Python v1.slice方法代码示例

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def _build_tiled_linear(self, inputs, input_name_and_sizes,
                          output_name_and_sizes, add_bias):
    # pylint: disable=missing-docstring
    def split_output(output):
      if len(output_name_and_sizes) == 1:
        return output
      elif len(set([size for _, size in output_name_and_sizes])) == 1:
        # This is a bit faster than several tf.slice calls.
        return tf.split(output, len(output_name_and_sizes), axis=1)
      else:
        outputs = []
        offset = 0
        for _, output_size in output_name_and_sizes:
          outputs.append(tf.slice(output, [0, offset], [-1, output_size]))
          offset += output_size
        return outputs

    weights = self._ensure_weights()
    if len(inputs) > 1:
      inputs = tf.concat(inputs, 1)
    if add_bias:
      biases = self._ensure_biases()
      return split_output(tf.nn.xw_plus_b(inputs, weights, biases))
    else:
      return split_output(tf.matmul(inputs, weights)) 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def update_hparams_for_vq_gating(hparams):
  """VQ Gating hparams."""
  hparams.add_hparam("z_size", 4)
  hparams.add_hparam("noise_dev", 0.5)
  # Bottleneck kinds supported: dense, vae, dvq.
  hparams.add_hparam("bottleneck_kind", "dvq")
  hparams.add_hparam("num_blocks", 1)
  hparams.add_hparam("num_residuals", 1)
  # Reshape method for DVQ: slice, project
  hparams.add_hparam("beta", 0.25)
  hparams.add_hparam("epsilon", 1e-5)
  hparams.add_hparam("decay", 0.999)
  hparams.add_hparam("ema", False)  # default is false until ema is implemented
  hparams.add_hparam("random_top_k", 1)
  hparams.add_hparam("soft_em", False)
  hparams.add_hparam("num_samples", 10)
  hparams.add_hparam("gating_type", "vq")
  hparams.add_hparam("use_scales", int(True))
  hparams.add_hparam("residual_centroids", int(False)) 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def combine(self, expert_out, multiply_by_gates=True):
    """Sum together the expert output, weighted by the gates.

    The slice corresponding to a particular batch element `b` is computed
    as the sum over all experts `i` of the expert output, weighted by the
    corresponding gate values.  If `multiply_by_gates` is set to False, the
    gate values are ignored.

    Args:
      expert_out: a list of `num_experts` `Tensor`s, each with shape
        `[expert_batch_size_i, <extra_output_dims>]`.
      multiply_by_gates: a boolean

    Returns:
      a `Tensor` with shape `[batch_size, <extra_output_dims>]`.
    """
    # see comments on convert_gradient_to_tensor
    stitched = common_layers.convert_gradient_to_tensor(
        tf.concat(expert_out, 0))
    if multiply_by_gates:
      stitched *= tf.expand_dims(self._nonzero_gates, 1)
    combined = tf.unsorted_segment_sum(stitched, self._batch_index,
                                       tf.shape(self._gates)[0])
    return combined 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def _attention(self, query, attn_states):
    conv2d = tf.nn.conv2d
    reduce_sum = tf.math.reduce_sum
    softmax = tf.nn.softmax
    tanh = tf.math.tanh

    with tf.variable_scope("attention"):
      k = tf.get_variable("attn_w",
                          [1, 1, self._attn_size, self._attn_vec_size])
      v = tf.get_variable("attn_v", [self._attn_vec_size])
      hidden = tf.reshape(attn_states,
                          [-1, self._attn_length, 1, self._attn_size])
      hidden_features = conv2d(hidden, k, [1, 1, 1, 1], "SAME")
      if self._linear3 is None:
        self._linear3 = _Linear(query, self._attn_vec_size, True)
      y = self._linear3(query)
      y = tf.reshape(y, [-1, 1, 1, self._attn_vec_size])
      s = reduce_sum(v * tanh(hidden_features + y), [2, 3])
      a = softmax(s)
      d = reduce_sum(
          tf.reshape(a, [-1, self._attn_length, 1, 1]) * hidden, [1, 2])
      new_attns = tf.reshape(d, [-1, self._attn_size])
      new_attn_states = tf.slice(attn_states, [0, 1, 0], [-1, -1, -1])
      return new_attns, new_attn_states 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def assign_to_slices(self, assign_fn, values, assign_to_tensor_list=None):
      """Assign to the slice variables.

      Args:
        assign_fn: a function from
          (mtf.Variable, tf.Variable, tf.Tensor) -> tf.Operation
        values: a list of tf.Tensor
        assign_to_tensor_list: an optional list of tf.Variable

      Returns:
        a tf.operation
      """
      if assign_to_tensor_list is None:
        assign_to_tensor_list = self._laid_out_tensor.all_slices
      # Handle both N -> 1 and N -> N cases.
      num_slices = min(len(assign_to_tensor_list), len(values))
      devices = [""] * num_slices
      return tf.group(
          mtf.parallel(devices, assign_fn,
                       [self._variable] * len(devices),
                       assign_to_tensor_list[:num_slices],
                       values[:num_slices])) 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def expanded_shape(orig_shape, start_dim, num_dims):
  """Inserts multiple ones into a shape vector.

  Inserts an all-1 vector of length num_dims at position start_dim into a shape.
  Can be combined with tf.reshape to generalize tf.expand_dims.

  Args:
    orig_shape: the shape into which the all-1 vector is added (int32 vector)
    start_dim: insertion position (int scalar)
    num_dims: length of the inserted all-1 vector (int scalar)
  Returns:
    An int32 vector of length tf.size(orig_shape) + num_dims.
  """
  with tf.name_scope('ExpandedShape'):
    start_dim = tf.expand_dims(start_dim, 0)  # scalar to rank-1
    before = tf.slice(orig_shape, [0], start_dim)
    add_shape = tf.ones(tf.reshape(num_dims, [1]), dtype=tf.int32)
    after = tf.slice(orig_shape, start_dim, [-1])
    new_shape = tf.concat([before, add_shape, after], 0)
    return new_shape 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def _minibatch_subsample_fn(self, inputs):
    """Randomly samples anchors for one image.

    Args:
      inputs: a list of 2 inputs. First one is a tensor of shape [num_anchors,
        num_classes] indicating targets assigned to each anchor. Second one
        is a tensor of shape [num_anchors] indicating the class weight of each
        anchor.

    Returns:
      batch_sampled_indicator: bool tensor of shape [num_anchors] indicating
        whether the anchor should be selected for loss computation.
    """
    cls_targets, cls_weights = inputs
    if self._add_background_class:
      # Set background_class bits to 0 so that the positives_indicator
      # computation would not consider background class.
      background_class = tf.zeros_like(tf.slice(cls_targets, [0, 0], [-1, 1]))
      regular_class = tf.slice(cls_targets, [0, 1], [-1, -1])
      cls_targets = tf.concat([background_class, regular_class], 1)
    positives_indicator = tf.reduce_sum(cls_targets, axis=1)
    return self._random_example_sampler.subsample(
        tf.cast(cls_weights, tf.bool),
        batch_size=None,
        labels=tf.cast(positives_indicator, tf.bool)) 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def _cross_suppression(boxes, box_slice, iou_threshold, inner_idx):
  """Bounding-boxes cross-suppression loop body.

  Args:
    boxes: A float Tensor of shape [1, anchors, 4], representing boxes.
    box_slice: A float Tensor of shape [1, _NMS_TILE_SIZE, 4], the box tile
      returned from last iteration
    iou_threshold: A scalar, representing IOU threshold.
    inner_idx: A scalar, representing inner index.

  Returns:
    boxes: A float Tensor of shape [1, anchors, 4], representing boxes.
    ret_slice: A float Tensor of shape [1, _NMS_TILE_SIZE, 4], the box tile
               after suppression
    iou_threshold: A scalar, representing IOU threshold.
    inner_idx: A scalar, inner index incremented.
  """
  new_slice = tf.slice(boxes, [0, inner_idx * _NMS_TILE_SIZE, 0],
                       [1, _NMS_TILE_SIZE, 4])
  iou = batch_iou(new_slice, box_slice)
  ret_slice = tf.expand_dims(
      tf.cast(tf.reduce_all(iou < iou_threshold, [1]), box_slice.dtype),
      2) * box_slice
  return boxes, ret_slice, iou_threshold, inner_idx + 1 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def __call__(self, inputs, state, scope=None):
    """Run this multi-layer cell on inputs, starting from state."""
    output = None
    with tf.variable_scope(scope or "skip_multi_rnn_cell"):
      cur_state_pos = 0
      cur_inp = inputs
      new_states = []
      for i, cell in enumerate(self._cells):
        with tf.variable_scope("cell_%d" % i):
          if self._state_is_tuple:
            if not nest.is_sequence(state):
              raise ValueError(
                  "Expected state to be a tuple of length %d, but received: %s"
                  % (len(self.state_size), state))
            cur_state = state[i]
          else:
            cur_state = tf.slice(
                state, [0, cur_state_pos], [-1, cell.state_size])
            cur_state_pos += cell.state_size
          cur_inp, new_state = cell(cur_inp, cur_state)
          new_states.append(new_state)
          if output is None:
            output = cur_inp
          else:
            output += cur_inp
    new_states = (tuple(new_states) if self._state_is_tuple else
                  tf.concat(new_states, 1))
    return output, new_states 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def __call__(self, inputs, state, scope=None):
    """Run this multi-layer cell on inputs, starting from state."""
    with tf.variable_scope(scope or "res_multi_rnn_cell"):
      cur_state_pos = 0
      cur_inp = inputs
      new_states = []
      for i, cell in enumerate(self._cells):
        with tf.variable_scope("cell_%d" % i):
          if self._state_is_tuple:
            if not nest.is_sequence(state):
              raise ValueError(
                  "Expected state to be a tuple of length %d, but received: %s"
                  % (len(self.state_size), state))
            cur_state = state[i]
          else:
            cur_state = tf.slice(
                state, [0, cur_state_pos], [-1, cell.state_size])
            cur_state_pos += cell.state_size
          cur_inp2, new_state = cell(cur_inp, cur_state)
          if i == 0:
            cur_inp = cur_inp2
          else:
            cur_inp = cur_inp + cur_inp2
          new_states.append(new_state)
    new_states = (tuple(new_states) if self._state_is_tuple else
                  tf.concat(new_states, 1))
    return cur_inp, new_states 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def _curvature_range(self):
    """Curvature range.

    Returns:
      h_max_t, h_min_t ops
    """
    self._curv_win = tf.get_variable("curv_win",
                                     dtype=tf.float32,
                                     trainable=False,
                                     shape=[self.curvature_window_width,],
                                     initializer=tf.zeros_initializer)
    # We use log smoothing for curvature range
    self._curv_win = tf.scatter_update(self._curv_win,
                                       self._step % self.curvature_window_width,
                                       tf.log(self._grad_norm_squared))
    # Note here the iterations start from iteration 0
    valid_window = tf.slice(self._curv_win,
                            tf.constant([0,]),
                            tf.expand_dims(
                                tf.minimum(
                                    tf.constant(self.curvature_window_width),
                                    self._step + 1), dim=0))
    self._h_min_t = tf.reduce_min(valid_window)
    self._h_max_t = tf.reduce_max(valid_window)

    curv_range_ops = []
    with tf.control_dependencies([self._h_min_t, self._h_max_t]):
      avg_op = self._moving_averager.apply([self._h_min_t, self._h_max_t])
      with tf.control_dependencies([avg_op]):
        self._h_min = tf.exp(
            tf.identity(self._moving_averager.average(self._h_min_t)))
        self._h_max = tf.exp(
            tf.identity(self._moving_averager.average(self._h_max_t)))
        if self._sparsity_debias:
          self._h_min *= self._sparsity_avg
          self._h_max *= self._sparsity_avg
    curv_range_ops.append(avg_op)
    return curv_range_ops  # h_max_t, h_min_t 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def call_fake_controller(push_values, pop_values, write_values, output_values):
  """Mock a RNN controller from a set of expected outputs.

  Args:
    push_values: Expected controller push values.
    pop_values: Expected controller pop values.
    write_values: Expected controller write values.
    output_values: Expected controller output values.

  Returns:
    A callable which behaves like the call method of an NeuralStackCell.
  """
  def call(cell, inputs, prev_read_values, controller_state, batch_size):
    del inputs
    del prev_read_values
    del batch_size
    next_step = tf.constant(0)
    if hasattr(cell, "current_step"):
      next_step = tf.assign_add(cell.current_step, tf.constant(1))
    return neural_stack.NeuralStackControllerInterface(
        push_strengths=tf.slice(tf.constant(push_values),
                                [next_step, 0, 0, 0],
                                [1, -1, -1, -1]),
        pop_strengths=tf.slice(tf.constant(pop_values),
                               [next_step, 0, 0, 0],
                               [1, -1, -1, -1]),
        write_values=tf.slice(tf.constant(write_values),
                              [next_step, 0, 0],
                              [1, -1, -1]),
        outputs=tf.slice(tf.constant(output_values),
                         [next_step, 0, 0],
                         [1, -1, -1]),
        state=controller_state
    )
  return call 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def tpu_conv1d(inputs, filters, kernel_size, padding="SAME", name="tpu_conv1d"):
  """Version of conv1d that works on TPU (as of 11/2017).

  Args:
    inputs: a Tensor with shape [batch, length, input_depth].
    filters: an integer.
    kernel_size: an integer.
    padding: a string - "SAME" or "LEFT".
    name: a string.

  Returns:
    a Tensor with shape [batch, length, filters].
  """
  if kernel_size == 1:
    return dense(inputs, filters, name=name, use_bias=True)
  if padding == "SAME":
    assert kernel_size % 2 == 1
    first_offset = -((kernel_size - 1) // 2)
  else:
    assert padding == "LEFT"
    first_offset = -(kernel_size - 1)
  last_offset = first_offset + kernel_size - 1
  results = []
  padded = tf.pad(inputs, [[0, 0], [-first_offset, last_offset], [0, 0]])
  for i in range(kernel_size):
    shifted = tf.slice(padded, [0, i, 0], tf.shape(inputs)) if i else inputs
    shifted.set_shape(inputs.get_shape())
    results.append(
        dense(shifted, filters, use_bias=(i == 0), name=name + "_%d" % i))
  ret = tf.add_n(results)
  ret *= kernel_size**-0.5
  return ret 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def dna_transformation(prev_image, dna_input, dna_kernel_size, relu_shift):
  """Apply dynamic neural advection to previous image.

  Args:
    prev_image: previous image to be transformed.
    dna_input: hidden lyaer to be used for computing DNA transformation.
    dna_kernel_size: dna kernel size.
    relu_shift: shift for ReLU function.
  Returns:
    List of images transformed by the predicted CDNA kernels.
  """
  # Construct translated images.
  prev_image_pad = tf.pad(prev_image, [[0, 0], [2, 2], [2, 2], [0, 0]])
  image_height = int(prev_image.get_shape()[1])
  image_width = int(prev_image.get_shape()[2])

  inputs = []
  for xkern in range(dna_kernel_size):
    for ykern in range(dna_kernel_size):
      inputs.append(
          tf.expand_dims(
              tf.slice(prev_image_pad, [0, xkern, ykern, 0],
                       [-1, image_height, image_width, -1]), [3]))
  inputs = tf.concat(axis=3, values=inputs)

  # Normalize channels to 1.
  kernel = tf.nn.relu(dna_input - relu_shift) + relu_shift
  kernel = tf.expand_dims(
      kernel / tf.reduce_sum(kernel, [3], keep_dims=True), [4])
  return tf.reduce_sum(kernel * inputs, [3], keep_dims=False) 

# 需要导入模块: from tensorflow.compat import v1 [as 别名]
# 或者: from tensorflow.compat.v1 import slice [as 别名]
def __init__(self, get_regularizer_to_slice, begin, size):
    """Creates an instance.

    Args:
      get_regularizer_to_slice: A callable, such that get_regularizer_to_slice()
        returns an OpRegularizer that has to be sliced.
      begin: An integer, where to begin the slice.
      size: An integer, the length of the slice (so the slice ends at
        begin + size).
    """
    self._get_regularizer_to_slice = get_regularizer_to_slice
    self._begin = begin
    self._size = size
    self._alive_vector = None
    self._regularization_vector = None