ComfyUI Studio patreon

How to Train a Flux LoRA Locally in 15 Minutes! Create an AI Influencer with ComfyUI Workflows.

🕑 Added 2024-12-20 19:16:30 +0000 UTC

Comments

Joel

I also got the same error when installing. I originally had done the manual install before trying this, but when you hit to train your LoRA, the model starts downloading and errors out for me after about half the download is done. I've been trying to fix this for a few days now.

Jeff Merilo

this helped! Thanks for the link bruno!

DigitAll

My workflows are dependent with studio

Joe C

i have a question. if I sign up am I able to get the workflows without downloading and installing comfy UI studio? I'm using comfy UI desktop and I like using that for now.

bruno davi

It works, I did the manual installation step by step and it works. follow what is written I'll link you a video where it explains. https://www.youtube.com/watch?v=nY5qsifYr_8&t=638s

bruno davi

in the insto trying with manual installation, because then in the interface it has the possibility to work with 20/16/12 vram, so it must work. https://github.com/cocktailpeanut/fluxgym as soon as it's ready I'll let you know if it works.

Landrodis

Did you figure this out? I'm having the same problem.

bruno davi

When I press to make captions

bruno davi

Hi, I was trying Fluxgym but I have this error: Unable to create tensor, you should probably activate padding with 'padding=True' to have batched tensors with the same length. I'm following your video.

DigitAll

yes for face use a closeup photo. only face no body.

Eusthace

Hey everyone, great to be part of this community! I'm new here and trying out Character v3 for the first time. I’ve noticed that the generated character looks quite different from the real person in the photo. Is there a specific parameter I can adjust to make it more accurate?