whisper-small-E30_Ypause_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: 0.1850
- Cer: 5.2632
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 |
|---|---|---|---|---|
| 0.2839 | 0.1289 | 100 | 0.2516 | 6.3499 |
| 0.1451 | 0.2579 | 200 | 0.2303 | 6.3440 |
| 0.167 | 0.3868 | 300 | 0.2301 | 6.3205 |
| 0.1308 | 0.5158 | 400 | 0.2191 | 5.9504 |
| 0.1233 | 0.6447 | 500 | 0.2171 | 5.8799 |
| 0.1086 | 0.7737 | 600 | 0.1987 | 5.3454 |
| 0.1031 | 0.9026 | 700 | 0.1966 | 5.3219 |
| 0.0557 | 1.0309 | 800 | 0.1917 | 5.1809 |
| 0.0461 | 1.1599 | 900 | 0.1952 | 5.4276 |
| 0.039 | 1.2888 | 1000 | 0.1918 | 5.1574 |
| 0.0398 | 1.4178 | 1100 | 0.1919 | 5.4629 |
| 0.0414 | 1.5467 | 1200 | 0.1892 | 5.3102 |
| 0.039 | 1.6757 | 1300 | 0.1866 | 5.2690 |
| 0.0414 | 1.8046 | 1400 | 0.1864 | 5.2455 |
| 0.0374 | 1.9336 | 1500 | 0.1850 | 5.2632 |
Framework versions
- Transformers 4.47.0.dev0
- Pytorch 2.5.0+cu121
- Datasets 3.1.0
- Tokenizers 0.20.3