CogVideoX
CogVideoX is an open-source video production model that supports video from text, video continuation, and video production tasks from visual. The project includes inference and fine-tuning components such as CogVideoX-5B, CogVideoX1.5 and CogKit. It offers infrastructures that optimize GPU memory usage with Lora fine-tuning code based on the Diffusers library and support for DDIM Inverse.
CO
THUDM · Video & Media Production
STARS12
FORKS1
COMMITS454
CONTRIBUTORS40
VERSIONv1.0
UPDATED12 mo ago
PythonAI VideoDitApache-2.0
Common commercial products in this category: Adobe Premiere Pro, Vimeo, Final Cut Pro
What can it do?
- You can create videos using text inputs and prompt words.
- You can use a given image as a background entry to create video from the visual.
- You can run the Lora fine-tuning code through the Diffusers version.
My note
Alternative pages
More in: Video & Media Production
- supervision★ 51,438 · today
- Video.js★ 40,050 · today
- Video Use (Browser Use)★ 30,200 · 22 days ago
- SRS★ 29,401 · today
- memvid★ 17,322 · 3 mo ago
- filmcraft★ 16,978 · today