# ComfyUI-RAVE

**Repository Path**: deplay/ComfyUI-RAVE

## Basic Information

- **Project Name**: ComfyUI-RAVE
- **Description**: No description available
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 0
- **Created**: 2024-03-18
- **Last Updated**: 2024-03-18

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

## ComfyUI-RAVE

[RAVE_00005.webm](https://github.com/spacepxl/ComfyUI-RAVE/assets/143970342/4d1b065d-f114-448b-ad40-ac7b337fbb67)

Unofficial ComfyUI implementation of [RAVE](https://rave-video.github.io/)

Use [BlenderNeko's Unsampler](https://github.com/BlenderNeko/ComfyUI_Noise) for noise inversion. A basic workflow is included, using the cupcake train example from the RAVE paper.

Most of the testing was done with SD1.5, but SDXL does work, although not as well (possibly because the multi-resolution training reduces the tiling effect?)