@Dr. Furkan Gözükara yes the same question, which face + clothes extension is the best, ip adapter 2
@Dr. Furkan Gözükara yes the same question, which face + clothes extension is the best, ip adapter 2?
Finstant id best for face
Fclothing idm vton
Fface instant id clothing idm vton
Fwow training of kohya way slower
Fthan before
Fon runpod
Vanything better than idm vton so far?
Ffor clothing transfer best
Fstarted 9 kohya training
F
Fnot clothes change. i mean having a face (just a face) and a dress, and compose image
Fin idm vton you need a full body picture right?
Fwtf
Fmulti gpu training?
Fhow much time it takes to train on 9x?
For 9 different trainings
Fdoes idm vton work with just a face?
Fwithout body*
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Dno
Dwell it swaps
D
Dbut the results are like
Flol
F9 different
Fi do train batch size 1 to find best
R@Dr. Furkan Gözükara what is the current best dreambooth training for kohya? You've done a lot of tests. Is it the 48gb_textencoder.json in your https://www.patreon.com/posts/very-best-for-of-89213064?
Fnot updated yet
Fcurrently working to update
Rcool. You're updating the TE training version of db adafactor ?
Ryou have it at TE lr = 1/3 Unet lr
RYoure testing for betterTE training?
RFi am testing both
FTE and Unet right now
Fwill share on patreon
Fstarted new 9 training too
Fso total will be 4+9+9
F
NYes. I'll work on getting that fixed tomorrow with the team
Fi think he is asking any way to access jupyterlab interface from his computer
Flike gradio share
For like runpod proxy
Fadding adetailer and reactor to kaggle notebook
FKaggle notebook updated
F
FNYes just need to double check that labs and notebook are opening the port and using host 0.0.0.0
I think notebook does automatically but need to check labs
I think notebook does automatically but need to check labs
Fnew kohya trainings just finished
Fwow took a lot
Frunpod is really really slow