Motion Diffusion Model – Text-to-Motion (50-step CPU demo)

Enter a natural-language prompt and receive a 3-D skeletal animation in BVH format.