LeRobot: train policy

Official LeRobot python main example for train policy. Inspect the pinned source and upstream setup before use.

lerobot-examples-training-train-policy-py | Record version: source-456639de7cc92465bde9499cc2e5e7186da526529f05f8edc52f7497bd2f8726

Record JSON | Markdown | Pinned upstream source

Access

Fetch the pinned repository, verify this source checksum, and use upstream installation and invocation. This record does not install or execute code.

Hub preparation adapter: false. Follow the upstream setup and invocation contract.

Source and license

Repository: https://github.com/huggingface/lerobot
Revision: 7e241bd630a3719a56157a497ce5d08f244784f1
Path: examples/training/train_policy.py
Source SHA256: 456639de7cc92465bde9499cc2e5e7186da526529f05f8edc52f7497bd2f8726
Original authors: See pinned upstream source and license notices.
License: Apache-2.0
License scope: Selected example source code only; preserve upstream copyright and per-file notices. Robot models, datasets, weights, simulator binaries and third-party dependencies retain their separate licenses.
License SHA256: 0583375a0ec642ec79c447781f34322b0cfa66baff31258214d040f1f6d86d7c
Hub commit: 88022cc5fa2396f47b1f2354e988189dc8e53bea

Requirements

{
  "python": "Python version and optional extras from the pinned LeRobot pyproject.toml",
  "packages": [
    "LeRobot from the pinned v0.6.1 checkout",
    "Task-specific torch, camera, robot and policy extras listed upstream"
  ],
  "display": "Examples may move real robots, access serial ports/cameras, download datasets or checkpoints, and train policies. Configure calibration, device, model and dataset identifiers first.",
  "execution_domain": "training-or-hardware",
  "source_imports": [
    "lerobot.configs",
    "lerobot.datasets",
    "lerobot.policies",
    "lerobot.policies.diffusion",
    "lerobot.utils.feature_utils",
    "pathlib",
    "torch"
  ]
}

Invocation

{
  "entrypoint": "python examples/training/train_policy.py",
  "notes": "Examples may move real robots, access serial ports/cameras, download datasets or checkpoints, and train policies. Configure calibration, device, model and dataset identifiers first. No Hub preparation adapter or measured runtime success is claimed."
}

Explicit compatibility

[]

Missing environment fields remain unknown. Discovery labels do not establish robot compatibility.

Evidence

Level: source_reviewed; runtime verification: not_run.

Fetched and SHA-256 checked this immutable source; statically parsed Python or located native main; verified pinned license bytes and checked explicit per-file license identifiers. No upstream code executed.

Source review and contributor reports do not establish task completion. Preserve upstream attribution and licenses.