IoanaLivia/real-voices-youtube-horoscope-synth-Google-Standard-A-whisper-small
This model is a fine-tuned version of openai/whisper-small on the IoanaLivia/real-voices-youtube-horoscope-synth-Google-Standard-A dataset. It achieves the following results on the evaluation set:
- Loss: 0.3638
- Wer: 19.6837
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: 4
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| No log | 0 | 0 | 0.4170 | 30.1014 |
| 0.1811 | 1.0 | 25 | 0.3390 | 20.6120 |
| 0.0653 | 2.0 | 50 | 0.3281 | 20.3885 |
| 0.0319 | 3.0 | 75 | 0.3355 | 19.9931 |
| 0.0166 | 4.0 | 100 | 0.3438 | 19.7524 |
| 0.0096 | 5.0 | 125 | 0.3521 | 19.8728 |
| 0.0066 | 6.0 | 150 | 0.3584 | 20.7323 |
| 0.0049 | 7.0 | 175 | 0.3638 | 19.6837 |
| 0.004 | 8.0 | 200 | 0.3686 | 19.7868 |
| 0.0034 | 9.0 | 225 | 0.3706 | 19.9416 |
| 0.003 | 10.0 | 250 | 0.3732 | 20.0103 |
| 0.0028 | 11.0 | 275 | 0.3754 | 19.9759 |
| 0.0025 | 12.0 | 300 | 0.3768 | 19.9759 |
| 0.0024 | 13.0 | 325 | 0.3780 | 19.9587 |
| 0.0024 | 14.0 | 350 | 0.3785 | 19.9759 |
| 0.0024 | 14.4082 | 360 | 0.3785 | 19.9416 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1