Pad audio instead of mel features to reduce word error rates

[MahmoudAshraf97]

This PR pads audio before feature extraction instead of padding the features, this is inline with how the whisper model was trained because reverting the zero-padded Mel spectrogram features back to time domain results in a white noise with moderate amplitude that causes hallucinations and wrong transcriptions
The distil-large-v3 WER dropped from 26.04 at 83a368e to 14.472
This figure can be reproduced by running benchmarks/yt_commons.py and switching the batched inference to sequential

as a side effect, it seems that the transcription is also faster
it took 27 minutes to evaluate the dataset vs 35 minutes before the fix

References:
openai/whisper#730 (comment)
openai/whisper#838 (comment)