Project AI

Vla Evaluation Harness

Releases tracked

31 days

Avg cadence

Jun 2026

First tracked

22d ago

Latest

Release history

Shared episode scheduling; Clearer shard failures; More reliable RoboTwin and container execution

More reliable recording and concurrent container runs, with CPU rendering for additional benchmarks

Recording on; LeRobot bridge ships with four reproduced LIBERO scores and reproducible Docker builds

Adds SQLite recording and experiment tracking for observability workflows