Whisper Small Fr - Joss
This model is a fine-tuned version of openai/whisper-small on the Common Voice 11.0 FR dataset. It achieves the following results on the evaluation set:
- Loss: 0.4212
- Wer: 24.0365
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.3803 | 0.99 | 1000 | 0.3992 | 23.9465 |
| 0.2214 | 1.99 | 2000 | 0.3902 | 22.8108 |
| 0.0986 | 2.98 | 3000 | 0.4028 | 22.4459 |
| 0.0478 | 3.98 | 4000 | 0.4212 | 24.0365 |
Framework versions
- Transformers 4.25.0.dev0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.2
- Downloads last month
- 3
Evaluation results
- Wer on Common Voice 11.0 FRself-reported24.037