Software Engineering Courses (SECourses)•2y ago

I was talking about @Dr. Furkan Gözükara’s optimizations, or this public post : https://www.reddit.c

I was talking about @Dr. Furkan Gözükara’s optimizations, or this public post : https://www.reddit.com/r/StableDiffusion/s/aK1PDlYVHP

From the StableDiffusion community on Reddit: SUPIR (Super Resoluti...

Explore this post and more from the StableDiffusion community

TTacite I was talking about @Dr. Furkan Gözükara’s optimizations, or this public post : ...

VAMP•3/1/24, 2:29 PM

Can someone make it more efficient.... I only have 8GB vram..but I do have 64gb RAM....

TTacite I was talking about @Dr. Furkan Gözükara’s optimizations, or this public post : ...

AIGambino•3/1/24, 2:30 PM

where is the optimizations that reduces vram useage on there? is it the --use_tile_vae?

TaciteOP•3/1/24, 2:38 PM

Loading half params + tile vae + no llava

VVAMP Can someone make it more efficient.... I only have 8GB vram..but I do have 64gb ...

TaciteOP•3/1/24, 2:38 PM

NVidia card ?

TTacite NVidia card ?

VAMP•3/1/24, 2:38 PM

Yes rtx 3070ti

TTacite NVidia card ?

VAMP•3/1/24, 2:39 PM

Amd right now is super shit . Nvidia is much more easy and convenient for AI purpose..

TaciteOP•3/1/24, 2:39 PM

And you have error message ? By default I think when vram is full it uses sysram

TTacite And you have error message ? By default I think when vram is full it uses sysram

VAMP•3/1/24, 2:39 PM

I haven't tried it yet .

VAMP•3/1/24, 2:39 PM

As it said it needed minimum of 12gb vram

TaciteOP•3/1/24, 3:07 PM

Try and tell us

TTonyHalik what about this error: c:\Supir\SUPIR\venv\lib\site-packages\open_clip\factory....

ts•3/1/24, 3:54 PM

Same problem here! Any ideas to solve it yet?

protector131090•3/1/24, 3:57 PM

Trying to laucn SUPIr and getting RuntimeError: PytorchStreamReader failed reading zip archive: failed finding central directory

KGTachi•3/1/24, 5:02 PM

@Dr. Furkan Gözükara what exactly does the weight_decay=0.01 do?

kareem9000•3/1/24, 5:02 PM

still getting that accelerate not founod error

kareem9000•3/1/24, 5:02 PM

anyone else have that issue in setting up dreambooth?

kareem9000•3/1/24, 5:02 PM

(locally)

VVAMP Can someone make it more efficient.... I only have 8GB vram..but I do have 64gb ...

Pelolisu•3/1/24, 9:52 PM

oww haha but you have how much for RAM? you shared gpu

KGTachi•3/1/24, 10:45 PM

@Dr. Furkan Gözükara is it possible for supir batch, to run both stage one and stage 2, batch only uses stage 2 at the moment.

AIGambino•3/1/24, 11:17 PM

Anyone experimenting with comfyui and supir?

pat._.fenis•3/1/24, 11:29 PM

Anyone here have a specific sdxl model or lora recommendation for making good hands?

Furkan Gözükara SECourses•3/2/24, 1:05 AM

sorry guys i will check messages now

KKGTachi @Dr. Furkan Gözükara is it possible for supir batch, to run both stage one and s...

Furkan Gözükara SECourses•3/2/24, 1:05 AM

sure i will add this hopefully today

Furkan Gözükara SECourses•3/2/24, 1:05 AM

starting to work now

AIGambino•3/2/24, 1:18 AM

https://github.com/layerdiffusion/sd-forge-layerdiffusion.git how do you install this? creates transparent backgrounds

AAIGambino Anyone experimenting with comfyui and supir?

RagnaSaurus•3/2/24, 1:45 AM

Just found this, although I haven't tried it: https://github.com/kijai/ComfyUI-SUPIR

Furkan Gözükara SECourses•3/2/24, 2:05 AM

has anyone solved how SUPIR upscale works? in terms of output resolution?

Furkan Gözükara SECourses•3/2/24, 2:05 AM

for example if i give 1024x1024 and upscale 1x i get 1024x1024 output
then i give 512x512 and do 1x upscale and i get output 1024x1024
then i give 256x256 and do 1x upscale and i get output 1024x1024

AAIGambino https://github.com/layerdiffusion/sd-forge-layerdiffusion.git how do you install...

Furkan Gözükara SECourses•3/2/24, 2:07 AM

this is revoluionary amazing

felixsanz•3/2/24, 2:11 AM

well is not that revolutionary

felixsanz•3/2/24, 2:11 AM

there is AI background removal already

felixsanz•3/2/24, 2:11 AM

https://github.com/Jcd1230/rembg-comfyui-node

felixsanz•3/2/24, 2:11 AM

just pass the result to this node, for example

felixsanz•3/2/24, 2:12 AM

it's just a plugin for sd forge?

Ffelixsanz it's just a plugin for sd forge?

Furkan Gözükara SECourses•3/2/24, 2:12 AM

nope

Furkan Gözükara SECourses•3/2/24, 2:12 AM

read it

Furkan Gözükara SECourses•3/2/24, 2:12 AM

felixsanz•3/2/24, 2:12 AM

i did. what i missed?

Furkan Gözükara SECourses•3/2/24, 2:12 AM

Furkan Gözükara SECourses•3/2/24, 2:13 AM

it generate image without background remove

Dazzastrous [4090]•3/2/24, 2:13 AM

nice

Furkan Gözükara SECourses•3/2/24, 2:13 AM

you cant get this image

Furkan Gözükara SECourses•3/2/24, 2:13 AM

with background remove ever

felixsanz•3/2/24, 2:13 AM

You can see that the native transparent diffusion can process transparent glass, semi-transparent glowing effects, etc, that are not possible with simple background removal methods. Native transparent diffusion also gives you detailed fur, hair, whiskers, and detailed structure like that skeleton.

felixsanz•3/2/24, 2:13 AM

this?

felixsanz•3/2/24, 2:13 AM

oh i see

felixsanz•3/2/24, 2:13 AM

so how the hell this work?

Furkan Gözükara SECourses•3/2/24, 2:13 AM

they trained new models for this

Furkan Gözükara SECourses•3/2/24, 2:13 AM

huge task

felixsanz•3/2/24, 2:13 AM

SD only uses 3 channels, red green and blue

Furkan Gözükara SECourses•3/2/24, 2:13 AM

i bet this will be paper 100%

I was talking about @Dr. Furkan Gözükara’s optimizations, or this public post : https://www.reddit.c

Similar Threads