search

lucidrains / audiolm-pytorch

Implementation of AudioLM, a SOTA Language Modeling Approach to Audio Generation out of Google Research, in Pytorch
MIT License
2.36k stars 255 forks source link

Pytorch 2? #130

Closed mrbusysky closed 1 year ago

mrbusysky commented 1 year ago

Was wondering if there was any plan to version bump to take advantage of the new speedups.

lucidrains commented 1 year ago

@mrbusysky it should work, with the exception of this complex valued stft discriminator in soundstream (sigh)

lucidrains commented 1 year ago

@mrbusysky have you tried it and seen any errors?

mrbusysky commented 1 year ago

Seems to be fine, i needed to change a few things to make it work better with my ide though.