tf.train.Checkpoint

Manages saving/restoring trackable values to disk.

tf.train.Checkpoint(
    root=None, **kwargs
)

Used in the notebooks

Used in the guide Used in the tutorials

TensorFlow objects may contain trackable state, such as tf.Variables, tf.keras.optimizers.Optimizer implementations, tf.data.Dataset iterators, tf.keras.Layer implementations, or tf.keras.Model implementations. These are called trackable objects.

A Checkpoint object can be constructed to save either a single or group of trackable objects to a checkpoint file. It maintains a save_counter for numbering checkpoints.

Example:

model = tf.keras.Model(...)
checkpoint = tf.train.Checkpoint(model)

# Save a checkpoint to /tmp/training_checkpoints-{save_counter}. Every time
# checkpoint.save is called, the save counter is increased.
save_path = checkpoint.save('/tmp/training_checkpoints')

# Restore the checkpointed values to the `model` object.
checkpoint.restore(save_path)

Example 2:

import tensorflow as tf
import os

checkpoint_directory = "/tmp/training_checkpoints"
checkpoint_prefix = os.path.join(checkpoint_directory, "ckpt")

# Create a Checkpoint that will manage two objects with trackable state,
# one we name "optimizer" and the other we name "model".
checkpoint = tf.train.Checkpoint(optimizer=optimizer, model=model)
status = checkpoint.restore(tf.train.latest_checkpoint(checkpoint_directory))
for _ in range(num_training_steps):
  optimizer.minimize( ... )  # Variables will be restored on creation.
status.assert_consumed()  # Optional sanity checks.
checkpoint.save(file_prefix=checkpoint_prefix)

Checkpoint.save() and Checkpoint.restore() write and read object-based checkpoints, in contrast to TensorFlow 1.x's tf.compat.v1.train.Saver which writes and reads variable.name based checkpoints. Object-based checkpointing saves a graph of dependencies between Python objects (Layers, Optimizers, Variables, etc.) with named edges, and this graph is used to match variables when restoring a checkpoint. It can be more robust to changes in the Python program, and helps to support restore-on-create for variables.

Checkpoint objects have dependencies on the objects passed as keyword arguments to their constructors, and each dependency is given a name that is identical to the name of the keyword argument for which it was created. TensorFlow classes like Layers and Optimizers will automatically add dependencies on their own variables (e.g. "kernel" and "bias" for tf.keras.layers.Dense). Inheriting from tf.keras.Model makes managing dependencies easy in user-defined classes, since Model hooks into attribute assignment. For example:

class Regress(tf.keras.Model):

  def __init__(self):
    super().__init__()
    self.input_transform = tf.keras.layers.Dense(10)
    # ...

  def call(self, inputs):
    x = self.input_transform(inputs)
    # ...

This Model has a dependency named "input_transform" on its Dense layer, which in turn depends on its variables. As a result, saving an instance of Regress using tf.train.Checkpoint will also save all the variables created by the Dense layer.

When variables are assigned to multiple workers, each worker writes its own section of the checkpoint. These sections are then merged/re-indexed to behave as a single checkpoint. This avoids copying all variables to one worker, but does require that all workers see a common filesystem.

This function differs slightly from the Keras Model save_weights function. tf.keras.Model.save_weights creates a checkpoint file with the name specified in filepath, while tf.train.Checkpoint numbers the checkpoints, using filepath as the prefix for the checkpoint file names. Aside from this, model.save_weights() and tf.train.Checkpoint(model).save() are equivalent.

See the guide to training checkpoints for details.

root The root object to checkpoint. root may be a trackable object or WeakRef of a trackable object.
**kwargs Keyword arguments are set as attributes of this object, and are saved with the checkpoint. All kwargs must be trackable objects, or a nested structure of trackable objects (list, dict, or tuple).

ValueError If root or the objects in kwargs are not trackable. A ValueError is also raised if the root object tracks different objects from the ones listed in attributes in kwargs (e.g. root.child = A and tf.train.Checkpoint(root, child=B) are incompatible).

save_counter Incremented when save() is called. Used to number checkpoints.

Methods

read

View source

read(
    save_path, options=None
)

Reads a training checkpoint written with write.

Reads this Checkpoint and any objects it depends on.

This method is just like restore() but does not expect the save_counter variable in the checkpoint. It only restores the objects that the checkpoint already depends on.

The method is primarily intended for use by higher level checkpoint management utilities that use write() instead of save() and have their own mechanisms to number and track checkpoints.

Example usage:

# Create a checkpoint with write()
ckpt = tf.train.Checkpoint(v=tf.Variable(1.))
path = ckpt.write('/tmp/my_checkpoint')

# Later, load the checkpoint with read()
# With restore() assert_consumed() would have failed.
checkpoint.read(path).assert_consumed()

# You can also pass options to read(). For example this
# runs the IO ops on the localhost:
options = tf.train.CheckpointOptions(
    experimental_io_device="/job:localhost")
checkpoint.read(path, options=options)

Args
save_path The path to the checkpoint as returned by write.
options Optional tf.train.CheckpointOptions object.

Returns
A load status object, which can be used to make assertions about the status of a checkpoint restoration. See restore for details.

restore

View source

restore(
    save_path, options=None
)

Restores a training checkpoint.

Restores this Checkpoint and any objects it depends on.

This method is intended to be used to load checkpoints created by save(). For checkpoints created by write() use the read() method which does not expect the save_counter variable added by save().

restore() either assigns values immediately if variables to restore have been created already, or defers restoration until the variables are created. Dependencies added after this call will be matched if they have a corresponding object in the checkpoint (the restore request will queue in any trackable object waiting for the expected dependency to be added).

checkpoint = tf.train.Checkpoint( ... )
checkpoint.restore(path)

# You can additionally pass options to restore():
options = tf.CheckpointOptions(experimental_io_device="/job:localhost")
checkpoint.restore(path, options=options)

To ensure that loading is complete and no more deferred restorations will take place, use the assert_consumed() method of the status object returned by restore():

checkpoint.restore(path, options=options).assert_consumed()

The assert will raise an error if any Python objects in the dependency graph were not found in the checkpoint, or if any checkpointed values do not have a matching Python object.

Name-based tf.compat.v1.train.Saver checkpoints from TensorFlow 1.x can be loaded using this method. Names are used to match variables. Re-encode name-based checkpoints using tf.train.Checkpoint.save as soon as possible.

Loading from SavedModel checkpoints

To load values from a SavedModel, just pass the SavedModel directory to checkpoint.restore:

model = tf.keras.Model(...)
tf.saved_model.save(model, path)  # or model.save(path, save_format='tf')

checkpoint = tf.train.Checkpoint(model)
checkpoint.restore(path).expect_partial()

This example calls expect_partial() on the loaded status, since SavedModels saved from Keras often generates extra keys in the checkpoint. Otherwise, the program prints a lot of warnings about unused keys at exit time.

Args
save_path The path to the checkpoint, as returned by save or tf.train.latest_checkpoint. If the checkpoint was written by the name-based tf.compat.v1.train.Saver, names are used to match variables. This path may also be a SavedModel directory.
options Optional tf.train.CheckpointOptions object.

Returns
A load status object, which can be used to make assertions about the status of a checkpoint restoration.

The returned status object has the following methods:

  • assert_consumed(): Raises an exception if any variables are unmatched: either checkpointed values which don't have a matching Python object or Python objects in the dependency graph with no values in the checkpoint. This method returns the status object, and so may be chained with other assertions.

  • assert_existing_objects_matched(): Raises an exception if any existing Python objects in the dependency graph are unmatched. Unlike assert_consumed, this assertion will pass if values in the checkpoint have no corresponding Python objects. For example a tf.keras.Layer object which has not yet been built, and so has not created any variables, will pass this assertion but fail assert_consumed. Useful when loading part of a larger checkpoint into a new Python program, e.g. a training checkpoint with a tf.compat.v1.train.Optimizer was saved but only the state required for inference is being loaded. This method returns the status object, and so may be chained with other assertions.

  • assert_nontrivial_match(): Asserts that something aside from the root object was matched. This is a very weak assertion, but is useful for sanity checking in library code where objects may exist in the checkpoint which haven't been created in Python and some Python objects may not have a checkpointed value.

  • expect_partial(): Silence warnings about incomplete checkpoint restores. Warnings are otherwise printed for unused parts of the checkpoint file or object when the Checkpoint object is deleted (often at program shutdown).

Raises
NotFoundError if the a checkpoint or SavedModel cannot be found at save_path.

save

View source

save(
    file_prefix, options=None
)

Saves a training checkpoint and provides basic checkpoint management.

The saved checkpoint includes variables created by this object and any trackable objects it depends on at the time Checkpoint.save() is called.

save is a basic convenience wrapper around the write method, sequentially numbering checkpoints using save_counter and updating the metadata used by tf.train.latest_checkpoint. More advanced checkpoint management, for example garbage collection and custom numbering, may be provided by other utilities which also wrap write and read. (tf.train.CheckpointManager for example).

step = tf.Variable(0, name="step")
checkpoint = tf.train.Checkpoint(step=step)
checkpoint.save("/tmp/ckpt")

# Later, read the checkpoint with restore()
checkpoint.restore("/tmp/ckpt-1")

# You can also pass options to save() and restore(). For example this
# runs the IO ops on the localhost:
options = tf.train.CheckpointOptions(experimental_io_device="/job:localhost")
checkpoint.save("/tmp/ckpt", options=options)

# Later, read the checkpoint with restore()
checkpoint.restore("/tmp/ckpt-1", options=options)

Args
file_prefix A prefix to use for the checkpoint filenames (/path/to/directory/and_a_prefix). Names are generated based on this prefix and Checkpoint.save_counter.
options Optional tf.train.CheckpointOptions object.

Returns
The full path to the checkpoint.

sync

View source

sync()

Wait for any outstanding save or restore operations.

write

View source

write(
    file_prefix, options=None
)

Writes a training checkpoint.

The checkpoint includes variables created by this object and any trackable objects it depends on at the time Checkpoint.write() is called.

write does not number checkpoints, increment save_counter, or update the metadata used by tf.train.latest_checkpoint. It is primarily intended for use by higher level checkpoint management utilities. save provides a very basic implementation of these features.

Checkpoints written with write must be read with read.

Example usage:

step = tf.Variable(0, name="step")
checkpoint = tf.Checkpoint(step=step)
checkpoint.write("/tmp/ckpt")

# Later, read the checkpoint with read()
checkpoint.read("/tmp/ckpt")

# You can also pass options to write() and read(). For example this
# runs the IO ops on the localhost:
options = tf.CheckpointOptions(experimental_io_device="/job:localhost")
checkpoint.write("/tmp/ckpt", options=options)

# Later, read the checkpoint with read()
checkpoint.read("/tmp/ckpt", options=options)

Args
file_prefix A prefix to use for the checkpoint filenames (/path/to/directory/and_a_prefix).
options Optional tf.train.CheckpointOptions object.

Returns
The full path to the checkpoint (i.e. file_prefix).