danielyeh2026/pi05-train-iter1

Model Card for pi05

π₀.₅ (Pi05) is a Vision-Language-Action model from Physical Intelligence designed for open-world generalization: it evolves π₀ to generalize to entirely new environments and situations that were never seen during training. The LeRobot implementation is adapted from their open-source OpenPI repository.

This policy has been trained and pushed to the Hub using LeRobot.

Learn how to train and run it in the LeRobot pi05 guide, or browse the full documentation.

---

Model Details

• License: apache-2.0
• Fine-tuned from: lerobot/pi05_base
• Robot type: so_follower
• Cameras: top, wrist

Inputs & Outputs

The policy consumes these observation features and produces these action features.

Inputs

Feature	Type	Shape
observation.state	STATE	(6,)
observation.images.top	VISUAL	(3, 480, 640)
observation.images.wrist	VISUAL	(3, 480, 640)

Outputs

Feature	Type	Shape
action	ACTION	(6,)

Training Dataset

• Repository: HenryZhang/VLAReplica_SFT_data
• Episodes: 501
• Frames: 266240
• Frame rate: 30 FPS
• Task(s): "Put the bread on the red plate.", "Put the bread on the blue plate.", "Put the yellow bowl on the purple coaster.", "Put the red bowl on the green coaster.", "Put the blue bowl on the orange coaster.", "Put the blue bowl on the green coaster.", "Put the red bowl on the purple coaster.", "Stack the red block on the blue block.", "Stack the blue block on the red block.", "Stack the blue block on the yellow block.", "Stack the yellow block on the blue block.", "Stack the red block on the yellow block.", "Fold the pink towel in half.", "Fold the yellow towel in half.", "Open the oven.", "Clean the whiteboard with the whiteboard eraser.", "Pour one shake of pepper on the plate.", "Pour two shakes of pepper on the plate.", "Pour three shakes of pepper on the plate.", "Lift the green bowl one time.", "Lift the green bowl three times.", "Lift the blue bowl one time.", "Lift the blue bowl three times.", "Lift the red bowl one time.", "Press the button one time.", "Press the button three times.", "Collect all the blocks into the pencil box."

Training Configuration

Setting	Value
Training steps	40000
Batch size	8
Optimizer	adamw
Learning rate	2.5e-05
Seed	1000
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

lerobot-rollout \
  --strategy.type=base \
  --robot.type=so_follower \
  --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=pi05-train-iter1 \
  --task="Put the bread on the red plate." \
  --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

This policy type is usually fine-tuned from the pretrained base model lerobot/pi05_base:

lerobot-train \
  --dataset.repo_id=${HF_USER}/<dataset> \
  --policy.path=lerobot/pi05_base \
  --output_dir=outputs/train/<policy_repo_id> \
  --job_name=lerobot_training \
  --policy.device=cuda \
  --policy.repo_id=${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:

@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}
}