tf.train.CheckpointOptions

Options for constructing a Checkpoint.

tf.train.CheckpointOptions(
    experimental_io_device=None, experimental_enable_async_checkpoint=False
)

Used as the options argument to either tf.train.Checkpoint.save() or tf.train.Checkpoint.restore() methods to adjust how variables are saved/restored.

Example: Run IO ops on "localhost" while saving a checkpoint:

step = tf.Variable(0, name="step")
checkpoint = tf.train.Checkpoint(step=step)
options = tf.train.CheckpointOptions(experimental_io_device="/job:localhost")
checkpoint.save("/tmp/ckpt", options=options)

experimental_io_device string. Applies in a distributed setting. Tensorflow device to use to access the filesystem. If None (default) then for each variable the filesystem is accessed from the CPU:0 device of the host where that variable is assigned. If specified, the filesystem is instead accessed from that device for all variables.

This is for example useful if you want to save to a local directory, such as "/tmp" when running in a distributed setting. In that case pass a device for the host where the "/tmp" directory is accessible.

experimental_enable_async_checkpoint bool Type. Indicates whether async checkpoint is enabled. Default is False, i.e., no async checkpoint.

Async checkpoint moves the checkpoint file writing off the main thread, so that the model can continue to train while the checkpoing file writing runs in the background. Async checkpoint reduces TPU device idle cycles and speeds up model training process, while memory consumption may increase.

experimental_enable_async_checkpoint

experimental_io_device