Our ML Tooling 2023 - Part 3: Modeling


8 min read

Fig 1. Our updated ML Tooling diagram, zoomed in on the Model section.

Fig 2. Check out this Tensorflow tutorial, served from GitHub pages and built with Jupyter Book.

  • 1Run unique and informative experiments

  • 2Reproduce those same experiments, including producing evaluation metrics and model artifacts

  • 3Store model artifacts and record their versions for internal use and comparison

  • 4Serve models in performant applications that offer easy deployment and flexibility for models with various compute resource requirements and detection types

What we're doing.