Transcript Files Not Being Generated under macOS (Printing to Console Instead)

[Zyniker13]

The title pretty much says it all: Under macOS (Sonoma 14.1.1), Whisper, via command line, is not generating transcript files even when they are specified, but, rather, is printing to the console.

Here is one of the commands I tried:

whisper audio.m4a --model large -f all -o .

[glangford]

(1) Make sure the current directory is writable.
(2) The default output directory is the current directory; my guess is that '.' for the directory is not being expanded and you don't need to specify it anyway. Try

whisper audio.m4a --model large -f all

(3) In addition, the default is also to produce all outputs, so even more straightforward is

whisper audio.m4a --model large

and if you don't want console output

whisper audio.m4a --model large --verbose False

[glangford]

p.s. Sonoma 14.1.1 works fine for me

[Zyniker13]

Going to mark your reply as an answer (in part because I do not know what fixed the issue), but:

After running a brew update, everything seems to be working again.