layers.py 55.1 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14
# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
#
# 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.

X
Xin Pan 已提交
15
import collections
16 17 18
import contextlib
import sys
import numpy as np
19
import six
20
import re
21 22 23
import copy
import weakref
import warnings
24
from copy import deepcopy
25 26
import inspect

27
import paddle
28

C
chengduo 已提交
29
from . import parallel_helper
X
Xin Pan 已提交
30
from .. import unique_name
31
from paddle.fluid import core
32
from .layer_object_helper import LayerObjectHelper
33
from .base import program_desc_tracing_guard, param_guard
34
from paddle.fluid import framework
35
from ..param_attr import ParamAttr
36 37 38
from paddle.fluid.executor import Executor, global_scope
from paddle.fluid.framework import in_dygraph_mode
from paddle.fluid.framework import _current_expected_place as _get_device
C
chentianyu03 已提交
39
from paddle.fluid.dygraph import no_grad
W
wanghuancoder 已提交
40
import paddle.utils.deprecated as deprecated
41

42
__all__ = ['Layer']
43

44 45 46 47 48 49 50 51
_first_cap_re = re.compile('(.)([A-Z][a-z]+)')
_all_cap_re = re.compile('([a-z])([A-Z])')


def _convert_camel_to_snake(name):
    s1 = _first_cap_re.sub(r'\1_\2', name)
    return _all_cap_re.sub(r'\1_\2', s1).lower()

52

53 54 55 56 57 58 59 60 61 62 63
def _addindent(string, indent):
    s1 = string.split('\n')
    if len(s1) == 1:
        return string
    s2 = []
    for idx, line in enumerate(s1):
        if idx > 0:
            s2.append(str((indent * ' ') + line))
    return s1[0] + '\n' + '\n'.join(s2)


64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79
class HookRemoveHelper(object):
    """ A HookRemoveHelper that can be used to remove hook. """

    next_hook_id = 0

    def __init__(self, hooks):
        self._hooks_ref = weakref.ref(hooks)
        self._hook_id = HookRemoveHelper.next_hook_id
        HookRemoveHelper.next_hook_id += 1

    def remove(self):
        hooks = self._hooks_ref()
        if hooks is not None and self._hook_id in hooks:
            del hooks[self._hook_id]


X
Xin Pan 已提交
80
class Layer(core.Layer):
81 82
    """
    Dynamic graph Layer based on OOD, includes the parameters of the layer, the structure of the forward graph and so on.
X
Xin Pan 已提交
83

84
    Parameters:
85 86
        name_scope (str, optional): prefix name used by the layer to name parameters.
            If prefix is "my_layer", parameter name in MyLayer
87 88 89
            can be "my_layer_0.w_n", where "w" is the parameter
            base name and "n" is an unique suffix auto-generated.
            If None, prefix name will be snake cased class name. Default: None.
90
        dtype(str, optional): data type of this parameter.
91 92
                If set str, it can be "bool",  "float16", "float32", "float64",
                "int8", "int16", "int32", "int64", "uint8" or "uint16".
93
                Default: "float32"
94 95 96
    
    Returns:
        None
X
Xin Pan 已提交
97
    """
X
Xin Pan 已提交
98

99
    def __init__(self, name_scope=None, dtype="float32"):
100
        self.training = True
101
        if name_scope is None:
102 103
            name_scope = _convert_camel_to_snake(self.__class__.__name__)
        self._full_name = unique_name.generate(name_scope)
104
        self._helper = LayerObjectHelper(self._full_name)
X
Xin Pan 已提交
105
        self._built = False
M
minqiyang 已提交
106
        self._dtype = dtype
107
        self._init_in_dynamic_mode = framework.in_dygraph_mode()
108

X
Xin Pan 已提交
109
        self._parameters = collections.OrderedDict()
110 111 112
        # Buffers the variable (not parameter) created in layer
        self._buffers = collections.OrderedDict()
        self._non_persistable_buffer_names_set = set()
X
Xin Pan 已提交
113
        self._sub_layers = collections.OrderedDict()
L
lujun 已提交
114
        self._loaddict_holder = collections.OrderedDict()
115

116 117 118
        self._forward_pre_hooks = collections.OrderedDict()
        self._forward_post_hooks = collections.OrderedDict()

M
minqiyang 已提交
119
    def train(self):
120 121 122 123 124 125
        """
        Sets this Layer and all its sublayers to training mode.
        This only effects certain modules like `Dropout` and `BatchNorm`.

        Returns:
            None
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149

        Example::
            .. code-block:: python

                import paddle

                class MyLayer(paddle.nn.Layer):
                    def __init__(self):
                        super(MyLayer, self).__init__()
                        self._linear = paddle.nn.Linear(1, 1)
                        self._dropout = paddle.nn.Dropout(p=0.5)

                    def forward(self, input):
                        temp = self._linear(input)
                        temp = self._dropout(temp)
                        return temp

                x = paddle.randn([10, 1], 'float32')
                mylayer = MyLayer()
                mylayer.eval()  # set mylayer._dropout to eval mode
                out = mylayer(x)
                mylayer.train()  # set mylayer._dropout to train mode
                out = mylayer(x)

150
        """
151 152 153 154 155
        # global setting in dygraph
        # NOTE(chenweihang): nn.Layer also can be used in static mode,
        # but _dygraph_tracer() can not be called in static mode
        if in_dygraph_mode():
            framework._dygraph_tracer().train_mode()
156 157 158
        # Layer-level setting
        self.training = True
        for layer in self.sublayers():
159
            layer.training = True
M
minqiyang 已提交
160 161

    def eval(self):
162 163 164 165 166 167
        """
        Sets this Layer and all its sublayers to evaluation mode.
        This only effects certain modules like `Dropout` and `BatchNorm`.

        Returns:
            None
168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190

        Example::
            .. code-block:: python

                import paddle

                class MyLayer(paddle.nn.Layer):
                    def __init__(self):
                        super(MyLayer, self).__init__()
                        self._linear = paddle.nn.Linear(1, 1)
                        self._dropout = paddle.nn.Dropout(p=0.5)

                    def forward(self, input):
                        temp = self._linear(input)
                        temp = self._dropout(temp)
                        return temp

                x = paddle.randn([10, 1], 'float32')
                mylayer = MyLayer()
                mylayer.eval()  # set mylayer._dropout to eval mode
                out = mylayer(x)
                print(out)

191
        """
192 193 194 195 196
        # global setting in dygraph
        # NOTE(chenweihang): nn.Layer also can be used in static mode,
        # but _dygraph_tracer() can not be called in static mode
        if in_dygraph_mode():
            framework._dygraph_tracer().eval_mode()
197 198 199
        # Layer-level setting
        self.training = False
        for layer in self.sublayers():
200
            layer.training = False
M
minqiyang 已提交
201

L
LielinJiang 已提交
202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
    def apply(self, fn):
        """
        Applies ``fn`` recursively to every sublayer (as returned by ``.sublayers()``)
        as well as self. Typical use includes initializing the parameters of a model.

        Parameters:
            fn (function): a function to be applied to each sublayer

        Returns:
            Layer: self

        Example::
            .. code-block:: python

              import paddle
              import paddle.nn as nn
218

L
LielinJiang 已提交
219 220 221 222 223
              net = nn.Sequential(nn.Linear(2, 2), nn.Linear(2, 2))

              def init_weights(layer):
                  if type(layer) == nn.Linear:
                      print('before init weight:', layer.weight.numpy())
224
                      new_weight = paddle.full(shape=layer.weight.shape, dtype=layer.weight.dtype, fill_value=0.9)
L
LielinJiang 已提交
225 226 227 228 229 230 231
                      layer.weight.set_value(new_weight)
                      print('after init weight:', layer.weight.numpy())

              net.apply(init_weights)

              print(net.state_dict())
        """
232
        for layer in self.children():
L
LielinJiang 已提交
233 234 235 236 237 238
            layer.apply(fn)

        fn(self)

        return self

X
Xin Pan 已提交
239
    def full_name(self):
240
        """Full name for this layer, composed by name_scope + "/" + MyLayer.__class__.__name__
X
Xin Pan 已提交
241

242 243
        Returns:
            str: full name of this layer.
244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260

        Example::
            .. code-block:: python

                import paddle

                class LinearNet(paddle.nn.Layer):
                    def __init__(self):
                        super(LinearNet, self).__init__(name_scope = "demo_linear_net")
                        self._linear = paddle.nn.Linear(1, 1)

                    def forward(self, x):
                        return self._linear(x)

                linear_net = LinearNet()
                print(linear_net.full_name())   # demo_linear_net_0

X
Xin Pan 已提交
261 262 263
        """
        return self._full_name

264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280
    def register_forward_post_hook(self, hook):
        """Register a forward post-hook for Layer. The hook will be called after `forward` function has been computed.

        It should have the following form, `input` and `output` of the `hook` is `input` and `output` of the `Layer` respectively.
        User can use forward post-hook to change the output of the Layer or perform information statistics tasks on the Layer.
 
        hook(Layer, input, output) -> None or modified output

        Parameters:
            hook(function): a function registered as a forward post-hook

        Returns:
            HookRemoveHelper: a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .

        Examples:
            .. code-block:: python

281 282 283 284 285 286
                import paddle
                import numpy as np

                # the forward_post_hook change the output of the layer: output = output * 2
                def forward_post_hook(layer, input, output):
                    # user can use layer, input and output for information statistis tasks
287

288 289
                    # change the output
                    return output * 2
290

291
                linear = paddle.nn.Linear(13, 5)
292

293 294
                # register the hook
                forward_post_hook_handle = linear.register_forward_post_hook(forward_post_hook)
295

296 297
                value1 = np.arange(26).reshape(2, 13).astype("float32")
                in1 = paddle.to_tensor(value1)
298

299
                out0 = linear(in1)
300

301 302 303 304 305 306 307
                # remove the hook
                forward_post_hook_handle.remove()

                out1 = linear(in1)

                # hook change the linear's output to output * 2, so out0 is equal to out1 * 2.
                assert (out0.numpy() == (out1.numpy()) * 2).any()
308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331
        """
        hook_remove_helper = HookRemoveHelper(self._forward_post_hooks)
        self._forward_post_hooks[hook_remove_helper._hook_id] = hook
        return hook_remove_helper

    def register_forward_pre_hook(self, hook):
        """Register a forward pre-hook for Layer. The hook will be called before `forward` function has been computed.
        
        It should have the following form, `input` of the `hook` is `input` of the `Layer`,
        hook can either return a tuple or a single modified value in the hook. We will wrap the value into a tuple if 
        a single value is returned(unless that value is already a tuple).
        User can use forward pre-hook to change the input of the Layer or perform information statistics tasks on the Layer.

        hook(Layer, input) -> None or modified input

        Parameters:
            hook(function): a function registered as a forward pre-hook

        Returns:
            HookRemoveHelper: a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .

        Examples:
            .. code-block:: python

332 333
                import paddle
                import numpy as np
334

335 336 337
                # the forward_post_hook change the input of the layer: input = input * 2
                def forward_pre_hook(layer, input):
                    # user can use layer and input for information statistis tasks
338

339 340 341
                    # change the input
                    input_return = (input[0] * 2)
                    return input_return
342

343
                linear = paddle.nn.Linear(13, 5)
344

345 346
                # register the hook
                forward_pre_hook_handle = linear.register_forward_pre_hook(forward_pre_hook)
347

348 349 350
                value0 = np.arange(26).reshape(2, 13).astype("float32")
                in0 = paddle.to_tensor(value0)
                out0 = linear(in0)
351

352 353
                # remove the hook
                forward_pre_hook_handle.remove()
354

355 356 357
                value1 = value0 * 2
                in1 = paddle.to_tensor(value1)
                out1 = linear(in1)
358

359 360
                # hook change the linear's input to input * 2, so out0 is equal to out1.
                assert (out0.numpy() == out1.numpy()).any()
361 362 363 364 365
        """
        hook_remove_helper = HookRemoveHelper(self._forward_pre_hooks)
        self._forward_pre_hooks[hook_remove_helper._hook_id] = hook
        return hook_remove_helper

366 367
    def create_parameter(self,
                         shape,
368
                         attr=None,
369
                         dtype=None,
370 371
                         is_bias=False,
                         default_initializer=None):
372 373 374
        """Create parameters for this layer.
        
        Parameters:
375
            shape(list): Shape of the parameter.
376 377
            attr(ParamAttr, optional): Parameter attribute of weight. Please refer to :ref:`api_paddle_ParamAttr`. Default: None.
            dtype(str, optional): Data type of this parameter.
378
                If set str, it can be "bool",  "float16", "float32", "float64",
379 380
                "int8", "int16", "int32", "int64", "uint8" or "uint16". Default: "float32".
            is_bias(bool, optional): if this is a bias parameter. Default: False.
381
            default_initializer(Initializer, optional): the default initializer for this parameter.
382
                If set None, default initializer will be set to paddle.nn.initializer.Xavier and paddle.nn.initializer.Constant
383
                for non-bias and bias parameter, respectively. Default: None.
384

385
        Returns:
386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406
            :Tensor, created parameter.

        Examples:
            .. code-block:: python

                import paddle

                class MyLayer(paddle.nn.Layer):
                    def __init__(self):
                        super(MyLayer, self).__init__()
                        self._linear = paddle.nn.Linear(1, 1)
                        w_tmp = self.create_parameter([1,1])
                        self.add_parameter("w_tmp", w_tmp)

                    def forward(self, input):
                        return self._linear(input)

                mylayer = MyLayer()
                for name, param in mylayer.named_parameters():
                    print(name, param)      # will print w_tmp,_linear.weight,_linear.bias

407
        """
H
hong 已提交
408 409 410 411
        temp_attr = copy.deepcopy(attr)
        if isinstance(temp_attr, six.string_types) and temp_attr == "":
            temp_attr = None
        return self._helper.create_parameter(temp_attr, shape, dtype, is_bias,
412 413
                                             default_initializer)

W
wanghuancoder 已提交
414 415 416 417
    @deprecated(
        since="2.0.0",
        update_to="paddle.nn.Layer.create_tensor",
        reason="New api in create_tensor, easier to use.")
418
    def create_variable(self, name=None, persistable=None, dtype=None):
W
wanghuancoder 已提交
419 420 421
        """

        Create Tensor for this layer.
422

423
        Parameters:
W
wanghuancoder 已提交
424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474
            name(str, optional): name of the tensor. Please refer to :ref:`api_guide_Name` . Default: None

            persistable(bool, optional): if set this tensor persistable. Default: False

            dtype(str, optional): data type of this parameter. If set str, it can be "bool", "float16", "float32", "float64","int8", "int16", "int32", "int64", "uint8" or "uint16". If set None, it will be "float32". Default: None

        Returns:
            Tensor, created Tensor.

        Examples:
            .. code-block:: python

                import paddle

                class MyLinear(paddle.nn.Layer):
                    def __init__(self,
                                in_features,
                                out_features):
                        super(MyLinear, self).__init__()
                        self.linear = paddle.nn.Linear( 10, 10)
                            
                        self.back_var = self.create_variable(name = "linear_tmp_0", dtype=self._dtype)
                    
                    def forward(self, input):
                        out = self.linear(input)
                        paddle.assign( out, self.back_var)
                        
                        return out

        """
        if name is not None:
            var_name = ".".join([self._full_name, name])
        else:
            var_name = unique_name.generate(".".join(
                [self._full_name, "_generated_var"]))

        return self._helper.main_program.current_block().create_var(
            name=var_name,
            persistable=persistable,
            dtype=dtype,
            type=core.VarDesc.VarType.LOD_TENSOR)

    # TODO: Add more parameter list when we need them
    def create_tensor(self, name=None, persistable=None, dtype=None):
        """

        Create Tensor for this layer.

        Parameters:
            name(str, optional): name of the tensor. Please refer to :ref:`api_guide_Name` . Default: None
            persistable(bool, optional): if set this tensor persistable. Default: False
475
            dtype(str, optional): data type of this parameter.
476 477
                If set str, it can be "bool",  "float16", "float32", "float64",
                "int8", "int16", "int32", "int64", "uint8" or "uint16".
478
                If set None, it will be "float32". Default: None
479

480
        Returns:
W
wanghuancoder 已提交
481
            Tensor, created Tensor.
482 483 484 485 486 487 488 489 490 491 492 493 494

        Examples:
            .. code-block:: python

                import paddle

                class MyLinear(paddle.nn.Layer):
                    def __init__(self,
                                in_features,
                                out_features):
                        super(MyLinear, self).__init__()
                        self.linear = paddle.nn.Linear( 10, 10)
                            
W
wanghuancoder 已提交
495
                        self.back_var = self.create_tensor(name = "linear_tmp_0", dtype=self._dtype)
496 497 498 499 500 501 502
                    
                    def forward(self, input):
                        out = self.linear(input)
                        paddle.assign( out, self.back_var)
                        
                        return out

503 504 505 506 507 508 509 510
        """
        if name is not None:
            var_name = ".".join([self._full_name, name])
        else:
            var_name = unique_name.generate(".".join(
                [self._full_name, "_generated_var"]))

        return self._helper.main_program.current_block().create_var(
511 512 513 514
            name=var_name,
            persistable=persistable,
            dtype=dtype,
            type=core.VarDesc.VarType.LOD_TENSOR)
515

X
polish  
Xin Pan 已提交
516
    def parameters(self, include_sublayers=True):
517
        """Returns a list of all Parameters from current layer and its sub-layers.
X
Xin Pan 已提交
518

519 520
        Parameters:
            include_sublayers(bool, optional): Whether include the parameters of sublayers. If True, also include the parameters from sublayers. Default: True
X
Xin Pan 已提交
521

522
        Returns:
523 524 525 526 527 528 529 530 531 532
            list of Tensor : a list of Parameters.

        Examples:
            .. code-block:: python

            import paddle

            linear = paddle.nn.Linear(1,1)
            print(linear.parameters())  # print linear_0.w_0 and linear_0.b_0

X
Xin Pan 已提交
533
        """
534 535 536 537 538
        ret = [
            param
            for _, param in self.named_parameters(
                include_sublayers=include_sublayers)
        ]
X
polish  
Xin Pan 已提交
539
        return ret
X
Xin Pan 已提交
540

541 542 543 544 545 546 547 548 549
    def children(self):
        """Returns an iterator over immediate children layers.

        Yields:
            Layer: a child layer

        Examples:
            .. code-block:: python

550
                import paddle
551

552 553 554 555 556
                linear1 = paddle.nn.Linear(10, 3)
                linear2 = paddle.nn.Linear(3, 10, bias_attr=False)
                model = paddle.nn.Sequential(linear1, linear2)

                layer_list = list(model.children())
557

558
                print(layer_list)   # [<paddle.nn.layer.common.Linear object at 0x7f7b8113f830>, <paddle.nn.layer.common.Linear object at 0x7f7b8113f950>]
559 560 561 562 563 564 565 566 567 568 569 570 571 572 573

        """
        for _, layer in self.named_children():
            yield layer

    def named_children(self):
        """Returns an iterator over immediate children layers, yielding both
        the name of the layer as well as the layer itself.

        Yields:
            (string, Layer): Tuple containing a name and child layer

        Examples:
            .. code-block:: python

574
                import paddle
575

576 577 578 579 580 581 582
                linear1 = paddle.nn.Linear(10, 3)
                linear2 = paddle.nn.Linear(3, 10, bias_attr=False)
                model = paddle.nn.Sequential(linear1, linear2)
                for prefix, layer in model.named_children():
                    print(prefix, layer)
                    # ('0', <paddle.nn.layer.common.Linear object at 0x7fb61ed85830>)
                    # ('1', <paddle.nn.layer.common.Linear object at 0x7fb61ed85950>)
583 584 585 586 587 588 589 590

        """
        memo = set()
        for name, layer in self._sub_layers.items():
            if layer is not None and layer not in memo:
                memo.add(layer)
                yield name, layer

X
Xin Pan 已提交
591 592 593
    def sublayers(self, include_sublayers=True):
        """Returns a list of sub layers.

594 595
        Parameters:
            include_sublayers(bool, optional): Whether return the sublayers of sublayers. If True, also include the sublayers of sublayers. Default: True
X
Xin Pan 已提交
596

597 598
        Returns:
            list of Layer : a list of sub layers.
599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618

        Examples:
            .. code-block:: python

                import paddle

                class MyLayer(paddle.nn.Layer):
                    def __init__(self):
                        super(MyLayer, self).__init__()
                        self._linear = paddle.nn.Linear(1, 1)
                        self._dropout = paddle.nn.Dropout(p=0.5)

                    def forward(self, input):
                        temp = self._linear(input)
                        temp = self._dropout(temp)
                        return temp

                mylayer = MyLayer()
                print(mylayer.sublayers())  # [<paddle.nn.layer.common.Linear object at 0x7f44b58977d0>, <paddle.nn.layer.common.Dropout object at 0x7f44b58978f0>]

X
Xin Pan 已提交
619
        """
620 621 622 623 624
        ret = [
            layer
            for _, layer in self.named_sublayers(
                include_sublayers=include_sublayers)
        ]
X
Xin Pan 已提交
625 626
        return ret

627 628 629 630 631 632 633 634 635 636 637 638 639 640 641
    def named_parameters(self, prefix='', include_sublayers=True):
        """
        Returns an iterator over all parameters in the Layer, yielding tuple of name and parameter.

        Parameters:
            prefix(str, optional): Prefix to prepend to all parameter names. Default: ''.
            include_sublayers(bool, optional): Whether include the parameters of sublayers.
                If True, also include the named parameters from sublayers. Default: True.

        Yields:
            (string, Parameter): Tuple of name and Parameter

        Examples:
            .. code-block:: python

642
                import paddle
643

644 645 646 647 648
                fc1 = paddle.nn.Linear(10, 3)
                fc2 = paddle.nn.Linear(3, 10, bias_attr=False)
                model = paddle.nn.Sequential(fc1, fc2)
                for name, param in model.named_parameters():
                    print(name, param)
649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685

        """
        params_set = set()
        named_sublayers = self.named_sublayers(
            prefix=prefix,
            include_sublayers=include_sublayers,
            include_self=True)
        for layer_prefix, sublayer in named_sublayers:
            params = sublayer._parameters.items()
            for key, param in params:
                if param is None or param in params_set:
                    continue
                params_set.add(param)
                name = layer_prefix + ('.' if layer_prefix else '') + key
                yield name, param

    def named_sublayers(self,
                        prefix='',
                        include_sublayers=True,
                        include_self=False,
                        layers_set=None):
        """
        Returns an iterator over all sublayers in the Layer, yielding tuple of name and sublayer.
        The duplicate sublayer will only be yielded once.

        Parameters:
            prefix(str, optional): Prefix to prepend to all parameter names. Default: ''.
            include_sublayers(bool, optional): Whether include the sublayers. Default: True.
            include_self(bool, optional): Whether include the Layer itself. Default: False.
            layers_set(set, optioanl): The set to record duplicate sublayers. Default: None.

        Yields:
            (string, Layer): Tuple of name and Layer

        Examples:
            .. code-block:: python

686
                import paddle
687

688 689 690 691 692
                fc1 = paddle.nn.Linear(10, 3)
                fc2 = paddle.nn.Linear(3, 10, bias_attr=False)
                model = paddle.nn.Sequential(fc1, fc2)
                for prefix, layer in model.named_sublayers():
                    print(prefix, layer)
693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711

        """
        if layers_set is None:
            layers_set = set()
        if include_self and self not in layers_set:
            layers_set.add(self)
            yield prefix, self
        if include_sublayers:
            for key, layer in self._sub_layers.items():
                if layer is None:
                    continue
                layer_prefix = prefix + ('.' if prefix else '') + key
                for p, l in layer.named_sublayers(
                        prefix=layer_prefix,
                        include_sublayers=include_sublayers,
                        include_self=True,
                        layers_set=layers_set):
                    yield p, l

712
    def register_buffer(self, name, tensor, persistable=True):
713
        """
714
        Registers a tensor as buffer into the layer.
715

716
        `buffer` is a non-trainable tensor and will not be updated by optimizer,
717 718 719 720 721 722 723 724 725 726
        but is necessary for evaluation and inference. For example, the mean and variance in BatchNorm layers.
        The registered buffer is persistable by default, and will be saved into
        `state_dict` alongside parameters. If set persistable=False, it registers
        a non-persistable buffer, so that it will not be a part of `state_dict` .

        Buffers can be accessed as attributes using given names.

        Parameters:
            name (string): name of the buffer. The buffer can be accessed
                from this layer using the given name
727
            tensor (Tensor): the tensor to be registered as buffer.
728 729 730 731 732 733 734 735 736 737
            persistable (bool): whether the buffer is part of this layer's
                state_dict.

        Returns:
            None
        
        Examples:
            .. code-block:: python

                import numpy as np
738
                import paddle
739

740 741 742 743 744 745 746
                linear = paddle.nn.Linear(10, 3)
                value = np.array([0]).astype("float32")
                buffer = paddle.to_tensor(value)
                linear.register_buffer("buf_name", buffer, persistable=True)

                # get the buffer by attribute.
                print(linear.buf_name)
747 748 749 750 751 752 753 754 755 756 757

        """

        if '_buffers' not in self.__dict__:
            raise ValueError(
                "super(YourLayer, self).__init__() should be called first")
        elif not isinstance(name, six.string_types):
            raise TypeError(
                "The name of buffer should be a string, but received {}.".
                format(type(name).__name__))
        elif '.' in name:
758 759 760 761
            raise KeyError(
                "The name of buffer can not contain `.`, "
                "because when you access the newly added buffer in the "
                "form of `self.**.**`, it will cause AttributeError.")
762 763 764 765
        elif name == '':
            raise KeyError("The name of buffer can not be empty.")
        elif hasattr(self, name) and name not in self._buffers:
            raise KeyError("attribute '{}' already exists.".format(name))
766
        elif tensor is not None and not type(tensor) == core.VarBase:
767 768
            raise TypeError(
                "The registered buffer should be a core.VarBase, but received {}.".
769
                format(type(tensor).__name__))
770
        else:
771
            self._buffers[name] = tensor
772 773 774 775 776 777 778 779 780 781 782 783 784
            if persistable:
                self._non_persistable_buffer_names_set.discard(name)
            else:
                self._non_persistable_buffer_names_set.add(name)

    def buffers(self, include_sublayers=True):
        """
        Returns a list of all buffers from current layer and its sub-layers.

        Parameters:
            include_sublayers(bool, optional): Whether include the buffers of sublayers. If True, also include the buffers from sublayers. Default: True

        Returns:
785 786 787 788 789 790 791 792 793 794 795 796 797 798 799
            list of Tensor : a list of buffers.

        Examples:
            .. code-block:: python

                import numpy as np
                import paddle

                linear = paddle.nn.Linear(10, 3)
                value = np.array([0]).astype("float32")
                buffer = paddle.to_tensor(value)
                linear.register_buffer("buf_name", buffer, persistable=True)

                print(linear.buffers())     # == print([linear.buf_name])

800 801 802 803 804 805 806 807 808 809
        """
        ret = [
            buffer
            for _, buffer in self.named_buffers(
                include_sublayers=include_sublayers)
        ]
        return ret

    def named_buffers(self, prefix='', include_sublayers=True):
        """
810
        Returns an iterator over all buffers in the Layer, yielding tuple of name and Tensor.
811 812 813 814 815 816 817

        Parameters:
            prefix(str, optional): Prefix to prepend to all buffer names. Default: ''.
            include_sublayers(bool, optional): Whether include the buffers of sublayers.
                If True, also include the named buffers from sublayers. Default: True.

        Yields:
818
            (string, Tensor): Tuple of name and tensor
819 820 821 822 823

        Examples:
            .. code-block:: python

                import numpy as np
824
                import paddle
825

826 827 828 829
                fc1 = paddle.nn.Linear(10, 3)
                buffer1 = paddle.to_tensor(np.array([0]).astype("float32"))
                # register a tensor as buffer by specific `persistable`
                fc1.register_buffer("buf_name_1", buffer1, persistable=True)
830

831 832 833 834 835
                fc2 = paddle.nn.Linear(3, 10)
                buffer2 = paddle.to_tensor(np.array([1]).astype("float32"))
                # register a buffer by assigning an attribute with Tensor.
                # The `persistable` can only be False by this way.
                fc2.buf_name_2 = buffer2
836

837
                model = paddle.nn.Sequential(fc1, fc2)
838

839 840 841
                # get all named buffers
                for name, buffer in model.named_buffers():
                    print(name, buffer)
842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857

        """
        buffers_set = set()
        named_sublayers = self.named_sublayers(
            prefix=prefix,
            include_sublayers=include_sublayers,
            include_self=True)
        for layer_prefix, sublayer in named_sublayers:
            buffers = sublayer._buffers.items()
            for key, buffer in buffers:
                if buffer is None or buffer in buffers_set:
                    continue
                buffers_set.add(buffer)
                name = layer_prefix + ('.' if layer_prefix else '') + key
                yield name, buffer

X
Xin Pan 已提交
858
    def clear_gradients(self):
859 860 861 862 863 864 865 866 867
        """
        Clear the gradients of all parameters for this layer.
        
        Returns:
            None
        
        Examples:
            .. code-block:: python

868
                import paddle
869 870
                import numpy as np

871 872 873 874 875 876 877 878 879
                value = np.arange(26).reshape(2, 13).astype("float32")
                a = paddle.to_tensor(value)
                linear = paddle.nn.Linear(13, 5)
                adam = paddle.optimizer.Adam(learning_rate=0.01,
                                            parameters=linear.parameters())
                out = linear(a)
                out.backward()
                adam.step()
                linear.clear_gradients()
880 881

        """
X
Xin Pan 已提交
882
        for p in self.parameters():
883 884
            if p.trainable:
                p.clear_gradient()
X
Xin Pan 已提交
885

886
    def _build_once(self, *args, **kwargs):
887 888
        pass

889
    def __call__(self, *inputs, **kwargs):
890
        with param_guard(self._parameters), param_guard(self._buffers):
891 892 893 894 895 896 897 898 899 900
            for forward_pre_hook in self._forward_pre_hooks.values():
                hook_result = forward_pre_hook(self, inputs)
                if hook_result is not None:
                    if not isinstance(hook_result, tuple):
                        hook_result = (hook_result, )
                    inputs = hook_result

            if not self._built:
                with program_desc_tracing_guard(False):
                    self._build_once(*inputs, **kwargs)
901 902 903 904 905 906

                    # TODO(liuyuhui) Only xpu broadcast parameters here. 
                    # The other device is to call _sync_params_buffers in DataParallel 
                    # to realize the parameter synchronization among multiply cards.
                    if parallel_helper._is_data_parallel_mode(
                    ) and paddle.is_compiled_with_xpu():
907 908
                        parallel_helper._broadcast_parameters(
                            self._parameters.values())
909

910 911
                self._built = True

912
            outputs = self.forward(*inputs, **kwargs)
913

914 915 916 917
            for forward_post_hook in self._forward_post_hooks.values():
                hook_result = forward_post_hook(self, inputs, outputs)
                if hook_result is not None:
                    outputs = hook_result
918

919
            return outputs
M
minqiyang 已提交
920

921
    def forward(self, *inputs, **kwargs):
922 923 924 925 926 927 928 929
        """
        Defines the computation performed at every call.
        Should be overridden by all subclasses.

        Parameters:
            *inputs(tuple): unpacked tuple arguments
            **kwargs(dict): unpacked dict arguments
        """
930
        raise NotImplementedError
X
Xin Pan 已提交
931 932 933 934

    def backward(self, *inputs):
        raise ValueError("Layer shouldn't implement backward")

X
Xin Pan 已提交
935 936 937
    def add_sublayer(self, name, sublayer):
        """Adds a sub Layer instance.

938
        Added sublayer can be accessed by self.name
X
Xin Pan 已提交
939

940 941 942
        Parameters:
            name(str): name of this sublayer.
            sublayer(Layer): an instance of Layer.
X
Xin Pan 已提交
943
        Returns:
944
            Layer: the sublayer passed in.
945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970
        
        Examples:
            .. code-block:: python

                import paddle

                class MySequential(paddle.nn.Layer):
                    def __init__(self, *layers):
                        super(MySequential, self).__init__()
                        if len(layers) > 0 and isinstance(layers[0], tuple):
                            for name, layer in layers:
                                self.add_sublayer(name, layer)
                        else:
                            for idx, layer in enumerate(layers):
                                self.add_sublayer(str(idx), layer)

                    def forward(self, input):
                        for layer in self._sub_layers.values():
                            input = layer(input)
                        return input

                fc1 = paddle.nn.Linear(10, 3)
                fc2 = paddle.nn.Linear(3, 10, bias_attr=False)
                model = MySequential(fc1, fc2)
                for prefix, layer in model.named_sublayers():
                    print(prefix, layer)
X
Xin Pan 已提交
971
        """
972
        assert (isinstance(sublayer, core.Layer) or sublayer == None)
973

X
Xin Pan 已提交
974 975 976 977 978 979
        self._sub_layers[name] = sublayer
        return sublayer

    def add_parameter(self, name, parameter):
        """Adds a Parameter instance.

980
        Added parameter can be accessed by self.name
X
Xin Pan 已提交
981

982 983 984
        Parameters:
            name(str): name of this sublayer.
            parameter(Parameter): an instance of Parameter.
X
Xin Pan 已提交
985
        Returns:
986
            Parameter: the parameter passed in.
987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005
        Examples:
            .. code-block:: python

                import paddle

                class MyLayer(paddle.nn.Layer):
                    def __init__(self):
                        super(MyLayer, self).__init__()
                        self._linear = paddle.nn.Linear(1, 1)
                        w_tmp = self.create_parameter([1,1])
                        self.add_parameter("w_tmp", w_tmp)

                    def forward(self, input):
                        return self._linear(input)

                mylayer = MyLayer()
                for name, param in mylayer.named_parameters():
                    print(name, param)      # will print w_tmp,_linear.weight,_linear.bias

X
Xin Pan 已提交
1006
        """
1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
        if '_parameters' not in self.__dict__:
            raise RuntimeError(
                "super(YourLayer, self).__init__() should be called firstly.")
        elif not isinstance(name, six.string_types):
            raise TypeError(
                "The name of parameter should be a string, but received {}.".
                format(type(name).__name__))
        elif '.' in name:
            raise KeyError(
                "The name of parameter can not contain `.`, "
                "because when you access the newly added parameter in the "
                "form of `self.**.**`, it will cause AttributeError.")
        elif name == '':
            raise KeyError("The name of parameter can not be empty.")
        elif hasattr(self, name) and name not in self._parameters:
            raise KeyError("The parameter '{}' already exists.".format(name))
        elif parameter is not None and not isinstance(parameter,
                                                      framework.Parameter):
1025
            raise TypeError(
1026 1027 1028 1029 1030
                "The parameter to be added should be a Parameter, but received {}.".
                format(type(parameter).__name__))
        else:
            if parameter is None:
                self._parameters[name] = None
1031

1032 1033 1034
            if len(self._loaddict_holder) > 0:
                assert parameter.name in self._loaddict_holder, "Parameter not found, Can't not find [ {} ] in state_dict".format(
                    parameter.name)
H
hong 已提交
1035

1036
                parameter.set_value(self._loaddict_holder[parameter.name])
1037

1038
            self._parameters[name] = parameter
X
Xin Pan 已提交
1039 1040
        return parameter

1041 1042 1043 1044 1045 1046
    def __getstate__(self):
        return self.__dict__

    def __setstate__(self, state):
        self.__dict__.update(state)

X
Xin Pan 已提交
1047
    def __getattr__(self, name):
1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060
        if '_parameters' in self.__dict__:
            _parameters = self.__dict__['_parameters']
            if name in self._parameters:
                return self._parameters[name]
        if '_sub_layers' in self.__dict__:
            _sub_layers = self.__dict__['_sub_layers']
            if name in self._sub_layers:
                return self._sub_layers[name]
        if '_buffers' in self.__dict__:
            _buffers = self.__dict__['_buffers']
            if name in _buffers:
                return _buffers[name]
        return object.__getattribute__(self, name)
X
Xin Pan 已提交
1061 1062

    def __setattr__(self, name, value):
S
songyouwei 已提交
1063 1064 1065 1066 1067
        def _remove_if_exist(*dicts):
            for d in dicts:
                if name in d:
                    del d[name]

1068 1069
        if isinstance(getattr(type(self), name, None), property):
            object.__setattr__(self, name, value)
1070
        params = self.__dict__.get('_parameters', None)
X
Xin Pan 已提交
1071 1072 1073 1074
        if isinstance(value, framework.Parameter):
            if params is None:
                raise ValueError(
                    "super(YourLayer, self).__init__() should be called first")
H
hong 已提交
1075
            if len(self._loaddict_holder) > 0:
1076
                assert value.name in self._loaddict_holder, "Parameter not found, Can't not find [ {} ] in state_dict".format(
H
hong 已提交
1077 1078 1079 1080
                    value.name)

                value.set_value(self._loaddict_holder[value.name])

1081
            _remove_if_exist(self.__dict__, self._buffers, self._sub_layers)
1082
            params[name] = value
1083 1084 1085 1086 1087 1088
        elif params is not None and name in params:
            if value is not None:
                raise TypeError(
                    "assignment to parameter '{}' should be of type Parameter or None, but got '{}'"
                    .format(name, type(value).__name__))
            params[name] = None
X
Xin Pan 已提交
1089
        else:
1090 1091 1092 1093 1094 1095 1096
            layers = self.__dict__.get('_sub_layers', None)
            if isinstance(value, core.Layer):
                if layers is None:
                    raise ValueError(
                        "super(YourLayer, self).__init__() should be called first"
                    )

1097
                _remove_if_exist(self.__dict__, self._parameters, self._buffers)
1098 1099 1100 1101 1102 1103 1104 1105
                layers[name] = value
            elif layers is not None and name in layers:
                if value is not None:
                    raise TypeError(
                        "assignment to sublayer '{}' should be of type Layer or None, but got '{}'"
                        .format(name, type(value).__name__))
                layers[name] = None
            else:
1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119
                _buffers = self.__dict__.get('_buffers', None)
                if type(value) == core.VarBase:
                    if _buffers is None:
                        raise ValueError(
                            "super(YourLayer, self).__init__() should be called first"
                        )
                    _remove_if_exist(self.__dict__, self._parameters,
                                     self._sub_layers)
                    # Set persistable=False by default. Only `register_buffer` can
                    # add a persistable buffer.
                    if name not in self._buffers:
                        self._non_persistable_buffer_names_set.add(name)
                    _buffers[name] = value
                elif _buffers is not None and name in _buffers:
1120 1121 1122 1123 1124
                    # Note(Aurelius84): In Dy2stat, the value of the Buffer may be modified in 
                    # decorated function, such as `self.buffer = new_tensor`. So we update its
                    # value via `assign`.
                    if type(value) == framework.Variable:
                        from paddle import assign
1125 1126 1127 1128 1129 1130 1131 1132 1133
                        # Note(zhhsplendid): the condition below happens in PaddleGan model,
                        # but should all non-Variable _buffers[name] be re-assign? We
                        # should consider it in the future. I current wrote this as
                        # conservative code.
                        if _buffers[name] is None or type(_buffers[
                                name]) == core.VarBase:
                            _buffers[name] = assign(value)
                        else:
                            assign(value, _buffers[name])
1134
                    elif value is not None:
1135 1136 1137
                        raise TypeError(
                            "assignment to buffers '{}' should be of type core.VarBase or None, but got '{}'"
                            .format(name, type(value).__name__))
1138 1139 1140 1141
                    else:
                        # Assigning None will remove the buffer, but if re-assign a new varBase to it,
                        # it will be remarked as a buffer with same `persistable` attribute.
                        _buffers[name] = None
1142 1143
                else:
                    object.__setattr__(self, name, value)
X
Xin Pan 已提交
1144 1145 1146 1147 1148 1149

    def __delattr__(self, name):
        if name in self._parameters:
            del self._parameters[name]
        elif name in self._sub_layers:
            del self._sub_layers[name]
1150 1151 1152
        elif name in self._buffers:
            del self._buffers[name]
            self._non_persistable_buffer_names_set.discard(name)
X
Xin Pan 已提交
1153 1154 1155
        else:
            object.__delattr__(self, name)

1156 1157
    def __dir__(self):
        """
W
wanghuancoder 已提交
1158
        Return a list. Get all parameters, buffers(non-parameter tensors), sublayers, method and attr of Layer.
1159 1160

        Examples:
1161 1162 1163
            .. code-block:: python
                import paddle
                import numpy as np
1164

1165 1166 1167 1168 1169
                class Mylayer(paddle.nn.Layer):
                    def __init__(self):
                        super(Mylayer, self).__init__()
                        self.linear1 = paddle.nn.Linear(10, 10)
                        self.linear2 = paddle.nn.Linear(5, 5)
C
cnn 已提交
1170
                        self.conv2d = paddle.nn.Conv2D(3, 2, 3)
1171 1172
                        self.embedding = paddle.nn.Embedding(128, 16)
                        self.h_0 = paddle.to_tensor(np.zeros([10, 10]).astype('float32'))
1173

1174 1175 1176 1177
                mylayer = Mylayer()
                print(dir(mylayer))
                # only parts are shown, because of list have too much content
                # ['__call__', '__class__',  ... , 'conv2d', 'embedding', 'h_0', 'linear1', 'linear2', ... , 'sublayers', 'train']
1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189

        """
        method = dir(self.__class__)
        attrs = list(self.__dict__.keys())
        parameters = list(self._parameters.keys())
        sublayers = list(self._sub_layers.keys())
        buffers = list(self._buffers.keys())

        keys = method + attrs + parameters + sublayers + buffers

        return keys

1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218
    def extra_repr(self):
        """
        Extra representation of this layer, you can have custom implementation
        of your own layer.
        """
        return ''

    def __repr__(self):
        extra_lines = []
        extra_repr = self.extra_repr()
        extra_lines = extra_repr.split('\n')
        sublayer_lines = []
        for name, layer in self._sub_layers.items():
            sublayer_str = repr(layer)
            sublayer_str = _addindent(sublayer_str, 2)
            sublayer_lines.append('(' + name + '): ' + sublayer_str)

        final_str = self.__class__.__name__ + '('
        if extra_lines:
            if len(extra_lines) > 1:
                final_str += '\n  ' + '\n  '.join(extra_lines) + '\n'
            elif len(extra_lines) == 1:
                final_str += extra_lines[0]
        if sublayer_lines:
            final_str += '\n  ' + '\n  '.join(sublayer_lines) + '\n'

        final_str += ')'
        return final_str

H
hong 已提交
1219 1220 1221 1222
    def state_dict(self,
                   destination=None,
                   include_sublayers=True,
                   structured_name_prefix=""):
H
hong 已提交
1223
        '''
1224
        Get all parameters and persistable buffers of current layer and its sub-layers. And set them into a dict
H
hong 已提交
1225

1226
        Parameters:
1227 1228
            destination(dict, optional) : If provide, all the parameters and persistable buffers will be set to this dict . Default: None
            include_sublayers(bool, optional) : If true, also include the parameters and persistable buffers from sublayers. Default: True
H
hong 已提交
1229 1230

        Retruns:
1231
            dict: a dict contains all the parameters and persistable buffers.
H
hong 已提交
1232 1233

        Examples:
1234 1235
            .. code-block:: python

1236
                import paddle
H
hong 已提交
1237

1238 1239 1240 1241
                emb = paddle.nn.Embedding(10, 10)

                state_dict = emb.state_dict()
                paddle.save( state_dict, "paddle_dy.pdparams")
H
hong 已提交
1242 1243 1244

        '''

1245 1246 1247 1248
        if destination is None:
            destination = collections.OrderedDict()
        for name, data in self._parameters.items():
            if data is not None:
H
hong 已提交
1249
                destination[structured_name_prefix + name] = data
1250 1251 1252
        for name, buffer in self._buffers.items():
            if buffer is not None and name not in self._non_persistable_buffer_names_set:
                destination[structured_name_prefix + name] = buffer
1253 1254 1255 1256 1257 1258

        if include_sublayers:
            for layer_name, layer_item in self._sub_layers.items():
                if layer_item is not None:
                    destination_temp = destination.copy()
                    destination_temp.update(
H
hong 已提交
1259 1260 1261
                        layer_item.state_dict(
                            destination_temp, include_sublayers,
                            structured_name_prefix + layer_name + "."))
1262 1263 1264
                    destination = destination_temp
        return destination

1265 1266 1267 1268 1269
    @framework.deprecate_stat_dict
    def set_state_dict(self,
                       state_dict,
                       include_sublayers=True,
                       use_structured_name=True):
H
hong 已提交
1270
        '''
1271
        Set parameters and persistable buffers from state_dict. All the parameters and buffers will be reset by the tensor in the state_dict
H
hong 已提交
1272

1273
        Parameters:
1274 1275 1276
            state_dict(dict) : Dict contains all the parameters and persistable buffers.
            include_sublayers(bool, optional) : If true, also include the parameters and peresistable buffers from sublayers. Default: True
            use_structured_name(bool, optional) : If true, use structured name as key, otherwise, use parameter or buffer name as key. 
H
hong 已提交
1277
                                                  Default: True
H
hong 已提交
1278 1279 1280 1281
        Returns:
            None

        Examples:
1282 1283
            .. code-block:: python

1284
                import paddle
1285

1286
                emb = paddle.nn.Embedding(10, 10)
H
hong 已提交
1287

1288
                state_dict = emb.state_dict()
1289 1290
                paddle.save(state_dict, "paddle_dy.pdparams")
                para_state_dict = paddle.load("paddle_dy.pdparams")
1291
                emb.set_state_dict(para_state_dict)
H
hong 已提交
1292

H
hong 已提交
1293 1294
        '''

1295 1296 1297 1298 1299
        def _check_match(key, param):
            state = state_dict.get(key, None)
            if state is None:
                raise ValueError("{} is not found in the provided dict.".format(
                    key))
1300 1301 1302
            state_shape = state.shape() if inspect.ismethod(
                state.shape) else state.shape
            if list(state_shape) != list(param.shape):
1303 1304
                raise ValueError(
                    "{} receives a shape {}, but the expected shape is {}.".
1305
                    format(key, list(state_shape), list(param.shape)))
1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320
            return param, state

        matched_param_state = []
        for key, param in self.state_dict().items():
            key_name = key if use_structured_name else param.name
            try:
                match_res = _check_match(key_name, param)
                matched_param_state.append(match_res)
            except ValueError as err:
                warnings.warn(("Skip loading for {}. ".format(key) + str(err)))

        if in_dygraph_mode():
            for param, state in matched_param_state:
                param.set_value(state)
        else:
H
hong 已提交
1321

1322 1323 1324 1325 1326 1327 1328
            def _set_var(var, ndarray):
                t = global_scope().find_var(var.name).get_tensor()
                p = t._place()
                if p.is_cpu_place():
                    place = core.CPUPlace()
                elif p.is_cuda_pinned_place():
                    place = core.CUDAPinnedPlace()
1329 1330 1331 1332
                elif p.is_xpu_place():
                    p = core.Place()
                    p.set_place(t._place())
                    place = core.XPUPlace(p.xpu_device_id())
1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346
                else:
                    p = core.Place()
                    p.set_place(t._place())
                    place = core.CUDAPlace(p.gpu_device_id())
                t.set(ndarray, place)

            executor = Executor(_get_device())._default_executor
            # restore parameter states
            core._create_loaded_parameter(
                [param for param, state in matched_param_state],
                global_scope(), executor)
            for param, state in matched_param_state:
                _set_var(param, state)

C
chentianyu03 已提交
1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454
    def _apply(self, func, device, dtype, blocking):
        for layer in self.children():
            layer._apply(func, device, dtype, blocking)

        for key, param in self._parameters.items():
            if param is not None:
                with no_grad():
                    param_applied = func(param, device, dtype, blocking)
                    assert param.is_leaf
                    param_applied.stop_gradient = param.stop_gradient
                    self._parameters[key] = param_applied

                if param.grad is not None:
                    with no_grad():
                        grad_applied = func(param._grad_ivar(), device, dtype,
                                            blocking)

                        grad_applied.stop_gradient = param._grad_ivar(
                        ).stop_gradient
                        self._parameters[key]._set_grad_ivar(grad_applied)

        for key, buf in self._buffers.items():
            self._buffers[key] = func(buf, device, dtype, blocking)

    def to(self, device=None, dtype=None, blocking=None):
        '''
        Cast the parameters and buffers of Layer by the give device, dtype and blocking.

        Parameters:
            device(str|paddle.CPUPlace()|paddle.CUDAPlace()|paddle.CUDAPinnedPlace()|paddle.XPUPlace()|None, optional): The device of the Layer which want to be stored. 
            If None, the device is the same with the original Tensor. If device is string, it can be ``cpu``, ``gpu:x`` and ``xpu:x``, where ``x`` is the 
            index of the GPUs or XPUs. Default: None. 
            
            dtype(str|core.VarDesc.VarType|None, optional): The type of the data. If None, the dtype is the same with the original Tensor. Default: None.

            blocking(bool|None, optional): If False and the source is in pinned memory, the copy will be 
              asynchronous with respect to the host. Otherwise, the argument has no effect. If None, the blocking is set True. Default: None.
            
        Returns:
            None

        Examples:
            .. code-block:: python

                import paddle

                linear=paddle.nn.Linear(2, 2)
                linear.weight
                #Parameter containing:
                #Tensor(shape=[2, 2], dtype=float32, place=CUDAPlace(0), stop_gradient=False,
                #       [[-0.32770029,  0.38653070],
                #        [ 0.46030545,  0.08158520]])

                linear.to(dtype='float64')
                linear.weight
                #Tenor(shape=[2, 2], dtype=float64, place=CUDAPlace(0), stop_gradient=False,
                #       [[-0.32770029,  0.38653070],
                #        [ 0.46030545,  0.08158520]])

                linear.to(device='cpu')
                linear.weight
                #Tensor(shape=[2, 2], dtype=float64, place=CPUPlace, stop_gradient=False,
                #       [[-0.32770029,  0.38653070],
                #        [ 0.46030545,  0.08158520]])
                linear.to(device=paddle.CUDAPinnedPlace(), blocking=False)
                linear.weight
                #Tensor(shape=[2, 2], dtype=float64, place=CUDAPinnedPlace, stop_gradient=False,
                #       [[-0.04989364, -0.56889004],
                #        [ 0.33960250,  0.96878713]])
    

        '''

        if device is None and dtype is None and blocking is None:
            return

        if device is not None:
            if isinstance(device, str):
                device = paddle.device._convert_to_place(device)
            elif isinstance(device, (core.CPUPlace, core.CUDAPlace,
                                     core.CUDAPinnedPlace, core.XPUPlace)):
                pass
            else:
                raise ValueError(
                    "device value error, must be str, paddle.CPUPlace(), paddle.CUDAPlace(), paddle.CUDAPinnedPlace() or paddle.XPUPlace(), but the type of device is "
                    + type(device).__name__)

        if blocking is None:
            blocking = True
        else:
            assert isinstance(
                blocking,
                bool), "blocking value error, must be the True, False or None"

        def transform(t, device, dtype, blocking):
            if device is None:
                device = t.place
            if dtype is None:
                dtype = t.dtype

            new_t = t._copy_to(device, blocking)
            if dtype is not None and dtype != t.dtype:
                new_t = new_t.cast(dtype=dtype)

            return new_t

        self._apply(transform, device, dtype, blocking)

1455 1456 1457
    # [aliases] Compatible with old method names
    set_dict = set_state_dict
    load_dict = set_state_dict