So101 Smolvla Dataset: Robot Arm Action Data

A dataset created using LeRobot, containing action data for a robotic arm. The dataset's meta information specifies a codebase version of v3.0 and a capture rate of 30 frames per second. It was authored by 'emectronics' and last updated on June 9, 2026.

Use Cases

• Train imitation learning models based on recorded robot joint positions.
• Benchmark reinforcement learning algorithms for robotic manipulation tasks.
• Develop inverse kinematics models using the provided joint position data.
• Simulate and validate robot arm movements in a virtual environment.

Strengths

• Action data includes 6 specific joint positions (shoulder_pan, shoulder_lift, elbow_flex, wrist_flex, wrist_roll, gripper).
• Recorded at a consistent 30 frames per second, suitable for temporal analysis.
• Created with a specified toolchain (LeRobot) and codebase version (v3.0).

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Row count, file formats, and license are unknown, which may limit suitability assessment.
• Column-level documentation beyond the action names is absent; field semantics must be inferred.

Provenance

Source

huggingface

Collection Method

Created using LeRobot.

Freshness

Last updated 2026-06-09 22:39:24; freshness should be verified.