Module: tf.estimator.experimental

Public API for tf.estimator.experimental namespace.


class InMemoryEvaluatorHook: Hook to run evaluation in training without a checkpoint.

class LinearSDCA: Stochastic Dual Coordinate Ascent helper for linear estimators.


build_raw_supervised_input_receiver_fn(...): Build a supervised_input_receiver_fn for raw features and labels.

call_logit_fn(...): Calls logit_fn (experimental).

dnn_logit_fn_builder(...): Function builder for a dnn logit_fn.

linear_logit_fn_builder(...): Function builder for a linear logit_fn.

make_early_stopping_hook(...): Creates early-stopping hook.

make_stop_at_checkpoint_step_hook(...): Creates a proper StopAtCheckpointStepHook based on chief status.

stop_if_higher_hook(...): Creates hook to stop if the given metric is higher than the threshold.