Python queue_runner.QueueRunner方法代碼示例

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def prefetch_op(self):
    """The op used to prefetch new data into the state saver.

    Running it once enqueues one new input example into the state saver.
    The first time this gets called, it additionally creates the prefetch_op.
    Subsequent calls simply return the previously created `prefetch_op`.

    It should be run in a separate thread via e.g. a `QueueRunner`.

    Returns:
      An `Operation` that performs prefetching.
    """
    if not self._prefetch_op:
      with ops.name_scope(None), ops.name_scope(
          self._scope, values=[self._barrier.barrier_ref]):
        self._create_prefetch_op()
    return self._prefetch_op 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _add_remote_queue_runner(self, queue, enq_ops):
    """Adds a remote queue runner to the graph.

    These queue runners differ from the standard in two ways: First,
    they never close their queue. Second, they are added to the
    `Feeder.REMOTE_QUEUE_RUNNERS` collection, rather than
    `ops.GraphKeys.QUEUE_RUNNERS`, so they can be started/stopped
    seperately.

    Args:
      queue: The queue.
      enq_ops: A list of ops which perform enqueues (each on its own thread).
    """

    runner = queue_runner.QueueRunner(
        queue,
        enq_ops,
        cancel_op=self._fake_op,
        close_op=self._fake_op)
    queue_runner.add_queue_runner(
        runner, collection=Feeder.REMOTE_QUEUE_RUNNERS) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def get_chief_queue_runner(self):
    """Returns the QueueRunner for the chief to execute.

    This includes the operations to synchronize replicas: aggregate gradients,
    apply to variables, increment global step, insert tokens to token queue.

    Note that this can only be called after calling apply_gradients() which
    actually generates this queuerunner.

    Returns:
      A `QueueRunner` for chief to execute.

    Raises:
      ValueError: If this is called before apply_gradients().
    """
    if self._gradients_applied is False:
      raise ValueError("Should be called after apply_gradients().")

    return self._chief_queue_runner 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def range_input_producer(limit, num_epochs=None, shuffle=True, seed=None,
                         capacity=32, shared_name=None, name=None):
  """Produces the integers from 0 to limit-1 in a queue.

  Note: if `num_epochs` is not `None`, this function creates local counter
  `epochs`. Use `local_variables_initializer()` to initialize local variables.

  Args:
    limit: An int32 scalar tensor.
    num_epochs: An integer (optional). If specified, `range_input_producer`
      produces each integer `num_epochs` times before generating an
      OutOfRange error. If not specified, `range_input_producer` can cycle
      through the integers an unlimited number of times.
    shuffle: Boolean. If true, the integers are randomly shuffled within each
      epoch.
    seed: An integer (optional). Seed used if shuffle == True.
    capacity: An integer. Sets the queue capacity.
    shared_name: (optional). If set, this queue will be shared under the given
      name across multiple sessions.
    name: A name for the operations (optional).

  Returns:
    A Queue with the output integers.  A `QueueRunner` for the Queue
    is added to the current `Graph`'s `QUEUE_RUNNER` collection.
  """
  with ops.name_scope(name, "input_producer", [limit]) as name:
    range_tensor = math_ops.range(limit)
    return input_producer(
        range_tensor, [], num_epochs, shuffle, seed, capacity,
        shared_name, "fraction_of_%d_full" % capacity, name) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _enqueue_join(queue, tensor_list_list, enqueue_many, keep_input):
  """Enqueue `tensor_list_list` in `queue`."""
  if enqueue_many:
    enqueue_fn = queue.enqueue_many
  else:
    enqueue_fn = queue.enqueue
  if keep_input.get_shape().ndims == 1:
    enqueue_ops = [enqueue_fn(_select_which_to_enqueue(x, keep_input))
                   for x in tensor_list_list]
  else:
    enqueue_ops = [_smart_cond(
        keep_input,
        lambda: enqueue_fn(tl),  # pylint:disable=cell-var-from-loop
        control_flow_ops.no_op) for tl in tensor_list_list]
  queue_runner.add_queue_runner(queue_runner.QueueRunner(queue, enqueue_ops)) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _enqueue(queue, tensor_list, threads, enqueue_many, keep_input):
  """Enqueue `tensor_list` in `queue`."""
  if enqueue_many:
    enqueue_fn = queue.enqueue_many
  else:
    enqueue_fn = queue.enqueue
  if keep_input.get_shape().ndims == 1:
    enqueue_ops = [
        enqueue_fn(_select_which_to_enqueue(tensor_list, keep_input))] * threads
  else:
    enqueue_ops = [_smart_cond(
        keep_input,
        lambda: enqueue_fn(tensor_list),
        control_flow_ops.no_op)] * threads
  queue_runner.add_queue_runner(queue_runner.QueueRunner(queue, enqueue_ops)) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def read(self, queue, name=None):
    """Returns the next record (key, value pair) produced by the reader.

    The multiple reader instances are all configured to `read()` from the
    filenames listed in `queue` and enqueue their output into the `common_queue`
    passed to the constructor, and this method returns the next record dequeued
    from that `common_queue`.


    Readers dequeue a work unit from `queue` if necessary (e.g. when a
    reader needs to start reading from a new file since it has finished with
    the previous file).

    A queue runner for enqueing in the `common_queue` is automatically added to
    the TF QueueRunners collection.

    Args:
      queue: A Queue or a mutable string Tensor representing a handle
        to a Queue, with string work items.
      name: A name for the operation (optional).

    Returns:
      The next record (i.e. (key, value pair)) from the common_queue.
    """

    enqueue_ops = []
    for reader in self._readers:
      enqueue_ops.append(self._common_queue.enqueue(reader.read(queue)))

    queue_runner.add_queue_runner(
        queue_runner.QueueRunner(self._common_queue, enqueue_ops))

    return self._common_queue.dequeue(name=name) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def set_fed_tensors(self, tensors):
    """Sets fed tensors."""
    enq_op = self._local_q.enqueue(tensors)
    queue_runner.add_queue_runner(queue_runner.QueueRunner(
        self._local_q, [enq_op])) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def set_many_fed_tensors(self, tensors):
    """Sets batches fed tensors."""
    enq_op = self._local_q.enqueue_many(tensors)
    queue_runner.add_queue_runner(queue_runner.QueueRunner(
        self._local_q, [enq_op])) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def next_batch(self):
    """The `NextQueuedSequenceBatch` providing access to batched output data.

    Also provides access to the `state` and `save_state` methods.
    The first time this gets called, it additionally prepares barrier reads
    and creates `NextQueuedSequenceBatch` / next_batch objects. Subsequent
    calls simply return the previously created `next_batch`.

    In order to access data in `next_batch` without blocking, the `prefetch_op`
    must have been run at least `batch_size` times (ideally in a separate
    thread, or launched via a `QueueRunner`). After processing a segment in
    `next_batch()`, `batch.save_state()` must be called which is done by the
    state_saving_rnn. Without this call, the dequeue op associated with the SQSS
    will not run.

    Returns:
      A cached `NextQueuedSequenceBatch` instance.
    """
    # This is needed to prevent errors if next_batch is called before
    # prefetch_op is created.
    if not self._prefetch_op:
      with ops.name_scope(None), ops.name_scope(
          self._scope, values=[self._barrier.barrier_ref]):
        self._create_prefetch_op()
    if not self._next_batch:
      with ops.name_scope(None), ops.name_scope(
          self._scope, values=[self._barrier.barrier_ref]):
        self._prepare_barrier_reads()
    return self._next_batch 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def range_input_producer(limit, num_epochs=None, shuffle=True, seed=None,
                         capacity=32, shared_name=None, name=None):
  """Produces the integers from 0 to limit-1 in a queue.

  Note: if `num_epochs` is not `None`, this function creates local counter
  `epochs`. Use `local_variables_initializer()` to initialize local variables.

  Args:
    limit: An int32 scalar tensor.
    num_epochs: An integer (optional). If specified, `range_input_producer`
      produces each integer `num_epochs` times before generating an
      OutOfRange error. If not specified, `range_input_producer` can cycle
      through the integers an unlimited number of times.
    shuffle: Boolean. If true, the integers are randomly shuffled within each
      epoch.
    seed: An integer (optional). Seed used if shuffle == True.
    capacity: An integer. Sets the queue capacity.
    shared_name: (optional). If set, this queue will be shared under the given
      name across multiple sessions.
    name: A name for the operations (optional).

  Returns:
    A Queue with the output integers.  A `QueueRunner` for the Queue
    is added to the current `Graph`'s `QUEUE_RUNNER` collection.
  """
  with ops.name_scope(name, "input_producer", [limit]) as name:
    range_tensor = math_ops.range(limit)
    return input_producer(
        range_tensor, [], num_epochs, shuffle, seed, capacity,
        shared_name, name, "fraction_of_%d_full" % capacity) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _enqueue_join(queue, tensor_list_list, enqueue_many, keep_input):
  """Enqueue `tensor_list_list` in `queue`."""
  if enqueue_many:
    enqueue_fn = queue.enqueue_many
  else:
    enqueue_fn = queue.enqueue
  if keep_input is None:
    enqueue_ops = [enqueue_fn(tl) for tl in tensor_list_list]
  else:
    enqueue_ops = [control_flow_ops.cond(
        keep_input,
        lambda: enqueue_fn(tl),
        control_flow_ops.no_op) for tl in tensor_list_list]
  queue_runner.add_queue_runner(queue_runner.QueueRunner(queue, enqueue_ops)) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _enqueue(queue, tensor_list, threads, enqueue_many, keep_input):
  """Enqueue `tensor_list` in `queue`."""
  if enqueue_many:
    enqueue_fn = queue.enqueue_many
  else:
    enqueue_fn = queue.enqueue
  if keep_input is None:
    enqueue_ops = [enqueue_fn(tensor_list)] * threads
  else:
    enqueue_ops = [control_flow_ops.cond(
        keep_input,
        lambda: enqueue_fn(tensor_list),
        control_flow_ops.no_op)] * threads
  queue_runner.add_queue_runner(queue_runner.QueueRunner(queue, enqueue_ops)) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _configure_readers_by(self, queue):
    enqueue_ops = []
    for reader in self._readers:
      enqueue_ops.append(self._common_queue.enqueue(reader.read(queue)))

    queue_runner.add_queue_runner(
        queue_runner.QueueRunner(self._common_queue, enqueue_ops)) 

# 需要導入模塊: from tensorflow.python.training import queue_runner [as 別名]
# 或者: from tensorflow.python.training.queue_runner import QueueRunner [as 別名]
def _enqueue_join(queue, tensor_list_list, enqueue_many, keep_input):
  """Enqueue `tensor_list_list` in `queue`."""
  if enqueue_many:
    enqueue_fn = queue.enqueue_many
  else:
    enqueue_fn = queue.enqueue
  if keep_input.shape.ndims == 1:
    enqueue_ops = [enqueue_fn(_select_which_to_enqueue(x, keep_input))
                   for x in tensor_list_list]
  else:
    enqueue_ops = [_smart_cond(
        keep_input,
        lambda: enqueue_fn(tl),  # pylint:disable=cell-var-from-loop
        control_flow_ops.no_op) for tl in tensor_list_list]
  queue_runner.add_queue_runner(queue_runner.QueueRunner(queue, enqueue_ops))