Original Source Here
Version control your dataset with DVC
Use DVC and Git for tracking changes and test your machine learning dataset
Data Version Control (DVC) is one of the most amazing projects in recent years. Before using it, we used to have trouble reproducing our models and experiments. We store our images and annotations in high-volume network attached storage where multiple people work every day, so there was no proper way to modify images or annotations while maintaining a correct and reproducible change history. You can imagine how often someone accidentally deletes an image, modifies some annotations, or infinitely more random problems that ended in trouble for properly reproducing our experiments.
In this post I will try to show how to configure DVC and how it can help us to maintain version of our datasets that can be easily integrated with Github.
If you want to read more, please visit the full article on my personal blog!
Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot