SUPIR Full Tutorial + 1 Click 12GB VRAM Windows & RunPod / Linux Installer + Batch Upscale + Comparison With Magnific - SUPIR Starts A New Era

[FurkanGozukara]

I have been working for several days, over 12 hours each day on SUPIR (Scaling-UP Image Restoration), a state-of-the-art image enhancing and upscaling model presented in the paper "Scaling Up to Excellence: Practicing Model Scaling for Photo-Realistic Image Restoration In the Wild.".

This model is just mind blowing. At the bottom of this post you will see 1 by 1 comparisons of SUPIR vs extremely expensive online service Magnific AI. Magnific is known to be best among community. However, SUPIR is by far better. SUPIR is also by far better than Topaz AI upscale as well. SUPIR is able to stay loyal to original image almost 100% meanwhile adding details and super upscaling with best realism.

I made a full 33 minute tutorial full chaptered with manually written captions. The chapters info posted at the very bottom.

You can watch video from here : https://youtu.be/PqREA6-bC3w

SUPIR: New SOTA Open Source Image Upscaler & Enhancer Model Better Than Magnific & Topaz AI Tutorial

Original repo of the SUPIR : https://github.com/Fanghua-Yu/SUPIR

I have worked hard to make a 1 click installer for Windows & RunPod. RunPod uses Linux, thus if you are a Linux user you can use RunPod files to install locally on Linux as well with 1 click install.

Full instructions are shared in this post along with the scripts : https://www.patreon.com/posts/supir-1-click-99176057

Here below installer files

The installer works with Python 3.10.11. It generates a new pip venv and install everything there. So you don't need Conda. Since it will generate its own VENV, it will not affect any other of your installations on your system.

The installer installs xFormers and Triton (we are using pre-compiled wheel) and Pytorch 2.2.0 automatically for you on Windows and Linux.

Gradio app launching interface like below

Currently with newest optimizations the SUPIR app works great on RTX 3060 without LLaVA. I have tested on my 12 GB single RTX 3060 GPU. So if you have a GPU that has 12GB or more VRAM you can use.

The installer download all models automatically as well. Also I changed base SDXL model with Juggernaut XL - V9 since it works better.

You can simply use any SDXL model. Instructions are on Patreon post.

I also greatly improved the base Gradio APP. Made the interface more useable.

Added number of images and randomized seed feature. Made the image upscale scaler 0.1 precision.

Moreover, I have added batch upscale feature as well.

You can see the improved advanced Gradio app interface below.

All the images the app generates will be automatically saved under outputs folder. You can define batch image processing outputs folder as well.

Here below the content of the Patreon post

SUPIR vs MAGNIFIC AI

Carefully look at the how much SUPIR can be loyal to the original image vs Magnific can be loyal to original image