https://github.com/dbpprt/ml-in-devcontainers
Immutable development environments for PyTorch powered by Visual Studio Code Dev Containers
https://github.com/dbpprt/ml-in-devcontainers
devcontainers machine-learning pytorch vscode
Last synced: 8 months ago
JSON representation
Immutable development environments for PyTorch powered by Visual Studio Code Dev Containers
- Host: GitHub
- URL: https://github.com/dbpprt/ml-in-devcontainers
- Owner: dbpprt
- License: mit
- Created: 2023-02-14T18:24:24.000Z (almost 3 years ago)
- Default Branch: main
- Last Pushed: 2023-02-15T22:26:27.000Z (almost 3 years ago)
- Last Synced: 2025-03-29T08:43:52.120Z (9 months ago)
- Topics: devcontainers, machine-learning, pytorch, vscode
- Language: Python
- Homepage:
- Size: 14.6 KB
- Stars: 10
- Watchers: 1
- Forks: 3
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
Immutable development environments for PyTorch powered by Visual Studio Code Dev Containers
#
This project provides a development environment for PyTorch using Visual Studio Code and Docker. It is based on the [Visual Studio Code Dev Container](https://marketplace.visualstudio.com/items?itemName=ms-vscode-remote.remote-containers) feature.
Being able to have **immutable** and **reproducible** development environments is a key part to successfully work with multiple projects.
## Features
- **devcontainer.json**: Configuration file for the Visual Studio Code Dev Container feature, using a Docker image, based on nvidia/cuda:11.8.0-base-ubuntu22.04
- **Optionally**: install AWS CLI V2 and reuse your AWS credentials from the host system
- **Optionally**: install Docker inside the container and reuse your Docker daemon from the host system (if you're training inside a container again)
- Sophisticated set of extensions for Python, Jupyter, etc.
- Sopthisticated set of settings for Python (Linter, Formatter, etc.)
- Sample MNIST training script (PyTorch Example)
## Getting Started
- **(Prerequisite)** Docker is installed on your system
- **(Prerequisite)** NVIDIA Container Toolkit is installed on your system
- **(Prerequisite)** Visual Studio Code Dev Container extension is installed
- Clone this repository
- Open the folder in Visual Studio Code
- Read comments in the `devcontainer.json` file
- Read comments in the `setup.sh` file
- When prompted, click "Reopen in Container"