Class GrpcDebugHook
Inherits From: SessionRunHook
Defined in tensorflow/python/debug/wrappers/hooks.py
.
A hook that streams debugger-related events to any grpc_debug_server.
For example, the debugger data server is a grpc_debug_server. The debugger data server writes debugger-related events it receives via GRPC to logdir. This enables debugging features in Tensorboard such as health pills.
When the arguments of debug_utils.watch_graph changes, strongly consider changing arguments here too so that features are available to tflearn users.
Can be used as a hook for tf.train.MonitoredSession
s and
tf.estimator.Estimator
s.
__init__
__init__(
grpc_debug_server_addresses,
watch_fn=None,
thread_name_filter=None,
log_usage=True
)
Constructs a GrpcDebugHook.
Args:
grpc_debug_server_addresses
: (list
ofstr
) A list of the gRPC debug server addresses, in the format of host:port, with or without the "grpc://" prefix. For example: ["localhost:7000", "192.168.0.2:8000"]watch_fn
: A function that allows for customizing which ops to watch at which specific steps. See doc ofdumping_wrapper.DumpingDebugWrapperSession.__init__
for details.thread_name_filter
: Regular-expression white list for threads on which the wrapper session will be active. See doc ofBaseDebugWrapperSession
for more details.log_usage
: (bool) Whether usage is to be logged.
Methods
tfdbg.GrpcDebugHook.after_create_session
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.
Args:
session
: A TensorFlow Session that has been created.coord
: A Coordinator object which keeps track of all threads.
tfdbg.GrpcDebugHook.after_run
after_run(
run_context,
run_values
)
Called after each call to run().
The run_values
argument contains results of requested ops/tensors by
before_run()
.
The run_context
argument is the same one send to before_run
call.
run_context.request_stop()
can be called to stop the iteration.
If session.run()
raises any exceptions then after_run()
is not called.
Args:
run_context
: ASessionRunContext
object.run_values
: A SessionRunValues object.
tfdbg.GrpcDebugHook.before_run
before_run(run_context)
Called right before a session is run.
Args:
run_context
: A session_run_hook.SessionRunContext. Encapsulates information on the run.
Returns:
A session_run_hook.SessionRunArgs object.
tfdbg.GrpcDebugHook.begin
begin()
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.
After the 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.
tfdbg.GrpcDebugHook.end
end(session)
Called at the end of session.
The session
argument can be used in case the hook wants to run final ops,
such as saving a last checkpoint.
If session.run()
raises exception other than OutOfRangeError or
StopIteration then end()
is not called.
Note the difference between end()
and after_run()
behavior when
session.run()
raises OutOfRangeError or StopIteration. In that case
end()
is called but after_run()
is not called.
Args:
session
: A TensorFlow Session that will be soon closed.