config_parser.py 124.9 KB
Newer Older
1
# Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved
Z
zhangjinchao01 已提交
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
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

from __future__ import print_function
'''
The following functions are available in the config file:

Bias: define bias. To be used as value of bias argument in Layer().

Data: define data provider.

Input: define input layer for a layer. To be used as element of inputs argument
       in Layer().

Conv: define a convolution operation for an input of a layer.

Norm: define a normalization operation for an input of a layer.

Pool: define a pooling operation for an input of a layer.

Layer: define a layer.

Parameter: define a parameter.

Import: import another config file. If the imported config file name is
        a relative path, then it will be searched under the directory of the
        current config file.

Inputs(layer_names...):
    Define the name of the input layers of the NeuralNetwork.
    The type of these layers must be "data".
    These layers will be provided with the DataBatch obtained
    from DataProvider. The data streams from DataProvider must
    have the same order.

Outputs(layer_names...):
    Define the name of the output layers of the NeuralNetwork.
    Usually the output is simply the cost layer.
    You can specify other layers as outputs and  calculate the
    cost (and its derivative) yourself.


default_initial_std(val)
default_initial_mean(val)
default_momentum(val):
default_decay_rate(val): Set the default value for these parameters


get_config_arg(name, type, default): Get the value for a config parameter.


*** customized extension to config_parser ***
The functionality of the config_parser can be extended.
If the config_arg_str for parse_config() contains
extension_module_name=[MODULE_NAME], then config_parser will call
MODULE_NAME.get_config_funcs(g_config)
MODULE_NAME.get_config_funcs() should return a dictionary of name to functions,
those functions will be available in the config file.
See trainer/tests/config_parser_test.py for example

To use this from paddle_trainer, paddle_trainer should be called with
--config_args=extension_module_name=[MODULE_NAME]

'''

import copy
import logging
import os
import sys
import traceback
import math
import shutil

try:
    from paddle.proto.DataConfig_pb2 import DataConfig
    from paddle.proto.ModelConfig_pb2 import ModelConfig
    from paddle.proto.ModelConfig_pb2 import LayerConfig
    from paddle.proto.ModelConfig_pb2 import LayerInputConfig
    from paddle.proto.ModelConfig_pb2 import ProjectionConfig
    from paddle.proto.ModelConfig_pb2 import OperatorConfig
    from paddle.proto.ModelConfig_pb2 import GeneratorConfig
    from paddle.proto.ModelConfig_pb2 import LinkConfig
    from paddle.proto.ParameterConfig_pb2 import ParameterConfig
    from paddle.proto.ParameterConfig_pb2 import ParameterUpdaterHookConfig
    from paddle.proto.TrainerConfig_pb2 import TrainerConfig

except Exception as e:
    traceback.print_exc()
    raise

logging.basicConfig(
Q
qijun 已提交
103
    format='[%(levelname)s %(asctime)s %(filename)s:%(lineno)s] %(message)s', )
Z
zhangjinchao01 已提交
104 105 106
logger = logging.getLogger('paddle')
logger.setLevel(logging.INFO)
__real_print__ = print
Q
qijun 已提交
107
print = logger.info
Z
zhangjinchao01 已提交
108 109 110 111

# from layer type name to layer class
g_layer_type_map = {}

Q
qijun 已提交
112

Z
zhangjinchao01 已提交
113 114 115
# Initialize global variables. We use this function so that we can
# call parse_config() multiple times
def init_config_environment(
Q
qijun 已提交
116 117 118 119 120 121 122 123 124 125 126 127 128 129 130
        g_default_momentum=None,
        g_default_decay_rate=None,
        g_default_initial_mean=0.,
        g_default_initial_std=0.01,
        g_default_num_batches_regularization=None,
        g_default_initial_strategy=0,
        g_default_initial_smart=False,
        g_default_gradient_clipping_threshold=None,
        g_default_device=None,
        g_default_update_hooks=None,
        g_default_compact_func=None,
        g_config=TrainerConfig(),
        g_layer_map={},
        g_parameter_map={},
        g_extended_config_funcs={},
Z
zhangjinchao01 已提交
131 132

        # store command args of paddle_trainer
Q
qijun 已提交
133
        g_command_config_args={},
Z
zhangjinchao01 已提交
134 135

        # Used for PyDataProvider to avoid duplicate module name
Q
qijun 已提交
136 137 138 139 140
        g_py_module_name_list=[],
        g_current_submodel=None,
        g_root_submodel=None,
        g_submodel_map={},
        g_submodel_stack=[],
L
Luo Tao 已提交
141 142 143
        g_add_submodel_suffix=False,

        # Whether current layer needs to pass the image height and width.
144 145 146
        # Default value is true, but if it encounters recurrent_layer_group,
        # it will be false. The reason is that image is converted to be sequence,
        # image height will be sequence length, and image width will be feature
L
Luo Tao 已提交
147 148
        # length of each timestep.
        g_pass_height_width=True, ):
Z
zhangjinchao01 已提交
149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164

    for k, v in locals().iteritems():
        globals()[k] = copy.deepcopy(v)


# Because type is widely used as a variable name in this code.
# we need a different function name for the builtin type()
def type_of(x):
    return type(x)


# Check a condition derived config file
def config_assert(b, msg):
    if not b:
        logger.fatal(msg)

Q
qijun 已提交
165

Z
zhangjinchao01 已提交
166 167
g_config_funcs = {}

Q
qijun 已提交
168

Z
zhangjinchao01 已提交
169 170 171 172 173
# decorator for indicating a function which can be used in config file
def config_func(func):
    g_config_funcs[func.func_name] = func
    return func

Q
qijun 已提交
174

Z
zhangjinchao01 已提交
175 176 177 178 179
# decorator for indicating a class which can be used in config file
def config_class(cls):
    g_config_funcs[cls.__name__] = cls
    return cls

Q
qijun 已提交
180

Z
zhangjinchao01 已提交
181 182 183 184 185 186
# decorator for indicating a class for a layer type
def config_layer(layer_type):
    def wrap(cls):
        g_config_funcs[cls.__name__] = cls
        g_layer_type_map[layer_type] = cls
        return cls
Q
qijun 已提交
187

Z
zhangjinchao01 已提交
188 189
    return wrap

Q
qijun 已提交
190

Z
zhangjinchao01 已提交
191 192 193
def gen_parameter_name(layer_name, input_index):
    return '_%s.w%d' % (layer_name, input_index)

Q
qijun 已提交
194

Z
zhangjinchao01 已提交
195 196 197
def gen_bias_parameter_name(layer_name):
    return '_%s.wbias' % layer_name

Q
qijun 已提交
198

Z
zhangjinchao01 已提交
199 200 201
def default(x, default_value):
    return default_value if x is None else x

Q
qijun 已提交
202

Z
zhangjinchao01 已提交
203 204 205 206 207 208
class Cfg(object):
    def add_keys(self, locals):
        for k, v in locals.iteritems():
            if not k.startswith('_'):
                self.__setattr__(k, v)

Q
qijun 已提交
209

Z
zhangjinchao01 已提交
210 211
# functions available in config file

Q
qijun 已提交
212

Z
zhangjinchao01 已提交
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230
# Define the name of the input layers of the NeuralNetwork.
# The type of these layers must be "data".
# These layers will be provided with the DataBatch obtained
# from DataProvider. The data streams from DataProvider must
# have the same order.
@config_func
def Inputs(*args):
    for name in args:
        name = MakeLayerNameInSubmodel(name)
        global g_current_submodel, g_root_submodel
        if g_current_submodel.is_recurrent_layer_group:
            config_assert(False, "Do not set Inputs in recurrent layer group")
        else:
            g_current_submodel.input_layer_names.append(name)

        if g_current_submodel is g_root_submodel:
            g_config.model_config.input_layer_names.append(name)

Q
qijun 已提交
231

232 233
@config_func
def HasInputsSet():
234
    return len(g_current_submodel.input_layer_names) != 0
235

Z
zhangjinchao01 已提交
236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259

# Define the name of the output layers of the NeuralNetwork.
# Usually the output is simply the cost layer.
# You can specify other layers as outputs and calculate the
# cost (and its derivative) yourself.
@config_func
def Outputs(*args):
    for name in args:
        name = MakeLayerNameInSubmodel(name)
        global g_current_submodel, g_root_submodel
        if g_current_submodel.is_recurrent_layer_group:
            config_assert(False, "Do not set Outputs in recurrent layer group")
        else:
            g_current_submodel.output_layer_names.append(name)

        if g_current_submodel is g_root_submodel:
            g_config.model_config.output_layer_names.append(name)


@config_func
def SubModelBegin(name):
    global g_current_submodel, g_root_submodel, g_submodel_stack
    g_submodel_stack.append(g_current_submodel)

Q
qijun 已提交
260
    name = MakeLayerNameInParentSubmodel(name)  #rename in nested submodel
Z
zhangjinchao01 已提交
261 262 263 264 265 266 267 268 269

    config_assert(name not in g_submodel_map,
                  'Duplicated submodel name: %s' % name)

    sub_model = g_config.model_config.sub_models.add()
    sub_model.name = name
    g_submodel_map[name] = sub_model
    g_current_submodel = sub_model

Q
qijun 已提交
270

Z
zhangjinchao01 已提交
271
@config_func
Q
qijun 已提交
272
def SubModelEnd(name=None):
Z
zhangjinchao01 已提交
273
    global g_current_submodel, g_root_submodel, g_submodel_stack
Q
qijun 已提交
274 275
    config_assert(g_current_submodel is not g_root_submodel,
                  "submodel not begin")
Z
zhangjinchao01 已提交
276
    if name is not None:
Q
qijun 已提交
277 278 279
        config_assert(
            g_current_submodel.name == MakeLayerNameInParentSubmodel(name),
            "submodel name error")
Z
zhangjinchao01 已提交
280 281 282

    g_current_submodel = g_submodel_stack.pop()

Q
qijun 已提交
283

Z
zhangjinchao01 已提交
284 285
def MakeLayerNameInParentSubmodel(name):
    suffix = ""
286 287
    if len(g_submodel_stack) > 1:
        suffix = "@" + g_submodel_stack[-1].name
Z
zhangjinchao01 已提交
288 289
    return name + suffix

Q
qijun 已提交
290

Z
zhangjinchao01 已提交
291 292 293
def GetLayerBaseName(name):
    return name.split('@')[0]

Q
qijun 已提交
294 295

def MakeLayerNameInSubmodel(name, submodel_name=None):
Z
zhangjinchao01 已提交
296 297
    global g_current_submodel
    global g_add_submodel_suffix
Q
qijun 已提交
298 299
    if (submodel_name is None and not g_add_submodel_suffix and
            not g_current_submodel.is_recurrent_layer_group):
Z
zhangjinchao01 已提交
300 301 302 303 304
        return name
    if submodel_name is None:
        submodel_name = g_current_submodel.name
    return name + "@" + submodel_name

Q
qijun 已提交
305

Z
zhangjinchao01 已提交
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
# Define a recurrent layer group begin with RecurrentLayerGroupBegin
# and end with RecurrentLayerGroupEnd.
# A recurrent layer group forward/backward one frame after previous frame
# forward/backward through all layers in layer group.
# in_links are names of layer used as input layer in the layer group.
# out_links are names of layer in layer group used as outside layer's input.
#
# If generator is set, the layer group need one or more than one outlinks.
# The first outlink should always be the generated token ids.
# If generator.num_results_per_sample is not set, the output for one sample is
# a ids sequence. Else if num_results_per_sample is more than one,
# the output for one sample is up to #num_results_per_sample generated
# sequences, which are packed in one sequence in output ids vector. Each
# generated sequence has a generation probability. The probabilities for one
# sample are stored in one row of output value matrix.
# Packed generated sequences format, for each i:
#   seq_i_length: one interger, seq_i content length,
#   [seq_i content], length = seq_i_length
#   seq_i_end_mark: one interger, for format check, always -1
# You can use "seq_text_printer" to print the output of the generator.
@config_func
def RecurrentLayerGroupWithoutOutLinksBegin(name,
                                            in_links,
329 330
                                            seq_reversed=False,
                                            target_inlinkname=""):
Z
zhangjinchao01 已提交
331 332 333 334 335 336 337
    global g_current_submodel
    config_assert(g_config.model_config.type == "recurrent_nn",
                  "RecurrentLayerGroup should be used only in recurrent_nn")
    RecurrentLayerGroup(name=name)  # add to father model
    SubModelBegin(name)
    g_current_submodel.is_recurrent_layer_group = True
    g_current_submodel.reversed = seq_reversed
338
    g_current_submodel.target_inlinkid = -1
Z
zhangjinchao01 已提交
339
    in_links_count = 0
340
    for linkid, link in enumerate(in_links):
Z
zhangjinchao01 已提交
341 342 343 344 345 346
        if isinstance(link, basestring):
            name = link
            has_subseq = False
        else:
            name = link.link_name
            has_subseq = link.has_subseq
347 348 349 350
        # assign target_inlinkid according to target_inlinkname
        if target_inlinkname == name:
            g_current_submodel.target_inlinkid = linkid

Z
zhangjinchao01 已提交
351 352 353
        if in_links_count == 0:
            in_links_has_subseq = has_subseq
        else:
Q
qijun 已提交
354 355 356 357
            config_assert(
                in_links_has_subseq == has_subseq,
                "The sequence type of in_links should be the same in RecurrentLayerGroup"
            )
Z
zhangjinchao01 已提交
358 359 360 361 362 363 364
        in_links_count += 1
        layer_name = MakeLayerNameInParentSubmodel(name)
        layer = g_layer_map[layer_name]
        if has_subseq:
            SequenceScatterAgentLayer(name=name, size=layer.size)
        else:
            ScatterAgentLayer(name=name, size=layer.size)
365

Z
zhangjinchao01 已提交
366 367 368 369 370
        pair = g_current_submodel.in_links.add()
        pair.layer_name = layer_name
        pair.link_name = MakeLayerNameInSubmodel(name)
        pair.has_subseq = has_subseq

Q
qijun 已提交
371

Z
zhangjinchao01 已提交
372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
@config_func
def RecurrentLayerGroupSetOutLink(link):
    if isinstance(link, basestring):
        name = link
        has_subseq = False
    else:
        name = link.link_name
        has_subseq = link.has_subseq
    layer_name = MakeLayerNameInParentSubmodel(name)
    pair = g_current_submodel.out_links.add()
    pair.layer_name = MakeLayerNameInSubmodel(name)
    pair.link_name = layer_name
    pair.has_subseq = has_subseq


def RecurrentLayerGroupSetGenerator(generator=None):
Q
qijun 已提交
388
    generator.eos_layer_name = MakeLayerNameInSubmodel(generator.eos_layer_name)
Z
zhangjinchao01 已提交
389 390 391 392 393 394 395 396
    g_current_submodel.generator.CopyFrom(generator)


@config_func
def RecurrentLayerGroupBegin(name,
                             in_links,
                             out_links,
                             generator=None,
397
                             target_inlinkname="",
Z
zhangjinchao01 已提交
398
                             seq_reversed=False):
Q
qijun 已提交
399
    RecurrentLayerGroupWithoutOutLinksBegin(name, in_links, seq_reversed,
400
                                            target_inlinkname)
Z
zhangjinchao01 已提交
401 402 403 404 405
    for link in out_links:
        RecurrentLayerGroupSetOutLink(link)

    if generator is not None:
        RecurrentLayerGroupSetGenerator(generator)
Q
qijun 已提交
406 407 408 409 410
        config_assert(
            len(in_links) == 0, "no in_links should be passed to generator")
        config_assert(
            len(out_links) >= 1,
            "one or more than one out_links should be passed to generator")
Z
zhangjinchao01 已提交
411 412 413 414 415 416 417


@config_func
def RecurrentLayerGroupEnd(name):
    global g_current_submodel
    config_assert(g_current_submodel.is_recurrent_layer_group,
                  "RecurrentLayerGroup not begin")
Q
qijun 已提交
418
    for pair in g_current_submodel.memories:  #check exist
Z
zhangjinchao01 已提交
419
        layer = g_layer_map[pair.layer_name]
Y
Yu Yang 已提交
420 421
        config_assert(layer is not None,
                      "memory declare wrong name:%s" % pair.layer_name)
Z
zhangjinchao01 已提交
422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439
        memory_link = g_layer_map[pair.link_name]
        config_assert(layer.size == memory_link.size,
                      "memory declare wrong size:%d" % memory_link.size)

    prev_submodel = g_current_submodel
    SubModelEnd(name)

    for pair in prev_submodel.out_links:
        layer = g_layer_map[pair.layer_name]
        # add out agent to father model
        agent_name = GetLayerBaseName(pair.link_name)
        if prev_submodel.HasField("generator"):
            DataLayer(name=agent_name, size=layer.size)
        elif pair.has_subseq:
            SequenceGatherAgentLayer(name=agent_name, size=layer.size)
        else:
            GatherAgentLayer(name=agent_name, size=layer.size)

Q
qijun 已提交
440

Z
zhangjinchao01 已提交
441 442 443 444 445 446
# Define the model type
# currently, the paddle supports "nn", "recurrent_nn", "recursive_nn" and "multi_nn"
@config_func
def model_type(name):
    g_config.model_config.type = name

Q
qijun 已提交
447

Z
zhangjinchao01 已提交
448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464
@config_class
class Bias(Cfg):
    def __init__(
            self,
            parameter_name=None,
            learning_rate=None,
            momentum=None,
            decay_rate=None,
            decay_rate_l1=None,
            initial_mean=None,
            initial_std=None,
            initial_strategy=None,
            initial_smart=None,
            num_batches_regularization=None,
            sparse_remote_update=None,
            gradient_clipping_threshold=None,
            is_static=None,
Q
qijun 已提交
465
            is_shared=None, ):
Z
zhangjinchao01 已提交
466 467
        self.add_keys(locals())

Q
qijun 已提交
468

Z
zhangjinchao01 已提交
469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488
# Define one input for a layer
@config_class
class Input(Cfg):
    def __init__(
            self,
            input_layer_name,
            parameter_name=None,
            learning_rate=None,
            momentum=None,
            decay_rate=None,
            decay_rate_l1=None,
            initial_mean=None,
            initial_std=None,
            initial_strategy=None,
            initial_smart=None,
            num_batches_regularization=None,
            sparse_remote_update=None,
            sparse_update=None,
            gradient_clipping_threshold=None,
            conv=None,
L
liaogang 已提交
489
            bilinear_interp=None,
Z
zhangjinchao01 已提交
490 491 492 493
            norm=None,
            pool=None,
            image=None,
            block_expand=None,
494
            maxout=None,
Q
qijun 已提交
495
            spp=None,
D
dangqingqing 已提交
496
            pad=None,
Z
zhangjinchao01 已提交
497 498 499 500 501
            format=None,
            nnz=None,
            is_static=None,
            is_shared=None,
            update_hooks=None,
502
            input_layer_argument=None,
D
dangqingqing 已提交
503 504 505 506 507
            make_layer_name_in_submodel=True, ):
        """
        @param make_layer_name_in_submodel True by defalut, you might need to
        set it carefully when adding Input in config_parser.py.
        """
Z
zhangjinchao01 已提交
508
        self.add_keys(locals())
D
dangqingqing 已提交
509 510 511
        self.input_layer_name = MakeLayerNameInSubmodel(
            input_layer_name
        ) if make_layer_name_in_submodel else input_layer_name
Z
zhangjinchao01 已提交
512

Q
qijun 已提交
513

Z
zhangjinchao01 已提交
514 515 516
# Define a projection for iexed layer
@config_class
class Projection(Input):
Q
qijun 已提交
517 518
    type = None  # subclass should set it correctly

Z
zhangjinchao01 已提交
519 520 521
    def __init__(
            self,
            input_layer_name,
Q
qijun 已提交
522
            size=0,  # projection output size
Z
zhangjinchao01 已提交
523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541
            parameter_name=None,
            learning_rate=None,
            momentum=None,
            decay_rate=None,
            decay_rate_l1=None,
            initial_mean=None,
            initial_std=None,
            initial_strategy=None,
            initial_smart=None,
            num_batches_regularization=None,
            sparse_remote_update=None,
            sparse_update=None,
            gradient_clipping_threshold=None,
            ptype=None,
            format=None,
            nnz=None,
            is_static=None,
            is_shared=None,
            update_hooks=None,
Q
qijun 已提交
542
            input_layer_argument=None, ):
Z
zhangjinchao01 已提交
543 544 545 546 547 548 549 550 551 552 553 554 555
        self.add_keys(locals())
        self.input_layer_name = MakeLayerNameInSubmodel(input_layer_name)

        self.proj_conf = ProjectionConfig()
        if ptype is not None:
            self.proj_conf.type = ptype
        else:
            self.proj_conf.type = self.type

    # calculate the output_size given input_size. return 0
    # to indicate using the size from Layer config
    def calc_output_size(self, input_layer_config):
        return self.size
Q
qijun 已提交
556

Z
zhangjinchao01 已提交
557 558
    def calc_parameter_size(self, input_size, output_size):
        raise NotimplementedError
Q
qijun 已提交
559

Z
zhangjinchao01 已提交
560 561 562 563 564 565 566 567 568 569
    def calc_parameter_dims(self, input_size, output_size):
        raise NotimplementedError


@config_class
class IdentityProjection(Projection):
    type = 'identity'

    def calc_output_size(self, input_layer_config):
        return input_layer_config.size
Q
qijun 已提交
570

Z
zhangjinchao01 已提交
571 572
    def calc_parameter_size(self, input_size, output_size):
        return 0
Q
qijun 已提交
573

Z
zhangjinchao01 已提交
574 575 576
    def calc_parameter_dims(self, input_size, output_size):
        return []

Q
qijun 已提交
577

Z
zhangjinchao01 已提交
578 579 580 581 582 583
# Like IdentityProjection, but layer size may smaller than input size,
# the projection select dimesions [offset, offset+layer_size) from input
@config_class
class IdentityOffsetProjection(Projection):
    type = 'identity_offset'

Q
qijun 已提交
584 585 586
    def __init__(self, input_layer_name, offset, **xargs):
        super(IdentityOffsetProjection, self).__init__(input_layer_name,
                                                       **xargs)
Z
zhangjinchao01 已提交
587 588 589 590
        self.proj_conf.offset = offset

    def calc_parameter_size(self, input_size, output_size):
        return 0
Q
qijun 已提交
591

Z
zhangjinchao01 已提交
592 593 594
    def calc_parameter_dims(self, input_size, output_size):
        return []

Q
qijun 已提交
595

Z
zhangjinchao01 已提交
596 597 598 599 600 601 602
# DotMulProjection performs element-wise multiplication with weight
@config_class
class DotMulProjection(Projection):
    type = 'dot_mul'

    def calc_output_size(self, input_layer_config):
        return input_layer_config.size
Q
qijun 已提交
603

Z
zhangjinchao01 已提交
604 605
    def calc_parameter_size(self, input_size, output_size):
        return output_size
Q
qijun 已提交
606

Z
zhangjinchao01 已提交
607 608 609
    def calc_parameter_dims(self, input_size, output_size):
        return [1, output_size]

L
Luo Tao 已提交
610

X
xuwei06 已提交
611 612 613 614 615 616 617 618 619 620 621 622 623 624
# ScalingProjection
@config_class
class ScalingProjection(Projection):
    type = 'scaling'

    def calc_output_size(self, input_layer_config):
        return input_layer_config.size

    def calc_parameter_size(self, input_size, output_size):
        return 1

    def calc_parameter_dims(self, input_size, output_size):
        return [1, 1]

Q
qijun 已提交
625

Z
zhangjinchao01 已提交
626 627 628 629 630 631
@config_class
class TableProjection(Projection):
    type = 'table'

    def calc_parameter_size(self, input_size, output_size):
        return input_size * output_size
Q
qijun 已提交
632

Z
zhangjinchao01 已提交
633 634 635
    def calc_parameter_dims(self, input_size, output_size):
        return [input_size, output_size]

Q
qijun 已提交
636

Z
zhangjinchao01 已提交
637 638 639 640 641 642
@config_class
class FullMatrixProjection(Projection):
    type = 'fc'

    def calc_parameter_size(self, input_size, output_size):
        return input_size * output_size
Q
qijun 已提交
643

Z
zhangjinchao01 已提交
644 645 646
    def calc_parameter_dims(self, input_size, output_size):
        return [input_size, output_size]

Q
qijun 已提交
647

Z
zhangjinchao01 已提交
648 649 650 651 652 653
@config_class
class TransposedFullMatrixProjection(Projection):
    type = 'trans_fc'

    def calc_parameter_size(self, input_size, output_size):
        return input_size * output_size
Q
qijun 已提交
654

Z
zhangjinchao01 已提交
655 656 657
    def calc_parameter_dims(self, input_size, output_size):
        return [output_size, input_size]

Q
qijun 已提交
658

Z
zhangjinchao01 已提交
659 660 661 662
@config_class
class ContextProjection(Projection):
    type = 'context'

Q
qijun 已提交
663 664
    def __init__(self, input_layer_name, context_start, context_length,
                 trainable_padding, **xargs):
Z
zhangjinchao01 已提交
665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687
        super(ContextProjection, self).__init__(input_layer_name, **xargs)
        self.proj_conf.context_start = context_start
        self.proj_conf.context_length = context_length
        self.proj_conf.trainable_padding = trainable_padding
        self._total_pad = max(0, -self.proj_conf.context_start) \
                          + max(0, self.proj_conf.context_start \
                                + self.proj_conf.context_length - 1)

    def calc_output_size(self, input_layer_config):
        return input_layer_config.size * self.proj_conf.context_length

    def calc_parameter_size(self, input_size, output_size):
        if self.proj_conf.trainable_padding == False:
            return 0
        else:
            return input_size * self._total_pad

    def calc_parameter_dims(self, input_size, output_size):
        return [self._total_pad, input_size]

    _total_pad = 0


688 689 690 691
@config_class
class ConvProjection(Projection):
    type = 'conv'

Q
qijun 已提交
692 693 694 695 696
    def __init__(self,
                 input_layer_name,
                 num_filters=None,
                 conv_conf=None,
                 **xargs):
697 698 699 700 701
        super(ConvProjection, self).__init__(input_layer_name, **xargs)

        if num_filters is not None:
            self.proj_conf.num_filters = num_filters

Q
qijun 已提交
702
        parse_conv(conv_conf, input_layer_name, self.proj_conf.conv_conf,
703
                   num_filters)
L
Luo Tao 已提交
704 705 706
        self.proj_conf.output_size = self.proj_conf.conv_conf.output_x * \
                                     self.proj_conf.conv_conf.output_y * \
                                     num_filters
707 708 709 710 711 712 713 714 715

    def calc_output_size(self, input_layer_config):
        return self.proj_conf.output_size

    def calc_parameter_size(self, input_size, output_size):
        co = self.proj_conf.num_filters
        ci = self.proj_conf.conv_conf.channels
        fh = self.proj_conf.conv_conf.filter_size
        fw = self.proj_conf.conv_conf.filter_size_y
716 717
        gr = self.proj_conf.conv_conf.groups
        return co * ci * fh * fw / gr
718 719 720 721 722 723 724

    def calc_bias_size(self):
        return self.proj_conf.num_filters

    def calc_parameter_dims(self, input_size, output_size):
        return None

Q
qijun 已提交
725

Z
zhangjinchao01 已提交
726 727 728
# Define a operator for mixed layer
@config_class
class Operator(Cfg):
Q
qijun 已提交
729 730
    type = None  # subclass should set it correctly

Z
zhangjinchao01 已提交
731 732
    def __init__(
            self,
Q
qijun 已提交
733
            input_layer_names, ):
Z
zhangjinchao01 已提交
734 735 736 737 738 739 740 741 742 743
        self.add_keys(locals())
        self.operator_conf = OperatorConfig()
        self.operator_conf.type = self.type

    def check_dims(self):
        pass

    def calc_output_size(self, input_sizes):
        return 0

Q
qijun 已提交
744

Z
zhangjinchao01 已提交
745 746 747
@config_class
class DotMulOperator(Operator):
    type = 'dot_mul'
Q
qijun 已提交
748 749 750

    def __init__(self, input_layer_names, scale=None, **xargs):
        super(DotMulOperator, self).__init__(input_layer_names, **xargs)
Z
zhangjinchao01 已提交
751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768
        if scale is not None:
            self.operator_conf.dotmul_scale = scale

        config_assert(len(input_layer_names) == 2, "DotMul is binary operator")

    def check_dims(self):
        for i in range(2):
            config_assert(self.operator_conf.input_sizes[i] ==
                          self.operator_conf.output_size,
                          "DotMul input_size != output_size")

    def calc_output_size(self, input_sizes):
        return input_sizes[0]


@config_class
class ConvOperator(Operator):
    type = 'conv'
Q
qijun 已提交
769 770 771 772 773 774 775

    def __init__(self,
                 input_layer_names,
                 num_filters=None,
                 conv_conf=None,
                 **xargs):
        super(ConvOperator, self).__init__(input_layer_names, **xargs)
Z
zhangjinchao01 已提交
776 777 778
        if num_filters is not None:
            self.operator_conf.num_filters = num_filters

779 780
        parse_conv(conv_conf,
                   MakeLayerNameInSubmodel(input_layer_names[0]),
Q
qijun 已提交
781
                   self.operator_conf.conv_conf, num_filters)
L
Luo Tao 已提交
782 783 784
        self.operator_conf.output_size = self.operator_conf.conv_conf.output_x * \
                                         self.operator_conf.conv_conf.output_y * \
                                         num_filters
Z
zhangjinchao01 已提交
785 786 787

        config_assert(len(input_layer_names) == 2, "Conv is binary operator")

788 789
    def calc_output_size(self, input_sizes):
        return self.operator_conf.output_size
Z
zhangjinchao01 已提交
790 791 792 793 794


# please refer to the comments in proto/ModelConfig.proto
@config_class
class Conv(Cfg):
Q
qijun 已提交
795 796 797 798 799 800 801 802 803 804 805 806 807
    def __init__(self,
                 filter_size,
                 channels,
                 padding=None,
                 stride=None,
                 groups=None,
                 filter_channels=None,
                 output_x=None,
                 img_size=None,
                 caffe_mode=True,
                 filter_size_y=None,
                 padding_y=None,
                 stride_y=None):
Z
zhangjinchao01 已提交
808 809
        self.add_keys(locals())
        if filter_size_y is None:
Q
qijun 已提交
810
            self.filter_size_y = filter_size
Z
zhangjinchao01 已提交
811
        if padding_y is None:
Q
qijun 已提交
812
            self.padding_y = padding
Z
zhangjinchao01 已提交
813
        if stride_y is None:
Q
qijun 已提交
814
            self.stride_y = stride
Z
zhangjinchao01 已提交
815
        if output_x is not None:
Q
qijun 已提交
816 817
            config_assert(output_x <= 0)

Z
zhangjinchao01 已提交
818

L
liaogang 已提交
819 820
@config_class
class BilinearInterp(Cfg):
L
Luo Tao 已提交
821
    def __init__(self, out_size_x=None, out_size_y=None, channels=None):
L
liaogang 已提交
822 823
        self.add_keys(locals())

Q
qijun 已提交
824

Z
zhangjinchao01 已提交
825 826
@config_class
class Pool(Cfg):
D
dangqingqing 已提交
827 828 829 830 831 832 833 834 835 836 837
    def __init__(
            self,
            pool_type,
            channels,
            size_x,
            size_y=None,
            start=None,
            stride=None,  # 1 by defalut in protobuf
            stride_y=None,
            padding=None,  # 0 by defalut in protobuf
            padding_y=None):
Z
zhangjinchao01 已提交
838
        self.add_keys(locals())
Q
qijun 已提交
839 840


Q
qijun 已提交
841
@config_class
Q
qijun 已提交
842
class SpatialPyramidPool(Cfg):
L
Luo Tao 已提交
843
    def __init__(self, pool_type, pyramid_height, channels):
Q
qijun 已提交
844
        self.add_keys(locals())
Z
zhangjinchao01 已提交
845

Q
qijun 已提交
846

D
dangqingqing 已提交
847 848 849 850 851 852
@config_class
class Pad(Cfg):
    def __init__(self, channels, pad_c, pad_h, pad_w):
        self.add_keys(locals())


Z
zhangjinchao01 已提交
853 854
@config_class
class Norm(Cfg):
Q
qijun 已提交
855 856 857 858 859 860 861 862 863
    def __init__(self,
                 norm_type,
                 channels,
                 size,
                 scale,
                 pow,
                 output_x=None,
                 img_size=None,
                 blocked=None):
Z
zhangjinchao01 已提交
864 865
        self.add_keys(locals())

Q
qijun 已提交
866

Z
zhangjinchao01 已提交
867 868
@config_class
class Image(Cfg):
Q
qijun 已提交
869
    def __init__(self, channels, img_size=None):
Z
zhangjinchao01 已提交
870 871
        self.add_keys(locals())

Q
qijun 已提交
872

Z
zhangjinchao01 已提交
873 874
@config_class
class BlockExpand(Cfg):
Q
qijun 已提交
875 876 877 878 879 880 881 882 883 884 885 886
    def __init__(self,
                 channels,
                 padding_x=0,
                 padding_y=0,
                 stride_x=0,
                 stride_y=0,
                 block_x=0,
                 block_y=0,
                 img_size_x=0,
                 img_size_y=0,
                 output_x=0,
                 output_y=0):
Z
zhangjinchao01 已提交
887 888
        self.add_keys(locals())

Q
qijun 已提交
889

890 891
@config_class
class MaxOut(Cfg):
Q
qijun 已提交
892
    def __init__(self, channels, groups, img_size_x=0, img_size_y=0):
893 894
        self.add_keys(locals())

Q
qijun 已提交
895

Z
zhangjinchao01 已提交
896 897 898 899 900 901 902 903 904 905 906 907 908
def DataBase(async_load_data=False,
             constant_slots=None,
             data_ratio=1,
             is_main_data=True,
             usage_ratio=None):
    # default: all sub dataproviders are treat as "main data".
    # see proto/DataConfig.proto for is_main_data
    data_config = DataConfig()

    data_config.async_load_data = async_load_data

    if constant_slots:
        data_config.constant_slots.extend(constant_slots)
Q
qijun 已提交
909 910
    data_config.data_ratio = data_ratio
    data_config.is_main_data = is_main_data
Z
zhangjinchao01 已提交
911

Q
qijun 已提交
912
    usage_ratio = default(usage_ratio, settings_deprecated["usage_ratio"])
Z
zhangjinchao01 已提交
913 914 915 916 917 918
    config_assert(usage_ratio >= 0 and usage_ratio <= 1,
                  "The range of usage_ratio is [0, 1]")
    data_config.usage_ratio = usage_ratio

    return data_config

Q
qijun 已提交
919

Z
zhangjinchao01 已提交
920
@config_func
Q
qijun 已提交
921 922 923 924 925
def SimpleData(files=None,
               feat_dim=None,
               context_len=None,
               buffer_capacity=None,
               **xargs):
Z
zhangjinchao01 已提交
926 927 928 929 930 931 932 933 934 935
    data_config = DataBase(**xargs)
    data_config.type = 'simple'
    data_config.files = files
    data_config.feat_dim = feat_dim
    if context_len is not None:
        data_config.context_len = context_len
    if buffer_capacity:
        data_config.buffer_capacity = buffer_capacity
    return data_config

Q
qijun 已提交
936

Z
zhangjinchao01 已提交
937
@config_func
Q
qijun 已提交
938 939 940 941 942 943 944 945 946 947
def PyData(files=None,
           type=None,
           file_group_queue_capacity=None,
           load_data_module=None,
           load_data_object=None,
           load_data_args="",
           load_file_count=None,
           constant_slots=None,
           load_thread_num=None,
           **xargs):
Z
zhangjinchao01 已提交
948 949 950
    data_config = DataBase(**xargs)
    data_config.type = 'py'
    if load_data_module in g_py_module_name_list:
Q
qijun 已提交
951

Z
zhangjinchao01 已提交
952 953 954
        def get_path(module):
            m = __import__(load_data_module)
            return os.path.split(os.path.realpath(m.__file__))[0]
Q
qijun 已提交
955

Z
zhangjinchao01 已提交
956 957 958
        # python C-api is not thread safe, one module can only be import once,
        # so here we nedd to copy the module with different names if it has to be
        # imported several times.
Q
qijun 已提交
959 960
        module_new_name = "%s_copy_%d" % (load_data_module,
                                          len(g_py_module_name_list))
Z
zhangjinchao01 已提交
961
        g_py_module_name_list.append(module_new_name)
Q
qijun 已提交
962 963 964 965
        module_path = "%s/%s.py" % (get_path(load_data_module),
                                    load_data_module)
        new_module_path = "%s/%s.py" % (get_path(load_data_module),
                                        module_new_name)
Z
zhangjinchao01 已提交
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989
        if os.path.isfile(module_path) == False:
            raise Exception("File %s is not exist." % module_path)
        shutil.copy2(module_path, new_module_path)
        load_data_module = module_new_name
    else:
        g_py_module_name_list.append(load_data_module)
    if load_data_module is not None and load_data_object is not None:
        data_config.load_data_module = load_data_module
        data_config.load_data_object = load_data_object
    else:
        raise ValueError('load_data_module, load_data_object is not defined.')
    data_config.load_data_args = load_data_args

    data_config.files = files or ''
    if file_group_queue_capacity is not None:
        data_config.file_group_conf.queue_capacity = file_group_queue_capacity
    if load_file_count is not None:
        data_config.file_group_conf.load_file_count = load_file_count
    if load_thread_num is not None:
        data_config.file_group_conf.load_thread_num = load_thread_num
    if constant_slots:
        data_config.constant_slots.extend(constant_slots)
    return data_config

Q
qijun 已提交
990

Z
zhangjinchao01 已提交
991
@config_func
Q
qijun 已提交
992 993 994 995 996 997 998
def ProtoData(files=None,
              type=None,
              file_group_queue_capacity=None,
              load_file_count=None,
              constant_slots=None,
              load_thread_num=None,
              **xargs):
Z
zhangjinchao01 已提交
999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018
    data_config = DataBase(**xargs)
    if type is None:
        data_config.type = 'proto'
    else:
        data_config.type = type
    data_config.files = files

    # When type="proto_group", one data provider contains at most
    # load_file_count files, and there are at most
    # (queue_capacity + load_thread_num + 1) data providers in memory
    if file_group_queue_capacity is not None:
        data_config.file_group_conf.queue_capacity = file_group_queue_capacity
    if load_file_count is not None:
        data_config.file_group_conf.load_file_count = load_file_count
    if load_thread_num is not None:
        data_config.file_group_conf.load_thread_num = load_thread_num
    if constant_slots:
        data_config.constant_slots.extend(constant_slots)
    return data_config

Q
qijun 已提交
1019

Z
zhangjinchao01 已提交
1020 1021
#real data for training is actually provided by "sub_data" data providers.
@config_func
Q
qijun 已提交
1022
def MultiData(sub_data=[]):
Z
zhangjinchao01 已提交
1023 1024 1025 1026 1027
    data_config = DataConfig()
    data_config.type = 'multi'
    data_config.sub_data_configs.extend(sub_data)
    return data_config

Q
qijun 已提交
1028

Z
zhangjinchao01 已提交
1029
@config_func
Q
qijun 已提交
1030 1031 1032 1033 1034 1035 1036
def Data(type,
         files=None,
         feat_dim=None,
         slot_dims=None,
         context_len=None,
         buffer_capacity=None,
         **xargs):
Z
zhangjinchao01 已提交
1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071

    data_config = DataBase(**xargs)
    data_config.type = type
    data_config.files = files
    data_config.feat_dim = feat_dim
    data_config.slot_dims.extend(slot_dims)
    if context_len is not None:
        data_config.context_len = context_len
    data_config.buffer_capacity = buffer_capacity
    return data_config


@config_func
def TrainData(data_config, async_load_data=None):
    config_assert(not g_config.HasField('data_config'),
                  'Only one TrainData definition is allowed')
    g_config.data_config.CopyFrom(data_config)
    g_config.data_config.for_test = False
    if async_load_data is not None:
        logger.warning("Deprecated: async_load_data should be used inside"
                       " Data definition")
        g_config.data_config.async_load_data = async_load_data


@config_func
def TestData(data_config, async_load_data=None):
    config_assert(not g_config.HasField('test_data_config'),
                  'Only one TestData definition is allowed')
    g_config.test_data_config.CopyFrom(data_config)
    g_config.test_data_config.for_test = True
    if async_load_data is not None:
        logger.warning("Deprecated: async_load_data should be used inside"
                       " Data definition")
        g_config.test_data_config.async_load_data = async_load_data

Q
qijun 已提交
1072

L
Luo Tao 已提交
1073 1074
#caffe_mode: compute the output size using floor instead of ceil,
#            which is consistent of caffe and CuDNN's convention.
1075 1076 1077 1078 1079 1080 1081
def cnn_output_size(img_size, filter_size, padding, stride, caffe_mode):
    output = (2 * padding + img_size - filter_size) / float(stride)
    if caffe_mode:
        return 1 + int(math.floor(output))
    else:
        return 1 + int(math.ceil(output))

Q
qijun 已提交
1082

1083
#calcualte image_size based on output_size for de-convolution (ConvTransLayer).
L
Luo Tao 已提交
1084
#It is the reverse function of cnn_output_size
1085
def cnn_image_size(output_size, filter_size, padding, stride, caffe_mode):
L
Luo Tao 已提交
1086 1087 1088
    img_size = (output_size - 1) * stride + filter_size - 2 * padding
    if not caffe_mode:
        img_size = img_size + 1
1089 1090
    return img_size

Q
qijun 已提交
1091

L
Luo Tao 已提交
1092
def get_img_size(input_layer_name, channels):
L
Luo Tao 已提交
1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110
    input = g_layer_map[input_layer_name]
    img_pixels = input.size / channels
    img_size = input.width if input.width > 0 else int(img_pixels**0.5)
    img_size_y = input.height if input.height > 0 else int(img_pixels /
                                                           img_size)
    config_assert(
        img_size * img_size_y == img_pixels,
        "Input layer %s: Incorrect input image size %d * %d for input image pixels %d"
        % (input_layer_name, img_size, img_size_y, img_pixels))
    return img_size, img_size_y


def parse_bilinear(bilinear, input_layer_name, bilinear_conf):
    parse_image(bilinear, input_layer_name, bilinear_conf.image_conf)
    bilinear_conf.out_size_x = bilinear.out_size_x
    bilinear_conf.out_size_y = bilinear.out_size_y


1111
def parse_pool(pool, input_layer_name, pool_conf, ceil_mode):
Z
zhangjinchao01 已提交
1112
    pool_conf.pool_type = pool.pool_type
Q
qijun 已提交
1113 1114 1115
    config_assert(pool.pool_type in [
        'max-projection', 'avg-projection', 'cudnn-max-pool', 'cudnn-avg-pool'
    ], "pool-type %s is not in "
Z
zhangjinchao01 已提交
1116
                  "['max-projection', 'avg-projection', "
Q
qijun 已提交
1117
                  "'cudnn-max-pool', 'cudnn-avg-pool']" % pool.pool_type)
Z
zhangjinchao01 已提交
1118 1119 1120 1121 1122 1123

    pool_conf.channels = pool.channels
    pool_conf.size_x = pool.size_x
    pool_conf.stride = pool.stride

    pool_conf.size_y = default(pool.size_y, pool_conf.size_x)
Q
qijun 已提交
1124
    pool_conf.stride_y = default(pool.stride_y, pool_conf.stride)
Z
zhangjinchao01 已提交
1125

L
Luo Tao 已提交
1126
    pool_conf.img_size, pool_conf.img_size_y = \
L
Luo Tao 已提交
1127
        get_img_size(input_layer_name, pool.channels)
Z
zhangjinchao01 已提交
1128

1129
    config_assert(not pool.start, "start is deprecated in pooling.")
Z
zhangjinchao01 已提交
1130

1131
    if pool.padding is not None:
Z
zhangjinchao01 已提交
1132
        pool_conf.padding = pool.padding
1133
    pool_conf.padding_y = default(pool.padding_y, pool_conf.padding)
D
dangqingqing 已提交
1134 1135
    pool_conf.output_x = cnn_output_size(pool_conf.img_size, pool_conf.size_x,
                                         pool_conf.padding, pool_conf.stride,
1136
                                         not ceil_mode)
D
dangqingqing 已提交
1137 1138
    pool_conf.output_y = cnn_output_size(pool_conf.img_size_y, pool_conf.size_y,
                                         pool_conf.padding_y,
1139
                                         pool_conf.stride_y, not ceil_mode)
Q
qijun 已提交
1140

Z
zhangjinchao01 已提交
1141

Q
qijun 已提交
1142
def parse_spp(spp, input_layer_name, spp_conf):
L
Luo Tao 已提交
1143
    parse_image(spp, input_layer_name, spp_conf.image_conf)
Q
qijun 已提交
1144 1145
    spp_conf.pool_type = spp.pool_type
    config_assert(spp.pool_type in ['max-projection', 'avg-projection'],
Q
qijun 已提交
1146 1147
                  "pool-type %s is not in "
                  "['max-projection', 'avg-projection']" % spp.pool_type)
Q
qijun 已提交
1148
    spp_conf.pyramid_height = spp.pyramid_height
Q
qijun 已提交
1149

Q
qijun 已提交
1150

Z
zhangjinchao01 已提交
1151 1152
def parse_image(image, input_layer_name, image_conf):
    image_conf.channels = image.channels
L
Luo Tao 已提交
1153
    image_conf.img_size, image_conf.img_size_y = \
L
Luo Tao 已提交
1154
        get_img_size(input_layer_name, image_conf.channels)
Q
qijun 已提交
1155

Z
zhangjinchao01 已提交
1156 1157 1158 1159

def parse_norm(norm, input_layer_name, norm_conf):
    norm_conf.norm_type = norm.norm_type
    config_assert(norm.norm_type in ['rnorm', 'cmrnorm-projection'],
Q
qijun 已提交
1160 1161
                  "norm-type %s is not in [rnorm, 'cmrnorm-projection']" %
                  norm.norm_type)
Z
zhangjinchao01 已提交
1162 1163 1164 1165 1166 1167
    norm_conf.channels = norm.channels
    norm_conf.size = norm.size
    norm_conf.scale = norm.scale
    norm_conf.pow = norm.pow
    norm_conf.blocked = norm.blocked

L
Luo Tao 已提交
1168
    norm_conf.img_size, norm_conf.img_size_y = \
L
Luo Tao 已提交
1169
        get_img_size(input_layer_name, norm.channels)
Z
zhangjinchao01 已提交
1170
    norm_conf.output_x = norm_conf.img_size
L
Luo Tao 已提交
1171
    norm_conf.output_y = norm_conf.img_size_y
Z
zhangjinchao01 已提交
1172 1173 1174
    if norm.norm_type in ['cmrnorm-projection']:
        norm_conf.scale /= norm.size
    else:
Q
qijun 已提交
1175 1176
        norm_conf.scale /= norm.size**2

1177

L
Luo Tao 已提交
1178 1179
#caffe_mode: compute the output size using floor instead of ceil,
#            which is consistent of caffe and CuDNN's convention.
1180
def parse_conv(conv, input_layer_name, conv_conf, num_filters, trans=False):
Z
zhangjinchao01 已提交
1181 1182 1183 1184 1185 1186 1187 1188 1189
    conv_conf.filter_size = conv.filter_size
    conv_conf.filter_size_y = conv.filter_size_y
    conv_conf.channels = conv.channels
    conv_conf.padding = conv.padding
    conv_conf.padding_y = conv.padding_y
    conv_conf.stride = conv.stride
    conv_conf.stride_y = conv.stride_y
    conv_conf.groups = conv.groups
    conv_conf.caffe_mode = conv.caffe_mode
Q
qijun 已提交
1190

1191
    if not trans:
1192
        conv_conf.filter_channels = conv.channels / conv.groups
L
Luo Tao 已提交
1193
        conv_conf.img_size, conv_conf.img_size_y = \
L
Luo Tao 已提交
1194
            get_img_size(input_layer_name, conv.channels)
1195
        conv_conf.output_x = cnn_output_size(
Q
qijun 已提交
1196 1197
            conv_conf.img_size, conv_conf.filter_size, conv_conf.padding,
            conv_conf.stride, conv_conf.caffe_mode)
L
Luo Tao 已提交
1198 1199 1200
        conv_conf.output_y = cnn_output_size(
            conv_conf.img_size_y, conv_conf.filter_size_y, conv_conf.padding_y,
            conv_conf.stride_y, conv_conf.caffe_mode)
1201
    else:
1202
        conv_conf.filter_channels = num_filters / conv.groups
L
Luo Tao 已提交
1203
        conv_conf.output_x, conv_conf.output_y = \
L
Luo Tao 已提交
1204
            get_img_size(input_layer_name, conv.channels)
1205
        conv_conf.img_size = cnn_image_size(
Q
qijun 已提交
1206 1207
            conv_conf.output_x, conv_conf.filter_size, conv_conf.padding,
            conv_conf.stride, conv_conf.caffe_mode)
L
Luo Tao 已提交
1208
        conv_conf.img_size_y = cnn_image_size(
L
Luo Tao 已提交
1209 1210
            conv_conf.output_y, conv_conf.filter_size_y, conv_conf.padding_y,
            conv_conf.stride_y, conv_conf.caffe_mode)
Q
qijun 已提交
1211

1212

Z
zhangjinchao01 已提交
1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225
def parse_block_expand(block_expand, input_layer_name, block_expand_conf):
    block_expand_conf.channels = block_expand.channels
    block_expand_conf.stride_x = block_expand.stride_x
    block_expand_conf.stride_y = block_expand.stride_y
    block_expand_conf.padding_x = block_expand.padding_x
    block_expand_conf.padding_y = block_expand.padding_y
    block_expand_conf.block_x = block_expand.block_x
    block_expand_conf.block_y = block_expand.block_y
    block_expand_conf.img_size_x = block_expand.img_size_x
    block_expand_conf.img_size_y = block_expand.img_size_y
    if block_expand_conf.img_size_x == 0:
        block_expand_conf.output_x = 0
    else:
1226
        block_expand_conf.output_x = cnn_output_size(
1227
            block_expand.img_size_x, block_expand.block_x,
1228
            block_expand.padding_x, block_expand.stride_x, False)
Z
zhangjinchao01 已提交
1229 1230

    if block_expand_conf.img_size_y == 0:
1231
        block_expand_conf.output_y = 0
Z
zhangjinchao01 已提交
1232
    else:
1233
        block_expand_conf.output_y = cnn_output_size(
1234
            block_expand.img_size_y, block_expand.block_y,
1235
            block_expand.padding_y, block_expand.stride_y, False)
Z
zhangjinchao01 已提交
1236

Q
qijun 已提交
1237

1238
def parse_maxout(maxout, input_layer_name, maxout_conf):
L
Luo Tao 已提交
1239
    parse_image(maxout, input_layer_name, maxout_conf.image_conf)
1240
    maxout_conf.groups = maxout.groups
1241

Q
qijun 已提交
1242

Z
zhangjinchao01 已提交
1243 1244 1245 1246 1247 1248
# Define an evaluator
@config_func
def Evaluator(
        name,
        type,
        inputs,
Q
qijun 已提交
1249 1250 1251 1252 1253 1254 1255
        chunk_scheme=None,
        num_chunk_types=None,
        classification_threshold=None,
        positive_label=None,
        dict_file=None,
        result_file=None,
        num_results=None,
1256 1257
        delimited=None,
        excluded_chunk_types=None, ):
Z
zhangjinchao01 已提交
1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271
    evaluator = g_config.model_config.evaluators.add()
    evaluator.type = type
    evaluator.name = MakeLayerNameInSubmodel(name)
    if type_of(inputs) == str:
        inputs = [inputs]

    evaluator.input_layers.extend(
        [MakeLayerNameInSubmodel(name) for name in inputs])

    if chunk_scheme is not None:
        evaluator.chunk_scheme = chunk_scheme
        evaluator.num_chunk_types = num_chunk_types
    g_current_submodel.evaluator_names.append(evaluator.name)

1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284
    if classification_threshold is not None:
        evaluator.classification_threshold = classification_threshold
    if positive_label is not None:
        evaluator.positive_label = positive_label
    if dict_file is not None:
        evaluator.dict_file = dict_file

    if result_file is not None:
        evaluator.result_file = result_file
    if num_results is not None:
        evaluator.num_results = num_results
    if delimited is not None:
        evaluator.delimited = delimited
Z
zhangjinchao01 已提交
1285

1286 1287 1288
    if excluded_chunk_types:
        evaluator.excluded_chunk_types.extend(excluded_chunk_types)

Q
qijun 已提交
1289

Z
zhangjinchao01 已提交
1290 1291 1292 1293 1294
class LayerBase(object):
    def __init__(
            self,
            name,
            type,
Q
qijun 已提交
1295
            size,  # size can be 0. In this case, subclass should set it.
Z
zhangjinchao01 已提交
1296 1297 1298 1299
            inputs,
            device=None,
            active_type="",
            drop_rate=0.,
1300
            coeff=None):
Z
zhangjinchao01 已提交
1301
        config_assert('@' not in name,
Q
qijun 已提交
1302
                      "layer name: %s contain special character @" % name)
Z
zhangjinchao01 已提交
1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317
        global g_current_submodel
        name = MakeLayerNameInSubmodel(name)

        config_assert(name not in g_layer_map,
                      'Duplicated layer name: %s' % name)

        self.inputs = copy.deepcopy(inputs)
        self.operators = []

        if self.inputs is None:
            self.inputs = []
        elif type_of(self.inputs) != list:
            self.inputs = [self.inputs]

        self.config = g_config.model_config.layers.add()
1318
        assert isinstance(self.config, LayerConfig)
Z
zhangjinchao01 已提交
1319 1320 1321
        self.config.name = name
        self.config.type = type
        self.config.active_type = active_type
1322 1323
        if coeff is not None:
            self.config.coeff = float(coeff)
Z
zhangjinchao01 已提交
1324 1325 1326 1327 1328 1329 1330
        if size != 0:
            self.config.size = size
        if drop_rate != 0:
            self.config.drop_rate = drop_rate

        if device is not None:
            self.config.device = device
1331
        elif g_default_device is not None:
Z
zhangjinchao01 已提交
1332 1333 1334 1335 1336 1337 1338 1339 1340
            self.config.device = g_default_device

        for input_index in xrange(len(self.inputs)):
            input = self.inputs[input_index]
            input_config = None
            input_layer_name = ''
            if type_of(input) == str:
                input_layer_name = input
                input_config = Input(
Q
qijun 已提交
1341 1342
                    input_layer_name=input,
                    parameter_name=gen_parameter_name(name, input_index))
Z
zhangjinchao01 已提交
1343 1344 1345 1346 1347 1348 1349 1350
                input_layer_name = input_config.input_layer_name
            elif isinstance(input, Input):
                input_layer_name = input.input_layer_name
                input_config = input
                if input_config.parameter_name is None:
                    input_config.parameter_name = \
                        gen_parameter_name(name, input_index)
            elif isinstance(input, Operator):
Q
qijun 已提交
1351
                self.operators.append(input)
Z
zhangjinchao01 已提交
1352 1353 1354 1355
                input.operator_conf.input_indices.append(input_index)
                input_config = Input(input.input_layer_names[0])
                input_layer_name = input_config.input_layer_name
            else:
Q
qijun 已提交
1356
                raise ValueError('Wrong type for inputs: %s' % type_of(input))
Z
zhangjinchao01 已提交
1357
            config_assert(input_layer_name in g_layer_map,
Q
qijun 已提交
1358 1359
                          "Unknown input layer '%s' for layer %s" %
                          (input_layer_name, name))
Z
zhangjinchao01 已提交
1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370
            self.inputs[input_index] = input_config
            layer_input = self.config.inputs.add()
            layer_input.input_layer_name = input_config.input_layer_name
            if input_config.input_layer_argument is not None:
                layer_input.input_layer_argument = \
                    input_config.input_layer_argument

        g_layer_map[name] = self.config

        g_current_submodel.layer_names.append(self.config.name)

L
Luo Tao 已提交
1371 1372 1373 1374 1375 1376
        if self.config.type != 'data' and g_pass_height_width:
            height = self.get_input_layer(0).height
            width = self.get_input_layer(0).width
            if height and width:
                self.set_layer_height_width(height, width)

Z
zhangjinchao01 已提交
1377 1378 1379 1380 1381 1382
    def get_input_layer(self, input_index):
        return g_layer_map[self.config.inputs[input_index].input_layer_name]

    # will return the bias created if not *for_self*
    def create_bias_parameter(
            self,
Q
qijun 已提交
1383
            bias,  # True/False or BiasCfg
Z
zhangjinchao01 已提交
1384
            size,
Q
qijun 已提交
1385 1386 1387
            dims=None,
            for_self=True,  # whether create bias for layer self
    ):
Z
zhangjinchao01 已提交
1388 1389 1390 1391 1392 1393

        if size == 0:
            return
        if dims is None:
            dims = [1, size]

Q
qijun 已提交
1394 1395 1396
        config_assert(
            type_of(bias) == bool or type_of(bias) == Bias,
            'Incorrect type for bias: %s' % type_of(bias))
Z
zhangjinchao01 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405

        if type_of(bias) == bool:
            if bias:
                bias = Bias()

        if type_of(bias) == Bias:
            if bias.parameter_name is None:
                bias.parameter_name = gen_bias_parameter_name(self.config.name)
            if bias.parameter_name not in g_parameter_map:
1406 1407
                assert isinstance(self.config, LayerConfig)

Z
zhangjinchao01 已提交
1408 1409 1410
                Parameter(
                    bias.parameter_name,
                    size,
Q
qijun 已提交
1411 1412
                    self.config.device
                    if self.config.HasField('device') else None,
Z
zhangjinchao01 已提交
1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423
                    dims,
                    bias.learning_rate,
                    bias.momentum,
                    decay_rate=bias.decay_rate,
                    decay_rate_l1=bias.decay_rate_l1,
                    initial_mean=bias.initial_mean,
                    initial_std=bias.initial_std,
                    initial_strategy=bias.initial_strategy,
                    initial_smart=bias.initial_smart,
                    num_batches_regularization=bias.num_batches_regularization,
                    sparse_remote_update=bias.sparse_remote_update,
Q
qijun 已提交
1424 1425
                    gradient_clipping_threshold=bias.
                    gradient_clipping_threshold,
Z
zhangjinchao01 已提交
1426
                    is_static=bias.is_static,
Q
qijun 已提交
1427
                    is_shared=bias.is_shared, )
Z
zhangjinchao01 已提交
1428 1429 1430 1431 1432
            if for_self:
                self.config.bias_parameter_name = bias.parameter_name
            else:
                return bias.parameter_name

Q
qijun 已提交
1433 1434 1435 1436 1437 1438
    def create_input_parameter(self,
                               input_index,
                               size,
                               dims=None,
                               sparse=None,
                               format=None):
Z
zhangjinchao01 已提交
1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452
        if dims is None:
            # TODO(yuyang18): print warning and callstack here!
            dims = list()

        if size == 0:
            return

        input_config = self.inputs[input_index]

        self.config.inputs[input_index].input_parameter_name = \
            input_config.parameter_name

        if input_config.parameter_name in g_parameter_map:
            para = g_parameter_map[input_config.parameter_name]
Q
qijun 已提交
1453 1454
            config_assert(size == para.size, (
                'Shared parameter "%s" does not ' + 'have same size: %s vs. %s')
Z
zhangjinchao01 已提交
1455 1456
                          % (input_config.parameter_name, para.size, size))

Q
qijun 已提交
1457 1458
            config_assert(dims == para.dims, (
                'Shared parameter "%s" does not ' + 'have same dims: %s vs. %s')
Z
zhangjinchao01 已提交
1459 1460 1461 1462 1463 1464
                          % (input_config.parameter_name, para.dims, dims))
            return

        Parameter(
            input_config.parameter_name,
            size,
1465
            self.config.device if self.config.HasField("device") else None,
Z
zhangjinchao01 已提交
1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477
            dims,
            input_config.learning_rate,
            input_config.momentum,
            decay_rate=input_config.decay_rate,
            decay_rate_l1=input_config.decay_rate_l1,
            initial_mean=input_config.initial_mean,
            initial_std=input_config.initial_std,
            initial_strategy=input_config.initial_strategy,
            initial_smart=input_config.initial_smart,
            num_batches_regularization=input_config.num_batches_regularization,
            sparse_remote_update=input_config.sparse_remote_update,
            sparse_update=input_config.sparse_update,
Q
qijun 已提交
1478 1479
            gradient_clipping_threshold=input_config.
            gradient_clipping_threshold,
Z
zhangjinchao01 已提交
1480 1481 1482 1483
            sparse=sparse,
            format=format,
            is_static=input_config.is_static,
            is_shared=input_config.is_shared,
Q
qijun 已提交
1484
            update_hooks=input_config.update_hooks)
Z
zhangjinchao01 已提交
1485 1486 1487 1488 1489 1490 1491 1492 1493

    def set_layer_size(self, size):
        if self.config.size == 0:
            self.config.size = size
        else:
            config_assert(self.config.size == size,
                          'Different inputs result in' +
                          'different layer size at layer %s' % self.config.name)

L
Luo Tao 已提交
1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510
    def set_layer_height_width(self, height, width):
        self.config.height = height
        self.config.width = width

    def set_cnn_layer(self,
                      input_layer_name,
                      height,
                      width,
                      channels,
                      is_print=True):
        size = height * width * channels
        self.set_layer_size(size)
        self.set_layer_height_width(height, width)
        if is_print:
            print("output for %s: c = %d, h = %d, w = %d, size = %d" %
                  (input_layer_name, channels, height, width, size))

Q
qijun 已提交
1511

Z
zhangjinchao01 已提交
1512 1513
@config_layer('multi_class_cross_entropy_with_selfnorm')
class MultiClassCrossEntropySelfNormCostLayer(LayerBase):
Q
qijun 已提交
1514 1515 1516
    def __init__(self, name, inputs, softmax_selfnorm_alpha=0.1, **xargs):
        super(MultiClassCrossEntropySelfNormCostLayer, self).__init__(
            name, 'multi_class_cross_entropy_with_selfnorm', 0, inputs, **xargs)
Z
zhangjinchao01 已提交
1517 1518
        self.config.softmax_selfnorm_alpha = softmax_selfnorm_alpha

Q
qijun 已提交
1519

Z
zhangjinchao01 已提交
1520 1521
@config_layer('fc')
class FCLayer(LayerBase):
Q
qijun 已提交
1522
    def __init__(self, name, size, inputs, bias=True, **xargs):
Z
zhangjinchao01 已提交
1523 1524 1525 1526 1527 1528 1529 1530 1531 1532
        super(FCLayer, self).__init__(name, 'fc', size, inputs=inputs, **xargs)
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            psize = self.config.size * input_layer.size
            dims = [input_layer.size, self.config.size]
            format = self.inputs[input_index].format
            sparse = format == "csr" or format == "csc"

            if sparse:
                psize = self.inputs[input_index].nnz
1533 1534
            else:
                sparse = None
Z
zhangjinchao01 已提交
1535

Q
qijun 已提交
1536 1537
            self.create_input_parameter(input_index, psize, dims, sparse,
                                        format)
Z
zhangjinchao01 已提交
1538 1539
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
1540

Z
zhangjinchao01 已提交
1541 1542
@config_layer('selective_fc')
class SelectiveFCLayer(LayerBase):
Q
qijun 已提交
1543 1544 1545 1546 1547 1548 1549 1550 1551 1552
    def __init__(self,
                 name,
                 size,
                 inputs,
                 bias=True,
                 selective_fc_pass_generation=False,
                 has_selected_colums=True,
                 selective_fc_full_mul_ratio=0.02,
                 selective_fc_parallel_plain_mul_thread_num=None,
                 **xargs):
Z
zhangjinchao01 已提交
1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572
        super(SelectiveFCLayer, self).__init__(
            name, 'selective_fc', size, inputs=inputs, **xargs)
        # user MUST know if selctive fc is used in training,
        # parameter matrices saved by this layer are automatically transposed,
        # BUT bias is not.

        # if selective_fc is used only in testing mode, and parameters for
        # this layer are trained by fully connected layers,
        # then TranposedFullMatrixProjectin MUST be used in training
        # to avoid manual transpose in testing.

        self.config.selective_fc_pass_generation = selective_fc_pass_generation
        self.config.has_selected_colums = has_selected_colums
        self.config.selective_fc_full_mul_ratio = selective_fc_full_mul_ratio
        if selective_fc_parallel_plain_mul_thread_num is not None:
            self.config.selective_fc_parallel_plain_mul_thread_num = selective_fc_parallel_plain_mul_thread_num

        input_num = len(self.inputs)
        if has_selected_colums:
            config_assert(input_num >= 2,
Q
qijun 已提交
1573 1574
                          ("if indices of selected columns are not specified, "
                           "selective_fc Layer has at least two inputs"))
Z
zhangjinchao01 已提交
1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586
            input_num -= 1

        for input_index in xrange(input_num):
            input_layer = self.get_input_layer(input_index)
            psize = self.config.size * input_layer.size
            dims = [input_layer.size, self.config.size]
            dims = dims[::-1]  # transpose the parameter
            format = self.inputs[input_index].format
            sparse = format == "csr" or format == "csc"
            if sparse:
                psize = self.inputs[input_index].nnz

Q
qijun 已提交
1587 1588
            self.create_input_parameter(input_index, psize, dims, sparse,
                                        format)
Z
zhangjinchao01 已提交
1589 1590
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
1591

1592 1593
@config_layer('print')
class PrintLayer(LayerBase):
Q
qijun 已提交
1594
    def __init__(self, name, inputs):
1595 1596
        super(PrintLayer, self).__init__(name, 'print', 0, inputs)

Q
qijun 已提交
1597

Y
yuan 已提交
1598 1599
@config_layer('priorbox')
class PriorBoxLayer(LayerBase):
G
gaoyuan 已提交
1600 1601
    def __init__(self, name, inputs, size, min_size, max_size, aspect_ratio,
                 variance):
Y
yuan 已提交
1602
        super(PriorBoxLayer, self).__init__(name, 'priorbox', 0, inputs)
G
gaoyuan 已提交
1603
        config_assert(len(inputs) == 2, 'PriorBoxLayer must have 2 inputs')
G
gaoyuan 已提交
1604 1605 1606 1607 1608 1609 1610
        input_layer = self.get_input_layer(1)
        config_assert(
            input_layer.type == 'data',
            'Expecting the second input layer of an priorbox layer to be '
            'a data layer')
        config_assert(input_layer.width > 0, 'The data layer must set width')
        config_assert(input_layer.height > 0, 'The data layer must set height')
G
gaoyuan 已提交
1611
        config_assert(len(variance) == 4, 'The variance must have 4 inputs')
Y
yuan 已提交
1612 1613 1614 1615 1616 1617
        self.config.inputs[0].priorbox_conf.min_size.extend(min_size)
        self.config.inputs[0].priorbox_conf.max_size.extend(max_size)
        self.config.inputs[0].priorbox_conf.aspect_ratio.extend(aspect_ratio)
        self.config.inputs[0].priorbox_conf.variance.extend(variance)
        self.config.size = size

Q
qijun 已提交
1618

Z
zhangjinchao01 已提交
1619 1620
@config_layer('data')
class DataLayer(LayerBase):
L
Luo Tao 已提交
1621
    def __init__(self, name, size, height=None, width=None, device=None):
Q
qijun 已提交
1622 1623
        super(DataLayer, self).__init__(
            name, 'data', size, inputs=[], device=device)
L
Luo Tao 已提交
1624 1625
        if height and width:
            self.set_layer_height_width(height, width)
Q
qijun 已提交
1626

Z
zhangjinchao01 已提交
1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653

'''
DataNormLayer: A layer for data normalization
Input: One and only one input layer is accepted. The input layer must
       be DataLayer with dense data type
Output: The normalization of the input data

Reference:
    LA Shalabi, Z Shaaban, B Kasasbeh. Data mining: A preprocessing engine

Example:
    Layer(
        name = "norm_input_layer",
        type = "data_norm",
        inputs = [Input("input_layer",
                        parameter_name = "_slot0.stats")],
        data_norm_strategy = "z-score",
    )

Note:
  (1) The parameter has been calculated in the preprocessing stage,
      and should be initialized by --init_model_path when training.
  (2) Three data normalization methoeds are considered
          z-score: y = (x-mean)/std
          min-max: y = (x-min)/(max-min)
          decimal-scaling: y = x/10^j, where j is the smallest integer such that max(|y|)<1
'''
Q
qijun 已提交
1654 1655


Z
zhangjinchao01 已提交
1656 1657
@config_layer('data_norm')
class DataNormLayer(LayerBase):
Q
qijun 已提交
1658
    def __init__(self, name, inputs, data_norm_strategy="z-score", device=None):
Z
zhangjinchao01 已提交
1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669
        super(DataNormLayer, self).__init__(
            name, 'data_norm', 0, inputs=inputs, device=device)
        self.config.data_norm_strategy = data_norm_strategy
        config_assert(len(inputs) == 1, 'DataNormLayer must have 1 input')
        input_layer = self.get_input_layer(0)
        self.set_layer_size(input_layer.size)
        para_size = 5 * input_layer.size
        para_dims = [5, input_layer.size]
        self.inputs[0].is_static = True
        self.create_input_parameter(0, para_size, para_dims)

Q
qijun 已提交
1670

Z
zhangjinchao01 已提交
1671 1672 1673
@config_layer('prelu')
class ParameterReluLayer(LayerBase):
    layer_type = 'prelu'
Q
qijun 已提交
1674 1675

    def __init__(self, name, inputs, partial_sum=1, **args):
Z
zhangjinchao01 已提交
1676 1677 1678 1679 1680 1681 1682 1683
        super(ParameterReluLayer, self).__init__(
            name, self.layer_type, 0, inputs=inputs, **args)
        config_assert(len(self.inputs) == 1)
        config_assert(self.input_layer.size % partial_sum == 0)
        input_layer = self.get_input_layer(0)
        self.set_layer_size(input_layer.size)
        self.create_input_parameter(0, input_layer.size / partial_sum)

Q
qijun 已提交
1684

Z
zhangjinchao01 已提交
1685 1686 1687
@config_layer('conv')
class ConvLayerBase(LayerBase):
    layer_type = 'conv'
Q
qijun 已提交
1688 1689 1690 1691 1692 1693 1694 1695

    def __init__(self,
                 name,
                 inputs=[],
                 bias=True,
                 num_filters=None,
                 shared_biases=False,
                 **xargs):
Z
zhangjinchao01 已提交
1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711
        super(ConvLayerBase, self).__init__(
            name, self.layer_type, 0, inputs=inputs, **xargs)

        if num_filters is not None:
            self.config.num_filters = num_filters

        use_gpu = int(g_command_config_args.get("use_gpu", 0))
        parallel_nn = int(g_command_config_args.get("parallel_nn", 0))

        # Automatically select cudnn_type for GPU and exconv for CPU
        # if set type=conv, but still reserve the way user specify
        # exconv or cudnn_conv manually.
        if self.layer_type == "cudnn_conv":
            config_assert(use_gpu, "cudnn_conv only support GPU")

        if (use_gpu == 1 and self.layer_type != "exconv" and
Q
qijun 已提交
1712
            (parallel_nn == 0 or self.config.device > -1)):
Z
zhangjinchao01 已提交
1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724
            self.layer_type = "cudnn_conv"
        else:
            self.layer_type = "exconv"
        # need to specify layer in config
        self.config.type = self.layer_type

        if shared_biases is not None:
            self.config.shared_biases = shared_biases

        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            conv_conf = self.config.inputs[input_index].conv_conf
L
Luo Tao 已提交
1725 1726
            parse_conv(self.inputs[input_index].conv, input_layer.name,
                       conv_conf, num_filters)
Z
zhangjinchao01 已提交
1727 1728
            psize = self.calc_parameter_size(conv_conf)
            self.create_input_parameter(input_index, psize)
L
Luo Tao 已提交
1729 1730
            self.set_cnn_layer(name, conv_conf.output_y, conv_conf.output_x,
                               self.config.num_filters)
Z
zhangjinchao01 已提交
1731 1732 1733 1734 1735 1736 1737 1738 1739 1740

        psize = self.config.size
        if shared_biases:
            psize = self.config.num_filters
        self.create_bias_parameter(bias, psize, [psize, 1])

    def calc_parameter_size(self, conv_conf):
        return self.config.num_filters * conv_conf.filter_channels \
                    * (conv_conf.filter_size * conv_conf.filter_size_y)

Q
qijun 已提交
1741

Z
zhangjinchao01 已提交
1742 1743 1744 1745
@config_layer('exconv')
class ConvLayer(ConvLayerBase):
    layer_type = 'exconv'

Q
qijun 已提交
1746

Z
zhangjinchao01 已提交
1747 1748 1749 1750
@config_layer('cudnn_conv')
class ConvLayer(ConvLayerBase):
    layer_type = 'cudnn_conv'

1751 1752 1753 1754

@config_layer('convt')
class ConvTransLayerBase(LayerBase):
    layer_type = 'convt'
Q
qijun 已提交
1755 1756 1757 1758 1759 1760 1761 1762

    def __init__(self,
                 name,
                 inputs=[],
                 bias=True,
                 num_filters=None,
                 shared_biases=False,
                 **xargs):
1763
        super(ConvTransLayerBase, self).__init__(
1764 1765 1766 1767 1768 1769 1770 1771
            name, self.layer_type, 0, inputs=inputs, **xargs)

        if num_filters is not None:
            self.config.num_filters = num_filters

        use_gpu = int(g_command_config_args.get("use_gpu", 0))
        parallel_nn = int(g_command_config_args.get("parallel_nn", 0))

1772 1773
        # cudnn_convt has not been implemented so use exconvt only
        self.layer_type = "exconvt"
1774 1775 1776 1777 1778 1779 1780 1781
        # need to specify layer in config
        self.config.type = self.layer_type

        if shared_biases is not None:
            self.config.shared_biases = shared_biases

        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
1782
            parse_conv(
1783 1784
                self.inputs[input_index].conv,
                input_layer.name,
1785
                self.config.inputs[input_index].conv_conf,
1786
                num_filters,
1787
                trans=True)
1788 1789 1790 1791 1792
            conv_conf = self.config.inputs[input_index].conv_conf
            psize = self.calc_parameter_size(conv_conf)
            print("output size for %s is %d " % (name, conv_conf.output_x))
            self.create_input_parameter(input_index, psize)
            self.set_layer_size(
Q
qijun 已提交
1793
                (conv_conf.img_size**2) * self.config.num_filters)
1794 1795 1796 1797 1798 1799 1800

        psize = self.config.size
        if shared_biases:
            psize = self.config.num_filters
        self.create_bias_parameter(bias, psize, [psize, 1])

    def calc_parameter_size(self, conv_conf):
1801
        return conv_conf.channels * conv_conf.filter_channels \
1802 1803
                    * (conv_conf.filter_size * conv_conf.filter_size_y)

Q
qijun 已提交
1804

1805 1806 1807 1808
@config_layer('exconvt')
class ConvTransLayer(ConvTransLayerBase):
    layer_type = 'exconvt'

Q
qijun 已提交
1809

Z
zhangjinchao01 已提交
1810 1811
@config_layer('norm')
class NormLayer(LayerBase):
1812 1813
    def __init__(self, name, inputs, **xargs):
        super(NormLayer, self).__init__(name, 'norm', 0, inputs=inputs, **xargs)
Z
zhangjinchao01 已提交
1814 1815 1816
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            norm_conf = self.config.inputs[input_index].norm_conf
L
Luo Tao 已提交
1817 1818 1819 1820
            parse_norm(self.inputs[input_index].norm, input_layer.name,
                       norm_conf)
            self.set_cnn_layer(name, norm_conf.output_y, norm_conf.output_x,
                               norm_conf.channels, False)
Q
qijun 已提交
1821

Z
zhangjinchao01 已提交
1822 1823 1824

@config_layer('pool')
class PoolLayer(LayerBase):
1825 1826
    def __init__(self, name, inputs, ceil_mode=True, **xargs):
        super(PoolLayer, self).__init__(name, 'pool', 0, inputs=inputs, **xargs)
Z
zhangjinchao01 已提交
1827 1828 1829
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            pool_conf = self.config.inputs[input_index].pool_conf
L
Luo Tao 已提交
1830
            parse_pool(self.inputs[input_index].pool, input_layer.name,
1831
                       pool_conf, ceil_mode)
L
Luo Tao 已提交
1832 1833
            self.set_cnn_layer(name, pool_conf.output_y, pool_conf.output_x,
                               pool_conf.channels)
Q
qijun 已提交
1834

Z
zhangjinchao01 已提交
1835

Q
qijun 已提交
1836 1837
@config_layer('spp')
class SpatialPyramidPoolLayer(LayerBase):
1838
    def __init__(self, name, inputs, **xargs):
Q
qijun 已提交
1839
        super(SpatialPyramidPoolLayer, self).__init__(
1840
            name, 'spp', 0, inputs=inputs, **xargs)
Q
qijun 已提交
1841 1842 1843
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            spp_conf = self.config.inputs[input_index].spp_conf
L
Luo Tao 已提交
1844 1845 1846
            parse_spp(self.inputs[input_index].spp, input_layer.name, spp_conf)
            output_x = (pow(4, spp_conf.pyramid_height) - 1) / (4 - 1)
            self.set_cnn_layer(name, 1, output_x, spp_conf.image_conf.channels)
Q
qijun 已提交
1847

Q
qijun 已提交
1848

D
dangqingqing 已提交
1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867
@config_layer('pad')
class PadLayer(LayerBase):
    def __init__(self, name, inputs, **xargs):
        super(PadLayer, self).__init__(name, 'pad', 0, inputs=inputs, **xargs)
        pad = self.inputs[0].pad
        self.config.inputs[0].pad_conf.pad_c.extend(pad.pad_c)
        self.config.inputs[0].pad_conf.pad_h.extend(pad.pad_h)
        self.config.inputs[0].pad_conf.pad_w.extend(pad.pad_w)

        input_layer = self.get_input_layer(0)
        image_conf = self.config.inputs[0].pad_conf.image_conf
        parse_image(pad, input_layer.name, image_conf)
        out_ch = pad.channels + pad.pad_c[0] + pad.pad_c[1]
        out_h = image_conf.img_size_y + pad.pad_h[0] + pad.pad_h[1]
        out_w = image_conf.img_size + pad.pad_w[0] + pad.pad_w[1]
        self.set_cnn_layer(name, out_h, out_w, out_ch)
        self.config.size = out_ch * out_h * out_w


Z
zhangjinchao01 已提交
1868 1869 1870
@config_layer('batch_norm')
class BatchNormLayer(LayerBase):
    layer_type = 'batch_norm'
Q
qijun 已提交
1871 1872 1873 1874 1875 1876 1877 1878 1879 1880

    def __init__(self,
                 name,
                 inputs,
                 active_type="linear",
                 bias=True,
                 use_global_stats=True,
                 moving_average_fraction=0.9,
                 batch_norm_type=None,
                 **xargs):
Z
zhangjinchao01 已提交
1881 1882 1883 1884
        if inputs is None:
            inputs = []
        elif not isinstance(inputs, list):
            inputs = [inputs]
Q
qijun 已提交
1885 1886
        config_assert(
            len(inputs) == 1, "BatchNormLayer must have one and only one input")
Z
zhangjinchao01 已提交
1887 1888 1889 1890 1891 1892 1893 1894
        # Create Input for moving mean and std,
        # in batch normalization layer.
        # These paras no need to update, so set is_static is true.
        # If not use is_static, even set learning_rate = 0, decay_rate = 0,
        # these paras will change if set average_window in configure.
        use_gpu = bool(int(g_command_config_args.get("use_gpu", 0)))
        is_shared = True if not use_gpu else False
        for i in xrange(2):
Q
qijun 已提交
1895 1896 1897 1898 1899 1900
            inputs.append(
                Input(
                    inputs[0].input_layer_name,
                    initial_std=0.0,
                    initial_mean=0.0,
                    is_static=True,
1901
                    is_shared=is_shared,
D
dangqingqing 已提交
1902
                    make_layer_name_in_submodel=False, ))
Z
zhangjinchao01 已提交
1903 1904 1905 1906 1907 1908 1909

        parallel_nn = bool(int(g_command_config_args.get("parallel_nn", 0)))
        cudnn_version = int(g_command_config_args.get("cudnn_version", 0))
        # Automatically select cudnn_batch_norm for GPU and batch_norm for CPU.
        # Also based on cudnn version.
        use_cudnn = use_gpu and batch_norm_type != "batch_norm" and \
            ((not parallel_nn) or self.config.device > -1) and \
1910
            cudnn_version >= 4007
Z
zhangjinchao01 已提交
1911
        self.layer_type = "cudnn_batch_norm" if use_cudnn else "batch_norm"
Q
qijun 已提交
1912 1913 1914 1915 1916 1917 1918
        super(BatchNormLayer, self).__init__(
            name,
            self.layer_type,
            0,
            active_type=active_type,
            inputs=inputs,
            **xargs)
Z
zhangjinchao01 已提交
1919 1920 1921 1922 1923 1924

        if use_global_stats is not None:
            self.config.use_global_stats = use_global_stats
        if moving_average_fraction is not None:
            self.config.moving_average_fraction = moving_average_fraction

Q
qijun 已提交
1925
        input_layer = self.get_input_layer(0)
Z
zhangjinchao01 已提交
1926
        image_conf = self.config.inputs[0].image_conf
L
Luo Tao 已提交
1927
        parse_image(self.inputs[0].image, input_layer.name, image_conf)
1928

1929 1930
        # Only pass the width and height of input to batch_norm layer
        # when either of it is non-zero.
1931 1932
        if input_layer.width != 0 or input_layer.height != 0:
            self.set_cnn_layer(name, image_conf.img_size_y, image_conf.img_size,
D
dangqingqing 已提交
1933
                               image_conf.channels, False)
1934 1935
        else:
            self.set_layer_size(input_layer.size)
Z
zhangjinchao01 已提交
1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947

        psize = self.calc_parameter_size(image_conf)
        dims = [1, psize]
        self.create_input_parameter(0, psize)
        self.create_input_parameter(1, psize, dims)
        self.create_input_parameter(2, psize, dims)

        self.create_bias_parameter(bias, psize)

    def calc_parameter_size(self, image_conf):
        return image_conf.channels

Q
qijun 已提交
1948

Z
zhangjinchao01 已提交
1949 1950
@config_layer('trans')
class TransLayer(LayerBase):
1951
    def __init__(self, name, inputs, **xargs):
Q
qijun 已提交
1952
        super(TransLayer, self).__init__(
1953
            name, 'trans', 0, inputs=inputs, **xargs)
Q
qijun 已提交
1954 1955 1956
        config_assert(
            len(self.inputs) == 1,
            'TransLayer must have one and only one input')
Z
zhangjinchao01 已提交
1957 1958
        self.set_layer_size(self.get_input_layer(0).size)

Q
qijun 已提交
1959

Z
zhangjinchao01 已提交
1960 1961
@config_layer('resize')
class ResizeLayer(LayerBase):
1962
    def __init__(self, name, size, inputs, **xargs):
Q
qijun 已提交
1963
        super(ResizeLayer, self).__init__(
1964
            name, 'resize', size=size, inputs=inputs, **xargs)
Q
qijun 已提交
1965 1966 1967 1968
        config_assert(
            len(self.inputs) == 1,
            'ResizeLayer must have one and only one input')

Z
zhangjinchao01 已提交
1969

1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981
@config_layer('rotate')
class RotateLayer(LayerBase):
    def __init__(self, name, inputs, height, device=None):
        super(RotateLayer, self).__init__(
            name, 'rotate', 0, inputs=inputs, device=device)
        config_assert(
            len(self.inputs) == 1,
            'RotateLayer must have one and only one input')
        self.config.height = height
        self.set_layer_size(self.get_input_layer(0).size)


Z
zhangjinchao01 已提交
1982 1983
@config_layer('blockexpand')
class BlockExpandLayer(LayerBase):
1984
    def __init__(self, name, inputs, **xargs):
Q
qijun 已提交
1985
        super(BlockExpandLayer, self).__init__(
1986
            name, 'blockexpand', 0, inputs=inputs, **xargs)
Z
zhangjinchao01 已提交
1987 1988
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
Q
qijun 已提交
1989 1990
            parse_block_expand(
                self.inputs[input_index].block_expand, input_layer.name,
Z
zhangjinchao01 已提交
1991
                self.config.inputs[input_index].block_expand_conf)
Q
qijun 已提交
1992 1993 1994 1995 1996 1997
            block_expand_conf = self.config.inputs[
                input_index].block_expand_conf
            self.set_layer_size(block_expand_conf.block_x *
                                block_expand_conf.block_y *
                                block_expand_conf.channels)

Z
zhangjinchao01 已提交
1998

1999 2000
@config_layer('maxout')
class MaxOutLayer(LayerBase):
Q
qijun 已提交
2001 2002 2003
    def __init__(self, name, inputs, **xargs):
        super(MaxOutLayer, self).__init__(
            name, 'maxout', 0, inputs=inputs, **xargs)
2004 2005
        input_layer = self.get_input_layer(0)
        maxout_conf = self.config.inputs[0].maxout_conf
L
Luo Tao 已提交
2006
        parse_maxout(self.inputs[0].maxout, input_layer.name, maxout_conf)
L
Luo Tao 已提交
2007 2008 2009
        out_channels = maxout_conf.image_conf.channels / maxout_conf.groups
        self.set_cnn_layer(name, g_layer_map[input_layer.name].height,
                           g_layer_map[input_layer.name].width, out_channels)
Q
qijun 已提交
2010

2011

Z
zhangjinchao01 已提交
2012 2013 2014 2015
# key: cost type
# value: cost class
g_cost_map = {}

Q
qijun 已提交
2016

Z
zhangjinchao01 已提交
2017 2018 2019
# define a cost layer without any parameters
def define_cost(class_name, cost_type):
    def init(cls, name, inputs, device=None, coeff=1.):
Q
qijun 已提交
2020 2021
        super(type(cls), cls).__init__(
            name, cost_type, 1, inputs, device=device, coeff=coeff)
Z
zhangjinchao01 已提交
2022

Q
qijun 已提交
2023
    cls = type(class_name, (LayerBase, ), dict(__init__=init))
Z
zhangjinchao01 已提交
2024 2025 2026
    global g_cost_map
    g_cost_map[cost_type] = cls

Q
qijun 已提交
2027

Z
zhangjinchao01 已提交
2028 2029 2030 2031 2032 2033 2034 2035
define_cost('MultiClassCrossEntropy', 'multi-class-cross-entropy')
define_cost('RankingCost', 'rank-cost')
define_cost('AucValidation', 'auc-validation')
define_cost('PnpairValidation', 'pnpair-validation')
define_cost('SumOfSquaresCostLayer', 'square_error')
define_cost('MultiBinaryLabelCrossEntropy', 'multi_binary_label_cross_entropy')
define_cost('SoftBinaryClassCrossEntropy', 'soft_binary_class_cross_entropy')
define_cost('HuberTwoClass', 'huber')
X
xuwei06 已提交
2036
define_cost('SumCost', 'sum_cost')
Z
zhangjinchao01 已提交
2037

Q
qijun 已提交
2038

Z
zhangjinchao01 已提交
2039 2040
@config_layer('hsigmoid')
class HierarchicalSigmoidLayer(LayerBase):
Q
qijun 已提交
2041
    def __init__(self, name, num_classes, inputs, device=None, bias=True):
Z
zhangjinchao01 已提交
2042 2043
        super(HierarchicalSigmoidLayer, self).__init__(
            name, 'hsigmoid', 1, inputs=inputs, device=device)
Q
qijun 已提交
2044 2045 2046
        config_assert(
            len(self.inputs) >= 2,
            'HierarchicalSigmoidLayer must have at least 2 inputs')
Z
zhangjinchao01 已提交
2047 2048 2049 2050 2051 2052 2053 2054
        self.config.num_classes = num_classes
        for input_index in xrange(len(self.inputs) - 1):
            input_layer = self.get_input_layer(input_index)
            psize = (num_classes - 1) * input_layer.size
            dims = [num_classes - 1, input_layer.size]
            self.create_input_parameter(input_index, psize, dims)
        self.create_bias_parameter(bias, num_classes - 1)

Q
qijun 已提交
2055

Z
zhangjinchao01 已提交
2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079
'''
lambdaCost for lambdaRank LTR approach

Usage:
  Example: Layer(name = "cost", type = "lambda_cost", NDCG_num = 8,
             max_sort_size = -1, inputs = ["output", "score"])

  Input data: Samples of the same query should be loaded as a sequence,
          by ProtoDataProvider or PyDataProvider etc.. User should provide
          scores for each sample. The score slot should be the 2nd
          input of lambdaRank layer.

  NDCG_num = the size of NDCG, e.g., 5 for NDCG@5.
    Note: NDCG_num must be less than or equal to the minimum
          size of lists.

  max_sort_size = the size of partial sorting in calculating gradient.
    Note: If max_sort_size = -1, then for each list, the algorithm will
          sort the entire list to get gradient.
          In other cases, max_sort_size must be greater than or equal
          to NDCG_num.
          max_sort_size can be greater than the size of a list, in which
          case the algorithm will sort the entire list to get gradient.
'''
Q
qijun 已提交
2080 2081


Z
zhangjinchao01 已提交
2082 2083
@config_layer('lambda_cost')
class LambdaCost(LayerBase):
Q
qijun 已提交
2084
    def __init__(self, name, inputs, NDCG_num=5, max_sort_size=-1, device=None):
Z
zhangjinchao01 已提交
2085 2086
        super(LambdaCost, self).__init__(
            name, 'lambda_cost', 1, inputs=inputs, device=device)
Q
qijun 已提交
2087
        config_assert(len(self.inputs) == 2, 'lambdaCost must have 2 inputs')
Z
zhangjinchao01 已提交
2088 2089
        self.config.NDCG_num = NDCG_num
        if max_sort_size != -1:
Q
qijun 已提交
2090 2091 2092
            config_assert(
                NDCG_num <= max_sort_size,
                'NDCG_num must be less than or equal to max_sort_size')
Z
zhangjinchao01 已提交
2093 2094
        self.config.max_sort_size = max_sort_size

Q
qijun 已提交
2095

Z
zhangjinchao01 已提交
2096 2097
@config_layer('nce')
class NCELayer(LayerBase):
Q
qijun 已提交
2098 2099 2100 2101 2102 2103 2104 2105
    def __init__(self,
                 name,
                 num_classes,
                 inputs,
                 num_neg_samples=10,
                 neg_sampling_dist=None,
                 bias=True,
                 **xargs):
Z
zhangjinchao01 已提交
2106
        super(NCELayer, self).__init__(name, 'nce', 1, inputs=inputs, **xargs)
Q
qijun 已提交
2107 2108
        config_assert(
            len(self.inputs) >= 2, 'NCELayer must have at least 2 inputs')
Z
zhangjinchao01 已提交
2109 2110
        self.config.num_classes = num_classes
        if neg_sampling_dist is not None:
Q
qijun 已提交
2111 2112 2113 2114
            config_assert(
                len(neg_sampling_dist) == num_classes,
                'len(neg_sampling_dist)(%s) is not same as num_classes (%s)' %
                (len(neg_sampling_dist), num_classes))
Z
zhangjinchao01 已提交
2115
            s = sum(neg_sampling_dist)
Q
qijun 已提交
2116 2117 2118
            config_assert(
                abs(s - 1) < 1e-5,
                'The sum of neg_sampling_dist (%s) is not 1' % s)
Z
zhangjinchao01 已提交
2119 2120 2121 2122 2123

            self.config.neg_sampling_dist.extend(neg_sampling_dist)

        self.config.num_neg_samples = num_neg_samples
        num_real_inputs = len(self.inputs) - 1
Q
qijun 已提交
2124
        input_layer = self.get_input_layer(num_real_inputs)
Z
zhangjinchao01 已提交
2125 2126 2127 2128
        config_assert(input_layer.type == 'data',
                      'Expecting the last input layer of an nce layer to be '
                      'a data layer')

Q
qijun 已提交
2129 2130
        if (num_real_inputs > 1 and input_layer.size == 1 and
                self.get_input_layer(num_real_inputs - 1).type == 'data'):
Z
zhangjinchao01 已提交
2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143
            # This input layer is assumed to be a sample weight layer
            num_real_inputs -= 1

        for input_index in xrange(num_real_inputs):
            input_layer = self.get_input_layer(input_index)
            psize = num_classes * input_layer.size
            dims = [num_classes, input_layer.size]
            self.create_input_parameter(input_index, psize, dims)
        self.create_bias_parameter(bias, num_classes)


@config_layer('addto')
class AddToLayer(LayerBase):
Q
qijun 已提交
2144
    def __init__(self, name, inputs, bias=True, **xargs):
Z
zhangjinchao01 已提交
2145 2146
        super(AddToLayer, self).__init__(
            name, 'addto', 0, inputs=inputs, **xargs)
Q
qijun 已提交
2147
        config_assert(len(inputs) > 0, 'inputs cannot be empty for AddToLayer')
Z
zhangjinchao01 已提交
2148 2149 2150 2151 2152
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
2153

Z
zhangjinchao01 已提交
2154 2155
@config_layer('agent')
class AgentLayer(LayerBase):
Q
qijun 已提交
2156 2157 2158 2159
    def __init__(self, name, size, device=None):
        super(AgentLayer, self).__init__(
            name, 'agent', size, inputs=[], device=device)

Z
zhangjinchao01 已提交
2160 2161 2162

@config_layer('sequence_agent')
class SequenceAgentLayer(LayerBase):
Q
qijun 已提交
2163
    def __init__(self, name, size, device=None):
Z
zhangjinchao01 已提交
2164 2165 2166
        super(SequenceAgentLayer, self).__init__(
            name, 'sequence_agent', size, inputs=[], device=device)

Q
qijun 已提交
2167

Z
zhangjinchao01 已提交
2168 2169
@config_layer('gather_agent')
class GatherAgentLayer(LayerBase):
Q
qijun 已提交
2170
    def __init__(self, name, size, device=None):
Z
zhangjinchao01 已提交
2171 2172 2173
        super(GatherAgentLayer, self).__init__(
            name, 'gather_agent', size, inputs=[], device=device)

Q
qijun 已提交
2174

Z
zhangjinchao01 已提交
2175 2176
@config_layer('scatter_agent')
class ScatterAgentLayer(LayerBase):
Q
qijun 已提交
2177
    def __init__(self, name, size, device=None):
Z
zhangjinchao01 已提交
2178 2179 2180
        super(ScatterAgentLayer, self).__init__(
            name, 'scatter_agent', size, inputs=[], device=device)

Q
qijun 已提交
2181

Z
zhangjinchao01 已提交
2182 2183
@config_layer('sequence_gather_agent')
class SequenceGatherAgentLayer(LayerBase):
Q
qijun 已提交
2184
    def __init__(self, name, size, device=None):
Z
zhangjinchao01 已提交
2185
        super(SequenceGatherAgentLayer, self).__init__(
Q
qijun 已提交
2186 2187
            name, 'sequence_gather_agent', size, inputs=[], device=device)

Z
zhangjinchao01 已提交
2188 2189 2190

@config_layer('sequence_scatter_agent')
class SequenceScatterAgentLayer(LayerBase):
Q
qijun 已提交
2191
    def __init__(self, name, size, device=None):
Z
zhangjinchao01 已提交
2192
        super(SequenceScatterAgentLayer, self).__init__(
Q
qijun 已提交
2193 2194
            name, 'sequence_scatter_agent', size, inputs=[], device=device)

Z
zhangjinchao01 已提交
2195 2196 2197

@config_layer('multiplex')
class MultiplexLayer(LayerBase):
Q
qijun 已提交
2198 2199 2200 2201 2202
    def __init__(self, name, inputs, size, device=None):
        super(MultiplexLayer, self).__init__(
            name, 'multiplex', size, inputs=inputs, device=device)
        config_assert(
            len(inputs) > 2, 'MultiplexLayer should have more than 2 inputs.')
Z
zhangjinchao01 已提交
2203
        for i in range(1, len(inputs)):
Q
qijun 已提交
2204 2205 2206 2207 2208
            config_assert(
                self.get_input_layer(i).size == size,
                "All the input layers except the first one should"
                "have the same size as the MultiplexLayer.")

Z
zhangjinchao01 已提交
2209 2210

@config_func
Q
qijun 已提交
2211 2212 2213
def Link(
        name,
        has_subseq=False, ):
Z
zhangjinchao01 已提交
2214 2215 2216 2217 2218
    link_config = LinkConfig()
    link_config.link_name = name
    link_config.has_subseq = has_subseq
    return link_config

Q
qijun 已提交
2219

Z
zhangjinchao01 已提交
2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233
# memory for recurrent layer group.
# *name* and *size* are actual layer's name and size.
# will return name of the memory,
# use this name if you assign the memory as other layer's input
#
# boot frame of memory is zeroed by default,
# or initialize by boot layer output if *boot_layer* set,
# or initialize by trainable bias if *boot_bias* set,
# or initialize by a constant id if *boot_with_const_id* set
#
# Memory can be a sequence if *is_sequence* set, this type of memory
# can only be initailized by a *boot_layer* which is a sequence.
#
@config_func
Q
qijun 已提交
2234 2235 2236 2237 2238 2239 2240 2241
def Memory(
        name,
        size,
        is_sequence=False,
        boot_layer=None,
        boot_bias=False,
        boot_bias_active_type="",
        boot_with_const_id=None, ):
Z
zhangjinchao01 已提交
2242 2243 2244 2245 2246 2247
    agent_name = name + "+delay1"
    if is_sequence:
        agent_layer = SequenceAgentLayer(agent_name, size)
    else:
        agent_layer = AgentLayer(agent_name, size)
    config_assert(g_current_submodel.is_recurrent_layer_group,
Q
qijun 已提交
2248
                  'Memory should be used in recurrent layer group only')
Z
zhangjinchao01 已提交
2249 2250 2251 2252
    memory = g_current_submodel.memories.add()
    memory.layer_name = MakeLayerNameInSubmodel(name)
    memory.link_name = MakeLayerNameInSubmodel(agent_name)
    memory.is_sequence = is_sequence
Q
qijun 已提交
2253
    options = sum((boot_layer is not None, bool(boot_bias),
Z
zhangjinchao01 已提交
2254
                   boot_with_const_id is not None))
Q
qijun 已提交
2255 2256 2257 2258
    config_assert(
        options <= 1,
        'take one option at most from boot_layer, boot_bias, or boot_with_const_id'
    )
Z
zhangjinchao01 已提交
2259 2260 2261
    if boot_layer is not None:
        boot_layer = MakeLayerNameInParentSubmodel(boot_layer)
        config_assert(boot_layer in g_layer_map,
Q
qijun 已提交
2262 2263
                      'boot_layer "%s" does not correspond to a layer name' %
                      boot_layer)
Z
zhangjinchao01 已提交
2264 2265 2266
        memory.boot_layer_name = boot_layer
    elif boot_bias:
        memory.boot_bias_parameter_name = agent_layer.create_bias_parameter(
Q
qijun 已提交
2267
            boot_bias, size, for_self=False)
Z
zhangjinchao01 已提交
2268 2269 2270 2271 2272
        memory.boot_bias_active_type = boot_bias_active_type
    elif boot_with_const_id is not None:
        memory.boot_with_const_id = boot_with_const_id
    return agent_name

Q
qijun 已提交
2273

Z
zhangjinchao01 已提交
2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284
# Generator for recurrent layer group, to use it:
#  1. define a id layer as output of layer group
#  2. define a memory of this id layer, and assign a boot id(begin of sequence)
#  3. define a eos check layer and fill its name in generator's *eos_layer_name*
# Sequence generation will stop when eos check return 1 or *max_num_frames* reached.
# If *beam_size* is greater than one, generator will use beam search.
#   in beam search, if *num_results_per_sample* set, one sample sequence can output
#   multiple results each with a probility.
@config_func
def Generator(
        max_num_frames,
Q
qijun 已提交
2285 2286 2287 2288
        eos_layer_name="eos_check",
        num_results_per_sample=1,
        beam_size=1,
        log_prob=None, ):
Z
zhangjinchao01 已提交
2289 2290 2291 2292 2293 2294 2295 2296 2297
    generator_config = GeneratorConfig()
    generator_config.max_num_frames = max_num_frames
    generator_config.eos_layer_name = eos_layer_name
    generator_config.num_results_per_sample = num_results_per_sample
    generator_config.beam_size = beam_size
    if log_prob is not None:
        generator_config.log_prob = log_prob
    return generator_config

Q
qijun 已提交
2298

Z
zhangjinchao01 已提交
2299 2300
@config_layer('expand')
class ExpandLayer(LayerBase):
Q
qijun 已提交
2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316
    def __init__(self,
                 name,
                 inputs,
                 trans_type='non-seq',
                 device=None,
                 bias=False):
        super(ExpandLayer, self).__init__(
            name, 'expand', 0, inputs=inputs, device=device)
        config_assert(
            len(self.inputs) == 2, 'ExpandLayer takes 2 and only 2 inputs')
        self.config.trans_type = trans_type
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
        self.set_layer_size(self.get_input_layer(0).size)
        self.create_bias_parameter(bias, self.config.size)

Z
zhangjinchao01 已提交
2317 2318 2319

@config_layer('featmap_expand')
class FeatMapExpandLayer(LayerBase):
Q
qijun 已提交
2320 2321 2322 2323 2324 2325
    def __init__(self, name, inputs, device=None, num_filters=None, bias=False):
        super(FeatMapExpandLayer, self).__init__(
            name, 'featmap_expand', 0, inputs=inputs, device=device)
        config_assert(
            len(self.inputs) == 1, 'ExpandLayer takes 1 and only 1 inputs')
        if num_filters is not None:
Z
zhangjinchao01 已提交
2326
            self.config.num_filters = num_filters
Q
qijun 已提交
2327
        else:
Z
zhangjinchao01 已提交
2328
            logger.fatal("FeatMapExpandLayer must specify num_filters.")
Q
qijun 已提交
2329
        self.set_layer_size(self.get_input_layer(0).size * num_filters)
Z
zhangjinchao01 已提交
2330 2331 2332 2333


@config_layer('max')
class MaxLayer(LayerBase):
Q
qijun 已提交
2334 2335 2336 2337 2338 2339 2340 2341 2342 2343
    def __init__(self,
                 name,
                 inputs,
                 trans_type='non-seq',
                 active_type='linear',
                 device=None,
                 bias=False,
                 output_max_index=None):
        super(MaxLayer, self).__init__(
            name, 'max', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2344
        config_assert(len(self.inputs) == 1, 'MaxLayer must have 1 input')
Q
qijun 已提交
2345 2346
        self.config.trans_type = trans_type
        self.config.active_type = active_type
Z
zhangjinchao01 已提交
2347 2348 2349 2350
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)
        self.create_bias_parameter(bias, self.config.size)
2351 2352
        if output_max_index is not None:
            self.config.output_max_index = output_max_index
Z
zhangjinchao01 已提交
2353 2354 2355 2356


@config_layer('maxid')
class MaxIdLayer(LayerBase):
Q
qijun 已提交
2357
    def __init__(self, name, inputs, beam_size=None, device=None):
Z
zhangjinchao01 已提交
2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374
        super(MaxIdLayer, self).__init__(
            name, 'maxid', 0, inputs=inputs, device=device)
        config_assert(len(self.inputs) == 1, 'MaxIdLayer must have 1 input')
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)

        if beam_size is None:
            global g_current_submodel
            if g_current_submodel.HasField("generator"):
                self.config.beam_size = g_current_submodel.generator.beam_size
        else:
            self.config.beam_size = beam_size


@config_layer('eos_id')
class EosIdLayer(LayerBase):
Q
qijun 已提交
2375
    def __init__(self, name, inputs, eos_id, device=None):
Z
zhangjinchao01 已提交
2376 2377 2378
        super(EosIdLayer, self).__init__(
            name, 'eos_id', 0, inputs=inputs, device=device)
        config_assert(len(self.inputs) == 1, 'EosIdLayer must have 1 input')
Q
qijun 已提交
2379
        self.set_layer_size(2)  # boolean output
Z
zhangjinchao01 已提交
2380 2381
        self.config.eos_id = eos_id

Q
qijun 已提交
2382

Z
zhangjinchao01 已提交
2383 2384
@config_layer('seqlastins')
class SequenceLastInstanceLayer(LayerBase):
Q
qijun 已提交
2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401
    def __init__(self,
                 name,
                 inputs,
                 active_type='linear',
                 trans_type='non-seq',
                 device=None,
                 bias=False):
        super(SequenceLastInstanceLayer, self).__init__(
            name,
            'seqlastins',
            0,
            inputs=inputs,
            device=device,
            active_type=active_type)
        config_assert(
            len(inputs) == 1, 'SequenceLastInstanceLayer must have 1 input')
        self.config.trans_type = trans_type
Z
zhangjinchao01 已提交
2402 2403 2404 2405 2406
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
2407

Z
zhangjinchao01 已提交
2408 2409 2410 2411 2412 2413 2414 2415 2416
@config_layer('seqfirstins')
class SequenceFirstInstanceLayer(SequenceLastInstanceLayer):
    def __init__(
            self,
            name,
            inputs,
            active_type='linear',
            trans_type='non-seq',
            device=None,
Q
qijun 已提交
2417 2418 2419 2420 2421 2422 2423 2424
            bias=False, ):
        super(SequenceFirstInstanceLayer, self).__init__(
            name,
            inputs=inputs,
            active_type=active_type,
            device=device,
            bias=bias)
        self.config.trans_type = trans_type
Z
zhangjinchao01 已提交
2425 2426
        self.config.select_first = True

Q
qijun 已提交
2427

Z
zhangjinchao01 已提交
2428 2429
@config_layer('seqconcat')
class SequenceConcatLayer(LayerBase):
Q
qijun 已提交
2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444
    def __init__(self,
                 name,
                 inputs,
                 active_type='linear',
                 device=None,
                 bias=False):
        super(SequenceConcatLayer, self).__init__(
            name,
            'seqconcat',
            0,
            inputs=inputs,
            device=device,
            active_type=active_type)
        config_assert(
            len(inputs) == 2, 'SequenceConcatLayer must have 2 inputs')
Z
zhangjinchao01 已提交
2445 2446 2447 2448 2449
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
2450

Z
zhangjinchao01 已提交
2451 2452
@config_layer('seqreshape')
class SequenceReshapeLayer(LayerBase):
Q
qijun 已提交
2453 2454 2455 2456 2457 2458 2459 2460 2461 2462
    def __init__(self,
                 name,
                 size,
                 inputs,
                 active_type='linear',
                 device=None,
                 bias=False):
        super(SequenceReshapeLayer, self).__init__(
            name,
            'seqreshape',
Z
zhangjinchao01 已提交
2463
            size,
Q
qijun 已提交
2464 2465 2466 2467 2468
            inputs=inputs,
            device=device,
            active_type=active_type)
        config_assert(
            len(inputs) == 1, 'SequenceReshapeLayer must have 1 inputs')
Z
zhangjinchao01 已提交
2469 2470 2471
        self.set_layer_size(size)
        self.create_bias_parameter(bias, size)

Q
qijun 已提交
2472

Z
zhangjinchao01 已提交
2473 2474
@config_layer('subseq')
class SubSequenceLayer(LayerBase):
Q
qijun 已提交
2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487
    def __init__(self,
                 name,
                 inputs,
                 active_type='linear',
                 device=None,
                 bias=False):
        super(SubSequenceLayer, self).__init__(
            name,
            'subseq',
            0,
            inputs=inputs,
            device=device,
            active_type=active_type)
Z
zhangjinchao01 已提交
2488 2489 2490 2491 2492 2493
        config_assert(len(inputs) == 3, 'SubSequenceLayer must have 3 inputs')
        input_layer0 = self.get_input_layer(0)
        size = input_layer0.size
        self.set_layer_size(size)
        self.create_bias_parameter(bias, size)

Q
qijun 已提交
2494

Z
zhangjinchao01 已提交
2495 2496
@config_layer('out_prod')
class OuterProdLayer(LayerBase):
Q
qijun 已提交
2497 2498 2499
    def __init__(self, name, inputs, device=None):
        super(OuterProdLayer, self).__init__(
            name, 'out_prod', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2500 2501 2502 2503 2504
        config_assert(len(inputs) == 2, 'OuterProdLayer must have 2 inputs')
        input_layer0 = self.get_input_layer(0)
        input_layer1 = self.get_input_layer(1)
        self.set_layer_size(input_layer0.size * input_layer1.size)

Q
qijun 已提交
2505

Z
zhangjinchao01 已提交
2506 2507
@config_layer('power')
class PowerLayer(LayerBase):
Q
qijun 已提交
2508 2509 2510
    def __init__(self, name, inputs, device=None):
        super(PowerLayer, self).__init__(
            name, 'power', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2511 2512 2513 2514
        config_assert(len(inputs) == 2, 'PowerLayer must have 2 inputs')
        input_layer1 = self.get_input_layer(1)
        self.set_layer_size(input_layer1.size)
        input_layer0 = self.get_input_layer(0)
Q
qijun 已提交
2515 2516 2517
        config_assert(1 == input_layer0.size,
                      'The left input is the exponent and should be of size 1')

Z
zhangjinchao01 已提交
2518 2519 2520

@config_layer('slope_intercept')
class SlopeInterceptLayer(LayerBase):
Q
qijun 已提交
2521 2522 2523
    def __init__(self, name, inputs, slope=1.0, intercept=0.0, device=None):
        super(SlopeInterceptLayer, self).__init__(
            name, 'slope_intercept', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2524 2525 2526 2527 2528 2529
        self.config.slope = slope
        self.config.intercept = intercept
        config_assert(len(inputs) == 1, 'SlopeInterceptLayer must have 1 input')
        input_layer0 = self.get_input_layer(0)
        self.set_layer_size(input_layer0.size)

Q
qijun 已提交
2530

Z
zhangjinchao01 已提交
2531 2532
@config_layer('scaling')
class ScalingLayer(LayerBase):
Q
qijun 已提交
2533 2534 2535
    def __init__(self, name, inputs, device=None):
        super(ScalingLayer, self).__init__(
            name, 'scaling', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2536 2537 2538 2539
        config_assert(len(inputs) == 2, 'ScalingLayer must have 2 inputs')
        input_layer1 = self.get_input_layer(1)
        self.set_layer_size(input_layer1.size)
        input_layer0 = self.get_input_layer(0)
Q
qijun 已提交
2540 2541 2542
        config_assert(1 == input_layer0.size,
                      'The left input should be of size 1')

Z
zhangjinchao01 已提交
2543 2544 2545

@config_layer('conv_shift')
class ConvShiftLayer(LayerBase):
Q
qijun 已提交
2546 2547 2548
    def __init__(self, name, inputs, device=None):
        super(ConvShiftLayer, self).__init__(
            name, 'conv_shift', 0, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2549 2550 2551 2552
        config_assert(len(inputs) == 2, 'ConvShiftLayer must have 2 inputs')
        input_layer0 = self.get_input_layer(0)
        self.set_layer_size(input_layer0.size)

Q
qijun 已提交
2553

Z
zhangjinchao01 已提交
2554 2555
@config_layer('convex_comb')
class ConvexCombinationLayer(LayerBase):
Q
qijun 已提交
2556
    def __init__(self, name, size, inputs, device=None):
Z
zhangjinchao01 已提交
2557
        super(ConvexCombinationLayer, self).__init__(
Q
qijun 已提交
2558 2559 2560
            name, 'convex_comb', size, inputs=inputs, device=device)
        config_assert(
            len(self.inputs) == 2, 'ConvexCombinationLayer must have 2 inputs')
2561 2562 2563
        config_assert(
            size * self.get_input_layer(0).size == self.get_input_layer(1).size,
            'Wrong input size for ConvexCombinationLayer')
Z
zhangjinchao01 已提交
2564 2565
        self.set_layer_size(size)

Q
qijun 已提交
2566

Z
zhangjinchao01 已提交
2567 2568
@config_layer('interpolation')
class InterpolationLayer(LayerBase):
Q
qijun 已提交
2569
    def __init__(self, name, inputs, device=None):
Z
zhangjinchao01 已提交
2570 2571
        super(InterpolationLayer, self).__init__(
            name, 'interpolation', 0, inputs=inputs, device=device)
Q
qijun 已提交
2572 2573
        config_assert(
            len(self.inputs) == 3, 'InterpolationLayer must have 3 inputs')
Z
zhangjinchao01 已提交
2574 2575 2576 2577 2578 2579 2580 2581
        input_layer0 = self.get_input_layer(0)
        input_layer1 = self.get_input_layer(1)
        input_layer2 = self.get_input_layer(2)
        self.set_layer_size(input_layer1.size)
        config_assert(input_layer0.size == 1, 'weight should be of size 1')
        config_assert(input_layer1.size == input_layer2.size,
                      'the two vector inputs should be of the same size')

Q
qijun 已提交
2582

L
liaogang 已提交
2583 2584
@config_layer('bilinear_interp')
class BilinearInterpLayer(LayerBase):
Q
qijun 已提交
2585
    def __init__(self, name, inputs, **xargs):
L
liaogang 已提交
2586
        super(BilinearInterpLayer, self).__init__(
L
liaogang 已提交
2587
            name, 'bilinear_interp', 0, inputs=inputs, **xargs)
L
liaogang 已提交
2588
        input_layer = self.get_input_layer(0)
L
Luo Tao 已提交
2589 2590 2591 2592
        conf = self.config.inputs[0].bilinear_interp_conf
        parse_bilinear(self.inputs[0].bilinear_interp, input_layer.name, conf)
        self.set_cnn_layer(name, conf.out_size_y, conf.out_size_x,
                           conf.image_conf.channels)
Q
qijun 已提交
2593

L
liaogang 已提交
2594

Z
zhangjinchao01 已提交
2595 2596
@config_layer('sum_to_one_norm')
class SumToOneNormLayer(LayerBase):
Q
qijun 已提交
2597
    def __init__(self, name, inputs, device=None):
Z
zhangjinchao01 已提交
2598
        super(SumToOneNormLayer, self).__init__(
Q
qijun 已提交
2599 2600 2601
            name, 'sum_to_one_norm', 0, inputs=inputs, device=device)
        config_assert(
            len(self.inputs) == 1, 'SumToOneNormLayer must have 1 input')
Z
zhangjinchao01 已提交
2602 2603 2604
        input_layer0 = self.get_input_layer(0)
        self.set_layer_size(input_layer0.size)

Q
qijun 已提交
2605

Z
zhangjinchao01 已提交
2606 2607
@config_layer('cos_vm')
class CosSimVecMatLayer(LayerBase):
Q
qijun 已提交
2608
    def __init__(self, name, size, inputs, cos_scale=1.0, device=None):
Z
zhangjinchao01 已提交
2609
        super(CosSimVecMatLayer, self).__init__(
Q
qijun 已提交
2610
            name, 'cos_vm', size, inputs=inputs, device=device)
Z
zhangjinchao01 已提交
2611
        self.config.cos_scale = cos_scale
Q
qijun 已提交
2612 2613
        config_assert(
            len(self.inputs) == 2, 'CosSimVecMatLayer must have 2 inputs')
2614 2615 2616
        config_assert(
            size * self.get_input_layer(0).size == self.get_input_layer(1).size,
            'Wrong input size for CosSimVecMatLayer')
Z
zhangjinchao01 已提交
2617

Q
qijun 已提交
2618

Z
zhangjinchao01 已提交
2619 2620
@config_layer('sampling_id')
class SamplingIdLayer(LayerBase):
Q
qijun 已提交
2621
    def __init__(self, name, inputs, device=None):
Z
zhangjinchao01 已提交
2622 2623
        super(SamplingIdLayer, self).__init__(
            name, 'sampling_id', 0, inputs=inputs, device=device)
Q
qijun 已提交
2624 2625
        config_assert(
            len(self.inputs) == 1, 'SamplingIdLayer must have 1 input')
Z
zhangjinchao01 已提交
2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)


# AverageLayer: "average" for each sample within a sequence.
# average_stratrgy: set to one of the following:
# 'average': plain average.
# 'sum': sum each sample instead of average (which is divide by sample_num).
# 'squarerootn': sum each sample, but divide by sqrt(sample_num).
@config_layer('average')
class AverageLayer(LayerBase):
Q
qijun 已提交
2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652
    def __init__(self,
                 name,
                 inputs,
                 average_strategy='average',
                 trans_type='non-seq',
                 active_type='linear',
                 device=None,
                 bias=False):
        super(AverageLayer, self).__init__(
            name,
            'average',
            0,
            inputs=inputs,
            device=device,
            active_type=active_type)
Z
zhangjinchao01 已提交
2653
        self.config.average_strategy = average_strategy
Q
qijun 已提交
2654
        self.config.trans_type = trans_type
Z
zhangjinchao01 已提交
2655 2656 2657 2658 2659 2660
        config_assert(len(inputs) == 1, 'AverageLayer must have 1 input')
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            self.set_layer_size(input_layer.size)
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
2661

Z
zhangjinchao01 已提交
2662 2663
@config_layer('cos')
class CosSimLayer(LayerBase):
2664
    def __init__(self, name, inputs, cos_scale=1, device=None):
Z
zhangjinchao01 已提交
2665 2666 2667 2668 2669 2670
        super(CosSimLayer, self).__init__(
            name, 'cos', 1, inputs=inputs, device=device)
        config_assert(len(self.inputs) == 2, 'CosSimLayer must have 2 inputs')
        config_assert(
            self.get_input_layer(0).size == self.get_input_layer(1).size,
            'inputs of CosSimLayer must have same dim')
2671
        self.config.cos_scale = cos_scale
Z
zhangjinchao01 已提交
2672 2673 2674 2675


@config_layer('tensor')
class TensorLayer(LayerBase):
Q
qijun 已提交
2676 2677 2678
    def __init__(self, name, size, inputs, device=None, bias=True, **xargs):
        super(TensorLayer, self).__init__(
            name, 'tensor', size, inputs=inputs, device=device, **xargs)
Z
zhangjinchao01 已提交
2679 2680
        config_assert(len(self.inputs) == 2, 'TensorLayer must have 2 inputs')
        config_assert(size > 0, 'size must be positive')
Q
qijun 已提交
2681 2682
        config_assert(inputs[1].parameter_name == None,
                      'second parameter should be None.')
Z
zhangjinchao01 已提交
2683 2684 2685 2686 2687 2688 2689 2690 2691 2692
        input_layer0 = self.get_input_layer(0)
        input_layer1 = self.get_input_layer(1)
        psize = size * input_layer0.size * input_layer1.size
        dims = [input_layer0.size, input_layer1.size, size]
        self.create_input_parameter(0, psize, dims)
        self.create_bias_parameter(bias, size)


@config_layer('mixed')
class MixedLayer(LayerBase):
Q
qijun 已提交
2693 2694 2695 2696 2697 2698 2699
    def __init__(self,
                 name,
                 inputs,
                 size=0,
                 bias=True,
                 error_clipping_threshold=None,
                 **xargs):
Z
zhangjinchao01 已提交
2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716
        config_assert(inputs, 'inputs cannot be empty')
        super(MixedLayer, self).__init__(
            name, 'mixed', size, inputs=inputs, **xargs)
        operator_input_index = []
        for operator in self.operators:
            operator_conf = operator.operator_conf
            for i in xrange(1, len(operator.input_layer_names)):
                input_index = len(self.config.inputs)
                operator_conf.input_indices.append(input_index)
                input_config = Input(operator.input_layer_names[i])
                self.inputs.append(input_config)
                layer_input = self.config.inputs.add()
                layer_input.input_layer_name = input_config.input_layer_name
            for input_index in operator_conf.input_indices:
                input_layer = self.get_input_layer(input_index)
                operator_conf.input_sizes.append(input_layer.size)
                operator_input_index.append(input_index)
2717
            if self.config.size == 0:
Z
zhangjinchao01 已提交
2718 2719 2720
                size = operator.calc_output_size(operator_conf.input_sizes)
                if size != 0:
                    self.set_layer_size(size)
2721
            else:
2722 2723
                sz = operator.calc_output_size(operator_conf.input_sizes)
                if sz != 0:
Q
qijun 已提交
2724 2725 2726 2727
                    config_assert(
                        sz == self.config.size,
                        "different inputs have different size: %s vs. %s" %
                        (sz, self.config.size))
Z
zhangjinchao01 已提交
2728 2729 2730 2731
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            input = self.inputs[input_index]
            if input_index not in operator_input_index:
Q
qijun 已提交
2732 2733 2734
                config_assert(
                    isinstance(input, Projection),
                    "input should be projection or operation")
2735
            if self.config.size == 0 and isinstance(input, Projection):
Z
zhangjinchao01 已提交
2736 2737 2738
                size = input.calc_output_size(input_layer)
                if size != 0:
                    self.set_layer_size(size)
2739
            elif isinstance(input, Projection):
Q
qijun 已提交
2740 2741 2742 2743 2744 2745
                sz = input.calc_output_size(input_layer)
                if sz != 0:
                    config_assert(
                        sz == self.config.size,
                        "different inputs have different size: %s vs. %s" %
                        (sz, self.config.size))
Z
zhangjinchao01 已提交
2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756
        config_assert(size != 0, "size is not set")

        for input_index in xrange(len(self.inputs)):
            input = self.inputs[input_index]
            if isinstance(input, Projection):
                input_layer = self.get_input_layer(input_index)
                input.proj_conf.input_size = input_layer.size
                input.proj_conf.output_size = size

                input_config = self.config.inputs[input_index]
                input_config.proj_conf.CopyFrom(input.proj_conf)
Q
qijun 已提交
2757 2758
                input_config.proj_conf.name = gen_parameter_name(name,
                                                                 input_index)
Z
zhangjinchao01 已提交
2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769
                psize = input.calc_parameter_size(input_layer.size, size)
                dims = input.calc_parameter_dims(input_layer.size, size)
                self.create_input_parameter(input_index, psize, dims)

        for operator in self.operators:
            operator_conf = operator.operator_conf
            operator_conf.output_size = self.config.size
            operator.check_dims()
            record_operator_conf = self.config.operator_confs.add()
            record_operator_conf.CopyFrom(operator_conf)

2770 2771 2772 2773 2774 2775
        psize = self.config.size
        if isinstance(self.inputs[0], ConvProjection):
            self.config.shared_biases = True
            psize = 0
            for input in self.inputs:
                psize += input.calc_bias_size()
Z
zhangjinchao01 已提交
2776

2777 2778 2779
        if bias:
            self.config.bias_size = psize
            self.create_bias_parameter(bias, psize)
Z
zhangjinchao01 已提交
2780

2781 2782
        if error_clipping_threshold is not None:
            self.config.error_clipping_threshold = error_clipping_threshold
Z
zhangjinchao01 已提交
2783

Q
qijun 已提交
2784

Z
zhangjinchao01 已提交
2785 2786
# like MixedLayer, but no bias parameter
@config_func
Q
qijun 已提交
2787
def ExpressionLayer(name, inputs, **xargs):
Z
zhangjinchao01 已提交
2788 2789
    MixedLayer(name, inputs, bias=False, **xargs)

Q
qijun 已提交
2790

Z
zhangjinchao01 已提交
2791 2792
@config_layer('concat')
class ConcatenateLayer(LayerBase):
Q
qijun 已提交
2793
    def __init__(self, name, inputs, bias=False, **xargs):
Z
zhangjinchao01 已提交
2794
        config_assert(inputs, 'inputs cannot be empty')
2795
        config_assert(not bias, 'ConcatenateLayer cannot support bias.')
Z
zhangjinchao01 已提交
2796 2797 2798 2799 2800 2801
        super(ConcatenateLayer, self).__init__(
            name, 'concat', 0, inputs=inputs, **xargs)
        size = 0
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            input = self.inputs[input_index]
Q
qijun 已提交
2802
            if self.config.size == 0:
Z
zhangjinchao01 已提交
2803 2804 2805 2806
                size += input_layer.size

        self.set_layer_size(size)

Q
qijun 已提交
2807

Z
zhangjinchao01 已提交
2808 2809 2810
# like concat layer, but each input layer was processed by a Projection.
@config_layer('concat2')
class ConcatenateLayer2(LayerBase):
Q
qijun 已提交
2811
    def __init__(self, name, inputs, bias=False, **xargs):
Z
zhangjinchao01 已提交
2812 2813 2814
        config_assert(inputs, 'inputs cannot be empty')
        super(ConcatenateLayer2, self).__init__(
            name, 'concat2', 0, inputs=inputs, **xargs)
2815 2816

        if isinstance(self.inputs[0], ConvProjection):
Q
qijun 已提交
2817 2818 2819 2820 2821 2822
            for input_index in xrange(len(self.inputs) - 1):
                input = self.inputs[input_index + 1]
                config_assert(
                    isinstance(input, ConvProjection),
                    "The first input of ConcatenateLayer2 is ConvProjection, "
                    "the other inputs should also be ConvProjection.")
2823

Z
zhangjinchao01 已提交
2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843
        size = 0
        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            input = self.inputs[input_index]
            output_size = input.calc_output_size(input_layer)
            config_assert(output_size != 0, "proj output size is not set")
            size += output_size

        self.set_layer_size(size)

        for input_index in xrange(len(self.inputs)):
            input_layer = self.get_input_layer(input_index)
            input = self.inputs[input_index]
            input.proj_conf.input_size = input_layer.size
            input.proj_conf.output_size = input.calc_output_size(input_layer)

            input_config = self.config.inputs[input_index]
            input_config.proj_conf.CopyFrom(input.proj_conf)
            input_config.proj_conf.name = gen_parameter_name(name, input_index)
            psize = input.calc_parameter_size(input.proj_conf.input_size,
Q
qijun 已提交
2844
                                              input.proj_conf.output_size)
Z
zhangjinchao01 已提交
2845
            dims = input.calc_parameter_dims(input.proj_conf.input_size,
Q
qijun 已提交
2846
                                             input.proj_conf.output_size)
Z
zhangjinchao01 已提交
2847 2848
            self.create_input_parameter(input_index, psize, dims)

2849 2850 2851 2852 2853 2854 2855
        psize = self.config.size
        if isinstance(self.inputs[0], ConvProjection):
            self.config.shared_biases = True
            psize = 0
            for input in self.inputs:
                psize += input.calc_bias_size()

2856 2857 2858
        if bias:
            self.config.bias_size = psize
            self.create_bias_parameter(bias, psize)
2859

Q
qijun 已提交
2860

Z
zhangjinchao01 已提交
2861 2862
@config_layer('recurrent')
class RecurrentLayer(LayerBase):
Q
qijun 已提交
2863
    def __init__(self, name, inputs, reversed=False, bias=True, **xargs):
Y
Yu Yang 已提交
2864 2865
        super(RecurrentLayer, self).__init__(name, 'recurrent', 0, inputs,
                                             **xargs)
Z
zhangjinchao01 已提交
2866 2867 2868 2869 2870 2871 2872 2873 2874
        config_assert(len(self.inputs) == 1, 'RecurrentLayer must have 1 input')
        input_layer = self.get_input_layer(0)
        size = input_layer.size
        self.set_layer_size(size)
        self.config.reversed = reversed
        dims = [size, size]
        self.create_input_parameter(0, size * size, dims)
        self.create_bias_parameter(bias, self.config.size)

Q
qijun 已提交
2875

Z
zhangjinchao01 已提交
2876 2877
@config_layer('lstmemory')
class LstmLayer(LayerBase):
Q
qijun 已提交
2878 2879 2880 2881 2882 2883 2884 2885
    def __init__(self,
                 name,
                 inputs,
                 reversed=False,
                 active_gate_type="sigmoid",
                 active_state_type="sigmoid",
                 bias=True,
                 **xargs):
Z
zhangjinchao01 已提交
2886 2887 2888 2889 2890 2891 2892 2893
        super(LstmLayer, self).__init__(name, 'lstmemory', 0, inputs, **xargs)
        config_assert(len(self.inputs) == 1, 'LstmLayer must have 1 input')
        input_layer = self.get_input_layer(0)
        #check input_layer.size is divided by 4
        config_assert(input_layer.size % 4 == 0, "size % 4 should be 0!")
        size = input_layer.size / 4
        self.set_layer_size(size)
        self.config.reversed = reversed
Q
qijun 已提交
2894
        self.config.active_gate_type = active_gate_type
Z
zhangjinchao01 已提交
2895 2896 2897 2898 2899
        self.config.active_state_type = active_state_type
        self.create_input_parameter(0, size * size * 4, [size, size, 4])
        #bias includes 3 kinds of peephole, 4 + 3 = 7
        self.create_bias_parameter(bias, size * 7)

Q
qijun 已提交
2900

Z
zhangjinchao01 已提交
2901 2902
@config_layer('lstm_step')
class LstmStepLayer(LayerBase):
Q
qijun 已提交
2903 2904 2905 2906 2907 2908 2909 2910 2911 2912
    def __init__(self,
                 name,
                 size,
                 inputs,
                 active_gate_type="sigmoid",
                 active_state_type="sigmoid",
                 bias=True,
                 **xargs):
        super(LstmStepLayer, self).__init__(name, 'lstm_step', size, inputs,
                                            **xargs)
Z
zhangjinchao01 已提交
2913 2914 2915
        config_assert(len(inputs) == 2, 'LstmStepLayer must have 2 inputs')
        input_layer0 = self.get_input_layer(0)
        input_layer1 = self.get_input_layer(1)
Q
qijun 已提交
2916 2917 2918 2919 2920
        config_assert(input_layer0.size == 4 * size,
                      'input_layer0.size != 4 * layer.size')
        config_assert(input_layer1.size == size,
                      'input_layer1.size != layer.size')
        self.config.active_gate_type = active_gate_type
Z
zhangjinchao01 已提交
2921 2922 2923
        self.config.active_state_type = active_state_type
        self.create_bias_parameter(bias, size * 3)

Q
qijun 已提交
2924

Z
zhangjinchao01 已提交
2925 2926 2927
# get the specific output from the input layer.
@config_layer('get_output')
class GetOutputLayer(LayerBase):
Q
qijun 已提交
2928 2929 2930 2931
    def __init__(self, name, size, inputs):
        super(GetOutputLayer, self).__init__(name, 'get_output', size, inputs)
        config_assert(
            len(self.inputs) == 1, 'GetOutputLayer must have 1 inputs')
Z
zhangjinchao01 已提交
2932 2933 2934 2935
        inputs = self.inputs[0]
        config_assert(inputs.input_layer_argument,
                      'input_layer_argument cannot be empty')

Q
qijun 已提交
2936

Z
zhangjinchao01 已提交
2937 2938
@config_layer('mdlstmemory')
class MDLstmLayer(LayerBase):
Q
qijun 已提交
2939 2940 2941 2942 2943 2944 2945 2946
    def __init__(self,
                 name,
                 inputs,
                 directions=True,
                 active_gate_type="sigmoid",
                 active_state_type="sigmoid",
                 bias=True,
                 **xargs):
Y
Yu Yang 已提交
2947 2948
        super(MDLstmLayer, self).__init__(name, 'mdlstmemory', 0, inputs,
                                          **xargs)
Z
zhangjinchao01 已提交
2949 2950 2951 2952
        config_assert(len(self.inputs) == 1, 'MDLstmLayer must have 1 input')
        input_layer = self.get_input_layer(0)
        dim_num = len(directions)
        #check input_layer.size is divided by (3+dim_num)
Y
Yu Yang 已提交
2953 2954
        config_assert(input_layer.size % (3 + dim_num) == 0,
                      "size % (dim_num) should be 0!")
Q
qijun 已提交
2955
        size = input_layer.size / (3 + dim_num)
Z
zhangjinchao01 已提交
2956
        self.set_layer_size(size)
Q
qijun 已提交
2957
        self.config.active_gate_type = active_gate_type
Z
zhangjinchao01 已提交
2958 2959 2960
        self.config.active_state_type = active_state_type
        for i in xrange(len(directions)):
            self.config.directions.append(int(directions[i]))
Y
Yu Yang 已提交
2961 2962
        self.create_input_parameter(0, size * size * (3 + dim_num),
                                    [size, size, 3 + dim_num])
Z
zhangjinchao01 已提交
2963
        #bias includes 3 kinds of peephole, 3+dim_num+2+dim_num
Q
qijun 已提交
2964 2965
        self.create_bias_parameter(bias, size * (5 + 2 * dim_num))

Z
zhangjinchao01 已提交
2966 2967 2968

@config_layer('gated_recurrent')
class GatedRecurrentLayer(LayerBase):
Q
qijun 已提交
2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979
    def __init__(self,
                 name,
                 inputs,
                 reversed=False,
                 active_gate_type="sigmoid",
                 bias=True,
                 **xargs):
        super(GatedRecurrentLayer, self).__init__(name, 'gated_recurrent', 0,
                                                  inputs, **xargs)
        config_assert(
            len(self.inputs) == 1, 'GatedRecurrentLayer must have 1 input')
Z
zhangjinchao01 已提交
2980 2981 2982 2983 2984 2985
        input_layer = self.get_input_layer(0)
        #check input_layer.size is divided by 3
        config_assert(input_layer.size % 3 == 0, "size % 3 should be 0!")
        size = input_layer.size / 3
        self.set_layer_size(size)
        self.config.reversed = reversed
Q
qijun 已提交
2986
        self.config.active_gate_type = active_gate_type
Z
zhangjinchao01 已提交
2987 2988 2989
        self.create_input_parameter(0, size * size * 3, [size, size * 3])
        self.create_bias_parameter(bias, size * 3)

Q
qijun 已提交
2990

Z
zhangjinchao01 已提交
2991 2992
@config_layer('gru_step')
class GruStepLayer(LayerBase):
Q
qijun 已提交
2993 2994 2995 2996 2997 2998 2999
    def __init__(self,
                 name,
                 size,
                 inputs,
                 active_gate_type="sigmoid",
                 bias=True,
                 **xargs):
Y
Yu Yang 已提交
3000 3001
        super(GruStepLayer, self).__init__(name, 'gru_step', size, inputs,
                                           **xargs)
Z
zhangjinchao01 已提交
3002 3003 3004
        config_assert(len(self.inputs) == 2, 'GruStepLayer must have 2 input')
        input_layer0 = self.get_input_layer(0)
        input_layer1 = self.get_input_layer(1)
Q
qijun 已提交
3005 3006 3007 3008 3009
        config_assert(input_layer0.size == 3 * size,
                      'input_layer0.size != 3 * layer.size')
        config_assert(input_layer1.size == size,
                      'input_layer1.size != layer.size')
        self.config.active_gate_type = active_gate_type
W
wangyang59 已提交
3010
        self.create_input_parameter(1, size * size * 3, [size, size * 3])
Z
zhangjinchao01 已提交
3011 3012
        self.create_bias_parameter(bias, size * 3)

Q
qijun 已提交
3013

Z
zhangjinchao01 已提交
3014 3015 3016 3017 3018 3019 3020
'''
 A layer for calculating the cost of sequential conditional random field model.
 Example: CRFLayer(name="crf_cost", size=label_num,
                   inputs=["output", "label", "weight"])
          where "weight" is optional, one weight for each sequence
 @param coeff: weight of the layer
'''
Q
qijun 已提交
3021 3022


Z
zhangjinchao01 已提交
3023 3024
@config_layer('crf')
class CRFLayer(LayerBase):
Q
qijun 已提交
3025
    def __init__(self, name, size, inputs, coeff=1.0, device=None):
Z
zhangjinchao01 已提交
3026
        super(CRFLayer, self).__init__(name, 'crf', size, inputs, device=device)
Q
qijun 已提交
3027 3028
        config_assert(2 <= len(self.inputs) <= 3,
                      'CRFLayer must have 2 or 3 inputs')
Z
zhangjinchao01 已提交
3029 3030 3031
        self.create_input_parameter(0, size * (size + 2), [size, size + 2])
        self.config.coeff = coeff

Q
qijun 已提交
3032

Z
zhangjinchao01 已提交
3033 3034 3035 3036 3037 3038 3039 3040
'''
 A layer for calculating the decoding sequence of sequential conditional
 random field model.
 The decoding sequence is stored in output_.ids
 If a second input is provided, it is treated as the ground-truth label, and
 this layer will also calculate error, output_.value[i] is 1 for incorrect
 decoding or 0 for correct decoding
'''
Q
qijun 已提交
3041 3042


Z
zhangjinchao01 已提交
3043 3044
@config_layer('crf_decoding')
class CRFDecodingLayer(LayerBase):
Q
qijun 已提交
3045
    def __init__(self, name, size, inputs, device=None):
Z
zhangjinchao01 已提交
3046 3047 3048 3049 3050 3051 3052
        super(CRFDecodingLayer, self).__init__(
            name, 'crf_decoding', size, inputs, device=device)
        config_assert(
            len(self.inputs) <= 2,
            'CRFDecodingLayer cannot have more than 2 inputs')
        self.create_input_parameter(0, size * (size + 2), [size, size + 2])

Q
qijun 已提交
3053

Z
zhangjinchao01 已提交
3054 3055
@config_layer('ctc')
class CTCLayer(LayerBase):
Q
qijun 已提交
3056
    def __init__(self, name, size, inputs, norm_by_times=False, device=None):
Z
zhangjinchao01 已提交
3057 3058 3059 3060
        super(CTCLayer, self).__init__(name, 'ctc', size, inputs, device=device)
        self.config.norm_by_times = norm_by_times
        config_assert(len(self.inputs) == 2, 'CTCLayer must have 2 inputs')

Q
qijun 已提交
3061

3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082
@config_layer('warp_ctc')
class WarpCTCLayer(LayerBase):
    def __init__(self,
                 name,
                 size,
                 inputs,
                 blank=0,
                 norm_by_times=False,
                 device=None):
        super(WarpCTCLayer, self).__init__(
            name, 'warp_ctc', size=size, inputs=inputs, device=device)
        self.config.blank = blank
        self.config.norm_by_times = norm_by_times
        config_assert(len(self.inputs) == 2, 'WarpCTCLayer must have 2 inputs')
        input_layer = self.get_input_layer(0)
        config_assert(
            (input_layer.active_type == '' or
             input_layer.active_type == 'linear'),
            "Expecting the active_type of input layer to be linear or null")


Z
zhangjinchao01 已提交
3083 3084
@config_layer('recurrent_layer_group')
class RecurrentLayerGroup(LayerBase):
Q
qijun 已提交
3085
    def __init__(self, name, device=None):
L
Luo Tao 已提交
3086 3087
        global g_pass_height_width
        g_pass_height_width = False
Z
zhangjinchao01 已提交
3088 3089 3090 3091 3092 3093
        super(RecurrentLayerGroup, self).__init__(
            name, 'recurrent_layer_group', 0, inputs=[], device=device)


# Deprecated, use a new layer specific class instead
@config_func
Q
qijun 已提交
3094
def Layer(name, type, **xargs):
Z
zhangjinchao01 已提交
3095 3096 3097 3098
    layers = {}
    layers.update(g_cost_map)
    layers.update(g_layer_type_map)
    layer_func = layers.get(type)
Q
qijun 已提交
3099
    config_assert(layer_func, "layer type '%s' not supported." % type)
X
xuwei06 已提交
3100
    return layer_func(name, **xargs)
Z
zhangjinchao01 已提交
3101

Q
qijun 已提交
3102

Z
zhangjinchao01 已提交
3103
@config_func
Q
qijun 已提交
3104
def ParameterHook(type, **kwargs):
Z
zhangjinchao01 已提交
3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116
    if type == 'pruning':
        mask_filename = kwargs.get('mask_filename', None)
        assert mask_filename is not None
        hook = ParameterUpdaterHookConfig()
        hook.type = type
        hook.purning_mask_filename = mask_filename
        return hook
    else:
        return None


@config_func
Q
qijun 已提交
3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138
def Parameter(name,
              size,
              device,
              dims,
              learning_rate=None,
              momentum=None,
              decay_rate=None,
              decay_rate_l1=None,
              initial_mean=None,
              initial_std=None,
              initial_strategy=None,
              initial_smart=None,
              num_batches_regularization=None,
              sparse_remote_update=None,
              sparse_update=None,
              gradient_clipping_threshold=None,
              sparse=None,
              format=None,
              need_compact=None,
              is_static=None,
              is_shared=None,
              update_hooks=None):
Z
zhangjinchao01 已提交
3139 3140 3141 3142 3143 3144 3145

    config_assert(name not in g_parameter_map,
                  'Duplicated parameter name: ' + name)

    para = g_config.model_config.parameters.add()
    para.name = name
    para.size = size
3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156
    if device is not None:
        para.device = int(device)
    para.dims.extend(dims)

    if learning_rate is not None:
        para.learning_rate = float(learning_rate)

    momentum = default(momentum, g_default_momentum)
    if momentum is not None:
        para.momentum = float(momentum)

Z
zhangjinchao01 已提交
3157 3158
    config_assert(not momentum or not decay_rate_l1,
                  "momentum and decay_rate_l1 cannot both be non-zero")
3159 3160 3161 3162 3163

    decay_rate = default(decay_rate, g_default_decay_rate)
    if decay_rate is not None:
        para.decay_rate = decay_rate

Z
zhangjinchao01 已提交
3164 3165 3166 3167
    if decay_rate_l1 is not None:
        para.decay_rate_l1 = decay_rate_l1
    para.initial_std = default(initial_std, g_default_initial_std)
    para.initial_mean = default(initial_mean, g_default_initial_mean)
3168

Q
qijun 已提交
3169 3170
    num_batches_regularization = default(num_batches_regularization,
                                         g_default_num_batches_regularization)
3171 3172 3173
    if num_batches_regularization is not None:
        para.num_batches_regularization = int(num_batches_regularization)

Z
zhangjinchao01 已提交
3174 3175 3176 3177 3178 3179
    if sparse_remote_update is not None:
        para.sparse_remote_update = sparse_remote_update
        if sparse_remote_update:
            g_config.opt_config.use_sparse_remote_updater = True
    if sparse_update is not None:
        para.sparse_update = sparse_update
Q
qijun 已提交
3180 3181
    gradient_clipping_threshold = default(gradient_clipping_threshold,
                                          g_default_gradient_clipping_threshold)
3182 3183
    if gradient_clipping_threshold is not None:
        para.gradient_clipping_threshold = gradient_clipping_threshold
Q
qijun 已提交
3184 3185
    para.initial_strategy = default(initial_strategy,
                                    g_default_initial_strategy)
Z
zhangjinchao01 已提交
3186 3187 3188 3189 3190 3191
    para.initial_smart = default(initial_smart, g_default_initial_smart)
    if para.initial_smart:
        para.initial_mean = 0.
        if len(para.dims) != 0:
            para.initial_std = 1. / math.sqrt(para.dims[0])
        else:
Q
qijun 已提交
3192 3193 3194
            print(
                "Use initial_smart, but dims not set. Initial_smart may not be used in this layer"
            )
Z
zhangjinchao01 已提交
3195 3196 3197 3198
            traceback.print_exc()
            para.initial_std = 1. / math.sqrt(para.size)
    if g_default_compact_func is not None:
        sparse, format, need_compact = g_default_compact_func(para.name)
3199 3200 3201 3202 3203 3204 3205

    if sparse is not None:
        para.is_sparse = sparse
    if format is not None:
        para.format = format
    if need_compact is not None:
        para.need_compact = need_compact
Z
zhangjinchao01 已提交
3206 3207 3208 3209
    if is_static is not None:
        para.is_static = is_static
    config_assert(not para.sparse_remote_update or not para.is_static,
                  "sparse_remote_update and is_static cannot both be true")
3210 3211
    if is_shared is not None:
        para.is_shared = is_shared
Z
zhangjinchao01 已提交
3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232

    update_hooks = default(update_hooks, g_default_update_hooks)

    if update_hooks is not None:
        if hasattr(update_hooks, '__call__'):
            update_hooks = update_hooks(para.name)

        if isinstance(update_hooks, list):
            for hook in update_hooks:
                para.update_hooks.extend([hook])
        else:
            para.update_hooks.extend(update_hooks)

    g_parameter_map[name] = para


@config_func
def default_initial_std(val):
    global g_default_initial_std
    g_default_initial_std = val

Q
qijun 已提交
3233

Z
zhangjinchao01 已提交
3234 3235 3236 3237 3238
@config_func
def default_initial_mean(val):
    global g_default_initial_mean
    g_default_initial_mean = val

Q
qijun 已提交
3239

Z
zhangjinchao01 已提交
3240 3241 3242 3243 3244
@config_func
def default_initial_strategy(val):
    global g_default_initial_strategy
    g_default_initial_strategy = val

Q
qijun 已提交
3245

Z
zhangjinchao01 已提交
3246 3247 3248 3249 3250
@config_func
def default_initial_smart(val):
    global g_default_initial_smart
    g_default_initial_smart = val

Q
qijun 已提交
3251

Z
zhangjinchao01 已提交
3252 3253 3254 3255 3256
@config_func
def default_momentum(val):
    global g_default_momentum
    g_default_momentum = val

Q
qijun 已提交
3257

Z
zhangjinchao01 已提交
3258 3259 3260 3261 3262
@config_func
def default_decay_rate(val):
    global g_default_decay_rate
    g_default_decay_rate = val

Q
qijun 已提交
3263

Z
zhangjinchao01 已提交
3264 3265 3266 3267 3268
@config_func
def default_num_batches_regularization(val):
    global g_default_num_batches_regularization
    g_default_num_batches_regularization = val

Q
qijun 已提交
3269

Z
zhangjinchao01 已提交
3270 3271 3272 3273 3274
@config_func
def default_gradient_clipping_threshold(val):
    global g_default_gradient_clipping_threshold
    g_default_gradient_clipping_threshold = val

Q
qijun 已提交
3275

Z
zhangjinchao01 已提交
3276 3277 3278 3279 3280
@config_func
def default_device(val):
    global g_default_device
    g_default_device = val

Q
qijun 已提交
3281

Z
zhangjinchao01 已提交
3282 3283 3284 3285 3286
@config_func
def default_update_hooks(val):
    global g_default_update_hooks
    g_default_update_hooks = val

Q
qijun 已提交
3287

Z
zhangjinchao01 已提交
3288 3289 3290 3291 3292
@config_func
def default_compact_func(val):
    global g_default_compact_func
    g_default_compact_func = val

Q
qijun 已提交
3293

Z
zhangjinchao01 已提交
3294 3295 3296 3297 3298
def make_importer(config_dir, config_args):
    def Import(config_file, local_args={}):
        if not config_file.startswith('/'):
            config_file = config_dir + '/' + config_file
            g_config.config_files.append(config_file)
Q
qijun 已提交
3299 3300 3301
        execfile(config_file,
                 make_config_environment(config_file, config_args), local_args)

Z
zhangjinchao01 已提交
3302 3303
    return Import

Q
qijun 已提交
3304

Z
zhangjinchao01 已提交
3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332
settings = dict(
    batch_size=None,
    mini_batch_size=None,
    algorithm='async_sgd',
    async_lagged_grad_discard_ratio=1.5,
    learning_method='momentum',
    num_batches_per_send_parameter=None,
    num_batches_per_get_parameter=None,
    center_parameter_update_method=None,
    learning_rate=1.,
    learning_rate_decay_a=0.,
    learning_rate_decay_b=0.,
    learning_rate_schedule='poly',
    learning_rate_args='',
    l1weight=0.1,
    l2weight=0.,
    l2weight_zero_iter=0,
    c1=0.0001,
    backoff=0.5,
    owlqn_steps=10,
    max_backoff=5,
    average_window=0,
    do_average_in_cpu=False,
    max_average_window=None,
    ada_epsilon=1e-6,
    ada_rou=0.95,
    delta_add_rate=1.0,
    shrink_parameter_value=0,
Q
qijun 已提交
3333 3334 3335
    adam_beta1=0.9,
    adam_beta2=0.999,
    adam_epsilon=1e-8, )
Z
zhangjinchao01 已提交
3336

Q
qijun 已提交
3337
settings_deprecated = dict(usage_ratio=1., )
Z
zhangjinchao01 已提交
3338 3339 3340 3341

trainer_settings = dict(
    save_dir="./output/model",
    init_model_path=None,
Q
qijun 已提交
3342 3343
    start_pass=0, )

Z
zhangjinchao01 已提交
3344 3345 3346 3347 3348

@config_func
def Settings(**args):
    for k, v in args.iteritems():
        if k == "usage_ratio":
Q
qijun 已提交
3349 3350
            logger.warning(
                "Deprecated: define usage_ratio in DataConfig instead")
Z
zhangjinchao01 已提交
3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361
            if g_config.HasField("data_config"):
                g_config.data_config.__setattr__(k, v)
            settings_deprecated[k] = v
            continue
        elif k in settings:
            settings[k] = v
        elif k in trainer_settings:
            trainer_settings[k] = v
        else:
            logger.fatal('Unkown setting: %s' % k)

Q
qijun 已提交
3362

Z
zhangjinchao01 已提交
3363 3364 3365 3366
@config_func
def cluster_config(**args):
    pass

Q
qijun 已提交
3367

Z
zhangjinchao01 已提交
3368 3369 3370 3371 3372 3373 3374 3375 3376
@config_func
def EnableSubmodelSuffix(flag=True):
    """
    If enabled, the layer and evaluator names in submodel will be automatically
    appended with @submodel_name
    """
    global g_add_submodel_suffix
    g_add_submodel_suffix = flag

Q
qijun 已提交
3377

Z
zhangjinchao01 已提交
3378 3379 3380 3381
def make_config_environment(config_file, config_args):
    def make_setter(k):
        def setter(v):
            logger.fatal("Obsolete: use Settings(%s=%s, ...) instead" % (k, v))
Q
qijun 已提交
3382

Z
zhangjinchao01 已提交
3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397
        return setter

    funcs = {}
    funcs.update(g_config_funcs)

    for k in settings.iterkeys():
        funcs[k] = make_setter(k)
    for k in settings_deprecated.iterkeys():
        funcs[k] = make_setter(k)
    config_dir = os.path.dirname(config_file)
    if not config_dir:
        config_dir = '.'

    funcs.update(
        Import=make_importer(config_dir, config_args),
Q
qijun 已提交
3398
        get_config_arg=make_get_config_arg(config_args), )
Z
zhangjinchao01 已提交
3399 3400 3401 3402 3403

    funcs.update(g_extended_config_funcs)

    return funcs

Q
qijun 已提交
3404

Z
zhangjinchao01 已提交
3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420
def make_get_config_arg(config_args):
    def get_config_arg(name, type, default=None):
        if type == bool:
            s = config_args.get(name)
            if not s:
                return default
            if s == 'True' or s == '1' or s == 'true':
                return True
            if s == 'False' or s == '0' or s == 'false':
                return False
            raise ValueError('Value of config_arg %s is not boolean' % name)
        else:
            return type(config_args.get(name, default))

    return get_config_arg

Q
qijun 已提交
3421

Z
zhangjinchao01 已提交
3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433
def importlib(name):
    __import__(name)
    return sys.modules[name]


def find_caller():
    stack = traceback.extract_stack()
    for s in stack[-4::-1]:
        if not s[0].endswith('config_parser.py'):
            return s[0], s[1], s[2]
    return "(unknown file)", 0, "(unknown function)"

Q
qijun 已提交
3434

Z
zhangjinchao01 已提交
3435 3436 3437 3438
def my_fatal(s):
    logger.critical(s)
    raise Exception()

Y
Yu Yang 已提交
3439

3440
_parse_config_hooks = set()
Y
Yu Yang 已提交
3441 3442


3443 3444 3445 3446 3447 3448 3449
def register_parse_config_hook(f):
    """
    Register a hook function for parse_config. parse_config will invoke the hook
    at the beginning of parse. This make it possible to reset global state for
    for constructing the model.
    """
    _parse_config_hooks.add(f)
Q
qijun 已提交
3450

Y
Yu Yang 已提交
3451

3452
def update_g_config():
Z
zhangjinchao01 已提交
3453
    '''
3454 3455 3456 3457 3458 3459 3460 3461 3462 3463 3464 3465 3466 3467 3468 3469 3470 3471 3472 3473 3474 3475 3476 3477 3478 3479 3480
    Update g_config after execute config_file or config_functions.
    '''
    for k, v in settings.iteritems():
        if v is None:
            continue
        g_config.opt_config.__setattr__(k, v)

    for k, v in trainer_settings.iteritems():
        if v is None:
            continue
        g_config.__setattr__(k, v)

    for name in g_config.model_config.input_layer_names:
        assert name in g_layer_map, \
            'input name "%s" does not correspond to a layer name' % name
        assert (g_layer_map[name].type == "data" or g_layer_map[name].type == "data_trim"), \
            'The type of input layer "%s" is not "data"' % name
    for name in g_config.model_config.output_layer_names:
        assert name in g_layer_map, \
            'input name "%s" does not correspond to a layer name' % name
    return g_config


def parse_config(trainer_config, config_arg_str):
    '''
    @param trainer_config: can be a string of config file name or a function name
    with config logic
Z
zhangjinchao01 已提交
3481 3482 3483 3484
    @param config_arg_str: a string of the form var1=val1,var2=val2. It will be
    passed to config script as a dictionary CONFIG_ARGS
    '''
    init_config_environment()
3485 3486
    for hook in _parse_config_hooks:
        hook()
Z
zhangjinchao01 已提交
3487 3488 3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500 3501 3502 3503 3504 3505 3506 3507 3508 3509 3510 3511 3512 3513

    config_args = {}

    logger.findCaller = find_caller
    logger.fatal = my_fatal

    g_config.model_config.type = "nn"
    if config_arg_str:
        config_args = dict([f.split('=') for f in config_arg_str.split(',')])

    global g_command_config_args
    g_command_config_args.update(config_args)

    extension_module_name = config_args.get('extension_module_name')
    if extension_module_name:
        global g_extended_config_funcs
        extension_module = importlib(extension_module_name)
        g_extended_config_funcs = extension_module.get_config_funcs(g_config)

    g_config.model_config.type = 'nn'

    global g_current_submodel, g_root_submodel
    g_root_submodel = g_config.model_config.sub_models.add()
    g_root_submodel.name = 'root'
    g_root_submodel.is_recurrent_layer_group = False
    g_current_submodel = g_root_submodel

3514 3515
    if hasattr(trainer_config, '__call__'):
        trainer_config.func_globals.update(
L
Luo Tao 已提交
3516
            make_config_environment("", config_args))
3517
        trainer_config()
H
hanchao 已提交
3518
    else:
3519 3520
        execfile(trainer_config,
                 make_config_environment(trainer_config, config_args))
Z
zhangjinchao01 已提交
3521

3522
    return update_g_config()
Z
zhangjinchao01 已提交
3523 3524


3525
def parse_config_and_serialize(trainer_config, config_arg_str):
Z
zhangjinchao01 已提交
3526
    try:
3527
        config = parse_config(trainer_config, config_arg_str)
Z
zhangjinchao01 已提交
3528 3529 3530 3531 3532 3533
        #logger.info(config)
        return config.SerializeToString()
    except:
        traceback.print_exc()
        raise

Q
qijun 已提交
3534

Z
zhangjinchao01 已提交
3535 3536 3537 3538 3539 3540 3541 3542
if __name__ == '__main__':
    try:
        config = parse_config(sys.argv[1], '')
        config.SerializeToString()
        __real_print__(str(config))
    except:
        traceback.print_exc()
        raise