跳至正文
OpenAgentSkill
浏览已开放 · 目录记录日期 2026-11-21 · 自动更新和账户服务暂时关闭。
开发与测试

robosuite-evosearch

Run Robosuite Fix Loop + Evolutionary Search with K=8 candidate search on seeds 101-125 and selected-code evaluation on seeds 1-100.

NVlabs · 已有 AI 审核记录

安装与使用

运行命令前先阅读来源与依赖说明。复制命令不会自动安装或执行。

npx skills add NVlabs/ASPIRE --skill robosuite-evosearch

技能说明


name: libero-evosearch description: "Run the LIBERO-Pro Fix Loop + Evolutionary Search experiment: iterative candidate search for low-performing tasks, validation selection, and final held-out evaluation."

LIBERO Fix Loop + Evolutionary Search

Use this skill for Experiment 2. Start with INSTRUCTIONS.md.

Run Order

  1. Start with INSTRUCTIONS.md.
  2. Confirm the Fix Loop experiment has produced fix_code.py for the target tasks.
  3. Follow main-agent-prompt.md as the coordinator.
  4. Fill subagent-prompt.md once per task below the success threshold and dispatch subagents on the available worker GPUs.
  5. Read experiment-specific companion skills under skills/ when generating candidates.
  6. Use clean-task-slate.md before rerunning a task or suite.

Files

FilePurpose
INSTRUCTIONS.mdEnd-to-end human/agent entrypoint
main-agent-prompt.mdCoordinator runbook for Evolutionary Search task dispatch and result collection
subagent-prompt.mdPer-task Evolutionary Search iteration prompt
clean-task-slate.mdReset checklist before reruns
skills/evosearch-iteration.mdCandidate/eval/keyframe iteration mechanics
../api-reference.mdLIBERO reduced API reference
../skills/Shared LIBERO robot skills