huggingface / huggingface/diffusers

Generated image is cropped when using smaller output size to reduce VRAM and processing time in Flux Kontext

Open
#12,406 5 comments 0 reactions 0 assignees View on GitHub
stale
Dominant language
Python
Stars
34.5k
Forks
7.3k
Avg merge
3d 3h
Merged PRs (30d)
91

Description

Use `max_area` in FluxKontextPipeline, default `max_area=1024**2`. I set `max_area=768**2`, and output is cropped

Original image:

![Image](https://github.com/user-attachments/assets/28103be5-e4c6-417c-abba-d75ed6e53031)

Output with `max_area=1024**2`:

Image

Output with `max_area=768**2`:

Image

Contributor guide

Open the contributing guide

Research direction

Start with FluxKontextPipeline and reproduce the reported behavior using max_area=1024**2 and max_area=768**2. Compare the generated outputs with the original image and trace how the smaller max_area is applied; done means the reduced-size output is not unexpectedly cropped.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
computer-vision, machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
38/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.