MLflow

Introduction

MLflow is an open source platform to manage the ML lifecycle, including experimentation, reproducibility, deployment, and a central model registry.

Property	Description
App Image	larribas/mlflow
Official Website	https://mlflow.org/

Screenshots

Usage

1. Create a MLflow app

2. There are two predefined environment variables

3. The default value of backend store uri sqlite:///$(PRIMEHUB_APP_ROOT)/mlflow.db is the sqlite db under the group volume

4. The default value of artifact root $(PRIMEHUB_APP_ROOT)/mlruns is the folder under the group volume

5. You can change these environment variables to use your own settings if you want

6. After created, MLflow is ready to manage your ML lifecycle