The latest articles on DEV Community by Akash Desarda (@akashdesarda). https://dev.to/akashdesarda https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F217930%2Fa22567db-93a3-433d-9859-3b2dc6dc3afe.png https://dev.to/akashdesarda en Akash Desarda Thu, 26 Dec 2019 06:14:44 +0000 https://dev.to/akashdesarda/version-control-system-for-data-using-dvc-20fd https://dev.to/akashdesarda/version-control-system-for-data-using-dvc-20fd <p>One of the biggest pain working on a machine learning project is <strong>data management</strong>. </p> <p>Using <em>VCS</em> (like git) helps a lot during the development phase. But git doesn't work for the large dataset (in the likes of 100s of gbs). This problem can be fixed using <a href="https://dvc.org">DVC</a>. <strong>DVC</strong> is like git for large dataset. This is not even the best part of DVC, in fact, the best part is it completely work like git (conceptually, syntactically)</p> <p>DVC has excellent &amp; intuitive <a href="https://dvc.org/doc">Documentation</a>. I would say it is one of the best out there.</p> <p>Follow this <a href="https://dvc.org/doc/tutorials">Official tutorial</a> guide to getting started.</p> <p>DVC works with local file storage as well as cloud storage like AWS S3, GCP storage, etc</p> machinelearning dvc data git