Picsart-AI-Research / Text2Video-Zero

[ICCV 2023 Oral] Text-to-Image Diffusion Models are Zero-Shot Video Generators
https://text2video-zero.github.io/
Other
3.91k stars 336 forks source link

Not reproducible #73

Closed robotNo201 closed 4 months ago

robotNo201 commented 8 months ago

As a developer, I would like to convince u that this repository is not reproducible on most devices. I use A100 gpu with at least 60 gb graphic memory, still out of memory, if u working on a device better than that, you can try to make prompt's size equal to 1, which means u can only put a word in the prompt, that will reduced graphic memory used, if luckily, you might get the result.