mlsetup 1.0.0

Project Structure for MLflow integrated ML Projects

This cli tool generates the following directory structure for quickstart ML projects

installaton:

pip install mlsetup

example use:

mlsetup create mytestproject

It generates the project with following structure

Project
|
+--- Input
|    |
|    +--- raw               Raw data here
|    |
|    +--- interim           Any intermediate data, to pause and continue experiments
|    |
|    +--- processed         Processed data ready for ML pipeline
|
+--- output
|    |
|    +--- models             Model pickle or model weights stored here
|    |
|    +--- artifacts         Serialized artifacts like LabelEncoder, Vectorizer etc
|    |
|    +--- images            All plots and visualizations goes here
|    |
|    +--- Results           If the results needs to be stored for review, save here
|
+--- notebooks              All notebooks and experiments resides here
|
+--- src                    Final program, with training and prediction pipeline
|
|
README.md                   Description and instruction about the project
mlsetup                   MLflow project file. If you want to use this directory as MLflow project
Requirements.txt            python dependencies
Config.yml                  configuration key values in yaml format