ap-8joXukl3mo8gfg5IDAwvaL
This model is a fine-tuned version of openai/whisper-tiny on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4830
- Model Preparation Time: 0.0025
- Wer: 0.1725
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: 0.0003
- 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: 8
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Model Preparation Time | Wer |
|---|---|---|---|---|---|
| 0.4806 | 1.0 | 250 | 0.4410 | 0.0025 | 0.2914 |
| 0.3596 | 2.0 | 500 | 0.5134 | 0.0025 | 0.2726 |
| 0.2834 | 3.0 | 750 | 0.4729 | 0.0025 | 0.2455 |
| 0.1924 | 4.0 | 1000 | 0.4203 | 0.0025 | 0.2076 |
| 0.1247 | 5.0 | 1250 | 0.4385 | 0.0025 | 0.1961 |
| 0.0797 | 6.0 | 1500 | 0.4320 | 0.0025 | 0.1910 |
| 0.0417 | 7.0 | 1750 | 0.4428 | 0.0025 | 0.1769 |
| 0.0167 | 7.9685 | 1992 | 0.4830 | 0.0025 | 0.1725 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.2.0
- Tokenizers 0.21.2