Software Engineering Courses (SECourses)•16mo ago

I have 800 different characters, between accessories and country, do you think that training with th

I have 800 different characters, between accessories and country, do you think that training with the 800 images is not the way to go?
I should choose 100 and work with them?
Because per country I have 4-6 versions which share the same flag color, but different accessories.

DEXperience SolutionsOP•9/7/24, 11:09 PM

The least I hope in this experiment is to get to occupy a prompt, that I can pass it for example “Ball on the beach” I hope one of those balls with a background on the beach.

DDEXperience Solutions I have 800 different characters, between accessories and country, do you think t...

Furkan Gözükara SECourses•9/7/24, 11:09 PM

as long as they consistent and not repeating i would say use all

Furkan Gözükara SECourses•9/7/24, 11:09 PM

also in your case captions may work better not sure though

DDEXperience Solutions I have 800 different characters, between accessories and country, do you think t...

xlads•9/7/24, 11:10 PM

This is interesting is it an nft project?

Xxlads This is interesting is it an nft project?

DEXperience SolutionsOP•9/7/24, 11:10 PM

yes.

FFurkan Gözükara SECourses as long as they consistent and not repeating i would say use all

DEXperience SolutionsOP•9/7/24, 11:12 PM

Do you recommend using them without background?

DDEXperience Solutions Do you recommend using them without background?

Furkan Gözükara SECourses•9/7/24, 11:14 PM

i say keep dont remove

Furkan Gözükara SECourses•9/7/24, 11:14 PM

it is how you want to generate right?

Furkan Gözükara SECourses•9/7/24, 11:15 PM

wow : https://github.com/lllyasviel/stable-diffusion-webui-forge/issues/1733

GitHub

Why is the flux realism sampler code hidden in a zipped blockly imp...

I was having a hell of a time trying to figure out what the new "flux realism" sampler was (I'm guessing it's similar to comfyui's dpmpp_2s_ancestral that was semi-recently ad...

FFurkan Gözükara SECourses it is how you want to generate right?

DEXperience SolutionsOP•9/7/24, 11:17 PM

I mean, I want the background to be on the FLUX side, if I ask for 'Ball in paris' I would want one of the cartoon with a realistic or non-realistic background of the paris background.

DDEXperience Solutions I mean, I want the background to be on the FLUX side, if I ask for 'Ball in pari...

Furkan Gözükara SECourses•9/7/24, 11:22 PM

not sure try it

xlads•9/7/24, 11:37 PM

After done training I need to download swarmui to generate images right? also what exactly are regularization images for?

Xxlads After done training I need to download swarmui to generate images right? also wh...

Furkan Gözükara SECourses•9/7/24, 11:38 PM

reg images not useful for flux i tested

Furkan Gözükara SECourses•9/7/24, 11:38 PM

you can use swarm ui forge comfy

Furkan Gözükara SECourses•9/7/24, 11:38 PM

i prefer swarm

FFurkan Gözükara SECourses i prefer swarm

xlads•9/7/24, 11:39 PM

https://youtu.be/HKX8_F1Er_w?feature=shared
I will follow this video after training is down just installing part and lora part

YouTubeSECourses

Zero to Hero Stable Diffusion 3 Tutorial with Amazing SwarmUI SD We...

Do not skip any part of this tutorial to master how to use Stable Diffusion 3 (SD3) with the most advanced generative AI open source APP SwarmUI. Automatic1111 SD Web UI or Fooocus are not supporting the #SD3 yet. Therefore, I am starting to make tutorials for SwarmUI as well. #StableSwarmUI is officially developed by the StabilityAI and your m...

Xxlads https://youtu.be/HKX8_F1Er_w?feature=shared I will follow this video after trai...

Furkan Gözükara SECourses•9/7/24, 11:39 PM

yep great one

Luciani•9/7/24, 11:39 PM

so does it not matter if the training images are kinda "high res" like 4000px instead of 1024 on the wide side?

LLuciani so does it not matter if the training images are kinda "high res" like 4000px in...

Furkan Gözükara SECourses•9/7/24, 11:39 PM

it will auto downscale

Furkan Gözükara SECourses•9/7/24, 11:39 PM

to your train res

DEXperience SolutionsOP•9/7/24, 11:41 PM

not available

Luciani•9/7/24, 11:42 PM

I now have unified the aspect ratios -- so now all pics are one of the following aspect ratio: 0.5, 0.666, 0.8, 1, 1.5, 2
I let it train over night and we´ll see -- also I add some more images for training

Furkan Gözükara SECourses•9/7/24, 11:44 PM

pick alt config but at least get 2 GPUs so to get more ram

DDEXperience Solutions not available 😦

Furkan Gözükara SECourses•9/7/24, 11:44 PM

@NicB we need more gpus

DEXperience SolutionsOP•9/7/24, 11:48 PM

96GB RAM, will it be enough?

DEXperience SolutionsOP•9/7/24, 11:49 PM

Coupon code is not valid for this GPU Type and/or Quantity. Please try again.

DDEXperience Solutions 96GB RAM, will it be enough?

Furkan Gözükara SECourses•9/7/24, 11:50 PM

yep greqat

DDEXperience Solutions Coupon code is not valid for this GPU Type and/or Quantity. Please try again.

Furkan Gözükara SECourses•9/7/24, 11:50 PM

only up to 4 :/

Furkan Gözükara SECourses•9/8/24, 12:14 AM

teesting ipndm

Furkan Gözükara SECourses•9/8/24, 12:14 AM

Furkan Gözükara SECourses•9/8/24, 12:15 AM

wow ipndm makes it much more realistic

DEXperience SolutionsOP•9/8/24, 1:07 AM

@Dr. Furkan Gözükara in my case instance Prompt is the keyword trigger?
but in class prompt what do I write? it is not a person

DEXperience SolutionsOP•9/8/24, 1:08 AM

I will try a first test using:
Rank_1_FAST_4x_GPU

DDEXperience Solutions @Dr. Furkan Gözükara in my case instance Prompt is the keyword trigger? but in ...

Furkan Gözükara SECourses•9/8/24, 1:12 AM

use joy captions and like chatgpt

Furkan Gözükara SECourses•9/8/24, 1:12 AM

to get your images described

Furkan Gözükara SECourses•9/8/24, 1:12 AM

from there get idea

Furkan Gözükara SECourses•9/8/24, 1:12 AM

like i used 3d rendering

Furkan Gözükara SECourses•9/8/24, 1:12 AM

rank 1 fast is good

Furkan Gözükara SECourses•9/8/24, 1:16 AM

lol i publish thousands of dollars worth article, lots of research info and still some people complains like why i use lora rank 128 instead of 32

you can use if you wish

DEXperience SolutionsOP•9/8/24, 1:24 AM

According to chatgpt it tells me that my image is a “Countryball”.
but using some caption models trained in replicate or huggingface it tells me it's a “cartoon illustration”.

DDEXperience Solutions According to chatgpt it tells me that my image is a “Countryball”. but using som...

Furkan Gözükara SECourses•9/8/24, 1:24 AM

try both on flux

DEXperience SolutionsOP•9/8/24, 1:24 AM

Furkan Gözükara SECourses•9/8/24, 1:24 AM

and see which is more accurate generating

Furkan Gözükara SECourses•9/8/24, 1:24 AM

cartoon illustration looks more accurate

DEXperience SolutionsOP•9/8/24, 1:25 AM

is it necessary to describe that it is in 2D?

DEXperience SolutionsOP•9/8/24, 1:29 AM

I have a doubt, and excuse my ignorance....
but when using the prompt is it necessary to write “cbznft cartoon illustration”? i.e. use both?
or am I wrong?

DDEXperience Solutions I have a doubt, and excuse my ignorance.... but when using the prompt is it nec...

Furkan Gözükara SECourses•9/8/24, 1:31 AM

ye like this

Luciani•9/8/24, 1:36 AM

stupid question: should or can training images also contain things to learn which are rare? like 98% of all the training images of the same person have blue eyes -- and 2% have green eyes (like they wearing colored contact lenses)?

Luciani•9/8/24, 1:38 AM

another example -- how the person looks like from "behind"

LLuciani stupid question: should or can training images also contain things to learn whic...

Furkan Gözükara SECourses•9/8/24, 1:39 AM

they should contain images

I have 800 different characters, between accessories and country, do you think that training with th

Similar Threads