Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[CheckpointSaver] Add saving listeners support for increment checkpoint saver. #915

Merged
merged 1 commit into from
Jul 26, 2023
Merged

[CheckpointSaver] Add saving listeners support for increment checkpoint saver. #915

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 13 additions & 3 deletions tensorflow/python/training/basic_session_run_hooks.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -610,9 +610,7 @@ def after_run(self, run_context, run_values):
global_step = run_context.session.run(self._global_step_tensor)
if self._incremental_timer.should_trigger_for_step(global_step):
self._incremental_timer.update_last_triggered_step(global_step)
logging.info("Start Save incremental checkpoints for %d into %s.", global_step, self._incremental_save_path)
self._get_incr_saver().incremental_save(run_context.session, self._incremental_save_path, global_step=global_step)
logging.info("Finish Save incremental checkpoints for %d into %s.", global_step, self._incremental_save_path)
self._incr_save(run_context.session, global_step)


def end(self, session):
Expand Down Expand Up @@ -666,6 +664,18 @@ def _get_saver(self):
self._saver = savers[0]
return savers[0]

def _incr_save(self, session, step):
logging.info("Saving incremental checkpoints for %d into %s.", step,
self._incremental_save_path)
for l in self._listeners:
l.before_save(session, step)

self._get_incr_saver().incremental_save(session,
self._incremental_save_path,
global_step=step)
for l in self._listeners:
l.after_save(session, step)

def _get_incr_saver(self):
if self._scaffold is not None:
return self._scaffold._incr_saver
Expand Down
7 changes: 6 additions & 1 deletion tensorflow/python/training/monitored_session.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -491,7 +491,8 @@ def MonitoredTrainingSession(
save_checkpoint_steps=USE_DEFAULT,
summary_dir=None,
save_incremental_checkpoint_secs=None,
target_nodes_or_tensors=None):
target_nodes_or_tensors=None,
saving_listeners=None):

"""Creates a `MonitoredSession` for training.

Expand Down Expand Up @@ -548,6 +549,9 @@ def MonitoredTrainingSession(
summaries. If None, checkpoint_dir is used instead.
target_nodes_or_tensors: list of tf.Tensor or tf.Operation indicates
targets, which determine graph transformation of 'smart-stage'
saving_listeners: List of `CheckpointSaverListener` subclass instances. Used
for callbacks that run immediately before or after this hook saves the
checkpoint.

Returns:
A `MonitoredSession` object.
Expand Down Expand Up @@ -648,6 +652,7 @@ def MonitoredTrainingSession(
save_steps=save_checkpoint_steps,
save_secs=save_checkpoint_secs,
scaffold=scaffold,
listeners=saving_listeners,
incremental_save_secs=save_incremental_checkpoint_secs))

if hooks:
Expand Down