DevContainer for CUDA + PyTorch

This repository contains a devcontainer for CUDA + PyTorch development. It is based on the nvidia/cuda image.

Features

You can use this devcontainer to develop CUDA + PyTorch applications in a container.

The container is based on the nvidia/cuda image and contains the following tools:

• CUDA 11.7.1
• Python 3.10
• PyTorch
• Windows Subsystem for Linux 2 (WSL2)

Usage

1. Install Docker and VSCode.
2. Install the Remote - Containers extension.
3. Clone this repository.
4. Open the repository in VSCode.
5. Click the green button in the lower left corner of the window.
6. Wait for the container to build and start.
7. Open a terminal in VSCode and run python to start the Python REPL.

License

This project is licensed under the MIT License.