openai/whisper-medium
This model is a fine-tuned version of openai/whisper-medium on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2775
- Wer: 9.7168
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: 32
- eval_batch_size: 32
- 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 |
|---|---|---|---|---|
| 0.2572 | 0.12 | 500 | 0.2180 | 10.4499 |
| 0.1542 | 1.08 | 1000 | 0.1995 | 10.2143 |
| 0.062 | 2.04 | 1500 | 0.2177 | 10.2560 |
| 0.0761 | 2.16 | 2000 | 0.2013 | 11.0693 |
| 0.0193 | 3.12 | 2500 | 0.2318 | 10.0235 |
| 0.0065 | 4.07 | 3000 | 0.2591 | 9.9646 |
| 0.0123 | 5.03 | 3500 | 0.2760 | 9.7453 |
| 0.0039 | 5.15 | 4000 | 0.2775 | 9.7168 |
Framework versions
- Transformers 4.27.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.9.1.dev0
- Tokenizers 0.13.2