Skip to content

Latest commit

 

History

History
51 lines (37 loc) · 1.46 KB

README.md

File metadata and controls

51 lines (37 loc) · 1.46 KB

MLOps CLI on CPD

This project includes an evolving design of the MLOps flow on CPD and the corresponding implementation as a CLI tool.

The current version covers a flow for deep learning models as the follows:

  • train: code development in WS, training job on WMLA
  • deploy: WMLA Elastic Distributed Inference
  • monitor: custom monitor for OpenScale, headless service provider & dummy subscription, only custom monitors enabled for subscription

Roadmap

Next steps:

  • add toy model, toy data, and toy custom monitor script for dev and test
  • set up unit tests
  • extend to WML deployments
  • extend to OOTB OpenScale monitors
  • move from config yaml to factsheets host metadata shared between services

Dependencies

Python: >= 3.8

Python packages:

  • ibm-cloud-sdk-core==3.10.1
  • ibm-watson-openscale>=3.0.14
  • ibm-watson-machine-learning>=1.0.246
  • click
  • cpd-sdk-plus>=1.1

Installation

No installation needed, but you can install the dependencies as follows:

pip install -r requirements.txt

Usage

Download the cli script and the dependency utility scripts. Now you can use it:

python cli_mlops.py --help

For example of available commands, see the cheat sheet.

How to Contribute

DCO is suggested to be used. See here for details on how to do it.

Contributors