lerobot/spaces/benchmark-leaderboard

title, emoji, colorFrom, colorTo, sdk, sdk_version, app_file, pinned, license, short_description

title	emoji	colorFrom	colorTo	sdk	sdk_version	app_file	pinned	license	short_description
LeRobot Benchmark Leaderboard	🤖	yellow	orange	gradio	5.29.0	app.py	false	apache-2.0	Benchmark history for LeRobot policy x benchmark runs

LeRobot Benchmark Leaderboard

This Space reads immutable benchmark rows from a Hugging Face dataset and shows:

• Latest result per policy and benchmark
• Historical trends over time
• Direct links to uploaded eval and config artifacts

Configuration

Set BENCHMARK_RESULTS_REPO in the Space settings if you want to point the UI at a different public dataset. The default is:

• lerobot/benchmark-history