GR00T-N1.7-AppleToPlate

Dataset Description: The GR00T-N1.7-AppleToPlate dataset is a multimodal collection of trajectories collected on a Unitree G1 humanoid robot. It supports a humanoid (G1) static loco-manipulation task in which the robot picks up an apple and places it on a plate. Each entry provides the full context (state, vision, language, action) needed to train and evaluate generalist robot policies for an apple pick-and-place task. | Dataset Name | # Trajectories | |-----------------------------|-----------------| | G1 Static AppleToPlate Task | 402 | This dataset is ideal for behavior cloning, policy learning, and generalist robotic manipulation research. It has been used for post-training the GR00T N1.7 model. This dataset is ready for commercial or non-commercial uses. Dataset Owner NVIDIA Corporation Dataset Creation Date: 05/20/2026 Version: v1.0 Previous Version(s): No previous version Relationship to Previous Version(s): N/A License/Terms of Use: This dataset is governed by the Creative Commons Attribution 4.0 International License (CC-BY-4.0). Intended Usage: This dataset is intended for: Training robot manipulation policies using behavior cloning. Research in generalist robotics and task-conditioned agents. Real-world manipulation and imitation learning research. Dataset Characterization: Data Collection Method Manually-Collected All demonstrations were manually collected through human teleoperation on a physical G1 robot using an XR headset. Each demo was recorded at 30 Hz. Labeling Method Manually-Labeled Dataset Format: The dataset is provided in LeRobot format, organized into the following directories: `data/` contains the time-indexed trajectories as parquet files. `videos/` contains the recorded RGB videos in mp4 format. `meta/` contains the dataset metadata. Each demo in GR00T-Lerobot datasets consists of a time-indexed sequence of the following modalities: Actions action (FP32): joint desired positions for all body joints (43 DoF) Observations observation.state