Feature	Type	Shape
`observation.images.robot0_eye_in_hand`	VISUAL	`(3, 256, 256)`
`observation.images.robot0_agentview_left`	VISUAL	`(3, 256, 256)`
`observation.images.robot0_agentview_right`	VISUAL	`(3, 256, 256)`
`observation.state`	STATE	`(16,)`

Outputs

Feature	Type	Shape
`action`	ACTION	`(12,)`

Training Dataset

Repository: pepijn223/atomic_4_annotated
Episodes: 2031
Frames: 559517
Frame rate: 20 FPS
Task(s): "Close the lid blender by securely placing the lid on top.", "CloseBlenderLid", "Close the fridge doors.", "Close the fridge door.", "CloseFridge", "Open the right drawer.", "Open the left drawer.", "OpenDrawer", "Fully slide the top dishwasher rack in.", "Fully slide the top dishwasher rack out.", "SlideDishwasherRack"

Training Configuration

Setting	Value
Training steps	12000
Batch size	24
Optimizer	adamw
Learning rate	2.5e-05
Seed	1001
LeRobot version	0.5.2

How to Get Started with the Model

New to LeRobot? These guides cover the full workflow:

Install LeRobot — set up the lerobot package.
Hardware setup — assemble, wire, and calibrate your robot and cameras.
Record data & train a policy — the end-to-end imitation-learning walkthrough.
CLI cheat-sheet — quick reference for the lerobot-* commands.

The short version to run and train this policy:

Run the policy on your robot

bash

lerobot-rollout \
  --strategy.type=base \
  --robot.type=PandaOmron \
  --robot.port=<your_robot_port> \
  --robot.cameras="{ <camera_1>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}, <camera_2>: {type: opencv, index_or_path: <index_or_path>, width: 640, height: 480, fps: 30}}" \
  --policy.path=pepijn223/pi052_atomic4_v2_01_baseline_seed1001 \
  --task="Close the lid blender by securely placing the lid on top." \
  --duration=60

Replace the remaining <...> placeholders with your own values: --robot.port and the camera names/indices are specific to your machine, and the camera names must match the observation keys this policy was trained on.

When --strategy.type=base is used the script doesn't record the episodes. Skipping duration will make the policy run indefinitely. For more information look at rollout documentation.

Train your own policy

bash

lerobot-train \
  --dataset.repo_id=${HF_USER}/<dataset> \
  --policy.type=pi052 \
  --output_dir=outputs/train/<policy_repo_id> \
  --job_name=lerobot_training \
  --policy.device=cuda \
  --policy.repo_id=${HF_USER}/<policy_repo_id> \
  --wandb.enable=true

Writes checkpoints to outputs/train/<policy_repo_id>/checkpoints/.

Evaluation

No evaluation results have been provided for this policy yet.

Citation

If you use this policy, please cite the method linked in the description above, along with LeRobot:

bibtex

@misc{cadene2024lerobot,
    author = {Cadene, Remi and Alibert, Simon and Soare, Alexander and Gallouedec, Quentin and Zouitine, Adil and Palma, Steven and Kooijmans, Pepijn and Aractingi, Michel and Shukor, Mustafa and Aubakirova, Dana and Russi, Martino and Capuano, Francesco and Pascal, Caroline and Choghari, Jade and Moss, Jess and Wolf, Thomas},
    title = {LeRobot: State-of-the-art Machine Learning for Real-World Robotics in Pytorch},
    howpublished = "\url{https://github.com/huggingface/lerobot}",
    year = {2024}
}