metadata
library_name: transformers
language:
- ar
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper tiny AR - BH
results: []
Whisper tiny AR - BH
This model is a fine-tuned version of openai/whisper-tiny on the quran-ayat-speech-to-text dataset. It achieves the following results on the evaluation set:
- Loss: 0.0191
- Wer: 0.1462
- Cer: 0.0494
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: 5e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Use 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: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.0148 | 1.0 | 250 | 0.0146 | 0.1685 | 0.0520 |
| 0.0109 | 2.0 | 500 | 0.0144 | 0.1765 | 0.0595 |
| 0.0082 | 3.0 | 750 | 0.0152 | 0.1634 | 0.0542 |
| 0.0065 | 4.0 | 1000 | 0.0155 | 0.1593 | 0.0505 |
| 0.0033 | 5.0 | 1250 | 0.0181 | 0.1644 | 0.0571 |
| 0.0025 | 6.0 | 1500 | 0.0182 | 0.1558 | 0.0533 |
| 0.0024 | 7.0 | 1750 | 0.0180 | 0.1544 | 0.0492 |
| 0.0014 | 8.0 | 2000 | 0.0190 | 0.1461 | 0.0508 |
| 0.001 | 9.0 | 2250 | 0.0196 | 0.1430 | 0.0479 |
| 0.0006 | 10.0 | 2500 | 0.0198 | 0.1467 | 0.0497 |
| 0.0005 | 11.0 | 2750 | 0.0198 | 0.1531 | 0.0503 |
| 0.0004 | 12.0 | 3000 | 0.0197 | 0.1446 | 0.0473 |
Framework versions
- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0