Transcription truncated with dots on medium model

[neilgd]

Hi,

I'm transcribing a 10Mb .mp3 file. The large model works fine, but with the medium.en model the transcription truncates early with .......
(ie a series of dots)

I was wondering how I could troubleshoot / fix this? I suspect it is something to do with long 'quiet' sections where only ambient music plays.

The file is here:

https://www.dropbox.com/s/u3ftu0xmzzf49qp/yanyu.mp3?dl=0

Thank you!