Evidentiary-operatingmicroscope154
- 1 skill
- 0 followers
- 23 hours ago last updated
- ▌ Embodied Eval Automation · evidentiary-operatingmicroscope154 bundlePlan, build, run, monitor, validate, transfer, and audit reproducible batch episode collection for policy, VLA, world-model, or hybrid models on embodied benchmarks and simulators. Use when a user provides or wants to connect a local/SSH/cloud GPU host, reuse existing model or benchmark assets, obtain GitHub or Hugging Face resources, compare native and unified episode formats, create adapters and validators, stage single-request through batch runs, recover interrupted jobs, manage disk and GPU limits, or deliver auditable episode datasets and reports.