search

lllyasviel / stable-diffusion-webui-forge

GNU Affero General Public License v3.0
8.26k stars 808 forks source link

Flux Text encoder loras don't work, results are either blurry or the clip is not loaded #1533

Open DarkViewAI opened 2 months ago

DarkViewAI commented 2 months ago

sample images when training it looks like subject.

when trying in forge it doesn't load. i trained the clip l text encoder along with it. but doesn't work in forge or comfy

DarkViewAI commented 2 months ago

it also doesn't patch the lora, it just ignores it

al-swaiti commented 2 months ago

follow this https://civitai.com/articles/6715 be careful use this vae not diffusers one image