tf.train.latest_checkpoint  |  TensorFlow v2.16.1 (original) (raw)

tf.train.latest_checkpoint

Stay organized with collections Save and categorize content based on your preferences.

Finds the filename of latest saved checkpoint file.

View aliases

Compat aliases for migration

SeeMigration guide for more details.

tf.compat.v1.train.latest_checkpoint

tf.train.latest_checkpoint(
    checkpoint_dir, latest_filename=None
)

Used in the notebooks

Used in the guide Used in the tutorials
Migrating model checkpoints Training checkpoints Estimators Custom training with tf.distribute.Strategy Distributed training with Keras Custom training loop with Keras and MultiWorkerMirroredStrategy Multi-worker training with Keras Deep Convolutional Generative Adversarial Network

Gets the checkpoint state given the provided checkpoint_dir and looks for a corresponding TensorFlow 2 (preferred) or TensorFlow 1.x checkpoint path. The latest_filename argument is only applicable if you are saving checkpoint using v1.train.Saver.save

See the Training Checkpoints Guide for more details and examples.`

Args
checkpoint_dir Directory where the variables were saved.
latest_filename Optional name for the protocol buffer file that contains the list of most recent checkpoint filenames. See the corresponding argument to v1.train.Saver.save.
Returns
The full path to the latest checkpoint or None if no checkpoint was found.

Except as otherwise noted, the content of this page is licensed under the Creative Commons Attribution 4.0 License, and code samples are licensed under the Apache 2.0 License. For details, see the Google Developers Site Policies. Java is a registered trademark of Oracle and/or its affiliates. Some content is licensed under the numpy license.

Last updated 2024-04-26 UTC.