GuyTevet/motion-diffusion-model
A PyTorch implementation of a text-conditioned diffusion model that generates 3D human motion from natural language prompts.
Not currently ranked — collecting fresh signals.
star history
This repository provides the official implementation of the Human Motion Diffusion Model (MDM), a denoising diffusion probabilistic model for text-to-motion generation. It trains on human motion data (HumanML3D dataset) and uses CLIP text embeddings as conditioning to generate diverse, natural human motions from text descriptions. The codebase includes the DiP (Diffusion Pipeline) for ultra-fast inference.
Frequently asked
- What is GuyTevet/motion-diffusion-model?
- A PyTorch implementation of a text-conditioned diffusion model that generates 3D human motion from natural language prompts.
- Is motion-diffusion-model open source?
- Yes — GuyTevet/motion-diffusion-model is open source, released under the MIT license.
- What language is motion-diffusion-model written in?
- GuyTevet/motion-diffusion-model is primarily written in Python.
- How popular is motion-diffusion-model?
- GuyTevet/motion-diffusion-model has 4.1k stars on GitHub.
- Where can I find motion-diffusion-model?
- GuyTevet/motion-diffusion-model is on GitHub at https://github.com/GuyTevet/motion-diffusion-model.