Software Engineering Courses (SECourses)•13mo ago

Oh, I know 😦 😦 😦

Oh, I know

DDrocsid Imperator Maximus Oh, I know 😦 😦 😦

Furkan Gözükara SECourses•12/11/24, 11:37 AM

let me make a fresh install

Furkan Gözükara SECourses•12/11/24, 11:38 AM

to tell you exact space

Furkan Gözükara SECourses•12/11/24, 11:38 AM

installing

Drocsid Imperator MaximusOP•12/11/24, 11:40 AM

fadounav•12/11/24, 11:47 AM

@Dr. Furkan Gözükara Hello I can't paste the training files into the folder linked to my computer, even though I have plenty of space. How can I download the elements to my computer? If I go through google drive it will take forever.

Capture_decran_2024-12-11_a_12.45.01_PM.png

fadounav•12/11/24, 12:01 PM

Je crois que c'est ok par hugginfaces @Dr. Furkan Gözükara

DDrocsid Imperator Maximus Oh, I know 😦 😦 😦

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

sorry for delay here

Ffadounav @Dr. Furkan Gözükara Hello I can't paste the training files into the folder link...

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

wont work for big files

Furkan Gözükara SECourses•12/11/24, 12:27 PM

use notebook shared here : https://youtu.be/X5WVZ0NMaTg

YouTubeSECourses

How to Download (wget) Models from CivitAI & Hugging Face (HF) & up...

Notebook updated to V5 with huge improvements. If you are having trouble to download models onto your cloud platforms such as RunPod, Google Colab, Kaggle, Massed Compute or anywhere, this tutorial is made for you. Moreover, you can upload your files / models into your private Hugging Face repository and save them permanently privately and downl...

Furkan Gözükara SECourses•12/11/24, 12:27 PM

it is also updated and now way faster better

Furkan Gözükara SECourses•12/11/24, 12:30 PM

finally completed twitch streaming app it took me huge time

auto downloads and stream our videos from youtube : https://www.twitch.tv/secourses

Twitch

SECourses - Twitch

FLUX Tools Outpainting, Inpainting (Fill), Redux, Depth & Canny Ultimate Tutorial Guide with SwarmUI

mikemenders•12/11/24, 12:36 PM

When I realise that the easiest way was right under my nose all along... Priceless. I thought I'd start the Prodigy sampler without any extra parameters or normalization for 100 images with cosine, see what happens. The end result was a flexible, you could say perfect person-Lora for Pony. After that, I wondered why I had previously set all the extra parameters for Prodigy in the first place and why I hadn't trained without them before?

DDrocsid Imperator Maximus Oh, I know 😦 😦 😦

Kallamamran•12/11/24, 12:38 PM

Why SUPIR? I know it's good, but I'm not sure it's so good it's better than just using tiled upscaling with TT or SDUpscale and flux. For sure I don't think it's worth the disk waste/space...

KKallamamran Why SUPIR? I know it's good, but I'm not sure it's so good it's better than just...

Drocsid Imperator MaximusOP•12/11/24, 12:40 PM

As @Dr. Furkan Gözükara himself mentioned, SUPIRSUPIR is way, way, more faithful to the original texture than any model that came before it. A LoRALoRA just rebuilds what it thinks the texture looks like.

Drocsid Imperator MaximusOP•12/11/24, 12:41 PM

That's why I originally went through the trouble. However, I've since had to delete it 3/4 of the way through due to low disk space, heh.

Drocsid Imperator MaximusOP•12/11/24, 12:41 PM

There might be an alternative, but I don't know of one yet.

Drocsid Imperator MaximusOP•12/11/24, 12:42 PM

Not that JoyCaptionJoyCaption works either... (after another few GBs too)

fadounav•12/11/24, 1:09 PM

I can't find segmentation sort order @Dr. Furkan Gözükara

Capture_decran_2024-12-11_a_2.08.09_PM.png

DDrocsid Imperator Maximus As @Dr. Furkan Gözükara himself mentioned, `SUPIR` is way, way, more faithful to...

Kallamamran•12/11/24, 1:19 PM

SUPIR is just upscaling. It's not magic

Don't get me wrong. It's great... It just takes a shitload of resources though and it's not fast. SUPIR is like every other "upscaler" nowadays only a "upscaler with an upscaler" together with latent diffusion of the input image. Basically like img2img. "Good", sure, "that good"... Well... I don't think so. It still doesn't "know" anything. It's just guessing, so there's no guarantees the end result will be fantastic. If I can get fantastic results with SUPIR, using ~80GB of drive space and a very long time. I'd rather get fantastic results with ComfyUI TT or SDU in a fraction of the time using models I already have installed on my drive

KKallamamran SUPIR is just upscaling. It's not magic 😊 Don't get me wrong. It's great... It ...

Drocsid Imperator MaximusOP•12/11/24, 1:20 PM

So what's your recommendation then for 8GB VRAM lol?

Kallamamran•12/11/24, 1:20 PM

8GB VRAM and SUPIR doesn't marry that good either huh?

Kallamamran•12/11/24, 1:21 PM

That's quite limiting

Kallamamran•12/11/24, 1:21 PM

Hell I'm struggling with 16GB VRAM on my laptop right now and usually I find 24GB to be on the low end as well

KKallamamran SUPIR is just upscaling. It's not magic 😊 Don't get me wrong. It's great... It ...

fadounav•12/11/24, 1:26 PM

very interesting thank you for sharing your experience

KKallamamran 8GB VRAM and SUPIR doesn't marry that good either huh?

Drocsid Imperator MaximusOP•12/11/24, 1:28 PM

No, I mean another model entirely then with somewhat similar results, heh.

Drocsid Imperator MaximusOP•12/11/24, 1:29 PM

I just wasted 60+GB on a non-working SUPIRSUPIR and JoyCaptionJoyCaption

FFurkan Gözükara SECourses sorry for delay here

Drocsid Imperator MaximusOP•12/11/24, 1:52 PM

Thanks. Woah!

Drocsid Imperator MaximusOP•12/11/24, 1:52 PM

Is AuraSRAuraSR any better?

Drocsid Imperator MaximusOP•12/11/24, 1:52 PM

It appears to be faster... But I don't know how much space that one takes...

yassinemenace•12/11/24, 1:58 PM

@Dr. Furkan Gözükara When decreasing ddim and alpha should I increase learning rate 5e-5 for 128 what for example 64 ?

fadounav•12/11/24, 3:41 PM

@Dr. Furkan Gözükara I have 2 observations, I have the impression that the generation is long if I compare it to the one on openart in turbo mode. I took 2 A6000, btw, You said in your video that the comparison between lora and finetuning is like night and day. I wanted to know at what level, because if it's rendering quality, I think it's pretty similar. the first image is made with openart and the second with finetuning , it's even worse in finetuning because I've lost the comic illustration aspect.

fadounav•12/11/24, 3:56 PM