search

city96 / ComfyUI_ExtraModels

Support for miscellaneous image models. Currently supports: DiT, PixArt, HunYuanDiT, MiaoBi, and a few VAEs.
Apache License 2.0
393 stars 34 forks source link

[Feature Request] Support for Lumina #56

Open GavChap opened 6 months ago

GavChap commented 6 months ago

https://github.com/Alpha-VLLM/Lumina-T2X

This looks like a promising variation on Text to Anything. It'd be nice to get support for it as at the moment it's just gradio demos or python code.

city96 commented 6 months ago

I've had a play around with Lumia as well as Large-DiT in the past, but they were pretty undertrained. Looks like it got updated recently and no longer uses llama as the text model, making it actually possible to run lol.

Might take a shot at it but not a priority at the moment.

festivus37 commented 5 months ago

So it looks like they just released a new version of Lumina, Lumina-Next-SFT which is another step better. I wasn't majorly impressed with the previous Lumina, but this one is actually rather impressive. One of their demos sites: http://106.14.2.150:10020/ image image

festivus37 commented 5 months ago

ComfyUI_0217 (1) Just wanted to poke a little... lumina-next-sft is seriously impressive (this is sdxl refined), but even without refinement it's looking better than sd3-medium raw in a lot of ways.

noyart commented 5 months ago

Lumina-Next-SFT looks amazing and would be awesome if comfyui would get support for it

LankyPoet commented 5 months ago

Agree - please add support for this. It has so much promise.

pandayummy commented 4 months ago

please add support for this.