[docs] docs: promote pre-built docker images to top-level section

- Promote 'Pre-built Docker Images' from a ### subsection of 'Option A:
  Docker Image' to a top-level ## section so it appears in the page TOC
  alongside the build and source-install options.
- Move 'Run the container' under 'Pre-built Docker Images' since the
  example tags (loongforge/loongforge:${LOONGFORGE_VERSION}[_lerobot])
  only exist on the pre-built path; pairing it with the preceding
  'docker pull' commands keeps the snippet self-consistent.
- Promote 'Dual Python Environment in the LeRobot Image' from #### to
  ### so it nests correctly under the new top-level section.
- Update cross-reference 'see Dual Python Environment above' -> 'below'
  to match the new ordering.
- Mirror the same restructuring in the Chinese translation
  (docs/source_zh/get_started/installation.md).

Author: Zachary-wW

docs/source/get_started/installation.md

@@ -83,7 +83,9 @@ After the build finishes, verify:
 docker images | grep loongforge
 ```
 
-### Pre-built Docker Images
+---
+
+## Pre-built Docker Images
 
 LoongForge Docker images are available on Docker Hub:
 [https://hub.docker.com/u/loongforge](https://hub.docker.com/u/loongforge).
@@ -108,7 +110,30 @@ docker pull loongforge/loongforge:${LOONGFORGE_VERSION}
 docker pull loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot
 ```
 
-#### Dual Python Environment in the LeRobot Image
+### Run the container
+
+```bash
+# Set the version tag you want to use, for example: 0.1.1
+LOONGFORGE_VERSION=<version>
+
+# Using the base image (LLM/VLM/Diffusion)
+docker run --runtime=nvidia --gpus all -itd --rm \
+  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
+  -v /path/to/data:/mnt/cluster/LoongForge/ \
+  loongforge/loongforge:${LOONGFORGE_VERSION} /bin/bash
+
+# Using the LeRobot image (VLA: Pi0.5 + GR00T)
+docker run --runtime=nvidia --gpus all -itd --rm \
+  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
+  -v /path/to/data:/mnt/cluster/LoongForge/ \
+  loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot /bin/bash
+```
+
+Once inside the container, navigate to `/workspace/LoongForge/examples/` and
+launch the desired training script. For GR00T training, remember to activate
+the GR00T virtual environment first (see Dual Python Environment below).
+
+### Dual Python Environment in the LeRobot Image
 
 The LeRobot image (`<version>_lerobot`) uses a **dual virtual-environment** setup to resolve
 dependency conflicts between Pi0.5 and GR00T:
@@ -138,29 +163,6 @@ deactivate
 /opt/venvs/gr00t/bin/torchrun ${DISTRIBUTED_ARGS[@]} ...
 ```
 
-### Run the container
-
-```bash
-# Set the version tag you want to use, for example: 0.1.1
-LOONGFORGE_VERSION=<version>
-
-# Using the base image (LLM/VLM/Diffusion)
-docker run --runtime=nvidia --gpus all -itd --rm \
-  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
-  -v /path/to/data:/mnt/cluster/LoongForge/ \
-  loongforge/loongforge:${LOONGFORGE_VERSION} /bin/bash
-
-# Using the LeRobot image (VLA: Pi0.5 + GR00T)
-docker run --runtime=nvidia --gpus all -itd --rm \
-  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
-  -v /path/to/data:/mnt/cluster/LoongForge/ \
-  loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot /bin/bash
-```
-
-Once inside the container, navigate to `/workspace/LoongForge/examples/` and
-launch the desired training script. For GR00T training, remember to activate
-the GR00T virtual environment first (see Dual Python Environment above).
-
 ---
 
 ## Option B: Install from Source

docs/source_zh/get_started/installation.md

@@ -75,7 +75,9 @@ docker build --build-arg COMPILE_ENV=hopper --build-arg ENABLE_LEROBOT=false \
 docker images | grep loongforge
 ```
 
-### 预构建 Docker 镜像
+---
+
+## 预构建 Docker 镜像
 
 LoongForge Docker 镜像保存在 Docker Hub：
 [https://hub.docker.com/u/loongforge](https://hub.docker.com/u/loongforge)。
@@ -99,7 +101,29 @@ docker pull loongforge/loongforge:${LOONGFORGE_VERSION}
 docker pull loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot
 ```
 
-#### LeRobot 镜像中的双 Python 环境
+### 运行容器
+
+```bash
+# 设置需要使用的版本 tag，例如：0.1.1
+LOONGFORGE_VERSION=<version>
+
+# 使用基础镜像（LLM/VLM/Diffusion）
+docker run --runtime=nvidia --gpus all -itd --rm \
+  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
+  -v /path/to/data:/mnt/cluster/LoongForge/ \
+  loongforge/loongforge:${LOONGFORGE_VERSION} /bin/bash
+
+# 使用 LeRobot 镜像（VLA: Pi0.5 + GR00T）
+docker run --runtime=nvidia --gpus all -itd --rm \
+  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
+  -v /path/to/data:/mnt/cluster/LoongForge/ \
+  loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot /bin/bash
+```
+
+进入容器后，导航到 `/workspace/LoongForge/examples/` 并启动所需的训练脚本。
+对于 GR00T 训练，请先激活 GR00T 虚拟环境（参见下方双 Python 环境说明）。
+
+### LeRobot 镜像中的双 Python 环境
 
 LeRobot 镜像（`<version>_lerobot`）使用**双虚拟环境**方案解决 Pi0.5 和 GR00T 之间的依赖冲突：
 
@@ -128,28 +152,6 @@ deactivate
 /opt/venvs/gr00t/bin/torchrun ${DISTRIBUTED_ARGS[@]} ...
 ```
 
-### 运行容器
-
-```bash
-# 设置需要使用的版本 tag，例如：0.1.1
-LOONGFORGE_VERSION=<version>
-
-# 使用基础镜像（LLM/VLM/Diffusion）
-docker run --runtime=nvidia --gpus all -itd --rm \
-  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
-  -v /path/to/data:/mnt/cluster/LoongForge/ \
-  loongforge/loongforge:${LOONGFORGE_VERSION} /bin/bash
-
-# 使用 LeRobot 镜像（VLA: Pi0.5 + GR00T）
-docker run --runtime=nvidia --gpus all -itd --rm \
-  -v /path/to/your/hf/models:/mnt/cluster/huggingface.co/ \
-  -v /path/to/data:/mnt/cluster/LoongForge/ \
-  loongforge/loongforge:${LOONGFORGE_VERSION}_lerobot /bin/bash
-```
-
-进入容器后，导航到 `/workspace/LoongForge/examples/` 并启动所需的训练脚本。
-对于 GR00T 训练，请先激活 GR00T 虚拟环境（参见上方双 Python 环境说明）。
-
 ---
 
 ## 方式 B：源码安装