it was taking too much vram and we have better llava
it was taking too much vram and we have better llava
Boh that's right. What about not using stage 1 at all
Fi am preparing a great cloud vm image. runpod users will love it
Fyep not useful :d
Fi dont know why they implemented it at all
Fit is also a short dirty method nothng more
Fnothing like AI some image manipulation function
Fby the way i added stop batch processing too
Fhopefully will add image comparison slider and restore face
Bhere is what am thinking, chop the source image with a big overlap, upscale as big as I can with my 3090 and stitch them back. using photoshop and mask the seams if I have to
Bbasicly post tiles method
![Dazzastrous [4090]](https://cdn.discordapp.com/avatars/871389071686135838/c4a5f4d81d9eeb18d856295d692e2c07.webp?size=40)
Dtooo much like hard work
Bwell with my 24gig 3090 i can only go to ~4000
Bwhat if I want 16k
Dwhy would u need 16k?
Bprint high res.
Bwhy do camera need 50 mega pixels
Vcomparing megapixels to max upscale model is interesting
Vi havent looked into it but it feels like you could go max upscale to parameters with ai, then move to photoshop to interpolate further. however, are you talking about like highway billboards?
Dhave you seen billboards up close lol
Vright lol, they look like ping pong balls
Dexactly
Bactually I just had a theory and want to see how difficult to pull it off
Vsounds like you are tinkering, welcome that for sure. but good to know
Bone thing i notice is each upscale does distort from original, if each tile is distort, it maybe hard to stitich them back together
Danyone seen the latest sora vids the pug wearing goggles lol
Bwonder if there an easy way to take an image and chop into 4 piece with 25% overlap
Vthe upscale distortion exists, but so does non-distortion upscaliing
Ban automate way!
Bam planing to use SUIR for upscale so there will ne distortion
Vsame, haven't played with it much
Dsora . sota . supir ......wonder what they will name the next thing
Vi'm a sora hater
Dwell thats just Supir
Vsome of these corporate ai platforms feel slimey. i like midjourney though.
BYouTubeGreenBox
Learn how to create high-resolution photos by stitching lower resolution photos (Gigapixel) using PTGUI
This video is a complete walk-through in which we edit a photo from start to finish.
SPECIAL VIDEO OFFER
Get a 10$ restream credit https://restream.grsm.io/Greenbox10
Restream Playlist https://www.youtube.com/watch?v=Sx0JdXr3DNM&list=PLBb...
This video is a complete walk-through in which we edit a photo from start to finish.
SPECIAL VIDEO OFFER
Get a 10$ restream credit https://restream.grsm.io/Greenbox10
Restream Playlist https://www.youtube.com/watch?v=Sx0JdXr3DNM&list=PLBb...
Vi think we may be in the wrong room to chat
Bwhy, am talking about suir
Bbreaking the vram limit on local machine
Dare u suir your in the right room?
Vthis room is very focused i think, concise
Blol
Bsupir oops
Vi suggest moving banter to the #chat room
Vto keep this timeline clean for Q&A
GGreetings, thank you very much for accepting me, an enthusiast from Colombia, I am in a challenging situation, I was asked to animate this character, but I have not had good results, I tried runpod, with comfyui, but I still do not achieve the consistency of the character. Any tips to polish the workflow?
Gthis character was done with Dall-e 3 and it's supposed to be part of a "trailer" with 3 other characters (an astronaut, a Roman legionnaire and a Centurion) with the setting of the Olympics, if any of you are interested in offering your services as a creator, you can write to me, this project is for Movistar Arena, 8K screens pixel Perfect
G
D
K@Dr. Furkan Gözükara Sent a DM regarding private lecturing/tutoring whenever u can check it please 
DThe cuDNN backend API uses less memory for many execution plans which should be beneficial for users who cache execution plans.
FP16 and BF16 fused flash attention engine performance has been significantly improved for NVIDIA GPUs:
Speed-up of up to 50% over cuDNN 8.9.7 on Hopper GPUs.
Speed-up of up to 100% over cuDNN 8.9.7 on Ampere GPUs.
FP16 and BF16 fused flash attention engine performance has been significantly improved for NVIDIA GPUs:
Speed-up of up to 50% over cuDNN 8.9.7 on Hopper GPUs.
Speed-up of up to 100% over cuDNN 8.9.7 on Ampere GPUs.
