Optimize data loading, preprocessing, and metric computation

huggingface / distil-whisper

Distilled variant of Whisper for speech recognition. 6x faster, 50% smaller, within 1% word error rate.

MIT License

3.32k stars 238 forks source link

Optimize data loading, preprocessing, and metric computation #120

Open arjun-234 opened 2 months ago

arjun-234 commented 2 months ago

Refactored code to improve efficiency and readability in the data loading, preprocessing, and metric computation functions. Optimizations include using list comprehensions, dictionary unpacking, and consolidating redundant operations.