Software Engineering Courses (SECourses)•16mo ago

Hi @Dr. Furkan Gözükara , After i generated lora of product images with flux, its gving amazing resu

Hi @Dr. Furkan Gözükara , After i generated lora of product images with flux, its gving amazing results. Thanks a lot for great explainer video, this is promising to solve my long term problems on realistic product images,

How can i Dictate swarm not to change the size and color of trained product image while generating the final images.?

FFurkan Gözükara SECourses you should post it as error on kohya meanwhile

GomezBro•9/5/24, 4:44 AM

Where do I post?

Tacite•9/5/24, 6:10 AM

I now use SOLAR-10.7B-Instruct-v1.0-uncensored-GGUF to generate prompt for Flux and it’s night and day, my generated images are incredible.

TTacite I now use SOLAR-10.7B-Instruct-v1.0-uncensored-GGUF to generate prompt for Flux ...

GomezBro•9/5/24, 7:29 AM

what is that

GomezBro•9/5/24, 7:29 AM

share samples

Tacite•9/5/24, 7:35 AM

It’s a LLM. Basically you type : generate prompt for « a man riding a bicycle in the street » and it generates a very long prompt with many details

Tacite•9/5/24, 7:35 AM

I use this node https://github.com/SeargeDP/ComfyUI_Searge_LLM

GitHub

GitHub - SeargeDP/ComfyUI_Searge_LLM: Custom nodes for ComfyUI that...

Custom nodes for ComfyUI that utilize a language model to generate text-to-image prompts - SeargeDP/ComfyUI_Searge_LLM

Tacite•9/5/24, 7:36 AM

The model I use is this one https://huggingface.co/TheBloke/SOLAR-10.7B-Instruct-v1.0-uncensored-GGUF

TheBloke/SOLAR-10.7B-Instruct-v1.0-uncensored-GGUF · Hugging Face

virginia•9/5/24, 8:43 AM

@Dr. Furkan Gözükara i still cannot understand your new post of kaggle swarmUI, I have sent you a private message

slipkn67•9/5/24, 9:08 AM

hello, guys, please tell me: to make cool realistic photos of myself flying like superman I need to use 2 loras? 1 of myself and 2nd is realism lora?

Jjhb Hi @Dr. Furkan Gözükara , After i generated lora of product images with flux, it...

Furkan Gözükara SECourses•9/5/24, 10:27 AM

you need a better dataset for it

GGomezBro Where do I post?

Furkan Gözükara SECourses•9/5/24, 10:27 AM

https://github.com/kohya-ss/sd-scripts/issues

GitHub

Issues · kohya-ss/sd-scripts

Contribute to kohya-ss/sd-scripts development by creating an account on GitHub.

Vvirginia @Dr. Furkan Gözükara i still cannot understand your new post of kaggle swarmUI, ...

Furkan Gözükara SECourses•9/5/24, 10:30 AM

i replied try to figure out logic easyt

Sslipkn67 hello, guys, please tell me: to make cool realistic photos of myself flying like...

Furkan Gözükara SECourses•9/5/24, 10:30 AM

you can just train on flux and sufficient

magical•9/5/24, 12:23 PM

hello if i want to train a flux lora and try 32 network rank should my network alpha be 32 or 1

magical•9/5/24, 12:23 PM

in sdxl i used to set it to 1

Mmagical hello if i want to train a flux lora and try 32 network rank should my network a...

Furkan Gözükara SECourses•9/5/24, 12:47 PM

Set equal

Furkan Gözükara SECourses•9/5/24, 12:47 PM

Since if alpha is not set equal a new LR needs to be researched

magical•9/5/24, 12:48 PM

thank you

magical•9/5/24, 2:30 PM

is "Rank_3_18950MB_9_05_Second_IT" config for gpus with 24gb vram bf16 or are only the configs for 48gb vram bf16?

Mmagical is "Rank_3_18950MB_9_05_Second_IT" config for gpus with 24gb vram bf16 or are on...

Furkan Gözükara SECourses•9/5/24, 3:03 PM

rank 3 is for 24 gb

Furkan Gözükara SECourses•9/5/24, 3:03 PM

rank 1 and 2 for 48gb

Websteria•9/5/24, 3:06 PM

@Dr. Furkan Gözükara Using the Rank5 and the latest Torch 2.5, I'm trying to train, it gets here and then just seems to not be doing the training. Also even though it says it uses less memory, my GPU is still using 23.37g out of 24.

Websteria•9/5/24, 3:07 PM

nevermind it just started training although extremely slowly. If the lora source images, and target direectory are on a slower drive would that cause a delay?

WWebsteria nevermind it just started training although extremely slowly. If the lora source...

Furkan Gözükara SECourses•9/5/24, 3:08 PM

it can cause delay in initial

Furkan Gözükara SECourses•9/5/24, 3:08 PM

start

Furkan Gözükara SECourses•9/5/24, 3:08 PM

but not after it

lilwashuu10•9/5/24, 3:08 PM

Im curious how much RAM are you running with, looks like FLUX is heavy on the PC RAM as well

Llilwashuu10 Im curious how much RAM are you running with, looks like FLUX is heavy on the PC...

Furkan Gözükara SECourses•9/5/24, 3:12 PM

ye it uses a lot

Furkan Gözükara SECourses•9/5/24, 3:12 PM

i have 64 gb so never cared :d

Furkan Gözükara SECourses•9/5/24, 3:12 PM

also you can set virtual ram

Furkan Gözükara SECourses•9/5/24, 3:12 PM

ram is nowadays easier

lilwashuu10•9/5/24, 3:13 PM

Yeah I have more coming in the mail, but I think paging is what is causing some peoples slow down

LuckyStrike•9/5/24, 3:14 PM

Hey, so I ran a couple of successful Flux lora's. with the rank three settings, but. the file size is like 2.3 gigs and I'm thinking I must have done something wrong to end up with that file size. I remember having that mistake before when I accidentally ran it on the Dreambooth tab instead of the Lora tab. But that's not what I did here. Is there anything apparent that I'm doing wrong?

lilwashuu10•9/5/24, 3:14 PM

From the video it looked like Furkans where about that size

LuckyStrike•9/5/24, 3:15 PM

Oh, so maybe that's the expected size using this approach. It's big, but if it works, then I'm totally fine with that.

lilwashuu10•9/5/24, 3:16 PM

Yeah I tried the rank 12 which does i think 512x512 training and the model was 525MB

lilwashuu10•9/5/24, 3:17 PM

I am sure it goes up in size with the resolution, since it is gathering more information

FFurkan Gözükara SECourses rank 3 is for 24 gb

magical•9/5/24, 3:18 PM

is rank 3 config bf16?

Mmagical is rank 3 config bf16?

Furkan Gözükara SECourses•9/5/24, 3:32 PM

all are bf16 but rank 5 and lower cast into fp8

Furkan Gözükara SECourses•9/5/24, 3:37 PM

some 16 gb configs doesnt cast but they are lower resolution

magical•9/5/24, 3:39 PM

I was using this extension https://github.com/cheald/sd-webui-loractl with a1111 to change the loras weight at different steps for example I can start one lora at first step and stop it at 15th step and start another lora at 15th step until the end. Now I'm looking if I can do it in SwarmUI and the only thing I found is Lora Block Weight node in ComfyUI-Inspire-Pack.

magical•9/5/24, 3:39 PM

So could I download this https://github.com/ltdrdata/ComfyUI-Inspire-Pack and use it inside SwarmUI?

Mmagical So could I download this https://github.com/ltdrdata/ComfyUI-Inspire-Pack and us...

Furkan Gözükara SECourses•9/5/24, 3:58 PM

not sure may not work

slipkn67•9/5/24, 4:21 PM

it has been running over 3hrs and only 844/2800 done. How do I pause it to continue later? Running 14 photos on RTX 3060 12GB.