diff --git a/docs/en/guide/installation-without-docker.md b/docs/en/guide/installation-without-docker.md
new file mode 100644
index 000000000..2d5382f12
--- /dev/null
+++ b/docs/en/guide/installation-without-docker.md
@@ -0,0 +1,370 @@
+# Installation without Docker
+
+::: warning
+This guide is derived from the official `docker/Dockerfile` and is intended for users who cannot use Docker (e.g., restricted shared servers, HPC environments without container runtime). The exact build steps are complex and version-sensitive — if you encounter compilation errors, please refer to the Dockerfile as the authoritative source and open an issue with your environment details.
+:::
+
+## Overview
+
+Relax relies on several heavyweight CUDA extensions that are built from source with specific commits and patches:
+
+- **Megatron-LM** (via Megatron-Bridge, with a custom patch)
+- **SGLang** (with a custom patch for R3 rollout routing)
+- **flash-attn** (v2 + v3 from a specific commit)
+- **transformer-engine** (v2.14.1)
+- **apex** (from a specific commit)
+- **TransferQueue** (from a specific commit)
+
+The official Docker image (`ghcr.io/redai-infra/relaxrl:latest`) is the recommended and most thoroughly tested installation path. Use this guide only when Docker is unavailable.
+
+## Prerequisites
+
+### Hardware
+
+- NVIDIA GPU with compute capability >= 8.0 (Ampere or newer; Hopper recommended for FA3)
+- At least 24 GB GPU memory for small-model training (Qwen3-0.6B / 4B)
+- At least 50 GB free disk space for source builds, models, and checkpoints
+
+### Software
+
+| Component | Required Version | Notes |
+|-----------|-----------------|-------|
+| OS | Ubuntu 22.04 (LTS) | Other Linux distros may work but are untested |
+| Python | 3.12 | Exact version used in the official image |
+| CUDA Toolkit | 12.9 | Must match PyTorch build |
+| NVIDIA Driver | >= 560 | Supports CUDA 12.9 |
+| GCC / G++ | 11.x | Required for CUDA extension compilation |
+| CMake | >= 3.18 | Required for apex and other builds |
+| Git | >= 2.30 | For cloning dependencies |
+| rsync | latest | For merging Megatron-Bridge sources |
+| jq | latest | For patch verification |
+
+### Environment Variables
+
+Set these before starting:
+
+```bash
+export CUDA_HOME=/usr/local/cuda-12.9
+export PATH=$CUDA_HOME/bin:$PATH
+export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+export MAX_JOBS=8  # Adjust based on your CPU cores; higher = faster build, more RAM
+```
+
+## Step 1: Install System Dependencies
+
+```bash
+sudo apt-get update && sudo apt-get install -y --no-install-recommends \
+    build-essential patchelf tmux net-tools htop nano iputils-ping \
+    autoconf automake libatlas-base-dev libgoogle-glog-dev libbz2-dev libc-ares2 \
+    libleveldb-dev liblmdb-dev libprotobuf-dev libsnappy-dev libtool nasm protobuf-compiler pkg-config unzip sox \
+    libsndfile1 libpng-dev libhdf5-dev gfortran rapidjson-dev ninja-build libedit-dev rsync jq \
+    locales tzdata
+```
+
+## Step 2: Install PyTorch 2.9.0 with CUDA 12.9
+
+The official image uses NVIDIA's PyTorch 2.9.0 build (from the 25.06 NGC container). Install the matching version:
+
+```bash
+pip install torch==2.9.0 torchvision==0.24.0 torchaudio==2.9.0 --index-url https://download.pytorch.org/whl/cu129
+```
+
+Verify:
+
+```bash
+python -c "import torch; print(f'PyTorch {torch.__version__}, CUDA {torch.version.cuda}, GPU: {torch.cuda.get_device_name(0)}')"
+```
+
+## Step 3: Build Core CUDA Extensions
+
+### 3.1 flash-attn v2
+
+```bash
+MAX_JOBS=64 pip -v install flash-attn==2.7.4.post1 --no-build-isolation --no-cache-dir
+pip install --no-cache-dir flash-linear-attention==0.4.1
+pip install --no-cache-dir tilelang -f https://tile-ai.github.io/whl/nightly/cu128/
+```
+
+### 3.2 flash-attn v3 (Hopper only)
+
+::: warning
+This step is only required for Hopper (H100/H800) GPUs. Skip on Ampere/Ada.
+:::
+
+```bash
+cd /opt
+git clone https://github.com/Dao-AILab/flash-attention.git
+cd flash-attention/
+git checkout 0f82fead0b2d59f6390c62a1d74e927159bc5a7b
+git submodule update --init
+cd hopper/
+MAX_JOBS=96 python setup.py install
+
+# Remove stray top-level module that conflicts with TE
+export python_path=$(python -c "import site; print(site.getsitepackages()[0])")
+mkdir -p $python_path/flash_attn_3
+cp flash_attn_interface.py $python_path/flash_attn_3/flash_attn_interface.py
+find $python_path -maxdepth 1 -name flash_attn_interface.py -delete
+find $python_path -maxdepth 2 -path '*.egg/flash_attn_interface.py' -delete
+
+# Verify
+cd / && python -c "import importlib.util as u; assert u.find_spec('flash_attn_interface') is None; import flash_attn_config"
+rm -rf /opt/flash-attention/
+```
+
+### 3.3 transformer-engine
+
+```bash
+pip -v install --no-cache-dir --no-build-isolation "transformer_engine[pytorch]==2.14.1"
+```
+
+### 3.4 torch_memory_saver
+
+```bash
+TMS_CUDA_MAJOR=$(python -c 'import torch; print(torch.version.cuda.split(".")[0])') \
+pip install git+https://github.com/redai-infra/torch_memory_saver.git@afc13785c50119048e2dd8ac497cc9e29ec75bd4 --no-cache-dir --force-reinstall
+```
+
+### 3.5 apex
+
+```bash
+NVCC_APPEND_FLAGS="--threads 32" \
+pip -v install --disable-pip-version-check --no-cache-dir \
+    --no-build-isolation \
+    --config-settings "--build-option=--cpp_ext --cuda_ext --parallel 8" \
+    git+https://github.com/NVIDIA/apex.git@10417aceddd7d5d05d7cbf7b0fc2daad1105f8b4
+```
+
+### 3.6 Additional packages
+
+```bash
+pip install nvidia-modelopt[torch]>=0.37.0 --no-build-isolation --no-cache-dir
+pip install "numpy<2" nvidia-cudnn-cu12==9.16.0.29 --no-cache-dir
+```
+
+## Step 4: Install Megatron-LM via Megatron-Bridge
+
+Relax uses a merged Megatron-LM tree built from Megatron-Bridge sources and its submodule, with a custom patch.
+
+```bash
+export MEGATRON_BRIDGE_COMMIT=2faedbf6fe3c422835a44b2b360cadcb2a116a54
+export MEGATRON_PATH=/root/Megatron-LM  # Adjust to your preferred path
+
+rm -rf $MEGATRON_PATH
+git clone https://github.com/NVIDIA-NeMo/Megatron-Bridge.git /root/Megatron-Bridge
+cd /root/Megatron-Bridge/
+git checkout ${MEGATRON_BRIDGE_COMMIT}
+git submodule update --init --recursive
+./scripts/switch_mcore.sh dev
+
+mkdir -p $MEGATRON_PATH
+cp -r src/megatron $MEGATRON_PATH/
+rsync -avP 3rdparty/Megatron-LM/megatron/ $MEGATRON_PATH/megatron/
+rm -rf /root/Megatron-Bridge
+```
+
+### Apply the megatron patch
+
+```bash
+# Copy the patch from the Relax repository
+cp /path/to/Relax/docker/patch/latest/megatron.patch $MEGATRON_PATH/
+
+cd $MEGATRON_PATH
+patch -p1 < megatron.patch
+
+# Verify patch applied cleanly
+if grep -R -n '^<<<<<<< ' .; then
+    echo "ERROR: Patch failed to apply cleanly. Please resolve conflicts."
+    exit 1
+fi
+rm megatron.patch
+```
+
+### Set PYTHONPATH
+
+```bash
+export PYTHONPATH=$MEGATRON_PATH:$PYTHONPATH
+```
+
+Verify:
+
+```bash
+python -c "import megatron; print(f'Megatron imported from {megatron.__file__}')"
+```
+
+## Step 5: Install SGLang with Custom Patch
+
+Relax uses SGLang v0.5.15.post1 with a custom patch for R3 (rollout routing replay).
+
+::: tip
+The base SGLang installation comes from the `lmsysorg/sglang:v0.5.15.post1-cu129` Docker image. Without Docker, you need to install SGLang from source at the matching version first.
+:::
+
+```bash
+# Install SGLang from source at the matching version
+pip install "sglang[all]==0.5.15.post1"
+
+# Apply the custom sglang patch
+export SGLANG_PATH=$(python -c "import sglang; print(sglang.__path__[0])")
+cp /path/to/Relax/docker/patch/latest/sglang.patch $SGLANG_PATH/
+
+cd $SGLANG_PATH
+git add .
+git update-index --refresh
+git apply sglang.patch --3way
+
+# Verify
+if grep -R -n '^<<<<<<< ' .; then
+    echo "ERROR: SGLang patch failed."
+    exit 1
+fi
+rm sglang.patch
+```
+
+### Install patched sglang-router
+
+The official `sglang-router` wheel drops the `routed_experts` field, breaking R3. Relax uses slime's patched fork:
+
+```bash
+pip install --no-cache-dir --force-reinstall --no-deps \
+    https://github.com/zhuzilin/sgl-router/releases/download/v0.3.2-9daabcd/sglang_router-0.3.2-cp38-abi3-manylinux_2_28_x86_64.whl
+
+# Verify the patched version
+python -c "import sglang_router; assert 'slime' in sglang_router.__version__, sglang_router.__version__"
+```
+
+## Step 6: Install TransferQueue
+
+```bash
+pip install "transferqueue @ git+https://github.com/redai-infra/TransferQueue.git@58054a33834aadbcf76aacd6b1e32e25c030f2c9" --no-deps
+```
+
+## Step 7: Install Python Dependencies and Relax
+
+```bash
+cd /path/to/Relax
+
+# Install pure-Python dependencies
+pip install --ignore-installed PyJWT
+pip install -r requirements.txt --no-cache-dir
+
+# Install additional packages
+pip install --no-cache-dir "compressed_tensors>=0.13.0" tensordict==0.10.0 pyvers==0.1.0 'nvidia-modelopt[hf]==0.44.0' --no-deps
+
+# Install Relax in development mode
+pip install -e .
+```
+
+## Step 8: Build INT4 QAT CUDA Extension
+
+```bash
+cd /path/to/Relax/relax/backends/megatron/kernels/int4_qat
+pip install . --no-build-isolation --no-cache-dir
+```
+
+## Environment Variables Summary
+
+Add these to your `~/.bashrc` or a dedicated environment script:
+
+```bash
+# CUDA
+export CUDA_HOME=/usr/local/cuda-12.9
+export PATH=$CUDA_HOME/bin:$PATH
+export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+
+# Megatron
+export MEGATRON_PATH=/path/to/Megatron-LM
+export PYTHONPATH=$MEGATRON_PATH:$PYTHONPATH
+
+# Relax
+export RELAX=/path/to/Relax
+export PYTHONPATH=$RELAX:$PYTHONPATH
+```
+
+## Verifying the Installation
+
+Run these checks in order:
+
+```bash
+# 1. Basic imports
+python -c "
+import torch
+import ray
+import sglang
+import megatron
+import relax
+print('All basic imports successful')
+print(f'  torch: {torch.__version__}')
+print(f'  CUDA available: {torch.cuda.is_available()}')
+print(f'  GPU: {torch.cuda.get_device_name(0)}')
+"
+
+# 2. CUDA extensions
+python -c "
+import flash_attn
+import transformer_engine
+print('CUDA extensions imported successfully')
+print(f'  flash_attn: {flash_attn.__version__}')
+"
+
+# 3. Relax components
+python -c "
+from relax.controller import Controller
+from relax.registry import ALGOS
+print(f'Registered algorithms: {list(ALGOS.keys())}')
+print('Relax components loaded successfully')
+"
+```
+
+::: warning
+Passing these import checks does not guarantee that full training will work. End-to-end training involves Ray cluster initialization, SGLang engine startup, NCCL communication, and Megatron training loops — each of which may have environment-specific issues. Start with a small model (Qwen3-0.6B) and a few training steps to validate.
+:::
+
+## Troubleshooting
+
+### Compilation errors during flash-attn / apex / TE builds
+
+- Ensure `CUDA_HOME` is set correctly and `nvcc --version` shows 12.9
+- Ensure GCC version is 11.x (`gcc --version`)
+- Reduce `MAX_JOBS` if you run out of RAM during compilation
+- Check that your NVIDIA driver supports CUDA 12.9 (`nvidia-smi`)
+
+### `ModuleNotFoundError: No module named 'megatron'`
+
+- Ensure `PYTHONPATH` includes your `$MEGATRON_PATH`
+- Verify the merged directory exists: `ls $MEGATRON_PATH/megatron/`
+
+### `ModuleNotFoundError: No module named 'sglang'`
+
+- SGLang must be installed before applying the patch
+- Verify: `python -c "import sglang; print(sglang.__version__)"` should show 0.5.15.post1
+
+### NCCL errors or Ray cluster failures
+
+- Ensure `--network=host` equivalent: all nodes must be reachable
+- Set `NCCL_SOCKET_IFNAME` to the correct network interface
+- For single-node training, ensure `ray start --head` can bind to localhost
+
+### Patch conflicts
+
+If `git apply` or `patch` reports conflicts:
+
+1. Check that you have the exact base version (SGLang 0.5.15.post1, Megatron-Bridge commit 2faedbf6)
+2. Try `git apply --3way` for three-way merge
+3. If conflicts persist, refer to the Dockerfile for the exact patch context
+
+## Getting Help
+
+If you encounter issues not covered here:
+
+1. Check the [official Dockerfile](https://github.com/redai-infra/Relax/blob/main/docker/Dockerfile) — it is the authoritative build reference
+2. Search existing [GitHub Issues](https://github.com/redai-infra/Relax/issues)
+3. Open a new issue with:
+   - Your OS, GPU model, NVIDIA driver version, CUDA version
+   - The exact error message and stack trace
+   - The step at which the failure occurred
+   - Whether you are using the exact versions specified in this guide
+
+::: tip
+For most users, the Docker image remains the recommended installation path. This guide is provided as a reference for constrained environments and may not be updated as frequently as the Dockerfile.
+:::
diff --git a/docs/zh/guide/installation-without-docker.md b/docs/zh/guide/installation-without-docker.md
new file mode 100644
index 000000000..52945e18c
--- /dev/null
+++ b/docs/zh/guide/installation-without-docker.md
@@ -0,0 +1,370 @@
+# 无 Docker 环境安装指南
+
+::: warning
+本指南基于官方 `docker/Dockerfile` 整理，面向无法使用 Docker 的用户（例如受限共享服务器、无容器运行时的 HPC 环境）。具体的编译步骤复杂且版本敏感——如果遇到编译错误，请以 Dockerfile 为准，并携带你的环境信息提交 issue。
+:::
+
+## 概述
+
+Relax 依赖多个需要从源码编译的重量级 CUDA 扩展，且锁定了特定 commit 和 patch：
+
+- **Megatron-LM**（通过 Megatron-Bridge 构建，带自定义 patch）
+- **SGLang**（带 R3 rollout 路由回放的自定义 patch）
+- **flash-attn**（v2 + 从特定 commit 编译的 v3）
+- **transformer-engine**（v2.14.1）
+- **apex**（从特定 commit 编译）
+- **TransferQueue**（从特定 commit 编译）
+
+官方 Docker 镜像（`ghcr.io/redai-infra/relaxrl:latest`）是推荐且经过最充分测试的安装方式。仅在无法使用 Docker 时才参考本指南。
+
+## 前置条件
+
+### 硬件
+
+- NVIDIA GPU，计算能力 >= 8.0（Ampere 及以上；Hopper 推荐用于 FA3）
+- 小模型训练（Qwen3-0.6B / 4B）至少需要 24 GB 显存
+- 至少 50 GB 可用磁盘空间，用于源码编译、模型和 checkpoint
+
+### 软件
+
+| 组件 | 要求版本 | 说明 |
+|------|---------|------|
+| 操作系统 | Ubuntu 22.04 (LTS) | 其他 Linux 发行版理论上可用但未测试 |
+| Python | 3.12 | 官方镜像使用的精确版本 |
+| CUDA Toolkit | 12.9 | 必须与 PyTorch 构建版本匹配 |
+| NVIDIA 驱动 | >= 560 | 支持 CUDA 12.9 |
+| GCC / G++ | 11.x | CUDA 扩展编译所需 |
+| CMake | >= 3.18 | apex 等编译所需 |
+| Git | >= 2.30 | 克隆依赖仓库 |
+| rsync | 最新版 | 合并 Megatron-Bridge 源码 |
+| jq | 最新版 | patch 验证 |
+
+### 环境变量
+
+开始前设置以下变量：
+
+```bash
+export CUDA_HOME=/usr/local/cuda-12.9
+export PATH=$CUDA_HOME/bin:$PATH
+export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+export MAX_JOBS=8  # 根据 CPU 核心数调整；越高编译越快，占用内存越多
+```
+
+## 步骤 1：安装系统依赖
+
+```bash
+sudo apt-get update && sudo apt-get install -y --no-install-recommends \
+    build-essential patchelf tmux net-tools htop nano iputils-ping \
+    autoconf automake libatlas-base-dev libgoogle-glog-dev libbz2-dev libc-ares2 \
+    libleveldb-dev liblmdb-dev libprotobuf-dev libsnappy-dev libtool nasm protobuf-compiler pkg-config unzip sox \
+    libsndfile1 libpng-dev libhdf5-dev gfortran rapidjson-dev ninja-build libedit-dev rsync jq \
+    locales tzdata
+```
+
+## 步骤 2：安装 PyTorch 2.9.0（CUDA 12.9）
+
+官方镜像使用 NVIDIA 的 PyTorch 2.9.0 构建（来自 25.06 NGC 容器）。安装匹配版本：
+
+```bash
+pip install torch==2.9.0 torchvision==0.24.0 torchaudio==2.9.0 --index-url https://download.pytorch.org/whl/cu129
+```
+
+验证：
+
+```bash
+python -c "import torch; print(f'PyTorch {torch.__version__}, CUDA {torch.version.cuda}, GPU: {torch.cuda.get_device_name(0)}')"
+```
+
+## 步骤 3：编译核心 CUDA 扩展
+
+### 3.1 flash-attn v2
+
+```bash
+MAX_JOBS=64 pip -v install flash-attn==2.7.4.post1 --no-build-isolation --no-cache-dir
+pip install --no-cache-dir flash-linear-attention==0.4.1
+pip install --no-cache-dir tilelang -f https://tile-ai.github.io/whl/nightly/cu128/
+```
+
+### 3.2 flash-attn v3（仅 Hopper 架构）
+
+::: warning
+此步骤仅适用于 Hopper（H100/H800）GPU。Ampere/Ada 架构请跳过。
+:::
+
+```bash
+cd /opt
+git clone https://github.com/Dao-AILab/flash-attention.git
+cd flash-attention/
+git checkout 0f82fead0b2d59f6390c62a1d74e927159bc5a7b
+git submodule update --init
+cd hopper/
+MAX_JOBS=96 python setup.py install
+
+# 移除与 TE 冲突的顶层模块
+export python_path=$(python -c "import site; print(site.getsitepackages()[0])")
+mkdir -p $python_path/flash_attn_3
+cp flash_attn_interface.py $python_path/flash_attn_3/flash_attn_interface.py
+find $python_path -maxdepth 1 -name flash_attn_interface.py -delete
+find $python_path -maxdepth 2 -path '*.egg/flash_attn_interface.py' -delete
+
+# 验证
+cd / && python -c "import importlib.util as u; assert u.find_spec('flash_attn_interface') is None; import flash_attn_config"
+rm -rf /opt/flash-attention/
+```
+
+### 3.3 transformer-engine
+
+```bash
+pip -v install --no-cache-dir --no-build-isolation "transformer_engine[pytorch]==2.14.1"
+```
+
+### 3.4 torch_memory_saver
+
+```bash
+TMS_CUDA_MAJOR=$(python -c 'import torch; print(torch.version.cuda.split(".")[0])') \
+pip install git+https://github.com/redai-infra/torch_memory_saver.git@afc13785c50119048e2dd8ac497cc9e29ec75bd4 --no-cache-dir --force-reinstall
+```
+
+### 3.5 apex
+
+```bash
+NVCC_APPEND_FLAGS="--threads 32" \
+pip -v install --disable-pip-version-check --no-cache-dir \
+    --no-build-isolation \
+    --config-settings "--build-option=--cpp_ext --cuda_ext --parallel 8" \
+    git+https://github.com/NVIDIA/apex.git@10417aceddd7d5d05d7cbf7b0fc2daad1105f8b4
+```
+
+### 3.6 附加包
+
+```bash
+pip install nvidia-modelopt[torch]>=0.37.0 --no-build-isolation --no-cache-dir
+pip install "numpy<2" nvidia-cudnn-cu12==9.16.0.29 --no-cache-dir
+```
+
+## 步骤 4：通过 Megatron-Bridge 安装 Megatron-LM
+
+Relax 使用从 Megatron-Bridge 源码及其子模块合并的 Megatron-LM 目录树，并应用自定义 patch。
+
+```bash
+export MEGATRON_BRIDGE_COMMIT=2faedbf6fe3c422835a44b2b360cadcb2a116a54
+export MEGATRON_PATH=/root/Megatron-LM  # 调整为你偏好的路径
+
+rm -rf $MEGATRON_PATH
+git clone https://github.com/NVIDIA-NeMo/Megatron-Bridge.git /root/Megatron-Bridge
+cd /root/Megatron-Bridge/
+git checkout ${MEGATRON_BRIDGE_COMMIT}
+git submodule update --init --recursive
+./scripts/switch_mcore.sh dev
+
+mkdir -p $MEGATRON_PATH
+cp -r src/megatron $MEGATRON_PATH/
+rsync -avP 3rdparty/Megatron-LM/megatron/ $MEGATRON_PATH/megatron/
+rm -rf /root/Megatron-Bridge
+```
+
+### 应用 megatron patch
+
+```bash
+# 从 Relax 仓库复制 patch
+cp /path/to/Relax/docker/patch/latest/megatron.patch $MEGATRON_PATH/
+
+cd $MEGATRON_PATH
+patch -p1 < megatron.patch
+
+# 验证 patch 干净应用
+if grep -R -n '^<<<<<<< ' .; then
+    echo "错误：patch 应用失败，请解决冲突。"
+    exit 1
+fi
+rm megatron.patch
+```
+
+### 设置 PYTHONPATH
+
+```bash
+export PYTHONPATH=$MEGATRON_PATH:$PYTHONPATH
+```
+
+验证：
+
+```bash
+python -c "import megatron; print(f'Megatron 导入自 {megatron.__file__}')"
+```
+
+## 步骤 5：安装带自定义 patch 的 SGLang
+
+Relax 使用 SGLang v0.5.15.post1，带 R3（rollout 路由回放）的自定义 patch。
+
+::: tip
+基础 SGLang 安装来自 `lmsysorg/sglang:v0.5.15.post1-cu129` Docker 镜像。无 Docker 环境下，需要先从源码安装匹配版本的 SGLang。
+:::
+
+```bash
+# 从源码安装匹配版本的 SGLang
+pip install "sglang[all]==0.5.15.post1"
+
+# 应用自定义 sglang patch
+export SGLANG_PATH=$(python -c "import sglang; print(sglang.__path__[0])")
+cp /path/to/Relax/docker/patch/latest/sglang.patch $SGLANG_PATH/
+
+cd $SGLANG_PATH
+git add .
+git update-index --refresh
+git apply sglang.patch --3way
+
+# 验证
+if grep -R -n '^<<<<<<< ' .; then
+    echo "错误：SGLang patch 应用失败。"
+    exit 1
+fi
+rm sglang.patch
+```
+
+### 安装打过 patch 的 sglang-router
+
+官方 `sglang-router` wheel 会丢弃 `routed_experts` 字段，导致 R3 失效。Relax 使用 slime 的打过 patch 的 fork：
+
+```bash
+pip install --no-cache-dir --force-reinstall --no-deps \
+    https://github.com/zhuzilin/sgl-router/releases/download/v0.3.2-9daabcd/sglang_router-0.3.2-cp38-abi3-manylinux_2_28_x86_64.whl
+
+# 验证打过 patch 的版本
+python -c "import sglang_router; assert 'slime' in sglang_router.__version__, sglang_router.__version__"
+```
+
+## 步骤 6：安装 TransferQueue
+
+```bash
+pip install "transferqueue @ git+https://github.com/redai-infra/TransferQueue.git@58054a33834aadbcf76aacd6b1e32e25c030f2c9" --no-deps
+```
+
+## 步骤 7：安装 Python 依赖和 Relax
+
+```bash
+cd /path/to/Relax
+
+# 安装纯 Python 依赖
+pip install --ignore-installed PyJWT
+pip install -r requirements.txt --no-cache-dir
+
+# 安装附加包
+pip install --no-cache-dir "compressed_tensors>=0.13.0" tensordict==0.10.0 pyvers==0.1.0 'nvidia-modelopt[hf]==0.44.0' --no-deps
+
+# 以开发模式安装 Relax
+pip install -e .
+```
+
+## 步骤 8：编译 INT4 QAT CUDA 扩展
+
+```bash
+cd /path/to/Relax/relax/backends/megatron/kernels/int4_qat
+pip install . --no-build-isolation --no-cache-dir
+```
+
+## 环境变量汇总
+
+将以下内容添加到 `~/.bashrc` 或专用的环境脚本中：
+
+```bash
+# CUDA
+export CUDA_HOME=/usr/local/cuda-12.9
+export PATH=$CUDA_HOME/bin:$PATH
+export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH
+
+# Megatron
+export MEGATRON_PATH=/path/to/Megatron-LM
+export PYTHONPATH=$MEGATRON_PATH:$PYTHONPATH
+
+# Relax
+export RELAX=/path/to/Relax
+export PYTHONPATH=$RELAX:$PYTHONPATH
+```
+
+## 验证安装
+
+按顺序运行以下检查：
+
+```bash
+# 1. 基础导入
+python -c "
+import torch
+import ray
+import sglang
+import megatron
+import relax
+print('所有基础导入成功')
+print(f'  torch: {torch.__version__}')
+print(f'  CUDA 可用: {torch.cuda.is_available()}')
+print(f'  GPU: {torch.cuda.get_device_name(0)}')
+"
+
+# 2. CUDA 扩展
+python -c "
+import flash_attn
+import transformer_engine
+print('CUDA 扩展导入成功')
+print(f'  flash_attn: {flash_attn.__version__}')
+"
+
+# 3. Relax 组件
+python -c "
+from relax.controller import Controller
+from relax.registry import ALGOS
+print(f'已注册算法: {list(ALGOS.keys())}')
+print('Relax 组件加载成功')
+"
+```
+
+::: warning
+通过这些导入检查并不保证完整训练可以正常运行。端到端训练涉及 Ray 集群初始化、SGLang 引擎启动、NCCL 通信和 Megatron 训练循环——每一步都可能遇到环境特定的问题。建议先用小模型（Qwen3-0.6B）跑几个 training step 来验证。
+:::
+
+## 常见问题排查
+
+### flash-attn / apex / TE 编译错误
+
+- 确保 `CUDA_HOME` 设置正确，`nvcc --version` 显示 12.9
+- 确保 GCC 版本为 11.x（`gcc --version`）
+- 如果编译时内存不足，降低 `MAX_JOBS`
+- 检查 NVIDIA 驱动是否支持 CUDA 12.9（`nvidia-smi`）
+
+### `ModuleNotFoundError: No module named 'megatron'`
+
+- 确保 `PYTHONPATH` 包含 `$MEGATRON_PATH`
+- 验证合并目录存在：`ls $MEGATRON_PATH/megatron/`
+
+### `ModuleNotFoundError: No module named 'sglang'`
+
+- 必须先安装 SGLang，再应用 patch
+- 验证：`python -c "import sglang; print(sglang.__version__)"` 应显示 0.5.15.post1
+
+### NCCL 错误或 Ray 集群失败
+
+- 确保所有节点网络互通（相当于 Docker 的 `--network=host`）
+- 设置 `NCCL_SOCKET_IFNAME` 为正确的网卡接口
+- 单节点训练时，确保 `ray start --head` 能绑定到 localhost
+
+### Patch 冲突
+
+如果 `git apply` 或 `patch` 报告冲突：
+
+1. 检查基础版本是否精确匹配（SGLang 0.5.15.post1、Megatron-Bridge commit 2faedbf6）
+2. 尝试 `git apply --3way` 进行三方合并
+3. 如果冲突持续，请参考 Dockerfile 中的精确 patch 上下文
+
+## 获取帮助
+
+如果遇到本指南未涵盖的问题：
+
+1. 查看[官方 Dockerfile](https://github.com/redai-infra/Relax/blob/main/docker/Dockerfile)——它是权威的构建参考
+2. 搜索已有的 [GitHub Issues](https://github.com/redai-infra/Relax/issues)
+3. 提交新 issue，包含：
+   - 操作系统、GPU 型号、NVIDIA 驱动版本、CUDA 版本
+   - 精确的错误信息和堆栈跟踪
+   - 失败发生的步骤
+   - 是否使用了本指南中指定的精确版本
+
+::: tip
+对大多数用户而言，Docker 镜像仍然是推荐的安装方式。本指南作为受限环境的参考提供，更新频率可能不如 Dockerfile。
+:::