search

openai / jukebox

Code for the paper "Jukebox: A Generative Model for Music"
https://openai.com/blog/jukebox/
Other
7.83k stars 1.41k forks source link

Extracting embeddings? #229

Open youssefavx opened 3 years ago

youssefavx commented 3 years ago

Hi, is it possible to get vectors for certain audio segments?

E.g. Given a 1 second wav file, give me an embedding for that?

If so, is there example code somewhere?

youssefavx commented 3 years ago

After looking at the code more closely, I wonder if this solves it: https://github.com/openai/jukebox/blob/08efbbc1d4ed1a3cef96e08a931944c8b4d63bb3/jukebox/prior/autoregressive.py#L177

hadisaadat commented 3 years ago

@youssefavx Can you share by any chance how did you manage to get use of it?

shoegazerstella commented 2 years ago

Try this: https://github.com/p-lambda/jukemir