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

Method _get_callback_model

tensorflow/python/keras/engine/training.py:1764–1777  ·  view source on GitHub ↗

Returns the Callback Model for this Model.

(self)

Source from the content-addressed store, hash-verified

1762 return total_loss
1763
1764 def _get_callback_model(self):
1765 """Returns the Callback Model for this Model."""
1766
1767 if hasattr(self, '_replicated_model') and self._replicated_model:
1768 # When using training_distributed, we set the callback model
1769 # to an instance of the `DistributedModel` that we create in
1770 # the `compile` call. The `DistributedModel` is initialized
1771 # with the first replicated model. We need to set the callback
1772 # model to a DistributedModel to allow us to override saving
1773 # and loading weights when we checkpoint the model during training.
1774 return self._replicated_model
1775 if hasattr(self, 'callback_model') and self.callback_model:
1776 return self.callback_model
1777 return self
1778
1779 def _make_callback_model(self, grouped_model):
1780 first_replicated_model = self._distribution_strategy.unwrap(

Callers 1

configure_callbacksFunction · 0.80

Calls

no outgoing calls

Tested by

no test coverage detected