https://github.com/nonsignificantp/torch-cuda-jupyter-environ

Jupyter Lab dev environment with torch an cuda out of the box
https://github.com/nonsignificantp/torch-cuda-jupyter-environ

Last synced: 1 day ago
JSON representation

Jupyter Lab dev environment with torch an cuda out of the box

• Host: GitHub
• URL: https://github.com/nonsignificantp/torch-cuda-jupyter-environ
• Owner: nonsignificantp
• Created: 2021-10-19T05:10:59.000Z (about 3 years ago)
• Default Branch: main
• Last Pushed: 2021-10-19T05:48:24.000Z (about 3 years ago)
• Last Synced: 2024-11-09T09:44:07.971Z (about 2 months ago)
• Language: Dockerfile
• Size: 2.93 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# JupyterLab environment w/ torch and [CUDA](https://www.youtube.com/watch?v=PeMvMNpvB5M)

## Getting started

### Pull from ghcr

```

docker pull ghcr.io/nonsignificantp/torch-cuda-jupyter-environ:main

```

```

docker run -it --rm -p 8888:8888 ghcr.io/nonsignificantp/torch-cuda-jupyter-environ:main

```

### Build from source

Clonar el repositorio

```

git clone [email protected]:nonsignificantp/torch-cuda-jupyter-environ.git

```

#### Con docker

Crear la imagen

```

docker build -t app/pytorch-cuda-jupyter:latest .

```

Arrancar un container

```

docker run -it --rm -p 8888:8888 app/pytorch-cuda-jupyter

```

Se levantara una instancia de jupyter lab accesible desde el navegador ([http://localhost:8888/lab/](http://localhost:8888/lab/)).

#### Con make

```

make jupyter

```

### Setting up a workdir

El mapeo de un directorio de trabajo debe apuntar a la carpeta `/workspace` dentro del container. A modo de ejemplo:

```

docker run -it --rm \

    -p 8888:8888 \

    -v /path/to/my/workspace:/workspace

    app/pytorch-cuda-jupyter

```