search

robotgradient / grasp_diffusion

Pytorch implementation of diffusion models on Lie Groups for 6D grasp pose generation https://sites.google.com/view/se3dif/home
MIT License
234 stars 24 forks source link

Ways to control # of sampled grasps #9

Closed kidpaul94 closed 11 months ago

kidpaul94 commented 11 months ago

Hi authors,

Thank you for sharing the great work!

I wonder whether there is a way to specify the number of grasps generated from a point cloud input in the grasp_diffusion? It seems like there is a parameter called "n_grasps" as shown below, but it is never used in the sample code.

https://github.com/TheCamusean/grasp_diffusion/blob/59b633c5bdfec1c9c5505ea0f4f8d963beeed252/scripts/sample/generate_pointcloud_6d_grasp_poses.py#L99

FYI, pip install theseus-ai did not cause any error when I prepared my environment for grasp_diffusion in conda.

Best,

kidpaul94 commented 11 months ago

Never mind, it seems like https://github.com/TheCamusean/grasp_diffusion/issues/4#issue-1428288856 is a way do to.