How to use Lumina 2.0 and Wan 2.1? #257
-
|
I saw support for Lumina 2.0 and WAN 2.1 in the README, but after browsing the wiki, code, and issue list, I’m still confused about how to use these two models in Ruined Fooocus. I tried downloading the models to the checkpoints and running inference, but the results were very broken, clearly indicating that something wasn’t configured properly. So let me ask this dumb question directly: How can I use Lumina 2.0 and WAN 2.1 in Ruined Fooocus? Thanks! |
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 1 reply
-
|
Lumina should work. But you might have to make a new Performance setting for it. I use: Custom Steps: 15 (probably need more to get good images) WAN 2.1 is explained (badly) at https://github.com/runew0lf/RuinedFooocus/wiki/WAN-&-Hunyuan-Video |
Beta Was this translation helpful? Give feedback.
-
|
Thank you! This setting works, but only for 1024*1024, I'm not sure why but that's enough |
Beta Was this translation helpful? Give feedback.
Lumina should work. But you might have to make a new Performance setting for it. I use:
Custom Steps: 15 (probably need more to get good images)
CFG: 5
Sampler: dpmpp_2m
Scheduler: sgm_uniform
Clip Skip: 1
WAN 2.1 is explained (badly) at https://github.com/runew0lf/RuinedFooocus/wiki/WAN-&-Hunyuan-Video
My gpu isn't fast enough to experiment properly with video so the wiki is a bit lacking.