One framework to evaluate any VLA model on any robot simulation benchmark.
This repository also configures its own agents. See what vla-evaluation-harness tells them →
allenai/vla-evaluation-harness
Skill Claude CodeCodex
Add or improve a simulation benchmark in vla-eval, including evaluation protocol, assets, adapter lifecycle, and runtime validation.
allenai/vla-evaluation-harness
Skill Claude CodeCodex
Add or improve a VLA model integration in vla-eval, covering checkpoint setup, inference parity, session state, and serving validation.
allenai/vla-evaluation-harness
Skill Claude CodeCodex needs its repo
Run, debug, and scale VLA evaluations in vla-eval while preserving evaluation conditions and verifying complete results.