Python v1.float64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 float64 [as 別名]
def two_class_log_likelihood(predictions, labels, weights_fn=None):
  """Log-likelihood for two class classification with 0/1 labels.

  Args:
    predictions: A float valued tensor of shape [`batch_size`].  Each
      component should be between 0 and 1.
    labels: An int valued tensor of shape [`batch_size`].  Each component
      should either be 0 or 1.
    weights_fn: unused.

  Returns:
    A pair, with the average log likelihood in the first component.
  """
  del weights_fn
  float_predictions = tf.cast(tf.squeeze(predictions), dtype=tf.float64)
  batch_probs = tf.stack([1. - float_predictions, float_predictions], axis=-1)
  int_labels = tf.cast(tf.squeeze(labels), dtype=tf.int32)
  onehot_targets = tf.cast(tf.one_hot(int_labels, 2), dtype=tf.float64)
  chosen_probs = tf.einsum(
      "ij,ij->i", batch_probs, onehot_targets, name="chosen_probs")
  avg_log_likelihood = tf.reduce_mean(tf.log(chosen_probs))
  return avg_log_likelihood, tf.constant(1.0) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def testTwoClassLogLikelihoodVersusOldImplementation(self):
    def alt_two_class_log_likelihood_impl(predictions, labels):
      float_labels = tf.cast(labels, dtype=tf.float64)
      float_predictions = tf.cast(tf.squeeze(predictions), dtype=tf.float64)
      # likelihood should be just p for class 1, and 1 - p for class 0.
      # signs is 1 for class 1, and -1 for class 0
      signs = 2 * float_labels - tf.ones_like(float_labels)
      # constant_term is 1 for class 0, and 0 for class 1.
      constant_term = tf.ones_like(float_labels) - float_labels
      likelihoods = constant_term + signs * float_predictions
      log_likelihoods = tf.log(likelihoods)
      avg_log_likelihood = tf.reduce_mean(log_likelihoods)
      return avg_log_likelihood
    predictions = np.random.rand(1, 10, 1)
    targets = np.random.randint(2, size=10)
    with self.test_session() as session:
      new_log_likelihood, _ = metrics.two_class_log_likelihood(
          predictions, targets)
      alt_log_likelihood = alt_two_class_log_likelihood_impl(
          predictions, targets)
      new_impl, alt_impl = session.run([new_log_likelihood, alt_log_likelihood])
    self.assertAlmostEqual(new_impl, alt_impl) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def _verify_infiniteness_ops(tf_op, name):
    """test operator infinity ops"""

    # Only float types are allowed in Tensorflow for isfinite and isinf
    # float16 is failing on cuda
    tf_dtypes = ["float32", "float64"]
    for tf_dtype in tf_dtypes:
        shape = (8, 8)
        data = np.random.uniform(size=shape).astype(tf_dtype)
        data.ravel()[np.random.choice(data.size, int(data.size * 0.5), replace=False)] = np.infty
        data.ravel()[np.random.choice(data.size, int(data.size * 0.5), replace=False)] = np.nan

        tf.reset_default_graph()
        in_data = tf.placeholder(tf_dtype, shape, name="in_data")
        tf_op(in_data, name=name)
        compare_tf_with_tvm([data], ['in_data:0'], '{}:0'.format(name)) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def simulate(self, action):
    """Step the batch of environments.

    The results of the step can be accessed from the variables defined below.

    Args:
      action: Tensor holding the batch of actions to apply.

    Returns:
      Operation.
    """
    with tf.name_scope("environment/simulate"):
      if action.dtype in (tf.float16, tf.float32, tf.float64):
        action = tf.check_numerics(action, "action")
      def step(action):
        step_response = self._batch_env.step(action)
        # Current env doesn't return `info`, but EnvProblem does.
        # TODO(afrozm): The proper way to do this is to make T2TGymEnv return
        # an empty info return value.
        if len(step_response) == 3:
          (observ, reward, done) = step_response
        else:
          (observ, reward, done, _) = step_response
        return (observ, reward.astype(np.float32), done)
      observ, reward, done = tf.py_func(
          step, [action],
          [self.observ_dtype, tf.float32, tf.bool], name="step")
      reward = tf.check_numerics(reward, "reward")
      reward.set_shape((len(self),))
      done.set_shape((len(self),))
      with tf.control_dependencies([self._observ.assign(observ)]):
        return tf.identity(reward), tf.identity(done) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def get_sari(source_ids, prediction_ids, target_ids, max_gram_size=4):
  """Computes the SARI scores from the given source, prediction and targets.

  Args:
    source_ids: A 2D tf.Tensor of size (batch_size , sequence_length)
    prediction_ids: A 2D tf.Tensor of size (batch_size, sequence_length)
    target_ids: A 3D tf.Tensor of size (batch_size, number_of_targets,
        sequence_length)
    max_gram_size: int. largest n-gram size we care about (e.g. 3 for unigrams,
        bigrams, and trigrams)

  Returns:
    A 4-tuple of 1D float Tensors of size (batch_size) for the SARI score and
        the keep, addition and deletion scores.
  """

  def get_sari_numpy(source_ids, prediction_ids, target_ids):
    """Iterate over elements in the batch and call the SARI function."""
    sari_scores = []
    keep_scores = []
    add_scores = []
    deletion_scores = []
    # Iterate over elements in the batch.
    for source_ids_i, prediction_ids_i, target_ids_i in zip(
        source_ids, prediction_ids, target_ids):
      sari, keep, add, deletion = get_sari_score(
          source_ids_i, prediction_ids_i, target_ids_i, max_gram_size,
          BETA_FOR_SARI_DELETION_F_MEASURE)
      sari_scores.append(sari)
      keep_scores.append(keep)
      add_scores.append(add)
      deletion_scores.append(deletion)
    return (np.asarray(sari_scores), np.asarray(keep_scores),
            np.asarray(add_scores), np.asarray(deletion_scores))

  sari, keep, add, deletion = tf.py_func(
      get_sari_numpy,
      [source_ids, prediction_ids, target_ids],
      [tf.float64, tf.float64, tf.float64, tf.float64])
  return sari, keep, add, deletion 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def get_dtype(hparams):
  if hparams.activation_dtype == "float32":
    return tf.float32
  elif hparams.activation_dtype == "float64":
    return tf.float64
  elif hparams.activation_dtype == "bfloat16":
    return tf.bfloat16
  else:
    return None 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_quantile_function(self):
    # Test that quantile function inverts cumulative.
    scale = tf.placeholder(tf.float64, [None])
    layer = self.subclass(scale, [1], dtype=tf.float64)
    with self.cached_session() as sess:
      sess.run(tf.global_variables_initializer())
      quantiles = np.array([1e-5, 1e-2, .1, .5, .6, .8])
      locations = layer._standardized_quantile(quantiles)
      locations = tf.constant(locations, tf.float64)
      values, = sess.run([layer._standardized_cumulative(locations)])
    self.assertAllClose(quantiles, values, rtol=1e-12, atol=0) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def testFloat64(self):
    # Given action values that are float 64:
    action_values = tf.convert_to_tensor([1., 2., 4., 3.], dtype=tf.float64)
    epsilon = 0.1

    expected = [0.025, 0.025, 0.925, 0.025]

    result = policy_ops.epsilon_greedy(action_values, epsilon).probs
    with self.test_session() as sess:
      self.assertAllClose(sess.run(result), expected) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_forward_matmul():
    """ MatMul op test"""
    _test_matmul(1, 3, 6, 'int32')
    _test_matmul(5, 3, 1, 'float64') 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_forward_unstack():
    '''test unstack layer'''
    _test_unstack((6,), 0, 'int32')
    _test_unstack((2, 6), 1, 'float64')
    # negative axis
    _test_unstack((1, 4), -1, 'int32')
    _test_unstack((3, 6, 4), -2, 'float32')


#######################################################################
# Tile
# ---- 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_forward_tile():
    '''test Tile'''
    _test_tile((2, ), (3, ), "int32")
    _test_tile((2, 2), (2, 3), "float32")
    _test_tile((2, 4, 6), (6, 7, 8), "float64")


#######################################################################
# ClipByValue
# ----------- 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_forward_reverse_v2():
    """test ReverseV2"""
    _test_forward_reverse_v2((2, 3), 0, "int32")
    _test_forward_reverse_v2((2, 3, 5), 2, "float32")
    _test_forward_reverse_v2((2, 3, 5, 7), 1, "float32")
    _test_forward_reverse_v2((2, 3, 5), -1, "float64")
    _test_forward_reverse_v2((2, 3, 5), -3, "float64") 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def _make_example(input_ids, problem, input_feature_name="inputs"):
  """Make a tf.train.Example for the problem.

  features[input_feature_name] = input_ids

  Also fills in any other required features with dummy values.

  Args:
    input_ids: list<int>.
    problem: Problem.
    input_feature_name: name of feature for input_ids.

  Returns:
    tf.train.Example
  """
  features = {
      input_feature_name:
          tf.train.Feature(int64_list=tf.train.Int64List(value=input_ids))
  }

  # Fill in dummy values for any other required features that presumably
  # will not actually be used for prediction.
  data_fields, _ = problem.example_reading_spec()
  for fname, ftype in data_fields.items():
    if fname == input_feature_name:
      continue
    if not isinstance(ftype, tf.FixedLenFeature):
      # Only FixedLenFeatures are required
      continue
    if ftype.default_value is not None:
      # If there's a default value, no need to fill it in
      continue
    num_elements = functools.reduce(lambda acc, el: acc * el, ftype.shape, 1)
    if ftype.dtype in [tf.int32, tf.int64]:
      value = tf.train.Feature(
          int64_list=tf.train.Int64List(value=[0] * num_elements))
    if ftype.dtype in [tf.float32, tf.float64]:
      value = tf.train.Feature(
          float_list=tf.train.FloatList(value=[0.] * num_elements))
    if ftype.dtype == tf.bytes:
      value = tf.train.Feature(
          bytes_list=tf.train.BytesList(value=[""] * num_elements))
    tf.logging.info("Adding dummy value for feature %s as it is required by "
                    "the Problem.", fname)
    features[fname] = value
  return tf.train.Example(features=tf.train.Features(feature=features)) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def op(
    name,
    data,
    bucket_count=None,
    display_name=None,
    description=None,
    collections=None,
):
    """Create a legacy histogram summary op.

    Arguments:
      name: A unique name for the generated summary node.
      data: A `Tensor` of any shape. Must be castable to `float64`.
      bucket_count: Optional positive `int`. The output will have this
        many buckets, except in two edge cases. If there is no data, then
        there are no buckets. If there is data but all points have the
        same value, then there is one bucket whose left and right
        endpoints are the same.
      display_name: Optional name for this summary in TensorBoard, as a
        constant `str`. Defaults to `name`.
      description: Optional long-form description for this summary, as a
        constant `str`. Markdown is supported. Defaults to empty.
      collections: Optional list of graph collections keys. The new
        summary op is added to these collections. Defaults to
        `[Graph Keys.SUMMARIES]`.

    Returns:
      A TensorFlow summary op.
    """
    # TODO(nickfelt): remove on-demand imports once dep situation is fixed.
    import tensorflow.compat.v1 as tf

    if display_name is None:
        display_name = name
    summary_metadata = metadata.create_summary_metadata(
        display_name=display_name, description=description
    )
    with tf.name_scope(name):
        tensor = _buckets(data, bucket_count=bucket_count)
        return tf.summary.tensor_summary(
            name="histogram_summary",
            tensor=tensor,
            collections=collections,
            summary_metadata=summary_metadata,
        ) 

# 需要導入模塊: from tensorflow.compat import v1 [as 別名]
# 或者: from tensorflow.compat.v1 import float64 [as 別名]
def test_forward_space_to_batch_nd():
    # test cases: https://www.tensorflow.org/api_docs/cc/class/tensorflow/ops/space-to-batch-n-d
    _test_space_to_batch_nd(
        input_shape=[1, 2, 2, 1],
        block_shape=[2, 2],
        paddings=[[0, 0], [0, 0]]
    )

    _test_space_to_batch_nd(
        input_shape=[1, 2, 2, 3],
        block_shape=[2, 2],
        paddings=[[0, 0], [0, 0]]
    )

    _test_space_to_batch_nd(
        input_shape=[1, 4, 4, 1],
        block_shape=[2, 2],
        paddings=[[0, 0], [0, 0]]
    )

    _test_space_to_batch_nd(
        input_shape=[2, 2, 4, 1],
        block_shape=[2, 2],
        paddings=[[0, 0], [2, 0]],
        dtype='int64'
    )

    # pylint: disable=line-too-long
    # https://github.com/tensorflow/tensorflow/blob/24f578/tensorflow/python/kernel_tests/spacetobatch_op_test.py
    _test_space_to_batch_nd(
        input_shape=[2, 3],
        block_shape=[2],
        paddings=[[1, 0]],
        dtype='float32'
    )

    _test_space_to_batch_nd(
        input_shape=[2, 3, 2],
        block_shape=[2],
        paddings=[[1, 0]],
        dtype='float64'
    )

    _test_space_to_batch_nd_infer_paddings(
        input_shape=[2, 3, 2],
        block_shape=[2]
    )

#######################################################################
# BatchToSpaceND
# --------------