Anthill Inside 2019

A conference on AI and Deep Learning

Machine Learning Model and Dataset Versioning

Submitted by Kurian Benoy (@kurianbenoy) on Jul 18, 2019

Section: Crisp talk Technical level: Intermediate Session type: Lecture Status: Under evaluation

Abstract

In this talk we will discuss about the current practices of organizing ML projects using traditional open-source tool set like Git and Git-LFS as well as this tool set limitation.
Thereby motivation for developing new ML specific version control systems will be explained.

Currently the life-cycle of any Machine learning model goes through following process:

  • a ML practitioner tries out new image classification algorithm with input dataset
  • He tweaks algorithms, tries other ideas and fix bugs. All in local system
  • Some of her training data might require long runs, and may change code while weights remains same
  • She keeps around the model weights and evaluation scores for all her runs, and picks which weights to release as the final model once she’s out of time to run more experiments.
  • She publishes her results, with code and the trained weights.

Git can’t handle large amount of data of GB’s of size. While Git-LFS comes with the in-build difficulty of supporting only 2 GBs of data at the maximum.(github limitations)

Data Version Control or DVC.ORG is an open source, command-line tool written in Python. We will show how to version datasets with dozens of gigabytes of data and version ML models, how to use your favorite cloud storage (S3, GCS, or bare metal SSH server) as a data file backend and how to embrace the best engineering practices in your ML projects.

Talk Outline

  • Why ML is different?
  • Problems on using git
  • About using MLFlow
  • Problems with git-LFS
  • Solving machine reproducibility crisis using DVC
  • Creating robust pipelines and rapid experimentation using DVC
  • How can DVC be added to your workflow
  • Conclusion

Outline

Talk Outline

  • Why ML is different?
  • Problems on using git
  • About using MLFlow
  • Problems with git-LFS
  • Solving machine reproducibility crisis using DVC
  • Creating robust pipelines and rapid experimentation using DVC
  • How can DVC be added to your workflow
  • Conclusion

Speaker bio

Kurian Benoy is an open source contributor at CloudCV, DVC. He is the lead organiser of School of AI, Kochi and is an AI enthusiast working on Deep Learning and Computer Vision. Kurian is FOSSASIA Open TechNights WInner and gave a talk in FOSSASIA Open Tech submit about the [keralarescue.in team] (https://www.youtube.com/watch?v=2RzImb5JwMA).

Kurian has been contributing to DVC for the past few months and has been a top 10 contributor to DVC.org and made an introductory kaggle kernel about dvc

Links

Slides

https://docs.google.com/presentation/d/16mbu71NqNH6ULPJWSMDheYwolRrIn1sSLD8JYy9s4ks/edit?usp=sharing

Comments

{{ gettext('Login to leave a comment') }}

{{ gettext('You need to be a participant to comment.') }}

{{ formTitle }}
{{ gettext('Post a comment...') }}
{{ gettext('New comment') }}

{{ errorMsg }}