Counting 3,384 Big Data & Machine Learning Frameworks, Toolsets, and Examples...
Suggestion? Feedback? Tweet @stkim1

Author
Contributors
Last Commit
Dec. 12, 2018
Created
Nov. 5, 2018

modelchimp-logo

Join ModelChimp Slack channel

What is ModelChimp?

ModelChimp is an experiment tracker for Deep Learning and Machine Learning experiments.

ModelChimp provides the following features:

  • Real-time tracking of parameters and metrics
  • Realtime charts for experiment metrics at epoch level
  • Code used for the experiment
  • Experiment comparison
  • Collaborate and share experiments with team members
  • Python objects storage such as data objects and model objects which can be used pulled for other experiments
  • Storage of test and validation images for computer vision use cases. Useful for post experiment forensics of deep learning models
  • Server based solution with user registration and authentication

modelchimp-gif

Why ModelChimp?

The idea for ModelChimp came up when I was building a recommendation algorithm for a large retail company based in India. Along with my 6 member team, we would store the meta information related to each experiment in an excel sheet. Two of the biggest problems we encountered while using this approach were:

  1. Sometimes, we would miss out on logging the details while fine-tuning and analysing the model
  2. Sharing these excel sheets over email amongst the team members and the client was a cumbersome process

ModelChimp is a solution to this problem faced by data scientists and machine learning engineers/enthusiasts. They can spend more time on experiments and not on managing the data related to the experiments.

Installation

Choose either Docker based installation or the manual approach.

  • Docker
  • Manual
  • Production Deployment

Docker

  1. Docker is a prerequisite. You can download it from here - https://docs.docker.com/install/
$ git clone https://github.com/ModelChimp/modelchimp
$ cd modelchimp
$ bash docker.sh --load-data

Use the following command to start ModelChimp without any preloaded data

$ bash docker.sh
  1. After starting ModelChimp server, you can access it at http://localhost:8000

  2. Use the following credentials to log in

username: admin@modelchimp.com
password: modelchimp123

Manual

  1. Create database and user in Postgres for ModelChimp and give privileges.
$ psql user=postgres
Password:
psql (10.4 (Ubuntu 10.4-0ubuntu0.18.04))
Type "help" for help.

postgres=# CREATE DATABASE modelchimp;

CREATE DATABASE
postgres=# CREATE USER modelchimp WITH PASSWORD 'modelchimp123';
CREATE ROLE
postgres=# GRANT ALL PRIVILEGES ON DATABASE modelchimp TO modelchimp;
GRANT
postgres=# \q
  1. Install Redis and check its live with the following command
$ redis-cli
127.0.0.1:6379> ping
PONG
  1. Clone and cd to the repository
$ git clone https://github.com/ModelChimp/modelchimp
$ cd modelchimp
  1. Copy .env-dev into .env and fill the db details
$ cp .env-dev .env
DB_HOST=localhost
DB_NAME=modelchimp
DB_USER=modelchimp
DB_PASSWORD=modelchimp123
DB_PORT=
  1. Create a virtual environment and instantiate it
$ virtualenv -p python3 venv
$ source venv/bin/activate
  1. Run the following to start ModelChimp server
$ python install -r requirements.txt
$ python manage.py collectstatic
$ python loaddata modelchimp.json
$ python manage.py runserver
  1. Access ModelChimp server at http://localhost:8000 and use the following credentials
username: admin@modelchimp.com
password: modelchimp123

Production Deployment

For production deployment, contact Karthik at karthik@modelchimp.com

Documentation