Turn off source splitting

[corneliusgerico]

Hi,

What is the simplest way to do this?

Thanks

[MahmoudAshraf97]

pass --no-stem in the cli arguments

[Oheed911]

can we do it from the python usage?

[MahmoudAshraf97]

can we do it from the python usage?

Yes, remove the following code

whisper-diarization/diarize.py

Lines 37 to 52 in 800d03a

	if args.stemming:
	# Isolate vocals from the rest of the audio
	return_code = os.system(
	f'python3 -m demucs.separate -n htdemucs_ft --two-stems=vocals "{args.audio}" -o "temp_outputs"'
	)
	if return_code != 0:
	print(
	"Source splitting failed, using original audio file. Use --no-stem argument to disable it."
	)
	vocal_target = args.audio
	else:
	vocal_target = f"temp_outputs/htdemucs_ft/{args.audio[:-4]}/vocals.wav"
	else:
	vocal_target = args.audio

and replace it with
vocal_target = args.audio or whatever your file name argument is