Software Engineering Courses (SECourses)•2y ago

i was told that supir would work if i had enough system ram, so i upgraded to 64 and comfyui is stil

i was told that supir would work if i had enough system ram, so i upgraded to 64 and comfyui is still saying out of memory, do i need to run comfy in low vram mode?

Zono50OP•3/12/24, 1:01 AM

@Furkan Gözükara SECourses

ZZono50 i was told that supir would work if i had enough system ram, so i upgraded to 64...

Furkan Gözükara SECourses•3/12/24, 1:03 AM

who said that?

Furkan Gözükara SECourses•3/12/24, 1:03 AM

yes it will work but if it uses RAM instead of VRAM it will be like 20x slower

Furkan Gözükara SECourses•3/12/24, 1:03 AM

so you have to wait

Furkan Gözükara SECourses•3/12/24, 1:03 AM

what is your VRAM?

Zono50OP•3/12/24, 1:03 AM

12 gb

Zono50OP•3/12/24, 1:03 AM

and 64 gb of system ram

ZZono50 12 gb

Furkan Gözükara SECourses•3/12/24, 1:03 AM

ok if you dont upscale too much

Furkan Gözükara SECourses•3/12/24, 1:03 AM

12 GB will work

Zono50OP•3/12/24, 1:04 AM

but still said oom

Furkan Gözükara SECourses•3/12/24, 1:04 AM

Zono50OP•3/12/24, 1:04 AM

do i need specific arguments when i run comfyui to tell it to use system ram

Furkan Gözükara SECourses•3/12/24, 1:04 AM

i dont know comfyui implementation

Furkan Gözükara SECourses•3/12/24, 1:04 AM

ours works

MrGTAmodsgerman•3/12/24, 7:37 AM

I am wondering if this technique could be used to upscale images by having high res reference photos https://realfill.github.io/
Like having an old image in front of a church but the church still looks the same, so you give it a today taken photo as reference to enhance the photo. Or you want to colorize an photo of the original Titanic, so you give it photos of the movie Titanic where they restored it by hand

RealFill: Reference-Driven Generation for Authentic Image Completion

MMrGTAmodsgerman I am wondering if this technique could be used to upscale images by having high ...

Furkan Gözükara SECourses•3/12/24, 12:07 PM

i think this is generative fill

Furkan Gözükara SECourses•3/12/24, 12:07 PM

controlnet already have this

FFurkan Gözükara SECourses i think this is generative fill

MrGTAmodsgerman•3/12/24, 12:07 PM

Reference based? I don't think so

MMrGTAmodsgerman Reference based? I don't think so

Furkan Gözükara SECourses•3/12/24, 12:11 PM

by reference arent they meaning original input image expanded?

FFurkan Gözükara SECourses by reference arent they meaning original input image expanded?

MrGTAmodsgerman•3/12/24, 12:13 PM

Have you read the paper? It's about training the AI on reference pictures from a simular scene with at least one image as training image to expand or inpaint an target image for a accurate representation.

MrGTAmodsgerman•3/12/24, 12:15 PM

Stable Diffusion plugins just recreate the images or inpaint randomly by it's text input. But RealFill uses actual reference images that are accurate to the target image, hence why it's called RealFill. Like in my mentioned examples above, this could be very useful for restoration.

MrGTAmodsgerman•3/12/24, 12:17 PM

I thought maybe this could be transformed to be used in upscaling by giving a high res reference image from the same scene/subject

MMrGTAmodsgerman Have you read the paper? It's about training the AI on reference pictures from a...

Furkan Gözükara SECourses•3/12/24, 12:18 PM

i didnt read yet but i have a feeling that image is used as prompt

Furkan Gözükara SECourses•3/12/24, 12:18 PM

thus it wont be loyal to input :d

FFurkan Gözükara SECourses thus it wont be loyal to input :d

MrGTAmodsgerman•3/12/24, 12:20 PM

wdym? That's the best case scenario you could have when you are allowed to give the AI the information to make an accurate restoration of an image or video or what ever instead of letting the AI think what it sees.

MrGTAmodsgerman•3/12/24, 12:21 PM

There was a video from TwoMinutePapers about it

MMrGTAmodsgerman There was a video from TwoMinutePapers about it

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

2 minute papers never uses himself just show the demo which are usually very misleading

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

but i can make an auto installer for this

FFurkan Gözükara SECourses 2 minute papers never uses himself just show the demo which are usually very mis...

MrGTAmodsgerman•3/12/24, 12:25 PM

Yea that's why Two minutes

FFurkan Gözükara SECourses but i can make an auto installer for this

MrGTAmodsgerman•3/12/24, 12:26 PM

I thought i mentioned it here because maybe it could be used in SUPIR? Or something else?

MMrGTAmodsgerman I thought i mentioned it here because maybe it could be used in SUPIR? Or someth...

Furkan Gözükara SECourses•3/12/24, 12:26 PM

maybe need testing

Zono50OP•3/12/24, 1:50 PM

is there a better version of xformers for comfyui than 0.0.24?

ZZono50 is there a better version of xformers for comfyui than 0.0.24?

Furkan Gözükara SECourses•3/12/24, 3:14 PM

xformers 0.0.24 is latest

Newlartz1•3/13/24, 7:03 PM

Hi , is there a working model to detect Deepfake images/vids?

NNewlartz1 Hi , is there a working model to detect Deepfake images/vids?

Furkan Gözükara SECourses•3/13/24, 8:45 PM

not that i know

Zono50OP•3/14/24, 1:20 PM

@Furkan Gözükara SECourses Has there been any kind of SUPIR implementation into automatic1111 or webui-forge yet?

ZZono50 <@205854764540362752> Has there been any kind of SUPIR implementation into auto...

Furkan Gözükara SECourses•3/14/24, 2:44 PM

i dont think so

Zono50OP•3/15/24, 6:26 PM

@Furkan Gözükara SECourses Do you know what model i need for controlnet media pipe for sdxl?

Zono50OP•3/15/24, 6:27 PM

I see an sd21 and sd 15 model so far

ZZono50 <@205854764540362752> Do you know what model i need for controlnet media pipe f...

Furkan Gözükara SECourses•3/16/24, 12:09 AM

what was media pipe doing?

Furkan Gözükara SECourses•3/16/24, 12:09 AM

i think my auto download has all the available SDXL controlnet models

Zono50OP•3/16/24, 12:10 AM

Mapped facial. Expressions basically to do non ordinary smiles

ZZono50 Mapped facial. Expressions basically to do non ordinary smiles

Furkan Gözükara SECourses•3/16/24, 12:11 AM

i never used sadly

Dazzastrous [4090]•3/16/24, 12:11 AM

I see facefusion has a new update, however im getting an error - RuntimeError: D:\a_work\1\s\onnxruntime\python\onnxruntime_pybind_state.cc:743 onnxruntime::python::CreateExecutionProviderInstance CUDA_PATH is set but CUDA wasn't able to be loaded. Please install the correct version of CUDA and cuDNN as mentioned in the GPU requirements page (https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html#requirements), make sure they're in the PATH, and that your GPU is supported.

onnxruntime

NVIDIA - CUDA

Instructions to execute ONNX Runtime applications with CUDA

DDazzastrous [4090]I see facefusion has a new update, however im getting an error - RuntimeError: D...

Furkan Gözükara SECourses•3/16/24, 12:12 AM

i am not using at the moment

Dazzastrous [4090]•3/16/24, 12:13 AM

Will updating the cuda and cuDNN to the latest versions break anything?

DDazzastrous [4090]Will updating the cuda and cuDNN to the latest versions break anything?

Furkan Gözükara SECourses•3/16/24, 12:14 AM

probably since it uses onnx

Csókosgéza•3/17/24, 3:33 PM

I would like to ask for some help. Have you known a comfyui AI that can create json files based on a prompt (or/and pic or/and video) that can be dropped into comfyUI? If not, how would it be worthwhile to start making one?

CCsókosgéza I would like to ask for some help. Have you known a comfyui AI that can create ...

Furkan Gözükara SECourses•3/17/24, 9:36 PM

sadly i dont know

iloveh8•3/18/24, 1:38 AM

hi all can i ask any free LLM API we can access? or easy open source model thats easy to deploy?

i was told that supir would work if i had enough system ram, so i upgraded to 64 and comfyui is stil

Similar Threads

Similar Threads

Similar Threads