def _get_batches_of_transformed_samples(self, index_array):
batch_x = np.zeros((len(index_array),) + self.image_shape, dtype=K.floatx())
grayscale = self.color_mode == 'grayscale'
# build batch of image data
for i, j in enumerate(index_array):
fname = self.filenames[j]
img = load_img(os.path.join(self.directory, fname),
grayscale=grayscale,
target_size=self.target_size)
x = img_to_array(img, data_format=self.data_format)
x = self.image_data_generator.random_transform(x)
x = self.image_data_generator.standardize(x)
batch_x[i] = x
# optionally save augmented images to disk for debugging purposes
if self.save_to_dir:
for i, j in enumerate(index_array):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix, index=j, hash=np.random.randint(1e7),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
# build batch of labels
if self.class_mode == 'input':
batch_y = batch_x.copy()
elif self.class_mode == 'sparse':
batch_y = self.classes[index_array]
elif self.class_mode == 'binary':
batch_y = self.classes[index_array].astype(K.floatx())
elif self.class_mode == 'categorical':
batch_y = np.zeros((len(batch_x), self.num_classes), dtype=K.floatx())
for i, label in enumerate(self.classes[index_array]):
batch_y[i, label] = 1.
else:
return batch_x
return batch_x, batch_y
def next(self):
"""For python 2.x.
Returns:
The next batch.
"""
# Keeps under lock only the mechanism which advances
# the indexing of each batch.
with self.lock:
index_array, current_index, current_batch_size = next(
self.index_generator)
# The transformation of images is not under thread lock
# so it can be done in parallel
batch_x = np.zeros(tuple([current_batch_size] +
list(self.x.shape)[1:]),
dtype=K.floatx())
for i, j in enumerate(index_array):
x = self.x[j]
x = self.image_data_generator.random_transform(x.astype(
K.floatx()))
x = self.image_data_generator.standardize(x)
batch_x[i] = x
if self.save_to_dir:
for i in range(current_batch_size):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix,
index=current_index + i,
hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
if self.y is None:
return batch_x
batch_y = self.y[index_array]
return batch_x, batch_y
def test_on_batch(model, inputs, targets, sample_weights=None):
"""Calculates the loss for one input batch.
Arguments:
model: Model whose loss has to be calculated.
inputs: Input batch data.
targets: Target batch data.
sample_weights: Sample weight batch data.
Returns:
total loss, loss and metrics associated with each output.
"""
inputs = [
ops.convert_to_tensor(val, dtype=backend.floatx()) for val in inputs
]
targets = [
ops.convert_to_tensor(val, dtype=backend.floatx()) for val in targets
]
sample_weights = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None for val in sample_weights
]
outs, loss, loss_metrics = _process_single_batch(
model, inputs, targets, sample_weights=sample_weights, training=False)
if not isinstance(outs, list):
outs = [outs]
metrics_results = _eager_metrics_fn(model, outs, targets)
if not isinstance(loss, list):
loss = [loss]
return loss + loss_metrics + metrics_results
def test_on_batch(model, inputs, targets, sample_weights=None):
"""Calculates the loss for one input batch.
Arguments:
model: Model whose loss has to be calculated.
inputs: Input batch data.
targets: Target batch data.
sample_weights: Sample weight batch data.
Returns:
total loss, loss and metrics associated with each output.
"""
inputs = [
ops.convert_to_tensor(val, dtype=backend.floatx()) for val in inputs]
targets = [
ops.convert_to_tensor(val, dtype=backend.floatx()) for val in targets]
sample_weights = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None for val in sample_weights]
outs, loss, loss_metrics = _process_single_batch(
model, inputs, targets, sample_weights=sample_weights, training=False)
if not isinstance(outs, list):
outs = [outs]
metrics_results = _eager_metrics_fn(
model, outs, targets)
if not isinstance(loss, list):
loss = [loss]
return loss + loss_metrics + metrics_results
def weighted(y_true, y_pred, weights, mask=None):
"""Wrapper function.
Arguments:
y_true: `y_true` argument of `fn`.
y_pred: `y_pred` argument of `fn`.
weights: Weights tensor.
mask: Mask tensor.
Returns:
Scalar tensor.
"""
# score_array has ndim >= 2
score_array = fn(y_true, y_pred)
if mask is not None:
# Cast the mask to floatX to avoid float64 upcasting in theano
mask = math_ops.cast(mask, K.floatx())
# mask should have the same shape as score_array
score_array *= mask
# the loss per batch should be proportional
# to the number of unmasked samples.
score_array /= K.mean(mask)
# apply sample weighting
if weights is not None:
# reduce score_array to same ndim as weight array
ndim = K.ndim(score_array)
weight_ndim = K.ndim(weights)
score_array = K.mean(score_array,
axis=list(range(weight_ndim, ndim)))
score_array *= weights
score_array /= K.mean(
math_ops.cast(math_ops.not_equal(weights, 0), K.floatx()))
return K.mean(score_array)
def next(self):
"""For python 2.x.
Returns:
The next batch.
"""
# Keeps under lock only the mechanism which advances
# the indexing of each batch.
with self.lock:
index_array, current_index, current_batch_size = next(
self.index_generator)
# The transformation of images is not under thread lock
# so it can be done in parallel
batch_x = np.zeros(
tuple([current_batch_size] + list(self.x.shape)[1:]), dtype=K.floatx())
for i, j in enumerate(index_array):
x = self.x[j]
x = self.image_data_generator.random_transform(x.astype(K.floatx()))
x = self.image_data_generator.standardize(x)
batch_x[i] = x
if self.save_to_dir:
for i in range(current_batch_size):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix,
index=current_index + i,
hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
if self.y is None:
return batch_x
batch_y = self.y[index_array]
return batch_x, batch_y
def weighted(y_true, y_pred, weights, mask=None):
"""Wrapper function.
Arguments:
y_true: `y_true` argument of `fn`.
y_pred: `y_pred` argument of `fn`.
weights: Weights tensor.
mask: Mask tensor.
Returns:
Scalar tensor.
"""
# score_array has ndim >= 2
score_array = fn(y_true, y_pred)
if mask is not None:
# Cast the mask to floatX to avoid float64 upcasting in theano
mask = math_ops.cast(mask, K.floatx())
# mask should have the same shape as score_array
score_array *= mask
# the loss per batch should be proportional
# to the number of unmasked samples.
score_array /= K.mean(mask)
# apply sample weighting
if weights is not None:
# reduce score_array to same ndim as weight array
ndim = K.ndim(score_array)
weight_ndim = K.ndim(weights)
score_array = K.mean(score_array, axis=list(range(weight_ndim, ndim)))
score_array *= weights
score_array /= K.mean(
math_ops.cast(math_ops.not_equal(weights, 0), K.floatx()))
return K.mean(score_array)
def next(self):
"""For python 2.x.
Returns:
The next batch.
"""
# Keeps under lock only the mechanism which advances
# the indexing of each batch.
with self.lock:
index_array, current_index, current_batch_size = next(
self.index_generator)
# The transformation of images is not under thread lock
# so it can be done in parallel
batch_x = np.zeros((current_batch_size, ) + self.image_shape,
dtype=K.floatx())
if self.training:
batch_y = np.zeros((current_batch_size, ) + self.image_shape[:2] +
(self.num_classes, ),
dtype=K.floatx())
for e, i in enumerate(index_array):
# load labels of the person in focus
file_tuple = self.file_tuples[i]
image = misc.imread(file_tuple[0])
if image.shape != self.image_shape:
image = misc.imresize(image, self.image_shape)
if not self.training:
image = preprocess_input(image.astype(np.float32))
batch_x[e, :, :, :] = image
continue
else:
gt_image = misc.imread(file_tuple[1]).clip(0, 1)
if gt_image.shape[0] != self.image_shape[0]:
gt_image = misc.imresize(gt_image, self.image_shape)
if self.shift_aug:
image, gt_image = shift_and_pad_augmentation(
image, gt_image)
image = preprocess_input(image.astype(np.float32))
batch_x[e, :, :, :] = image
batch_y[e, :, :, :] = gt_image
if not self.training:
return batch_x
else:
return batch_x, batch_y
def next(self):
"""For python 2.x.
Returns:
The next batch.
"""
with self.lock:
index_array, current_index, current_batch_size = next(
self.index_generator)
# The transformation of images is not under thread lock
# so it can be done in parallel
batch_x = np.zeros((current_batch_size, ) + self.image_shape,
dtype=K.floatx())
grayscale = self.color_mode == 'grayscale'
# build batch of image data
for i, j in enumerate(index_array):
fname = self.filenames[j]
img = load_img(os.path.join(self.directory, fname),
grayscale=grayscale,
target_size=self.target_size)
x = img_to_array(img, data_format=self.data_format)
x = self.image_data_generator.random_transform(x)
x = self.image_data_generator.standardize(x)
batch_x[i] = x
# optionally save augmented images to disk for debugging purposes
if self.save_to_dir:
for i in range(current_batch_size):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix,
index=current_index + i,
hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
# build batch of labels
if self.class_mode == 'input':
batch_y = batch_x.copy()
elif self.class_mode == 'sparse':
batch_y = self.classes[index_array]
elif self.class_mode == 'binary':
batch_y = self.classes[index_array].astype(K.floatx())
elif self.class_mode == 'categorical':
batch_y = np.zeros((len(batch_x), self.num_class),
dtype=K.floatx())
for i, label in enumerate(self.classes[index_array]):
batch_y[i, label] = 1.
else:
return batch_x
return batch_x, batch_y
def next(self):
"""For python 2.x.
Returns:
The next batch.
"""
with self.lock:
index_array, current_index, current_batch_size = next(
self.index_generator)
# The transformation of images is not under thread lock
# so it can be done in parallel
batch_x = np.zeros(
(current_batch_size,) + self.image_shape, dtype=K.floatx())
grayscale = self.color_mode == 'grayscale'
# build batch of image data
for i, j in enumerate(index_array):
fname = self.filenames[j]
img = load_img(
os.path.join(self.directory, fname),
grayscale=grayscale,
target_size=self.target_size)
x = img_to_array(img, data_format=self.data_format)
x = self.image_data_generator.random_transform(x)
x = self.image_data_generator.standardize(x)
batch_x[i] = x
# optionally save augmented images to disk for debugging purposes
if self.save_to_dir:
for i in range(current_batch_size):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix,
index=current_index + i,
hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
# build batch of labels
if self.class_mode == 'input':
batch_y = batch_x.copy()
elif self.class_mode == 'sparse':
batch_y = self.classes[index_array]
elif self.class_mode == 'binary':
batch_y = self.classes[index_array].astype(K.floatx())
elif self.class_mode == 'categorical':
batch_y = np.zeros((len(batch_x), self.num_class), dtype=K.floatx())
for i, label in enumerate(self.classes[index_array]):
batch_y[i, label] = 1.
else:
return batch_x
return batch_x, batch_y
def add_weight(self,
name,
shape,
dtype=None,
initializer=None,
regularizer=None,
trainable=True,
constraint=None):
"""Adds a weight variable to the layer.
Arguments:
name: String, the name for the weight variable.
shape: The shape tuple of the weight.
dtype: The dtype of the weight.
initializer: An Initializer instance (callable).
regularizer: An optional Regularizer instance.
trainable: A boolean, whether the weight should
be trained via backprop or not (assuming
that the layer itself is also trainable).
constraint: An optional Constraint instance.
Returns:
The created weight variable.
"""
if dtype is None:
dtype = K.floatx()
weight = self.add_variable(name, shape,
dtype=dtype,
initializer=initializers.get(initializer),
regularizer=regularizers.get(regularizer),
constraint=constraints.get(constraint),
trainable=trainable)
return weight
def add_weight(self,
name,
shape,
dtype=None,
initializer=None,
regularizer=None,
trainable=True,
constraint=None):
"""Adds a weight variable to the layer.
Arguments:
name: String, the name for the weight variable.
shape: The shape tuple of the weight.
dtype: The dtype of the weight.
initializer: An Initializer instance (callable).
regularizer: An optional Regularizer instance.
trainable: A boolean, whether the weight should
be trained via backprop or not (assuming
that the layer itself is also trainable).
constraint: An optional Constraint instance.
Returns:
The created weight variable.
"""
if dtype is None:
dtype = K.floatx()
weight = self.add_variable(name,
shape,
dtype=dtype,
initializer=initializers.get(initializer),
regularizer=regularizers.get(regularizer),
constraint=constraints.get(constraint),
trainable=trainable)
return weight
def __init__(self,
input_dim,
output_dim,
embeddings_initializer='uniform',
embeddings_regularizer=None,
activity_regularizer=None,
embeddings_constraint=None,
mask_zero=False,
input_length=None,
**kwargs):
if 'input_shape' not in kwargs:
if input_length:
kwargs['input_shape'] = (input_length,)
else:
kwargs['input_shape'] = (None,)
dtype = kwargs.pop('dtype', K.floatx())
super(Embedding, self).__init__(dtype=dtype, **kwargs)
self.input_dim = input_dim
self.output_dim = output_dim
self.embeddings_initializer = initializers.get(embeddings_initializer)
self.embeddings_regularizer = regularizers.get(embeddings_regularizer)
self.activity_regularizer = regularizers.get(activity_regularizer)
self.embeddings_constraint = constraints.get(embeddings_constraint)
self.mask_zero = mask_zero
self.input_length = input_length
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [K.update_add(self.iterations, 1)]
lr = self.lr
if self.initial_decay > 0:
lr *= (1. /
(1. +
self.decay * K.cast(self.iterations, K.dtype(self.decay))))
t = K.cast(self.iterations, K.floatx()) + 1
lr_t = lr * (K.sqrt(1. - K.pow(self.beta_2, t)) /
(1. - K.pow(self.beta_1, t)))
ms = [K.zeros(K.int_shape(p), dtype=K.dtype(p)) for p in params]
vs = [K.zeros(K.int_shape(p), dtype=K.dtype(p)) for p in params]
self.weights = [self.iterations] + ms + vs
for p, g, m, v in zip(params, grads, ms, vs):
m_t = (self.beta_1 * m) + (1. - self.beta_1) * g
v_t = (self.beta_2 * v) + (1. - self.beta_2) * K.square(g)
p_t = p - lr_t * m_t / (K.sqrt(v_t) + self.epsilon)
self.updates.append(K.update(m, m_t))
self.updates.append(K.update(v, v_t))
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(K.update(p, new_p))
return self.updates
def __init__(self, **kwargs):
# These properties should be set by the user via keyword arguments.
# note that 'dtype', 'input_shape' and 'batch_input_shape'
# are only applicable to input layers: do not pass these keywords
# to non-input layers.
allowed_kwargs = {
'activity_regularizer',
'input_shape',
'batch_input_shape',
'batch_size',
'dtype',
'name',
'trainable',
'weights',
}
# Validate optional keyword arguments.
for kwarg in kwargs:
if kwarg not in allowed_kwargs:
raise TypeError('Keyword argument not understood:', kwarg)
# Get layer name.
name = kwargs.get('name')
# Get `trainable` status.
trainable = kwargs.get('trainable', True)
# Get `dtype`.
dtype = kwargs.get('dtype')
if dtype is None:
dtype = K.floatx()
# Call super, which will set all properties common to Keras layers
# and core TF layers.
super(Layer, self).__init__(
name=name, dtype=dtype, trainable=trainable,
activity_regularizer=kwargs.get('activity_regularizer'))
self._uses_inputs_arg = True
# Add properties that are Keras-only for now.
self.supports_masking = False
# Manage input shape information if passed.
if 'input_shape' in kwargs or 'batch_input_shape' in kwargs:
# In this case we will later create an input layer
# to insert before the current layer
if 'batch_input_shape' in kwargs:
batch_input_shape = tuple(kwargs['batch_input_shape'])
elif 'input_shape' in kwargs:
if 'batch_size' in kwargs:
batch_size = kwargs['batch_size']
else:
batch_size = None
batch_input_shape = (batch_size,) + tuple(kwargs['input_shape'])
self._batch_input_shape = batch_input_shape
# Manage initial weight values if passed.
if 'weights' in kwargs:
self._initial_weights = kwargs['weights']
else:
self._initial_weights = None
def img_to_array(img, data_format=None):
"""Converts a PIL Image instance to a Numpy array.
Arguments:
img: PIL Image instance.
data_format: Image data format.
Returns:
A 3D Numpy array.
Raises:
ValueError: if invalid `img` or `data_format` is passed.
"""
if data_format is None:
data_format = K.image_data_format()
if data_format not in {'channels_first', 'channels_last'}:
raise ValueError('Unknown data_format: ', data_format)
# Numpy array x has format (height, width, channel)
# or (channel, height, width)
# but original PIL image has format (width, height, channel)
x = np.asarray(img, dtype=K.floatx())
if len(x.shape) == 3:
if data_format == 'channels_first':
x = x.transpose(2, 0, 1)
elif len(x.shape) == 2:
if data_format == 'channels_first':
x = x.reshape((1, x.shape[0], x.shape[1]))
else:
x = x.reshape((x.shape[0], x.shape[1], 1))
else:
raise ValueError('Unsupported image shape: ', x.shape)
return x
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [K.update_add(self.iterations, 1)]
lr = self.lr
if self.initial_decay > 0:
lr *= (1. /
(1. + self.decay * K.cast(self.iterations, K.dtype(self.decay))))
t = K.cast(self.iterations, K.floatx()) + 1
lr_t = lr / (1. - K.pow(self.beta_1, t))
shapes = [K.int_shape(p) for p in params]
# zero init of 1st moment
ms = [K.zeros(shape) for shape in shapes]
# zero init of exponentially weighted infinity norm
us = [K.zeros(shape) for shape in shapes]
self.weights = [self.iterations] + ms + us
for p, g, m, u in zip(params, grads, ms, us):
m_t = (self.beta_1 * m) + (1. - self.beta_1) * g
u_t = K.maximum(self.beta_2 * u, K.abs(g))
p_t = p - lr_t * m_t / (u_t + self.epsilon)
self.updates.append(K.update(m, m_t))
self.updates.append(K.update(u, u_t))
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(K.update(p, new_p))
return self.updates
def __init__(self,
input_dim,
output_dim,
embeddings_initializer='uniform',
embeddings_regularizer=None,
activity_regularizer=None,
embeddings_constraint=None,
mask_zero=False,
input_length=None,
**kwargs):
if 'input_shape' not in kwargs:
if input_length:
kwargs['input_shape'] = (input_length,)
else:
kwargs['input_shape'] = (None,)
dtype = kwargs.pop('dtype', K.floatx())
super(Embedding, self).__init__(dtype=dtype, **kwargs)
self.input_dim = input_dim
self.output_dim = output_dim
self.embeddings_initializer = initializers.get(embeddings_initializer)
self.embeddings_regularizer = regularizers.get(embeddings_regularizer)
self.activity_regularizer = regularizers.get(activity_regularizer)
self.embeddings_constraint = constraints.get(embeddings_constraint)
self.mask_zero = mask_zero
self.input_length = input_length
def test_on_batch(model, ins):
"""Calculates the loss for one input batch.
Arguments:
model: Given model on which loss is calculated.
ins: Input and output batch numpy arrays.
Returns:
total loss, loss and metrics associated with each output.
"""
ins_batch_converted = []
for ib in ins:
ins_batch_converted.append(ops.convert_to_tensor(ib, dtype=K.floatx()))
eager_model_inputs = []
eager_model_outputs = []
for i in range(len(model.inputs)):
eager_model_inputs.append(ins_batch_converted[i])
for i in range(len(model.inputs), len(ins_batch_converted)):
eager_model_outputs.append(ins_batch_converted[i])
outs, loss, loss_metrics = _process_single_batch(eager_model_inputs,
eager_model_outputs,
model,
training=False)
if not isinstance(outs, list):
outs = [outs]
metric_names, metrics_results = _eager_metrics_fn(model, outs,
eager_model_outputs)
model.metrics_names.append(metric_names)
if not isinstance(loss, list):
loss = [loss]
return loss + loss_metrics + metrics_results
def test_on_batch(model, ins):
"""Calculates the loss for one input batch.
Arguments:
model: Given model on which loss is calculated.
ins: Input and output batch numpy arrays.
Returns:
total loss, loss and metrics associated with each output.
"""
ins_batch_converted = []
for ib in ins:
ins_batch_converted.append(ops.convert_to_tensor(ib, dtype=K.floatx()))
eager_model_inputs = []
eager_model_outputs = []
for i in range(len(model.inputs)):
eager_model_inputs.append(ins_batch_converted[i])
for i in range(len(model.inputs), len(ins_batch_converted)):
eager_model_outputs.append(ins_batch_converted[i])
outs, loss, loss_metrics = _process_single_batch(
eager_model_inputs, eager_model_outputs, model, training=False)
if not isinstance(outs, list):
outs = [outs]
metric_names, metrics_results = _eager_metrics_fn(
model, outs, eager_model_outputs)
model.metrics_names.append(metric_names)
if not isinstance(loss, list):
loss = [loss]
return loss + loss_metrics + metrics_results
def img_to_array(img, data_format=None):
"""Converts a PIL Image instance to a Numpy array.
Arguments:
img: PIL Image instance.
data_format: Image data format.
Returns:
A 3D Numpy array.
Raises:
ValueError: if invalid `img` or `data_format` is passed.
"""
if data_format is None:
data_format = K.image_data_format()
if data_format not in {'channels_first', 'channels_last'}:
raise ValueError('Unknown data_format: ', data_format)
# Numpy array x has format (height, width, channel)
# or (channel, height, width)
# but original PIL image has format (width, height, channel)
x = np.asarray(img, dtype=K.floatx())
if len(x.shape) == 3:
if data_format == 'channels_first':
x = x.transpose(2, 0, 1)
elif len(x.shape) == 2:
if data_format == 'channels_first':
x = x.reshape((1, x.shape[0], x.shape[1]))
else:
x = x.reshape((x.shape[0], x.shape[1], 1))
else:
raise ValueError('Unsupported image shape: ', x.shape)
return x
def _preprocess_symbolic_input(x, data_format, mode):
"""Preprocesses a tensor encoding a batch of images.
Arguments:
x: Input tensor, 3D or 4D.
data_format: Data format of the image tensor.
mode: One of "caffe", "tf" or "torch".
- caffe: will convert the images from RGB to BGR,
then will zero-center each color channel with
respect to the ImageNet dataset,
without scaling.
- tf: will scale pixels between -1 and 1,
sample-wise.
- torch: will scale pixels between 0 and 1 and then
will normalize each channel with respect to the
ImageNet dataset.
Returns:
Preprocessed tensor.
"""
global _IMAGENET_MEAN
if mode == 'tf':
x /= 127.5
x -= 1.
return x
if mode == 'torch':
x /= 255.
mean = [0.485, 0.456, 0.406]
std = [0.229, 0.224, 0.225]
else:
if data_format == 'channels_first':
# 'RGB'->'BGR'
if K.ndim(x) == 3:
x = x[::-1, ...]
else:
x = x[:, ::-1, ...]
else:
# 'RGB'->'BGR'
x = x[..., ::-1]
mean = [103.939, 116.779, 123.68]
std = None
if _IMAGENET_MEAN is None:
_IMAGENET_MEAN = constant_op.constant(-np.array(mean),
dtype=K.floatx())
# Zero-center by mean pixel
if K.dtype(x) != K.dtype(_IMAGENET_MEAN):
x = K.bias_add(x, math_ops.cast(_IMAGENET_MEAN, K.dtype(x)),
data_format)
else:
x = K.bias_add(x, _IMAGENET_MEAN, data_format)
if std is not None:
x /= std
return x
def batch_predict_loop(model, inputs, batch_size, verbose=0):
"""Predict function for eager execution when input is arrays or tensors.
Arguments:
model: Instance of `Model`.
inputs: List of input arrays.
batch_size: Integer batch size.
verbose: Verbosity mode.
Returns:
Array of predictions (if the model has a single output)
or list of arrays of predictions (if the model has multiple outputs).
"""
outs = []
num_samples = training_utils.check_num_samples(inputs, batch_size)
if verbose == 1:
progbar = generic_utils.Progbar(target=num_samples)
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch
]
if len(inputs_batch) == 1:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch[0], training=False)
else:
batch_outs = model.call(inputs_batch[0])
else:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch, training=False)
else:
batch_outs = model.call(inputs_batch)
if not isinstance(batch_outs, list):
batch_outs = [batch_outs]
if batch_index == 0:
# Pre-allocate the results arrays.
for batch_out in batch_outs:
dims = batch_out.shape[1:].dims
dims_list = [d.value for d in dims]
shape = (num_samples, ) + tuple(dims_list)
outs.append(
np.zeros(shape, dtype=batch_out.dtype.as_numpy_dtype))
for i, batch_out in enumerate(batch_outs):
outs[i][batch_start:batch_end] = batch_out
if verbose == 1:
progbar.update(batch_end)
if len(outs) == 1:
return outs[0]
return outs
def dropped_inputs(inputs=inputs, rate=self.rate, seed=self.seed):
alpha_p = -alpha * scale
kept_idx = K.greater_equal(K.random_uniform(noise_shape, seed=seed),
rate)
kept_idx = K.cast(kept_idx, K.floatx())
a = ((1 - rate) * (1 + rate * alpha_p ** 2)) ** -0.5
b = -a * alpha_p * rate
x = inputs * kept_idx + alpha_p * (1 - kept_idx)
return a * x + b
def dropped_inputs(inputs=inputs, rate=self.rate, seed=self.seed):
alpha_p = -alpha * scale
kept_idx = K.greater_equal(K.random_uniform(noise_shape, seed=seed),
rate)
kept_idx = K.cast(kept_idx, K.floatx())
a = ((1 - rate) * (1 + rate * alpha_p ** 2)) ** -0.5
b = -a * alpha_p * rate
x = inputs * kept_idx + alpha_p * (1 - kept_idx)
return a * x + b
def batch_predict_loop(model, inputs, batch_size, verbose=0):
"""Predict function for eager execution when input is arrays or tensors.
Arguments:
model: Instance of `Model`.
inputs: List of input arrays.
batch_size: Integer batch size.
verbose: Verbosity mode.
Returns:
Array of predictions (if the model has a single output)
or list of arrays of predictions (if the model has multiple outputs).
"""
outs = []
num_samples = training_utils.check_num_samples(inputs, batch_size)
if verbose == 1:
progbar = generic_utils.Progbar(target=num_samples)
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch
]
if len(inputs_batch) == 1:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch[0], training=False)
else:
batch_outs = model.call(inputs_batch[0])
else:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch, training=False)
else:
batch_outs = model.call(inputs_batch)
if not isinstance(batch_outs, list):
batch_outs = [batch_outs]
if batch_index == 0:
# Pre-allocate the results arrays.
for batch_out in batch_outs:
dims = batch_out.shape[1:].dims
dims_list = [d.value for d in dims]
shape = (num_samples,) + tuple(dims_list)
outs.append(np.zeros(shape, dtype=batch_out.dtype.as_numpy_dtype))
for i, batch_out in enumerate(batch_outs):
outs[i][batch_start:batch_end] = batch_out
if verbose == 1:
progbar.update(batch_end)
if len(outs) == 1:
return outs[0]
return outs
def _get_batches_of_transformed_samples(self, index_array):
batch_x = np.zeros(tuple([len(index_array)] + list(self.x.shape)[1:]),
dtype=K.floatx())
for i, j in enumerate(index_array):
x = self.x[j]
x = self.image_data_generator.random_transform(x.astype(K.floatx()))
x = self.image_data_generator.standardize(x)
batch_x[i] = x
if self.save_to_dir:
for i, j in enumerate(index_array):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix, index=j, hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
if self.y is None:
return batch_x
batch_y = self.y[index_array]
return batch_x, batch_y
def _get_batches_of_transformed_samples(self, index_array):
batch_x = np.zeros(tuple([len(index_array)] + list(self.x.shape)[1:]),
dtype=K.floatx())
for i, j in enumerate(index_array):
x = self.x[j]
x = self.image_data_generator.random_transform(x.astype(K.floatx()))
x = self.image_data_generator.standardize(x)
batch_x[i] = x
if self.save_to_dir:
for i, j in enumerate(index_array):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix, index=j, hash=np.random.randint(1e4),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
if self.y is None:
return batch_x
batch_y = self.y[index_array]
return batch_x, batch_y
def _preprocess_symbolic_input(x, data_format, mode):
"""Preprocesses a tensor encoding a batch of images.
Arguments:
x: Input tensor, 3D or 4D.
data_format: Data format of the image tensor.
mode: One of "caffe", "tf" or "torch".
- caffe: will convert the images from RGB to BGR,
then will zero-center each color channel with
respect to the ImageNet dataset,
without scaling.
- tf: will scale pixels between -1 and 1,
sample-wise.
- torch: will scale pixels between 0 and 1 and then
will normalize each channel with respect to the
ImageNet dataset.
Returns:
Preprocessed tensor.
"""
global _IMAGENET_MEAN
if mode == 'tf':
x /= 127.5
x -= 1.
return x
if mode == 'torch':
x /= 255.
mean = [0.485, 0.456, 0.406]
std = [0.229, 0.224, 0.225]
else:
if data_format == 'channels_first':
# 'RGB'->'BGR'
if K.ndim(x) == 3:
x = x[::-1, ...]
else:
x = x[:, ::-1, ...]
else:
# 'RGB'->'BGR'
x = x[..., ::-1]
mean = [103.939, 116.779, 123.68]
std = None
if _IMAGENET_MEAN is None:
_IMAGENET_MEAN = constant_op.constant(-np.array(mean), dtype=K.floatx())
# Zero-center by mean pixel
if K.dtype(x) != K.dtype(_IMAGENET_MEAN):
x = K.bias_add(x, math_ops.cast(_IMAGENET_MEAN, K.dtype(x)), data_format)
else:
x = K.bias_add(x, _IMAGENET_MEAN, data_format)
if std is not None:
x /= std
return x
def _get_batches_of_transformed_samples(self, index_array):
batch_x = np.zeros((len(index_array), ) + self.image_shape,
dtype=K.floatx())
grayscale = self.color_mode == 'grayscale'
# build batch of image data
for i, j in enumerate(index_array):
fname = self.filenames[j]
img = load_img(os.path.join(self.directory, fname),
grayscale=grayscale,
target_size=self.target_size)
x = img_to_array(img, data_format=self.data_format)
x = self.image_data_generator.random_transform(x)
x = self.image_data_generator.standardize(x)
batch_x[i] = x
# optionally save augmented images to disk for debugging purposes
if self.save_to_dir:
for i, j in enumerate(index_array):
img = array_to_img(batch_x[i], self.data_format, scale=True)
fname = '{prefix}_{index}_{hash}.{format}'.format(
prefix=self.save_prefix,
index=j,
hash=np.random.randint(1e7),
format=self.save_format)
img.save(os.path.join(self.save_to_dir, fname))
# build batch of labels
if self.class_mode == 'input':
batch_y = batch_x.copy()
elif self.class_mode == 'sparse':
batch_y = self.classes[index_array]
elif self.class_mode == 'binary':
batch_y = self.classes[index_array].astype(K.floatx())
elif self.class_mode == 'categorical':
batch_y = np.zeros((len(batch_x), self.num_classes),
dtype=K.floatx())
for i, label in enumerate(self.classes[index_array]):
batch_y[i, label] = 1.
else:
return batch_x
return batch_x, batch_y
def build(self, input_shape=None):
if input_shape and not self.inputs:
batch_shape = tuple(input_shape)
dtype = K.floatx()
x = Input(
batch_shape=batch_shape, dtype=dtype, name=self.name + '_input')
self.inputs = [x]
for layer in self._layers:
x = layer(x)
self.outputs = [x]
if self.inputs:
self._init_graph_network(self.inputs, self.outputs, name=self.name)
self.built = True
def build(self, input_shape=None):
if input_shape and not self.inputs:
batch_shape = tuple(input_shape)
dtype = K.floatx()
x = Input(
batch_shape=batch_shape, dtype=dtype, name=self.name + '_input')
self.inputs = [x]
for layer in self._layers:
x = layer(x)
self.outputs = [x]
if self.inputs:
self._init_graph_network(self.inputs, self.outputs, name=self.name)
self.built = True
def array_to_img(x, data_format=None, scale=True):
"""Converts a 3D Numpy array to a PIL Image instance.
Arguments:
x: Input Numpy array.
data_format: Image data format.
scale: Whether to rescale image values
to be within [0, 255].
Returns:
A PIL Image instance.
Raises:
ImportError: if PIL is not available.
ValueError: if invalid `x` or `data_format` is passed.
"""
if pil_image is None:
raise ImportError('Could not import PIL.Image. '
'The use of `array_to_img` requires PIL.')
x = np.asarray(x, dtype=K.floatx())
if x.ndim != 3:
raise ValueError(
'Expected image array to have rank 3 (single image). '
'Got array with shape:', x.shape)
if data_format is None:
data_format = K.image_data_format()
if data_format not in {'channels_first', 'channels_last'}:
raise ValueError('Invalid data_format:', data_format)
# Original Numpy array x has format (height, width, channel)
# or (channel, height, width)
# but target PIL image has format (width, height, channel)
if data_format == 'channels_first':
x = x.transpose(1, 2, 0)
if scale:
x = x + max(-np.min(x), 0) # pylint: disable=g-no-augmented-assignment
x_max = np.max(x)
if x_max != 0:
x /= x_max
x *= 255
if x.shape[2] == 3:
# RGB
return pil_image.fromarray(x.astype('uint8'), 'RGB')
elif x.shape[2] == 1:
# grayscale
return pil_image.fromarray(x[:, :, 0].astype('uint8'), 'L')
else:
raise ValueError('Unsupported channel number: ', x.shape[2])
def array_to_img(x, data_format=None, scale=True):
"""Converts a 3D Numpy array to a PIL Image instance.
Arguments:
x: Input Numpy array.
data_format: Image data format.
scale: Whether to rescale image values
to be within [0, 255].
Returns:
A PIL Image instance.
Raises:
ImportError: if PIL is not available.
ValueError: if invalid `x` or `data_format` is passed.
"""
if pil_image is None:
raise ImportError('Could not import PIL.Image. '
'The use of `array_to_img` requires PIL.')
x = np.asarray(x, dtype=K.floatx())
if x.ndim != 3:
raise ValueError('Expected image array to have rank 3 (single image). '
'Got array with shape:', x.shape)
if data_format is None:
data_format = K.image_data_format()
if data_format not in {'channels_first', 'channels_last'}:
raise ValueError('Invalid data_format:', data_format)
# Original Numpy array x has format (height, width, channel)
# or (channel, height, width)
# but target PIL image has format (width, height, channel)
if data_format == 'channels_first':
x = x.transpose(1, 2, 0)
if scale:
x = x + max(-np.min(x), 0) # pylint: disable=g-no-augmented-assignment
x_max = np.max(x)
if x_max != 0:
x /= x_max
x *= 255
if x.shape[2] == 3:
# RGB
return pil_image.fromarray(x.astype('uint8'), 'RGB')
elif x.shape[2] == 1:
# grayscale
return pil_image.fromarray(x[:, :, 0].astype('uint8'), 'L')
else:
raise ValueError('Unsupported channel number: ', x.shape[2])
def __init__(self,
x,
y,
image_data_generator,
batch_size=32,
shuffle=False,
seed=None,
data_format=None,
save_to_dir=None,
save_prefix='',
save_format='png'):
if y is not None and len(x) != len(y):
raise ValueError('X (images tensor) and y (labels) '
'should have the same length. '
'Found: X.shape = %s, y.shape = %s' %
(np.asarray(x).shape, np.asarray(y).shape))
if data_format is None:
data_format = K.image_data_format()
self.x = np.asarray(x, dtype=K.floatx())
if self.x.ndim != 4:
raise ValueError(
'Input data in `NumpyArrayIterator` '
'should have rank 4. You passed an array '
'with shape', self.x.shape)
channels_axis = 3 if data_format == 'channels_last' else 1
if self.x.shape[channels_axis] not in {1, 3, 4}:
logging.warning('NumpyArrayIterator is set to use the '
'data format convention "' + data_format + '" '
'(channels on axis ' + str(channels_axis) +
'), i.e. expected '
'either 1, 3 or 4 channels on axis ' +
str(channels_axis) + '. '
'However, it was passed an array with shape ' +
str(self.x.shape) + ' (' +
str(self.x.shape[channels_axis]) + ' channels).')
if y is not None:
self.y = np.asarray(y)
else:
self.y = None
self.image_data_generator = image_data_generator
self.data_format = data_format
self.save_to_dir = save_to_dir
self.save_prefix = save_prefix
self.save_format = save_format
super(NumpyArrayIterator, self).__init__(x.shape[0], batch_size,
shuffle, seed)
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [K.update_add(self.iterations, 1)]
t = K.cast(self.iterations, K.floatx()) + 1
# Due to the recommendations in [2], i.e. warming momentum schedule
momentum_cache_t = self.beta_1 * (
1. - 0.5 *
(K.pow(K.cast_to_floatx(0.96), t * self.schedule_decay)))
momentum_cache_t_1 = self.beta_1 * (
1. - 0.5 * (K.pow(K.cast_to_floatx(0.96),
(t + 1) * self.schedule_decay)))
m_schedule_new = self.m_schedule * momentum_cache_t
m_schedule_next = self.m_schedule * momentum_cache_t * momentum_cache_t_1
self.updates.append((self.m_schedule, m_schedule_new))
shapes = [K.int_shape(p) for p in params]
ms = [K.zeros(shape) for shape in shapes]
vs = [K.zeros(shape) for shape in shapes]
self.weights = [self.iterations] + ms + vs
for p, g, m, v in zip(params, grads, ms, vs):
# the following equations given in [1]
g_prime = g / (1. - m_schedule_new)
m_t = self.beta_1 * m + (1. - self.beta_1) * g
m_t_prime = m_t / (1. - m_schedule_next)
v_t = self.beta_2 * v + (1. - self.beta_2) * K.square(g)
v_t_prime = v_t / (1. - K.pow(self.beta_2, t))
m_t_bar = (1. - momentum_cache_t
) * g_prime + momentum_cache_t_1 * m_t_prime
self.updates.append(K.update(m, m_t))
self.updates.append(K.update(v, v_t))
p_t = p - self.lr * m_t_bar / (K.sqrt(v_t_prime) + self.epsilon)
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(K.update(p, new_p))
return self.updates
def __init__(self,
x,
y,
image_data_generator,
batch_size=32,
shuffle=False,
seed=None,
data_format=None,
save_to_dir=None,
save_prefix='',
save_format='png'):
if y is not None and len(x) != len(y):
raise ValueError('X (images tensor) and y (labels) '
'should have the same length. '
'Found: X.shape = %s, y.shape = %s' %
(np.asarray(x).shape, np.asarray(y).shape))
if data_format is None:
data_format = K.image_data_format()
self.x = np.asarray(x, dtype=K.floatx())
if self.x.ndim != 4:
raise ValueError('Input data in `NumpyArrayIterator` '
'should have rank 4. You passed an array '
'with shape', self.x.shape)
channels_axis = 3 if data_format == 'channels_last' else 1
if self.x.shape[channels_axis] not in {1, 3, 4}:
logging.warning(
'NumpyArrayIterator is set to use the '
'data format convention "' + data_format + '" '
'(channels on axis ' + str(channels_axis) + '), i.e. expected '
'either 1, 3 or 4 channels on axis ' + str(channels_axis) + '. '
'However, it was passed an array with shape ' + str(self.x.shape) +
' (' + str(self.x.shape[channels_axis]) + ' channels).')
if y is not None:
self.y = np.asarray(y)
else:
self.y = None
self.image_data_generator = image_data_generator
self.data_format = data_format
self.save_to_dir = save_to_dir
self.save_prefix = save_prefix
self.save_format = save_format
super(NumpyArrayIterator, self).__init__(x.shape[0], batch_size, shuffle,
seed)
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [state_ops.assign_add(self.iterations, 1)]
t = math_ops.cast(self.iterations, K.floatx()) + 1
# Due to the recommendations in [2], i.e. warming momentum schedule
momentum_cache_t = self.beta_1 * (
1. - 0.5 *
(math_ops.pow(K.cast_to_floatx(0.96), t * self.schedule_decay)))
momentum_cache_t_1 = self.beta_1 * (
1. - 0.5 *
(math_ops.pow(K.cast_to_floatx(0.96), (t + 1) * self.schedule_decay)))
m_schedule_new = self.m_schedule * momentum_cache_t
m_schedule_next = self.m_schedule * momentum_cache_t * momentum_cache_t_1
self.updates.append((self.m_schedule, m_schedule_new))
shapes = [K.int_shape(p) for p in params]
ms = [K.zeros(shape) for shape in shapes]
vs = [K.zeros(shape) for shape in shapes]
self.weights = [self.iterations] + ms + vs
for p, g, m, v in zip(params, grads, ms, vs):
# the following equations given in [1]
g_prime = g / (1. - m_schedule_new)
m_t = self.beta_1 * m + (1. - self.beta_1) * g
m_t_prime = m_t / (1. - m_schedule_next)
v_t = self.beta_2 * v + (1. - self.beta_2) * math_ops.square(g)
v_t_prime = v_t / (1. - math_ops.pow(self.beta_2, t))
m_t_bar = (
1. - momentum_cache_t) * g_prime + momentum_cache_t_1 * m_t_prime
self.updates.append(state_ops.assign(m, m_t))
self.updates.append(state_ops.assign(v, v_t))
p_t = p - self.lr * m_t_bar / (K.sqrt(v_t_prime) + self.epsilon)
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(state_ops.assign(p, new_p))
return self.updates
def dropped_inputs(inputs=inputs, rate=self.rate, seed=self.seed): # pylint: disable=missing-docstring
alpha = 1.6732632423543772848170429916717
scale = 1.0507009873554804934193349852946
alpha_p = -alpha * scale
kept_idx = K.greater_equal(
K.random_uniform(noise_shape, seed=seed), rate)
kept_idx = K.cast(kept_idx, K.floatx())
# Get affine transformation params
a = ((1 - rate) * (1 + rate * alpha_p**2))**-0.5
b = -a * alpha_p * rate
# Apply mask
x = inputs * kept_idx + alpha_p * (1 - kept_idx)
# Do affine transformation
return a * x + b
def dropped_inputs(inputs=inputs, rate=self.rate, seed=self.seed): # pylint: disable=missing-docstring
alpha = 1.6732632423543772848170429916717
scale = 1.0507009873554804934193349852946
alpha_p = -alpha * scale
kept_idx = K.greater_equal(
K.random_uniform(noise_shape, seed=seed), rate)
kept_idx = K.cast(kept_idx, K.floatx())
# Get affine transformation params
a = ((1 - rate) * (1 + rate * alpha_p**2))**-0.5
b = -a * alpha_p * rate
# Apply mask
x = inputs * kept_idx + alpha_p * (1 - kept_idx)
# Do affine transformation
return a * x + b
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [state_ops.assign_add(self.iterations, 1)]
lr = self.lr
if self.initial_decay > 0:
lr = lr * ( # pylint: disable=g-no-augmented-assignment
1. / (1. + self.decay * math_ops.cast(self.iterations,
K.dtype(self.decay))))
t = math_ops.cast(self.iterations, K.floatx()) + 1
lr_t = lr * (
K.sqrt(1. - math_ops.pow(self.beta_2, t)) /
(1. - math_ops.pow(self.beta_1, t)))
ms = [K.zeros(K.int_shape(p), dtype=K.dtype(p)) for p in params]
vs = [K.zeros(K.int_shape(p), dtype=K.dtype(p)) for p in params]
if self.amsgrad:
vhats = [K.zeros(K.int_shape(p), dtype=K.dtype(p)) for p in params]
else:
vhats = [K.zeros(1) for _ in params]
self.weights = [self.iterations] + ms + vs + vhats
for p, g, m, v, vhat in zip(params, grads, ms, vs, vhats):
m_t = (self.beta_1 * m) + (1. - self.beta_1) * g
v_t = (self.beta_2 * v) + (1. - self.beta_2) * math_ops.square(g)
if self.amsgrad:
vhat_t = math_ops.maximum(vhat, v_t)
p_t = p - lr_t * m_t / (K.sqrt(vhat_t) + self.epsilon)
self.updates.append(state_ops.assign(vhat, vhat_t))
else:
p_t = p - lr_t * m_t / (K.sqrt(v_t) + self.epsilon)
self.updates.append(state_ops.assign(m, m_t))
self.updates.append(state_ops.assign(v, v_t))
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(state_ops.assign(p, new_p))
return self.updates
def get_updates(self, loss, params):
grads = self.get_gradients(loss, params)
self.updates = [K.update_add(self.iterations, 1)]
lr = self.lr
if self.initial_decay > 0:
lr = lr * (
1. / # pylint: disable=g-no-augmented-assignment
(1. +
self.decay * K.cast(self.iterations, K.dtype(self.decay))))
t = K.cast(self.iterations, K.floatx()) + 1
lr_t = lr / (1. - K.pow(self.beta_1, t))
shapes = [K.int_shape(p) for p in params]
# zero init of 1st moment
ms = [K.zeros(shape) for shape in shapes]
# zero init of exponentially weighted infinity norm
us = [K.zeros(shape) for shape in shapes]
self.weights = [self.iterations] + ms + us
for p, g, m, u in zip(params, grads, ms, us):
m_t = (self.beta_1 * m) + (1. - self.beta_1) * g
u_t = K.maximum(self.beta_2 * u, K.abs(g))
p_t = p - lr_t * m_t / (u_t + self.epsilon)
self.updates.append(K.update(m, m_t))
self.updates.append(K.update(u, u_t))
new_p = p_t
# Apply constraints.
if getattr(p, 'constraint', None) is not None:
new_p = p.constraint(new_p)
self.updates.append(K.update(p, new_p))
return self.updates
def __call__(self, w): w *= K.cast(w >= 0., K.floatx()) return w
def fit_loop(model,
inputs,
targets,
sample_weights=None,
val_inputs=None,
val_targets=None,
val_sample_weights=None,
batch_size=None,
epochs=100,
verbose=1,
callbacks=None,
shuffle=True,
callback_metrics=None,
initial_epoch=0,
steps_per_epoch=None,
validation_steps=None):
"""Abstract fit function for eager execution.
Arguments:
model: Instance of the model that is being executed in Eager mode.
inputs: List of input arrays.
targets: List of target arrays.
sample_weights: Optional list of sample weight arrays.
val_inputs: Input data for validation.
val_targets: Target data for validation.
val_sample_weights: Sample weight data for validation.
batch_size: Integer batch size or None if unknown.
epochs: Number of times to iterate over the data
verbose: Verbosity mode, 0, 1 or 2
callbacks: List of callbacks to be called during training
shuffle: Whether to shuffle the data at the beginning of each epoch
callback_metrics: List of strings, the display names of the metrics
passed to the callbacks. They should be the
concatenation of list the display names of the outputs of
`f` and the list of display names of the outputs of `f_val`.
initial_epoch: Epoch at which to start training
(useful for resuming a previous training run)
steps_per_epoch: Total number of steps (batches of samples)
before declaring one epoch finished and starting the
next epoch. Ignored with the default value of `None`.
validation_steps: Number of steps to run validation for (only if doing
validation from data tensors). Ignored with default value of `None`.
Returns:
`History` object.
Raises:
ValueError: In case of invalid argument values.
"""
if not batch_size:
raise ValueError(
'With eager execution, `batch_size` should be specified.')
if steps_per_epoch or validation_steps:
raise ValueError('With eager execution, `steps_per_epoch` and '
'`validation_steps` are not valid arguments '
'(set `batch_size` instead).')
# Required for Eager mode
with backend.learning_phase_scope(1):
do_validation = False
if val_inputs:
do_validation = True
if (verbose and inputs and hasattr(inputs[0], 'shape')
and hasattr(val_inputs[0], 'shape')):
print('Train on %d samples, validate on %d samples' %
(inputs[0].shape[0], val_inputs[0].shape[0]))
if validation_steps:
if steps_per_epoch is None:
raise ValueError(
'Can only use `validation_steps` when doing step-wise '
'training, i.e. `steps_per_epoch` must be set.')
do_validation = True
out_labels = model.metrics_names
if do_validation:
callback_metrics = copy.copy(out_labels) + [
'val_' + n for n in out_labels
]
else:
callback_metrics = copy.copy(out_labels)
if sample_weights:
feed_data = inputs + targets + sample_weights
else:
feed_data = inputs + targets
num_train_samples = training_utils.check_num_samples(
feed_data,
batch_size=batch_size,
steps=steps_per_epoch,
steps_name='steps_per_epoch')
if num_train_samples is not None:
index_array = np.arange(num_train_samples)
model.history = cbks.History()
callbacks = [cbks.BaseLogger()] + (callbacks or []) + [model.history]
if verbose:
if steps_per_epoch is not None:
count_mode = 'steps'
else:
count_mode = 'samples'
callbacks += [cbks.ProgbarLogger(count_mode)]
callbacks = cbks.CallbackList(callbacks)
# it's possible to callback a different model than self
# (used by Sequential models)
if hasattr(model, 'callback_model') and model.callback_model:
callback_model = model.callback_model
else:
callback_model = model
callbacks.set_model(callback_model)
callbacks.set_params({
'batch_size': batch_size,
'epochs': epochs,
'steps': steps_per_epoch,
'samples': num_train_samples,
'verbose': verbose,
'do_validation': do_validation,
'metrics': callback_metrics or [],
})
callbacks.on_train_begin()
callback_model.stop_training = False
for cbk in callbacks:
if not val_inputs:
cbk.validation_data = []
elif val_sample_weights:
cbk.validation_data = val_inputs + val_targets + val_sample_weights
else:
cbk.validation_data = val_inputs + val_targets
for epoch in range(initial_epoch, epochs):
callbacks.on_epoch_begin(epoch)
epoch_logs = {}
if shuffle == 'batch':
index_array = model._batch_shuffle(index_array, batch_size)
elif shuffle:
np.random.shuffle(index_array)
batches = generic_utils.make_batches(num_train_samples, batch_size)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
try:
inputs_batch = slice_arrays(inputs,
batch_ids,
contiguous=not shuffle)
targets_batch = slice_arrays(targets,
batch_ids,
contiguous=not shuffle)
if sample_weights:
sample_weights_batch = slice_arrays(
sample_weights, batch_ids, contiguous=not shuffle)
else:
sample_weights_batch = None
except TypeError:
raise TypeError('TypeError while preparing batch. '
'If using HDF5 input data, '
'pass shuffle="batch".')
batch_logs = {}
batch_logs['batch'] = batch_index
batch_logs['size'] = len(batch_ids)
callbacks.on_batch_begin(batch_index, batch_logs)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch
]
targets_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in targets_batch
]
if sample_weights:
sample_weights_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None
for val in sample_weights_batch
]
outs, loss, loss_metrics = _process_single_batch(
model,
inputs_batch,
targets_batch,
sample_weights=sample_weights_batch,
training=True)
if not isinstance(outs, list):
outs = [outs]
for l, o in zip(out_labels, outs):
batch_logs[l] = o
# Required for Eager mode
metrics_results = _eager_metrics_fn(model, outs, targets_batch)
batch_logs['loss'] = tensor_util.constant_value(
backend.mean(loss))
for k, v in zip(model.metrics_names, [backend.mean(loss)] +
loss_metrics + metrics_results):
batch_logs[k] = tensor_util.constant_value(v)
callbacks.on_batch_end(batch_index, batch_logs)
if callback_model.stop_training:
break
if batch_index == len(batches) - 1: # Last batch.
if do_validation:
val_outs = test_loop(model,
val_inputs,
val_targets,
sample_weights=val_sample_weights,
batch_size=batch_size,
verbose=0)
if not isinstance(val_outs, list):
val_outs = [val_outs]
# Same labels assumed.
for l, o in zip(out_labels, val_outs):
epoch_logs['val_' + l] = o
callbacks.on_epoch_end(epoch, epoch_logs)
if callback_model.stop_training:
break
callbacks.on_train_end()
return model.history
def __call__(self, w):
w *= K.cast(K.greater_equal(w, 0.), K.floatx())
return w
def _cast_tensor_to_floatx(x):
"""Cast tensor to keras's floatx dtype if it is not already the same dtype."""
if x.dtype == K.floatx():
return x
else:
return math_ops.cast(x, K.floatx())
def test_loop(model,
inputs,
targets,
sample_weights=None,
batch_size=None,
verbose=0,
steps=None):
"""Abstract method to loop over some data in batches.
Arguments:
model: Model instance that is being evaluated in Eager mode.
inputs: List of input arrays.
targets: List of target arrays.
sample_weights: Optional list of sample weight arrays.
batch_size: integer batch size or `None`.
verbose: verbosity mode.
steps: Total number of steps (batches of samples)
before declaring predictions finished.
Ignored with the default value of `None`.
Returns:
Scalar loss (if the model has a single output and no metrics)
or list of scalars (if the model has multiple outputs
and/or metrics). The attribute `model.metrics_names` will give you
the display labels for the scalar outputs.
"""
with backend.learning_phase_scope(0):
feed_data = inputs + targets
if sample_weights:
feed_data += sample_weights
num_samples = training_utils.check_num_samples(feed_data,
batch_size=batch_size,
steps=steps,
steps_name='steps')
outs = []
if verbose == 1:
progbar = generic_utils.Progbar(target=num_samples)
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
targets_batch = slice_arrays(targets, batch_ids)
if sample_weights:
sample_weights_batch = slice_arrays(sample_weights, batch_ids)
else:
sample_weights_batch = None
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch
]
targets_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in targets_batch
]
if sample_weights:
sample_weights_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None
for val in sample_weights_batch
]
loss_outs, loss, loss_metrics = _model_loss(
model,
inputs_batch,
targets_batch,
sample_weights=sample_weights_batch,
training=False)
metrics_results = _eager_metrics_fn(model, loss_outs,
targets_batch)
batch_outs = []
for _, v in zip(model.metrics_names, [backend.mean(loss)] +
loss_metrics + metrics_results):
batch_outs.append(tensor_util.constant_value(v))
if isinstance(batch_outs, list):
if batch_index == 0:
for batch_out in enumerate(batch_outs):
outs.append(0.)
for i, batch_out in enumerate(batch_outs):
outs[i] += batch_out * len(batch_ids)
else:
if batch_index == 0:
outs.append(0.)
outs[0] += batch_outs * len(batch_ids)
if verbose == 1:
progbar.update(batch_end)
for i in range(len(outs)):
outs[i] /= num_samples
if len(outs) == 1:
return outs[0]
return outs
def predict_loop(model, inputs,
batch_size=32,
verbose=0,
steps=None):
"""Abstract method to loop over some data in batches.
Arguments:
model:
inputs: List of input arrays.
batch_size: integer batch size.
verbose: verbosity mode.
steps: Total number of steps (batches of samples)
before declaring `_predict_loop` finished.
Ignored with the default value of `None`.
Returns:
Array of predictions (if the model has a single output)
or list of arrays of predictions
(if the model has multiple outputs).
"""
with backend.learning_phase_scope(0):
num_samples = training_utils.check_num_samples(
inputs, batch_size, steps, 'steps')
if verbose == 1:
if steps is not None:
progbar = generic_utils.Progbar(target=steps)
else:
progbar = generic_utils.Progbar(target=num_samples)
outs = []
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch]
if len(inputs_batch) == 1:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch[0], training=False)
else:
batch_outs = model.call(inputs_batch[0])
else:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch, training=False)
else:
batch_outs = model.call(inputs_batch)
if not isinstance(batch_outs, list):
batch_outs = [batch_outs]
if batch_index == 0:
# Pre-allocate the results arrays.
for batch_out in batch_outs:
dims = batch_out.shape[1:].dims
dims_list = [d.value for d in dims]
shape = (num_samples,) + tuple(dims_list)
outs.append(np.zeros(shape, dtype=batch_out.dtype.as_numpy_dtype))
for i, batch_out in enumerate(batch_outs):
outs[i][batch_start:batch_end] = batch_out
if verbose == 1:
progbar.update(batch_end)
if len(outs) == 1:
return outs[0]
return outs
def fit(self, x, augment=False, rounds=1, seed=None):
"""Fits internal statistics to some sample data.
Required for featurewise_center, featurewise_std_normalization
and zca_whitening.
Arguments:
x: Numpy array, the data to fit on. Should have rank 4.
In case of grayscale data,
the channels axis should have value 1, and in case
of RGB data, it should have value 3.
augment: Whether to fit on randomly augmented samples
rounds: If `augment`,
how many augmentation passes to do over the data
seed: random seed.
Raises:
ValueError: in case of invalid input `x`.
ImportError: if Scipy is not available.
"""
x = np.asarray(x, dtype=K.floatx())
if x.ndim != 4:
raise ValueError('Input to `.fit()` should have rank 4. '
'Got array with shape: ' + str(x.shape))
if x.shape[self.channel_axis] not in {3, 4}:
logging.warning(
'Expected input to be images (as Numpy array) '
'following the data format convention "' + self.data_format + '" '
'(channels on axis ' + str(self.channel_axis) + '), i.e. expected '
'either 1, 3 or 4 channels on axis ' + str(self.channel_axis) + '. '
'However, it was passed an array with shape ' + str(x.shape) + ' (' +
str(x.shape[self.channel_axis]) + ' channels).')
if seed is not None:
np.random.seed(seed)
x = np.copy(x)
if augment:
ax = np.zeros(
tuple([rounds * x.shape[0]] + list(x.shape)[1:]), dtype=K.floatx())
for r in range(rounds):
for i in range(x.shape[0]):
ax[i + r * x.shape[0]] = self.random_transform(x[i])
x = ax
if self.featurewise_center:
self.mean = np.mean(x, axis=(0, self.row_axis, self.col_axis))
broadcast_shape = [1, 1, 1]
broadcast_shape[self.channel_axis - 1] = x.shape[self.channel_axis]
self.mean = np.reshape(self.mean, broadcast_shape)
x -= self.mean
if self.featurewise_std_normalization:
self.std = np.std(x, axis=(0, self.row_axis, self.col_axis))
broadcast_shape = [1, 1, 1]
broadcast_shape[self.channel_axis - 1] = x.shape[self.channel_axis]
self.std = np.reshape(self.std, broadcast_shape)
x /= (self.std + K.epsilon())
if self.zca_whitening:
if linalg is None:
raise ImportError('Scipy is required for zca_whitening.')
flat_x = np.reshape(x, (x.shape[0], x.shape[1] * x.shape[2] * x.shape[3]))
sigma = np.dot(flat_x.T, flat_x) / flat_x.shape[0]
u, s, _ = linalg.svd(sigma)
self.principal_components = np.dot(
np.dot(u, np.diag(1. / np.sqrt(s + self.zca_epsilon))), u.T)
def fit_loop(
model,
inputs,
targets,
sample_weights=None,
val_inputs=None,
val_targets=None,
val_sample_weights=None,
batch_size=None,
epochs=100,
verbose=1,
callbacks=None,
shuffle=True,
callback_metrics=None,
initial_epoch=0,
steps_per_epoch=None,
validation_steps=None):
"""Abstract fit function for eager execution.
Arguments:
model: Instance of the model that is being executed in Eager mode.
inputs: List of input arrays.
targets: List of target arrays.
sample_weights: Optional list of sample weight arrays.
val_inputs: Input data for validation.
val_targets: Target data for validation.
val_sample_weights: Sample weight data for validation.
batch_size: Integer batch size or None if unknown.
epochs: Number of times to iterate over the data
verbose: Verbosity mode, 0, 1 or 2
callbacks: List of callbacks to be called during training
shuffle: Whether to shuffle the data at the beginning of each epoch
callback_metrics: List of strings, the display names of the metrics
passed to the callbacks. They should be the
concatenation of list the display names of the outputs of
`f` and the list of display names of the outputs of `f_val`.
initial_epoch: Epoch at which to start training
(useful for resuming a previous training run)
steps_per_epoch: Total number of steps (batches of samples)
before declaring one epoch finished and starting the
next epoch. Ignored with the default value of `None`.
validation_steps: Number of steps to run validation for (only if doing
validation from data tensors). Ignored with default value of `None`.
Returns:
`History` object.
Raises:
ValueError: In case of invalid argument values.
"""
if not batch_size:
raise ValueError('With eager execution, `batch_size` should be specified.')
if steps_per_epoch or validation_steps:
raise ValueError('With eager execution, `steps_per_epoch` and '
'`validation_steps` are not valid arguments '
'(set `batch_size` instead).')
# Required for Eager mode
with backend.learning_phase_scope(1):
do_validation = False
if val_inputs:
do_validation = True
if (verbose and inputs and hasattr(inputs[0], 'shape') and
hasattr(val_inputs[0], 'shape')):
print('Train on %d samples, validate on %d samples' %
(inputs[0].shape[0], val_inputs[0].shape[0]))
if validation_steps:
if steps_per_epoch is None:
raise ValueError('Can only use `validation_steps` when doing step-wise '
'training, i.e. `steps_per_epoch` must be set.')
do_validation = True
out_labels = model.metrics_names
if do_validation:
callback_metrics = copy.copy(out_labels) + [
'val_' + n for n in out_labels
]
else:
callback_metrics = copy.copy(out_labels)
if sample_weights:
feed_data = inputs + targets + sample_weights
else:
feed_data = inputs + targets
num_train_samples = training_utils.check_num_samples(
feed_data,
batch_size=batch_size,
steps=steps_per_epoch,
steps_name='steps_per_epoch')
if num_train_samples is not None:
index_array = np.arange(num_train_samples)
model.history = cbks.History()
callbacks = [cbks.BaseLogger()] + (callbacks or []) + [model.history]
if verbose:
if steps_per_epoch is not None:
count_mode = 'steps'
else:
count_mode = 'samples'
callbacks += [cbks.ProgbarLogger(count_mode)]
callbacks = cbks.CallbackList(callbacks)
# it's possible to callback a different model than self
# (used by Sequential models)
if hasattr(model, 'callback_model') and model.callback_model:
callback_model = model.callback_model
else:
callback_model = model
callbacks.set_model(callback_model)
callbacks.set_params({
'batch_size': batch_size,
'epochs': epochs,
'steps': steps_per_epoch,
'samples': num_train_samples,
'verbose': verbose,
'do_validation': do_validation,
'metrics': callback_metrics or [],
})
callbacks.on_train_begin()
callback_model.stop_training = False
for cbk in callbacks:
if not val_inputs:
cbk.validation_data = []
elif val_sample_weights:
cbk.validation_data = val_inputs + val_targets + val_sample_weights
else:
cbk.validation_data = val_inputs + val_targets
for epoch in range(initial_epoch, epochs):
callbacks.on_epoch_begin(epoch)
epoch_logs = {}
if shuffle == 'batch':
index_array = model._batch_shuffle(index_array, batch_size)
elif shuffle:
np.random.shuffle(index_array)
batches = generic_utils.make_batches(num_train_samples, batch_size)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
try:
inputs_batch = slice_arrays(inputs, batch_ids,
contiguous=not shuffle)
targets_batch = slice_arrays(targets, batch_ids,
contiguous=not shuffle)
if sample_weights:
sample_weights_batch = slice_arrays(sample_weights, batch_ids,
contiguous=not shuffle)
else:
sample_weights_batch = None
except TypeError:
raise TypeError('TypeError while preparing batch. '
'If using HDF5 input data, '
'pass shuffle="batch".')
batch_logs = {}
batch_logs['batch'] = batch_index
batch_logs['size'] = len(batch_ids)
callbacks.on_batch_begin(batch_index, batch_logs)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch]
targets_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in targets_batch]
if sample_weights:
sample_weights_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None
for val in sample_weights_batch]
outs, loss, loss_metrics = _process_single_batch(
model,
inputs_batch,
targets_batch,
sample_weights=sample_weights_batch,
training=True)
if not isinstance(outs, list):
outs = [outs]
for l, o in zip(out_labels, outs):
batch_logs[l] = o
# Required for Eager mode
metrics_results = _eager_metrics_fn(model, outs, targets_batch)
batch_logs['loss'] = tensor_util.constant_value(backend.mean(loss))
for k, v in zip(model.metrics_names,
[backend.mean(loss)] + loss_metrics + metrics_results):
batch_logs[k] = tensor_util.constant_value(v)
callbacks.on_batch_end(batch_index, batch_logs)
if callback_model.stop_training:
break
if batch_index == len(batches) - 1: # Last batch.
if do_validation:
val_outs = test_loop(
model, val_inputs, val_targets,
sample_weights=val_sample_weights,
batch_size=batch_size,
verbose=0)
if not isinstance(val_outs, list):
val_outs = [val_outs]
# Same labels assumed.
for l, o in zip(out_labels, val_outs):
epoch_logs['val_' + l] = o
callbacks.on_epoch_end(epoch, epoch_logs)
if callback_model.stop_training:
break
callbacks.on_train_end()
return model.history
def predict_loop(model, inputs, batch_size=32, verbose=0, steps=None):
"""Abstract method to loop over some data in batches.
Arguments:
model:
inputs: List of input arrays.
batch_size: integer batch size.
verbose: verbosity mode.
steps: Total number of steps (batches of samples)
before declaring `_predict_loop` finished.
Ignored with the default value of `None`.
Returns:
Array of predictions (if the model has a single output)
or list of arrays of predictions
(if the model has multiple outputs).
"""
with backend.learning_phase_scope(0):
num_samples = training_utils.check_num_samples(inputs, batch_size,
steps, 'steps')
if verbose == 1:
if steps is not None:
progbar = generic_utils.Progbar(target=steps)
else:
progbar = generic_utils.Progbar(target=num_samples)
outs = []
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch
]
if len(inputs_batch) == 1:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch[0], training=False)
else:
batch_outs = model.call(inputs_batch[0])
else:
if model._expects_training_arg:
batch_outs = model.call(inputs_batch, training=False)
else:
batch_outs = model.call(inputs_batch)
if not isinstance(batch_outs, list):
batch_outs = [batch_outs]
if batch_index == 0:
# Pre-allocate the results arrays.
for batch_out in batch_outs:
dims = batch_out.shape[1:].dims
dims_list = [d.value for d in dims]
shape = (num_samples, ) + tuple(dims_list)
outs.append(
np.zeros(shape, dtype=batch_out.dtype.as_numpy_dtype))
for i, batch_out in enumerate(batch_outs):
outs[i][batch_start:batch_end] = batch_out
if verbose == 1:
progbar.update(batch_end)
if len(outs) == 1:
return outs[0]
return outs
def __call__(self, w): return w * K.cast(K.greater_equal(w, 0.), K.floatx())
def fit(self, x, augment=False, rounds=1, seed=None):
"""Fits internal statistics to some sample data.
Required for featurewise_center, featurewise_std_normalization
and zca_whitening.
Arguments:
x: Numpy array, the data to fit on. Should have rank 4.
In case of grayscale data,
the channels axis should have value 1, and in case
of RGB data, it should have value 3.
augment: Whether to fit on randomly augmented samples
rounds: If `augment`,
how many augmentation passes to do over the data
seed: random seed.
Raises:
ValueError: in case of invalid input `x`.
ImportError: if Scipy is not available.
"""
x = np.asarray(x, dtype=K.floatx())
if x.ndim != 4:
raise ValueError('Input to `.fit()` should have rank 4. '
'Got array with shape: ' + str(x.shape))
if x.shape[self.channel_axis] not in {1, 3, 4}:
logging.warning(
'Expected input to be images (as Numpy array) '
'following the data format convention "' + self.data_format + '" '
'(channels on axis ' + str(self.channel_axis) + '), i.e. expected '
'either 1, 3 or 4 channels on axis ' + str(self.channel_axis) + '. '
'However, it was passed an array with shape ' + str(x.shape) + ' (' +
str(x.shape[self.channel_axis]) + ' channels).')
if seed is not None:
np.random.seed(seed)
x = np.copy(x)
if augment:
ax = np.zeros(
tuple([rounds * x.shape[0]] + list(x.shape)[1:]), dtype=K.floatx())
for r in range(rounds):
for i in range(x.shape[0]):
ax[i + r * x.shape[0]] = self.random_transform(x[i])
x = ax
if self.featurewise_center:
self.mean = np.mean(x, axis=(0, self.row_axis, self.col_axis))
broadcast_shape = [1, 1, 1]
broadcast_shape[self.channel_axis - 1] = x.shape[self.channel_axis]
self.mean = np.reshape(self.mean, broadcast_shape)
x -= self.mean
if self.featurewise_std_normalization:
self.std = np.std(x, axis=(0, self.row_axis, self.col_axis))
broadcast_shape = [1, 1, 1]
broadcast_shape[self.channel_axis - 1] = x.shape[self.channel_axis]
self.std = np.reshape(self.std, broadcast_shape)
x /= (self.std + K.epsilon())
if self.zca_whitening:
if linalg is None:
raise ImportError('Scipy is required for zca_whitening.')
flat_x = np.reshape(x, (x.shape[0], x.shape[1] * x.shape[2] * x.shape[3]))
num_examples = flat_x.shape[0]
_, s, vt = linalg.svd(flat_x / np.sqrt(num_examples))
s_expand = np.hstack(
(s, np.zeros(vt.shape[0] - num_examples, dtype=flat_x.dtype)))
self.principal_components = (
vt.T / np.sqrt(s_expand**2 + self.zca_epsilon)).dot(vt)
def call(self, inputs, mask=None): return inputs * K.cast(inputs > self.theta, K.floatx())
def predict_loop(model, ins, batch_size=32, verbose=0, steps=None):
"""Abstract method to loop over some data in batches.
Arguments:
model:
ins: list of tensors to be fed to `f`.
batch_size: integer batch size.
verbose: verbosity mode.
steps: Total number of steps (batches of samples)
before declaring `_predict_loop` finished.
Ignored with the default value of `None`.
Returns:
Array of predictions (if the model has a single output)
or list of arrays of predictions
(if the model has multiple outputs).
"""
K.set_learning_phase(False)
num_samples = model._check_num_samples(ins, batch_size, steps, 'steps')
if verbose == 1:
if steps is not None:
progbar = Progbar(target=steps)
else:
progbar = Progbar(target=num_samples)
outs = []
batches = _make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
if ins and isinstance(ins[-1], float):
# Do not slice the training phase flag.
ins_batch = _slice_arrays(ins[:-1], batch_ids) + [ins[-1]]
else:
ins_batch = _slice_arrays(ins, batch_ids)
ins_batch_converted = []
for ib in ins_batch:
ins_batch_converted.append(
ops.convert_to_tensor(ib, dtype=K.floatx()))
eager_model_inputs = []
for i in range(len(model.inputs)):
eager_model_inputs.append(ins_batch_converted[i])
batch_outs = model(eager_model_inputs)
if not isinstance(batch_outs, list):
batch_outs = [batch_outs]
if batch_index == 0:
# Pre-allocate the results arrays.
for batch_out in batch_outs:
dims = batch_out.shape[1:].dims
dims_list = [d.value for d in dims]
shape = (num_samples, ) + tuple(dims_list)
outs.append(
np.zeros(shape, dtype=batch_out.dtype.as_numpy_dtype))
for i, batch_out in enumerate(batch_outs):
outs[i][batch_start:batch_end] = batch_out
if verbose == 1:
progbar.update(batch_end)
if len(outs) == 1:
return outs[0]
return outs
def test_loop(model, inputs, targets,
sample_weights=None,
batch_size=None,
verbose=0,
steps=None):
"""Abstract method to loop over some data in batches.
Arguments:
model: Model instance that is being evaluated in Eager mode.
inputs: List of input arrays.
targets: List of target arrays.
sample_weights: Optional list of sample weight arrays.
batch_size: integer batch size or `None`.
verbose: verbosity mode.
steps: Total number of steps (batches of samples)
before declaring predictions finished.
Ignored with the default value of `None`.
Returns:
Scalar loss (if the model has a single output and no metrics)
or list of scalars (if the model has multiple outputs
and/or metrics). The attribute `model.metrics_names` will give you
the display labels for the scalar outputs.
"""
with backend.learning_phase_scope(0):
feed_data = inputs + targets
if sample_weights:
feed_data += sample_weights
num_samples = training_utils.check_num_samples(
feed_data, batch_size=batch_size, steps=steps, steps_name='steps')
outs = []
if verbose == 1:
progbar = generic_utils.Progbar(target=num_samples)
batches = generic_utils.make_batches(num_samples, batch_size)
index_array = np.arange(num_samples)
for batch_index, (batch_start, batch_end) in enumerate(batches):
batch_ids = index_array[batch_start:batch_end]
inputs_batch = slice_arrays(inputs, batch_ids)
targets_batch = slice_arrays(targets, batch_ids)
if sample_weights:
sample_weights_batch = slice_arrays(sample_weights, batch_ids)
else:
sample_weights_batch = None
inputs_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in inputs_batch]
targets_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
for val in targets_batch]
if sample_weights:
sample_weights_batch = [
ops.convert_to_tensor(val, dtype=backend.floatx())
if val is not None else None
for val in sample_weights_batch]
loss_outs, loss, loss_metrics = _model_loss(
model,
inputs_batch,
targets_batch,
sample_weights=sample_weights_batch,
training=False)
metrics_results = _eager_metrics_fn(model, loss_outs, targets_batch)
batch_outs = []
for _, v in zip(model.metrics_names,
[backend.mean(loss)] + loss_metrics + metrics_results):
batch_outs.append(tensor_util.constant_value(v))
if isinstance(batch_outs, list):
if batch_index == 0:
for batch_out in enumerate(batch_outs):
outs.append(0.)
for i, batch_out in enumerate(batch_outs):
outs[i] += batch_out * len(batch_ids)
else:
if batch_index == 0:
outs.append(0.)
outs[0] += batch_outs * len(batch_ids)
if verbose == 1:
progbar.update(batch_end)
for i in range(len(outs)):
outs[i] /= num_samples
if len(outs) == 1:
return outs[0]
return outs
def sparse_categorical_accuracy(y_true, y_pred):
return K.cast(
K.equal(
K.max(y_true, axis=-1), K.cast(K.argmax(y_pred, axis=-1),
K.floatx())), K.floatx())
def _cast_tensor_to_floatx(x):
"""Cast tensor to keras's floatx dtype if it is not already the same dtype."""
if x.dtype == K.floatx():
return x
else:
return math_ops.cast(x, K.floatx())