Whisper Small for Quran Recognition
This model is a fine-tuned version of openai/whisper-small on the Quran_Reciters dataset. It achieves the following results on the evaluation set:
- Loss: 0.0210
- Wer: 3.2835
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: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 12000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0073 | 1.62 | 500 | 0.0249 | 5.0026 |
| 0.0014 | 3.24 | 1000 | 0.0214 | 4.1086 |
| 0.0008 | 4.85 | 1500 | 0.0221 | 3.9883 |
| 0.0 | 6.47 | 2000 | 0.0180 | 2.9740 |
| 0.0 | 8.09 | 2500 | 0.0177 | 3.0944 |
| 0.0 | 9.71 | 3000 | 0.0178 | 3.0944 |
| 0.0 | 11.33 | 3500 | 0.0179 | 3.1288 |
| 0.0 | 12.94 | 4000 | 0.0179 | 3.1288 |
| 0.0 | 14.56 | 4500 | 0.0181 | 2.8881 |
| 0.0 | 16.18 | 5000 | 0.0184 | 2.9225 |
| 0.0 | 17.8 | 5500 | 0.0186 | 3.0256 |
| 0.0 | 19.42 | 6000 | 0.0188 | 3.1803 |
| 0.0 | 21.04 | 6500 | 0.0190 | 3.1631 |
| 0.0 | 22.65 | 7000 | 0.0191 | 3.1631 |
| 0.0 | 24.27 | 7500 | 0.0192 | 3.1803 |
| 0.0 | 25.89 | 8000 | 0.0192 | 3.1631 |
| 0.0 | 27.51 | 8500 | 0.0196 | 3.2491 |
| 0.0 | 29.13 | 9000 | 0.0199 | 3.2491 |
| 0.0 | 30.74 | 9500 | 0.0202 | 3.2835 |
| 0.0 | 32.36 | 10000 | 0.0204 | 3.2319 |
| 0.0 | 33.98 | 10500 | 0.0207 | 3.2835 |
| 0.0 | 35.6 | 11000 | 0.0209 | 3.2663 |
| 0.0 | 37.22 | 11500 | 0.0210 | 3.2835 |
| 0.0 | 38.83 | 12000 | 0.0210 | 3.2835 |
Framework versions
- Transformers 4.40.0.dev0
- Pytorch 2.1.2
- Datasets 2.17.1
- Tokenizers 0.15.1