Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore the evolving landscape of machine learning workflows and tools in this 37-minute conference talk by Dmitry Petrov and Ruslan Kuprieiev from Iterative AI. Delve into the unique challenges faced by ML teams, including data versioning, pipeline versioning, and experiment metrics visualization. Examine the limitations of traditional software engineering tools like Git and Git-LFS in addressing ML-specific needs. Discover the motivation behind developing new ML-focused experiment and data management systems. Learn about the open-source tool DVC (Data Version Control) and its approach to managing ML experiments, large datasets, and models. Gain insights into the distinct nature of ML workflows compared to traditional software engineering, and understand the importance of specialized tools in supporting the trial-and-error nature of ML projects.
