Hm too many Lora videos open at once haha. Will you explore it in the future?
Hm too many Lora videos open at once haha. Will you explore it in the future?
Fpossibly. the drembooth extension developer also working on lora
I@Dr. Furkan Gözükara do you think on doing any tutorials on using other tools like invokeAi, and maybe detailed tutorials on inpainting - i was just doing an indepth dive into creating very large detailed pieces with step by step inpainting - and the process is very interesting and rewarding in the end - its not fast - but oh so cool 
LHi friends. Can anyone explain the correct settings and prompt style for inpainting new subjects into a scene? I am trying to make a second kitten in this photo with the prompt "close up of a kitten in a field" but it just changes the grass and the trees? I am using an inpainting model, and denoise strength is set to 1.
C
Cpaint white in black area
Cprompt: Cat
C
CSomething like that
Celse the sampler will work on the layer already behind you, which so happen to be GREEN like grass
CAutomatic111 inpainting 2/10
Invoke AI : 7/10
Invoke AI : 7/10
Fi should make. just cant find time to make all atm. also earthquake really affected me as well
Fi would say first upscale to 4x
Flet me test and show
Fye i couldnt make it work either. looks like it is expecting kitten like thing there
LI put two kittens in the prompt and it worked. It seems like you have to describe the whole image when creating something new, instead of only the inpainted area.
L
LIf you want a puppy instead of a kitten, the prompt needs to be "close up of two puppies in a field"
LIf you prompt with "close up of a kitten and puppy in a field" it will give you another kitten
Fwow nice info i didnt know
LIt looks like if you want anything other than the original subject, you have to write two first, or else it won't register. Example:
close up of two laptops in a field
close up of two laptops in a field
Fhaha very good
Aanyone know how to get Automatic1111 SD to run under a different IP address/port? i am running kohya concurrently, and if i run them both same time... they interfere and only one ends up available
乂nice paint 
Fez. --port 7777
Fit will start as 127.0.0.1:7777
Athat worked. thanks!
Jim getting this error in google colab
Jand the path to my drive is correct
Cno, it cant count.
It just a random seed
It just a random seed
Cor maybe it could see kittens as two
Chmm
乂I think this refers to selecting a model to fine-tune
乂like v1-5-pruned.ckpt from huggingface.com
C
乂love this 
C
CThis
Ccould probably fix the small eye problem
Icheck this out
I(open it in the browser to appreciate all the details)
FYouTubeSoftware Engineering Courses - SE Courses
Discord : https://discord.gg/HbqgGaZVmr. This is the video where you will learn how to use Google Colab for Stable Diffusion. If I have been of assistance to you and you would like to show your support for my work, please consider becoming a patron on 
https://www.patreon.com/SECourses
Google colab notebook link: https://colab.research.google....
https://www.patreon.com/SECoursesGoogle colab notebook link: https://colab.research.google....
Qanyone know how to convert ckpt to onnx? the converter in nmkd says it failed to convert.
Fwhat is onnx used for
Fi am gonna look for controlnet extension of automatic1111 now
Qonnx is so you can use it on an amd gpu on windows
Fi see. nmkd supports amd. but you say it failed. i think you can ask nmkd developer
Fthis control net extension of automatic1111 is really problematic
Fmade it work as automatic1111 extension
HHave you been able to try training 2.1 768 yet?
HI see some good models on Civitai now in 768