Software Engineering Courses (SECourses)•2y ago

Hello. I'm currently trying to create a face lora after watching your video: https://www.youtube.com

Hello. I'm currently trying to create a face lora after watching your video: https://www.youtube.com/watch?v=TpuDOsuKIBo&t=721s. The results are somewhat disappointing, possibly due to dataset issues, but it's definitely improved since I started experimenting on my own after watching the video. However, I'm wondering if cropping a person's face or full body shot from the background and turning it into a white background would be significantly beneficial? I actually feel like it's ruining the train and result looks weird right now

YouTubeSECourses

Generate Studio Quality Realistic Photos By Kohya LoRA Stable Diffu...

#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...

Generate Studio Quality Realistic Photos By Kohya LoRA Stable Diffu...

FFurkan Gözükara SECourses i replied your message on twitter too. that photo need upscale and overal improv...

andysingal•12/28/23, 1:37 PM

I tried adetail for upscaling but no success

B0nd•12/28/23, 1:42 PM

Whats the best way to upscale your dataset for maximum detail?

FFurkan Gözükara SECourses it is true

Digital [Starburst]•12/28/23, 5:50 PM

Any chance a group of us could try and develop code to win $700,000.00 by deciphering 4 ancient documents found near Pompeii? https://www.msn.com/en-us/news/technology/unlocking-ancient-wisdom-ai-decodes-greek-word-from-herculaneum-scrolls/ar-AA1m3ZUO?ocid=msedgntp&pc=EDGEESS&cvid=5b89fbd3e75548e799f8543b42b70568&ei=16

Digital [Starburst]•12/28/23, 5:51 PM

They are called the Herculaneum Scrolls, and since we are so hyped by AI and that is the key to unlocking the text it might be fun to do it even if we lose.

shortdwarf•12/28/23, 7:19 PM

Has anyone tried training a LoRA or Dreambooth using PNGs of a subject with a transparent background (essentially subject cut out and nothing behind)? I've tried with white background and backgrounds of random colors, both successful to remove associating the training with the background and focus more on the subject, however im curious if anyone has already tried this and could speak about the results
EDIT: When using ohwx man within a prompt, the random color background works quite well. However I just tried a simple "(ohwx man:1.3)" prompt and it mostly worked well however I did receive one photo that was just a purple background, so thats worth noting

BB0nd Whats the best way to upscale your dataset for maximum detail?

shortdwarf•12/28/23, 7:33 PM

From what I've seen online it seems .dat upscalers are the best. However I can't personally speak to this since those are not yet supported by A1111. But looking at this, https://openmodeldb.info/models/4x-FaceUpDAT it seems exceptional

SSquid Hello. I'm currently trying to create a face lora after watching your video: htt...

Furkan Gözükara SECourses•12/28/23, 7:48 PM

what is your GPU?

Furkan Gözükara SECourses•12/28/23, 7:48 PM

SD 1.5 is really weak. SDXL makes huge difference

BB0nd Whats the best way to upscale your dataset for maximum detail?

Furkan Gözükara SECourses•12/28/23, 7:49 PM

you shouldnt upscale your dataset that is bad

Furkan Gözükara SECourses•12/28/23, 7:49 PM

but probably hd ultimate upscaler script

Aandysingal I tried adetail for upscaling but no success

Furkan Gözükara SECourses•12/28/23, 7:49 PM

ye you need to find a way to upscale with adding more details

DDigital [Starburst]Any chance a group of us could try and develop code to win $700,000.00 by deciph...

Furkan Gözükara SECourses•12/28/23, 7:50 PM

probably you have to be expert on that field

Sshortdwarf Has anyone tried training a LoRA or Dreambooth using PNGs of a subject with a tr...

Furkan Gözükara SECourses•12/28/23, 7:50 PM

currently SDXL or SD 1.5 doesnt have transparent support feature

Furkan Gözükara SECourses•12/28/23, 7:51 PM

there is a new update to multi gpu training

Furkan Gözükara SECourses•12/28/23, 7:51 PM

therefore now i am doing a new training on kaggle

SSquid Hello. I'm currently trying to create a face lora after watching your video: htt...

mikemenders•12/28/23, 7:52 PM

I train Lora on SD 1.5 (use Realistic Vision 5.1) with absolutely great results, check back for my comments!

soymarcelojs•12/28/23, 7:54 PM

hi people! any news about the course banned from udemy? it´s the course in other place? i talk whit udemy and dont lisent to me for reupload the course

FFurkan Gözükara SECourses what is your GPU?

SquidOP•12/28/23, 7:59 PM

hi I lately bought 3090ti

Digital [Starburst]•12/28/23, 7:59 PM

They wouldn't listen to me either.

Mmikemenders I train Lora on SD 1.5 (use Realistic Vision 5.1) with absolutely great results,...

SquidOP•12/28/23, 8:00 PM

Thanks!! I will check it, I need to train Asian face so might need other base model to train

Digital [Starburst]•12/28/23, 8:00 PM

They refunded me the fee. For some reason I can't spell today.

SSquid Thanks!! I will check it, I need to train Asian face so might need other base mo...

mikemenders•12/28/23, 8:02 PM

I think it will run on Realistic Vision 5.1, and the trained Lora will also run well on other SD 1.5 models. However, if you have such a card, SDXL might still be a better alternative in the long run.

Ssoymarcelojs hi people! any news about the course banned from udemy? it´s the course in other...

Furkan Gözükara SECourses•12/28/23, 8:15 PM

hello welcome. thanks for contacting them

Furkan Gözükara SECourses•12/28/23, 8:15 PM

currently you can use here + patreon subscription

Furkan Gözükara SECourses•12/28/23, 8:16 PM

https://icarus-ai.enki.tech/course/zero-to-hero-generative-ai-become-a-master/

ICARUS AI

Zero To Hero Generative AI - Become A Master

Expert-Level Tutorials on Stable Diffusion & SDXL: Master Advanced Techniques and Strategies

Furkan Gözükara SECourses•12/28/23, 8:16 PM

courses are at the very bottom

Furkan Gözükara SECourses•12/28/23, 8:16 PM

i am testing new kaggle notebook feature

Kallamamran•12/28/23, 8:41 PM

Diffusers 0.25.0 released? Is that in any way relevant for us A1111-users?

Kallamamran•12/28/23, 8:43 PM

I'm currently at xformers: 0.0.22.dev597. Any experience with newer version @Dr. Furkan Gözükara ?

KKallamamran I'm currently at xformers: 0.0.22.dev597. Any experience with newer version @Dr....

Furkan Gözükara SECourses•12/28/23, 8:44 PM

well in my tests

Furkan Gözükara SECourses•12/28/23, 8:44 PM

i didnt get better

Furkan Gözükara SECourses•12/28/23, 8:44 PM

actually i recently tested

Kallamamran•12/28/23, 8:44 PM

Yeah thought so

Kallamamran•12/28/23, 8:44 PM

Same with older versions. Not much difference

Furkan Gözükara SECourses•12/28/23, 8:45 PM

yes

Furkan Gözükara SECourses•12/28/23, 8:46 PM

currently testing multi gpu training improvement on kaggle

Kallamamran•12/28/23, 8:47 PM

Just noticed torch 2.1.2 is realeased as well

Kallamamran•12/28/23, 8:47 PM

And Gradio 4.12.0

Furkan Gözükara SECourses•12/28/23, 9:05 PM

probably will break your existing apps

Furkan Gözükara SECourses•12/28/23, 9:05 PM

new kaggle training

Furkan Gözükara SECourses•12/28/23, 9:05 PM

i am expecting 2100 steps to train good enough

Furkan Gözükara SECourses•12/28/23, 9:05 PM

Mmikemenders I think it will run on Realistic Vision 5.1, and the trained Lora will also run ...

SquidOP•12/28/23, 9:15 PM

many thanks, I must try SDXL train soon too :nodblue: do you usually use reg imgs from DR.furkan's patron?

lieutenant_mac•12/28/23, 9:21 PM

Hello, just joined this Discord server. Not sure if this is old news that for lora training captions were not found due to wrong file extension? I was following this turorial: https://www.youtube.com/watch?v=AY6DMBCIZ3A

YouTubeSECourses

First Ever SDXL Training With Kohya LoRA - Stable Diffusion XL Trai...

Updated for SDXL 1.0. How to install #Kohya SS GUI trainer and do #LoRA training with Stable Diffusion XL (#SDXL) this is the video you are looking for. I have shown how to install Kohya from scratch. The best parameters to do LoRA training with SDXL. How to use Kohya SDXL LoRAs with ComfyUI. How to do checkpoint comparison with SDXL LoRAs and m...