i made a clean install of KOhya and now i dont get error when LORA extract. Git pulling oldd version
i made a clean install of KOhya and now i dont get error when LORA extract. Git pulling oldd version didnt help. Only clean reinstall for ome reason
Fnice at least solved
FWas that the config in the Patreon post that I linked?
NModel called epiCPhotoGasm last version is pretty good
Bshould all stage1 references be remove from the GUI?
FWe may remove not useful
FPatreon
Patreon is empowering a new generation of creators.
Support and engage with artists and creators as they live out their passions!
Support and engage with artists and creators as they live out their passions!
FNice can be tested
A@Dr. Furkan Gözükara is there any way to rollback to supir v14 ?
Aits taking twice as long for me on the latest version and worse results
Rhi @Dr. Furkan Gözükara what's this gradio error?
Ralso no module called requests.
Ri installed your v19 and threw in the models
BI don't have v1.4 installed, but v1.7 quality is better than v1.9
Fthere shouldnt be this stuff
Fare you aware not it default selects Linear CFG?
Fi can turn it off
Flooks like same folder name repeat error
Finstall into different folde rname
Fcan you test turning off Linear CFG and compare?
Fthat could be reason
Fcan you test turning off Linear CFG and compare?
that could be reason
that could be reason
Fplease let me know i will change default according to that
Fcan you provide me your tested image and prompt? i can compare locally and see
Fwe have to fix if there is a quality drop
K@Dr. Furkan Gözükara Sora is using Diffusion transformers, but the use of those is nothing new. The quality of the output is something else though.
Pixart-alpha is also using diffusion transformers. Have you looked anything into this? https://github.com/PixArt-alpha/PixArt-alpha
Pixart-alpha is also using diffusion transformers. Have you looked anything into this? https://github.com/PixArt-alpha/PixArt-alpha
GitHub
Fast Training of Diffusion Transformer for Photorealistic Text-to-Image Synthesis - PixArt-alpha/PixArt-alpha
Fi know pixart alpha but didnt look carefully yet
Fi am adding dynamically shown target output resolution
F
P@Dr. Furkan Gözükara Dose the new Patreon Supir setup install a fully functional and up to date version of comfy ui?
Can I install it with out effecting the current stable diffusion a1111 install I have?
Will installing it change the current venv and python versions on my pc?
I only ask because I am in the middle of a large project and I would like to test the upscaling of supir but if I mess up my current A111 install it would be catastrophic.
Can I install it with out effecting the current stable diffusion a1111 install I have?
Will installing it change the current venv and python versions on my pc?
I only ask because I am in the middle of a large project and I would like to test the upscaling of supir but if I mess up my current A111 install it would be catastrophic.
F1 : we have stand alone gradio not comfy ui
F2 : yes . install into new folder
F3 : nope everything stays use python 3.10.11
Fupdated to v20
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Dclaude 3 released today
Dgotta love a nuclear "exposition" hehe
P@Dr. Furkan GözükaraThanks, did you make patreon that dose install an updated comfy UI ?
Fwhat you mean
Fby the way we are adding dark theme to supir app
Dyesss
F
Dmuch much better
Dso easier i think you should do all your videos in dark mode lol
PDid you make an install.bat for comfy ui?
Rwhat does this mean? When i reinstall the script because its not working there is already a SUPIR inside SUPIR where venv is.
RYou mean i have to delete the SUPIR inside SUPIR when re-instlaling?
Dnew folder
Fmake parent folder name different
Fi saw this error
Fnope i didnt
Fwe dont have supir for comfy ui yet
