Module: tf.contrib.autograph

Defined in tensorflow/contrib/autograph/

This is the legacy module for AutoGraph, kept for backward compatibility.

New users should instead use tensorflow.python.autograph.


class AutographParseError: Invalid syntax.

class ConversionOptions: Immutable container for global conversion flags.

class Feature: Represents conversion options that can be toggled on or off.

class GraphConstructionError: Error for graph construction errors from AutoGraph generated code.

class RunMode: Specifies the way a converted function or method should be executed in TF.

class TfRuntimeError: Error wrapper for runtime errors raised by AutoGraph generated code.

class Verbosity: Represents conversion verbosity levels.


convert(...): Decorator that compiles a function to use TensorFlow ops.

converted_call(...): Compiles a function call inline. For internal use only.

do_not_convert(...): Decorator that suppresses the conversion of a function.

improved_errors(...): Context manager that rewrites runtime errors.

set_element_type(...): Indicates that the entity is expected hold items of specified type/shape.

set_loop_options(...): Specifies additional arguments to be passed to the enclosing while_loop.

stack(...): Stacks the input, if it admits the notion of stacking.

tensor_list(...): Creates an tensor list and populates it with the given elements.

to_code(...): Similar to to_graph, but returns Python source code as a string.

to_graph(...): Converts a Python entity into a TensorFlow graph.

Other Members