TF 2.0 is out! Get hands-on practice at TF World, Oct 28-31. Use code TF20 for 20% off select passes. Register now


View source on GitHub

Get a function that returns a SupervisedInputReceiver matching an input_fn.


Note that this function calls the input_fn in a local graph in order to extract features and labels. Placeholders are then created from those features and labels in the default graph.


  • input_fn: An Estimator input_fn, which is a function that returns one of:

    • A '' object: Outputs of Dataset object must be a tuple (features, labels) with same constraints as below.
    • A tuple (features, labels): Where features is a Tensor or a dictionary of string feature name to Tensor and labels is a Tensor or a dictionary of string label name to Tensor. Both features and labels are consumed by model_fn. They should satisfy the expectation of model_fn from inputs.
  • **input_fn_args: set of kwargs to be passed to the input_fn. Note that these will not be checked or validated here, and any errors raised by the input_fn will be thrown to the top.


A function taking no arguments that, when called, returns a SupervisedInputReceiver. This function can be passed in as part of the input_receiver_map when exporting SavedModels from Estimator with multiple modes.