Python logging_ops.merge_all_summaries函数代码示例

  def _init_summary_op(self, summary_op=USE_DEFAULT):
    """Initilizes summary_op.

    Args:
      summary_op: An Operation that returns a Summary for the event logs.
        If set to USE_DEFAULT, create an op that merges all the summaries.
    """
    if summary_op is Supervisor.USE_DEFAULT:
      summary_op = self._get_first_op_from_collection(ops.GraphKeys.SUMMARY_OP)
      if summary_op is None:
        summary_op = logging_ops.merge_all_summaries()
        if summary_op is not None:
          ops.add_to_collection(ops.GraphKeys.SUMMARY_OP, summary_op)
    self._summary_op = summary_op

def evaluation_loop(master,
                    checkpoint_dir,
                    logdir,
                    num_evals=1,
                    initial_op=None,
                    initial_op_feed_dict=None,
                    eval_op=None,
                    eval_op_feed_dict=None,
                    final_op=None,
                    final_op_feed_dict=None,
                    summary_op=_USE_DEFAULT,
                    summary_op_feed_dict=None,
                    variables_to_restore=None,
                    eval_interval_secs=60,
                    max_number_of_evaluations=None,
                    session_config=None,
                    timeout=None):
  """Runs TF-Slim's Evaluation Loop.

  Args:
    master: The BNS address of the TensorFlow master.
    checkpoint_dir: The directory where checkpoints are stored.
    logdir: The directory where the TensorFlow summaries are written to.
    num_evals: The number of times to run `eval_op`.
    initial_op: An operation run at the beginning of evaluation.
    initial_op_feed_dict: A feed dictionary to use when executing `initial_op`.
    eval_op: A operation run `num_evals` times.
    eval_op_feed_dict: The feed dictionary to use when executing the `eval_op`.
    final_op: An operation to execute after all of the `eval_op` executions. The
      value of `final_op` is returned.
    final_op_feed_dict: A feed dictionary to use when executing `final_op`.
    summary_op: The summary_op to evaluate after running TF-Slims metric ops. By
      default the summary_op is set to tf.merge_all_summaries().
    summary_op_feed_dict: An optional feed dictionary to use when running the
      `summary_op`.
    variables_to_restore: A list of TensorFlow variables to restore during
      evaluation. If the argument is left as `None` then
      slim.variables.GetVariablesToRestore() is used.
    eval_interval_secs: The minimum number of seconds between evaluations.
    max_number_of_evaluations: the max number of iterations of the evaluation.
      If the value is left as 'None', the evaluation continues indefinitely.
    session_config: An instance of `tf.ConfigProto` that will be used to
      configure the `Session`. If left as `None`, the default will be used.
    timeout: The maximum amount of time to wait between checkpoints. If left as
      `None`, then the process will wait indefinitely.

  Returns:
    The value of `final_op` or `None` if `final_op` is `None`.
  """
  if summary_op == _USE_DEFAULT:
    summary_op = logging_ops.merge_all_summaries()

  global_step = variables.get_or_create_global_step()

  saver = tf_saver.Saver(variables_to_restore or
                         variables.get_variables_to_restore())

  summary_writer = summary_io.SummaryWriter(logdir)

  sv = supervisor.Supervisor(graph=ops.get_default_graph(),
                             logdir=logdir,
                             summary_op=None,
                             summary_writer=None,
                             global_step=None,
                             saver=saver)

  number_of_evaluations = 0
  for checkpoint_path in checkpoints_iterator(checkpoint_dir,
                                              eval_interval_secs,
                                              timeout):
    logging.info('Starting evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                           time.gmtime()))

    with sv.managed_session(
        master, start_standard_services=False, config=session_config) as sess:
      sv.saver.restore(sess, checkpoint_path)
      sv.start_queue_runners(sess)
      final_op_value = evaluation(sess,
                                  num_evals=num_evals,
                                  initial_op=initial_op,
                                  initial_op_feed_dict=initial_op_feed_dict,
                                  eval_op=eval_op,
                                  eval_op_feed_dict=eval_op_feed_dict,
                                  final_op=final_op,
                                  final_op_feed_dict=final_op_feed_dict,
                                  summary_op=summary_op,
                                  summary_op_feed_dict=summary_op_feed_dict,
                                  summary_writer=summary_writer,
                                  global_step=global_step)

    logging.info('Finished evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                           time.gmtime()))
    number_of_evaluations += 1
    if (max_number_of_evaluations and
        number_of_evaluations >= max_number_of_evaluations):
      logging.info('Reached max_number_of_evaluations=%s. Exit',
                   max_number_of_evaluations)
      return final_op_value

  logging.info(
#.........这里部分代码省略.........

def evaluate_once(master,
                  checkpoint_path,
                  logdir,
                  num_evals=1,
                  initial_op=None,
                  initial_op_feed_dict=None,
                  eval_op=None,
                  eval_op_feed_dict=None,
                  final_op=None,
                  final_op_feed_dict=None,
                  summary_op=_USE_DEFAULT,
                  summary_op_feed_dict=None,
                  variables_to_restore=None,
                  session_config=None):
  """Evaluates the model at the given checkpoint path.

  Args:
    master: The BNS address of the TensorFlow master.
    checkpoint_path: The path to a checkpoint to use for evaluation.
    logdir: The directory where the TensorFlow summaries are written to.
    num_evals: The number of times to run `eval_op`.
    initial_op: An operation run at the beginning of evaluation.
    initial_op_feed_dict: A feed dictionary to use when executing `initial_op`.
    eval_op: A operation run `num_evals` times.
    eval_op_feed_dict: The feed dictionary to use when executing the `eval_op`.
    final_op: An operation to execute after all of the `eval_op` executions. The
      value of `final_op` is returned.
    final_op_feed_dict: A feed dictionary to use when executing `final_op`.
    summary_op: The summary_op to evaluate after running TF-Slims metric ops. By
      default the summary_op is set to tf.merge_all_summaries().
    summary_op_feed_dict: An optional feed dictionary to use when running the
      `summary_op`.
    variables_to_restore: A list of TensorFlow variables to restore during
      evaluation. If the argument is left as `None` then
      slim.variables.GetVariablesToRestore() is used.
    session_config: An instance of `tf.ConfigProto` that will be used to
      configure the `Session`. If left as `None`, the default will be used.

  Returns:
    The value of `final_op` or `None` if `final_op` is `None`.
  """
  if summary_op == _USE_DEFAULT:
    summary_op = logging_ops.merge_all_summaries()

  global_step = variables.get_or_create_global_step()

  saver = tf_saver.Saver(
      variables_to_restore or variables.get_variables_to_restore(),
      write_version=saver_pb2.SaverDef.V1)

  summary_writer = summary_io.SummaryWriter(logdir)

  sv = supervisor.Supervisor(graph=ops.get_default_graph(),
                             logdir=logdir,
                             summary_op=None,
                             summary_writer=None,
                             global_step=None,
                             saver=None)

  logging.info('Starting evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                         time.gmtime()))
  with sv.managed_session(
      master, start_standard_services=False, config=session_config) as sess:
    saver.restore(sess, checkpoint_path)
    sv.start_queue_runners(sess)
    final_op_value = evaluation(sess,
                                num_evals=num_evals,
                                initial_op=initial_op,
                                initial_op_feed_dict=initial_op_feed_dict,
                                eval_op=eval_op,
                                eval_op_feed_dict=eval_op_feed_dict,
                                final_op=final_op,
                                final_op_feed_dict=final_op_feed_dict,
                                summary_op=summary_op,
                                summary_op_feed_dict=summary_op_feed_dict,
                                summary_writer=summary_writer,
                                global_step=global_step)

  logging.info('Finished evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                         time.gmtime()))

  return final_op_value

    def _setup_training(self):
        """Sets up graph, model and trainer."""
        # Create config if not given.
        if self._config is None:
            self._config = RunConfig(verbose=self.verbose)
        # Create new graph.
        self._graph = ops.Graph()
        self._graph.add_to_collection("IS_TRAINING", True)
        with self._graph.as_default():
            random_seed.set_random_seed(self._config.tf_random_seed)
            self._global_step = variables.Variable(
                0, name="global_step", trainable=False)

            # Setting up inputs and outputs.
            self._inp, self._out = self._data_feeder.input_builder()

            # If class weights are provided, add them to the graph.
            # Different loss functions can use this tensor by name.
            if self.class_weight:
                self._class_weight_node = constant_op.constant(
                    self.class_weight, name='class_weight')

            # Add histograms for X and y if they are floats.
            if self._data_feeder.input_dtype in (np.float32, np.float64):
                logging_ops.histogram_summary("X", self._inp)
            if self._data_feeder.output_dtype in (np.float32, np.float64):
                logging_ops.histogram_summary("y", self._out)

            # Create model's graph.
            self._model_predictions, self._model_loss = self.model_fn(
                self._inp, self._out)

            # Create trainer and augment graph with gradients and optimizer.
            # Additionally creates initialization ops.
            learning_rate = self.learning_rate
            optimizer = self.optimizer
            if callable(learning_rate):
                learning_rate = learning_rate(self._global_step)
            if callable(optimizer):
                optimizer = optimizer(learning_rate)
            self._train = optimizers.optimize_loss(self._model_loss, self._global_step,
                learning_rate=learning_rate,
                optimizer=optimizer, clip_gradients=self.clip_gradients)

            # Update ops during training, e.g. batch_norm_ops
            self._train = control_flow_ops.group(self._train, *ops.get_collection('update_ops'))

            # Merge all summaries into single tensor.
            self._summaries = logging_ops.merge_all_summaries()
 
            # Get all initializers for all trainable variables.
            self._initializers = variables.initialize_all_variables()

            # Create model's saver capturing all the nodes created up until now.
            self._saver = train.Saver(
                max_to_keep=self._config.keep_checkpoint_max,
                keep_checkpoint_every_n_hours=self._config.keep_checkpoint_every_n_hours)

            # Enable monitor to create validation data dict with appropriate tf placeholders
            self._monitor.create_val_feed_dict(self._inp, self._out)

            # Create session to run model with.
            self._session = session.Session(self._config.tf_master, config=self._config.tf_config)

            # Run parameter initializers.
            self._session.run(self._initializers)

def evaluation_loop(master,
                    checkpoint_dir,
                    logdir,
                    num_evals=1,
                    eval_op=None,
                    eval_op_feed_dict=None,
                    final_op=None,
                    final_op_feed_dict=None,
                    summary_op=_USE_DEFAULT,
                    summary_op_feed_dict=None,
                    variables_to_restore=None,
                    eval_interval_secs=60,
                    max_number_of_evaluations=None):
  """Runs TF-Slim's Evaluation Loop.

  Args:
    master: The BNS address of the TensorFlow master.
    checkpoint_dir: The directory where checkpoints are stored.
    logdir: The directory where the TensorFlow summaries are written to.
    num_evals: The number of times to run `eval_op`.
    eval_op: A operation run `num_evals` times.
    eval_op_feed_dict: The feed dictionary to use when executing the `eval_op`.
    final_op: An operation to execute after all of the `eval_op` executions. The
      value of `final_op` is returned.
    final_op_feed_dict: A feed dictionary to use when executing `final_op`.
    summary_op: The summary_op to evaluate after running TF-Slims metric ops. By
      default the summary_op is set to tf.merge_all_summaries().
    summary_op_feed_dict: An optional feed dictionary to use when running the
      `summary_op`.
    variables_to_restore: A list of TensorFlow variables to restore during
      evaluation. If the argument is left as `None` then
      slim.variables.GetVariablesToRestore() is used.
    eval_interval_secs: The minimum number of seconds between evaluations.
    max_number_of_evaluations: the max number of iterations of the evaluation.
      If the value is left as 'None', the evaluation continues indefinitely.
  """
  if summary_op == _USE_DEFAULT:
    summary_op = logging_ops.merge_all_summaries()

  global_step = variables.get_or_create_global_step()

  init_op = control_flow_ops.group(tf_variables.initialize_all_variables(),
                                   tf_variables.initialize_local_variables(),
                                   data_flow_ops.initialize_all_tables())

  saver = tf_saver.Saver(variables_to_restore or
                         variables.get_variables_to_restore())

  summary_writer = summary_io.SummaryWriter(logdir)

  sv = supervisor.Supervisor(graph=ops.get_default_graph(),
                             logdir=logdir,
                             init_op=init_op,
                             summary_op=None,
                             summary_writer=None,
                             global_step=None,
                             saver=saver)

  last_checkpoint = None
  number_of_evaluations = 0
  while True:
    last_checkpoint = wait_for_new_checkpoint(checkpoint_dir, last_checkpoint)
    start = time.time()
    logging.info('Starting evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                           time.gmtime()))

    with sv.managed_session(master, start_standard_services=False) as sess:
      sv.saver.restore(sess, last_checkpoint)
      sv.start_queue_runners(sess)
      evaluation(sess,
                 num_evals=num_evals,
                 eval_op=eval_op,
                 eval_op_feed_dict=eval_op_feed_dict,
                 final_op=final_op,
                 final_op_feed_dict=final_op_feed_dict,
                 summary_op=summary_op,
                 summary_op_feed_dict=summary_op_feed_dict,
                 summary_writer=summary_writer,
                 global_step=global_step)

    logging.info('Finished evaluation at ' + time.strftime('%Y-%m-%d-%H:%M:%S',
                                                           time.gmtime()))
    number_of_evaluations += 1
    if (max_number_of_evaluations and
        number_of_evaluations >= max_number_of_evaluations):
      logging.info('Reached max_number_of_evaluations=%s. Exit',
                   max_number_of_evaluations)
      break

    time_to_next_eval = start + eval_interval_secs - time.time()
    if time_to_next_eval > 0:
      time.sleep(time_to_next_eval)

def train(
    train_op,
    logdir,
    log_every_n_steps=1,
    graph=None,
    master='',
    is_chief=True,
    global_step=None,
    number_of_steps=None,
    init_op=_USE_DEFAULT,
    init_feed_dict=None,
    init_fn=None,
    summary_op=_USE_DEFAULT,
    save_summaries_secs=600,
    startup_delay_steps=0,
    saver=None,
    save_interval_secs=600,
    sync_optimizer=None):
  """Runs a training loop using a TensorFlow supervisor.

  When the sync_optimizer is supplied, gradient updates are applied
  synchronously. Otherwise, gradient updates are applied asynchronous.

  Args:
    train_op: A `Tensor` that, when executed, will apply the gradients and
      return the loss value.
    logdir: the directory where training logs are written to.
    log_every_n_steps: The frequency, in terms of global steps, that the loss
      and global step and logged.
    graph: The graph to pass to the supervisor. If no graph is supplied the
      default graph is used.
    master: The BNS name of the tensorflow master.
    is_chief: Specifies whether or not the training is being run by the primary
      replica during replica training.
    global_step: The `Tensor` representing the global step. If left as `None`,
      then slim.variables.get_or_create_global_step() is used.
    number_of_steps: The max number of gradient steps to take during training.
      If the value is left as None, training proceeds indefinitely.
    init_op: The initialization operation.
    init_feed_dict: A feed dictionary to use when executing the `init_op`.
    init_fn: An optional callable to be executed after `init_op` is called. The
      callable must accept one argument, the session being initialized.
    summary_op: The summary operation.
    save_summaries_secs: How often, in seconds, to save summaries.
    startup_delay_steps: The number of steps to wait for before beginning. Note
      that this must be 0 if a sync_optimizer is supplied.
    saver: Saver to save checkpoints. If none, a default one will be created
      and used.
    save_interval_secs: How often, in seconds, to save the model to `logdir`.
    sync_optimizer: an instance of tf.train.SyncReplicasOptimizer. If the
      argument is supplied, gradient updates will be synchronous. If left as
      `None`, gradient updates will be asynchronous.

  Returns:
    the value of the loss function after training.

  Raises:
    ValueError: if `train_op` is empty or if `startup_delay_steps` is
      non-zero when `sync_optimizer` is supplied, or if `number_of_steps` is
      negative.
  """
  if train_op is None:
    raise ValueError('train_op cannot be None.')

  if sync_optimizer and startup_delay_steps > 0:
    raise ValueError(
        'startup_delay_steps must be zero when sync_optimizer is supplied.')

  if number_of_steps is not None and number_of_steps <= 0:
    raise ValueError(
        '`number_of_steps` must be either None or a positive number.')

  graph = graph or ops.get_default_graph()
  if global_step is None:
    global_step = variables.get_or_create_global_step()
  saver = saver or tf_saver.Saver()

  if init_op is None:
    init_op = control_flow_ops.group(
        tf_variables.initialize_all_variables(),
        tf_variables.initialize_local_variables(),
        tf_variables.initialize_all_tables())

  if summary_op == _USE_DEFAULT:
    summary_op = logging_ops.merge_all_summaries()

  local_init_op = None
  cleanup_op = None

  if is_chief and sync_optimizer:
    if not isinstance(sync_optimizer,
                      sync_replicas_optimizer.SyncReplicasOptimizer):
      raise ValueError(
          '`sync_optimizer` must be a tf.train.SyncReplicasOptimizer')

    # Need to create these BEFORE the supervisor finalizes the graph:
    local_init_op = sync_optimizer.get_init_tokens_op()
    chief_queue_runner = sync_optimizer.get_chief_queue_runner()
    cleanup_op = sync_optimizer.get_clean_up_op()

#.........这里部分代码省略.........

#.........这里部分代码省略.........
      non-zero when `sync_optimizer` is supplied, or if `number_of_steps` is
      negative.
  """
  if train_op is None:
    raise ValueError('train_op cannot be None.')

  if logdir is None:
    if summary_op != _USE_DEFAULT:
      raise ValueError('Cannot provide summary_op because logdir=None')
    if saver is not None:
      raise ValueError('Cannot provide saver because logdir=None')

  if sync_optimizer and startup_delay_steps > 0:
    raise ValueError(
        'startup_delay_steps must be zero when sync_optimizer is supplied.')

  if number_of_steps is not None and number_of_steps <= 0:
    raise ValueError(
        '`number_of_steps` must be either None or a positive number.')

  graph = graph or ops.get_default_graph()
  with graph.as_default():
    if global_step is None:
      global_step = variables.get_or_create_global_step()
    saver = saver or tf_saver.Saver()

    if init_op == _USE_DEFAULT:
      init_op = tf_variables.initialize_all_variables()

    if ready_op == _USE_DEFAULT:
      ready_op = tf_variables.report_uninitialized_variables()

    if summary_op == _USE_DEFAULT:
      summary_op = logging_ops.merge_all_summaries()

    if summary_writer == _USE_DEFAULT:
      summary_writer = supervisor.Supervisor.USE_DEFAULT

    if local_init_op == _USE_DEFAULT:
      local_init_op = control_flow_ops.group(
          tf_variables.initialize_local_variables(),
          data_flow_ops.initialize_all_tables())

    cleanup_op = None

    if is_chief and sync_optimizer:
      if not isinstance(sync_optimizer,
                        sync_replicas_optimizer.SyncReplicasOptimizer):
        raise ValueError(
            '`sync_optimizer` must be a tf.train.SyncReplicasOptimizer')

      # Need to create these BEFORE the supervisor finalizes the graph:
      with ops.control_dependencies([init_op]):
        init_tokens_op = sync_optimizer.get_init_tokens_op()
      init_op = init_tokens_op
      chief_queue_runner = sync_optimizer.get_chief_queue_runner()
      cleanup_op = sync_optimizer.get_clean_up_op()

    if train_step_kwargs == _USE_DEFAULT:
      train_step_kwargs = {}

      if number_of_steps:
        should_stop_op = math_ops.greater_equal(global_step, number_of_steps)
      else:
        should_stop_op = constant_op.constant(False)
      train_step_kwargs['should_stop'] = should_stop_op