File size: 17,321 Bytes
eb90a8a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 | """
Model for making Neural Network models
for classification of source code.
The module provides factory class for making
Neural Network models
"""
import sys
import tensorflow
from tensorflow.keras import models
from tensorflow.keras import layers
from tensorflow.keras import initializers
class UniqueSeed():
"""
Class for generating unique seed for
model initializing DNN models
"""
#Initial value of seed
seed = 1234
#Step at which seed is incrtemented after its each query
step = 17
@classmethod
def getSeed(cls):
"""
Get unique seed
"""
cls.seed += cls.step
return cls.seed
@classmethod
def setSeed(cls, seed, step = 17):
"""
Initialize seed externally
Parameters:
- seed -- Initial value of seed
- step -- Step at which seed is incretemented
after its each query
"""
cls.seed = seed
cls.step = step
def __init__(self):
pass
class BaseModelFactory():
"""
Base class for factories of DNN models
Supplies utility methods of other DNN factories
"""
@classmethod
def computeInitializer(cls, seed = None):
"""
Compute layer initializer
Parameters:
- seed -- seed for random generator used by analyzer
If it is None a unique repeatable seed is generated
"""
_seed = seed if seed else UniqueSeed.getSeed()
#_initializer=initializers.GlorotNormal(seed=seed)
_initializer=initializers.GlorotUniform(seed=seed)
return _initializer
@classmethod
def computeRegularizer(cls, regular):
"""
Compute l1 and l2 regulizer
Parameters:
- regular -- regularizer as a pair (l1, l2)
or None if no regularization is applied.
"""
return None if regular is None \
else tensorflow.keras.regularizers.l1_l2(l1=regular[0],
l2=regular[1])
@classmethod
def compClassification(cls, labels):
"""
Compute classification function
Parameters:
- labels -- number of labels (classes)
"""
return "sigmoid" if labels == 1 \
else "softmax"
@classmethod
def compLoss(cls, labels):
"""
Compute loss function
Parameters:
- labels -- number of labels (classes)
"""
return "binary_crossentropy" if labels == 1 \
else "sparse_categorical_crossentropy"
@classmethod
def compileDNN(cls, dnn, loss, optimizer = "rmsprop"):
"""
Compile sequential DNN
Parameters:
- dnn -- DNN to compile
- loss -- loss to use for training
- optimizer -- training optimizer
Returns compiled DNN
"""
dnn.compile(optimizer = optimizer, loss = loss,
metrics = ['accuracy'])
print("Summary of DNN constructed")
dnn.summary()
return dnn
def __init__(self):
pass
#---------------- End of class BaseModelFactory ---------------------------
class SeqBlockFactory(BaseModelFactory):
"""
Factory of sequential DNN blocks for processing sequences of tokens
for code classification and similarity analysis
Provides functions for creating sequential DNN building blocks
"""
def __init__(self, tokens):
"""
Initialize factory of DNN sequential building blocks
Parameters:
- tokens -- number of types of tokens
It is the size (dimensionality) of sequence vectors
or number of values for cathegorical input coding
"""
super(SeqBlockFactory, self).__init__()
self.n_tokens = tokens
def addConv1Layer(self, dnn, convolution, activation = None,
bias = True, regularizer = None):
"""
Add to the dnn a convolutional layer
Parameters:
- dnn -- dnn to buld
- convolution -- specification of the convolutional layer
as list of tuples:
either (n filters, kernel width, stride)
or (n filters, kernel width) if stride = 1
- activation -- activation on convolutional layer
- bias -- use or do not use bias on convolutional layer
- regularizer -- regularizer of convolution kernels
Returns exit code:
- 0 if the layer is successfully added
- 1 if the layer is specified incorrectly
"""
if len(convolution) == 3:
dnn.add(layers.Conv1D(convolution[0], convolution[1],
strides=convolution[2],
activation = activation,
use_bias = bias,
kernel_regularizer = regularizer,
kernel_initializer = BaseModelFactory.computeInitializer()))
return 0
elif len(convolution) == 2:
dnn.add(layers.Conv1D(convolution[0], convolution[1],
activation = activation,
use_bias = bias,
kernel_regularizer = regularizer,
kernel_initializer = BaseModelFactory.computeInitializer()))
return 0
else:
return 1
def addConv1(self, dnn, convolutions, activation = None,
bias = True, out_no_act = True,
regularizer = None):
"""
Add to the dnn a stack of convolutional layers
Parameters:
- dnn -- dnn to buld
- convolutions -- specification of convolutional layers
as list of tuples:
either (n filters, kernel width, stride)
or (n filters, kernel width) if stride = 1
- activation -- activation on all convolutional layers
but the last one
- bias -- use or do not use bias on convolutional layers
- out_no_act -- flag of no activation on last layer
- regularizer -- regularizer of convolution kernels
"""
for _i in range(len(convolutions)):
_act = None if _i == len(convolutions) - 1 and out_no_act \
else activation
if self.addConv1Layer(dnn, convolutions[_i], activation = _act,
bias = bias, regularizer = regularizer):
print(f"Wrong specification of convolutional layer {-i}")
_n_conv_parms = len(convolutions[_i])
print(f"It has incorrect number {_n_conv_parm} of parameters")
sys.exit(0)
def addDense(self, dnn, dense, bias = True, regularizer = None):
"""
Add a stack of dense layers to the dnn
Parameters:
- dnn -- dnn to buld
- dense -- specification of dense layers as list of
their width
- bias -- use or do not use bias on dense layers
- regularizer -- regularizer of dense weight matrix
"""
if not dense: return
for _w in dense:
dnn.add(layers.Dense(_w, activation = 'relu',
use_bias = bias,
kernel_regularizer=regularizer,
kernel_initializer = BaseModelFactory.computeInitializer()))
def addPooling(self, dnn, pool):
"""
Add pooling layer to the dnn
Parameters:
- dnn -- dnn to buld
- pool -- pooling operation; either
'max' for maximum or 'aver' for average
"""
if pool == 'max':
dnn.add(layers.GlobalMaxPooling1D())
elif pool == 'aver':
dnn.add(layers.GlobalAveragePooling1D())
else:
print(f"Error: Wrong pooling operation {pool}\n" +
"It should be either max or aver")
sys.exit(0)
def addDropout(self, dnn, rate):
"""
Add droput layer to dnn
Parameters:
- dnn -- dnn to buld
- rate -- dropout rate
If it is 0 or None no droput layer is inserted
"""
if rate:
dnn.add(layers.Dropout(rate, seed = UniqueSeed.getSeed()))
def addEmbedding(self, dnn, trainable = False, width = None,
regularizer = None):
"""
Add embedding layer to the DNN
Parameters:
- dnn -- dnn to buld
- trainable -- flag defining that embedding should be trainable
If it is False embedding layer transforms
categorical coding into one hot coding
- width -- dimension of embedding vectors
It effects trainable embedding only.
If it is None self.n_tokens is used.
- regularizer -- regularizer of convolution and dense kernels
effects on trainable embedding
"""
if trainable:
_output_dim = width if width else self.n_tokens
_initializer = \
tensorflow.keras.initializers.RandomUniform(seed = \
UniqueSeed.getSeed())
dnn.add(layers.Embedding(
input_dim = self.n_tokens + 1,
output_dim = _output_dim,
embeddings_initializer = _initializer,
mask_zero = True,
embeddings_regularizer = regularizer))
else:
#Transformation of categorical coding into one-hot
_initializer = tensorflow.keras.initializers.Identity()
dnn.add(layers.Embedding(
input_dim = self.n_tokens + 1,
output_dim = self.n_tokens,
embeddings_initializer = _initializer,
mask_zero = True))
def convTower(self, convolutions,
conv_act = None, out_no_act = True,
conv_bias = True,
input_type = "one_hot",
regularizer = None, embedding_dim = None,
regul_dense_only = True):
"""
Make sequential DNN building block of stack
of convolutional layers
Parameters:
- convolutions -- specification of convolutional layers
as list of tuples:
either (kernel width, n filters, stride)
or (kernel width, n filters) if stride = 1
- conv_act -- activation of convolutional layers
- out_no_act -- flag of no activation on last convolutional layer
- conv_bias -- use bias for convolutional layers
- input_type -- type of input layer:
- one_hot - one hot coding of input samples
- categorical - categorical coding
decoded by preprocessing layer
- trainable - categorical coding processed
by trainable embedding layer
- regularizer -- regularizer applied to weights of dense layer
and to convolutionand embedding layers if
parameters regul_dense_only = False
- embedding_dim -- size of embedding vectors
It effects trainable embedding only,
i.e. input_type == "trainable"
If it is None self.n_tokens is used.
If it is None self.n_tokens is used
- regul_dense_only -- flag to regulirize only dense layers
Returns:
- dnn -- constructed DNN building block
"""
_conv_regul = None if regul_dense_only else regularizer
_embed_regul = None if regul_dense_only else regularizer
_dnn = models.Sequential()
if input_type == "one_hot":
_dnn.add(layers.Input(shape=(None, self.n_tokens,)))
elif input_type == "categorical" or input_type == "trainable":
_trainable = (input_type == "trainable")
_dnn.add(layers.Input(shape=(None,)))
self.addEmbedding(_dnn, trainable = _trainable,
width = embedding_dim,
regularizer = _embed_regul)
else:
sys.exit(f"Wrong type {input_type} of input_type parameter of sequentinal convolution block")
self.addConv1(_dnn, convolutions, activation = conv_act,
bias = conv_bias,
out_no_act = out_no_act,
regularizer = _conv_regul)
return _dnn
def convBlock(self, convolutions, dense = [], pool = 'max',
conv_act = None, conv_bias = True,
dense_bias = True, input_type = "one_hot",
regularizer = None, embedding_dim = None,
dropout_rate = 0,
regul_dense_only = True):
"""
Make sequential convolutional DNN building block
Parameters:
- convolutions -- specification of convolutional layers
as list of tuples:
either (kernel width, n filters, stride)
or (kernel width, n filters) if stride = 1
- dense -- specification of dense layers as list of
their width except the last one
- pool -- pooling operation either:
'max' for maximum or
'aver' for average
- conv_act -- activation of convolutional layers
- conv_bias -- use bias for convolutional layers
- dense_bias -- use bias for dense layers
- input_type -- type of input layer:
- one_hot - one hot coding of input samples
- categorical - categorical coding
decoded by preprocessing layer
- trainable - categorical coding processed
by trainable embedding layer
- regularizer -- regularizer applied to weights of dense layer
and to convolutionand embedding layers if
parameters regul_dense_only = False
- embedding_dim -- size of embedding vectors
It effects trainable embedding only,
i.e. input_type == "trainable"
If it is None self.n_tokens is used.
If it is None self.n_tokens is used
- dropout_rate -- dropout rate for inserted dropout layer
If it is 0 or None no droput layer is inserted
- regul_dense_only -- flag to regulirize only dense layers
Returns:
- dnn -- constructed DNN building block
"""
_conv_regul = None if regul_dense_only else regularizer
_embed_regul = None if regul_dense_only else regularizer
_dnn = models.Sequential()
if input_type == "one_hot":
_dnn.add(layers.Input(shape=(None, self.n_tokens,)))
elif input_type == "categorical" or input_type == "trainable":
_trainable = (input_type == "trainable")
_dnn.add(layers.Input(shape=(None,)))
self.addEmbedding(_dnn, trainable = _trainable,
width = embedding_dim,
regularizer = _embed_regul)
else:
sys.exit(f"Wrong type {input_type} of input_type parameter of functional convolution block")
self.addConv1(_dnn, convolutions, activation = conv_act,
bias = conv_bias,
out_no_act = (pool == 'max'),
regularizer = _conv_regul)
self.addPooling(_dnn, pool)
self.addDense(_dnn, dense, bias = dense_bias,
regularizer = regularizer)
self.addDropout(_dnn, dropout_rate)
return _dnn
def addClassifier(self, dnn, labels, bias = True,
regularizer = None):
"""
Add to the dnn a dense layer with classifier
Classifier is either softmax for more than 2 classes
or sigmoid for 2 classes
It is expected that input shape is known
Parameters:
- dnn -- dnn to complete
- labels -- number of labels (classes)
- bias -- use or do not use bias on dense layers
- regularizer -- regularizer applied to weights of
convolution, dense, and embedding layers
"""
dnn.add(layers.Dense(labels,
activation = self.compClassification(labels),
use_bias = bias,
kernel_regularizer=regularizer,
kernel_initializer = BaseModelFactory.computeInitializer()))
#---------------- End of class SeqBlockFactory ----------------------
|