RhythmicML

A tool for ML models deployment and versioning.

Quick Start:

Requirements:

`pip install flask requests rhythmic`

Launch the client-side server (see below), use browser for UI (http://localhost:5000);
Press "Add model";
Navigate to the folder with your model and artifacts
Press "Add"

A `ModelWrapper.py` class will be copied to that folder - wrap your model and scoring into it. This is compulsory.

Launch deployment server (see below)

You are ready to go.

NOTICE: a configuration is stored in configuration.py file in a helpers folder both for client and server side.

Client-side preparation:

cd ... /src/rhml_client/rhml_client
python setup_db.py

Launching UI:

cd ... /src/rhml_client/rhml_client
./run.sh

or

cd ... /src/rhml_client/rhml_client
python -m UI

Server-side preparation:

cd ... /src/rhml_server/rhml_server
python setup_db.py

Launching Server:

cd ... /src/rhml_server/rhml_server
./run.sh

or

cd ... /src/rhml_server/rhml_server
python -m Deploy