跳转到主要内容
此内容尚未提供您的语言版本,正在以英文显示。

Evaluating Cosmos Policy

技能 已验证 活跃

Evaluates NVIDIA Cosmos Policy on LIBERO and RoboCasa simulation environments. Use when setting up cosmos-policy for robot manipulation evaluation, running headless GPU evaluations with EGL rendering, or profiling inference latency on cluster or local GPU machines.

目的

To enable researchers and engineers to effectively set up, run, and profile NVIDIA Cosmos Policy evaluations in LIBERO and RoboCasa simulation environments.

功能

  • Evaluates NVIDIA Cosmos Policy on LIBERO and RoboCasa.
  • Supports headless GPU evaluation with EGL rendering.
  • Profiles inference latency and throughput.
  • Provides detailed setup and execution workflows.
  • Includes troubleshooting for common issues.

使用场景

  • Setting up cosmos-policy for robot manipulation evaluation.
  • Running headless GPU evaluations with EGL rendering.
  • Profiling inference latency on cluster or local GPU machines.

非目标

  • Training or fine-tuning Cosmos Policy models.
  • Working with physical robots or real-world evaluation.
  • Evaluating other vision-language-action models.

安装

请先添加 Marketplace

/plugin marketplace add Orchestra-Research/AI-Research-SKILLs
/plugin install AI-Research-SKILLs@ai-research-skills

质量评分

已验证
97 /100
1 day ago 分析

信任信号

最近提交17 days ago
星标8.3k
许可证MIT
状态
查看源代码

类似扩展

OpenVLA OFT Fine Tuning and Evaluation

98

Fine-tunes and evaluates OpenVLA-OFT and OpenVLA-OFT+ policies for robot action generation with continuous action heads, LoRA adaptation, and FiLM conditioning on LIBERO simulation and ALOHA real-world setups. Use when reproducing OpenVLA-OFT paper results, training custom VLA action heads (L1 or diffusion), deploying server-client inference for ALOHA, or debugging normalization, LoRA merge, and cross-GPU issues.

技能
Orchestra-Research

Context Compression

100

This skill should be used when the user asks to "compress context", "summarize conversation history", "implement compaction", "reduce token usage", or mentions context compression, structured summarization, tokens-per-task optimization, or long-running agent sessions exceeding context limits.

技能
muratcankoylan

Sentry React SDK

100

Full Sentry SDK setup for React. Use when asked to "add Sentry to React", "install @sentry/react", or configure error monitoring, tracing, session replay, profiling, or logging for React applications. Supports React 16+, React Router v5-v7 non-framework mode, TanStack Router, Redux, Vite, and webpack.

技能
getsentry

Sentry React Router Framework Sdk

100

Full Sentry SDK setup for React Router Framework mode. Use when asked to "add Sentry to React Router Framework", "install @sentry/react-router", or configure error monitoring, tracing, profiling, session replay, logs, or user feedback for a React Router v7 framework app.

技能
getsentry

Sentry Python SDK

100

Full Sentry SDK setup for Python. Use when asked to "add Sentry to Python", "install sentry-sdk", "setup Sentry in Python", or configure error monitoring, tracing, profiling, logging, metrics, crons, or AI monitoring for Python applications. Supports Django, Flask, FastAPI, Celery, Starlette, AIOHTTP, Tornado, and more.

技能
getsentry

Simulate Cpu Architecture

100

Design and simulate a minimal CPU from scratch: define an instruction set architecture (ISA), build the datapath (ALU, register file, program counter, memory interface), design the control unit (hardwired or microprogrammed), implement the fetch-decode-execute cycle, and verify by tracing a small program clock cycle by clock cycle. The capstone "computer inside a computer" exercise that composes combinational and sequential building blocks into a complete processor.

技能
pjt222