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

Method forward

python/singa/model.py:186–197  ·  view source on GitHub ↗

Defines the computation performed in every forward propagation. Should be overridden by all subclasses. Args: *input: the input training data for the model Returns: out: the outputs of the forward propagation.

(self, *input)

Source from the content-addressed store, hash-verified

184 self.sequential = sequential
185
186 def forward(self, *input):
187 """Defines the computation performed in every forward propagation.
188
189 Should be overridden by all subclasses.
190
191 Args:
192 *input: the input training data for the model
193
194 Returns:
195 out: the outputs of the forward propagation.
196 """
197 raise NotImplementedError
198
199 def train_one_batch(self, *input, **kwargs):
200 """Defines the computation performed in every training iteration

Callers 2

compileMethod · 0.95
__call__Method · 0.95

Calls

no outgoing calls

Tested by

no test coverage detected