whisper-small-Yfreq_pause_speed
This model is a fine-tuned version of openai/whisper-small on the aihub old adult freq speed pause changed dataset. It achieves the following results on the evaluation set:
- Loss: 0.2878
- Cer: 7.5012
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch 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: 50
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 0.2807 | 0.1289 | 100 | 0.2981 | 8.1767 |
| 0.1583 | 0.2579 | 200 | 0.2730 | 7.4777 |
| 0.1217 | 0.3868 | 300 | 0.2803 | 7.9182 |
| 0.1291 | 0.5158 | 400 | 0.2744 | 7.7890 |
| 0.1021 | 0.6447 | 500 | 0.2840 | 8.0416 |
| 0.0941 | 0.7737 | 600 | 0.2933 | 8.1356 |
| 0.1047 | 0.9026 | 700 | 0.2888 | 7.8066 |
| 0.0386 | 1.0309 | 800 | 0.2798 | 7.4013 |
| 0.0268 | 1.1599 | 900 | 0.2794 | 7.2545 |
| 0.0349 | 1.2888 | 1000 | 0.2858 | 7.2780 |
| 0.0292 | 1.4178 | 1100 | 0.2873 | 7.4072 |
| 0.0373 | 1.5467 | 1200 | 0.2876 | 7.4248 |
| 0.0276 | 1.6757 | 1300 | 0.2857 | 7.4542 |
| 0.0287 | 1.8046 | 1400 | 0.2901 | 7.6069 |
| 0.0295 | 1.9336 | 1500 | 0.2878 | 7.5012 |
Framework versions
- Transformers 4.47.0.dev0
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3