---
title: Machine learning datasets on demand tutorial with Python model example
description: Python code to create a Windocks container from an image, connect to that container, and run an ML algorithm to train and test the data.
---

 

 

 TECHNICAL DOCS

# Machine learning datasets on demand tutorial with Python model example

These instructions describe how to use a Python code sample to create a Windocks container from an image, connect to that container, and run an ML algorithm to train and test the data. 

## Steps

Step 1: Set up the Windocks server (see [Getting Started with Windocks](https://23302346.hs-sites.com/getting-started))

Step 2: Set up an MLOps tool like MLFlow (see [Installing MlFlow](https://www.mlflow.org/docs/latest/quickstart.html))

Step 3: Build an image with the data needed for the experiments. See [the build steps here](https://23302346.hs-sites.com/doc-simple-build-clone)

Step 4: Collect information on the Windocks machine IP, MLFlow machine IP, user name, password, etc

Step 5: Use the file windocks\samples\machinelearning\train1sql.py as the starting point for your model. It has the Python code to create a Windocks container from the image in step 2, connecting to that container and reading the data, and then running an ML algorithm on that data to train and test the data. Follow the steps in windocks\samples\machinelearning\README.txt

Email [support@windocks.com](mailto:support@windocks.com) with any questions

[Back to Technical Docs Gallery ](https://23302346.hs-sites.com/resources)

## Extend CI/CD to your data layer

[GET A DEMO ](https://23302346.hs-sites.com/schedule-demo) [TRY FREE ](https://23302346.hs-sites.com/try-free)