vishalbansal-1650 / MLflow-Demo

MLFLOW Demo

Install Requirements/Libs

pip install -r requiements.txt

Approach 1 - Running without setting mlflow tracking server

Note : It will create a new folder mlruns & store all the related information in that folder

Train Model

python train.py <alpha> <l1_ratio>

MLflow UI

mlflow ui

Approach 2 - Running with setting mlflow tracking server

Note : 
1.It would store model file into a newly created folder artifacts and all related logs would be available on mentioned server.

2. MLflow entities are inserted in a SQLite database file mlflow.db

Run MLflow Server

mlflow server \
--backend-store-uri sqlite:///mlflow.db \
--default-artifact-root ./artifacts \
--host 127.0.0.1 -p 1234

Train Model

python train.py <alpha> <l1_ratio>