def get_model_from_layers(model_layers,
input_shape=None,
input_dtype=None,
name=None,
input_ragged=None,
input_sparse=None):
"""Builds a model from a sequence of layers.
Args:
model_layers: The layers used to build the network.
input_shape: Shape tuple of the input or 'TensorShape' instance.
input_dtype: Datatype of the input.
name: Name for the model.
input_ragged: Boolean, whether the input data is a ragged tensor.
input_sparse: Boolean, whether the input data is a sparse tensor.
Returns:
A Keras model.
"""
model_type = get_model_type()
if model_type == 'subclass':
inputs = None
if input_ragged or input_sparse:
inputs = layers.Input(shape=input_shape,
dtype=input_dtype,
ragged=input_ragged,
sparse=input_sparse)
return _SubclassModel(model_layers, name=name, input_tensor=inputs)
if model_type == 'subclass_custom_build':
layer_generating_func = lambda: model_layers
return _SubclassModelCustomBuild(layer_generating_func, name=name)
if model_type == 'sequential':
model = models.Sequential(name=name)
if input_shape:
model.add(
layers.InputLayer(input_shape=input_shape,
dtype=input_dtype,
ragged=input_ragged,
sparse=input_sparse))
for layer in model_layers:
model.add(layer)
return model
if model_type == 'functional':
if not input_shape:
raise ValueError(
'Cannot create a functional model from layers with no '
'input shape.')
inputs = layers.Input(shape=input_shape,
dtype=input_dtype,
ragged=input_ragged,
sparse=input_sparse)
outputs = inputs
for layer in model_layers:
outputs = layer(outputs)
return models.Model(inputs, outputs, name=name)
raise ValueError('Unknown model type {}'.format(model_type))
def test_dynamic_loss_scaling(self,
strategy_fn,
experimental_run_tf_function=True):
if not self._is_strategy_supported(strategy_fn):
return
strategy = strategy_fn()
initial_loss_scale = 2.
batch_size = 4
expected_gradient = backend.variable([initial_loss_scale / batch_size],
dtype=dtypes.float16)
# If this variable is set to True, the model below will have NaN gradients
have_nan_gradients = backend.variable(False, dtype=dtypes.bool)
with strategy.scope():
with policy.policy_scope(policy.Policy('infer_float32_vars')):
x = layers.Input(
shape=(1,), batch_size=batch_size, dtype=dtypes.float16)
layer = AddLayer(assert_type=dtypes.float16)
y = layer(x)
identity_with_nan_grads = (
mp_test_util.create_identity_with_nan_gradients_fn(
have_nan_gradients))
y = core.Lambda(identity_with_nan_grads)(y)
identity_with_grad_check_fn = (
mp_test_util.create_identity_with_grad_check_fn(
expected_dtype=dtypes.float16,
expected_gradient=expected_gradient))
y = core.Lambda(identity_with_grad_check_fn)(y)
y = math_ops.cast(y, dtypes.float32)
model = models.Model(inputs=x, outputs=y)
def loss_fn(y_true, y_pred):
del y_true
return math_ops.reduce_mean(y_pred)
opt = gradient_descent.SGD(1.)
loss_scale = loss_scale_module.DynamicLossScale(
initial_loss_scale=initial_loss_scale, increment_period=2)
opt = loss_scale_optimizer.LossScaleOptimizer(opt, loss_scale)
model.compile(
opt,
loss=loss_fn,
run_eagerly=testing_utils.should_run_eagerly(),
experimental_run_tf_function=testing_utils.should_run_tf_function())
self.assertEqual(backend.eval(layer.v), 1)
x = np.ones((batch_size, 1))
y = np.ones((batch_size, 1))
dataset = dataset_ops.Dataset.from_tensor_slices((x, y)).batch(batch_size)
model.fit(dataset)
# The variables starts with 1 and has a gradient of 1, so will go down by 1
# each step.
self.assertEqual(backend.eval(layer.v), 0)
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -1)
# There have been two steps without NaNs, so the loss scale will double
backend.set_value(expected_gradient,
backend.get_value(expected_gradient * 2))
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -2)
# Next test with NaN gradients.
backend.set_value(have_nan_gradients, True)
model.fit(dataset)
# Variable should not be updated
self.assertEqual(backend.eval(layer.v), -2)
# Test with finite gradients again
backend.set_value(have_nan_gradients, False)
# The loss scale will be halved due to the NaNs, so the gradient will also
# be halved
backend.set_value(expected_gradient,
backend.get_value(expected_gradient / 2))
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -3)
def resnet50(num_classes, batch_size=None, use_l2_regularizer=True):
"""Instantiates the ResNet50 architecture.
Args:
num_classes: `int` number of classes for image classification.
batch_size: Size of the batches for each step.
use_l2_regularizer: whether to use L2 regularizer on Conv/Dense layer.
Returns:
A Keras model instance.
"""
input_shape = (224, 224, 3)
img_input = layers.Input(shape=input_shape, batch_size=batch_size)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channels_last
x = img_input
bn_axis = 3
x = layers.ZeroPadding2D(padding=(3, 3), name='conv1_pad')(x)
x = layers.Conv2D(
64, (7, 7),
strides=(2, 2),
padding='valid',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=_gen_l2_regularizer(use_l2_regularizer),
name='conv1')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name='bn_conv1')(x)
x = layers.Activation('relu')(x)
x = layers.MaxPooling2D((3, 3), strides=(2, 2), padding='same')(x)
x = conv_block(x,
3, [64, 64, 256],
stage=2,
block='a',
strides=(1, 1),
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [64, 64, 256],
stage=2,
block='b',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [64, 64, 256],
stage=2,
block='c',
use_l2_regularizer=use_l2_regularizer)
x = conv_block(x,
3, [128, 128, 512],
stage=3,
block='a',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [128, 128, 512],
stage=3,
block='b',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [128, 128, 512],
stage=3,
block='c',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [128, 128, 512],
stage=3,
block='d',
use_l2_regularizer=use_l2_regularizer)
x = conv_block(x,
3, [256, 256, 1024],
stage=4,
block='a',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [256, 256, 1024],
stage=4,
block='b',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [256, 256, 1024],
stage=4,
block='c',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [256, 256, 1024],
stage=4,
block='d',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [256, 256, 1024],
stage=4,
block='e',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [256, 256, 1024],
stage=4,
block='f',
use_l2_regularizer=use_l2_regularizer)
x = conv_block(x,
3, [512, 512, 2048],
stage=5,
block='a',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [512, 512, 2048],
stage=5,
block='b',
use_l2_regularizer=use_l2_regularizer)
x = identity_block(x,
3, [512, 512, 2048],
stage=5,
block='c',
use_l2_regularizer=use_l2_regularizer)
rm_axes = [1, 2
] if backend.image_data_format() == 'channels_last' else [2, 3]
x = layers.Lambda(lambda x: backend.mean(x, rm_axes),
name='reduce_mean')(x)
x = layers.Dense(
num_classes,
kernel_initializer=initializers.RandomNormal(stddev=0.01),
kernel_regularizer=_gen_l2_regularizer(use_l2_regularizer),
bias_regularizer=_gen_l2_regularizer(use_l2_regularizer),
name='fc1000')(x)
# A softmax that is followed by the model loss must be done cannot be done
# in float16 due to numeric issues. So we pass dtype=float32.
x = layers.Activation('softmax', dtype='float32')(x)
# Create model.
return models.Model(img_input, x, name='resnet50')
target_size=(28, 28),
color_mode='grayscale',
batch_size=250,
subset='validation')
# %%
import tensorflow as tf
from tensorflow.python.keras.models import Model
from tensorflow.python.keras import layers
from tensorflow.python.keras import backend as K
from tensorflow.python.keras.optimizers import RMSprop, Adam
from tensorflow.python.keras.callbacks import TensorBoard
K.clear_session()
img_input = layers.Input(shape=(28, 28, 1))
x = layers.Conv2D(32, 3, activation='relu')(img_input)
x = layers.Conv2D(64, 3, activation='relu')(x)
x = layers.MaxPooling2D(2, 2)(x)
x = layers.Dropout(0.25)(x)
x = layers.Flatten()(x)
x = layers.Dense(128, activation='relu')(x)
x = layers.Dropout(0.5)(x)
output = layers.Dense(26, activation='softmax')(x)
model = Model(img_input, output)
model.compile(loss='categorical_crossentropy',
optimizer=Adam(),
metrics=['acc'])
def resnet101(self, num_classes, dtype='float32', batch_size=None):
# TODO(tfboyd): add training argument, just lik resnet56.
"""Instantiates the ResNet50 architecture.
Args:
num_classes: `int` number of classes for image classification.
Returns:
A Keras model instance.
"""
input_shape = (384, 384, 3)
img_input = layers.Input(shape=input_shape,
dtype=dtype,
batch_size=batch_size)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channels_last
x = img_input
bn_axis = 3
x = layers.ZeroPadding2D(padding=(3, 3), name='conv1_pad')(x)
x = layers.Conv2D(64, (7, 7),
strides=(2, 2),
padding='valid',
use_bias=False,
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(
self.L2_WEIGHT_DECAY),
name='conv1')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=self.BATCH_NORM_DECAY,
epsilon=self.BATCH_NORM_EPSILON,
name='bn_conv1')(x)
x = layers.Activation('relu')(x)
x = layers.MaxPooling2D((3, 3), strides=(2, 2), padding='same')(x)
x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1))
x = identity_block(x, 3, [64, 64, 256], stage=2, block='b')
x = identity_block(x, 3, [64, 64, 256], stage=2, block='c')
x = conv_block(x, 3, [128, 128, 512], stage=3, block='a')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='b')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='c')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='d')
x = conv_block(x, 3, [256, 256, 1024], stage=4, block='a')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='b')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='c')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='d')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='e')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='f')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='g')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='h')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='i')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='j')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='k')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='l')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='m')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='n')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='o')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='p')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='q')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='r')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='s')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='t')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='u')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='v')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='w')
x = conv_block(x, 3, [512, 512, 2048], stage=5, block='a')
x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b')
x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c')
rm_axes = [
1, 2
] if backend.image_data_format() == 'channels_last' else [2, 3]
x = layers.Lambda(lambda x: backend.mean(x, rm_axes),
name='reduce_mean')(x)
x = layers.Dense(
num_classes,
kernel_regularizer=regularizers.l2(self.L2_WEIGHT_DECAY),
bias_regularizer=regularizers.l2(self.L2_WEIGHT_DECAY),
name='fc1000')(x)
# TODO(reedwm): Remove manual casts once mixed precision can be enabled with a
# single line of code.
x = backend.cast(x, 'float32')
x = layers.Activation('softmax')(x)
# Create model.
return models.Model(img_input, x, name='resnet50')
def ResNet(stack_fn,
preact,
use_bias,
model_name='resnet',
include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
classifier_activation='softmax',
**kwargs):
"""Instantiates the ResNet, ResNetV2, and ResNeXt architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Caution: Be sure to properly pre-process your inputs to the application.
Please see `applications.resnet.preprocess_input` for an example.
Arguments:
stack_fn: a function that returns output tensor for the
stacked residual blocks.
preact: whether to use pre-activation or not
(True for ResNetV2, False for ResNet and ResNeXt).
use_bias: whether to use biases for convolutional layers or not
(True for ResNet and ResNetV2, False for ResNeXt).
model_name: string, model name.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)` (with `channels_last` data format)
or `(3, 224, 224)` (with `channels_first` data format).
It should have exactly 3 inputs channels.
pooling: optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional layer.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional layer, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
**kwargs: For backwards compatibility only.
Returns:
A `keras.Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if 'layers' in kwargs:
global layers
layers = kwargs.pop('layers')
if kwargs:
raise ValueError('Unknown argument(s): %s' % (kwargs, ))
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=224,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
x = layers.ZeroPadding2D(padding=((3, 3), (3, 3)),
name='conv1_pad')(img_input)
x = layers.Conv2D(64, 7, strides=2, use_bias=use_bias,
name='conv1_conv')(x)
if not preact:
x = layers.BatchNormalization(axis=bn_axis,
epsilon=1.001e-5,
name='conv1_bn')(x)
x = layers.Activation('relu', name='conv1_relu')(x)
x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)), name='pool1_pad')(x)
x = layers.MaxPooling2D(3, strides=2, name='pool1_pool')(x)
x = stack_fn(x)
if preact:
x = layers.BatchNormalization(axis=bn_axis,
epsilon=1.001e-5,
name='post_bn')(x)
x = layers.Activation('relu', name='post_relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Dense(classes,
activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name=model_name)
# Load weights.
if (weights == 'imagenet') and (model_name in WEIGHTS_HASHES):
if include_top:
file_name = model_name + '_weights_tf_dim_ordering_tf_kernels.h5'
file_hash = WEIGHTS_HASHES[model_name][0]
else:
file_name = model_name + '_weights_tf_dim_ordering_tf_kernels_notop.h5'
file_hash = WEIGHTS_HASHES[model_name][1]
weights_path = data_utils.get_file(file_name,
BASE_WEIGHTS_PATH + file_name,
cache_subdir='models',
file_hash=file_hash)
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def InceptionResNetV2(include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000):
"""Instantiates the Inception-ResNet v2 architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Arguments:
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor (i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is `False` (otherwise the input shape
has to be `(299, 299, 3)` (with `'channels_last'` data format)
or `(3, 299, 299)` (with `'channels_first'` data format).
It should have exactly 3 inputs channels,
and width and height should be no smaller than 75.
E.g. `(150, 150, 3)` would be one valid value.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the last convolutional block.
- `'avg'` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `'max'` means that global max pooling will be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is `True`, and
if no `weights` argument is specified.
Returns:
A Keras `Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
"""
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=299,
min_size=75,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
# Stem block: 35 x 35 x 192
x = conv2d_bn(img_input, 32, 3, strides=2, padding='valid')
x = conv2d_bn(x, 32, 3, padding='valid')
x = conv2d_bn(x, 64, 3)
x = layers.MaxPooling2D(3, strides=2)(x)
x = conv2d_bn(x, 80, 1, padding='valid')
x = conv2d_bn(x, 192, 3, padding='valid')
x = layers.MaxPooling2D(3, strides=2)(x)
# Mixed 5b (Inception-A block): 35 x 35 x 320
branch_0 = conv2d_bn(x, 96, 1)
branch_1 = conv2d_bn(x, 48, 1)
branch_1 = conv2d_bn(branch_1, 64, 5)
branch_2 = conv2d_bn(x, 64, 1)
branch_2 = conv2d_bn(branch_2, 96, 3)
branch_2 = conv2d_bn(branch_2, 96, 3)
branch_pool = layers.AveragePooling2D(3, strides=1, padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 64, 1)
branches = [branch_0, branch_1, branch_2, branch_pool]
channel_axis = 1 if backend.image_data_format() == 'channels_first' else 3
x = layers.Concatenate(axis=channel_axis, name='mixed_5b')(branches)
# 10x block35 (Inception-ResNet-A block): 35 x 35 x 320
for block_idx in range(1, 11):
x = inception_resnet_block(x,
scale=0.17,
block_type='block35',
block_idx=block_idx)
# Mixed 6a (Reduction-A block): 17 x 17 x 1088
branch_0 = conv2d_bn(x, 384, 3, strides=2, padding='valid')
branch_1 = conv2d_bn(x, 256, 1)
branch_1 = conv2d_bn(branch_1, 256, 3)
branch_1 = conv2d_bn(branch_1, 384, 3, strides=2, padding='valid')
branch_pool = layers.MaxPooling2D(3, strides=2, padding='valid')(x)
branches = [branch_0, branch_1, branch_pool]
x = layers.Concatenate(axis=channel_axis, name='mixed_6a')(branches)
# 20x block17 (Inception-ResNet-B block): 17 x 17 x 1088
for block_idx in range(1, 21):
x = inception_resnet_block(x,
scale=0.1,
block_type='block17',
block_idx=block_idx)
# Mixed 7a (Reduction-B block): 8 x 8 x 2080
branch_0 = conv2d_bn(x, 256, 1)
branch_0 = conv2d_bn(branch_0, 384, 3, strides=2, padding='valid')
branch_1 = conv2d_bn(x, 256, 1)
branch_1 = conv2d_bn(branch_1, 288, 3, strides=2, padding='valid')
branch_2 = conv2d_bn(x, 256, 1)
branch_2 = conv2d_bn(branch_2, 288, 3)
branch_2 = conv2d_bn(branch_2, 320, 3, strides=2, padding='valid')
branch_pool = layers.MaxPooling2D(3, strides=2, padding='valid')(x)
branches = [branch_0, branch_1, branch_2, branch_pool]
x = layers.Concatenate(axis=channel_axis, name='mixed_7a')(branches)
# 10x block8 (Inception-ResNet-C block): 8 x 8 x 2080
for block_idx in range(1, 10):
x = inception_resnet_block(x,
scale=0.2,
block_type='block8',
block_idx=block_idx)
x = inception_resnet_block(x,
scale=1.,
activation=None,
block_type='block8',
block_idx=10)
# Final convolution block: 8 x 8 x 1536
x = conv2d_bn(x, 1536, 1, name='conv_7b')
if include_top:
# Classification block
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = layers.Dense(classes, activation='softmax', name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name='inception_resnet_v2')
# Load weights.
if weights == 'imagenet':
if include_top:
fname = 'inception_resnet_v2_weights_tf_dim_ordering_tf_kernels.h5'
weights_path = data_utils.get_file(
fname,
BASE_WEIGHT_URL + fname,
cache_subdir='models',
file_hash='e693bd0210a403b3192acc6073ad2e96')
else:
fname = ('inception_resnet_v2_weights_'
'tf_dim_ordering_tf_kernels_notop.h5')
weights_path = data_utils.get_file(
fname,
BASE_WEIGHT_URL + fname,
cache_subdir='models',
file_hash='d19885ff4a710c122648d3b5c3b684e4')
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def VGG19(include_top=True,
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000):
""" Instantiates the LeNet-5 architecture.
Args:
include_top: whether to include the 3 fully-connected
layers at the top of the network.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)`
(with `channels_last` data format)
or `(3, 224, 224)` (with `channels_first` data format).
It should have exactly 3 input channels,
and width and height should be no smaller than 32.
E.g. `(200, 200, 3)` would be one valid value.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
Returns:
A Keras model instance.
"""
if classes == 1000:
raise ValueError('If use dataset is `imagenet`, please use it'
'otherwise please use classifier images classes.')
if input_shape == (32, 32, 1):
raise ValueError(
'If use input shape is `32 * 32 * 1`, please don`t use it! '
'So you should change network architecture '
'or use input shape is `224 * 224 * 3`.')
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
# Block 1
x = layers.Conv2D(64, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block1_conv1')(img_input)
x = layers.Conv2D(64, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block1_conv2')(x)
x = layers.MaxPooling2D((2, 2), strides=(2, 2), name='block1_pool')(x)
# Block 2
x = layers.Conv2D(128, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block2_conv1')(x)
x = layers.Conv2D(128, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block2_conv2')(x)
x = layers.MaxPooling2D((2, 2), strides=(2, 2), name='block2_pool')(x)
# Block 3
x = layers.Conv2D(256, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block3_conv1')(x)
x = layers.Conv2D(256, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block3_conv2')(x)
x = layers.Conv2D(256, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block3_conv3')(x)
x = layers.Conv2D(256, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block3_conv4')(x)
x = layers.MaxPooling2D((2, 2), strides=(2, 2), name='block3_pool')(x)
# Block 4
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block4_conv1')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block4_conv2')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block4_conv3')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block4_conv4')(x)
x = layers.MaxPooling2D((2, 2), strides=(2, 2), name='block4_pool')(x)
# Block 5
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block5_conv1')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block5_conv2')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block5_conv3')(x)
x = layers.Conv2D(512, (3, 3),
activation=tf.nn.relu,
padding='same',
name='block5_conv4')(x)
x = layers.MaxPooling2D((2, 2), strides=(2, 2), name='block5_pool')(x)
if include_top:
# Classification block
x = layers.Flatten(name='flatten')(x)
x = layers.Dense(4096, activation=tf.nn.relu, name='fc1')(x)
x = layers.Dense(4096, activation=tf.nn.relu, name='fc2')(x)
x = layers.Dense(classes, activation=tf.nn.softmax,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = models.Model(inputs, x, name='vgg16')
return model
def main(cluster_mode, max_epoch, file_path, batch_size):
if cluster_mode == "local":
init_orca_context(cluster_mode="local", cores=4, memory="3g")
elif cluster_mode == "yarn":
init_orca_context(cluster_mode="yarn-client",
num_nodes=2,
cores=2,
driver_memory="3g")
load_data(file_path)
img_dir = os.path.join(file_path, "train")
label_dir = os.path.join(file_path, "train_masks")
# Here we only take the first 1000 files for simplicity
df_train = pd.read_csv(os.path.join(file_path, 'train_masks.csv'))
ids_train = df_train['img'].map(lambda s: s.split('.')[0])
ids_train = ids_train[:1000]
x_train_filenames = []
y_train_filenames = []
for img_id in ids_train:
x_train_filenames.append(os.path.join(img_dir,
"{}.jpg".format(img_id)))
y_train_filenames.append(
os.path.join(label_dir, "{}_mask.gif".format(img_id)))
x_train_filenames, x_val_filenames, y_train_filenames, y_val_filenames = \
train_test_split(x_train_filenames, y_train_filenames, test_size=0.2, random_state=42)
def load_and_process_image(path):
array = mpimg.imread(path)
result = np.array(Image.fromarray(array).resize(size=(128, 128)))
result = result.astype(float)
result /= 255.0
return result
def load_and_process_image_label(path):
array = mpimg.imread(path)
result = np.array(Image.fromarray(array).resize(size=(128, 128)))
result = np.expand_dims(result[:, :, 1], axis=-1)
result = result.astype(float)
result /= 255.0
return result
train_images = np.stack(
[load_and_process_image(filepath) for filepath in x_train_filenames])
train_label_images = np.stack([
load_and_process_image_label(filepath)
for filepath in y_train_filenames
])
val_images = np.stack(
[load_and_process_image(filepath) for filepath in x_val_filenames])
val_label_images = np.stack([
load_and_process_image_label(filepath) for filepath in y_val_filenames
])
train_shards = XShards.partition({
"x": train_images,
"y": train_label_images
})
val_shards = XShards.partition({"x": val_images, "y": val_label_images})
# Build the U-Net model
def conv_block(input_tensor, num_filters):
encoder = layers.Conv2D(num_filters, (3, 3),
padding='same')(input_tensor)
encoder = layers.Activation('relu')(encoder)
encoder = layers.Conv2D(num_filters, (3, 3), padding='same')(encoder)
encoder = layers.Activation('relu')(encoder)
return encoder
def encoder_block(input_tensor, num_filters):
encoder = conv_block(input_tensor, num_filters)
encoder_pool = layers.MaxPooling2D((2, 2), strides=(2, 2))(encoder)
return encoder_pool, encoder
def decoder_block(input_tensor, concat_tensor, num_filters):
decoder = layers.Conv2DTranspose(num_filters, (2, 2),
strides=(2, 2),
padding='same')(input_tensor)
decoder = layers.concatenate([concat_tensor, decoder], axis=-1)
decoder = layers.Activation('relu')(decoder)
decoder = layers.Conv2D(num_filters, (3, 3), padding='same')(decoder)
decoder = layers.Activation('relu')(decoder)
decoder = layers.Conv2D(num_filters, (3, 3), padding='same')(decoder)
decoder = layers.Activation('relu')(decoder)
return decoder
inputs = layers.Input(shape=(128, 128, 3)) # 128
encoder0_pool, encoder0 = encoder_block(inputs, 16) # 64
encoder1_pool, encoder1 = encoder_block(encoder0_pool, 32) # 32
encoder2_pool, encoder2 = encoder_block(encoder1_pool, 64) # 16
encoder3_pool, encoder3 = encoder_block(encoder2_pool, 128) # 8
center = conv_block(encoder3_pool, 256) # center
decoder3 = decoder_block(center, encoder3, 128) # 16
decoder2 = decoder_block(decoder3, encoder2, 64) # 32
decoder1 = decoder_block(decoder2, encoder1, 32) # 64
decoder0 = decoder_block(decoder1, encoder0, 16) # 128
outputs = layers.Conv2D(1, (1, 1), activation='sigmoid')(decoder0)
net = models.Model(inputs=[inputs], outputs=[outputs])
# Define custom metrics
def dice_coeff(y_true, y_pred):
smooth = 1.
# Flatten
y_true_f = tf.reshape(y_true, [-1])
y_pred_f = tf.reshape(y_pred, [-1])
intersection = tf.reduce_sum(y_true_f * y_pred_f)
score = (2. * intersection + smooth) / \
(tf.reduce_sum(y_true_f) + tf.reduce_sum(y_pred_f) + smooth)
return score
# Define custom loss function
def dice_loss(y_true, y_pred):
loss = 1 - dice_coeff(y_true, y_pred)
return loss
def bce_dice_loss(y_true, y_pred):
loss = losses.binary_crossentropy(y_true, y_pred) + dice_loss(
y_true, y_pred)
return loss
# compile model
net.compile(optimizer=tf.keras.optimizers.Adam(2e-3), loss=bce_dice_loss)
print(net.summary())
# create an estimator from keras model
est = Estimator.from_keras(keras_model=net)
# fit with estimator
est.fit(data=train_shards, batch_size=batch_size, epochs=max_epoch)
# evaluate with estimator
result = est.evaluate(val_shards)
print(result)
# predict with estimator
val_shards.cache()
val_image_shards = val_shards.transform_shard(
lambda val_dict: {"x": val_dict["x"]})
pred_shards = est.predict(data=val_image_shards, batch_size=batch_size)
pred = pred_shards.collect()[0]["prediction"]
val_image_label = val_shards.collect()[0]
val_image = val_image_label["x"]
val_label = val_image_label["y"]
# visualize 5 predicted results
plt.figure(figsize=(10, 20))
for i in range(5):
img = val_image[i]
label = val_label[i]
predicted_label = pred[i]
plt.subplot(5, 3, 3 * i + 1)
plt.imshow(img)
plt.title("Input image")
plt.subplot(5, 3, 3 * i + 2)
plt.imshow(label[:, :, 0], cmap='gray')
plt.title("Actual Mask")
plt.subplot(5, 3, 3 * i + 3)
plt.imshow(predicted_label, cmap='gray')
plt.title("Predicted Mask")
plt.suptitle("Examples of Input Image, Label, and Prediction")
plt.show()
stop_orca_context()
def PINet_CIFAR10():
## model
input_shape = [32,32,3]
initial_conv_width=3
initial_stride=1
initial_filters=64
initial_pool_width=3
initial_pool_stride=2
use_global_pooling = True
dropout_rate = 0.2
model_input = layers.Input(shape=input_shape)
x = layers.Conv2D(
128,
initial_conv_width,
strides=initial_stride,
padding="same")(model_input)
x = layers.BatchNormalization()(x)
x = layers.Activation("relu")(x)
x = layers.MaxPooling2D(
pool_size=initial_pool_width,
strides=initial_pool_stride,
padding="same")(x)
x = layers.Conv2D(
256,
initial_conv_width,
strides=initial_stride,
padding="same")(x)
x = layers.BatchNormalization()(x)
x = layers.Activation("relu")(x)
x = layers.MaxPooling2D(
pool_size=initial_pool_width,
strides=initial_pool_stride,
padding="same")(x)
x = layers.Conv2D(
512,
initial_conv_width,
strides=initial_stride,
padding="same")(x)
x = layers.BatchNormalization()(x)
x = layers.Activation("relu")(x)
x = layers.MaxPooling2D(
pool_size=initial_pool_width,
strides=initial_pool_stride,
padding="same")(x)
x = layers.Conv2D(
1024,
initial_conv_width,
strides=initial_stride,
padding="same")(x)
x = layers.BatchNormalization()(x)
x = layers.Activation("relu")(x)
if use_global_pooling:
x = layers.GlobalAveragePooling2D()(x)
x_logits1 = layers.Dense(2500, activation="relu")(x)
x_logits1_reshape = layers.Reshape((1,50,50))(x_logits1)
x_logits1_reshape = layers.Permute((2,3,1))(x_logits1_reshape)
x_logits2 = layers.Conv2DTranspose(
3,
50,
strides=initial_stride,
padding="same")(x_logits1_reshape)
x_logits2 = layers.BatchNormalization()(x_logits2)
x_logits2 = layers.Activation("relu")(x_logits2)
model_output = layers.Flatten()(x_logits2)
model = models.Model(model_input, model_output)
return model
def EfficientNet(width_coefficient,
depth_coefficient,
default_size,
dropout_rate=0.2,
drop_connect_rate=0.2,
depth_divisor=8,
activation='swish',
blocks_args='default',
model_name='efficientnet',
include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
classifier_activation='softmax'):
"""Instantiates the EfficientNet architecture using given scaling coefficients.
Reference paper:
- [EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks](
https://arxiv.org/abs/1905.11946) (ICML 2019)
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Arguments:
width_coefficient: float, scaling coefficient for network width.
depth_coefficient: float, scaling coefficient for network depth.
default_size: integer, default input image size.
dropout_rate: float, dropout rate before final classifier layer.
drop_connect_rate: float, dropout rate at skip connections.
depth_divisor: integer, a unit of network width.
activation: activation function.
blocks_args: list of dicts, parameters to construct block modules.
model_name: string, model name.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False.
It should have exactly 3 inputs channels.
pooling: optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional layer.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional layer, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
Returns:
A `keras.Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if blocks_args == 'default':
blocks_args = DEFAULT_BLOCKS_ARGS
if not (weights in {'imagenet', None} or file_io.file_exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
bn_axis = 3 if backend.image_data_format() == 'channels_last' else 1
def round_filters(filters, divisor=depth_divisor):
"""Round number of filters based on depth multiplier."""
filters *= width_coefficient
new_filters = max(divisor,
int(filters + divisor / 2) // divisor * divisor)
# Make sure that round down does not go down by more than 10%.
if new_filters < 0.9 * filters:
new_filters += divisor
return int(new_filters)
def round_repeats(repeats):
"""Round number of repeats based on depth multiplier."""
return int(math.ceil(depth_coefficient * repeats))
# Build stem
x = img_input
x = layers.Rescaling(1. / 255.)(x)
x = layers.Normalization(axis=bn_axis)(x)
x = layers.ZeroPadding2D(padding=imagenet_utils.correct_pad(x, 3),
name='stem_conv_pad')(x)
x = layers.Conv2D(round_filters(32),
3,
strides=2,
padding='valid',
use_bias=False,
kernel_initializer=CONV_KERNEL_INITIALIZER,
name='stem_conv')(x)
x = layers.BatchNormalization(axis=bn_axis, name='stem_bn')(x)
x = layers.Activation(activation, name='stem_activation')(x)
# Build blocks
blocks_args = copy.deepcopy(blocks_args)
b = 0
blocks = float(sum(args['repeats'] for args in blocks_args))
for (i, args) in enumerate(blocks_args):
assert args['repeats'] > 0
# Update block input and output filters based on depth multiplier.
args['filters_in'] = round_filters(args['filters_in'])
args['filters_out'] = round_filters(args['filters_out'])
for j in range(round_repeats(args.pop('repeats'))):
# The first block needs to take care of stride and filter size increase.
if j > 0:
args['strides'] = 1
args['filters_in'] = args['filters_out']
x = block(x,
activation,
drop_connect_rate * b / blocks,
name='block{}{}_'.format(i + 1, chr(j + 97)),
**args)
b += 1
# Build top
x = layers.Conv2D(round_filters(1280),
1,
padding='same',
use_bias=False,
kernel_initializer=CONV_KERNEL_INITIALIZER,
name='top_conv')(x)
x = layers.BatchNormalization(axis=bn_axis, name='top_bn')(x)
x = layers.Activation(activation, name='top_activation')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
if dropout_rate > 0:
x = layers.Dropout(dropout_rate, name='top_dropout')(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Dense(classes,
activation=classifier_activation,
kernel_initializer=DENSE_KERNEL_INITIALIZER,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name=model_name)
# Load weights.
if weights == 'imagenet':
if include_top:
file_suffix = '.h5'
file_hash = WEIGHTS_HASHES[model_name[-2:]][0]
else:
file_suffix = '_notop.h5'
file_hash = WEIGHTS_HASHES[model_name[-2:]][1]
file_name = model_name + file_suffix
weights_path = data_utils.get_file(file_name,
BASE_WEIGHTS_PATH + file_name,
cache_subdir='models',
file_hash=file_hash)
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def test_dynamic_loss_scaling(self,
strategy_fn,
pass_loss_scale_to_policy=False,
get_config=False):
strategy = strategy_fn()
initial_loss_scale = 2.
batch_size = 4
loss_scale = loss_scale_module.DynamicLossScale(
initial_loss_scale=initial_loss_scale, increment_period=2)
expected_gradient = backend.variable([initial_loss_scale / batch_size],
dtype=dtypes.float16)
# If this variable is set to True, the model below will have NaN gradients
have_nan_gradients = backend.variable(False, dtype=dtypes.bool)
with strategy.scope():
opt = gradient_descent.SGD(1.)
if pass_loss_scale_to_policy:
p = policy.Policy('mixed_float16', loss_scale=loss_scale)
else:
p = policy.Policy('mixed_float16', loss_scale=None)
opt = loss_scale_optimizer.LossScaleOptimizer(opt, loss_scale)
with policy.policy_scope(p):
x = layers.Input(
shape=(1,), batch_size=batch_size, dtype=dtypes.float16)
layer = mp_test_util.MultiplyLayer(assert_type=dtypes.float16)
y = layer(x)
identity_with_nan_grads = (
mp_test_util.create_identity_with_nan_gradients_fn(
have_nan_gradients))
y = core.Lambda(identity_with_nan_grads)(y)
identity_with_grad_check_fn = (
mp_test_util.create_identity_with_grad_check_fn(
expected_dtype=dtypes.float16,
expected_gradient=expected_gradient))
y = core.Lambda(identity_with_grad_check_fn)(y)
model = models.Model(inputs=x, outputs=y)
if get_config:
config = model.get_config()
model = model.__class__.from_config(
config,
custom_objects={'MultiplyLayer': mp_test_util.MultiplyLayer})
(layer,) = (layer for layer in model.layers
if isinstance(layer, mp_test_util.MultiplyLayer))
def loss_fn(y_true, y_pred):
del y_true
return math_ops.reduce_mean(y_pred)
model.compile(
opt,
loss=loss_fn,
run_eagerly=testing_utils.should_run_eagerly())
self.assertEqual(backend.eval(layer.v), 1)
x = np.ones((batch_size, 1))
y = np.ones((batch_size, 1))
dataset = dataset_ops.Dataset.from_tensor_slices((x, y)).batch(batch_size)
model.fit(dataset)
# The variables starts with 1 and has a gradient of 1, so will go down by 1
# each step.
self.assertEqual(backend.eval(layer.v), 0)
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -1)
# There have been two steps without NaNs, so the loss scale will double
backend.set_value(expected_gradient,
backend.get_value(expected_gradient * 2))
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -2)
# Next test with NaN gradients.
backend.set_value(have_nan_gradients, True)
model.fit(dataset)
# Variable should not be updated
self.assertEqual(backend.eval(layer.v), -2)
# Test with finite gradients again
backend.set_value(have_nan_gradients, False)
# The loss scale will be halved due to the NaNs, so the gradient will also
# be halved
backend.set_value(expected_gradient,
backend.get_value(expected_gradient / 2))
model.fit(dataset)
self.assertEqual(backend.eval(layer.v), -3)
def Xception(include_top=True,
weights='imagenet',
input_tensor=None,
input_shape=None,
pooling=None,
classes=1000,
classifier_activation='softmax'):
"""Instantiates the Xception architecture.
Reference:
- [Xception: Deep Learning with Depthwise Separable Convolutions](
https://arxiv.org/abs/1610.02357) (CVPR 2017)
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
Note that the default input image size for this model is 299x299.
Caution: Be sure to properly pre-process your inputs to the application.
Please see `applications.xception.preprocess_input` for an example.
Arguments:
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor
(i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(299, 299, 3)`.
It should have exactly 3 inputs channels,
and width and height should be no smaller than 71.
E.g. `(150, 150, 3)` would be one valid value.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True,
and if no `weights` argument is specified.
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
Returns:
A `keras.Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if not (weights in {'imagenet', None} or file_io.file_exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
# Determine proper input shape
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=299,
min_size=71,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
channel_axis = 1 if backend.image_data_format() == 'channels_first' else -1
x = layers.Conv2D(32, (3, 3),
strides=(2, 2),
use_bias=False,
padding="same",
name='block1_conv1')(img_input)
x = layers.BatchNormalization(axis=channel_axis, name='block1_conv1_bn')(x)
x = layers.Activation('relu', name='block1_conv1_act')(x)
x = layers.Conv2D(64, (3, 3),
padding="same",
use_bias=False,
name='block1_conv2')(x)
x = layers.BatchNormalization(axis=channel_axis, name='block1_conv2_bn')(x)
x = layers.Activation('relu', name='block1_conv2_act')(x)
residual = layers.Conv2D(128, (1, 1),
strides=(2, 2),
padding='same',
use_bias=False)(x)
residual = layers.BatchNormalization(axis=channel_axis)(residual)
x = layers.SeparableConv2D(128, (3, 3),
padding='same',
use_bias=False,
name='block2_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block2_sepconv1_bn')(x)
x = layers.Activation('relu', name='block2_sepconv2_act')(x)
x = layers.SeparableConv2D(128, (3, 3),
padding='same',
use_bias=False,
name='block2_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block2_sepconv2_bn')(x)
x = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='same',
name='block2_pool')(x)
x = layers.add([x, residual])
residual = layers.Conv2D(256, (1, 1),
strides=(2, 2),
padding='same',
use_bias=False)(x)
residual = layers.BatchNormalization(axis=channel_axis)(residual)
x = layers.Activation('relu', name='block3_sepconv1_act')(x)
x = layers.SeparableConv2D(256, (3, 3),
padding='same',
use_bias=False,
name='block3_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block3_sepconv1_bn')(x)
x = layers.Activation('relu', name='block3_sepconv2_act')(x)
x = layers.SeparableConv2D(256, (3, 3),
padding='same',
use_bias=False,
name='block3_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block3_sepconv2_bn')(x)
x = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='same',
name='block3_pool')(x)
x = layers.add([x, residual])
residual = layers.Conv2D(728, (1, 1),
strides=(2, 2),
padding='same',
use_bias=False)(x)
residual = layers.BatchNormalization(axis=channel_axis)(residual)
x = layers.Activation('relu', name='block4_sepconv1_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name='block4_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block4_sepconv1_bn')(x)
x = layers.Activation('relu', name='block4_sepconv2_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name='block4_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block4_sepconv2_bn')(x)
x = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='same',
name='block4_pool')(x)
x = layers.add([x, residual])
for i in range(8):
residual = x
prefix = 'block' + str(i + 5)
x = layers.Activation('relu', name=prefix + '_sepconv1_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name=prefix + '_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name=prefix + '_sepconv1_bn')(x)
x = layers.Activation('relu', name=prefix + '_sepconv2_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name=prefix + '_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name=prefix + '_sepconv2_bn')(x)
x = layers.Activation('relu', name=prefix + '_sepconv3_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name=prefix + '_sepconv3')(x)
x = layers.BatchNormalization(axis=channel_axis,
name=prefix + '_sepconv3_bn')(x)
x = layers.add([x, residual])
residual = layers.Conv2D(1024, (1, 1),
strides=(2, 2),
padding='same',
use_bias=False)(x)
residual = layers.BatchNormalization(axis=channel_axis)(residual)
x = layers.Activation('relu', name='block13_sepconv1_act')(x)
x = layers.SeparableConv2D(728, (3, 3),
padding='same',
use_bias=False,
name='block13_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block13_sepconv1_bn')(x)
x = layers.Activation('relu', name='block13_sepconv2_act')(x)
x = layers.SeparableConv2D(1024, (3, 3),
padding='same',
use_bias=False,
name='block13_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block13_sepconv2_bn')(x)
x = layers.MaxPooling2D((3, 3),
strides=(2, 2),
padding='same',
name='block13_pool')(x)
x = layers.add([x, residual])
x = layers.SeparableConv2D(1536, (3, 3),
padding='same',
use_bias=False,
name='block14_sepconv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block14_sepconv1_bn')(x)
x = layers.Activation('relu', name='block14_sepconv1_act')(x)
x = layers.SeparableConv2D(2048, (3, 3),
padding='same',
use_bias=False,
name='block14_sepconv2')(x)
x = layers.BatchNormalization(axis=channel_axis,
name='block14_sepconv2_bn')(x)
x = layers.Activation('relu', name='block14_sepconv2_act')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Dense(classes,
activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name='xception')
# Load weights.
if weights == 'imagenet':
if include_top:
weights_path = data_utils.get_file(
'xception_weights_tf_dim_ordering_tf_kernels.h5',
TF_WEIGHTS_PATH,
cache_subdir='models',
file_hash='0a58e3b7378bc2990ea3b43d5981f1f6')
else:
weights_path = data_utils.get_file(
'xception_weights_tf_dim_ordering_tf_kernels_notop.h5',
TF_WEIGHTS_PATH_NO_TOP,
cache_subdir='models',
file_hash='b0042744bf5b25fce3cb969f33bebb97')
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def layer_test(layer_cls,
kwargs=None,
input_shape=None,
input_dtype=None,
input_data=None,
expected_output=None,
expected_output_dtype=None,
expected_output_shape=None,
validate_training=True,
adapt_data=None,
custom_objects=None,
test_harness=None):
"""Test routine for a layer with a single input and single output.
Arguments:
layer_cls: Layer class object.
kwargs: Optional dictionary of keyword arguments for instantiating the
layer.
input_shape: Input shape tuple.
input_dtype: Data type of the input data.
input_data: Numpy array of input data.
expected_output: Numpy array of the expected output.
expected_output_dtype: Data type expected for the output.
expected_output_shape: Shape tuple for the expected shape of the output.
validate_training: Whether to attempt to validate training on this layer.
This might be set to False for non-differentiable layers that output
string or integer values.
adapt_data: Optional data for an 'adapt' call. If None, adapt() will not
be tested for this layer. This is only relevant for PreprocessingLayers.
custom_objects: Optional dictionary mapping name strings to custom objects
in the layer class. This is helpful for testing custom layers.
test_harness: The Tensorflow test, if any, that this function is being
called in.
Returns:
The output data (Numpy array) returned by the layer, for additional
checks to be done by the calling code.
Raises:
ValueError: if `input_shape is None`.
"""
if input_data is None:
if input_shape is None:
raise ValueError('input_shape is None')
if not input_dtype:
input_dtype = 'float32'
input_data_shape = list(input_shape)
for i, e in enumerate(input_data_shape):
if e is None:
input_data_shape[i] = np.random.randint(1, 4)
input_data = 10 * np.random.random(input_data_shape)
if input_dtype[:5] == 'float':
input_data -= 0.5
input_data = input_data.astype(input_dtype)
elif input_shape is None:
input_shape = input_data.shape
if input_dtype is None:
input_dtype = input_data.dtype
if expected_output_dtype is None:
expected_output_dtype = input_dtype
if dtypes.as_dtype(expected_output_dtype) == dtypes.string:
if test_harness:
assert_equal = test_harness.assertAllEqual
else:
assert_equal = string_test
else:
if test_harness:
assert_equal = test_harness.assertAllClose
else:
assert_equal = numeric_test
# instantiation
kwargs = kwargs or {}
layer = layer_cls(**kwargs)
# Test adapt, if data was passed.
if adapt_data is not None:
layer.adapt(adapt_data)
# test get_weights , set_weights at layer level
weights = layer.get_weights()
layer.set_weights(weights)
# test and instantiation from weights
if 'weights' in tf_inspect.getargspec(layer_cls.__init__):
kwargs['weights'] = weights
layer = layer_cls(**kwargs)
# test in functional API
x = layers.Input(shape=input_shape[1:], dtype=input_dtype)
y = layer(x)
if backend.dtype(y) != expected_output_dtype:
raise AssertionError(
'When testing layer %s, for input %s, found output '
'dtype=%s but expected to find %s.\nFull kwargs: %s' %
(layer_cls.__name__, x, backend.dtype(y), expected_output_dtype,
kwargs))
def assert_shapes_equal(expected, actual):
"""Asserts that the output shape from the layer matches the actual shape."""
if len(expected) != len(actual):
raise AssertionError(
'When testing layer %s, for input %s, found output_shape='
'%s but expected to find %s.\nFull kwargs: %s' %
(layer_cls.__name__, x, actual, expected, kwargs))
for expected_dim, actual_dim in zip(expected, actual):
if isinstance(expected_dim, tensor_shape.Dimension):
expected_dim = expected_dim.value
if isinstance(actual_dim, tensor_shape.Dimension):
actual_dim = actual_dim.value
if expected_dim is not None and expected_dim != actual_dim:
raise AssertionError(
'When testing layer %s, for input %s, found output_shape='
'%s but expected to find %s.\nFull kwargs: %s' %
(layer_cls.__name__, x, actual, expected, kwargs))
if expected_output_shape is not None:
assert_shapes_equal(tensor_shape.TensorShape(expected_output_shape),
y.shape)
# check shape inference
model = models.Model(x, y)
computed_output_shape = tuple(
layer.compute_output_shape(
tensor_shape.TensorShape(input_shape)).as_list())
computed_output_signature = layer.compute_output_signature(
tensor_spec.TensorSpec(shape=input_shape, dtype=input_dtype))
actual_output = model.predict(input_data)
actual_output_shape = actual_output.shape
assert_shapes_equal(computed_output_shape, actual_output_shape)
assert_shapes_equal(computed_output_signature.shape, actual_output_shape)
if computed_output_signature.dtype != actual_output.dtype:
raise AssertionError(
'When testing layer %s, for input %s, found output_dtype='
'%s but expected to find %s.\nFull kwargs: %s' %
(layer_cls.__name__, x, actual_output.dtype,
computed_output_signature.dtype, kwargs))
if expected_output is not None:
assert_equal(actual_output, expected_output)
# test serialization, weight setting at model level
model_config = model.get_config()
recovered_model = models.Model.from_config(model_config, custom_objects)
if model.weights:
weights = model.get_weights()
recovered_model.set_weights(weights)
output = recovered_model.predict(input_data)
assert_equal(output, actual_output)
# test training mode (e.g. useful for dropout tests)
# Rebuild the model to avoid the graph being reused between predict() and
# See b/120160788 for more details. This should be mitigated after 2.0.
layer_weights = layer.get_weights(
) # Get the layer weights BEFORE training.
if validate_training:
model = models.Model(x, layer(x))
if _thread_local_data.run_eagerly is not None:
model.compile('rmsprop',
'mse',
weighted_metrics=['acc'],
run_eagerly=should_run_eagerly())
else:
model.compile('rmsprop', 'mse', weighted_metrics=['acc'])
model.train_on_batch(input_data, actual_output)
# test as first layer in Sequential API
layer_config = layer.get_config()
layer_config['batch_input_shape'] = input_shape
layer = layer.__class__.from_config(layer_config)
# Test adapt, if data was passed.
if adapt_data is not None:
layer.adapt(adapt_data)
model = models.Sequential()
model.add(layers.Input(shape=input_shape[1:], dtype=input_dtype))
model.add(layer)
layer.set_weights(layer_weights)
actual_output = model.predict(input_data)
actual_output_shape = actual_output.shape
for expected_dim, actual_dim in zip(computed_output_shape,
actual_output_shape):
if expected_dim is not None:
if expected_dim != actual_dim:
raise AssertionError(
'When testing layer %s **after deserialization**, '
'for input %s, found output_shape='
'%s but expected to find inferred shape %s.\nFull kwargs: %s'
% (layer_cls.__name__, x, actual_output_shape,
computed_output_shape, kwargs))
if expected_output is not None:
assert_equal(actual_output, expected_output)
# test serialization, weight setting at model level
model_config = model.get_config()
recovered_model = models.Sequential.from_config(model_config,
custom_objects)
if model.weights:
weights = model.get_weights()
recovered_model.set_weights(weights)
output = recovered_model.predict(input_data)
assert_equal(output, actual_output)
# for further checks in the caller function
return actual_output
def InceptionV3(input_shape=None):
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=299,
min_size=75,
data_format=backend.image_data_format(),
require_flatten=True)
img_input = layers.Input(shape=input_shape)
if backend.image_data_format() == 'channels_first':
channel_axis = 1
else:
channel_axis = 3
x = conv2d_bn(img_input, 32, 3, 3, strides=(2, 2), padding='valid')
x = conv2d_bn(x, 32, 3, 3, padding='valid')
x = conv2d_bn(x, 64, 3, 3)
x = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
x = conv2d_bn(x, 80, 1, 1, padding='valid')
x = conv2d_bn(x, 192, 3, 3, padding='valid')
x = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
# mixed 0: 35 x 35 x 256
branch1x1 = conv2d_bn(x, 64, 1, 1)
branch5x5 = conv2d_bn(x, 48, 1, 1)
branch5x5 = conv2d_bn(branch5x5, 64, 5, 5)
branch3x3dbl = conv2d_bn(x, 64, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 32, 1, 1)
x = layers.concatenate([branch1x1, branch5x5, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed0')
# mixed 1: 35 x 35 x 288
branch1x1 = conv2d_bn(x, 64, 1, 1)
branch5x5 = conv2d_bn(x, 48, 1, 1)
branch5x5 = conv2d_bn(branch5x5, 64, 5, 5)
branch3x3dbl = conv2d_bn(x, 64, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 64, 1, 1)
x = layers.concatenate([branch1x1, branch5x5, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed1')
# mixed 2: 35 x 35 x 288
branch1x1 = conv2d_bn(x, 64, 1, 1)
branch5x5 = conv2d_bn(x, 48, 1, 1)
branch5x5 = conv2d_bn(branch5x5, 64, 5, 5)
branch3x3dbl = conv2d_bn(x, 64, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 64, 1, 1)
x = layers.concatenate([branch1x1, branch5x5, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed2')
# mixed 3: 17 x 17 x 768
branch3x3 = conv2d_bn(x, 384, 3, 3, strides=(2, 2), padding='valid')
branch3x3dbl = conv2d_bn(x, 64, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 96, 3, 3)
branch3x3dbl = conv2d_bn(branch3x3dbl,
96,
3,
3,
strides=(2, 2),
padding='valid')
branch_pool = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
x = layers.concatenate([branch3x3, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed3')
# mixed 4: 17 x 17 x 768
branch1x1 = conv2d_bn(x, 192, 1, 1)
branch7x7 = conv2d_bn(x, 128, 1, 1)
branch7x7 = conv2d_bn(branch7x7, 128, 1, 7)
branch7x7 = conv2d_bn(branch7x7, 192, 7, 1)
branch7x7dbl = conv2d_bn(x, 128, 1, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 128, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 128, 1, 7)
branch7x7dbl = conv2d_bn(branch7x7dbl, 128, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 1, 7)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 192, 1, 1)
x = layers.concatenate([branch1x1, branch7x7, branch7x7dbl, branch_pool],
axis=channel_axis,
name='mixed4')
# mixed 5, 6: 17 x 17 x 768
for i in range(2):
branch1x1 = conv2d_bn(x, 192, 1, 1)
branch7x7 = conv2d_bn(x, 160, 1, 1)
branch7x7 = conv2d_bn(branch7x7, 160, 1, 7)
branch7x7 = conv2d_bn(branch7x7, 192, 7, 1)
branch7x7dbl = conv2d_bn(x, 160, 1, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 160, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 160, 1, 7)
branch7x7dbl = conv2d_bn(branch7x7dbl, 160, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 1, 7)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 192, 1, 1)
x = layers.concatenate(
[branch1x1, branch7x7, branch7x7dbl, branch_pool],
axis=channel_axis,
name='mixed' + str(5 + i))
# mixed 7: 17 x 17 x 768
branch1x1 = conv2d_bn(x, 192, 1, 1)
branch7x7 = conv2d_bn(x, 192, 1, 1)
branch7x7 = conv2d_bn(branch7x7, 192, 1, 7)
branch7x7 = conv2d_bn(branch7x7, 192, 7, 1)
branch7x7dbl = conv2d_bn(x, 192, 1, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 1, 7)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 7, 1)
branch7x7dbl = conv2d_bn(branch7x7dbl, 192, 1, 7)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 192, 1, 1)
x = layers.concatenate([branch1x1, branch7x7, branch7x7dbl, branch_pool],
axis=channel_axis,
name='mixed7')
# mixed 8: 8 x 8 x 1280
branch3x3 = conv2d_bn(x, 192, 1, 1)
branch3x3 = conv2d_bn(branch3x3,
320,
3,
3,
strides=(2, 2),
padding='valid')
branch7x7x3 = conv2d_bn(x, 192, 1, 1)
branch7x7x3 = conv2d_bn(branch7x7x3, 192, 1, 7)
branch7x7x3 = conv2d_bn(branch7x7x3, 192, 7, 1)
branch7x7x3 = conv2d_bn(branch7x7x3,
192,
3,
3,
strides=(2, 2),
padding='valid')
branch_pool = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
x = layers.concatenate([branch3x3, branch7x7x3, branch_pool],
axis=channel_axis,
name='mixed8')
# mixed 9: 8 x 8 x 2048
for i in range(2):
branch1x1 = conv2d_bn(x, 320, 1, 1)
branch3x3 = conv2d_bn(x, 384, 1, 1)
branch3x3_1 = conv2d_bn(branch3x3, 384, 1, 3)
branch3x3_2 = conv2d_bn(branch3x3, 384, 3, 1)
branch3x3 = layers.concatenate([branch3x3_1, branch3x3_2],
axis=channel_axis,
name='mixed9_' + str(i))
branch3x3dbl = conv2d_bn(x, 448, 1, 1)
branch3x3dbl = conv2d_bn(branch3x3dbl, 384, 3, 3)
branch3x3dbl_1 = conv2d_bn(branch3x3dbl, 384, 1, 3)
branch3x3dbl_2 = conv2d_bn(branch3x3dbl, 384, 3, 1)
branch3x3dbl = layers.concatenate([branch3x3dbl_1, branch3x3dbl_2],
axis=channel_axis)
branch_pool = layers.AveragePooling2D((3, 3),
strides=(1, 1),
padding='same')(x)
branch_pool = conv2d_bn(branch_pool, 192, 1, 1)
x = layers.concatenate(
[branch1x1, branch3x3, branch3x3dbl, branch_pool],
axis=channel_axis,
name='mixed' + str(9 + i))
# Classification block
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
imagenet_utils.validate_activation('softmax', None)
x = layers.Dense(NUM_CLASSES, activation='softmax', name='predictions')(x)
# Create model.
model = training.Model(img_input, x, name='inception_v3')
return model
def test():
x = layers.Input(shape=(4, 100, 100, 3))
y = layers.Concatenate(axis=0)([x, x])
m = Model(x, y)
m.summary()
def MobileNetV2(input_shape=None,
alpha=1.0,
include_top=True,
weights='imagenet',
input_tensor=None,
pooling=None,
classes=1000,
**kwargs):
"""Instantiates the MobileNetV2 architecture.
Arguments:
input_shape: optional shape tuple, to be specified if you would
like to use a model with an input img resolution that is not
(224, 224, 3).
It should have exactly 3 inputs channels (224, 224, 3).
You can also omit this option if you would like
to infer input_shape from an input_tensor.
If you choose to include both input_tensor and input_shape then
input_shape will be used if they match, if the shapes
do not match then we will throw an error.
E.g. `(160, 160, 3)` would be one valid value.
alpha: controls the width of the network. This is known as the
width multiplier in the MobileNetV2 paper, but the name is kept for
consistency with MobileNetV1 in Keras.
- If `alpha` < 1.0, proportionally decreases the number
of filters in each layer.
- If `alpha` > 1.0, proportionally increases the number
of filters in each layer.
- If `alpha` = 1, default number of filters from the paper
are used at each layer.
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor (i.e. output of
`layers.Input()`)
to use as image input for the model.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model
will be the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a
2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
**kwargs: For backwards compatibility only.
Returns:
A Keras model instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape or invalid alpha, rows when
weights='imagenet'
"""
if 'layers' in kwargs:
global layers
layers = kwargs.pop('layers')
if kwargs:
raise ValueError('Unknown argument(s): %s' % (kwargs, ))
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top` '
'as true, `classes` should be 1000')
# Determine proper input shape and default size.
# If both input_shape and input_tensor are used, they should match
if input_shape is not None and input_tensor is not None:
try:
is_input_t_tensor = backend.is_keras_tensor(input_tensor)
except ValueError:
try:
is_input_t_tensor = backend.is_keras_tensor(
layer_utils.get_source_inputs(input_tensor))
except ValueError:
raise ValueError('input_tensor: ', input_tensor,
'is not type input_tensor')
if is_input_t_tensor:
if backend.image_data_format == 'channels_first':
if backend.int_shape(input_tensor)[1] != input_shape[1]:
raise ValueError(
'input_shape: ', input_shape, 'and input_tensor: ',
input_tensor,
'do not meet the same shape requirements')
else:
if backend.int_shape(input_tensor)[2] != input_shape[1]:
raise ValueError(
'input_shape: ', input_shape, 'and input_tensor: ',
input_tensor,
'do not meet the same shape requirements')
else:
raise ValueError('input_tensor specified: ', input_tensor,
'is not a keras tensor')
# If input_shape is None, infer shape from input_tensor
if input_shape is None and input_tensor is not None:
try:
backend.is_keras_tensor(input_tensor)
except ValueError:
raise ValueError('input_tensor: ', input_tensor, 'is type: ',
type(input_tensor), 'which is not a valid type')
if input_shape is None and not backend.is_keras_tensor(input_tensor):
default_size = 224
elif input_shape is None and backend.is_keras_tensor(input_tensor):
if backend.image_data_format() == 'channels_first':
rows = backend.int_shape(input_tensor)[2]
cols = backend.int_shape(input_tensor)[3]
else:
rows = backend.int_shape(input_tensor)[1]
cols = backend.int_shape(input_tensor)[2]
if rows == cols and rows in [96, 128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
# If input_shape is None and no input_tensor
elif input_shape is None:
default_size = 224
# If input_shape is not None, assume default size
else:
if backend.image_data_format() == 'channels_first':
rows = input_shape[1]
cols = input_shape[2]
else:
rows = input_shape[0]
cols = input_shape[1]
if rows == cols and rows in [96, 128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if backend.image_data_format() == 'channels_last':
row_axis, col_axis = (0, 1)
else:
row_axis, col_axis = (1, 2)
rows = input_shape[row_axis]
cols = input_shape[col_axis]
if weights == 'imagenet':
if alpha not in [0.35, 0.50, 0.75, 1.0, 1.3, 1.4]:
raise ValueError('If imagenet weights are being loaded, '
'alpha can be one of `0.35`, `0.50`, `0.75`, '
'`1.0`, `1.3` or `1.4` only.')
if rows != cols or rows not in [96, 128, 160, 192, 224]:
rows = 224
logging.warning('`input_shape` is undefined or non-square, '
'or `rows` is not in [96, 128, 160, 192, 224].'
' Weights for input shape (224, 224) will be'
' loaded as the default.')
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
channel_axis = 1 if backend.image_data_format() == 'channels_first' else -1
first_block_filters = _make_divisible(32 * alpha, 8)
x = layers.ZeroPadding2D(padding=imagenet_utils.correct_pad(img_input, 3),
name='Conv1_pad')(img_input)
x = layers.Conv2D(first_block_filters,
kernel_size=3,
strides=(2, 2),
padding='valid',
use_bias=False,
name='Conv1')(x)
x = layers.BatchNormalization(axis=channel_axis,
epsilon=1e-3,
momentum=0.999,
name='bn_Conv1')(x)
x = layers.ReLU(6., name='Conv1_relu')(x)
x = _inverted_res_block(x,
filters=16,
alpha=alpha,
stride=1,
expansion=1,
block_id=0)
x = _inverted_res_block(x,
filters=24,
alpha=alpha,
stride=2,
expansion=6,
block_id=1)
x = _inverted_res_block(x,
filters=24,
alpha=alpha,
stride=1,
expansion=6,
block_id=2)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=2,
expansion=6,
block_id=3)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=1,
expansion=6,
block_id=4)
x = _inverted_res_block(x,
filters=32,
alpha=alpha,
stride=1,
expansion=6,
block_id=5)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=2,
expansion=6,
block_id=6)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=7)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=8)
x = _inverted_res_block(x,
filters=64,
alpha=alpha,
stride=1,
expansion=6,
block_id=9)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=10)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=11)
x = _inverted_res_block(x,
filters=96,
alpha=alpha,
stride=1,
expansion=6,
block_id=12)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=2,
expansion=6,
block_id=13)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=1,
expansion=6,
block_id=14)
x = _inverted_res_block(x,
filters=160,
alpha=alpha,
stride=1,
expansion=6,
block_id=15)
x = _inverted_res_block(x,
filters=320,
alpha=alpha,
stride=1,
expansion=6,
block_id=16)
# no alpha applied to last conv as stated in the paper:
# if the width multiplier is greater than 1 we
# increase the number of output channels
if alpha > 1.0:
last_block_filters = _make_divisible(1280 * alpha, 8)
else:
last_block_filters = 1280
x = layers.Conv2D(last_block_filters,
kernel_size=1,
use_bias=False,
name='Conv_1')(x)
x = layers.BatchNormalization(axis=channel_axis,
epsilon=1e-3,
momentum=0.999,
name='Conv_1_bn')(x)
x = layers.ReLU(6., name='out_relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D()(x)
x = layers.Dense(classes,
activation='softmax',
use_bias=True,
name='Logits')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs,
x,
name='mobilenetv2_%0.2f_%s' % (alpha, rows))
# Load weights.
if weights == 'imagenet':
if include_top:
model_name = ('mobilenet_v2_weights_tf_dim_ordering_tf_kernels_' +
str(alpha) + '_' + str(rows) + '.h5')
weight_path = BASE_WEIGHT_PATH + model_name
weights_path = data_utils.get_file(model_name,
weight_path,
cache_subdir='models')
else:
model_name = ('mobilenet_v2_weights_tf_dim_ordering_tf_kernels_' +
str(alpha) + '_' + str(rows) + '_no_top' + '.h5')
weight_path = BASE_WEIGHT_PATH + model_name
weights_path = data_utils.get_file(model_name,
weight_path,
cache_subdir='models')
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def ssd_model(feature_fn,
cls_head_fn,
rgr_head_fn,
input_shape,
num_classes,
specs: List[FeatureSpec],
max_gt_num=100,
positive_iou_threshold=0.5,
negative_iou_threshold=0.4,
negatives_per_positive=3,
min_negatives_per_image=5,
score_threshold=0.5,
iou_threshold=0.3,
max_detections_per_class=100,
max_total_detections=100,
stage='train'):
image_input = layers.Input(shape=input_shape, name='input_image')
anchors = generate_anchors(specs, input_shape[0])
feature_list = feature_fn(image_input)
num_anchors_list = [len(spec.aspect_ratios) + 2 for spec in specs]
predict_logits = cls_head_fn(feature_list, num_anchors_list, num_classes)
predict_deltas = rgr_head_fn(feature_list, num_anchors_list)
if stage == 'train':
gt_boxes = layers.Input(shape=(max_gt_num, 5),
dtype='float32',
name='input_gt_boxes')
gt_class_ids = layers.Input(shape=(max_gt_num, 2),
dtype='int32',
name='input_gt_class_ids')
# 分类和回归目标
# for i, anchors in enumerate(anchors_list):
# target = SSDTarget(anchors, positive_iou_threshold, negative_iou_threshold,
# name='target_{}'.format(i))
# deltas_list[i], labels_list[i], anchors_tag_list[i] = target([gt_boxes, gt_class_ids])
#
# # 计算loss
# cls_loss = MultiClsLoss(name='multi_cls_loss')([predict_logits_list,
# labels_list, anchors_tag_list])
# rgr_loss = MultiRegressLoss(name='multi_rgr_loss')([predict_deltas_list,
# deltas_list, anchors_tag_list])
deltas, cls_ids, anchors_tag = SSDTarget(
anchors,
positive_iou_threshold,
negative_iou_threshold,
name='ssd_target')([gt_boxes, gt_class_ids])
cls_losses = layers.Lambda(
lambda x: cls_loss_v2(
*x,
negatives_per_positive=negatives_per_positive,
min_negatives_per_image=min_negatives_per_image),
name='class_loss')([predict_logits, cls_ids, anchors_tag])
rgr_losses = layers.Lambda(lambda x: regress_loss(*x),
name='bbox_loss')(
[predict_deltas, deltas, anchors_tag])
m = Model([image_input, gt_boxes, gt_class_ids],
[cls_losses, rgr_losses])
elif stage == 'debug':
gt_boxes = layers.Input(shape=(max_gt_num, 5),
dtype='float32',
name='input_gt_boxes')
gt_class_ids = layers.Input(shape=(max_gt_num, 2),
dtype='int32',
name='input_gt_class_ids')
deltas, cls_ids, anchors_tag = SSDTarget(
anchors,
positive_iou_threshold,
negative_iou_threshold,
name='ssd_target')([gt_boxes, gt_class_ids])
m = Model([image_input, gt_boxes, gt_class_ids],
[cls_ids, anchors_tag, predict_logits])
else:
boxes, class_ids, scores = DetectBox(
anchors,
score_threshold=score_threshold,
iou_threshold=iou_threshold,
max_detections_per_class=max_detections_per_class,
max_total_detections=max_total_detections)(
[predict_deltas, predict_logits])
m = Model(image_input, [boxes, class_ids, scores])
return m
def MobileNet(input_shape=None,
alpha=1.0,
depth_multiplier=1,
dropout=1e-3,
include_top=True,
weights='imagenet',
input_tensor=None,
pooling=None,
classes=1000,
classifier_activation='softmax',
**kwargs):
"""Instantiates the MobileNet architecture.
Reference paper:
- [MobileNets: Efficient Convolutional Neural Networks for Mobile Vision
Applications](https://arxiv.org/abs/1704.04861)
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in the `tf.keras.backend.image_data_format()`.
Arguments:
input_shape: Optional shape tuple, only to be specified if `include_top`
is False (otherwise the input shape has to be `(224, 224, 3)` (with
`channels_last` data format) or (3, 224, 224) (with `channels_first`
data format). It should have exactly 3 inputs channels, and width and
height should be no smaller than 32. E.g. `(200, 200, 3)` would be one
valid value. Default to `None`.
`input_shape` will be ignored if the `input_tensor` is provided.
alpha: Controls the width of the network. This is known as the width
multiplier in the MobileNet paper. - If `alpha` < 1.0, proportionally
decreases the number of filters in each layer. - If `alpha` > 1.0,
proportionally increases the number of filters in each layer. - If
`alpha` = 1, default number of filters from the paper are used at each
layer. Default to 1.0.
depth_multiplier: Depth multiplier for depthwise convolution. This is
called the resolution multiplier in the MobileNet paper. Default to 1.0.
dropout: Dropout rate. Default to 0.001.
include_top: Boolean, whether to include the fully-connected layer at the
top of the network. Default to `True`.
weights: One of `None` (random initialization), 'imagenet' (pre-training
on ImageNet), or the path to the weights file to be loaded. Default to
`imagenet`.
input_tensor: Optional Keras tensor (i.e. output of `layers.Input()`) to
use as image input for the model. `input_tensor` is useful for sharing
inputs between multiple different networks. Default to None.
pooling: Optional pooling mode for feature extraction when `include_top`
is `False`.
- `None` (default) means that the output of the model will be
the 4D tensor output of the last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will be applied.
classes: Optional number of classes to classify images into, only to be
specified if `include_top` is True, and if no `weights` argument is
specified. Defaults to 1000.
classifier_activation: A `str` or callable. The activation function to use
on the "top" layer. Ignored unless `include_top=True`. Set
`classifier_activation=None` to return the logits of the "top" layer.
**kwargs: For backwards compatibility only.
Returns:
A `keras.Model` instance.
Raises:
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
ValueError: if `classifier_activation` is not `softmax` or `None` when
using a pretrained top layer.
"""
if 'layers' in kwargs:
global layers
layers = kwargs.pop('layers')
if kwargs:
raise ValueError('Unknown argument(s): %s' % (kwargs, ))
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top` '
'as true, `classes` should be 1000')
# Determine proper input shape and default size.
if input_shape is None:
default_size = 224
else:
if backend.image_data_format() == 'channels_first':
rows = input_shape[1]
cols = input_shape[2]
else:
rows = input_shape[0]
cols = input_shape[1]
if rows == cols and rows in [128, 160, 192, 224]:
default_size = rows
else:
default_size = 224
input_shape = imagenet_utils.obtain_input_shape(
input_shape,
default_size=default_size,
min_size=32,
data_format=backend.image_data_format(),
require_flatten=include_top,
weights=weights)
if backend.image_data_format() == 'channels_last':
row_axis, col_axis = (0, 1)
else:
row_axis, col_axis = (1, 2)
rows = input_shape[row_axis]
cols = input_shape[col_axis]
if weights == 'imagenet':
if depth_multiplier != 1:
raise ValueError('If imagenet weights are being loaded, '
'depth multiplier must be 1')
if alpha not in [0.25, 0.50, 0.75, 1.0]:
raise ValueError('If imagenet weights are being loaded, '
'alpha can be one of'
'`0.25`, `0.50`, `0.75` or `1.0` only.')
if rows != cols or rows not in [128, 160, 192, 224]:
rows = 224
logging.warning('`input_shape` is undefined or non-square, '
'or `rows` is not in [128, 160, 192, 224]. '
'Weights for input shape (224, 224) will be'
' loaded as the default.')
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
x = _conv_block(img_input, 32, alpha, strides=(2, 2))
x = _depthwise_conv_block(x, 64, alpha, depth_multiplier, block_id=1)
x = _depthwise_conv_block(x,
128,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=2)
x = _depthwise_conv_block(x, 128, alpha, depth_multiplier, block_id=3)
x = _depthwise_conv_block(x,
256,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=4)
x = _depthwise_conv_block(x, 256, alpha, depth_multiplier, block_id=5)
x = _depthwise_conv_block(x,
512,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=6)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=7)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=8)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=9)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=10)
x = _depthwise_conv_block(x, 512, alpha, depth_multiplier, block_id=11)
x = _depthwise_conv_block(x,
1024,
alpha,
depth_multiplier,
strides=(2, 2),
block_id=12)
x = _depthwise_conv_block(x, 1024, alpha, depth_multiplier, block_id=13)
if include_top:
if backend.image_data_format() == 'channels_first':
shape = (int(1024 * alpha), 1, 1)
else:
shape = (1, 1, int(1024 * alpha))
x = layers.GlobalAveragePooling2D()(x)
x = layers.Reshape(shape, name='reshape_1')(x)
x = layers.Dropout(dropout, name='dropout')(x)
x = layers.Conv2D(classes, (1, 1), padding='same',
name='conv_preds')(x)
x = layers.Reshape((classes, ), name='reshape_2')(x)
imagenet_utils.validate_activation(classifier_activation, weights)
x = layers.Activation(activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D()(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D()(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs,
x,
name='mobilenet_%0.2f_%s' % (alpha, rows))
# Load weights.
if weights == 'imagenet':
if alpha == 1.0:
alpha_text = '1_0'
elif alpha == 0.75:
alpha_text = '7_5'
elif alpha == 0.50:
alpha_text = '5_0'
else:
alpha_text = '2_5'
if include_top:
model_name = 'mobilenet_%s_%d_tf.h5' % (alpha_text, rows)
weight_path = BASE_WEIGHT_PATH + model_name
weights_path = data_utils.get_file(model_name,
weight_path,
cache_subdir='models')
else:
model_name = 'mobilenet_%s_%d_tf_no_top.h5' % (alpha_text, rows)
weight_path = BASE_WEIGHT_PATH + model_name
weights_path = data_utils.get_file(model_name,
weight_path,
cache_subdir='models')
model.load_weights(weights_path)
elif weights is not None:
model.load_weights(weights)
return model
def ResNet50Beta(batch_input_shape,
include_top=True,
weights='imagenet',
fusion_mode='bgr_hha',
classes=1000,
dilated=True,
multi_grid=False,
multi_dilation=None,
conv_trainable=True):
"""Instantiates the ResNet50 architecture.
Optionally loads weights pre-trained on ImageNet.
Note that the data format convention used by the model is
the one specified in your Keras config at `~/.keras/keras.json`.
# Arguments
include_top: whether to include the fully-connected
layer at the top of the network.
weights: one of `None` (random initialization),
'imagenet' (pre-training on ImageNet),
or the path to the weights file to be loaded.
input_tensor: optional Keras tensor (i.e. output of `layers.Input()`)
to use as image input for the model.
input_shape: optional shape tuple, only to be specified
if `include_top` is False (otherwise the input shape
has to be `(224, 224, 3)` (with `channels_last` data format)
or `(3, 224, 224)` (with `channels_first` data format).
It should have exactly 3 inputs channels,
and width and height should be no smaller than 32.
E.g. `(200, 200, 3)` would be one valid value.
pooling: Optional pooling mode for feature extraction
when `include_top` is `False`.
- `None` means that the output of the model will be
the 4D tensor output of the
last convolutional block.
- `avg` means that global average pooling
will be applied to the output of the
last convolutional block, and thus
the output of the model will be a 2D tensor.
- `max` means that global max pooling will
be applied.
classes: optional number of classes to classify images
into, only to be specified if `include_top` is True, and
if no `weights` argument is specified.
# Returns
A Keras model instance.
# Raises
ValueError: in case of invalid argument for `weights`,
or invalid input shape.
"""
if not (weights in {'imagenet', None} or os.path.exists(weights)):
raise ValueError('The `weights` argument should be either '
'`None` (random initialization), `imagenet` '
'(pre-training on ImageNet), '
'or the path to the weights file to be loaded.')
if weights == 'imagenet' and include_top and classes != 1000:
raise ValueError(
'If using `weights` as `"imagenet"` with `include_top`'
' as true, `classes` should be 1000')
img_input = klayers.Input(shape=batch_input_shape[1:4],
batch_size=batch_input_shape[0])
if fusion_mode == 'bgr_hha_gw':
x, xyz = ilayers.GWConv(sizes=(3, 3),
strides=(2, 2),
rates=(1, 1),
padding='same',
delta=0.5,
out_c=64,
activation=None,
name='gw_conv1_1')(img_input)
x = klayers.Conv2D(64, (3, 3),
strides=(3, 3),
name='custom_entry_flow_conv1_1',
trainable=conv_trainable,
use_bias=False,
padding='valid')(x)
else:
# x = klayers.ZeroPadding2D(padding=(3, 3), name='conv1_pad')(img_input)
x = klayers.Conv2D(64, (3, 3),
strides=(2, 2),
padding='same',
kernel_initializer='he_uniform',
trainable=True,
name='conv1' if batch_input_shape[-1] == 3 else
'custom_conv1')(img_input)
x = klayers.BatchNormalizationV2(name='bn_conv1')(x)
x = klayers.Activation('relu')(x)
x = klayers.ZeroPadding2D(padding=(1, 1), name='pool1_pad')(x)
x = klayers.MaxPooling2D((3, 3), strides=(2, 2))(x)
x = resnet_block(x,
num_units=3,
base_filters=64,
stage=2,
stride=1,
conv_trainable=conv_trainable)
x = resnet_block(x,
num_units=4,
base_filters=128,
stage=3,
stride=2,
conv_trainable=conv_trainable)
if dilated:
if multi_grid:
x = resnet_block(x,
num_units=6,
base_filters=256,
stage=4,
dilation=2,
stride=1,
conv_trainable=conv_trainable)
x = resnet_block(x,
num_units=3,
base_filters=512,
stage=5,
dilation=4,
stride=1,
multi_grid=multi_grid,
multi_dilation=multi_dilation,
conv_trainable=conv_trainable)
else:
x = resnet_block(x,
num_units=6,
base_filters=256,
stage=4,
dilation=2,
stride=1,
conv_trainable=conv_trainable)
x = resnet_block(x,
num_units=3,
base_filters=512,
stage=5,
dilation=4,
stride=1,
conv_trainable=conv_trainable)
else:
x = resnet_block(x,
num_units=6,
base_filters=256,
stage=4,
stride=2,
conv_trainable=conv_trainable)
x = resnet_block(x,
num_units=3,
base_filters=512,
stage=5,
stride=2,
conv_trainable=conv_trainable)
if include_top:
x = klayers.GlobalAveragePooling2D(name='avg_pool')(x)
x = klayers.Dense(classes, activation='softmax', name='fc1000')(x)
# Create model.
model = tf.keras.models.Model(img_input, x, name='resnet50_beta')
# Load weights.
if weights == 'imagenet':
if include_top:
weights_path = tf.keras.utils.get_file(
'resnet50_weights_tf_dim_ordering_tf_kernels.h5',
WEIGHTS_PATH,
cache_subdir='models',
md5_hash='a7b3fe01876f51b976af0dea6bc144eb')
else:
weights_path = tf.keras.utils.get_file(
'resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5',
WEIGHTS_PATH_NO_TOP,
cache_subdir='models',
md5_hash='a268eb855778b3df3c7506639542a6af')
model.load_weights(weights_path, by_name=True)
elif weights is not None:
model.load_weights(weights, by_name=True)
return model
def resnet50(num_classes, input_shape):
img_input = layers.Input(shape=input_shape)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channels_last
x = img_input
bn_axis = 3
# Conv1 (7x7,64,stride=2)
#x = layers.ZeroPadding2D(padding=(3, 3))(x)
x = layers.Conv2D(64, (7, 7),
strides=(2, 2),
padding='valid',
use_bias=False,
kernel_initializer='he_normal')(x)
x = layers.BatchNormalization()(x)
x = layers.Activation('relu')(x)
x = layers.ZeroPadding2D(padding=(1, 1))(x)
# 3x3 max pool,stride=2
x = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
# Conv2_x
# 1×1, 64
# 3×3, 64
# 1×1, 256
x = conv_block(x, 3, [64, 64, 256], strides=(1, 1))
########x = identity_block(x, 3, [64, 64, 256])
x = identity_block(x, 3, [64, 64, 256])
# Conv3_x
#
# 1×1, 128
# 3×3, 128
# 1×1, 512
x = conv_block(x, 3, [128, 128, 512])
######## x = identity_block(x, 3, [128, 128, 512])
########x = identity_block(x, 3, [128, 128, 512])
x = identity_block(x, 3, [128, 128, 512])
# Conv4_x
# 1×1, 256
# 3×3, 256
# 1×1, 1024
x = conv_block(x, 3, [256, 256, 1024])
x = identity_block(x, 3, [256, 256, 1024])
########x = identity_block(x, 3, [256, 256, 1024])
########x = identity_block(x, 3, [256, 256, 1024])
x = identity_block(x, 3, [256, 256, 1024])
x = identity_block(x, 3, [256, 256, 1024])
# 1×1, 512
# 3×3, 512
# 1×1, 2048
x = conv_block(x, 3, [512, 512, 2048])
########x = identity_block(x, 3, [512, 512, 2048])
x = identity_block(x, 3, [512, 512, 2048])
# average pool, 1000-d fc, softmax
x = layers.GlobalAveragePooling2D()(x)
x = layers.Dense(2, activation='softmax')(x)
# Create model.
return models.Model(img_input, x, name='resnet50')
def resnet56(classes=100, training=None):
"""Instantiates the ResNet56 architecture.
Arguments:
classes: optional number of classes to classify images into
training: Only used if training keras model with Estimator. In other
scenarios it is handled automatically.
Returns:
A Keras model instance.
"""
input_shape = (32, 32, 3)
img_input = layers.Input(shape=input_shape)
if backend.image_data_format() == 'channels_first':
x = layers.Lambda(
lambda x: backend.permute_dimensions(x, (0, 3, 1, 2)),
name='transpose')(img_input)
bn_axis = 1
else: # channel_last
x = img_input
bn_axis = 3
x = tf.keras.layers.ZeroPadding2D(padding=(1, 1), name='conv1_pad')(x)
x = tf.keras.layers.Conv2D(
16, (3, 3),
strides=(1, 1),
padding='valid',
kernel_initializer='he_normal',
kernel_regularizer=tf.keras.regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=tf.keras.regularizers.l2(L2_WEIGHT_DECAY),
name='conv1')(x)
x = tf.keras.layers.BatchNormalization(axis=bn_axis,
name='bn_conv1',
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON)(
x, training=training)
x = tf.keras.layers.Activation('relu')(x)
x = conv_building_block(x,
3, [16, 16],
stage=2,
block='a',
strides=(1, 1),
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='b',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='c',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='d',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='e',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='f',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='g',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='h',
training=training)
x = identity_building_block(x,
3, [16, 16],
stage=2,
block='i',
training=training)
x = conv_building_block(x,
3, [32, 32],
stage=3,
block='a',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='b',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='c',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='d',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='e',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='f',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='g',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='h',
training=training)
x = identity_building_block(x,
3, [32, 32],
stage=3,
block='i',
training=training)
x = conv_building_block(x,
3, [64, 64],
stage=4,
block='a',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='b',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='c',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='d',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='e',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='f',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='g',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='h',
training=training)
x = identity_building_block(x,
3, [64, 64],
stage=4,
block='i',
training=training)
x = tf.keras.layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = tf.keras.layers.Dense(
classes,
activation='softmax',
kernel_initializer='he_normal',
kernel_regularizer=tf.keras.regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=tf.keras.regularizers.l2(L2_WEIGHT_DECAY),
name='fc10')(x)
inputs = img_input
# Create model.
model = tf.keras.models.Model(inputs, x, name='resnet56')
return model
def ResNet50(num_classes):
"""Instantiates the ResNet50 architecture.
Args:
num_classes: `int` number of classes for image classification.
Returns:
A Keras model instance.
"""
# Determine proper input shape
if backend.image_data_format() == 'channels_first':
input_shape = (3, 32, 32)
bn_axis = 1
else:
input_shape = (32, 32, 3)
bn_axis = 3
img_input = layers.Input(shape=input_shape)
x = layers.ZeroPadding2D(padding=(3, 3), name='conv1_pad')(img_input)
x = layers.Conv2D(64, (3, 3), use_bias=False,
strides=(1, 1),
padding='valid',
kernel_initializer='he_normal',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='conv1')(x)
x = layers.BatchNormalization(axis=bn_axis,
momentum=BATCH_NORM_DECAY,
epsilon=BATCH_NORM_EPSILON,
name='bn_conv1')(x)
x = layers.Activation('relu')(x)
x = layers.ZeroPadding2D(padding=(1, 1), name='pool1_pad')(x)
x = layers.MaxPooling2D((3, 3), strides=(2, 2))(x)
x = conv_block(x, 3, [64, 64, 256], stage=2, block='a', strides=(1, 1))
x = identity_block(x, 3, [64, 64, 256], stage=2, block='b')
x = identity_block(x, 3, [64, 64, 256], stage=2, block='c')
x = conv_block(x, 3, [128, 128, 512], stage=3, block='a')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='b')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='c')
x = identity_block(x, 3, [128, 128, 512], stage=3, block='d')
x = conv_block(x, 3, [256, 256, 1024], stage=4, block='a')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='b')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='c')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='d')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='e')
x = identity_block(x, 3, [256, 256, 1024], stage=4, block='f')
x = conv_block(x, 3, [512, 512, 2048], stage=5, block='a')
x = identity_block(x, 3, [512, 512, 2048], stage=5, block='b')
x = identity_block(x, 3, [512, 512, 2048], stage=5, block='c')
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = layers.Dense(
num_classes, activation='softmax',
kernel_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
bias_regularizer=regularizers.l2(L2_WEIGHT_DECAY),
name='fc1000')(x)
# Create model.
return models.Model(img_input, x, name='resnet50')
code_snippet = np.expand_dims(processed['body_tokens'], -1)
label_name = np.expand_dims(processed['name_tokens'], axis=-1)
print("Vocab Size: {} number of Code snippet: {} number of labels: {}".format(vocabulary_size, len(code_snippet),
len(label_name)))
print("Label_name shape: {}\nCode_snippet shape: {}".format(label_name.shape, code_snippet.shape))
# TODO make the input a json file and parse it
hyperparameter = {'batch_size': 1, 'k1': 8, 'k2': 8, 'w1': 24, 'w2': 29, 'w3': 10, 'dropout_rate': 0.5,
'max_chunk_length': max_chunk_length, 'vocabulary_size': vocabulary_size, 'embedding_dim': 128}
# Optimised hyperparameter are reported in page 5 of the paper
batch_size = hyperparameter['batch_size']
# define layers
main_input = layers.Input(shape=(max_chunk_length, 1),
batch_size=batch_size,
dtype=tf.int32, name='main_input',
)
cnn_layer = ConvAttention(hyperparameter)
optimizer = keras.optimizers.Nadam() # RMSprop with Nesterov momentum
# loss_func = masked_sparse_cross_entropy_loss
loss_func = keras.losses.sparse_categorical_crossentropy
# define execution
cnn_output = cnn_layer(main_input)
model = keras.Model(inputs=[main_input], outputs=cnn_output)
model.compile(optimizer=optimizer,
loss=loss_func,
metrics=['accuracy'],
)
def test_advanced_model(self, strategy_fn, use_loss_scaling=False):
# The advanced model tests mixed-precision-related features that would occur
# in a resnet50 model. It tests a model that has:
# * Multiple layers, some which use auto-cast variables and some which do
# not
# * Regularization on some variables and not others.
# * A fixed loss scale (if use_loss_scaling is True)
if not self._is_strategy_supported(strategy_fn):
return
strategy = strategy_fn()
if use_loss_scaling:
loss_scale = 8.
learning_rate = 2**-14
with strategy.scope():
with policy.policy_scope(policy.Policy('infer_float32_vars')):
x = layers.Input(shape=(1,), batch_size=2, dtype=dtypes.float16)
layer1 = AddLayer(
assert_type=dtypes.float16,
regularizer=IdentityRegularizer(),
use_operator=True)
layer2 = AddLayerWithoutAutoCast(
assert_type=dtypes.float16, use_operator=True)
layer3 = AddLayer(assert_type=dtypes.float16, use_operator=False)
layer4 = AddLayerWithoutAutoCast(
assert_type=dtypes.float16,
regularizer=IdentityRegularizer(),
use_operator=False)
y = layer1(x)
y = layer2(y)
y = layer3(y)
y = layer4(y)
if use_loss_scaling:
# The gradient of 'y' at this point is 1. With loss scaling, the
# gradient is 'loss_scale'. We divide by the batch size of 2 since the
# loss is averaged across batch elements.
expected_gradient = loss_scale / 2
identity_with_grad_check_fn = (
mp_test_util.create_identity_with_grad_check_fn(
expected_dtype=dtypes.float16,
expected_gradient=[expected_gradient]))
y = core.Lambda(identity_with_grad_check_fn)(y)
y = math_ops.cast(y, dtypes.float32)
model = models.Model(inputs=x, outputs=y)
def loss_fn(y_true, y_pred):
self.assertEqual(y_true.dtype, dtypes.float32)
self.assertEqual(y_pred.dtype, dtypes.float32)
return math_ops.reduce_mean(y_pred)
opt = gradient_descent.SGD(learning_rate)
if use_loss_scaling:
opt = loss_scale_optimizer.LossScaleOptimizer(opt, loss_scale)
model.compile(
opt,
loss=loss_fn,
run_eagerly=testing_utils.should_run_eagerly(),
experimental_run_tf_function=testing_utils.should_run_tf_function())
x = np.ones((2, 1))
y = np.ones((2, 1))
dataset = dataset_ops.Dataset.from_tensor_slices((x, y)).batch(2)
model.fit(dataset)
for layer in (layer1, layer2, layer3, layer4):
if layer.losses:
# Layer has weight regularizer
self.assertEqual(backend.eval(layer.v), 1 - 2 * learning_rate)
else:
# Layer does not have weight regularizer
self.assertEqual(backend.eval(layer.v), 1 - learning_rate)
def create_network(
Zt,
Ct,
channels=1,
upscaling_filters=[512, 256, 256, 128, 64],
upscaling_ks=[5, 5, 5, 5, 5],
upscaling_strides=[2, 2, 2, 2, 2, 1],
encoder_filters=[64, 128, 256],
encoder_ks=[3, 3, 3],
resblock_filters=[256, 256, 256, 256, 256],
resblock_ks=[3, 3, 3, 3, 3, 3],
decoder_filters=[256, 256, 128, 64],
decoder_ks=[5, 5, 5, 5, 5],
):
with tf.name_scope("Gen"):
Z = kl.Input((
None,
None,
channels,
), tensor=Zt, name="Z")
C = kl.Input((
None,
None,
channels,
), tensor=Ct, name="C")
layer = Z
# Upscaling
for l in range(len(upscaling_filters)):
layer = kl.Conv2DTranspose(filters=upscaling_filters[l],
kernel_size=upscaling_ks[l],
padding="same",
strides=upscaling_strides[l],
kernel_regularizer=kr.l2(),
activation="relu")(layer)
layer = kl.BatchNormalization()(layer)
layer = kl.Conv2DTranspose(filters=channels,
kernel_size=upscaling_ks[-1],
padding="same",
strides=1,
activation="relu",
kernel_regularizer=kr.l2())(layer)
layer = kl.concatenate([layer, C])
# Encoder
skips = [C]
for l in range(len(encoder_filters)):
layer = kl.Conv2D(
filters=encoder_filters[l],
kernel_size=encoder_ks[l],
padding="same",
activation="relu",
# strides=2,
kernel_regularizer=kr.l2())(layer)
layer = kl.AveragePooling2D()(layer)
layer = kl.BatchNormalization()(layer)
skips.append(layer)
# Residual blocks
for l in range(len(resblock_filters)):
layer = ResidualBlock(resblock_filters[l],
nb_layers=3,
kernel_size=resblock_ks[l])(layer)
# Decoder
layer = kl.Conv2DTranspose(filters=decoder_filters[0],
kernel_size=decoder_ks[0],
padding="same",
strides=2,
kernel_regularizer=kr.l2(),
activation="relu")(layer)
layer = kl.BatchNormalization()(layer)
skips = skips[::-1]
for l in range(1, len(decoder_filters)):
layer = kl.concatenate([layer, skips[l]])
layer = kl.Conv2DTranspose(filters=decoder_filters[l],
kernel_size=decoder_ks[l],
padding="same",
strides=2,
kernel_regularizer=kr.l2(),
activation="relu")(layer)
layer = kl.BatchNormalization()(layer)
G_out = kl.Conv2D(filters=channels,
kernel_size=5,
activation="tanh",
padding="same",
kernel_regularizer=kr.l2())(layer)
model = k.Model(inputs=[Z, C], outputs=G_out, name="G")
return model
def DenseNet(blocks,
classes,
filters,
dropout_rate,
include_top=True,
weights=None,
input_tensor=None,
input_shape=None,
pooling=None,
classifier_activation=None):
# Determine proper input shape
if input_tensor is None:
img_input = layers.Input(shape=input_shape)
else:
if not backend.is_keras_tensor(input_tensor):
img_input = layers.Input(tensor=input_tensor, shape=input_shape)
else:
img_input = input_tensor
# x = layers.ZeroPadding2D(padding=((3, 3), (3, 3)))(x)
x = layers.Conv2D(2 * filters,
3,
strides=2,
padding='same',
name='conv1/conv')(img_input)
# x = layers.BatchNormalization(axis=bn_axis,
# epsilon=1.001e-5,
# name='conv1/bn')(x)
# x = layers.Activation('relu', name='conv1/relu')(x)
# x = layers.ZeroPadding2D(padding=((1, 1), (1, 1)))(x)
x = layers.MaxPooling2D(pool_size=(3, 3), strides=2, name='pool1')(x)
# x = dense_block(x,
# blocks=6,
# filters=filters,
# dropout_rate=dropout_rate,
# name='dense_1')
# x = transition_block(x, dropout_rate=dropout_rate, name='trans_2')
# x = dense_block(x,
# blocks=12,
# filters=filters,
# dropout_rate=dropout_rate,
# name='dense_2')
# x = transition_block(x, dropout_rate=dropout_rate, name='trans_3')
# x = dense_block(x,
# blocks=24,
# filters=filters,
# dropout_rate=dropout_rate,
# name='dense_3')
# x = transition_block(x, dropout_rate=dropout_rate, name='trans_4')
# x = dense_block(x,
# blocks=16,
# filters=filters,
# dropout_rate=dropout_rate,
# name='dense_4')
for i in range(blocks):
x = dense_block(x,
blocks=6,
filters=filters,
dropout_rate=dropout_rate,
name='dense_' + str(i))
x = transition_block(x,
dropout_rate=dropout_rate,
name='trans_' + str(i))
x = dense_block(x,
blocks=12,
filters=filters,
dropout_rate=dropout_rate,
name='dense_final')
x = layers.BatchNormalization()(x)
x = layers.Activation('relu', name='relu')(x)
if include_top:
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
x = layers.Flatten()(x)
x = layers.Dense(classes,
activation=classifier_activation,
name='predictions')(x)
else:
if pooling == 'avg':
x = layers.GlobalAveragePooling2D(name='avg_pool')(x)
elif pooling == 'max':
x = layers.GlobalMaxPooling2D(name='max_pool')(x)
# Ensure that the model takes into account
# any potential predecessors of `input_tensor`.
if input_tensor is not None:
inputs = layer_utils.get_source_inputs(input_tensor)
else:
inputs = img_input
# Create model.
model = training.Model(inputs, x, name='NET.model')
return model
padding="same")(input_tensor)
decoder = layers.concatenate([concat_tensor, decoder], axis=-1)
decoder = layers.BatchNormalization()(decoder)
decoder = layers.Activation("relu")(decoder)
decoder = layers.Conv2D(num_filters, (3, 3), padding="same")(decoder)
decoder = layers.BatchNormalization()(decoder)
decoder = layers.Activation("relu")(decoder)
decoder = layers.Conv2D(num_filters, (3, 3), padding="same")(decoder)
decoder = layers.BatchNormalization()(decoder)
decoder = layers.Activation("relu")(decoder)
return decoder
#Define Structures
inputs = layers.Input(shape=img_shape)
#256
encoder0_pool, encoder0 = encoder_block(inputs, 32)
#128
encoder1_pool, encoder1 = encoder_block(encoder0_pool, 64)
#64
encoder2_pool, encoder2 = encoder_block(encoder1_pool, 128)
#32
encoder3_pool, encoder3 = encoder_block(encoder2_pool, 256)
#16
encoder4_pool, encoder4 = encoder_block(encoder3_pool, 512)
def build(self, del_layer_dict):
prev_prune_idx = None
first_dense = True
#output processing for keeping the number of classes
if self.is_lastConv2D:
del_layer_dict[self.lastConv2D] = []
#initialize input
input = layers.Input(shape=self.input_shape)
x = input
for idx, layer in enumerate(self.model.layers):
if idx == 0:
self.layer_info[id(layer.output)].update({'out': x})
continue
# reconstruct each layer
if type(layer) is layers.convolutional.Conv2D:
prev_prune_idx = self.get_prev_pruned_idx(
layer.input, self.layer_info)
pruned_weights = self.get_conv_pruned_weights(
layer, del_layer_dict[idx], prev_prune_idx)
cutted_channl_num = len(del_layer_dict[idx])
attr_name = inspect.getargspec(
layers.convolutional.Conv2D.__init__)[0]
attr_name.pop(0)
attr_value = {attr: getattr(layer, attr) for attr in attr_name}
attr_value['filters'] -= cutted_channl_num
recon_layer = layers.convolutional.Conv2D(**attr_value)
self.layer_info[id(layer.output)].update(
{'pruned_idx': del_layer_dict[idx]})
elif type(layer) is layers.convolutional.DepthwiseConv2D:
prev_prune_idx = self.get_prev_pruned_idx(
layer.input, self.layer_info)
pruned_weights = self.get_depthwiseconv_pruned_weights(
layer, prev_prune_idx)
recon_layer = layer.__class__.from_config(layer.get_config())
elif type(layer) is layers.normalization_v2.BatchNormalization:
prev_prune_idx = self.get_prev_pruned_idx(
layer.input, self.layer_info)
pruned_weights = self.get_batchnorm_pruned_weights(
layer, prev_prune_idx)
cutted_channl_num = len(prev_prune_idx)
config = layer.get_config()
config['gamma_regularizer'] = None
recon_layer = layers.normalization_v2.BatchNormalization.from_config(
config)
elif type(layer) is layers.Dense:
if first_dense == True:
prev_prune_idx = self.get_prev_pruned_idx(
layer.input, self.layer_info)
pruned_weights = self.get_dense_pruned_weights(
layer, prev_prune_idx)
first_dense = False
recon_layer = layer.__class__.from_config(layer.get_config())
elif type(layer) is layers.Reshape:
prev_prune_idx = self.get_prev_pruned_idx(
layer.input, self.layer_info)
prev_prune_num = len(prev_prune_idx)
config = layer.get_config()
original_shape = config['target_shape']
original_shape = list(original_shape)
original_shape[-1] = original_shape[-1] - prev_prune_num
new_shape = tuple(original_shape)
config['target_shape'] = new_shape
recon_layer = layer.__class__.from_config(config)
else:
if type(layer) is layers.Add or \
type(layer) is layers.Subtract or \
type(layer) is layers.Multiply or \
type(layer) is layers.Average or \
type(layer) is layers.Maximum or \
type(layer) is layers.Minimum:
pruned_idx = self.set_pruned_idx(layer.input,
self.layer_info)
self.layer_info[id(layer.output)].update(
{'pruned_idx': pruned_idx})
if type(layer) is layers.Concatenate:
pruned_idx = self.set_pruned_idx_for_concat(
layer.input, self.layer_info)
self.layer_info[id(layer.output)].update(
{'pruned_idx': pruned_idx})
recon_layer = layer.__class__.from_config(layer.get_config())
# connect layers
if isinstance(layer.input, list):
input_list = []
for in_layer in layer.input:
input_list.append(self.layer_info[id(in_layer)]['out'])
self.layer_info[id(in_layer)]['count'] -= 1
self.del_key(in_layer, self.layer_info)
x = input_list
else:
x = self.layer_info[id(layer.input)]['out']
self.layer_info[id(layer.input)]['count'] -= 1
self.del_key(layer.input, self.layer_info)
x = recon_layer(x)
self.layer_info[id(layer.output)]['out'] = x
try:
recon_layer.set_weights(pruned_weights)
except:
pass
model = Model(inputs=input, outputs=x)
return model
def get_small_functional_mlp(num_hidden, num_classes, input_dim):
inputs = layers.Input(shape=(input_dim, ))
outputs = layers.Dense(num_hidden, activation='relu')(inputs)
activation = 'sigmoid' if num_classes == 1 else 'softmax'
outputs = layers.Dense(num_classes, activation=activation)(outputs)
return models.Model(inputs, outputs)
