Function conv2d_same has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def conv2d_same(inputs, num_outputs, kernel_size, stride, rate=1, scope=None):

Function preprocess_image has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def preprocess_image(image,

Function pix2pix_discriminator has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def pix2pix_discriminator(net, num_filters, padding=2, pad_mode='REFLECT',

Function _build_pnasnet_base has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def _build_pnasnet_base(images,

Avoid deeply nested control flow statements.

Open

            with self.test_session() as sess:
              tf.set_random_seed(0)
              inputs = create_test_input(1, height, width, 3)
              # Dense feature extraction followed by subsampling.
              output = resnet_utils.stack_blocks_dense(inputs,

Avoid deeply nested control flow statements.

Open

            if self._use_bounded_activation:
              h = tf.nn.relu6(h)

          # Add hiddenstate to the list of hiddenstates we can choose from
          net.append(h)

Avoid deeply nested control flow statements.

Open

            if output_stride is None:
              factor = 1
            else:
              factor = nominal_stride // output_stride
            output = resnet_utils.subsample(output, factor)

Avoid deeply nested control flow statements.

Open

          if output_stride % 2 != 0:
            raise ValueError('The output_stride needs to be a multiple of 2.')
          output_stride //= 2

Avoid deeply nested control flow statements.

Open

            if current_config[_INPUT] < 0:
              operation_input = features
            else:
              operation_input = branch_logits[current_config[_INPUT]]
            if current_config[_OP] == _CONV:

Avoid deeply nested control flow statements.

Open

            if current_config[_OP] == _CONV:
              if current_config[_KERNEL] == [1, 1] or current_config[
                  _KERNEL] == 1:
                branch_logits.append(
                    slim.conv2d(operation_input, depth, 1, scope=scope))

Function __call__ has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

  def __call__(self, net, scope, filter_scaling, stride, prev_layer, cell_num):

Function lenet has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def lenet(images, num_classes=10, is_training=False,

Function build_nasnet_mobile has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def build_nasnet_mobile(images, num_classes,

Function restore_model has a Cognitive Complexity of 8 (exceeds 5 allowed). Consider refactoring.

Open

def restore_model(sess, checkpoint_path, enable_ema=True):
  """Restore variables from the checkpoint into the provided session.

  Args:
    sess: A tensorflow session where the checkpoint will be loaded.

• Read up
  Read up

Function cifarnet has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def cifarnet(images, num_classes=10, is_training=False,

Function resolve_shape has a Cognitive Complexity of 8 (exceeds 5 allowed). Consider refactoring.

Open

def resolve_shape(tensor, rank=None, scope=None):
  """Fully resolves the shape of a Tensor.

  Use as much as possible the shape components already known during graph
  creation and resolve the remaining ones during runtime.

• Read up
  Read up

Function preprocess_for_train has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def preprocess_for_train(image,

Function _combine_unused_states has a Cognitive Complexity of 8 (exceeds 5 allowed). Consider refactoring.

Open

  def _combine_unused_states(self, net):
    """Concatenate the unused hidden states of the cell."""
    used_hiddenstates = self._used_hiddenstates

    final_height = int(net[-1].shape[2])

• Read up
  Read up

Function build_nasnet_large has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def build_nasnet_large(images, num_classes,

Function preprocess_for_train has 6 arguments (exceeds 4 allowed). Consider refactoring.

Open

def preprocess_for_train(image,