Software Engineering Courses (SECourses)•17mo ago

I'll look at the full video tonight with some popcorn 🙂 Is there a part for style/concept training,

I'll look at the full video tonight with some popcorn

Is there a part for style/concept training, or is it mostly for character ?

PPhenyX I'll look at the full video tonight with some popcorn 🙂 Is there a part for sty...

Furkan Gözükara SECourses•8/29/24, 1:01 PM

for style training i got a good dataset

Furkan Gözükara SECourses•8/29/24, 1:01 PM

will make a new video for it and share results

Furkan Gözükara SECourses•8/29/24, 1:01 PM

but watching this super important

J2_Nedo•8/29/24, 1:03 PM

i still didnt startet to scan my original paintings from the 1990, but now i will because of Flux, yes a style/concept video would be interessting too.

pixelscience•8/29/24, 1:22 PM

I just made one from 1700s Japanese wood block prints and it came out amazing!!!!

JJ2_Nedo i still didnt startet to scan my original paintings from the 1990, but now i wil...

Furkan Gözükara SECourses•8/29/24, 1:29 PM

yep

Ppixelscience I just made one from 1700s Japanese wood block prints and it came out amazing!!!...

Furkan Gözükara SECourses•8/29/24, 1:29 PM

awesome

PhenyXOP•8/29/24, 2:13 PM

PhenyXOP•8/29/24, 2:13 PM

@Dr. Furkan Gözükara is that a big error or i can move forward with it ?

protector131090•8/29/24, 2:16 PM

if you dont use xfromers - ignore

PhenyXOP•8/29/24, 2:22 PM

ok thanks. I indeed done use it in comfy or forge

PhenyXOP•8/29/24, 2:23 PM

PhenyXOP•8/29/24, 2:23 PM

But if i have this, i should install the cu124 ? @protector131090

dsienra•8/29/24, 2:23 PM

@Dr. Furkan Gözükara I still have the bleeding problem, is imposible to train two or more people, may be is because the lr of TE and unet are incorrect, all bleeds, there is something missing, the base model can differentiate between "Michael Jackson" and "Michael Keaton" both man both Michael and the model can differentiate so is something fundamental that is missing, something did to train the base model that is missing now

Jehex•8/29/24, 2:24 PM

@Dr. Furkan Gözükara Hi, I successfully full finetuned flux but now I got diffusion_pytorch_model-00001-of-00003.safetensors, diffusion_pytorch_model-00002-of-00003.safetensors and diffusion_pytorch_model-00003-of-00003.safetensors after the training, I guess at this point I will have to fusion/merge theses 3 parts for created the whole checkpoint usable for Forge ( so reshard theses 3 shards ) . Would like to know a way to do it, thanks

JJehex @Dr. Furkan Gözükara Hi, I successfully full finetuned flux but now I got diffus...

�

𝔸𝖨ΞӾ•8/29/24, 2:26 PM

He claims to have finetuned Flux @Dr. Furkan Gözükara I thought it wasn't even possible yet?

�𝔸𝖨ΞӾ He claims to have finetuned Flux @Dr. Furkan Gözükara I thought it wasn't even p...

Jehex•8/29/24, 2:27 PM

it's honestly easy with runpod, I think for now I need a tool or a script for "reshards " the 3 parts I get from the training

JJehex it's honestly easy with runpod, I think for now I need a tool or a script for "r...

�

𝔸𝖨ΞӾ•8/29/24, 2:27 PM

Was it through Dreambooth tab in Kohya?

�𝔸𝖨ΞӾ Was it through Dreambooth tab in Kohya?

Jehex•8/29/24, 2:28 PM

I tried with kohya but switch to ostris ai tool kit

�

𝔸𝖨ΞӾ•8/29/24, 2:28 PM

Very interesting

PPhenyX Click to see attachment

Furkan Gözükara SECourses•8/29/24, 2:57 PM

it is because of cuda torch versions

Furkan Gözükara SECourses•8/29/24, 2:57 PM

you can ignore xformers we dont use it

Furkan Gözükara SECourses•8/29/24, 2:57 PM

for flux

Ddsienra @Dr. Furkan Gözükara I still have the bleeding problem, is imposible to train t...

Furkan Gözükara SECourses•8/29/24, 2:57 PM

someone said he trained 2 person perfectly with no bleed

Furkan Gözükara SECourses•8/29/24, 2:58 PM

use 5 image person a 5 person b and 5 person a and b in same image

Furkan Gözükara SECourses•8/29/24, 2:58 PM

did you follow my latest advice? no class token

JJehex @Dr. Furkan Gözükara Hi, I successfully full finetuned flux but now I got diffus...

Furkan Gözükara SECourses•8/29/24, 2:58 PM

what softwrae did you use? kohya doesnt do that

�𝔸𝖨ΞӾ He claims to have finetuned Flux @Dr. Furkan Gözükara I thought it wasn't even p...

Furkan Gözükara SECourses•8/29/24, 2:58 PM

fine tuning fully possible

Furkan Gözükara SECourses•8/29/24, 2:58 PM

i am gonna do with kohya

FFurkan Gözükara SECourses i am gonna do with kohya

�

𝔸𝖨ΞӾ•8/29/24, 3:00 PM

Oh ok it's becasue I haven't really seen anyone yet with an actual working model. I look forward to your research in that area though

�𝔸𝖨ΞӾ Oh ok it's becasue I haven't really seen anyone yet with an actual working model...

Furkan Gözükara SECourses•8/29/24, 3:00 PM

ye it is next research

Furkan Gözükara SECourses•8/29/24, 3:00 PM

with 8 gpus :d

Furkan Gözükara SECourses•8/29/24, 3:00 PM

i expect to go as low as 12 gb

Furkan Gözükara SECourses•8/29/24, 3:00 PM

but we will see

Furkan Gözükara SECourses•8/29/24, 3:00 PM

there are some serious optimizations

FFurkan Gözükara SECourses did you follow my latest advice? no class token

dsienra•8/29/24, 3:03 PM

yes I did, no class, now I balanced better the dataset, and I will make a test with regularization images from the class person because I have different ages and genders, and let's see, my dataset was a little unbalanced all man looked like me even I didn't used any class

pixelscience•8/29/24, 3:05 PM

Dude on reddit suposedly fine tuned it. Link is a discord link that doesn't work :/
https://www.reddit.com/r/StableDiffusion/comments/1f40dvc/finetunning_the_flux1_for_your_own_data/

From the StableDiffusion community on Reddit: Finetunning the FLUX....

Explore this post and more from the StableDiffusion community

Maxivi•8/29/24, 3:05 PM

@Dr. Furkan Gözükara have you tried this prompt Anime illustration portrait of ohwx man, forest with lake in the background, high quality, beautiful face, action pose, clear focus, detailed and fantasy setting

dsienra•8/29/24, 3:07 PM

@Dr. Furkan Gözükara but I have to say same dataset same caption including class on sdxl no bleeding at all, I trained 20 or more people at the same time on a single sdxl lora and no bleeding at all, something is missing, idk

Ppixelscience Dude on reddit suposedly fine tuned it. Link is a discord link that doesn't work...

Maxivi•8/29/24, 3:08 PM

this is not a finetune, this is lora training, i went on his discord its just 512x512 lora training with fal

pixelscience•8/29/24, 3:08 PM

lol, I figured it might be something shady

Maxivi•8/29/24, 3:09 PM

his discord is quite shady too i left

Ddsienra yes I did, no class, now I balanced better the dataset, and I will make a test w...

Furkan Gözükara SECourses•8/29/24, 3:10 PM

balance matters