Why is self.feature_extractor.nb_max_frames subtracted from total frames

[stri8ed]

I'm trying to understand the logic behind this subtraction:

faster-whisper/faster_whisper/transcribe.py

Line 469 in 1eb9a80

content_frames = features.shape[-1] - self.feature_extractor.nb_max_frames

It looks like it's discarding the end of the audio, and not transcribing it. Why?

[trungkienbkhn]

Same as this logic from original whisper project. You can refer to this commit.