Originally shared on GitHub by guoyww
Learn about how to run this model to create animated images on GitHub.
Choose the version that aligns with the version your desired model was based on.
Description
FAQ
Comments (32)
although im still unable to run this I'm still happy to see technologies like this becoming more accessible 😍
amazing!! can it run on 8GB VRAM??
12gb with xformers on linux
60gb on windows
@DorotaLuna wrong, im on windows and running it perfectly fine.
@boyetosekuji how did you set it up on windows?
@Brouz it was straight forward as mentioned in the README, what problems are you having
@boyetosekuji I had problems with the install on windows since it's using conda and xformers cannot be installed with conda on windows. Also installing with the environment file caused torch to be installed with CPU rather than GPU. In the end I just used a normal python venv and installed everything manually.
@boyetosekuji What's your VRAM amount on Windows?
Yes, it can
Can you please add all the presets? On the main repo page, they are currently not downloadable from google drive due to being rate limited
Which presets are you talking about? I haven't personally had a chance to play with this yet.
Installation instructions seem to be for linux? How would you go about installing this on windows, if possible?
you would go to your command prompt and clone the repo there. Coding knowledge is helpful here
Hi there! I have a guide here which I'll be updating over time - https://education.civitai.com/beginners-guide-to-animatediff/
I'm interested in trying it, but honestly, I got a little confused in the installation github... could you help me a little? I am using VAST to run 1111
For anyone downloading this file you must rename it to "mm_sd_v14.ckpt", or "mm_sd_v15.ckpt" depending on which one you get. As it downloads with a different name. Uploader please change file name if possible.
Awesome model by the way, thanks!
谢谢
What happens if it stays the downloaded name? Is this potentially creating a problem with function? I am testing to see if changing the name fixed any issue I have currently.
Make video guides how, do you get the animation so smooth? I've tried it and it's just rubbish. There are no settings and recommendations in the app! After the first generation, you can not run the second, so SD Automatic 11111 just hangs and only reloading resumes its work - fix the bug.
How about instead of being entitled, you contribute to the project and try fixing the bugs yourself. These people work hard on these things and post them for free, then entitled pricks like you come along and bitch that shit didn't work. I have it running in a Google Colab, no problems, and even moved it to my Google Drive so all my files are saved.
PEBCAK
I'll ask his question in the nicer way, Hi I know nothing about this extension except that I can't seem to get it to do anything. Does anyone have any idea where to find a how to make it work, not a how to install and place the chkpt file for it, because anyone who uses automatic 1111 knows how to do that but how to actually use the extension to do something? I run it with the settings I saw in a youtube video and nothing happens. The command liine shows it trying to do something but then it just hangs at 0%. Mostly people that are asking for help are not able to make extensions and they just want help to get it working like you have it. We need your help and appreciate your help. Thank you. I like the extension and WANT to use it but something is stopping it for me.
error,ImportError: Imageio Pillow requires Pillow, not PIL!
pip install pillow. pillow is PIL
I take it then that this can't be used in ComfyUI?
It do work well in ComfyUI
Install at manager, then download the model rename it as mm_sd_v15.ckpt.
at ComfyUI\custom_nodes\comfyui-animatediff\models
restart comfyui
@kuma660224 There are two problems with comfyui, one is that the picture won't move, and the other is that the 16 pictures are split into two different scenes, 8 pictures each~~weird!
it can
All the flickering ones here are made with SD webui extension, use the standalone version its better.
The extension was fixed, it's fine now