ComfyUI automatic Installers for Windows, RunPod, Massed Compute, Linux (Patreon)

Patreon exclusive posts index

Join discord to get help, chat, discuss and also tell me your discord username to get your special rank : SECourses Discord

• It is best to install ComfyUI into a Python venv so that it will not break your system

• These installers will generate a venv for ComfyUI and install into that

• None other application on your system will be affected

• Download Comfy_UI zip file from attachments

• I will try to make it work on a free Kaggle account as well hopefully

• Use Massed Compute installer for local Ubuntu / Linux machine

9 June 2024 Update

• IP_Adapter_Downloader .py file added to the scripts

• This file will download all IP Adapter models into the accurate ComfyUI folders with accurate naming

• Open a cmd in the folder where ComfyUI installed and run IP_Adapter_Downloader .py as below

• Python IP_Adapter_Downloader.py

• 

31 May 2024 Update

• Installers updated to install ONNX runtime GPU so that insightface related libraries will work faster

• For this to work, your system must have CUDA 11.8 installed

• Massed Compute already have that

• On RunPod select RunPod Pytorch 2.1 , runpod/pytorch:2.1.0-py3.10-cuda11.8.0-devel-ubuntu22.04 template

Requirements

• For Windows you need to install Python 3.10.11, Git, C++ tools, CUDA 11.8 and FFmpeg

• Watch this tutorial to install them (step by step) https://youtu.be/-NjNy7afOQ0

• For RunPod and Massed Compute just follow the instructions - copy paste nothing else

Features

• Always installs latest version of ComfyUI

• Automatically installs ComfyUI manager latest version

• Automatically installs latest version of ComfyUI_IPAdapter_plus

• Works perfectly on Windows, Massed Compute and RunPod

• Massed Compute has ngrok feature to use it on your computer

• RunPod works with RunPod proxy feature perfectly

• Automatically downloads SDXL base with fixed VAE, RealVis XL V4, Realistic Vision V6 and Hyper Realism V3 models

• Installs latest Torch 2.3.0 and xFormers 0.0.26.post1

• Installs insightface library - you need to have C++ tools