MMMich's picture
Update README.md
b7b8270 verified
|
Raw History Blame Contribute Delete
2.15 kB
metadata
license: mit
tags:
  - robotics
  - in-context-learning

ICL-oriented-Training-Set

This repository releases the training resources of the paper “In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks”

Training Set

We do not construct our dataset using the full set of RoboTwin tasks, as fragile objects such as bottle are particularly prone to collisions in the dense-object settings designed for ICL, which can interfere with task execution.

The training data is collected via our tailored SimpleICL data pipeline, which is not the official RoboTwin setting. The archive ICL-oriented-training-set.tar.gz is extracted to obtain the training set organized into LeRobot v2.1 directory structure.

Directory structure

After extraction, you will get:

ICL-oriented-training-set/
  conditions/
  data/
  meta/
  videos/
  dataset_stats.json
  • conditions contains the visual conditioning videos for in-context learning, which is a discrepancy with LeRobot v2.1.
  • meta/conditions.json describes the mapping between each training chunk and conditioning video.
  • dataset_stats.json is the statistics file of RoboTwin.

Project

Citation

@article{li2026context,
  title={In-context Robot Learning Made Simple: A Democratized Recipe for Manipulation Tasks},
  author={Li, Minxing and Han, Minghao and Zhao, Weizhi and Wang, Hanwen and Liu, Xiangshuo and Shang, Shuyao and Zhou, Jingxiang and Sun, Mingchao and Pan, Hongyu and Xu, Mu and others},
  journal={arXiv preprint arXiv:2609.38173},
  year={2026}
}

License

This release is a derivative of RoboTwin 2.0 framework for ICL-oriented data collection.

It is released under the MIT license, consistent with the upstream RoboTwin 2.0 release. Please also refer to the original RoboTwin project and dataset for upstream attribution and terms.

RoboTwin repository: https://github.com/RoboTwin-Platform/RoboTwin