Welcome to the Elasticsearch Linear Regression Model Project

This project uses MLflow and DAGSflow to monitor a Linear Regression model trained on Elasticsearch data. The project is hosted on GitHub at https://github.com/harshvasisht/mlflow_dagshub.

Features

• The model is trained on a dataset of Elasticsearch documents.
• The model is monitored locally using MLflow.
• The model is monitored remotely using DAGSflow.

Benefits

• The MLflow tracking system provides a centralized view of all experiments and models.
• The DAGSflow monitoring system provides real-time alerts and notifications.
• The project is hosted on GitHub, which makes it easy to collaborate and share with others.

Getting started

To get started with this project, you can follow these steps:

1. Clone the repository.

2. Install the dependencies.

   bash setup.sh

3. Activate the environment.

       conda activate venv

4. Run the 'app.py' script to train the model. With the alpha & l1_ratio.

       python app.py 0.3 0.7

       Elasticnet model (alpha=0.300000, l1_ratio=0.700000):
       RMSE: 0.774713648356711
       MAE: 0.607967853255621 
       R2: 0.14961391810397706

5. To view MLflow dashboard.

       mlflow ui

Resources

• MLflow documentation
• DAGSflow documentation
• GitHub repository

Community

The MLflow and DAGSflow communities are vibrant communities of machine learning practitioners and developers. The communities provide support, resources, and collaboration opportunities.

• MLflow community forum
• MLflow community Slack channel
• DAGSflow community forum
• DAGSflow community Slack channel

I hope you enjoy this project!