Fail to process a BBC podcast mp3 with a dead loop

[jhpzju]

Latest whisper update from Python PIP install;
Used file downloaded (128lbps version, Windows + Chrome) from this link https://www.bbc.co.uk/programmes/b07gh4pg
Used command: whisper audio.mp3 --fp16 False --language English --model large
The processing got some sort of dead loop around 36m (see screenshot)

It worked for other podcast files with similar topics and length.