search

lucidrains / audiolm-pytorch

Implementation of AudioLM, a SOTA Language Modeling Approach to Audio Generation out of Google Research, in Pytorch
MIT License
2.32k stars 249 forks source link

pretrained soundstream weights? #227

Open muazhuda opened 10 months ago

muazhuda commented 10 months ago

Are there pretrained weights compatible with this repo's soundstream?

so far I found (quantized) weights here https://github.com/google/lyra/tree/main/lyra/model_coeffs but not sure how to transfer these weights to this pytorch implementation.

limchr commented 6 months ago

I would be also interested in pretrained weights