Gaiaflow is a local-first MLOps infrastructure python package tool that simplifies the process of building, testing, and deploying ML workflows. It provides an opinionated CLI for managing Airflow, MLflow, and other dependencies, abstracting away complex configurations, and giving you a smooth developer experience.
NOTE: Currently this library is released as an experimental version. Stable releases will follow later
Gaiaflow is a tool that
-
provides you with a local MLOps infrastructure via a CLI tool with some prerequisites already installed.
-
handles the complex Airflow configuration and Xcom handling and provides the user a simpler interface for creating DAGs.
-
provides a cookiecutter template to get started with your projects with a standardized structure.
-
provides tools to deploy models locally and in production (in future)
-
provides clear documentation on how to setup production environment to run your workflows at scale (in future, private?)
Prerequisites:
- Docker
- Docker compose
- Miniforge
- Mamba/Conda
To install it, you can do it via:
pip install gaiaflow
Check installation:
gaiaflow --help
You can read the documentation here