Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Training new artists / genres on colab #55

Open
RegattaBurrata opened this issue May 9, 2020 · 5 comments
Open

Training new artists / genres on colab #55

RegattaBurrata opened this issue May 9, 2020 · 5 comments

Comments

@RegattaBurrata
Copy link

Is it possible to train on new artists or genres on colab? For example, if I'd like to create samples in the style of an artist that wasn't in the original dataset and therefore has no artist id tag.

@diffractometer
Copy link

+1

@heewooj
Copy link
Contributor

heewooj commented May 15, 2020

We added more instructions on how to fine-tune to new styles or train from scratch, but they are from command line.

@diffractometer
Copy link

@heewooj thanks for the update. Do you know if there's a simple way to run command line scripts from a colab runtime?

@canokaue
Copy link

I'm also interested on training new layers on Colab, since p100 are kinda off my league outside Colab Pro.

Well, I guess you could do something like:

import os os.popen('command you want)

or just preface commands with ! or %%shell

But, then again, these are just guesses. I'm sure trying them out myself later this week.

ps: you need to mount your drive in order to load new training data. SMarioMan has an open PR solving this: https://github.com/openai/jukebox/pull/72/commits

@diffractometer
Copy link

@heewooj nm, I got it !...

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants