Skip to content

diontimmer/sample-diffusion-gui

Repository files navigation

sample-diffusion-gui/Vextra Diffusion Toolkit

GUI toolkit for various audio diffusion operations. Run/Install using start_windows.bat or start_unix.sh.
This will create a venv and install the requirements.
This is mainly tested on Windows. OSX, Linux and WSL SHOULD work but there could be uncaught errors.
WSL/Linux users might need to sudo apt install nvidia-cudnn

GIT & FFMPEG is required.

Linux users can install ffmpeg by updating packages and running sudo apt install ffmpeg
OSX users can install by using homebrew: brew install ffmpeg

for windows: start_windows.bat will automatically check for ffmpeg and download icedterminals MSI package.
https://github.com/icedterminal/ffmpeg-installer.

Requires Python 3.10

Includes additional features such as:

  • Model merging
  • Instant input load for variation
  • Batch looping
  • Local model training
  • Input Wave Generation
  • Model Trimming

Features are made accessible through an extension API; look in the extensions folder for examples!

Note:

  • Local Model Training is not recommended for systems with less than 8GB of VRAM.
  • Please import your models through the included model importer. This ensures the models are trimmed and have the correct format for the filenames for use in auto-complete.
  • launch_script.py can be ran directly from your python interpreter if you do not want to use a venv.

alt text

Citation

Various libs and snippets taken from:
https://github.com/Harmonai-org/sample-generator
https://github.com/sudosilico/sample-diffusion

Special Thanks to:
https://github.com/twobob
https://github.com/zqevans
https://github.com/drscotthawley

About

GUI toolkit using various audio diffusion repos.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published