Whisper Base Ar Tashkeel - AzeemX
This model is a fine-tuned version of tarteel-ai/whisper-base-ar-quran on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0762
- Wer: 11.1681
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: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 7000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1686 | 1.0320 | 1000 | 0.1651 | 26.5255 |
| 0.0828 | 2.0640 | 2000 | 0.1216 | 20.2028 |
| 0.0459 | 3.0960 | 3000 | 0.1020 | 16.3712 |
| 0.0237 | 4.1280 | 4000 | 0.0898 | 14.2751 |
| 0.0142 | 5.1600 | 5000 | 0.0808 | 12.4992 |
| 0.009 | 6.1920 | 6000 | 0.0772 | 11.4605 |
| 0.0054 | 7.2239 | 7000 | 0.0762 | 11.1681 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.3.1
- Datasets 3.0.1
- Tokenizers 0.20.0