domains / families / data / i4h-workflow-train-rl

i4h-workflow-train-rl

separate

Use when training, evaluating, or exporting Workflow policies with online RSL-RL or RLinf, including RL checkpoint and Workflow handoff.

When to use

• Resolve a maintained online-RL profile, verify its Scene/objective/model contracts, run the selected vectorized trainer, evaluate and export its artifact, and hand that artifact to normal Workflow policy validation.

When not to use

• The maintained profiles are `ultrasound_probe_reach` with RSL-RL and `assemble_trocar` with RLinf. RSL-RL resume, RSL-RL evaluation video, and distributed multi-GPU launch are not yet exposed. Trocar currently requires two visible local GP…

Skill page

Source