Skip to content

Instantly share code, notes, and snippets.

@javidcf
Last active April 20, 2018 05:13
Show Gist options
  • Save javidcf/445d0578e9e749af5779880393e90853 to your computer and use it in GitHub Desktop.
Save javidcf/445d0578e9e749af5779880393e90853 to your computer and use it in GitHub Desktop.
Freeze an active TensorFlow session
def freeze_session(session, keep_var_names=None, output_names=None, clear_devices=True):
"""
Freezes the state of a session into a prunned computation graph.
Creates a new computation graph where variable nodes are replaced by
constants taking their current value in the session. The new graph will be
prunned so subgraphs that are not neccesary to compute the requested
outputs are removed.
@param session The TensorFlow session to be frozen.
@param keep_var_names A list of variable names that should not be frozen,
or None to freeze all the variables in the graph.
@param output_names Names of the relevant graph outputs.
@param clear_devices Remove the device directives from the graph for better portability.
@return The frozen graph definition.
"""
from tensorflow.python.framework.graph_util import convert_variables_to_constants
graph = session.graph
with graph.as_default():
freeze_var_names = list(set(v.op.name for v in tf.global_variables()).difference(keep_var_names or []))
output_names = output_names or []
output_names += [v.op.name for v in tf.global_variables()]
input_graph_def = graph.as_graph_def()
if clear_devices:
for node in input_graph_def.node:
node.device = ""
frozen_graph = convert_variables_to_constants(session, input_graph_def,
output_names, freeze_var_names)
return frozen_graph
@gandhivaibhav1996
Copy link

what is the name of the input variable and output variable for given frozen_graph?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment