setup-ml

Sqlite3

Steps

0. cd setup

1. Build docker

docker build -t qureshi/ml .

This can be achieved by running the script image-build.sh.

Note this line

Successfully tagged qureshi/ml:latest

which identifies that the image has been successfully built. You'll need image tag to run the container.

2. Set external variables and spool up the container

User id

export user_id=$(id -u $USER)

and group id.

export group_id=$(id -g $USER)

Now spool up the container

docker-compose -f docker-compose.yml up rr

This can be achieved by running the script container-run.sh.

Note this line

Attaching to setup-ml_charlie_1

that identifies the container tag. You'll need this to attach to this container.

4. Attach to this container

docker exec -it setup-ml_rr_1 bash

This can be achieved by running the script container-attach.sh.

5. Within container prompt, run jupyter if needed

jupyter notebook --ip 0.0.0.0 --port 8888 --no-browser

Then from another machine, you can connect to this jupyter instance via SSH tunneling.

On another machine X

ssh panther -NL 8888:localhost:8888

Then on X, simply go to localhost:8888 to access the jupyter notebook running within the container that you just created. You will need to unique token to access it. Note it down when jupyter notebook spits it out.