在 VScode中使用docker和remote跑一个GPU项目How to use docker to build a python program in Visual Studio

How to use docker to build a python program in Visual Studio

1. EXTENSION to get dev exploer and other jupyter…
   dev container is not that good
   
   

2. build a folder in the left and use ctrl shift +p to dev container: reopen in container or dev container: open folder in container

3. devcontainer.json

{
    "name": "pytorch-aria",
    "dockerComposeFile": "docker-compose.yml",
    "runServices": ["devcontainer"],
    "service": "devcontainer",
    "workspaceFolder": "/home/workspace",
    "customizations": {"vscode":{"extensions": [
        "ms-python.python",
        "esbenp.prettier-vscode",
        "eamodio.gitlens"
    ]}}

}

4. docker-compose.yml

version:  "3.7"
services: 
  devcontainer:
    build: .
    image: work1:v1
    container_name: work1
    volumes:
      - ..:/home/workspace
      - ~/.ssh:/root/.ssh:ro
      - ~/.gitconfig:/root/.gitconfig:ro
    deploy:
      resources:
        reservations:
          devices:
            - capabilities: [gpu]
    entrypoint: bash
    stdin_open: true
    tty: true

6. dockerfile
   if I add some new pip in dockerfile, you should command shift +p dev containers:rebuild container

FROM pytorch/pytorch:2.0.0-cuda11.7-cudnn8-devel
RUN apt-get update && apt-get install -y \
    git \
    libsparsehash-dev

RUN pip install --upgrade git+https://github.com/mit-han-lab/torchsparse.git@v1.4.0
RUN pip install matplotlib

7. other exploer
   
   8.docker
   
   dev container中打开