ml-pipelines
Summed up quite nicely by this image taken from the CD foundation, used in the Coursera [[ml-ops]] course:
Some examples of ML pipeline frameworks are like Tensorflow Extended or #tfx uses the Tensorflow ecosystem for end-to-end applications, e.g. Keras for models, Tensorflow Transform for feature engineering, and Tensorflow Serving for model serving.
Backlinks
ml-metadata
Part of the [[ml-pipelines]] process, we want to understand how raw data is transformed by the pipeline, including transformation + schema + model + metrics, part of *data provenance/lineage*.