LIBERO-Object

seq init libero-object gives the stable version, seq init libero-object@beta the beta, and seq init libero-object@v<N> any version below that is not withdrawn. A benchmark registered from this template names the version it runs in seq benchmark ls; seq benchmark upgrade moves it to the newer stable one.

1. v12026-10-17stablebeta

   LIBERO's libero_object suite as a benchmark of its own, each task listed with its instruction.

   Needs sequence-ai 0.37.0 or later.

   • seq init libero-object my-libero-object; it registers as libero-object, and libero-object-cpu on CPUs.
   • --param eval.max_steps=N sets how many steps an episode may take.
   • libero-spatial, libero-object, libero-goal and libero-10 are LIBERO's four suites, a template each.
   Use v1

   libero-object v1shell

   # start a benchmark from this version, then register it
   seq init libero-object@v1 my-benchmark
   seq deploy my-benchmark/benchmark.py
   # its CPU version, on CPU machines
   seq deploy my-benchmark/benchmark_cpu.py
   
   # or move a benchmark registered from this template to it
   seq benchmark upgrade my-benchmark --to v1

   In the console, New benchmark registers this version: the console registers the stable one.