CLARU
LeWAM

LeWAM PointMaze Large

Simulation training data for LeWAM, consisting of 1,000 episodes with 1,000,000 frames collected from a PointMaze environment with 2D action space.

Downloads17
Episodes1,000

Why This Matters for Physical AI

This dataset provides large-scale simulated navigation data for training reinforcement learning agents, supporting research in spatial reasoning and maze-solving policies.

Technical Profile

Modalities
rgb
Action Space
2D continuous
Environment
simulation
Task Types
navigation
Episodes
1,000
Data Format
HDF5
Part of the LeWAM family

Access

View on HuggingFace

Need custom rgb data?

Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.

Request a Sample Pack

Related Datasets

LeWAM Eval Tape Stack Diffusion Wide

Real-robot evaluation rollouts of a Diffusion Policy on a Rebot B601 7-DoF arm performing tape stacking tasks. Dataset contains 30 policy rollouts (not teleoperated demonstrations) scored by human operators with 21,299 frames recorded at 30 fps.

rgbproprioception
52 downloadsOct 2026apache-2.0

LeWAM Evaluation Tape Stack PI05 Frozen

Real-robot evaluation rollouts of policy π0.5 on a Rebot B601 7-DoF arm performing tape stacking tasks, with 30 episodes scored by human operators achieving 30% success rate.

rgbproprioception
50 downloadsOct 2026apache-2.0

LeWAM Evaluation Pick-and-Place Diffusion Wide

Real-robot evaluation rollouts of a diffusion policy performing pick-and-place tasks on a Rebot B601 7-DoF arm, with 30 episodes achieving 63.3% success rate.

rgbproprioception
48 downloadsOct 2026apache-2.0

LeWAM Evaluation PushT π0.5 Frozen

Real-robot evaluation rollouts of the π0.5 policy (with frozen VLM backbone) on a Rebot B601 7-DoF arm performing non-prehensile PushT tasks. Contains 30 policy rollouts scored by human operators with a 10% success rate.

rgbproprioception
48 downloadsOct 2026apache-2.0

LeWAM Eval PushT Diffusion Wide

Real-robot evaluation rollouts of a Diffusion Policy on a Rebot B601 7-DoF arm performing non-prehensile PushT tasks, with human-annotated success labels.

rgbproprioception
48 downloadsOct 2026apache-2.0

LeWAM Eval PNPT PI05 Frozen

Real-robot evaluation rollouts on a Rebot B601 7-DoF arm performing pick-and-place tasks, with policy rollouts scored by human operators. Contains 30 episodes of policy inference on physical hardware with synchronized multi-camera observations.

rgbproprioception
47 downloadsOct 2026apache-2.0