Setup collaborative MLflow with PostgreSQL as Tracking Server and MinIO as Artifact Store using docker containers

This is the repo for these blog posts:

• My Website: Setup collaborative MLflow with PostgreSQL as Tracking Server and MinIO as Artifact Store using docker containers
• Medium Post

Preparing the environment

Note: I have tested the codes on Linux. It can certainly be run on Windows and MacOS with small modifications.

1. Clone the repository, and navigate to the downloaded folder.

git clone https://github.com/iamirmasoud/mlflow_postgres_minio.git
cd mlflow_postgres_minio

2. Create (and activate) a new environment, named mlflow_env with Python 3.7. If prompted to proceed with the install (Proceed [y]/n) type y.

   conda create -n mlflow_env python=3.7
   source activate mlflow_env

3. Install the prerequisites for building the psycopg2 package from source on Ubuntu:

sudo apt install libpq-dev python3-dev

4. Install requirements for the environment:

pip install -r requirements.txt