New Supir Workflow ComfyUI
Vložit
- čas přidán 19. 03. 2024
- New Supir Workflow ComfyUI with split nodes
Download workflow:
drive.google.com/file/d/1GWLm...
Supir Github:
github.com/kijai/ComfyUI-SUPIR
Join and Support me ###
Support me on Patreon:
/ aifuzz
Let’s be Instagram friends:
/ aifuzz1
Discord
This is AI voice, that's very incredible :O !
Where is the color match node?
Comfy just crashes without error message it appears: [rgthree] Using rgthree's optimized recursive execution.
[rgthree] First run patching recursive_output_delete_if_changed and recursive_will_execute.
[rgthree] Note: If execution seems broken due to forward ComfyUI changes, you can disable the optimization from rgthree settings in ComfyUI.
Diffusion using fp16..... and reconnect button appears. Any advise?
Yo, nice content, looking for an upscaler, algorithm delivers. What's that slick comfyui theme btw ?
I think it’s “GitHub” under the themes
What settings makes it render the fastest?
Missing notes, ColorMatch and ImageResize
I wonder if anyone told you that. You have the most beatiful voice. ❤
Aww thank you sweetie
0:50 - building the workflow
Is there any way to make this run faster? Took me 3 hours to upscale one image with a 2070
One more issue I noticed was that, when I upscaled the image of a person, the face and hair were upscaled but the background remained blurry
Unfortunately not much. Supir is meant to run on a lot of VRAM. Remember when it originally came out a month ago it required MINIMUM 32GB VRAM. They got it down to 12 now. But it’s a memory hog.
I guess that’s how. Run with more vram.
Was wondering how you got straight nodes connect lines instead of default "noodles"?
It's in the settings under "Link Render Mode"
@@boleslawciesielski6199 Thanks👍👍
@@boleslawciesielski6199 I did not know either, now I know 😊 thanks.
So it's definitely an AI voice.. But I feel it could be an AI Voice filter... Are you using Voicemod by any chance?
90% sure its rvc
No voice filter or voice mod
How to Install Supir into Comfyui?
works only with gpu
Yep. With a strong gpu unfortunately
wow thats impressive text to speech/ speech to speech voice actually hard to differentiate
It's Ai voice?😐
@@DJVARAO yes indeed lol
😊
@@AIFuzz59 Was it really an AI voice?! I can't even believe now.
@@AIFuzz59 Can you please make a video on which models you used to synthesize the audio for this video?
how much VRAM for SUPIR?
it used to be like 30GB
Yes it’s much lower now
4070 12GB crashes I have tested comfy local and standalone.@@AIFuzz59
Yeah I think 12 minimum now. But it eats a lot. I’m running 24 and whenever I run SUPIR it eats up around 20+