Module: tfr.extension.pipeline

Stay organized with collections Save and categorize content based on your preferences.

Provides a RankingPipeline for running a TF-Ranking model.

This class contains the boilerplate required to run a TF-Ranking model, which reduces a few replicated setups (e.g., input function, serving input function, model export strategies) for running TF-Ranking models. Advanced users can also derive from this class and further tailor for their needs.


class RankingPipeline: Class to set up the input, train and eval processes for a TF Ranking model.

absolute_import Instance of __future__._Feature
division Instance of __future__._Feature
print_function Instance of __future__._Feature