ndizi-whisper-large-v3-turbo-optimized_1
This model is a fine-tuned version of openai/whisper-large-v3-turbo on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6903
- Wer: 0.3248
- Cer: 0.1195
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: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 4
- 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: 200
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.6851 | 1.0 | 531 | 0.5861 | 0.4269 | 0.1586 |
| 0.4673 | 2.0 | 1062 | 0.5280 | 0.3708 | 0.1349 |
| 0.2927 | 3.0 | 1593 | 0.5199 | 0.3554 | 0.1319 |
| 0.1981 | 4.0 | 2124 | 0.5311 | 0.3418 | 0.1250 |
| 0.1331 | 5.0 | 2655 | 0.5563 | 0.3430 | 0.1251 |
| 0.0878 | 6.0 | 3186 | 0.5840 | 0.3473 | 0.1286 |
| 0.0553 | 7.0 | 3717 | 0.6089 | 0.3387 | 0.1251 |
| 0.0252 | 8.0 | 4248 | 0.6602 | 0.3368 | 0.1220 |
| 0.0108 | 9.0 | 4779 | 0.6788 | 0.3307 | 0.1229 |
| 0.0038 | 10.0 | 5310 | 0.6903 | 0.3248 | 0.1195 |
Framework versions
- Transformers 4.57.3
- Pytorch 2.8.0+cu128
- Datasets 3.6.0
- Tokenizers 0.22.2