Automate your ML model with DVC and CMl

Automate your ML model with DVC and CMl

Table of contents

No heading

No headings in the article.

When a data Scientist or ml engineer develop a ML model for solving businesses related problems they create a model but sometimes need to retrain your model because more type of data are generated in a every single second.

The Tools for automate your workflows.

  1. GitHub Actions
  2. Jenkins
  3. Bitbucket
  4. Grafava (Advance level)
  5. CML (Continues Machine Learning)

and many more these are the best.

For you ML model GitHub Actions is best Tool for automate your workflows

  1. Create a small project and push in your GitHub.

  2. Then go to Actions tab.

  3. Select a workflows according to your project suppose i created a ML model i selected a python workflows. You can also write a custom workflows just directly write a CML.yaml file click here for instructions.

  4. If you are using GitHub Actions no manual writing is required just select a workflow.

  5. Above below On after runs: be sure that you run install dependencies file.

    like : pip install -r requirments.txt Then step by step execute your file like this :

  6. Python main.py # suppose in this file you collect a data from online and export to csv.

Then you need to train your model.

  1. Python train.py #after running this file have some accuracy write in any file your accuracy.

  2. Automate your workflows and with any tool you like.

Note : For every push which is in your main or any other branch whenever a push in you repo automatically build start and jobs are in run for checking you can write your new model accuracy.

Did you find this article valuable?

Support Aman by becoming a sponsor. Any amount is appreciated!