Software Engineering Courses (SECourses)•14mo ago

Hi <@205854764540362752> have you tried training a style with dreambooth. I wonder what is different

Hi @Furkan Gözükara SECourses have you tried training a style with dreambooth. I wonder what is different when a checkpoint learning a style compare with lora. Does it learn faster so we can train with lower epoch

harap.•11/11/24, 5:00 AM

hi, I trained this sdxl checkpoint in the past https://civitai.com/models/570998/cocaxl-01 but it was impossible to generate character sheets with it. So I trained another checkpoint based on https://civitai.com/models/293331/wildcardx-xl-turbo, the previous one had been trained on juggernaut xl. Not character sheets work better, but there is a problem.

harap.•11/11/24, 5:01 AM

for the same prompt the new model outputs this

harap.•11/11/24, 5:01 AM

the old one outputed this

harap.•11/11/24, 5:02 AM

I don't mind the output, but it doesn't render the clothing as well as the old model

harap.•11/11/24, 5:03 AM

I am talking about this

harap.•11/11/24, 5:04 AM

for example for the footwear, when I input "man wearing ((opinci-like moccasins)) " the old model creates pretty accurate opinci, the new one, not as much, also it creates 2 people where as I only asked for one. And in general it doesn't stick to the prompt as well.

harap.•11/11/24, 5:05 AM

So my question is: is there something I did wrong with the training? can I do anything to make the new model be more like the old model with respect to accurately represanting the clothing

LLeolis78 I would use a single folder, 1 nameofrock bandxyz people (1 repeat_triggerword ...

Furkan Gözükara SECourses•11/11/24, 10:43 AM

yes but i wont read folder caption if you use txt captions

Wwardensc2 Hi <@205854764540362752> have you tried training a style with dreambooth. I wond...

Furkan Gözükara SECourses•11/11/24, 10:44 AM

yes i did read here : https://huggingface.co/MonsterMMORPG/3D-Cartoon-Style-FLUX

MonsterMMORPG/3D-Cartoon-Style-FLUX · Hugging Face

Hharap.So my question is: is there something I did wrong with the training? can I do an...

Furkan Gözükara SECourses•11/11/24, 10:45 AM

impossible to know without looking everything it requires full time debug. you can also try flux did you try?

harap.•11/11/24, 11:47 AM

it takes way to much time, I have 340 training images scanned from old children's story books, it took 27 hours to train on my wife's hp omen

FFurkan Gözükara SECourses yes i did read here : https://huggingface.co/MonsterMMORPG/3D-Cartoon-Style-FLUX

wardensc2OP•11/11/24, 12:10 PM

But this is lora not a full model, I already read your writing many times but it's only about lora training

Wwardensc2 But this is lora not a full model, I already read your writing many times but it...

Furkan Gözükara SECourses•11/11/24, 12:26 PM

it is exactly same for fine tuning

Furkan Gözükara SECourses•11/11/24, 12:26 PM

you only change config

FFurkan Gözükara SECourses you only change config

wardensc2OP•11/11/24, 1:03 PM

Yes but i wonder how about learning speed does full model require less epoch to learn style than lora.

Wwardensc2 Yes but i wonder how about learning speed does full model require less epoch to ...

Furkan Gözükara SECourses•11/11/24, 1:09 PM

it should be around same

FFurkan Gözükara SECourses impossible to know without looking everything it requires full time debug. you c...

harap.•11/11/24, 1:10 PM

if I increase the text encoding rate and train again could that improve the results? I used the low vram settings for your tutorial, I just increased the repetitions from 40 to 100 on the first screen

Hharap.if I increase the text encoding rate and train again could that improve the resu...

Furkan Gözükara SECourses•11/11/24, 1:11 PM

it can cook text encoder

Felipe Machado•11/11/24, 1:35 PM

Hi Boss @Furkan Gözükara SECourses , for SDXL training I can more or less use the same settings for flux, just changing at the beginning ? Or you recommend different tutorial?

FFelipe Machado Hi Boss <@205854764540362752> , for SDXL training I can more or less use the sam...

Furkan Gözükara SECourses•11/11/24, 1:44 PM

we have sdxl config

Furkan Gözükara SECourses•11/11/24, 1:44 PM

you should use it as a base

Felipe Machado•11/11/24, 1:46 PM

yes, but you dont have like sdxl config file by any chance?

FFelipe Machado yes, but you dont have like sdxl config file by any chance?

Furkan Gözükara SECourses•11/11/24, 1:50 PM

yes i have here : https://www.patreon.com/posts/very-best-for-of-89213064

Patreon

The Very Best Workflow For SDXL DreamBooth / Full Fine Tuning - Res...

Get more from SECourses: FLUX, Tutorials, Guides, Resources, Training, Scripts on Patreon

Felipe Machado•11/11/24, 1:50 PM

YES thank you

Felipe Machado•11/11/24, 1:51 PM

have you ever done something like 250 images for sdxl fine tuning?

FFelipe Machado have you ever done something like 250 images for sdxl fine tuning?

Furkan Gözükara SECourses•11/11/24, 1:56 PM

yes i had done style

Furkan Gözükara SECourses•11/11/24, 1:56 PM

even more than that

Felipe Machado•11/11/24, 2:06 PM

cool

Felipe Machado•11/11/24, 2:06 PM

uow so much faster then flux right?

Felipe Machado•11/11/24, 2:07 PM

impressed with the range of emotions in a 256 data set

Felipe Machado•11/11/24, 2:07 PM

me as a dwarf

FFelipe Machado impressed with the range of emotions in a 256 data set

Furkan Gözükara SECourses•11/11/24, 2:08 PM

nice

Felipe Machado•11/11/24, 2:10 PM

if you turn down guidence you can stilize really well

FFelipe Machado if you turn down guidence you can stilize really well

Furkan Gözükara SECourses•11/11/24, 2:48 PM

nice

Xcorpion•11/12/24, 12:48 AM

just trained a lora on a recent sdxl model (noobAI V-pred0.5), and it turned out like this. but some loras trained on epsilon versions works with vpred checkpoints.
this is just a speculation, since the checkpoint is not epsilon structure (dont know what is epsilon or vpred meaning) kohya can't train it or smthing?

03523-451806595-masterpiece20best20quality20absurdres20highly20detailed_lora_style1noobai0.png

XXcorpion just trained a lora on a recent sdxl model (noobAI V-pred0.5), and it turned out...

Furkan Gözükara SECourses•11/12/24, 12:52 AM

very likely. try to train on sdxl base first

Furkan Gözükara SECourses•11/12/24, 12:52 AM

and see if your config works

Xcorpion•11/12/24, 6:05 AM

@Furkan Gözükara SECourses a friend asked me to train using his gpu too, so is it possible in kohya or in any other way? I got 4070, he got 4060ti.

Xcorpion•11/12/24, 6:07 AM

and I see these checkpoint models trained on 2x A100 GPU, 4x A100 GPU... so we can combine gpus or smthing?

XXcorpion <@205854764540362752> a friend asked me to train using his gpu too, so is it pos...

Furkan Gözükara SECourses•11/12/24, 11:47 AM

nope you cant combine

Furkan Gözükara SECourses•11/12/24, 11:47 AM

like that

Furkan Gözükara SECourses•11/12/24, 11:47 AM

on linux 2x 3090 may work for lora

Felipe Machado•11/13/24, 1:51 PM

hello again, how do I resume a Lora training in Khoya? I know how to pickup a training I left when training a finetuned checkpoint by selecting the latest saved safetensor in the " Pretrained model name or path" but I dont know hot to resume a previews Lora training , thanks for the help

FFelipe Machado hello again, how do I resume a Lora training in Khoya? I know how to pickup a tr...

Furkan Gözükara SECourses•11/13/24, 1:56 PM

tips-info

wardensc2OP•11/15/24, 1:03 AM

Hi @Furkan Gözükara SECourses have you ever train hyper flux or turbo flux model

wardensc2OP•11/15/24, 1:04 AM

I trained model and merge with hyper or turbo 8 step lora. The image result for 8 step for normal gen is ok but the upscale quality is worst than normal

wardensc2OP•11/15/24, 1:06 AM

I wonder if i train on an already merged hyper or turbo does image will be better or not

Wwardensc2 Hi <@205854764540362752> have you ever train hyper flux or turbo flux model

Furkan Gözükara SECourses•11/15/24, 12:19 PM

nope i havent

yassinemenace•11/15/24, 4:39 PM

@Furkan Gözükara SECourses do you do face inpainting for close up photos ?

yassinemenace•11/15/24, 4:40 PM

Sometimes it messes up with eyes looking direction

Hi <@205854764540362752> have you tried training a style with dreambooth. I wonder what is different

Similar Threads