Comfy-Org / Comfy-Org/comfy-kitchen

MiniMax ref2va Noise output with comfy-kitchen attention

Open
#140 10 comments 0 reactions 0 assignees View on GitHub
Dominant language
Python
Stars
220
Forks
91
Avg merge
1d 7h
Merged PRs (30d)
12

Description

Hi.
Im on Rocm 7.2 with Amd Radeon AI PRO r9700.
When i try to use comfy-kitchen attention i get just pure noise video when i push resolution to 0.9 megapixels and video duration of 15seconds.
But if i put 5seconds instead the video works fine. Something its causing it to output pure noise above certeing point of duration.
If i use normal pytorch attention the video works fine at 15seconds again.

Im using the v2v workflow for editing video.
Also Im using the fp8_scaled pruned model.
I don't get any log in the console af any kind of error, it looks like a normal generation.

Regards

Contributor guide

Open the contributing guide

Research direction

Start with the v2v workflow described in the issue and reproduce generation at 0.9 megapixels for 5 and 15 seconds on ROCm 7.2 with the fp8_scaled pruned model. Compare comfy-kitchen attention with normal PyTorch attention; done means the 15-second output is no longer pure noise or the failure has been isolated to a documented condition.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning, performance
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Active
Clarity
Mostly clear
Newbie friendliness
52/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.