MCPcopy Create free account
hub / github.com/apache/singa / compile

Method compile

examples/model_selection/Trails/singa_pkg_code/model.py:179–207  ·  view source on GitHub ↗

Compile and initialize the model This function will automatically derive the shape of parameters in each sublayer based on the shape of input placeholders. It will also do some settings. Args: inputs(list): the list of input tensors(placeholders)

(self, inputs, is_train=True, use_graph=False, sequential=False)

Source from the content-addressed store, hash-verified

177 self._results = None
178
179 def compile(self, inputs, is_train=True, use_graph=False, sequential=False):
180 """ Compile and initialize the model
181
182 This function will automatically derive the shape of parameters
183 in each sublayer based on the shape of input placeholders. It will
184 also do some settings.
185
186 Args:
187 inputs(list): the list of input tensors(placeholders)
188 is_train(bool): when is_trainis True, this model will enter
189 training mode, otherwise it will enter the evaluation mode
190 use_graph(bool): when use_graph is True, computational graph
191 will be used to train this model
192 sequential(bool): when sequential is True, model will execute ops
193 in the graph follow the order of joining the graph
194 """
195 assert len(inputs) > 0 and isinstance(inputs[0], Tensor), (
196 'compile function expects PlaceHolders or Tensors')
197
198 dev = inputs[0].device
199 dev.EnableGraph(True)
200 self.forward(*inputs)
201 dev.EnableGraph(False)
202 dev.ResetGraph()
203
204 autograd.training = is_train
205 self.training = is_train
206 self.graph_mode = use_graph
207 self.sequential = sequential
208
209 def forward(self, *input):
210 """Defines the computation performed in every forward propagation.

Callers 4

runFunction · 0.45
model.pyFile · 0.45
model.pyFile · 0.45
model.pyFile · 0.45

Calls 3

forwardMethod · 0.95
EnableGraphMethod · 0.80
ResetGraphMethod · 0.80

Tested by

no test coverage detected