metadata
library_name: transformers
language:
- ar
license: apache-2.0
base_model: openai/whisper-base
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper base AR - BA
results: []
Whisper base AR - BA
This model is a fine-tuned version of openai/whisper-base on the quran-ayat-speech-to-text dataset. It achieves the following results on the evaluation set:
- Loss: 0.1315
- Wer: 0.3034
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: 0.0001
- 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: 500
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1243.1483 | 1.0 | 157 | 0.2708 | 1.0528 |
| 61.6092 | 2.0 | 314 | 0.1723 | 0.6490 |
| 32.3819 | 3.0 | 471 | 0.1496 | 0.4943 |
| 14.9798 | 4.0 | 628 | 0.1389 | 0.3908 |
| 10.9898 | 5.0 | 785 | 0.1385 | 0.3569 |
| 6.9766 | 6.0 | 942 | 0.1362 | 0.3239 |
| 5.8893 | 7.0 | 1099 | 0.1359 | 0.3176 |
| 4.4691 | 8.0 | 1256 | 0.1313 | 0.3050 |
| 3.9368 | 9.0 | 1413 | 0.1310 | 0.3077 |
| 3.4552 | 10.0 | 1570 | 0.1282 | 0.2896 |
| 3.2193 | 11.0 | 1727 | 0.1281 | 0.2922 |
| 2.991 | 12.0 | 1884 | 0.1267 | 0.2829 |
| 2.7838 | 13.0 | 2041 | 0.1268 | 0.2955 |
| 2.7073 | 14.0 | 2198 | 0.1269 | 0.2806 |
| 2.5728 | 14.9088 | 2340 | 0.1266 | 0.2875 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1