Python v1.TensorArray方法代碼示例

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_write_read():
    def run(dtype_str, infer_shape, element_shape):
        with tf.Graph().as_default():
            dtype = tf_dtypes[dtype_str]
            np_data = np.array([[1.0, 2.0], [3.0, 4.0]]).astype(dtype_str)
            in_data = [np_data, np_data]
            t1 = tf.constant(np_data, dtype=dtype)
            t2 = tf.constant(np_data, dtype=dtype)
            ta1 = tf.TensorArray(dtype=dtype, size=2, infer_shape=infer_shape,
                                 element_shape=element_shape)
            ta2 = ta1.write(0, t1)
            ta3 = ta2.write(1, t2)
            out = ta3.read(0)
            g = tf.get_default_graph()
            compare_tf_with_tvm([], [], 'TensorArrayReadV3:0', mode='vm')

    for dtype in ["float32", "int8"]:
        run(dtype, False, None)
        run(dtype, False, tf.TensorShape([None, 2]))
        run(dtype, True, None) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_scatter():
    def run(dtype_str, infer_shape):
        with tf.Graph().as_default():
            dtype =  tf_dtypes[dtype_str]
            if infer_shape:
                element_shape = tf.TensorShape([tf.Dimension(None)])
            else:
                element_shape = None
            t = tf.constant(np.array([[1.0], [2.0], [3.0]]).astype(dtype_str), dtype=dtype)
            indices = tf.constant([2, 1, 0])
            ta1 = tf.TensorArray(dtype=dtype, size=3,
                                 infer_shape=infer_shape,
                                 element_shape=element_shape)
            ta2 = ta1.scatter(indices, t)
            out0 = ta2.read(0)
            out1 = ta2.read(1)
            out2 = ta2.read(2)
            g = tf.get_default_graph()
            compare_tf_with_tvm([], [], ['TensorArrayReadV3:0'], mode='vm')
            compare_tf_with_tvm([], [], ['TensorArrayReadV3_1:0'], mode='vm')
            compare_tf_with_tvm([], [], ['TensorArrayReadV3_2:0'], mode='vm')
    for dtype in ["float32", "int8"]:
        run(dtype, False)
        run(dtype, True) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_split():
    def run(dtype_str, infer_shape):
        with tf.Graph().as_default():
            dtype =  tf_dtypes[dtype_str]
            t = tf.constant(np.array([[1.0], [2.0], [3.0], [4.0], [5.0], [6.0], [7.0], [8.0]]).astype(dtype_str), dtype=dtype)
            split_length = tf.constant([2, 2, 2, 2], dtype=tf.int32)
            ta1 = tf.TensorArray(dtype=dtype, size=4, infer_shape=infer_shape)
            ta2 = ta1.split(t, split_length)
            out0 = ta2.read(0)
            out1 = ta2.read(1)
            out2 = ta2.read(2)
            out3 = ta2.read(3)
            g = tf.get_default_graph()
            compare_tf_with_tvm([], [], ['TensorArrayReadV3:0'], mode='debug')
            compare_tf_with_tvm([], [], ['TensorArrayReadV3_1:0'], mode='debug')
            compare_tf_with_tvm([], [], ['TensorArrayReadV3_2:0'], mode='debug')
            compare_tf_with_tvm([], [], ['TensorArrayReadV3_3:0'], mode='debug')
    for dtype in ["float32", "int8"]:
        run(dtype, False)
        run(dtype, True) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_size():
    if package_version.parse(tf.VERSION) >= package_version.parse('1.15.0'):
            pytest.skip("Needs fixing for tflite >= 1.15.0")

    def run(dtype_str, infer_shape):
        with tf.Graph().as_default():
            dtype =  tf_dtypes[dtype_str]
            np_data = np.array([[1.0, 2.0], [3.0, 4.0]]).astype(dtype_str)
            in_data = [np_data, np_data]
            t1 = tf.constant(np_data, dtype=dtype)
            t2 = tf.constant(np_data, dtype=dtype)
            ta1 = tf.TensorArray(dtype=dtype, size=2, infer_shape=infer_shape)
            ta2 = ta1.write(0, t1)
            ta3 = ta2.write(1, t2)
            out = ta3.size()
            g = tf.get_default_graph()
            compare_tf_with_tvm([], [], 'TensorArraySizeV3:0', mode='debug')
    for dtype in ["float32", "int8"]:
        run(dtype, False)
        run(dtype, True) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_stack():
    def run(dtype_str, infer_shape):
        if package_version.parse(tf.VERSION) >= package_version.parse('1.15.0'):
            pytest.skip("Needs fixing for tflite >= 1.15.0")

        with tf.Graph().as_default():
            dtype =  tf_dtypes[dtype_str]
            t = tf.constant(np.array([[1.0], [2.0], [3.0]]).astype(dtype_str))
            scatter_indices = tf.constant([2, 1, 0])
            ta1 = tf.TensorArray(dtype=dtype, size=3, infer_shape=infer_shape)
            ta2 = ta1.scatter(scatter_indices, t)
            t1 = ta2.stack()
            print(t1)
            g = tf.get_default_graph()

            compare_tf_with_tvm([], [], ['TensorArrayStack/TensorArrayGatherV3:0'], mode='vm')
    for dtype in ["float32", "int8"]:
        run(dtype, True) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_unstack():
    def run(dtype_str, input_shape, infer_shape):
        if package_version.parse(tf.VERSION) >= package_version.parse('1.15.0'):
            pytest.skip("Needs fixing for tflite >= 1.15.0")

        with tf.Graph().as_default():
            dtype = tf_dtypes[dtype_str]
            t = tf.constant(np.random.choice([0, 1, 2, 3],
                                             size=input_shape).astype(dtype.name))
            ta1 = tf.TensorArray(dtype=dtype, infer_shape=infer_shape, size=input_shape[0])
            ta2 = ta1.unstack(t)
            out0 = ta2.size()
            out1 = ta2.read(0)
            compare_tf_with_tvm([], [], 'TensorArraySizeV3:0', mode='debug')
            compare_tf_with_tvm([], [], 'TensorArrayReadV3:0', mode='debug')
    for dtype in ["float32", "int8"]:
        run(dtype, (5,), False)
        run(dtype, (5, 5), True)
        run(dtype, (5, 5, 5), False)
        run(dtype, (5, 5, 5, 5), True)


#######################################################################
# ConcatV2
# -------- 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def _unstack_ta(inp):
  return tf.TensorArray(
      dtype=inp.dtype, size=tf.shape(inp)[0],
      element_shape=inp.get_shape()[1:]).unstack(inp) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def nms_tf(dets, thresh):
  """Non-maximum suppression with tf graph mode."""
  x1 = dets[:, 0]
  y1 = dets[:, 1]
  x2 = dets[:, 2]
  y2 = dets[:, 3]
  scores = dets[:, 4]

  areas = (x2 - x1 + 1) * (y2 - y1 + 1)
  order = tf.argsort(scores, direction='DESCENDING')

  keep = tf.TensorArray(tf.int32, size=0, dynamic_size=True)
  index = 0
  while tf.size(order) > 0:
    i = order[0]
    keep = keep.write(index, i)
    xx1 = tf.maximum(x1[i], tf.gather(x1, order[1:]))
    yy1 = tf.maximum(y1[i], tf.gather(y1, order[1:]))
    xx2 = tf.minimum(x2[i], tf.gather(x2, order[1:]))
    yy2 = tf.minimum(y2[i], tf.gather(y2, order[1:]))

    w = tf.maximum(0.0, xx2 - xx1 + 1)
    h = tf.maximum(0.0, yy2 - yy1 + 1)
    intersection = w * h
    overlap = intersection / (
        areas[i] + tf.gather(areas, order[1:]) - intersection)

    inds = tf.where_v2(overlap <= thresh)
    order = tf.concat(tf.gather(order, inds + 1), axis=1)
    order = tf.squeeze(order, axis=-1)
    index += 1
  return keep.stack() 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def batch_image_files_decode(image_files):
  #raw_images = tf.TensorArray(tf.uint8, size=0, dynamic_size=True)
  raw_images = tf.TensorArray(tf.float32, size=0, dynamic_size=True)
  for i in tf.range(tf.shape(image_files)[0]):
    #image = tf.io.decode_image(image_files[i])
    image = tf.io.decode_image(image_files[i], dtype=tf.float32)
    image.set_shape([None, None, None])
    raw_images = raw_images.write(i, image)
  return raw_images.stack()
############################################################################### 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_forward_squeeze():
    """ Squeeze """

    # Nothing to squeeze.
    _test_squeeze(np.arange(2).reshape((2)))
    _test_squeeze(np.arange(6).reshape((2, 3)))

    # Squeeze the middle element away.
    _test_squeeze(np.arange(4).reshape((2, 1, 2)))

    # Squeeze on both ends.
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)))

    # Positive squeeze dim index.
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [0])
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [2, 4])
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [0, 4, 2])

    # Negative squeeze dim index.
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [-1])
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [-3, -5])
    _test_squeeze(np.arange(6).reshape((1, 2, 1, 3, 1)), [-3, -5, -1])


#######################################################################
# TensorArray
# ----------- 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def test_tensor_array_gather():
    def run(dtype_str, infer_shape):
        with tf.Graph().as_default():
            dtype =  tf_dtypes[dtype_str]
            t = tf.constant(np.array([[1.0], [2.0], [3.0]]).astype(dtype_str))
            scatter_indices = tf.constant([2, 1, 0])
            gather_indices = tf.constant([1, 2])
            ta1 = tf.TensorArray(dtype=dtype, size=3, infer_shape=infer_shape)
            ta2 = ta1.scatter(scatter_indices, t)
            t1 = ta2.gather(gather_indices)
            g = tf.get_default_graph()
            compare_tf_with_tvm([], [], ['TensorArrayGatherV3:0'], mode='vm')
    for dtype in ["float32", "int8"]:
        run(dtype, True) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def _scan_initial_state(self):
    """Create TensorArrays and indices to track bin assignment.

    availability: TensorArray[queue_size, num_sequences]
      This represents the number of tokens available in the ith bin.
      See implementation note below.

    contents: TensorArray[queue_size, num_sequences * 2]
      This holds the actual contents of the packed strings as well as a bit
      mask indicating where sequences begin. It is stored in a flat vector and
      is accessed in offsets of packed_length.

    top_index: scalar [0, queue_size)
      Integer tensor indicating which index is the "top" bin. See implementation
      note below.

    IMPLEMENTATION_NOTE:
      The FFD algorithm periodically pops the topmost queue and pushes a new
      one to replace it. In order to replicate those semantics with a fixed size
      TensorArray, indexing operations are shifted by top_index. For example,
      instead of:
        `queue_available.read(i)`

      a read is instead performed as:
        `queue_available.read((i - top_index) % queue_size)`

      to account for the fact that the "ith" logical FFD queue is stored at
      position j. This means that the pop / push update can be performed by
      simply incrementing top_index. (And zeroing the old top_index position.)

    Returns:
      The state for the binning scan.
    """

    all_available = tf.ones((self._queue_size, self._num_sequences),
                            dtype=INDEX_DTYPE) * self._packed_length
    total_size = self._packed_length * self._queue_size
    total_size_range = tf.range(total_size, dtype=INDEX_DTYPE)
    empty = tf.zeros((total_size, self._num_sequences * 2),
                     dtype=self._token_dtype)

    availability = tf.TensorArray(
        dtype=INDEX_DTYPE, size=self._queue_size, dynamic_size=False,
        clear_after_read=False, element_shape=(self._num_sequences,)
        ).scatter(tf.range(self._queue_size, dtype=INDEX_DTYPE), all_available)

    contents = tf.TensorArray(
        dtype=self._token_dtype, size=total_size, dynamic_size=False,
        clear_after_read=False, element_shape=(self._num_sequences * 2,)
        ).scatter(total_size_range, empty)

    # Which index should be considered the "top" bucket for the purpose of
    # the first-fit descending algorithm.
    top_index = tf.zeros((), dtype=INDEX_DTYPE)

    return availability, contents, top_index 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def _benchmark_handwritten_dynamic_rnn(self, batch_size, max_seq_len):

    def my_dynamic_rnn(rnn_cell,
                       input_data,
                       initial_state,
                       sequence_length=None):
      """A handwritten reimplementation of dynamic_rnn."""
      input_data = tf.transpose(input_data, [1, 0, 2])
      outputs = tf.TensorArray(tf.float32, input_data.shape[0])
      if sequence_length is None:
        max_seq_len = input_data.shape[0]
      else:
        max_seq_len = tf.reduce_max(sequence_length)

      def while_body(i, state, outputs):
        new_output, new_state = rnn_cell(input_data[i], state)
        output = tf.where(i < sequence_length, new_output,
                          tf.zeros(new_output.shape))
        state = tf.where(i < sequence_length, new_state, state)
        outputs = outputs.write(i, output)
        return i + 1, state, outputs

      def while_cond(i, unused_state, unused_outputs):
        return i < max_seq_len

      _, state, outputs = tf.while_loop(
          while_cond,
          while_body,
          loop_vars=(tf.constant(0), initial_state, outputs))
      return tf.transpose(outputs.stack(), [1, 0, 2]), state

    with tf.Graph().as_default():
      input_data, sequence_lengths = self._generate_fake_rnn_inputs(
          batch_size=batch_size, max_seq_len=max_seq_len)
      rnn_cell, initial_state = self._create_rnn_cell(batch_size=batch_size)
      graph_output_t = my_dynamic_rnn(rnn_cell, input_data, initial_state,
                                      sequence_lengths)

      with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        def target():
          sess.run(graph_output_t)

        self.time_execution(
            ('Handwritten', batch_size, max_seq_len),
            target,
            iter_volume=batch_size,
            iter_unit='examples',
            extras={
                'max_seq_len': max_seq_len,
                'batch_size': batch_size,
            }) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import TensorArray [as 別名]
def _benchmark_ag_dynamic_rnn(self, batch_size, max_seq_len):

    def ag_dynamic_rnn(rnn_cell,
                       input_data,
                       initial_state,
                       sequence_length=None):
      """An autograph-able reimplementation of subset of dynamic_rnn."""
      # [batch, time, features] -> [time, batch, features]
      input_data = tf.transpose(input_data, [1, 0, 2])
      if sequence_length is None:
        max_seq_len = input_data.shape[0]
      else:
        max_seq_len = tf.reduce_max(sequence_length)

      outputs = tf.TensorArray(tf.float32, size=max_seq_len)
      state = initial_state
      for i in tf.range(max_seq_len):
        new_output, new_state = rnn_cell(input_data[i], state)
        output = tf.where(i < sequence_length, new_output,
                          tf.zeros(new_output.shape))
        state = tf.where(i < sequence_length, new_state, state)
        outputs = outputs.write(i, output)
      return tf.transpose(outputs.stack(), [1, 0, 2]), state

    ag_dynamic_rnn = tf.autograph.to_graph(ag_dynamic_rnn)

    with tf.Graph().as_default():
      input_data, sequence_lengths = self._generate_fake_rnn_inputs(
          batch_size=batch_size, max_seq_len=max_seq_len)
      rnn_cell, initial_state = self._create_rnn_cell(batch_size=batch_size)
      rnn_output = ag_dynamic_rnn(rnn_cell, input_data, initial_state,
                                  sequence_lengths)

      with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())

        def target():
          sess.run(rnn_output)

        self.time_execution(
            ('AutoGraph', batch_size, max_seq_len),
            target,
            iter_volume=batch_size,
            iter_unit='examples',
            extras={
                'max_seq_len': max_seq_len,
                'batch_size': batch_size,
            })