Add Dockerfile

docker/Dockerfile

@@ -0,0 +1,34 @@
+FROM nvidia/cuda:12.6.3-cudnn-devel-ubuntu22.04
+
+# Set environment variables
+ENV DEBIAN_FRONTEND=noninteractive
+ENV NVIDIA_DRIVER_CAPABILITIES=all
+ENV DISPLAY=unix$DISPLAY
+
+# Install necessary packages
+RUN apt-get update && apt-get upgrade -y \
+    && apt-get install -y --no-install-recommends \
+        sudo \
+        software-properties-common \
+        python-is-python3 \
+        git \
+        python3-pip \
+        libgl1-mesa-glx \
+        libglu1-mesa \
+        libxrandr2 \
+        libxinerama1 \
+        libxcursor1 \
+        libxi6 \
+        libx11-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# Clone the repository
+WORKDIR /root/
+RUN git clone https://github.com/LeCar-Lab/dial-mpc.git --depth 1
+
+# Install the package
+WORKDIR /root/dial-mpc
+RUN pip3 install -e .
+
+# Default command
+CMD ["/bin/bash"]

docker/README.md

@@ -0,0 +1,45 @@
+# Docker Tutorial for [DIAL-MPC](https://github.com/LeCAR-Lab/dial-mpc)
+
+This folder contains the Dockerfile of the DIAL-MPC.
+DIAL-MPC is an impressive framework for legged robot ***full-order torque-level*** control with both precision and agility in a ***training-free*** manner.
+You can find the DIAL-MPC in [here](https://github.com/LeCAR-Lab/dial-mpc).
+
+## Requirenments
+> [!CAUTION]
+> You need a Linux or Ubuntu system with Nvidia-driver (>=525.60.13) to support CUDA 12.*
+
+## Docker and Nvidia-toolkit installation
+### Step 1. Docker install
+``` bash
+curl -fsSL https://get.docker.com -o get-docker.sh
+sudo bash get-docker.sh
+``` 
+### Step 2. Docker Postinstall(Follow [here](https://docs.docker.com/engine/install/linux-postinstall/))
+### Step 3. Docker Nvidia Toolkit(Follow [here](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html))
+
+### Step 4. (Optional) Install xhost
+``` bash
+sudo apt install x11-xserver-utils
+```
+## Build Docker Env
+``` bash
+git clone https://github.com/XuXinhangNTU/dial-mpc-docker
+cd dial-mpc-docker
+docker build -t dial-mpc .
+```
+### First Time Run (Create the container)
+```bash
+xhost +
+docker run -it -v /tmp/.X11-unix:/tmp/.X11-unix -e DISPLAY=unix$DISPLAY --gpus all --name <Your Name> dial-mpc
+```
+### Second Time Run
+```bash
+xhost +
+docker start <Your Name> && docker exec -it <Your Name> bash
+```
+## Tips
+The vscode plugin "Dev Containers" is very easy to use when you do docker development.
+
+## Notice
+> [!CAUTION]
+> If the demo looks odd, check your GPU usage! RTX 3090 is tested.