Whisper Medium Sr Fleurs
This model is a fine-tuned version of openai/whisper-medium on combined Google Fleurs and Mozilla Common Volice 13 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1947
- Wer Ortho: 0.1874
- Wer: 0.0788
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: 4
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 1500
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
|---|---|---|---|---|---|
| 0.072 | 1.34 | 500 | 0.1769 | 0.1896 | 0.0912 |
| 0.0223 | 2.67 | 1000 | 0.1774 | 0.1993 | 0.0832 |
| 0.0101 | 4.01 | 1500 | 0.1947 | 0.1874 | 0.0788 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3