ml-docker-example

This repository is for the simple example with docker environment. It contains how to build sample image and how to access docker container in various research environment form(jupyter notebook, jupyter lab, ssh).

How to build sample image.

1. Download docker in https://www.docker.com/

2. Pull base image

docker pull nvidia/cuda:11.3.1-cudnn8-devel-ubuntu20.04

3. Build sample image with Makefile. It actually build two images(Dockerfile, Dockerfile.opencv)

make all

4. Execute image with container

make dry-run

Access to docker container with different port number.

with make dry-run command, port forwarding can be set in container and each external port number can be assigned to each research environment in Dockerfile.

1. jupyter notebook

host_server_ip_address:21501

2. jupyter lab

host_server_ip_address:21502

3. ssh

ssh -i private_key_location -p 25000 omnious@host_server_ip_address