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: 0.0125
- Wer: 0.1057
- Cer: 0.0420
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-06
- 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: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 0.0116 | 1.0 | 469 | 0.0084 | 0.0929 | 0.0328 |
| 0.0066 | 2.0 | 938 | 0.0075 | 0.0923 | 0.0344 |
| 0.0058 | 3.0 | 1407 | 0.0074 | 0.0914 | 0.0346 |
| 0.0041 | 4.0 | 1876 | 0.0077 | 0.0883 | 0.0336 |
| 0.0043 | 5.0 | 2345 | 0.0082 | 0.0892 | 0.0330 |
| 0.0028 | 6.0 | 2814 | 0.0089 | 0.0890 | 0.0344 |
| 0.0024 | 7.0 | 3283 | 0.0094 | 0.0887 | 0.0329 |
| 0.0011 | 8.0 | 3752 | 0.0099 | 0.0901 | 0.0341 |
| 0.0015 | 9.0 | 4221 | 0.0104 | 0.0874 | 0.0316 |
| 0.0017 | 10.0 | 4690 | 0.0108 | 0.0887 | 0.0336 |
| 0.0005 | 11.0 | 5159 | 0.0111 | 0.0847 | 0.0307 |
| 0.0003 | 12.0 | 5628 | 0.0113 | 0.0885 | 0.0331 |
| 0.0003 | 13.0 | 6097 | 0.0115 | 0.0896 | 0.0333 |
| 0.0005 | 14.0 | 6566 | 0.0131 | 0.1070 | 0.0423 |
| 0.001 | 15.0 | 7035 | 0.0116 | 0.0900 | 0.0333 |
Framework versions
- Transformers 4.47.0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0