jnogga/

bridge_data_v2_teleop

Teleoperated demonstrations in BridgeData V2. Episodes with inconsistent data or lacking language instructions were discarded, leaving ~86% of all teleoperated episodes. Ported from raw demos_8_17 data at full resolution to LeRobotDataset v3.0 format (0.04 TiB | 0.5k inodes), down from (0.49 TiB | 4883.9k inodes) raw.

Your browser does not support the video tag.





	


	Dataset Structure

Note that the available cameras… See the full description on the dataset page: https://huggingface.co/datasets/jnogga/bridge_data_v2_teleop.