Working in Jupyter Lab/Notebook on one’s own computer is fairly straightforward. When you add add more team members; however, the equation becomes more difficult. Keeping a shared set of utility functions, libraries, package versions, etc. standardized across multiple projects and people can lead to obstacles. By mapping out how the workflows can effectively coalesce to form a body of work beforehand and through continuous iterations, multiple team members can work in harmony to produce cohesive machine learning and natural language processing code, processes, and results. Above is a diagram showing the coordination of team members and various Google Cloud Platform tools for an ML/NLP project.
Mari Bolton 
You must be logged in to post a comment.