Gallery · LIBERO-Pro

put the red mug on the plate and put the chocolate pudding to the right of the plate

libero_10_task, task 6 · seed 1. DeepSeek-V4.1-Flash, code mode: 10 decisions, 40 motions, solved.

2× · 114 s of motion
  1. 1 of 10 · step1_mug.py
  2. 2 of 10 · step1b_mug.py
  3. 3 of 10 · mug_grasp2.py
  4. 4 of 10 · mug_grasp3.py
  5. 5 of 10 · mug_grasp4.py
  6. 6 of 10 · mug_place.py
  7. 7 of 10 · pud_grasp.py
  8. 8 of 10 · pud_place.py
  9. 9 of 10 · pud_place2.py
  10. 10 of 10 · pud_lower.py
  1. move_pose
  2. move_pose
  3. set_gripper
  4. move_pose
  5. move_pose
  6. move_pose
  7. set_gripper
  8. set_gripper
  9. move_pose
  10. move_pose
  11. set_gripper
  12. move_pose
  13. move_pose
  14. move_pose
  15. move_pose
  16. set_gripper
  17. move_pose
  18. move_pose
  19. move_pose
  20. move_pose
  21. move_pose
  22. move_pose
  23. set_gripper
  24. move_pose
  25. set_gripper
  26. move_pose
  27. move_pose
  28. move_pose
  29. move_pose
  30. move_pose
  31. set_gripper
  32. move_pose
  33. move_pose
  34. move_pose
  35. move_pose
  36. move_pose
  37. move_pose
  38. move_pose
  39. move_pose
  40. set_gripper
Decisions

The recording is the robot's own cameras at 2×; the time between motions, while the model thinks, is not filmed. The 3D replay is the camera's picture in depth after each decision, with the paths it planned, to turn and step through.