huggingface/diffusers

Add SUPIR Upscaler

Open

#7,219 opened on Mar 5, 2024

View on GitHub
 (36 comments) (14 reactions) (0 assignees)Python (22,190 stars) (4,562 forks)batch import
New pipeline/modelcontributions-welcomehelp wanted

Description

Model/Pipeline/Scheduler description

SUPIR is a super-resolution model that looks like it produces excellent results

Github Repo: https://github.com/Fanghua-Yu/SUPIR

The model is quite memory intensive, so the optimisation features available in diffusers might be quite helpful in making this accessible to lower resource GPUs.

Open source status

  • The model implementation is available.
  • The model weights are available (Only relevant if addition is not a scheduler).

Provide useful links for the implementation

No response

Contributor guide

Tech stack
pythonpytorch
Domain
machine learningai
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
over 1 week
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
clear
Prerequisites
PyTorch knowledgediffusers codebase familiaritydiffusion model basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
25
Research direction
Start by examining existing diffusion pipelines in the `src/diffusers/pipelines` directory, particularly similar upscaling pipelines like `stable diffusion` or `controlnet`. Review the SUPIR model architecture in its GitHub repository to understand the components needed. The issue mentions memory constraints, so consider implementing with support for feature extraction offloading or using `torch.compile` for optimization.
Add SUPIR Upscaler · huggingface/diffusers#7219 | Good First Issue