// Docs

ALOHA HDF5

Convert ALOHA-style HDF5 episodes into LeRobot

ALOHA-style datasets often store one episode per HDF5 file.

import refiner as mdr

pipeline = (
    mdr.read_hdf5(
        "/data/aloha/*.hdf5",
        datasets={
            "action": "action",
            "qpos": "observations/qpos",
            "qvel": "observations/qvel",
            "cam_high": "observations/images/cam_high",
            "cam_left_wrist": "observations/images/cam_left_wrist",
        },
        file_path_column="episode_file",
    )
    .to_robot_rows(
        episode_id_key="episode_file",
        action_key="action",
        state_key="qpos",
        extra_observation_keys=("qvel",),
        video_keys=("cam_high", "cam_left_wrist"),
        fps=50.0,
        robot_type="aloha",
    )
    .write_lerobot("hf://buckets/acme-robotics/aloha-converted")
)

Run a small local test first:

print(pipeline.take(1)[0].num_frames)

Then launch:

pipeline.launch_local(
    name="convert-aloha",
    num_workers=2,
)

Set HF_TOKEN in your local environment before launching so the workers can write to the Hugging Face bucket.