sony/ai-research-code

[X-UMX] Bad performance when using --targets

guillemcortes opened this issue · 1 comments

Hi,

I am doing some tests on the Google Collab you provide and I've seen that the performance varies a lot if I use the flag --targets vocals respect to if I just run the default test command:
!python test.py --inputs $filename --out-dir results --model models/x-umx.h5

Why is this happening? My goal is to have audio + accompaniment

Many thanks in advance,

Guillem

Sorry, now I see that when using the --targets flag the following files are assumed to be present:
mymodel/separator.json
mymodel/vocals.pth
mymodel/vocals.json