multitask-dit-robosuite-wrist-diffusion-064000

LeRobot Multi-task DiT policy trained on robosuite Lift, PickPlaceCan, and NutAssemblySquare with agentview and wrist camera inputs.

Source checkpoint: /home/anthonyliang/Documents/rfm_rl/outputs/multitask_dit_robosuite_wrist_diffusion_gpu0/checkpoints/064000/pretrained_model

Rollout Eval

Overall success rate: 0.9555555555555556 over 90 episodes.

task	episodes	success_rate	avg_return	avg_steps
can	30	1.0	26.727644266768134	105.33333333333333
lift	30	1.0	11.27891640541154	48.13333333333333
square	30	0.8666666666666667	45.49792050940882	189.7

Loading

Use this as a standard LeRobot pretrained policy checkpoint.

from lerobot.configs import PreTrainedConfig
from lerobot.policies import get_policy_class

cfg = PreTrainedConfig.from_pretrained("TRI-ML/multitask-dit-robosuite-wrist-diffusion-064000")
policy = get_policy_class(cfg.type).from_pretrained("{repo_id}", config=cfg)