Contents Menu Expand Light mode Dark mode Auto light/dark, in light mode Auto light/dark, in dark mode Skip to content
UniLab 0.0.0 documentation
UniLab 0.0.0 documentation
  • UniLab Documentation
    • Getting Started
      • Quick Demo
      • Installation
      • Evaluation and Playback
      • Project Structure
      • FAQ
    • User Guide
      • Training
        • CLI Reference
        • Hydra Config
        • Logging
        • Multi-GPU
        • Resume And Checkpoints
        • Docker
      • Algorithms
        • PPO
        • APPO
        • SAC
        • TD3
        • FlashSAC
        • HIM-PPO
        • HORA
        • MLX PPO
      • Simulation Backends
        • MuJoCo Backend
        • Motrix Backend
        • Choosing a Backend
      • Tasks
        • Locomotion
        • Motion Tracking
        • Manipulation
        • Manip-Loco
      • Domain Randomization
        • Configuration
        • Writing Providers
      • Terrain
        • Procedural Terrain
        • Heightfield Import
      • Tooling
        • ONNX Export
        • W&B and TensorBoard
        • NaN Visualizer
        • Scene Export
        • Robot Import
      • Manipulation Notes
        • Dexterous In-Hand Manipulation
        • Manip-Loco
    • Deployment
      • Sim-to-Real
        • Sim-to-Real Overview
        • G1 Whole-Body Motion Tracking on Hardware
        • Go2 / Go2W Locomotion Deployment
        • Allegro / Sharpa In-Hand Manipulation Deployment
        • ONNX Runtime
        • Domain Randomization for Real-World Transfer
        • Hardware Safety Layers
        • Latency Budget
        • Sim-to-Real Troubleshooting
      • Sim-to-Sim
        • Backend Swap
        • Adding a Backend YAML to an Existing Task
        • Aligning Contact and Friction Between Backends
        • Reward Parity Across Backends
        • Playback and Snapshot Differences
        • Capability Gaps
      • Framework Migration
        • Migrating from Isaac Lab
        • Migrating from Legged Gym
        • Migrating from RSL-RL
        • Migrating from skrl
        • Task Config Translation Cheatsheet
        • Reward Porting
    • Developer Guide
      • Architecture
        • Architecture Overview
        • Runtime Model
        • Layer Boundaries
        • Scene Composition
        • Registry Bootstrap
      • Contracts
        • Env Contract
        • Backend Capability Contract
        • Task Owner Config Contract
        • Domain Randomization Contract
        • Runner Lifecycle
      • Extending
        • Extending UniLab: New Task
        • Extending UniLab: New Backend
        • Extending UniLab: New Algorithm
        • Extending UniLab: New Terrain
      • Contributing To UniLab
      • Collaboration Workflow
      • Agent Quick Reference
      • Motion Asset Migration (Hugging Face)
      • Motrix Contact Sensor Notes
    • Reference
      • API Reference
      • Glossary
      • Changelog
      • Architecture Decision Records
      • Support Matrix
  • UniLab 文档
    • 快速上手
      • 快速演示
      • 安装
      • 评估与回放
      • 项目结构
      • 常见问题
    • 用户指南
      • 训练
        • CLI 参考
        • Hydra 配置
        • 日志
        • 多 GPU
        • 续训与检查点
        • Docker
      • 算法
        • PPO
        • APPO
        • SAC
        • TD3
        • FlashSAC
        • HIM-PPO
        • HORA
        • MLX PPO
      • 仿真后端
        • MuJoCo 后端
        • Motrix 后端
        • 选择后端
      • 任务
        • 运动控制
        • 动作追踪
        • 操作
        • Manip-Loco
      • 域随机化
        • 配置
        • 编写 Provider
      • 地形
        • 程序化地形
        • 高度场导入
      • 工具
        • ONNX 导出
        • W&B 与 TensorBoard
        • NaN 可视化工具
        • 场景导出
        • 机器人导入
      • 操作说明
        • 灵巧手内操作
        • Manip-Loco
    • 部署
      • 仿真到真机
        • 仿真到真机总览
        • 硬件上的 G1 全身运动跟踪
        • Go2 / Go2W 运动部署
        • Allegro / Sharpa 手内操作部署
        • ONNX 运行时
        • 面向真机迁移的域随机化
        • 硬件安全层
        • 延迟预算
        • 仿真到真机故障排查
      • Sim-to-Sim
        • 切换后端
        • 为已有任务添加后端 YAML
        • 在后端之间对齐接触与摩擦
        • 跨后端的 Reward 一致性
        • 回放与快照差异
        • 能力缺口
      • 框架迁移
        • 从 Isaac Lab 迁移
        • 从 Legged Gym 迁移
        • 从 RSL-RL 迁移
        • 从 skrl 迁移
        • 任务配置翻译速查表
        • Reward 移植
    • Developer 指南
      • 架构
        • 架构概览
        • 运行时模型
        • 分层边界
        • 场景组合
        • Registry Bootstrap
      • 契约
        • Env 契约
        • Backend 能力契约
        • 任务 owner config 契约
        • Domain Randomization 契约
        • Runner 生命周期
      • 扩展
        • 扩展 UniLab：新任务
        • 扩展 UniLab：新后端
        • 扩展 UniLab：新算法
        • 扩展 UniLab：新地形
      • 为 UniLab 做贡献
      • 协作工作流
      • Agent 速查
      • Motion 资产迁移指南（Hugging Face）
      • Motrix Contact Sensor 适配笔记
    • 参考
      • API 参考
      • 术语表
      • 变更日志
      • 架构决策记录
      • 后端支持矩阵
Back to top
Copyright © 2026, UniLab Sim Authors
Made with Sphinx and @pradyunsg's Furo