Skip to content

Latest commit

 

History

History
47 lines (31 loc) · 1.29 KB

README.md

File metadata and controls

47 lines (31 loc) · 1.29 KB

luigi-ml

A data/model agnostic ML experiment system implemented by luigi, redis, postgres and docker. This project was spun off from a talk I gave virtually at Instituto Tecnológico Autónomo de México of course over Zoom. The slide of the talk and the part of this tool is here

Quick Start

Copy luigi.cfg.example to luigi.cfg

cp luigi.cfg.example luigi.cfg

Build with docker

docker-compose -f docker-compose-dev.yml build

Modify model_config.yaml

Run the modeling experiment

docker exec -it luigi_service python -m luigi --module pipeline.tasks Experiment --model-config=model_config.yaml --workers=4

Open the dashboard by typing localhost:9091 on the browser

Use results schema for post-modeling anlaysis

docker exec -it luigi_db psql -d luigid -U luigid

Run Locally

Start Postgres Database

Start Redis Service

Configure luigi.cfg with correct connection credential for Postgres and Redis

Run luigid luigi scheduler daemon

luigid --background --port=9091 --pidfile <PATH_TO_PIDFILE> --logdir <PATH_TO_LOGDIR> --state-path <PATH_TO_STATEFILE>

Open the dashboard by typing localhost:9091 on the browser