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: 0.0894
- Wer: 0.2482
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: 4
- total_train_batch_size: 32
- 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 |
|---|---|---|---|---|
| 2.8029 | 1.0 | 540 | 0.0903 | 0.2405 |
| 2.2364 | 2.0 | 1080 | 0.0882 | 0.2167 |
| 1.9642 | 3.0 | 1620 | 0.0849 | 0.2300 |
| 1.5845 | 4.0 | 2160 | 0.0837 | 0.2223 |
| 1.3608 | 5.0 | 2700 | 0.0831 | 0.2430 |
| 1.1521 | 6.0 | 3240 | 0.0804 | 0.2405 |
| 1.0722 | 7.0 | 3780 | 0.0838 | 0.2438 |
| 0.9285 | 8.0 | 4320 | 0.0813 | 0.2468 |
| 0.8415 | 9.0 | 4860 | 0.0812 | 0.2333 |
| 0.8056 | 10.0 | 5400 | 0.0793 | 0.2243 |
| 0.7192 | 11.0 | 5940 | 0.0799 | 0.2142 |
| 0.6929 | 12.0 | 6480 | 0.0803 | 0.2208 |
| 0.5994 | 13.0 | 7020 | 0.0806 | 0.2421 |
| 0.5793 | 14.0 | 7560 | 0.0812 | 0.2277 |
| 0.551 | 15.0 | 8100 | 0.0810 | 0.2265 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1