本文整理汇总了Python中tensorflow.python.pywrap_tensorflow.TF_ExtendGraph方法的典型用法代码示例。如果您正苦于以下问题：Python pywrap_tensorflow.TF_ExtendGraph方法的具体用法？Python pywrap_tensorflow.TF_ExtendGraph怎么用？Python pywrap_tensorflow.TF_ExtendGraph使用的例子？那么, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类tensorflow.python.pywrap_tensorflow的用法示例。
在下文中一共展示了pywrap_tensorflow.TF_ExtendGraph方法的2个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: _extend_graph
# 需要导入模块: from tensorflow.python import pywrap_tensorflow [as 别名]
# 或者: from tensorflow.python.pywrap_tensorflow import TF_ExtendGraph [as 别名]
def _extend_graph(self):
# Ensure any changes to the graph are reflected in the runtime.
with self._extend_lock:
if self._graph.version > self._current_version:
# pylint: disable=protected-access
graph_def, self._current_version = self._graph._as_graph_def(
from_version=self._current_version,
add_shapes=self._add_shapes)
# pylint: enable=protected-access
with errors.raise_exception_on_not_ok_status() as status:
tf_session.TF_ExtendGraph(
self._session, graph_def.SerializeToString(), status)
self._opened = True
# The threshold to run garbage collection to delete dead tensors. 示例2: _extend_graph
# 需要导入模块: from tensorflow.python import pywrap_tensorflow [as 别名]
# 或者: from tensorflow.python.pywrap_tensorflow import TF_ExtendGraph [as 别名]
def _extend_graph(self):
# Nothing to do if we're using the new session interface
# TODO(skyewm): remove this function altogether eventually
if self._created_with_new_api: return
# Ensure any changes to the graph are reflected in the runtime.
with self._extend_lock:
if self._graph.version > self._current_version:
# pylint: disable=protected-access
graph_def, self._current_version = self._graph._as_graph_def(
from_version=self._current_version,
add_shapes=self._add_shapes)
# pylint: enable=protected-access
with errors.raise_exception_on_not_ok_status() as status:
tf_session.TF_ExtendGraph(
self._session, graph_def.SerializeToString(), status)
self._opened = True
# The threshold to run garbage collection to delete dead tensors. 开发者ID:PacktPublishing，项目名称:Serverless-Deep-Learning-with-TensorFlow-and-AWS-Lambda，代码行数:22，代码来源:session.py