000 | 0.9372 | 40.1339 |
| 0.7053 | 27.0006 | 12000 | 0.9368 | 42.7727 |
| 0.515 | 31.0118 | 14000 | 0.9355 | 40.0945 |
| 0.5938 | 36.0008 | 16000 | 0.9333 | 39.5825 |
| 0.734 | 40.012 | 18000 | 0.9320 | 39.3068 |
| 0.6472 | 45.001 | 20000 | 0.9327 | 39.1887 |
Framework versions
- Transformers 4.54.0
- Pytorch 2.8.0.dev20250319+cu128
- Datasets 3.6.0
- Tokenizers 0.21.2