From 1b77b594f3e1ddc8af0df8939f4e6cae65155938 Mon Sep 17 00:00:00 2001 From: File_xor Date: Tue, 30 Jun 2026 02:48:55 +0900 Subject: [PATCH] Add more description to README.md --- README.md | 23 +++++++++++++++++++++-- 1 file changed, 21 insertions(+), 2 deletions(-) diff --git a/README.md b/README.md index 93d1aed..8f4dbbe 100644 --- a/README.md +++ b/README.md @@ -1,8 +1,27 @@ # SEGA Anima -SEGA anima is implementation of ["SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers" (arXiv:2605.22668)](https://arxiv.org/abs/2605.22668) for Anima model in Comfyui. +SEGA anima is implementation of ["SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers" (arXiv:2605.22668)](https://arxiv.org/abs/2605.22668) for Anima model in ComfyUI. -## What should I do when output image is bad +## Installation + +### Using git + +From root folder of ComfyUI, goto `./ComfyUI/custom_nodes` and open terminal or command prompt and type following: +`git clone https://github.com/Filexor/SEGA_Anima.git` + +### Via downloading zip + +Click green "\<\> Code ▾" button and then click "Download ZIP" and extract folder to `./ComfyUI/custom_nodes` . + +### Via ComfyUI Manager + +This method is currently unavailable. + +## Usage + +Insert "SEGA Anima" node to between "Load Diffusion Model" node and "KSampler" node. + +### What should I do when output image is bad 1. If image does not appear, first thing you have to do is decreasing "training_resolution". Too low "training_resolution" results structual inaccuracy. 2. Adjust CFG. Normally, higher CFG is needed for high resolution output.