Software Engineering Courses (SECourses)•3y ago

how to train with config is here https://youtu.be/EEV8RPohsbw

YouTubeSECourses

How To Do Stable Diffusion XL (SDXL) DreamBooth Training (Full Fine...

Patreon post for config files

https://www.patreon.com/posts/89213064

SECourses Discord To Get Full Support

https://discord.com/servers/software-engineering-courses-secourses-772774097734074388

My LinkedIn

https://www.linkedin.com/in/furkangozukara/

My Twitter

https://twitter.com/GozukaraFurkan

My Instagram

https://www.instagra...

FFurkan Gözükara SECourses how to train with config is here https://youtu.be/EEV8RPohsbw

Chip•12/20/23, 9:41 PM

that is what I did using Kohya but the results are not as good as yours

FFurkan Gözükara SECourses how to train with config is here https://youtu.be/EEV8RPohsbw

Chip•12/20/23, 9:42 PM

do I need regularisation images?

CChip do I need regularisation images?

Furkan Gözükara SECoursesOP•12/20/23, 9:44 PM

reg images makes it improved

Furkan Gözükara SECoursesOP•12/20/23, 9:44 PM

where did you train? did you use my config?

Chip•12/20/23, 9:45 PM

yes, used your config. used rundiffusion server with i think enough ram

Chip•12/20/23, 9:46 PM

still better results than when i train a quick model one dreamlook.ai but not as good as yours

CChip yes, used your config. used rundiffusion server with i think enough ram

Furkan Gözükara SECoursesOP•12/20/23, 9:49 PM

rundiffusion server supports kohya?=

Furkan Gözükara SECoursesOP•12/20/23, 9:49 PM

you should train on runpod

Chip•12/20/23, 9:49 PM

yes

Chip•12/20/23, 9:49 PM

yes will try in the future

Furkan Gözükara SECoursesOP•12/20/23, 9:49 PM

i doubt that they are properly supporting

Furkan Gözükara SECoursesOP•12/20/23, 9:49 PM

you can also message me your training dataset i can comment

Chip•12/20/23, 9:51 PM

ok, sure. thanks so much

FFurkan Gözükara SECourses how are you trying can you show screenshot

StarBuffer•12/20/23, 10:48 PM

I am just using diffusers to implement those things.

SStarBuffer Click to see attachment

StarBuffer•12/20/23, 10:48 PM

This is the basic result.

SStarBuffer I am just using diffusers to implement those things.

Furkan Gözükara SECoursesOP•12/20/23, 11:04 PM

diffusers is

Furkan Gözükara SECoursesOP•12/20/23, 11:04 PM

very different than automatic1111 or kohya

Furkan Gözükara SECoursesOP•12/20/23, 11:04 PM

very primitive

StarBuffer•12/20/23, 11:23 PM

Is there a way to use WebUI extension directly in diffusers?

StarBuffer•12/20/23, 11:23 PM

Because I think what I am trying to do is outpainting more than inpainting, and seems inpainting models can't work properly(because they make changes from side.)

SStarBuffer Is there a way to use WebUI extension directly in diffusers?

Furkan Gözükara SECoursesOP•12/20/23, 11:26 PM

web ui has api

Furkan Gözükara SECoursesOP•12/20/23, 11:26 PM

you can use it

StarBuffer•12/20/23, 11:27 PM

So running WebUI and install extensions and access them via APIs?

SStarBuffer So running WebUI and install extensions and access them via APIs?

Furkan Gözükara SECoursesOP•12/20/23, 11:29 PM

yep

StarBuffer•12/20/23, 11:30 PM

Understood.
are there any outpainting models available in diffusers?

SStarBuffer Understood. are there any outpainting models available in diffusers?

Furkan Gözükara SECoursesOP•12/20/23, 11:31 PM

i dont know sadly

StarBuffer•12/20/23, 11:32 PM

Could you explain some details of inpainting parameters and initialization?

StarBuffer•12/20/23, 11:33 PM

I think adjust noise can affect the final result.

SStarBuffer I think adjust noise can affect the final result.

Furkan Gözükara SECoursesOP•12/21/23, 12:11 AM

sadly i only used automatic1111 implementation

FFurkan Gözükara SECourses also what is your python version @Motoko_Shi

Motoko_Shi•12/21/23, 2:33 AM

3.10.11

nad•12/21/23, 6:01 AM

Hello @Dr. Furkan Gözükara I have been following your tutorial where you setup C++, ffmpeg and CUDA.

Unfortunately, I am on mac and CUDA setup is not possible. Should I use online Google colab as an alternative or do you have better suggestion ?

AIGambino•12/21/23, 6:24 AM

Do you have any tutorials on using different controlnets? Like CN tile and upscaling?

FFurkan Gözükara SECourses this is not my notebook right?

eyepatch_2022•12/21/23, 7:43 AM

Yes.
Actually after that I'm not able to start ngrok link.

eyepatch_2022•12/21/23, 7:44 AM

rn_image_picker_lib_temp_48acedd3-d260-4fe7-bc98-53bf361e925f.jpg

Vvv will really apprciate the help; am trying to get something where my model is hel...

vv•12/21/23, 9:22 AM

can anyone help ;_;

Lorelai AI•12/21/23, 9:22 AM

Training 1.5 is a lot faster. first time trying it. 10 minutes vs 1 hour same data set except the checkpoint

FFurkan Gözükara SECourses i am yet to explore it

Imogen Fox•12/21/23, 10:04 AM

Which is best way to get gif/mp4 output using stable diffusion with your own style trained model?

Nnad Hello @Dr. Furkan Gözükara I have been following your tutorial where you setup C...

Furkan Gözükara SECoursesOP•12/21/23, 10:10 AM

i suggest to use runpod

Furkan Gözükara SECoursesOP•12/21/23, 10:11 AM

it is really good. i also have tutorials and auto installers

AAIGambino Do you have any tutorials on using different controlnets? Like CN tile and upsca...

Furkan Gözükara SECoursesOP•12/21/23, 10:11 AM

i have a big tutorial where i covered a lot of stuff you can look chapters

Furkan Gözükara SECoursesOP•12/21/23, 10:11 AM

https://youtu.be/3E5fhFQUVLo

YouTubeSECourses

Zero to Hero ControlNet Tutorial: Stable Diffusion Web UI Extension...

I show how to install Automatic1111 Web UI & ControlNet extension installation from scratch in this video. Moreover I show how to make amazing QR codes and inpainting and out painting of ControlNet which are very similar to Photoshop generative fill and Midjourney zoom out. Furthermore, I explain and show what are Canny, Depth, Normal, OpenPose...