DAGsHub · raw details

DAGsHub is a web platform for data version control and collaboration for data scientists and machine learning engineers. The platform is built on DVC, an open-source version control system for machine learning projects that works seamlessly with Git. The platform is free for open-source projects and completely agnostic to the users ecosystem, programming language, and libraries. The DAGsHub platform enables the reproduction of any experiment from previous versions of the users data pipeline. It also allows for experimentation with different hyperparameters in parallel, with each run automatically saving its results in an organized, easy-to-access location. The platform also allows users to maintain giant spreadsheets of hyperparameters and to collaborate more effectively by leveraging the same best practices used in software engineering to get more done. DAGsHub also helps users automate their workflow and share intermediate results from the pipeline with any collaborator, rerunning only those parts of the pipeline that the user has changed instead of rerunning all of the preprocessing code.