https://github.com/shawonashraf/mlcontainers
https://github.com/shawonashraf/mlcontainers
devcontainers docker pytorch
Last synced: 7 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/shawonashraf/mlcontainers
- Owner: ShawonAshraf
- License: mit
- Created: 2024-12-02T21:33:30.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-03-17T04:00:51.000Z (about 1 year ago)
- Last Synced: 2025-03-25T10:31:54.916Z (about 1 year ago)
- Topics: devcontainers, docker, pytorch
- Language: Dockerfile
- Homepage:
- Size: 29.3 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# mlcontainers
A collection of devcontainer configurations I usually use for my deep learning / machine learning projects.
## usage
> [!IMPORTANT]
> For GPU passthrough to work in some configurations, you'll need to install
> the `nvidia-container-toolkit` first. Please refer to the official [docs](https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html).
Install the `devcontainer-cli` via `npm`
```bash
npm install -g @devcontainers/cli
# you may have to use sudo in some linux distros
# especially if npm installed as a system package
```
Then run the configuration you want by
```bash
devcontainer up --workspace-folder
# example
devcontainer up --workspace-folder ./torch-uv
```
If you only wish to build a devcontainer to run it later,
```bash
devcontainer build --workspace-folder
# example
devcontainer build --workspace-folder ./jax-uv
```
> [!IMPORTANT]
> These configurations are for Nvidia GPUs and x86_64 based systems.
> Besides, GPU passthrough isn't there for macOS/Asahi Linux running Apple Silicon yet.
Once up and running, you can connect to the devcontainer through VS Code or Jetbrains IDEs (these are the ones I use actually!).