CUDA Development Container

This repo configures a CUDA enabled python development container. If you are unfamiliar with development containers, check out this page.

Note: This repository is a fork of mirzai/cuda-devcontainer, but it closely follows the structure and configuration of raphaelschwinger/cuda-python-devcontainer, from which most of the setup has been adapted.

Installation

1. Add repo as submodule

git submodule add git@github.com:PariaValizadeh/cuda-devcontainer.git .devcontainer

2. Set CUDA_VERSION and VARIANT in .devcontainer/devcontainer.json to your desired values. The CUDA and Ubuntu version has to match with image tags found here.

3. Reopen in container

4. Init poetry project

poetry init

or create your own pyproject.toml file.

[tool.poetry]
name = "pytorch-lightning-devcontainer-example"
version = "0.1.0"
description = ""
authors = ["Raphael Schwinger"]
readme = "README.md"

[tool.poetry.dependencies]
python = "^3.8"
pathlib = "^1.0.1"
argparse = "^1.4.0"
torch = [
    {url = "https://download.pytorch.org/whl/cu117/torch-2.0.1%2Bcu117-cp38-cp38-linux_x86_64.whl", platform = "linux", python = ">=3.8 <3.9"}
]
torchvision = [
    {url = "https://download.pytorch.org/whl/cu117/torchvision-0.15.2%2Bcu117-cp38-cp38-linux_x86_64.whl", platform = "linux", python = ">=3.8 <3.9"}
]
torchaudio = [
    {url = "https://download.pytorch.org/whl/cu117/torchaudio-2.0.2%2Bcu117-cp38-cp38-linux_x86_64.whl", platform = "linux", python = ">=3.8 <3.9"}
]
ipykernel = "^6.25.2"
lightning = "^2.0.8"
jupyter = "^1.0.0"
pillow = "^10.0.0"



[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"