caio-freitas / GraphDiffusionImitate

Diffusion-based graph generative policies for imitation learning in robotics tasks 🧠🤖
MIT License
1 stars 1 forks source link

Diffusion #31

Closed caio-freitas closed 1 year ago

caio-freitas commented 1 year ago

Fully adapts diffusion_policy to be interchangeable among other policies, requiring changes on

caio-freitas commented 1 year ago

The environments have to be expanded to support for the keys obs_horizon, action_horizon and pred_horizon to be ran with diffusion_policy