Skip to content

spsanps/diffwave-sashimi

 
 

Repository files navigation

Experiments with S4/Sashimi models

What is Sashmi: https://arxiv.org/abs/2202.09729

This repository is a fork of : https://github.com/albertfgu/diffwave-sashimi

Each branch is a different experiment

Please view the branches for more information.

  • cold_diffusion2: Using Cold Diffusion to transform midi inputs to audio waveforms using Sashimi
  • freq_diffusion2: Diffusion in the frequency domain using Sashimi
  • no_diffusion: Using Sashmi without diffusion
  • proll_as_mel: Using piano roll for conditional diffusion

Other branches are not in use.

About

Series of experiments with DiffWave and SaShiMi audio generation models

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages

  • Python 80.7%
  • Cuda 11.7%
  • Jupyter Notebook 3.7%
  • C++ 1.8%
  • C 1.5%
  • Shell 0.6%