pi0.5-pick

pi0.5-pick is a Pi0.5 policy trained via imitation learning for a tabletop object pickup task.

Usage

lerobot-eval \
  --policy.repo_id=sachin604/pi0.5-pick \
  --policy.device=cuda