Slide 1

Slide 1 text

No content

Slide 2

Slide 2 text

MLOps Exploration with Git & DVC for Machine learning project on DagsHub

Slide 3

Slide 3 text

Who am I? - Sport Data Scientist (Basketball). - Founder of The African Data Community Newsletter. - Technical Writer, Public Speaker, Social Media Content creator, Machine Learning Thought Leader (Global AI Hub), Open-source & Community Advocate. Gift Ojeabulu CBBAnalytics @GiftOjeabulu_

Slide 4

Slide 4 text

Learning Objective - A prelude to MLops. - DagsHub & DVC Workflow. - A Diagrammatic explanation of Code, Model, Data parameters & metrics in respect to DVC & DagsHub. - Connect, Migrate, Create your DagsHub Repo. - A practical Guide to build your first MLops project with DVC & DagsHub. - Conclusion.

Slide 5

Slide 5 text

A prelude to MLOps.

Slide 6

Slide 6 text

An Overture - What is MLOps - Why MLOps. - CRISP-DM VS ASUM-DM Methodology - MLops tools Categorization/class. - The Future of MLOps.

Slide 7

Slide 7 text

Open Source Festival 2022 What is MLOps? MLOps is a set of practices that aims to deploy and maintain machine learning models in production reliably and efficiently. The word is a compound of "machine learning" and the continuous development practice of DevOps in the software development field.

Slide 8

Slide 8 text

Open Source Festival 2022 Why MLOps? MLOps is a set of practices for collaboration and communication between data scientists and operations professionals. Applying these practices increases the quality, simplifies the management process, and automates the deployment of Machine Learning and Deep Learning models in large-scale production environments.

Slide 9

Slide 9 text

Open Source Festival 2022 The Future of MLOps MLOps is the future of machine learning, and it brings a host of benefits to organizations looking to deliver high-quality models continuously. It also offers many other benefits to organizations, including improved collaboration between data scientists and developers, faster time-to-market for new models, and increased model accuracy.

Slide 10

Slide 10 text

DVC is to Machine learning Engineers what Git is to Software Engineers.

Slide 11

Slide 11 text

Photo comes here DVC & DagsHub Workflow Data Version Control) is an ideal tool for versioning your binary data. However, you cannot view the DVC tracked files on GitHub.Wouldn’t it be nice if there is a platform that is like GitHub, but is more optimized for data scientists and machine learning engineers? That is when DagsHub comes in handy Git - DVC - DagsHub

Slide 12

Slide 12 text

DagsHub is to Machine learning Engineers what GitHub is to Software Engineers.

Slide 13

Slide 13 text

Photo comes here Code, Model, Data parameters & metrics As a Data Scientist you might want to version your code, model, data and metrics so you can reproduce a certain experiment. Github is a great platform to version control your code but not ideal to version your data, model and metrics. DagsHub is ideal. DagsHub is Github Supplement

Slide 14

Slide 14 text

Open Source Festival 2022 Connect your Repo If you want to manage your repository through both GitHub and DagsHub, Connect A Repo.

Slide 15

Slide 15 text

Migrate your Repo If you already have a repository on GitHub and you want to migrate your repository to DagsHub, Migrate A Repo.

Slide 16

Slide 16 text

Create your Repo DagsHub allows you to either create a new repository on their platform or connect to a repository on GitHub. If you want to create a new repository on DagsHub, click New Repository.

Slide 17

Slide 17 text

What Next?

Slide 18

Slide 18 text

Go ye therefore & build your first MLOps project with DVC & DagsHub 1. Khuyen Tran Article 2. Abid Ali Awan Article. 3. Dagshub Documentation

Slide 19

Slide 19 text

In Conclusion

Slide 20

Slide 20 text

Alessya Visnjic Data Scientist need to think about their models in post-production because only when the model is in production is when it starts generating value. What the CEO of Whylabs said at DataFramed Podcast about MLOps…

Slide 21

Slide 21 text

Thank you! Gift Ojeabulu CBB Analytics @GiftOjeabulu_

Slide 22

Slide 22 text

No content