Thanks for the kohya video @Furkan Gözükara SECourses you absolute legend. ❤️
Thanks for the kohya video @Furkan Gözükara SECourses you absolute legend. 
FSorry guys I have been away will read messages
FTy so much for the feedback
FYou don't need. () will make prompt strength 1.1
FYe I also found 128 performed best with alpha value 1. I plan to test even higher values
FI would try with same settings and even with higher network rank and compare
FYouTubeSECourses
#Kohya SS web GUI DreamBooth #LoRA training full tutorial. You don't need technical knowledge to follow this tutorial. In this tutorial I have explained how to generate professional photo studio quality portrait / self images for free with Stable Diffusion training.
Our Discord server
https://bit.ly/SECoursesDiscord
If I have been of assis...
Our Discord server
https://bit.ly/SECoursesDiscord
If I have been of assis...
FThere is Linux instructions on the github page but I haven't tested yet. I plan to install on runpod and hopefully make a tutorial
FI think dreambooth still better
FI use the default one that comes with automatic1111
FImg2img works great when you use low denoise like 30% and upscale image to like 2x with same prompts
FCan you show above part of the screen
FYou can ignore it
FI was gonna ask this therefore asked above part of screen haha 
FYep. I find that no captions works better for teaching person or a style. Captions I think need for fine tuning
FNot possible atm. It's architecture is different. We need modified script. Kandinsky has training scripts on github but I haven't tested them yet
Zhave you noticed much of a difference between the results from kohya and auto1111?
FIt makes a venv so shouldn't affect your other installations
Iooh shoot can i continue to use 2 or do i need to uninstall everything now
HVery cool, I don’t think my machine can handle kohya training but I will watch
Ihow do you train a lora in auto1111?
FSomething must be wrong. You are using realistic vision v2? With man word?
FVery weird you are using text 2 img?
FI don't use captions for faces. I find that they degrade teaching a person
FThe logic is different. You can watch my this video to understand token logic https://youtu.be/Bdl-jWR3Ukc
YouTubeSECourses
Our Discord : https://discord.gg/HbqgGaZVmr. The most advanced tutorial of Stable Diffusion Dreambooth Training. 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
Playlist of Stable Diffusion Tutorials, Automatic1111 and Google Colab ...
https://www.patreon.com/SECoursesPlaylist of Stable Diffusion Tutorials, Automatic1111 and Google Colab ...
FThank you so much for the amazing comment
FI haven't compared yet but planning to do
FYou can continue. It is just slower. Only important thing is xformers version. Xformers 0.0.16 don't work
Ii have 0.0.18
FYou are welcome. It supports cpu training too. How much vram you had?
FOK it works 
FHopefully I will make a video waiting them to merge dev into main
D@Furkan Gözükara SECourses Possibly another issue for me, I am not running "Darwin" debian but a debian derivative, POP_OS by system76, so when I run the "upgrade.sh" script it errors out. I'm going to look for another solution to train on a GNU/Linux system. Maybe one with more detailed instructions to make the system work properly. Thank you!
FYes hopefully I will make a tutorial for Runpod. It uses Linux so you should be able to follow instructions
Iwhy is torch 2 slower on kohya ss?
FYe that is interesting. Maybe because I tested with Fp16 and not bf16
FCan you also test?
Ii selected no there
Ii have 24GB VRAM
FYes makes sense
FMaybe that is reason
Ialmost ready to train, working through your tutorial, thank you
Ihow are you running stable diffusion and kohya ss at the same time, it told me i could only run one at a time
Invm got it
IHey guys, i need to use SD+ControlNet+ Temporal Kit, do you have any colab i can use please? i used colab only for dreambooth
FChange port
FUnfortunately don't have colab for that :/
IThanks anyways, i'm doing some test for making a decent video but i need more vram
Imy 8gb are decent but with more i can have more stability
Ilol now to gen 600 images of "man"