This project sets up a Dockerized environment for machine learning development with Jupyter Notebook and a Java service. It uses NVIDIA CUDA for GPU acceleration and provides a clean, portable setup for your development needs.
In reality i got tired for setting up python, and jupyter on several differnet os's windows, mac linux and all the computers. I can clone this and go.
The other project is java, and ubuntu mode includes cuda. in most cases for me since i have nvidia cards i want cuda for models/infrence. Just follow mac-style if you want jupyter only.
This started because I use linux a lot; I don't like polluting my machine with dependnecies like cuda and various python libraries and whatever else gets pulled in so I use docker.
- Docker
- Docker Compose
- NVIDIA Docker (for GPU support)
- Make
Clone the repository and navigate to the project directory.
git clone <your-repo-url>
cd machine_learning
To build the Docker images, run:
make build
To start the Docker containers, run:
make up
Open your web browser and go to http://localhost:8888. You should see the Jupyter Notebook interface.
The Makefile provides several useful targets for managing and diagnosing the application:
make build
make up
make down
make clean
make rmi
make ps
make logs
make export
make setup
make init
Folders: machine_learning/ ├── docker-compose.yml ├── docker-compose.linux.yml ├── docker-compose.mac.yml ├── Dockerfile.base ├── Dockerfile.jupyter ├── Dockerfile.jupyter.mac ├── Makefile ├── requirements-common.txt ├── requirements-jupyter.txt ├── requirements-jupyter-mac.txt ├── java/ └── jupyter/