MCPcopy Create free account
hub / github.com/DeepRec-AI/DeepRec / push

Method push

tensorflow/python/eager/context.py:214–232  ·  view source on GitHub ↗

Push metadata about a context switch onto the stack. A context switch can take any one of the two forms: installing a graph as the default graph, or entering the eager context. For each context switch, we record whether or not the entered context is building a function. Args:

(self, is_building_function, enter_context_fn, device_stack)

Source from the content-addressed store, hash-verified

212 device_stack=None)
213
214 def push(self, is_building_function, enter_context_fn, device_stack):
215 """Push metadata about a context switch onto the stack.
216
217 A context switch can take any one of the two forms: installing a graph as
218 the default graph, or entering the eager context. For each context switch,
219 we record whether or not the entered context is building a function.
220
221 Args:
222 is_building_function: (bool.) Whether the context is building a function.
223 enter_context_fn: (function.) A callable that executes the context switch.
224 For example, `graph.as_default` or `eager_mode`.
225 device_stack: If applicable, the device function stack for this
226 graph. When breaking out of graphs in init_scope, the innermost nonempty
227 device stack is used. Eager contexts put `None` here and the value is
228 never used.
229 """
230
231 self.stack.append(
232 ContextSwitch(is_building_function, enter_context_fn, device_stack))
233
234 def pop(self):
235 """Pop the stack."""

Callers 3

__init__Method · 0.95
get_controllerMethod · 0.45
_modeMethod · 0.45

Calls 1

appendMethod · 0.45

Tested by

no test coverage detected