whisper-tiny-indov2
This model is a fine-tuned version of openai/whisper-tiny on the generator dataset. It achieves the following results on the evaluation set:
- Loss: 0.3805
- Wer: 30.6093
- Cer: 14.6600
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: 16
- seed: 42
- 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: 500
- training_steps: 5180
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.5573 | 1.0998 | 1036 | 0.4492 | 33.2732 | 13.9096 |
| 0.2444 | 3.0994 | 2072 | 0.3950 | 31.1077 | 13.3610 |
| 0.1119 | 5.0990 | 3108 | 0.3854 | 29.8101 | 13.8502 |
| 0.0641 | 7.0986 | 4144 | 0.3863 | 28.9336 | 12.5746 |
| 0.0456 | 9.0983 | 5180 | 0.3805 | 30.6093 | 14.6600 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 4.8.5
- Tokenizers 0.22.2