base.py 25.5 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13
# 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.
14
from ..wrapped_decorator import signature_safe_contextmanager, wrap_decorator
S
songyouwei 已提交
15
import decorator
16
import contextlib
17 18
import functools
import inspect
19
import sys
20 21 22
import numpy as np
from paddle.fluid import core
from paddle.fluid import framework
H
hong 已提交
23
from paddle.fluid.multiprocess_utils import CleanupFuncRegistrar
M
minqiyang 已提交
24
from .tracer import Tracer
Z
Zeng Jinle 已提交
25
import logging
26
from ..data_feeder import convert_dtype
L
Leo Chen 已提交
27
import warnings
28
from ..framework import _get_paddle_place
29
import paddle
30

31
__all__ = [
32 33
    'no_grad', 'no_grad_', 'grad', 'guard', 'enable_dygraph', 'disable_dygraph',
    'enabled', 'to_variable'
34
]
35

36 37 38 39 40 41 42 43 44 45 46
# Flag that indicates whether running code under `@declarative`
_in_declarative_mode_ = False


def in_declarative_mode():
    """
    Return a bool value that indicates whether running code under `@declarative`

    """
    return _in_declarative_mode_

47

48 49 50 51 52 53 54 55 56 57 58
def _switch_to_static_graph_(func):
    def __impl__(*args, **kwargs):
        with framework._dygraph_guard(None):
            return func(*args, **kwargs)

    return __impl__


switch_to_static_graph = wrap_decorator(_switch_to_static_graph_)


59 60 61 62 63 64 65 66 67 68
@signature_safe_contextmanager
def _switch_declarative_mode_guard_(is_declarative=True):

    global _in_declarative_mode_
    original_val = _in_declarative_mode_
    _in_declarative_mode_ = is_declarative
    yield
    _in_declarative_mode_ = original_val


69 70 71 72 73 74
@signature_safe_contextmanager
def program_desc_tracing_guard(enable):
    tracer = framework._dygraph_tracer()
    if tracer:
        original_val = tracer._enable_program_desc_tracing
        tracer._enable_program_desc_tracing = enable
75 76 77 78 79
    try:
        yield
    finally:
        if tracer:
            tracer._enable_program_desc_tracing = original_val
80 81


82 83 84
_functional_dygraph_context_manager = None


85 86
@signature_safe_contextmanager
def param_guard(parameters):
87
    # Note: parameters is a reference of self._parameters or self._buffers
88
    if in_declarative_mode() and not framework.in_dygraph_mode() and parameters:
89 90
        origin_parameters = parameters.copy()
        for name, var_base in parameters.items():
91 92 93 94 95
            if isinstance(var_base, list):
                new_var = [_convert_into_variable(var) for var in var_base]
            else:
                new_var = _convert_into_variable(var_base)
            parameters[name] = new_var
96 97 98 99 100 101
        yield
        parameters.update(origin_parameters)
    else:
        yield


102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131
def _convert_into_variable(var_base):
    """
    Convert Varbase into Variable.
    """
    if isinstance(var_base, core.VarBase):
        # Check whether has been created before.
        new_var = var_base.block._find_var_recursive(var_base.name)
        if new_var is not None:
            assert isinstance(new_var, framework.Variable)
        # Convert ParamBase into Parameter with same attributes in dy2stat.
        elif isinstance(var_base, framework.ParamBase):
            new_var = var_base._to_static_var(to_parameter=True)
        else:
            # Note(Aurelius84): Convert VarBase in self._buffers into Variable with
            # same attributes and set persistable=True to allow saving this var.
            # Because users can create a VarBase in `__init__`  like a
            # `mask` Tensor or `hidden_0` in RNN layers, which is equivalent to a Parameter
            # and necessary for inferring. It will be pruned if it's not necessary for inferring.

            # But if its shape is empty while created from `create_variable()`, we consider this buffer
            # non-persistable. See case of `drop_state` in lstm api.
            is_persistable = len(var_base.shape) > 0

            new_var = var_base._to_static_var(
                to_parameter=False, persistable=is_persistable)
        return new_var
    else:
        return var_base


132
def enabled():
133 134 135
    """
    This function checks whether the program runs in dynamic graph mode or not.
    You can enter dynamic graph mode with :ref:`api_fluid_dygraph_guard` api,
136 137
    or enable and disable dynamic graph mode with :ref:`api_fluid_dygraph_enable_dygraph`
    and :ref:`api_fluid_dygraph_disable_dygraph` api .
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155

    **Note**:
        ``fluid.dygraph.enabled`` is the alias of ``fluid.in_dygraph_mode``, and
        ``fluid.in_dygraph_mode`` is recommended to use.

    Returns:
        bool: Whether the program is running in dynamic graph mode.

    Examples:
        .. code-block:: python

            import paddle.fluid as fluid

            fluid.enable_dygraph()  # Now we are in dygragh mode
            print(fluid.dygraph.enabled())  # True
            fluid.disable_dygraph()
            print(fluid.dygraph.enabled())  # False
    """
L
lujun 已提交
156
    return framework.in_dygraph_mode()
157 158


159 160
def enable_dygraph(place=None):
    """
161 162 163 164 165

    .. note::
        Dynamic graph mode is turn ON by default since paddle 2.0.0

    This API turn OFF static graph mode. You can turn ON static graph mode by `enable_static <./disable_dygraph_en.html>`_ .
166 167

    Parameters:
168 169 170
        place(paddle.CPUPlace|paddle.CUDAPlace|str, optional): Place to run dynamic graph. Default: None. Which means that the running place will be 
            determined according to the way of paddle compilation. If ``place`` is string, It can be ``cpu``, and ``gpu:x``, where ``x`` is the
            index of the GPUs.
171 172 173 174 175 176 177

    return:
        None

    Examples:
        .. code-block:: python

178 179 180 181 182 183 184 185
            import paddle
            print(paddle.in_dynamic_mode())  # True, dynamic mode is turn ON by default since paddle 2.0.0

            paddle.enable_static()
            print(paddle.in_dynamic_mode())  # False, Now we are in static mode

            paddle.disable_static()
            print(paddle.in_dynamic_mode())  # True, Now we are in dynamic mode
186 187 188

    """
    global _functional_dygraph_context_manager
S
songyouwei 已提交
189
    if _functional_dygraph_context_manager is None:
190 191
        _functional_dygraph_context_manager = guard(
            place=_get_paddle_place(place))
S
songyouwei 已提交
192
        _functional_dygraph_context_manager.__enter__()
193

H
hong 已提交
194 195 196
        # call disable_dygraph when Python exit
        CleanupFuncRegistrar.register(disable_dygraph)

197 198 199

def disable_dygraph():
    """
200 201 202 203 204

    .. note::
        Dynamic graph mode is turn ON by default since paddle 2.0.0

    This API turn ON static graph mode. You can turn ON static graph mode by `disable_static <./enable_dygraph_en.html>`_ .
205 206 207 208 209 210 211

    return:
        None

    Examples:
        .. code-block:: python

212 213 214 215 216 217 218 219
            import paddle
            print(paddle.in_dynamic_mode())  # True, dynamic mode is turn ON by default since paddle 2.0.0

            paddle.enable_static()
            print(paddle.in_dynamic_mode())  # False, Now we are in static mode

            paddle.disable_static()
            print(paddle.in_dynamic_mode())  # True, Now we are in dynamic mode
220 221 222 223 224 225 226 227

    """
    global _functional_dygraph_context_manager
    if _functional_dygraph_context_manager is not None:
        _functional_dygraph_context_manager.__exit__(*sys.exc_info())
        _functional_dygraph_context_manager = None


228 229 230 231
@signature_safe_contextmanager
def _switch_tracer_mode_guard_(is_train=True):
    tracer = framework._dygraph_tracer()
    if tracer:
232 233
        has_grad = tracer._has_grad
        tracer._has_grad = is_train
234 235 236
        try:
            yield
        finally:
237
            tracer._has_grad = has_grad
238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301
    else:
        yield


def no_grad(func=None):
    """
    :api_attr: imperative

    Create a context which disables dygraph gradient calculation.
    In this mode, the result of every computation will have `stop_gradient=True`.

    Also functions as a decorator. (Make sure to instantiate without parenthesis.)

    Examples:

     .. code-block:: python

        import numpy as np
        import paddle.fluid as fluid

        # use as generator

        data = np.array([[2, 3], [4, 5]]).astype('float32')
        with fluid.dygraph.guard():
            l0 = fluid.Linear(2, 2)  # l0.weight.gradient() is None
            l1 = fluid.Linear(2, 2)
            with fluid.dygraph.no_grad():
                # l1.weight.stop_gradient is False
                tmp = l1.weight * 2  # tmp.stop_gradient is True
            x = fluid.dygraph.to_variable(data)
            y = l0(x) + tmp
            o = l1(y)
            o.backward()
            print(tmp.gradient() is None)  # True
            print(l0.weight.gradient() is None)  # False

        # use as decorator

        @fluid.dygraph.no_grad
        def test_layer():
            with fluid.dygraph.guard():
                inp = np.ones([3, 1024], dtype='float32')
                t = fluid.dygraph.base.to_variable(inp)
                linear1 = fluid.Linear(1024, 4, bias_attr=False)
                linear2 = fluid.Linear(4, 4)
                ret = linear1(t)
                dy_ret = linear2(ret)

        test_layer()

    """
    if func is None:
        return _switch_tracer_mode_guard_(is_train=False)
    else:

        @decorator.decorator
        def __impl__(func, *args, **kwargs):
            with _switch_tracer_mode_guard_(is_train=False):
                return func(*args, **kwargs)

        return __impl__(func)


class no_grad_:
302
    """
303 304
    :api_attr: imperative

305
    Create a context which disables dygraph gradient calculation.
306 307
    In this mode, the result of every computation will have `stop_gradient` set
    to `True`.
308

309
    Also functions as a decorator. (Make sure to use an instance.)
310 311 312 313 314 315

    Examples:

     .. code-block:: python

        import numpy as np
316
        import paddle
317

318 319 320
        # use as generator

        data = np.array([[2, 3], [4, 5]]).astype('float32')
321 322 323
        l0 = paddle.nn.Linear(2, 2)  # l0.weight.gradient() is None
        l1 = paddle.nn.Linear(2, 2)
        with paddle.no_grad():
324 325
            # l1.weight.stop_gradient is False
            tmp = l1.weight * 2  # tmp.stop_gradient is True
326
        x = paddle.to_tensor(data)
327 328 329 330 331
        y = l0(x) + tmp
        o = l1(y)
        o.backward()
        print(tmp.gradient() is None)  # True
        print(l0.weight.gradient() is None)  # False
332 333 334

        # use as decorator

335
        @paddle.no_grad()
336
        def test_layer():
337
            inp = np.ones([3, 1024], dtype='float32')
338 339 340
            t = paddle.to_tensor(inp)
            linear1 = paddle.nn.Linear(1024, 4, bias_attr=False)
            linear2 = paddle.nn.Linear(4, 4)
341 342
            ret = linear1(t)
            dy_ret = linear2(ret)
343 344 345 346

        test_layer()
    """

347
    def __call__(self, func):
S
songyouwei 已提交
348
        @decorator.decorator
349 350
        def _decorate_function(func, *args, **kwargs):
            with self:
351
                return func(*args, **kwargs)
352

353 354 355 356 357 358 359 360 361 362 363 364 365 366 367
        @decorator.decorator
        def _decorate_generator(func, *args, **kwargs):
            gen = func(*args, **kwargs)
            with self:
                for x in gen:
                    yield x

        if inspect.isgeneratorfunction(func):
            return _decorate_generator(func)
        else:
            return _decorate_function(func)

    def __enter__(self):
        tracer = framework._dygraph_tracer()
        if tracer:
368 369
            self.orig = tracer._has_grad
            tracer._has_grad = False
370 371 372 373

    def __exit__(self, *args):
        tracer = framework._dygraph_tracer()
        if tracer:
374
            tracer._has_grad = self.orig
375 376


S
rename  
sneaxiy 已提交
377
@signature_safe_contextmanager
P
Paddle CI 已提交
378
def guard(place=None):
379
    """
380 381
    :api_attr: imperative

382
    This context will create a dygraph context for dygraph to run, using python ``with`` statement.
383

384
    Parameters:
385 386 387 388
        place(fluid.CPUPlace| fluid.CUDAPlace|str, optional): Place to execute dygraph. 
            If None, the running place will be determined according to the way of paddle compilation.
            If ``place`` is string, It can be ``cpu``, ``gpu:x`` and ``xpu:x``, where ``x`` is the
            index of the GPUs or XPUs. Default: None
389 390 391 392 393 394 395 396 397 398 399 400

    return:
        None

    Examples:

     .. code-block:: python

        import numpy as np
        import paddle.fluid as fluid

        with fluid.dygraph.guard():
401
            inp = np.ones([3, 1024], dtype='float32')
402
            t = fluid.dygraph.base.to_variable(inp)
403 404 405 406
            linear1 = fluid.Linear(1024, 4, bias_attr=False)
            linear2 = fluid.Linear(4, 4)
            ret = linear1(t)
            dy_ret = linear2(ret)
407 408

    """
409 410
    train = framework.Program()
    startup = framework.Program()
J
Jiabin Yang 已提交
411
    tracer = Tracer()
412
    VarBase = core.VarBase
M
minqiyang 已提交
413

414
    if place is not None:
415
        expected_place = _get_paddle_place(place)
416 417
    else:
        expected_place = framework._current_expected_place()
M
minqiyang 已提交
418

419 420
    with framework.program_guard(train, startup):
        with framework.unique_name.guard():
L
lujun 已提交
421
            with framework._dygraph_guard(tracer):
422
                with framework._dygraph_place_guard(expected_place):
P
Paddle CI 已提交
423
                    yield
424 425


426 427 428 429
@framework.dygraph_only
def grad(outputs,
         inputs,
         grad_outputs=None,
Z
Zeng Jinle 已提交
430
         retain_graph=None,
431
         create_graph=False,
Z
Zeng Jinle 已提交
432 433
         only_inputs=True,
         allow_unused=False,
434
         no_grad_vars=None):
Z
Zeng Jinle 已提交
435 436
    ''' 
    .. note::
437
        **This API is ONLY available in imperative mode.**
Z
Zeng Jinle 已提交
438 439 440 441

    This API computes the sum of gradients of `outputs` with respect to each `inputs` .

    Parameters:
442 443 444 445
        outputs (Tensor|list(Tensor)|tuple(Tensor)): the output Tensor or 
            Tensor list/tuple of the graph to compute gradients.
        inputs (Tensor|list(Tensor)|tuple(Tensor)): the input Tensor or 
            Tensor list/tuple of the graph to compute gradients. The returned
Z
Zeng Jinle 已提交
446
            values of this API are the gradients of `inputs` . 
447
        grad_outputs (Tensor|list(Tensor|None)|tuple(Tensor|None), optional): 
Z
Zeng Jinle 已提交
448 449 450 451 452 453
            initial gradient values of `outputs` . If `grad_outputs` is None, 
            the initial gradient values of `outputs` would be Tensors filled with 1; 
            if `grad_outputs` is not None, it must have the same length as `outputs` , 
            and in this case, the initial gradient value of the i-th `outputs` would
            be: (1) a Tensor filled with 1 when the i-th element of `grad_outputs` 
            is None; (2) the i-th element of `grad_outputs` when the i-th element of
454
            `grad_outputs` is a Tensor. Default None.
Z
Zeng Jinle 已提交
455 456 457 458 459 460 461 462 463 464 465
        retain_graph (bool, optional): whether to retain the forward graph which 
            is used to calculate the gradient. When it is True, the graph would 
            be retained, in which way users can calculate backward twice for the 
            same graph. When it is False, the graph would be freed. Default None,
            which means it is equal to `create_graph` . 
        create_graph (bool, optional): whether to create the gradient graphs of
            the computing process. When it is True, higher order derivatives are
            supported to compute; when it is False, the gradient graphs of the
            computing process would be discarded. Default False.
        only_inputs (bool, optional): whether to only compute the gradients of
            `inputs` . If it is False, the gradients of all remaining leaf 
466
            Tensors in the graph would be also computed and accumulated. 
Z
Zeng Jinle 已提交
467 468 469 470
            If it is True, only the gradients of `inputs` would be computed.
            Default True. only_inputs=False is under development, and it is
            not supported yet.    
        allow_unused (bool, optional): whether to raise error or return None if some 
471
            Tensors of `inputs` are unreachable in the graph. If some Tensors of 
Z
Zeng Jinle 已提交
472 473 474
            `inputs` are unreachable in the graph (i.e., their gradients are None),  
            error would be raised if allow_unused=False, or None would be returned as
            their gradients if allow_unused=True. Default False.
475 476
        no_grad_vars (Tensor|list(Tensor)|tuple(Tensor)|set(Tensor), optional): 
            the Tensors whose gradients are not needed to compute. Default None.
Z
Zeng Jinle 已提交
477 478

    Returns:
L
levi131 已提交
479
        list: a list of Tensors, whose length is the same as the Tensor number 
480
        inside `inputs`, and the i-th returned Tensor is the sum of gradients of 
Z
Zeng Jinle 已提交
481 482 483 484 485
        `outputs` with respect to the i-th `inputs`.

    Examples 1:
        .. code-block:: python

486
            import paddle
Z
Zeng Jinle 已提交
487 488

            def test_dygraph_grad(create_graph):
489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514
                x = paddle.ones(shape=[1], dtype='float32')
                x.stop_gradient = False
                y = x * x

                # Since y = x * x, dx = 2 * x
                dx = paddle.grad(
                        outputs=[y],
                        inputs=[x],
                        create_graph=create_graph,
                        retain_graph=True)[0]

                z = y + dx

                # If create_graph = False, the gradient of dx
                # would not be backpropagated. Therefore,
                # z = x * x + dx, and x.gradient() = 2 * x = 2.0

                # If create_graph = True, the gradient of dx
                # would be backpropagated. Therefore,
                # z = x * x + dx = x * x + 2 * x, and
                # x.gradient() = 2 * x + 2 = 4.0

                z.backward()
                return x.gradient()

            print(test_dygraph_grad(create_graph=False)) # [2.]
Z
Zeng Jinle 已提交
515 516 517 518 519
            print(test_dygraph_grad(create_graph=True)) # [4.]

    Examples 2:
        .. code-block:: python

520
            import paddle
Z
Zeng Jinle 已提交
521 522

            def test_dygraph_grad(grad_outputs=None):
523
                x = paddle.to_tensor(2.0)
Z
Zeng Jinle 已提交
524 525 526 527 528 529 530 531 532 533 534 535 536 537 538
                x.stop_gradient = False

                y1 = x * x
                y2 = x * 3 

                # If grad_outputs=None, dy1 = [1], dy2 = [1].
                # If grad_outputs=[g1, g2], then:
                #    - dy1 = [1] if g1 is None else g1
                #    - dy2 = [1] if g2 is None else g2

                # Since y1 = x * x, dx = 2 * x * dy1.
                # Since y2 = x * 3, dx = 3 * dy2.
                # Therefore, the final result would be:
                # dx = 2 * x * dy1 + 3 * dy2 = 4 * dy1 + 3 * dy2.

539
                dx = paddle.grad(
Z
Zeng Jinle 已提交
540 541 542 543 544 545
                    outputs=[y1, y2], 
                    inputs=[x],
                    grad_outputs=grad_outputs)[0]

                return dx.numpy()

546
            grad_value = paddle.to_tensor(4.0)
Z
Zeng Jinle 已提交
547 548 549 550
            # dy1 = [1], dy2 = [1]
            print(test_dygraph_grad(None)) # [7.]

            # dy1 = [1], dy2 = [4]
551
            print(test_dygraph_grad([None, grad_value])) # [16.]
Z
Zeng Jinle 已提交
552 553

            # dy1 = [4], dy2 = [1]
554
            print(test_dygraph_grad([grad_value, None])) # [19.]
Z
Zeng Jinle 已提交
555 556

            # dy1 = [3], dy2 = [4]
557
            grad_y1 = paddle.to_tensor(3.0)
558
            print(test_dygraph_grad([grad_y1, grad_value])) # [24.]
Z
Zeng Jinle 已提交
559 560
	'''

561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597
    def check_in_out(in_out_list, name):
        assert in_out_list is not None, "{} should not be None".format(name)

        if isinstance(in_out_list, (list, tuple)):
            assert len(in_out_list) > 0, "{} cannot be empty".format(name)
            for each_var in in_out_list:
                assert isinstance(
                    each_var,
                    core.VarBase), "Elements of {} must be Variable".format(
                        name)
            return in_out_list
        else:
            assert isinstance(
                in_out_list,
                core.VarBase), "{} must be Variable or list of Variable".format(
                    name)
            return [in_out_list]

    outputs = check_in_out(outputs, 'outputs')
    inputs = check_in_out(inputs, 'inputs')

    if grad_outputs is not None:
        if not isinstance(grad_outputs, (list, tuple)):
            grad_outputs = [grad_outputs]

        for each_var in grad_outputs:
            if each_var is not None:
                assert isinstance(
                    each_var, core.VarBase
                ), "grad_outputs must be None, a Variable or a list containing None or Variables"
    else:
        grad_outputs = []

    if len(grad_outputs) > 0:
        assert len(grad_outputs) == len(
            outputs), "The length of grad_outputs must be equal to outputs"

Z
Zeng Jinle 已提交
598 599 600 601 602 603 604
    if no_grad_vars is None:
        no_grad_vars = []
    elif isinstance(no_grad_vars, core.VarBase):
        no_grad_vars = [no_grad_vars]
    elif isinstance(no_grad_vars, (list, tuple, set)):
        no_grad_vars = list(no_grad_vars)
        for var in no_grad_vars:
605
            assert isinstance(
Z
Zeng Jinle 已提交
606
                var, core.VarBase), "no_grad_vars can only contains Variable"
607 608
    else:
        raise AssertionError(
Z
Zeng Jinle 已提交
609
            "no_grad_vars must be None, Variable or list/tuple/set of Variables")
610 611 612

    assert isinstance(create_graph, bool), "create_graph must be True or False"

Z
Zeng Jinle 已提交
613 614 615 616 617 618 619 620 621 622 623
    if retain_graph is None:
        retain_graph = create_graph

    assert isinstance(retain_graph,
                      bool), "retain_graph must be None, True or False"

    assert isinstance(allow_unused, bool), "allow_unused must be True or False"

    assert isinstance(only_inputs, bool), "only_inputs must be True or False"
    assert only_inputs, "only_inputs=False is not supported yet"

624 625
    place = core.Place()
    place.set_place(framework._current_expected_place())
626 627 628
    return core.dygraph_partial_grad(inputs, outputs, grad_outputs,
                                     no_grad_vars, place, create_graph,
                                     retain_graph, allow_unused, only_inputs)
629 630


631
@framework.dygraph_only
632
def to_variable(value, name=None, zero_copy=None, dtype=None):
633
    r"""
634 635
    :api_attr: imperative

C
chentianyu03 已提交
636 637
    The API will create a ``Variable`` object from 
    tuple, list, numpy\.ndarray or Variable object.
638

639
    Parameters:
C
chentianyu03 已提交
640 641
        value(tuple|list|ndarray|Variable|Tensor): Initial data. 
            Can be a list, tuple, NumPy ndarray, Variable, Tensor.
642 643 644
            The shape can be multi-dimensional. The data type is one of 
            numpy\.{float16, float32, float64, int16, int32, int64, 
            uint8, uint16, complex64, complex128}.
645 646
        name(str, optional): The default value is None. Normally there is no 
            need for user to set this property. For more information, please 
L
Leo Chen 已提交
647
            refer to :ref:`api_guide_Name` . 
648 649
        zero_copy(bool, optional): Whether to share memory with the input numpy 
            array. This parameter only works with CPUPlace and will be set to 
L
Leo Chen 已提交
650
            True when it is None. Default: None. (Note: zero_copy is discarded temporally for some reason.)
651 652 653
        dtype(str, optional): The desired data type of returned ``Variable`` .
            Can be 'bool' , 'float16' , 'float32' , 'float64' , 'int8' , 'int16' , 
            'int32' , 'int64' , 'uint8' . Default: None.
654

655
    Returns:
C
chentianyu03 已提交
656
        Variable : If ``value`` is a tuple/list/numpy\.ndarray object, 
657
            return ``Tensor`` created from the corresponding numpy\.ndarray object, which has 
C
chentianyu03 已提交
658
            same data type and shape with ``value``. 
659

660 661 662 663 664 665 666 667

    Examples:

     .. code-block:: python

        import numpy as np
        import paddle.fluid as fluid

668
        with fluid.dygraph.guard(fluid.CPUPlace()):
669
            x = np.ones([2, 2], np.float32)
670 671 672
            y = fluid.dygraph.to_variable(x, zero_copy=False)
            x[0][0] = -1
            y[0][0].numpy()  # array([1.], dtype=float32)
673
            y = fluid.dygraph.to_variable(x)
674 675
            x[0][0] = 0
            y[0][0].numpy()  # array([0.], dtype=float32)
676 677 678 679
            c = np.array([2+1j, 2])
            z = fluid.dygraph.to_variable(c)
            z.numpy() # array([2.+1.j, 2.+0.j])
            z.dtype # 'complex128'
680 681 682 683 684 685 686

            y = fluid.dygraph.to_variable([[0.1, 1.2], [2.2, 3.1], [4.9, 5.2]])
            y.shape     # [3L, 2L]

            y = fluid.dygraph.to_variable(((0.1, 1.2), (2.2, 3.1), (4.9, 5.2)), dtype='int32')
            y.shape     # [3L, 2L]

687
    """
688
    support_type = (list, tuple, np.ndarray, core.VarBase, framework.Variable,
C
chentianyu03 已提交
689
                    core.Tensor, core.LoDTensor)
690 691 692 693
    if not isinstance(value, support_type):
        raise TypeError(
            "The type of 'value' in fluid.dygraph.to_variable must be %s, but received %s."
            % (support_type, type(value)))
C
chentianyu03 已提交
694
    if isinstance(value, (core.VarBase, framework.Variable)):
695 696 697 698
        return value
    elif isinstance(value, (core.Tensor, core.LoDTensor)):
        return core.VarBase(value)
    else:
699 700
        if isinstance(framework._current_expected_place(),
                      framework.core.CPUPlace):
L
Leo Chen 已提交
701
            #TODO(zhiqiu): we found two problems when enable zero_copy on CPUPlace.
702
            # (1): eigen requires 16-bytes alignments, but the data of numpy array may not statisfy.
L
Leo Chen 已提交
703 704 705 706 707 708 709 710 711
            # Details: https://eigen.tuxfamily.org/dox/group__TopicUnalignedArrayAssert.html
            # (2): when used in flask framework, it may result in hang.
            # Details: https://github.com/PaddlePaddle/Paddle/issues/26635
            # So, we temporally diable the zero_copy strategy.
            if zero_copy == True:
                warnings.warn(
                    "Currently, zero_copy is not supported, and it will be discarded."
                )
                zero_copy = False
712 713
        else:
            assert not zero_copy, "zero_copy mode can only be used with CPUPlace"
714 715 716 717 718 719 720 721 722

        if not isinstance(value, np.ndarray):
            value = np.array(value)

        if dtype is not None:
            dtype = convert_dtype(dtype)
            if value.dtype != dtype:
                value = value.astype(dtype)

C
chentianyu03 已提交
723 724 725 726 727 728 729
        py_var = core.VarBase(
            value=value,
            place=framework._current_expected_place(),
            persistable=False,
            zero_copy=zero_copy,
            name=name if name else '')
        return py_var