def __init__(self,
mode,
graph_fn,
loss=None,
optimizer=None,
summaries='all',
metrics=None,
clip_gradients=0.5,
clip_embed_gradients=0.1,
one_hot_encode=None,
n_classes=None,
name="Classifier"):
if one_hot_encode and (n_classes is None
or not isinstance(n_classes, int)):
raise ValueError(
'`n_classes` must be an integer non negative value '
'when `one_hot_encode` is set to `True`, '
'received instead: {}'.format(n_classes))
self.one_hot_encode = one_hot_encode
self.n_classes = n_classes
loss = loss or SigmoidCrossEntropyConfig()
super(Classifier,
self).__init__(mode=mode,
name=name,
model_type=self.Types.CLASSIFIER,
graph_fn=graph_fn,
loss=loss,
optimizer=optimizer,
metrics=metrics,
summaries=summaries,
clip_gradients=clip_gradients,
clip_embed_gradients=clip_embed_gradients)
def __init__(self,
mode,
encoder_fn,
decoder_fn,
bridge_fn,
loss=None,
optimizer=None,
summaries='all',
metrics=None,
clip_gradients=0.5,
clip_embed_gradients=0.1,
name="Generator"):
optimizer = optimizer or AdadeltaConfig(learning_rate=0.4)
loss = loss or SigmoidCrossEntropyConfig()
self._check_subgraph_fn(function=encoder_fn, function_name='encoder_fn')
self._check_subgraph_fn(function=decoder_fn, function_name='decoder_fn')
self._check_bridge_fn(function=bridge_fn)
self._encode_fn = encoder_fn
self._decoder_fn = decoder_fn
self._bridge_fn = bridge_fn
graph_fn = self._build_graph_fn()
super(Generator, self).__init__(
mode=mode,
name=name,
model_type=self.Types.GENERATOR,
graph_fn=graph_fn,
loss=loss,
optimizer=optimizer,
metrics=metrics,
summaries=summaries,
clip_gradients=clip_gradients,
clip_embed_gradients=clip_embed_gradients)
def test_sigmoid_crossentropy_loss_config(self):
config_dict = {
'input_layer': 'images',
'output_layer': 'relu_1',
'label_smoothing': 0.,
'weights': 1.0,
'name': 'l',
'collect': False
}
config = SigmoidCrossEntropyConfig.from_dict(config_dict)
self.assert_equal_losses(config.to_dict(), config_dict)
def model_fn(features, labels, params, mode, config):
model = plx.models.Classifier(mode=mode,
graph_fn=graph_fn,
loss=SigmoidCrossEntropyConfig(),
optimizer=AdamConfig(learning_rate=0.001),
metrics=[AccuracyConfig()],
summaries='all',
one_hot_encode=True,
n_classes=10)
return model(features=features,
labels=labels,
params=params,
config=config)