QuranWhisper

This is a project to evaluate the performance of different fine-tuned neural ASR systems.

Evaluation Metrics

1. Model Inference Time
2. Word error rate (WER)
3. Match error rate (MER)
4. Word information Preserved (WIP)
5. Character error rate (CER)

Included Models

1. OpenAI Whisper Large
2. arbml/whisper-largev2-ar
3. tarteel-ai/whisper-base-ar-quran
4. tarteel-ai/whisper-tiny-ar-quran

Environment setup

1. Assuming Conda is installed, create a new conda environment:

conda create -n quranwhisper python=3.10 
conda activate quranwhisper 

2. Install the required packages:

pip install -r requirements.txt 

3. Add the recitation audio files in the root directory of the project, as shown below:

    .
    ├── ...
    ├── recitations                    
    │   ├── aya1.wav          
    │   ├── aya2.wav         
    │   └── ...                
    └── ...