tf.train.experimental.PythonState

TensorFlow 1 version View source on GitHub

A mixin for putting Python state in an object-based checkpoint.

This is an abstract class which allows extensions to TensorFlow's object-based checkpointing (see tf.train.Checkpoint). For example a wrapper for NumPy arrays:

import io
import numpy

class NumpyWrapper(tf.train.experimental.PythonState):

  def __init__(self, array):
    self.array = array

  def serialize(self):
    string_file = io.BytesIO()
    try:
      numpy.save(string_file, self.array, allow_pickle=False)
      serialized = string_file.getvalue()
    finally:
      string_file.close()
    return serialized

  def deserialize(self, string_value):
    string_file = io.BytesIO(string_value)
    try:
      self.array = numpy.load(string_file, allow_pickle=False)
    finally:
      string_file.close()

Instances of NumpyWrapper are checkpointable objects, and will be saved and restored from checkpoints along with TensorFlow state like variables.

root = tf.train.Checkpoint(numpy=NumpyWrapper(numpy.array([1.])))
save_path = root.save(prefix)
root.numpy.array *= 2.
assert [2.] == root.numpy.array
root.restore(save_path)
assert [1.] == root.numpy.array

Methods

deserialize

View source

@abc.abstractmethod
deserialize(
    string_value
)

Callback to deserialize the object.

serialize

View source

@abc.abstractmethod
serialize()

Callback to serialize the object. Returns a string.