See the guide: Training > Training Hooks
Saves summaries every N steps.
__init__( save_steps=None, save_secs=None, output_dir=None, summary_writer=None, scaffold=None, summary_op=None )
int, save summaries every N steps. Exactly one of
save_stepsshould be set.
int, save summaries every N seconds.
string, the directory to save the summaries to. Only used if no
output_dirwas passed, one will be created accordingly.
Scaffoldto get summary_op if it's not provided.
stringcontaining the serialized
Summaryprotocol buffer or a list of
Tensor. They are most likely an output by TF summary methods like
tf.summary.merge_all. It can be passed in as one tensor; if more than one, they must be passed in as a list.
ValueError: Exactly one of scaffold or summary_op should be set.
after_create_session( session, coord )
Called when new TensorFlow session is created.
This is called to signal the hooks that a new session has been created. This
has two essential differences with the situation in which
begin is called:
- When this is called, the graph is finalized and ops can no longer be added to the graph.
- This method will also be called as a result of recovering a wrapped session, not only at the beginning of the overall session.
session: A TensorFlow Session that has been created.
coord: A Coordinator object which keeps track of all threads.
after_run( run_context, run_values )
Called after each call to run().
run_values argument contains results of requested ops/tensors by
run_context argument is the same one send to
run_context.request_stop() can be called to stop the iteration.
session.run() raises any exceptions then
after_run() is not called.
run_values: A SessionRunValues object.
Called before each call to run().
You can return from this call a
SessionRunArgs object indicating ops or
tensors to add to the upcoming
run() call. These ops/tensors will be run
together with the ops/tensors originally passed to the original run() call.
The run args you return can also contain feeds to be added to the run()
run_context argument is a
SessionRunContext that provides
information about the upcoming
run() call: the originally requested
op/tensors, the TensorFlow Session.
At this point graph is finalized and you can not add ops.
None or a
Called once before using the session.
When called, the default graph is the one that will be launched in the
session. The hook can modify the graph by adding new operations to it.
begin() call the graph will be finalized and the other callbacks
can not modify the graph anymore. Second call of
begin() on the same
graph, should not change the graph.
Called at the end of session.
session argument can be used in case the hook wants to run final ops,
such as saving a last checkpoint.
session.run() raises exception other than OutOfRangeError or
end() is not called.
Note the difference between
after_run() behavior when
session.run() raises OutOfRangeError or StopIteration. In that case
end() is called but
after_run() is not called.
session: A TensorFlow Session that will be soon closed.