ap-u2ve4vNEqG8dHywtw4QDVP
This model is a fine-tuned version of openai/whisper-large-v3-turbo on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3547
- Model Preparation Time: 0.0073
- Wer: 0.1169
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: 3e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.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: 400
- num_epochs: 16
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer |
|---|---|---|---|---|---|
| 0.4516 | 1.0 | 250 | 0.2822 | 0.0073 | 0.1329 |
| 0.1981 | 2.0 | 500 | 0.2433 | 0.0073 | 0.1390 |
| 0.1437 | 3.0 | 750 | 0.2299 | 0.0073 | 0.1183 |
| 0.1073 | 4.0 | 1000 | 0.2392 | 0.0073 | 0.1261 |
| 0.0825 | 5.0 | 1250 | 0.2526 | 0.0073 | 0.1339 |
| 0.063 | 6.0 | 1500 | 0.2545 | 0.0073 | 0.1262 |
| 0.0496 | 7.0 | 1750 | 0.2777 | 0.0073 | 0.1244 |
| 0.0375 | 8.0 | 2000 | 0.2883 | 0.0073 | 0.1308 |
| 0.0289 | 9.0 | 2250 | 0.2757 | 0.0073 | 0.1129 |
| 0.0216 | 10.0 | 2500 | 0.3027 | 0.0073 | 0.1218 |
| 0.0153 | 11.0 | 2750 | 0.3247 | 0.0073 | 0.1199 |
| 0.0117 | 12.0 | 3000 | 0.3193 | 0.0073 | 0.1147 |
| 0.0059 | 13.0 | 3250 | 0.3347 | 0.0073 | 0.1158 |
| 0.0023 | 14.0 | 3500 | 0.3520 | 0.0073 | 0.1178 |
| 0.0011 | 15.0 | 3750 | 0.3593 | 0.0073 | 0.1160 |
| 0.0007 | 15.9365 | 3984 | 0.3547 | 0.0073 | 0.1169 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.2