Yes! Got it working on a 4080 super using q4. I noticed it never used more than 12gb vram. Image was kinda blurry but not terrible, I'm just impressed I'm running an open source video model on this PC already, thought it'd be another year or two at least considering I've only got 16gb vram.
Alas not yet, that’s what I was referring to, there’s no image to video yet. But considering it’s open source, I imagine it won’t be long before we have it
This looks awesome, but what I'm here for... is to ask if this type of open model can be installed on our aws stage maker space where we can "rent" GPUs because my machine only has a 4 GB GPU. I feel like you're the guy that can answer this and teach us how to do it if you haven't already. Sorry I'm new to your channel and early in my AI software development bootcamp so I don't know these things yet.
Ya This guy is a pric about that heres the workflow to make iteasy and it's not that much of one... oh its free in the link (join patrion or pay for this post 6 bucks)
Not there yet. I think Frame by frame models will take over soon. As long as one can generate images, they will make quality vids with only 2gb vram. The tech nerds are making this stuff way too complicated. IMO. They used to animate with stop motion and a cell phone. LOL. Plus frame by frame will not ever warp.
Yes! Got it working on a 4080 super using q4. I noticed it never used more than 12gb vram. Image was kinda blurry but not terrible, I'm just impressed I'm running an open source video model on this PC already, thought it'd be another year or two at least considering I've only got 16gb vram.
nice. encouragement lol
Kindly update when it is released for Low VRAM
How low is low for you?
@@EndangeredAI 8 GB
This thing required FOUR H100 96GB VRAM each (requiring 386 GB VRAM in total) to run literally last week🤷♂
@ a week? That’s like a decade in AI time 🫠
Waiting for 6 or 8gb VRAM version, because if it needs 6, you can run it on 8, 10, 11, 12gb vram for sure and at full speed.
can you give the model the initial frame/picture?
Alas not yet, that’s what I was referring to, there’s no image to video yet. But considering it’s open source, I imagine it won’t be long before we have it
This looks awesome, but what I'm here for... is to ask if this type of open model can be installed on our aws stage maker space where we can "rent" GPUs because my machine only has a 4 GB GPU. I feel like you're the guy that can answer this and teach us how to do it if you haven't already. Sorry I'm new to your channel and early in my AI software development bootcamp so I don't know these things yet.
Anyone have the workflow? The video says it's available free via the link in the description, but it's behind a Patreon paywall, is this is an error?
Ya This guy is a pric about that heres the workflow to make iteasy and it's not that much of one... oh its free in the link (join patrion or pay for this post 6 bucks)
👋 hi
Not there yet. I think Frame by frame models will take over soon. As long as one can generate images, they will make quality vids with only 2gb vram. The tech nerds are making this stuff way too complicated. IMO. They used to animate with stop motion and a cell phone. LOL. Plus frame by frame will not ever warp.
First comment. Finally I did it 🎉
Wohoooooo
congratulations lol
It takes 1 hour to generate a video with your recommend default settings(in workflow json) in a RTX 3090 ? how to speedup in the same GPU?
you can't
@@ronbere ooh. And I get milky ghosting like videos. Why?
How long is the video?
How long is the video, and which quantized model are you using?
@@EndangeredAI it's the default length 5 secs
And model is as you recommended
can we use it on pinokio or mimicpc?
Have you eaten at Bodhigreen for lunch today? 😅
I was sitting with my girl at the table next to yours
@@JorgeGisbertSoler omg no way! Not lunch, it was dinner last week! Did my family keep asking you for recommendations?
Bodhigreen restaurant at Alicante, Spain ?
Oh sorry! Yes I was in bodhigreen in Alicante! And the week prior bodhi began in Paris 😝 so many bodhi restaurants
@@rageshantony2182 yes
Id love more vram but id also like a new car.
Hi sir I'm from India, how about 6gb vram? Is that possible or im just pathetic 😅
You’re better off using rubpod
Is there any technical barriers preventing me from running this on a macbook pro with 32GB unified memory?
What's with the bad background music?