loss.py 93.7 KB
Newer Older
1
# -*- coding: utf-8 -*
2 3 4 5 6 7 8 9 10 11 12 13 14 15
#   Copyright (c) 2020 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.

16
import paddle
17
from ...fluid.data_feeder import check_variable_and_dtype
18

19
# TODO: define loss functions of neural network
20
import numpy as np
21 22 23
import paddle
import paddle.fluid as fluid
from ...fluid.layers.nn import _elementwise_op_in_dygraph
Z
zhiboniu 已提交
24 25 26
from ...fluid.layers import dice_loss  # noqa: F401
from ...fluid.layers import log_loss  # noqa: F401
from ...fluid.layers import npair_loss  # noqa: F401
27
from ...tensor.manipulation import reshape
Z
zhiboniu 已提交
28 29
from ...fluid.layers import softmax_with_cross_entropy as fluid_softmax_with_cross_entropy
from ...fluid.layers import square_error_cost  # noqa: F401
30

Z
zhiboniu 已提交
31
from ...fluid.layers import edit_distance  # noqa: F401
32
from ...fluid.layers import huber_loss
33
from ...fluid.layer_helper import LayerHelper
34
from ...fluid.framework import _varbase_creator
35
from ...static import Variable
36
from paddle.utils import deprecated
W
wanghuancoder 已提交
37
from paddle import _C_ops
Z
zhiboniu 已提交
38
from paddle import in_dynamic_mode
J
Jiabin Yang 已提交
39
from paddle.framework import core
H
hong 已提交
40
from ...fluid.framework import _in_legacy_dygraph, in_dygraph_mode, _non_static_mode
41 42
__all__ = []

43

44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
def binary_cross_entropy(input, label, weight=None, reduction='mean',
                         name=None):
    """
    This op measures the binary_cross_entropy loss between input predictions ``input``
    and target labels ``label`` . The binary_cross_entropy loss can be described as:

    If :attr:`weight` is set, the loss is:

    .. math::
        Out = -1 * weight * (label * log(input) + (1 - label) * log(1 - input))

    If :attr:`weight` is None, the loss is:

    .. math::
        Out = -1 * (label * log(input) + (1 - label) * log(1 - input))

    If :attr:`reduction` set to ``'none'``, the interface will return the original loss `Out`.

    If :attr:`reduction` set to ``'mean'``, the reduced mean loss is:

    .. math::
        Out = MEAN(Out)

    If :attr:`reduction` set to ``'sum'``, the reduced sum loss is:

    .. math::
        Out = SUM(Out)

    Note that the input predictions ``input`` always be the output of sigmoid, and the target labels ``label``
    should be numbers between 0 and 1.

    Parameters:
        input (Tensor): The input predications tensor. 2-D tensor with shape: [N, *],
            N is batch_size, `*` means number of additional dimensions. The ``input``
            should always be the output of sigmod.  Available dtype is float32, float64.
        label (Tensor): The target labels tensor. 2-D tensor with the same shape as
            ``input``. The target labels which values should be numbers between 0 and 1.
            Available dtype is float32, float64.
        weight (Tensor, optional): A manual rescaling weight given to the loss of each
            batch element. If given, has to be a Tensor of size nbatch and the data type
            is float32, float64. Default is ``'None'``.
        reduction (str, optional): Indicate how to average the loss by batch_size,
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned;
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
            If :attr:`reduction` is ``'sum'``, the summed loss is returned.
            Default is ``'mean'``.
        name (str, optional): Name for the operation (optional, default is None).
            For more information, please refer to :ref:`api_guide_Name`.


    Returns:
        output (Tensor): If ``reduction`` is ``'none'``, the shape of output is
            same as ``input`` , else the shape of output is scalar.

    Examples:
        .. code-block:: python

            import paddle

104 105
            input = paddle.to_tensor([0.5, 0.6, 0.7], 'float32')
            label = paddle.to_tensor([1.0, 0.0, 1.0], 'float32')
106
            output = paddle.nn.functional.binary_cross_entropy(input, label)
N
Noel 已提交
107
            print(output)  # [0.65537095]
108 109 110 111 112 113 114 115

    """
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in binary_cross_entropy should be 'sum', "
            "'mean' or 'none', but received %s, which is not allowed." %
            reduction)

J
Jiabin Yang 已提交
116 117
    if in_dygraph_mode():
        out = _C_ops.final_state_bce_loss(input, label)
118
        if weight is not None:
W
wanghuancoder 已提交
119
            out = _C_ops.elementwise_mul(out, weight, 'axis', -1)
120 121

        if reduction == 'sum':
W
wanghuancoder 已提交
122 123
            return _C_ops.reduce_sum(out, 'dim', [0], 'keep_dim', False,
                                     "reduce_all", True)
124
        elif reduction == 'mean':
W
wanghuancoder 已提交
125
            return _C_ops.mean(out)
126 127 128
        else:
            return out
    else:
J
Jiabin Yang 已提交
129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170
        if _in_legacy_dygraph():
            out = _C_ops.bce_loss(input, label)
            if weight is not None:
                out = _C_ops.elementwise_mul(out, weight, 'axis', -1)
            if reduction == 'sum':
                return _C_ops.reduce_sum(out, 'dim', [0], 'keep_dim', False,
                                         "reduce_all", True)
            elif reduction == 'mean':
                return _C_ops.mean(out)
            else:
                return out
        else:
            fluid.data_feeder.check_variable_and_dtype(
                input, 'input', ['float32', 'float64'], 'binary_cross_entropy')
            fluid.data_feeder.check_variable_and_dtype(
                label, 'label', ['float32', 'float64'], 'binary_cross_entropy')

            sub_name = name if weight is None and reduction == 'none' else None
            helper = LayerHelper("binary_cross_entropy", name=sub_name)
            out = helper.create_variable_for_type_inference(dtype=input.dtype)
            helper.append_op(
                type='bce_loss',
                inputs={
                    'X': [input],
                    'Label': [label],
                },
                outputs={'Out': [out]})

            if weight is not None:
                if isinstance(weight, paddle.static.Variable):
                    weight_name = name if reduction == 'none' else None
                    out = paddle.multiply(out, weight, name=weight_name)
                else:
                    raise ValueError(
                        "The weight is not a Tensor, please convert to Tensor.")

            if reduction == 'sum':
                return paddle.sum(out, name=name)
            elif reduction == 'mean':
                return paddle.mean(out, name=name)
            else:
                return out
171 172


173 174 175 176 177 178
def binary_cross_entropy_with_logits(logit,
                                     label,
                                     weight=None,
                                     reduction='mean',
                                     pos_weight=None,
                                     name=None):
179
    r"""
180 181 182 183 184 185 186 187 188 189 190 191 192
    This operator combines the sigmoid layer and the :ref:`api_nn_loss_BCELoss` layer.
    Also, we can see it as the combine of ``sigmoid_cross_entropy_with_logits``
    layer and some reduce operations.

    This measures the element-wise probability error in classification tasks
    in which each class is independent.
    This can be thought of as predicting labels for a data-point, where labels
    are not mutually exclusive. For example, a news article can be about
    politics, technology or sports at the same time or none of these.

    First this operator calculate loss function as follows:

    .. math::
193
           Out = -Labels * \log(\sigma(Logit)) - (1 - Labels) * \log(1 - \sigma(Logit))
194

195
    We know that :math:`\sigma(Logit) = \frac{1}{1 + e^{-Logit}}`. By substituting this we get:
196 197

    .. math::
198
           Out = Logit - Logit * Labels + \log(1 + e^{-Logit})
199

N
Noel 已提交
200
    For stability and to prevent overflow of :math:`e^{-Logit}` when Logit < 0,
201 202 203
    we reformulate the loss as follows:

    .. math::
204
           Out = \max(Logit, 0) - Logit * Labels + \log(1 + e^{-\|Logit\|})
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248

    Then, if ``weight`` or ``pos_weight`` is not None, this operator multiply the
    weight tensor on the loss `Out`. The ``weight`` tensor will attach different
    weight on every items in the batch. The ``pos_weight`` will attach different
    weight on the positive label of each class.

    Finally, this operator applies reduce operation on the loss.
    If :attr:`reduction` set to ``'none'``, the operator will return the original loss `Out`.
    If :attr:`reduction` set to ``'mean'``, the reduced mean loss is :math:`Out = MEAN(Out)`.
    If :attr:`reduction` set to ``'sum'``, the reduced sum loss is :math:`Out = SUM(Out)`.

    Note that the target labels ``label`` should be numbers between 0 and 1.

    Args:
        logit (Tensor): The input predications tensor. 2-D tensor with shape: [N, *],
            N is batch_size, `*` means number of additional dimensions. The ``logit``
            is usually the output of Linear layer. Available dtype is float32, float64.
        label (Tensor): The target labels tensor. 2-D tensor with the same shape as
            ``logit``. The target labels which values should be numbers between 0 and 1.
            Available dtype is float32, float64.
        weight (Tensor, optional): A manual rescaling weight given to the loss of each
            batch element. If given, it has to be a 1D Tensor whose size is `[N, ]`,
            The data type is float32, float64. Default is ``'None'``.
        reduction (str, optional): Indicate how to average the loss by batch_size,
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned;
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
            If :attr:`reduction` is ``'sum'``, the summed loss is returned.
            Default is ``'mean'``.
        pos_weight (Tensor, optional): A weight of positive examples. Must be a vector
            with length equal to the number of classes. The data type is float32, float64.
            Default is ``'None'``.
        name (str, optional): Name for the operation (optional, default is None).
            For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        output (Tensor): If ``reduction`` is ``'none'``, the shape of output is
            same as ``logit`` , else the shape of output is scalar.

    Examples:

        .. code-block:: python

            import paddle
N
Noel 已提交
249

250 251
            logit = paddle.to_tensor([5.0, 1.0, 3.0])
            label = paddle.to_tensor([1.0, 0.0, 1.0])
252
            output = paddle.nn.functional.binary_cross_entropy_with_logits(logit, label)
N
Noel 已提交
253
            print(output)  # [0.45618808]
254 255 256 257 258 259 260 261

    """
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in binary_cross_entropy_with_logits "
            "should be 'sum', 'mean' or 'none', but received %s, which is not allowed."
            % reduction)

Z
zhiboniu 已提交
262
    if in_dynamic_mode():
263
        one = _varbase_creator(dtype=logit.dtype)
W
wanghuancoder 已提交
264 265 266 267
        _C_ops.fill_constant(one, 'value',
                             float(1.0), 'force_cpu', False, 'dtype', one.dtype,
                             'str_value', '1.0', 'shape', [1])
        out = _C_ops.sigmoid_cross_entropy_with_logits(logit, label)
268
        if pos_weight is not None:
W
wanghuancoder 已提交
269 270 271 272 273
            log_weight = _C_ops.elementwise_add(
                _C_ops.elementwise_mul(label,
                                       _C_ops.elementwise_sub(pos_weight, one)),
                one)
            out = _C_ops.elementwise_mul(out, log_weight)
274
        if weight is not None:
W
wanghuancoder 已提交
275
            out = _C_ops.elementwise_mul(out, weight)
276 277

        if reduction == "sum":
W
wanghuancoder 已提交
278
            return _C_ops.reduce_sum(out, 'reduce_all', True)
279
        elif reduction == "mean":
W
wanghuancoder 已提交
280
            return _C_ops.mean(out)
281 282 283 284 285 286 287 288 289 290 291 292 293
        else:
            return out

    fluid.data_feeder.check_variable_and_dtype(
        logit, 'logit', ['float32', 'float64'],
        'binary_cross_entropy_with_logits')
    fluid.data_feeder.check_variable_and_dtype(
        label, 'label', ['float32', 'float64'],
        'binary_cross_entropy_with_logits')
    sigmoid_name = None
    if reduction == 'none' and pos_weight is None and weight is None:
        sigmoid_name = name

294
    out = paddle.fluid.layers.sigmoid_cross_entropy_with_logits(
295 296
        logit, label, name=sigmoid_name)

Z
zhiboniu 已提交
297
    one = paddle.full(shape=[1], fill_value=1.0, dtype=logit.dtype)
298 299 300 301 302
    if pos_weight is not None:
        fluid.data_feeder.check_variable_and_dtype(
            pos_weight, 'pos_weight', ['float32', 'float64'],
            'binary_cross_entropy_with_logits')
        log_weight = paddle.add(
303
            paddle.multiply(label, paddle.subtract(pos_weight, one)), one)
304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320
        pos_weight_name = name if reduction == 'none' and weight is None else None
        out = paddle.multiply(out, log_weight, name=pos_weight_name)

    if weight is not None:
        fluid.data_feeder.check_variable_and_dtype(
            weight, 'weight', ['float32', 'float64'],
            'binary_cross_entropy_with_logits')
        weight_name = name if reduction == 'none' else None
        out = paddle.multiply(out, weight, name=weight_name)

    if reduction == "sum":
        return paddle.sum(out, name=name)
    elif reduction == "mean":
        return paddle.mean(out, name=name)
    return out


321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403
def hsigmoid_loss(input,
                  label,
                  num_classes,
                  weight,
                  bias=None,
                  path_table=None,
                  path_code=None,
                  is_sparse=False,
                  name=None):
    """
    The hierarchical sigmoid organizes the classes into a complete binary tree to reduce the computational complexity
    and speed up the model training, especially the training of language model.
    Each leaf node of the complete binary tree represents a class(word) and each non-leaf node acts as a binary classifier.
    For each class(word), there's a unique path from root to itself, hsigmoid calculate the cost for each non-leaf node on
    the path, and sum them to get a total cost.
    Comparing to softmax, the OP can reduce the computational complexity from :math:`O(N)` to :math:`O(logN)`, where :math:`N`
    represents the number of classes or the size of word dict.

    The OP supports default tree and custom tree. For the default tree, you can refer to `Hierarchical Probabilistic Neural
    Network Language Model <http://www.iro.umontreal.ca/~lisa/pointeurs/hierarchical-nnlm-aistats05.pdf>`_. For the custom
    tree, you need to set :attr:`is_custom` to True, and do the following steps (take the language model as an example):

    1. Using a custom word dict to build a binary tree, each leaf node should be an word in the word dict.
    2. Creating a dict map word_id -> path that from the word to the root node, we call it path_table.
    3. Creating a dict map word_id -> code of path that from the word to the root node, we call it path_code.
       Code means the label of each binary classifier, 1 indicate true, 0 indicate false.
    4. Now, each word should has its path and code along the path, you can pass a batch of path and code related
       to the same batch of inputs.

    Parameters:
        input (Tensor): A tensor with the shape [N, D], where N is the size of mini-batch,
            and D is the feature size. Its data type supports float32 or float64.
        label (Tensor): A tensor contains the labels of training data. Its shape is [N, 1]
            and data type is int64.
        num_classes (int): The number of classes or the size of word dict, must be greater than 2.
            If the default tree is used (path_code and path_table is None are None), `num_classes`
            should not be None. If the custom tree is used (path_code and path_table is None are not None),
            `num_classes` should be the number of non-leaf nodes, which indicates the num of
            classes using by the binary classifier.
        weight (Tensor): A tensor with shape (num_classes - 1, D), with the same data type as `input`.
        bias (Tensor, optional): A tensor with shape (num_classes - 1, 1), with the same data type as `input`.
            If `bias` is None, no bias will be add. Default is None.
        path_table (Tensor, optional): A tensor that stores each batch of samples' path from leaf to root
            node, its shape is [N, L] and data type is int64, where L is the length of path. For each sample i,
            path_table[i] is a np.array like structure and each element in this array is the indexes in parent
            nodes' weight matrix. If `path_table` and `path_code` are None, the default tree will be used.
            Default is None.
        path_code (Tensor, optional): A tensor that stores each batch of samples' code of path from leaf
            to root node, its shape is [N, L] and data type is int64, which is the same as :attr:`path_table`.
            Each code of path is consisted with the code of nodes from leaf to root node. If `path_table` and
            `path_code` are None, the default tree will be used. Default is None.
        is_sparse (bool, optional): Whether use sparse updating instead of dense updating. If `is_sparse` is True,
            the gradient of `weight` and `input` will be sparse. Default is False.
        name (str, optional): Name for the operation (optional, default is None).
            For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        A tensor with the cost of hierarchical sigmoid, its shape is [N, 1] and data type is the same as `input`.

    Examples:
        .. code-block:: python

            import paddle
            import paddle.nn.functional as F

            paddle.set_device('cpu')

            input = paddle.uniform([2, 3])
            # [[-0.8018668   0.8736385  -0.9064771 ] # random
            #  [-0.10228515 -0.87188244 -0.8783718 ]] # random
            label = paddle.to_tensor([0, 1, 4, 5])
            num_classes = 5
            weight=paddle.uniform([num_classes-1, 3])
            # [[-0.24148715  0.8449961  -0.7399121 ] # random
            #  [-0.9800559   0.43509364  0.9091208 ] # random
            #  [ 0.60194826  0.10430074 -0.4521166 ] # random
            #  [-0.4469818  -0.01536179 -0.604454  ]] # random

            out=F.hsigmoid_loss(input, label, num_classes, weight)
            # [[3.0159328]
            #  [2.2407534]]
    """

Z
zhiboniu 已提交
404
    if in_dynamic_mode():
W
wanghuancoder 已提交
405
        out, _, _ = _C_ops.hierarchical_sigmoid(
406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452
            input, weight, label, path_table, path_code, bias, 'num_classes',
            num_classes, 'is_sparse', is_sparse, 'remote_prefetch', is_sparse)
        return out

    check_variable_and_dtype(input, 'input', ['float32', 'float64'],
                             'hsigmoid_loss')
    check_variable_and_dtype(label, 'label', ['int64'], 'hsigmoid_loss')
    check_variable_and_dtype(weight, 'weight', ['float32', 'float64'],
                             'hsigmoid_loss')
    if bias is not None:
        check_variable_and_dtype(bias, 'bias', ['float32', 'float64'],
                                 'hsigmoid_loss')
    if path_table is not None:
        check_variable_and_dtype(path_table, 'path_table', ['int64'],
                                 'hsigmoid_loss')
    if path_code is not None:
        check_variable_and_dtype(path_code, 'path_code', ['int64'],
                                 'hsigmoid_loss')

    attrs = {
        "num_classes": num_classes,
        "is_sparse": is_sparse,
        "remote_prefetch": is_sparse
    }

    inputs = {
        "X": input,
        "W": weight,
        "Bias": bias,
        "PathTable": path_table,
        "PathCode": path_code,
        "Label": label
    }

    helper = LayerHelper('hsigmoid_loss', **locals())
    out = helper.create_variable_for_type_inference(input.dtype)
    pre_out = helper.create_variable_for_type_inference(input.dtype)
    outputs = {"Out": out, "PreOut": pre_out, "W_Out": weight}

    helper.append_op(
        type="hierarchical_sigmoid",
        inputs=inputs,
        outputs=outputs,
        attrs=attrs)
    return out


453
def smooth_l1_loss(input, label, reduction='mean', delta=1.0, name=None):
454
    r"""
455 456 457 458 459 460 461
    This operator calculates smooth_l1_loss. Creates a criterion that uses a squared
    term if the absolute element-wise error falls below 1 and an L1 term otherwise.
    In some cases it can prevent exploding gradients and it is more robust and less
    sensitivity to outliers. Also known as the Huber loss:

    .. math::

462
         loss(x,y) = \frac{1}{n}\sum_{i}z_i
463 464 465 466 467 468


    where z_i is given by:

    .. math::

469 470
        \mathop{z_i} = \left\{\begin{array}{rcl}
        0.5(x_i - y_i)^2 & & {if |x_i - y_i| < delta} \\
471
        delta * |x_i - y_i| - 0.5 * delta^2 & & {otherwise}
472
        \end{array} \right.
473 474 475 476 477 478 479 480 481 482 483 484 485

    Parameters:
        input (Tensor): Input tensor, the data type is float32 or float64. Shape is
            (N, C), where C is number of classes, and if shape is more than 2D, this
            is (N, C, D1, D2,..., Dk), k >= 1.
        label (Tensor): Label tensor, the data type is float32 or float64. The shape of label
            is the same as the shape of input.
        reduction (str, optional): Indicate how to average the loss by batch_size,
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
            If :attr:`reduction` is ``'sum'``, the reduced sum loss is returned.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned.
            Default is ``'mean'``.
486
        delta (float, optional): Specifies the hyperparameter delta to be used.
487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507
            The value determines how large the errors need to be to use L1. Errors
            smaller than delta are minimized with L2. Parameter is ignored for
            negative/zero values. Default = 1.0
        name (str, optional): Name for the operation (optional, default is
            None). For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        The tensor variable storing the smooth_l1_loss of input and label.

    Return type: Tensor.

    Examples:
        .. code-block:: python

            import paddle
            import numpy as np

            input_data = np.random.rand(3,3).astype("float32")
            label_data = np.random.rand(3,3).astype("float32")
            input = paddle.to_tensor(input_data)
            label = paddle.to_tensor(label_data)
C
Chen Long 已提交
508
            output = paddle.nn.functional.smooth_l1_loss(input, label)
G
Guanghua Yu 已提交
509
            print(output)
510 511 512 513 514 515 516 517 518 519 520 521 522 523 524
    """
    fluid.data_feeder.check_variable_and_dtype(
        input, 'input', ['float32', 'float64'], 'smooth_l1_loss')
    fluid.data_feeder.check_variable_and_dtype(
        label, 'label', ['float32', 'float64'], 'smooth_l1_loss')

    out = huber_loss(input=input, label=label, delta=delta)

    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in smooth_l1_loss should be 'sum', 'mean' or"
            " 'none', but received %s, which is not allowed." % reduction)
    if reduction == 'none':
        return out
    elif reduction == 'mean':
525
        return paddle.mean(out)
526
    elif reduction == 'sum':
527
        return paddle.sum(out)
528 529


530 531
def margin_ranking_loss(input,
                        other,
532
                        label,
533 534 535
                        margin=0.0,
                        reduction='mean',
                        name=None):
536
    r"""
537

538
    This op the calcluate the the margin rank loss between the input, other and label, use the math function as follows.
539

540
    .. math::
541
        margin\_rank\_loss = max(0, -label * (input - other) + margin)
542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557

    If :attr:`reduction` set to ``'mean'``, the reduced mean loss is:

    .. math::
        Out = MEAN(margin\_rank\_loss)

    If :attr:`reduction` set to ``'sum'``, the reduced sum loss is:

    .. math::
        Out = SUM(margin\_rank\_loss)

    If :attr:`reduction` set to ``'none'``, just return the origin ``margin_rank_loss``.

    Parameters:
        input(Tensor): the first input tensor, it's data type should be float32, float64.
        other(Tensor): the second input tensor, it's data type should be float32, float64.
558
        label(Tensor): the label value corresponding to input, it's data type should be float32, float64.
559 560 561 562 563 564 565 566 567 568
        margin (float, optional): The margin value to add, default value is 0;
        reduction (str, optional): Indicate the reduction to apply to the loss, the candicates are ``'none'``, ``'mean'``, ``'sum'``.If :attr:`reduction` is ``'none'``, the unreduced loss is returned; If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned. If :attr:`reduction` is ``'sum'``, the reduced sum loss is returned. Default is ``'mean'``.
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.

    Returns: Tensor, if :attr:`reduction` is ``'mean'`` or ``'sum'``, the out shape is :math:`[1]`, otherwise the shape is the same as `input` .The same dtype as input tensor.

    Examples:

        .. code-block:: python

569 570
            import paddle

Z
Zhong Hui 已提交
571 572 573
            input = paddle.to_tensor([[1, 2], [3, 4]], dtype='float32')
            other = paddle.to_tensor([[2, 1], [2, 4]], dtype='float32')
            label = paddle.to_tensor([[1, -1], [-1, -1]], dtype='float32')
574
            loss = paddle.nn.functional.margin_ranking_loss(input, other, label)
N
Noel 已提交
575
            print(loss) # [0.75]
576
    """
577 578 579 580
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in MarginRankingLoss should be 'sum', 'mean' or 'none', but "
            "received %s, which is not allowed." % reduction)
Z
zhiboniu 已提交
581
    if in_dynamic_mode():
W
wanghuancoder 已提交
582 583
        out = _C_ops.elementwise_sub(other, input)
        out = _C_ops.elementwise_mul(out, label)
584 585
        if margin != 0.0:
            margin = fluid.dygraph.base.to_variable([margin], dtype=out.dtype)
W
wanghuancoder 已提交
586 587
            out = _C_ops.elementwise_add(out, margin)
        out = _C_ops.relu(out)
588
        if reduction == 'sum':
W
wanghuancoder 已提交
589
            return _C_ops.reduce_sum(out, 'reduce_all', True)
590
        elif reduction == 'mean':
W
wanghuancoder 已提交
591
            return _C_ops.mean(out)
592 593 594 595 596 597 598 599
        return out

    helper = LayerHelper("margin_ranking_loss", **locals())
    fluid.data_feeder.check_variable_and_dtype(
        input, 'input', ['float32', 'float64'], 'margin_rank_loss')
    fluid.data_feeder.check_variable_and_dtype(
        other, 'other', ['float32', 'float64'], 'margin_rank_loss')
    fluid.data_feeder.check_variable_and_dtype(
600
        label, 'label', ['float32', 'float64'], 'margin_rank_loss')
601

602
    out = paddle.subtract(other, input)
603
    out = paddle.multiply(out, label)
604 605 606

    if margin != 0.0:
        margin_var = out.block.create_var(dtype=out.dtype)
Z
zhiboniu 已提交
607
        margin_var = paddle.full(shape=[1], fill_value=margin, dtype=out.dtype)
608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634
        out = paddle.add(out, margin_var)

    result_out = helper.create_variable_for_type_inference(input.dtype)

    if reduction == 'none':
        helper.append_op(
            type="relu", inputs={"X": out}, outputs={"Out": result_out})
        return result_out
    elif reduction == 'sum':
        out = paddle.nn.functional.relu(out)
        attrs = {"dim": [0], "keep_dim": False, "reduce_all": True}
        helper.append_op(
            type="reduce_sum",
            inputs={"X": out},
            outputs={"Out": result_out},
            attrs=attrs)
        return result_out
    elif reduction == 'mean':
        out = paddle.nn.functional.relu(out)
        helper.append_op(
            type="mean",
            inputs={"X": out},
            outputs={"Out": result_out},
            attrs={})
        return result_out


635
def l1_loss(input, label, reduction='mean', name=None):
636
    r"""
637
    This operator computes the L1 Loss of Tensor ``input`` and ``label`` as follows.
638

639
    If `reduction` set to ``'none'``, the loss is:
640 641

    .. math::
642
        Out = \lvert input - label \rvert
643

644
    If `reduction` set to ``'mean'``, the loss is:
645 646

    .. math::
647
        Out = MEAN(\lvert input - label \rvert)
648

649
    If `reduction` set to ``'sum'``, the loss is:
650 651

    .. math::
652
        Out = SUM(\lvert input - label \rvert)
653

654

655
    Parameters:
N
Noel 已提交
656 657
        input (Tensor): The input tensor. The shapes is [N, `*`], where N is batch size and `*` means any number of additional dimensions. It's data type should be float32, float64, int32, int64.
        label (Tensor): label. The shapes is [N, `*`], same shape as ``input`` . It's data type should be float32, float64, int32, int64.
658
        reduction (str, optional): Indicate the reduction to apply to the loss,
659
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
660 661 662
            If `reduction` is ``'none'``, the unreduced loss is returned;
            If `reduction` is ``'mean'``, the reduced mean loss is returned.
            If `reduction` is ``'sum'``, the reduced sum loss is returned.
663 664
            Default is ``'mean'``.
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
N
Noel 已提交
665

666
    Returns:
667 668 669
        Tensor, the L1 Loss of Tensor ``input`` and ``label``.
            If `reduction` is ``'none'``, the shape of output loss is [N, *], the same as ``input`` .
            If `reduction` is ``'mean'`` or ``'sum'``, the shape of output loss is [1].
N
Noel 已提交
670

671 672
    Examples:
        .. code-block:: python
N
Noel 已提交
673

674
            import paddle
675

676 677
            input = paddle.to_tensor([[1.5, 0.8], [0.2, 1.3]])
            label = paddle.to_tensor([[1.7, 1], [0.4, 0.5]])
678

679
            l1_loss = paddle.nn.functional.l1_loss(input, label)
680
            print(l1_loss.numpy())
681 682
            # [0.35]

683
            l1_loss = paddle.nn.functional.l1_loss(input, label, reduction='none')
684
            print(l1_loss.numpy())
685 686 687
            # [[0.20000005 0.19999999]
            # [0.2        0.79999995]]

688
            l1_loss = paddle.nn.functional.l1_loss(input, label, reduction='sum')
689
            print(l1_loss.numpy())
690 691 692 693 694 695 696
            # [1.4]
    """
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in L1Loss should be 'sum', 'mean' or 'none', but "
            "received %s, which is not allowed." % reduction)

Z
zhiboniu 已提交
697
    if in_dynamic_mode():
698
        unreduced = _elementwise_op_in_dygraph(
699
            input, label, axis=-1, act='abs', op_name='elementwise_sub')
700
        if reduction == 'mean':
W
wanghuancoder 已提交
701
            return _C_ops.mean(unreduced)
702
        elif reduction == 'sum':
W
wanghuancoder 已提交
703 704
            return _C_ops.reduce_sum(unreduced, 'dim', [0], 'keep_dim', False,
                                     'reduce_all', True)
705 706 707 708
        else:
            return unreduced

    fluid.data_feeder.check_variable_and_dtype(
709
        input, 'input', ['float32', 'float64', 'int32', 'int64'], 'l1_loss')
710 711 712 713
    fluid.data_feeder.check_variable_and_dtype(
        label, 'label', ['float32', 'float64', 'int32', 'int64'], 'l1_loss')

    if reduction == 'sum':
714
        unreduced = paddle.fluid.layers.elementwise_sub(input, label, act='abs')
715 716
        return paddle.sum(unreduced, name=name)
    elif reduction == 'mean':
717
        unreduced = paddle.fluid.layers.elementwise_sub(input, label, act='abs')
718 719
        return paddle.mean(unreduced, name=name)
    else:
720 721
        return paddle.fluid.layers.elementwise_sub(
            input, label, act='abs', name=name)
722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759


def nll_loss(input,
             label,
             weight=None,
             ignore_index=-100,
             reduction='mean',
             name=None):
    """
    This api returns negative log likelihood.
    See more detail in :ref:`api_nn_loss_NLLLoss` .

    Parameters:
         input (Tensor): Input tensor, the shape is :math:`[N, C]`, `C` is the number of classes.
             But in K-dimension situation, the shape is :math:`[N, C, d_1, d_2, ..., d_K]`.
             The data type is float32, float64.
         label (Tensor): Label tensor, the shape is :math:`[N,]` or :math:`[N, d_1, d_2, ..., d_K]`.
             The data type is int64.
         weight (Tensor, optional): Weight tensor, a manual rescaling weight given
             to each class. If given, it has to be a 1D Tensor whose size is `[C, ]`. Otherwise,
             it treated as if having all ones. the data type is
             float32, float64, Default is ``'None'``.
         ignore_index (int64, optional): Specifies a target value that is ignored
             and does not contribute to the input gradient.
         reduction (str, optional): Indicate how to average the loss,
             the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
             If `reduction` is ``'mean'``, the reduced mean loss is returned;
             if `reduction` is ``'sum'``, the reduced sum loss is returned;
             if `reduction` is ``'none'``, no reduction will be apllied.
             Default is ``'mean'``.
         name (str, optional): Name for the operation (optional, default is None).
             For more information, please refer to :ref:`api_guide_Name`.

    Returns:
         `Tensor`, the value of negative log likelihood loss.

    Examples:
        .. code-block:: python
760

761 762 763 764
                import paddle
                from paddle.nn.functional import nll_loss
                log_softmax = paddle.nn.LogSoftmax(axis=1)

765 766 767 768 769
                input = paddle.to_tensor([[0.88103855, 0.9908683 , 0.6226845 ],
                          [0.53331435, 0.07999352, 0.8549948 ],
                          [0.25879037, 0.39530203, 0.698465  ],
                          [0.73427284, 0.63575995, 0.18827209],
                          [0.05689114, 0.0862954 , 0.6325046 ]], "float32")
770
                log_out = log_softmax(input)
771
                label = paddle.to_tensor([0, 2, 1, 1, 0], "int64")
772
                result = nll_loss(log_out, label)
773
                print(result) # Tensor(shape=[1], dtype=float32, place=CPUPlace, stop_gradient=True, [1.07202101])
774 775 776 777 778 779 780 781 782 783 784 785 786
    """
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in nll_loss should be 'sum', 'mean' or "
            "'none', but received %s, which is not allowed." % reduction)

    input_shape = list(input.shape)
    input_dims = len(input_shape)
    if input_dims < 2:
        raise ValueError('Expected 2 or more dimensions (got {})'.format(
            input_dims))
    n = input_shape[0]
    c = input_shape[1]
Z
zyfncg 已提交
787 788 789 790 791 792 793 794 795 796 797
    if in_dygraph_mode():
        if input_dims != 2 and input_dims != 4:
            input, _ = _C_ops.reshape2(input, None, 'shape', [n, c, 1, -1])
            label, _ = _C_ops.reshape2(label, None, 'shape', [n, 1, -1])
            out_shape = [n] + input_shape[2:]
        out, total_weight = _C_ops.final_state_nll_loss(input, label, weight,
                                                        ignore_index, reduction)
        if input_dims != 2 and input_dims != 4 and reduction == 'none':
            out, _ = _C_ops.reshape2(out, None, 'shape', out_shape)
        return out
    if _in_legacy_dygraph():
798
        if input_dims != 2 and input_dims != 4:
W
wanghuancoder 已提交
799 800
            input, _ = _C_ops.reshape2(input, None, 'shape', [n, c, 1, -1])
            label, _ = _C_ops.reshape2(label, None, 'shape', [n, 1, -1])
801
            out_shape = [n] + input_shape[2:]
H
hong 已提交
802

W
wanghuancoder 已提交
803 804 805
        out, total_weight = _C_ops.nll_loss(input, label, weight,
                                            'ignore_index', ignore_index,
                                            'reduction', reduction)
806
        if input_dims != 2 and input_dims != 4 and reduction == 'none':
W
wanghuancoder 已提交
807
            out, _ = _C_ops.reshape2(out, None, 'shape', out_shape)
808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836
        return out

    helper = LayerHelper('nll_loss', **locals())

    if input_dims != 2 and input_dims != 4:
        input = reshape(input, shape=[n, c, 1, -1])
        label = reshape(label, shape=[n, 1, -1])
        out_shape = [n] + input_shape[2:]

    fluid.data_feeder.check_variable_and_dtype(
        input, 'input', ['float32', 'float64'], 'nll_loss')
    fluid.data_feeder.check_variable_and_dtype(label, 'label', ['int64'],
                                               'nll_loss')
    inputs = {'X': input, 'Label': label}
    attrs = {'reduction': reduction, 'ignore_index': ignore_index}
    if weight is not None:
        if isinstance(weight, Variable):
            inputs['Weight'] = weight

    out = helper.create_variable_for_type_inference(dtype=input.dtype)
    total_weight = helper.create_variable_for_type_inference(dtype=input.dtype)
    outputs = {'Out': out, 'Total_weight': total_weight}

    helper.append_op(
        type='nll_loss', inputs=inputs, outputs=outputs, attrs=attrs)
    if input_dims != 2 and input_dims != 4 and reduction == 'none':
        out = reshape(out, shape=out_shape)

    return out
837 838


839
def kl_div(input, label, reduction='mean', name=None):
840
    r"""
841 842 843 844 845 846 847 848 849 850 851
    This operator calculates the Kullback-Leibler divergence loss
    between Input(X) and Input(Target). Notes that Input(X) is the
    log-probability and Input(Target) is the probability.

    KL divergence loss is calculated as follows:

    $$l(x, y) = y * (\log(y) - x)$$

    While :math:`x` is input and :math:`y` is label.

    While :attr:`reduction` is :attr:`none`, output loss is in
852
    the same shape as input, loss in each point is calculated
853
    seperately and no reduction is applied.
854

855 856
    While :attr:`reduction` is :attr:`mean`, output loss is in
    shape of [1] and loss value is the mean value of all losses.
857

858 859
    While :attr:`reduction` is :attr:`sum`, output loss is in
    shape of [1] and loss value is the sum value of all losses.
860 861

    While :attr:`reduction` is :attr:`batchmean`, output loss is
862 863 864 865
    in shape of [1] and loss value is the sum value of all losses
    divided by batch size.

    Args:
866
        input (Tensor): The input tensor. The shapes is [N, *], where N is batch size and `*` means
867 868 869 870 871 872 873 874 875
             any number of additional dimensions. It's data type should be float32, float64.
        label (Tensor): label. The shapes is [N, *], same shape as ``input`` . It's data type should be float32, float64.
        reduction (Tensor): Indicate how to average the loss,
             the candicates are ``'none'`` | ``'batchmean'`` | ``'mean'`` | ``'sum'``.
             If `reduction` is ``'mean'``, the reduced mean loss is returned;
             If `reduction` is ``'batchmean'``, the sum loss divided by batch size is returned;
             if `reduction` is ``'sum'``, the reduced sum loss is returned;
             if `reduction` is ``'none'``, no reduction will be apllied.
             Default is ``'mean'``.
876
        name(str, optional): Name for the operation (optional, default is None). For more information,
877 878 879 880 881 882 883 884 885 886 887
            please refer to :ref:`api_guide_Name`.

    Returns:
        Tensor: The KL divergence loss. The data type is same as input tensor

    Examples:
        .. code-block:: python

            import paddle
            import numpy as np
            import paddle.nn.functional as F
888

889 890 891 892
            shape = (5, 20)
            input = np.random.uniform(-10, 10, shape).astype('float32')
            target = np.random.uniform(-10, 10, shape).astype('float32')

L
LielinJiang 已提交
893
            # 'batchmean' reduction, loss shape will be [1]
894 895
            pred_loss = F.kl_div(paddle.to_tensor(input),
                                 paddle.to_tensor(target), reduction='batchmean')
L
LielinJiang 已提交
896
            # shape=[1]
897

898
            # 'mean' reduction, loss shape will be [1]
899 900
            pred_loss = F.kl_div(paddle.to_tensor(input),
                                 paddle.to_tensor(target), reduction='mean')
901 902 903
            # shape=[1]

            # 'sum' reduction, loss shape will be [1]
904 905
            pred_loss = F.kl_div(paddle.to_tensor(input),
                                 paddle.to_tensor(target), reduction='sum')
906 907 908
            # shape=[1]

            # 'none' reduction, loss shape is same with input shape
909 910
            pred_loss = F.kl_div(paddle.to_tensor(input),
                                 paddle.to_tensor(target), reduction='none')
911 912 913
            # shape=[5, 20]

    """
L
LielinJiang 已提交
914 915 916 917
    # ugly type promotion
    if fluid.data_feeder.convert_dtype(
            input.dtype) == 'float32' and fluid.data_feeder.convert_dtype(
                label.dtype) == 'float64':
918
        input = paddle.cast(input, 'float64')
L
LielinJiang 已提交
919 920 921
    elif fluid.data_feeder.convert_dtype(
            input.dtype) == 'float64' and fluid.data_feeder.convert_dtype(
                label.dtype) == 'float32':
922
        label = paddle.cast(label, 'float64')
L
LielinJiang 已提交
923

924
    if paddle.in_dynamic_mode():
925 926 927 928 929 930 931 932 933
        out = _C_ops.kldiv_loss(input, label, 'reduction', 'none')
        if reduction == 'mean':
            out = paddle.mean(out)
        elif reduction == 'sum':
            out = paddle.sum(out)
        elif reduction == 'batchmean':
            if len(input.shape) > 0:
                batch_size = input.shape[0]
                out = paddle.sum(out) / batch_size
934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949
        return out

    helper = LayerHelper('kl_div', **locals())

    fluid.data_feeder.check_variable_and_dtype(input, 'input',
                                               ['float32', 'float64'], 'kl_div')
    fluid.data_feeder.check_variable_and_dtype(label, 'label',
                                               ['float32', 'float64'], 'kl_div')
    fluid.data_feeder.check_type(reduction, 'reduction', str, 'kl_div')

    loss = helper.create_variable_for_type_inference(dtype=input.dtype)
    helper.append_op(
        type='kldiv_loss',
        inputs={'X': input,
                'Target': label},
        outputs={'Loss': loss},
950 951 952 953 954 955 956 957 958
        attrs={'reduction': 'none'})

    if reduction == 'mean':
        loss = paddle.mean(loss)
    elif reduction == 'sum':
        loss = paddle.sum(loss)
    elif reduction == 'batchmean':
        batch_size = paddle.shape(input)[0]
        loss = paddle.sum(loss) / batch_size
959 960 961
    return loss


962
def mse_loss(input, label, reduction='mean', name=None):
963
    r"""
964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996
    This op accepts input predications and label and returns the mean square error.

    If :attr:`reduction` is set to ``'none'``, loss is calculated as:

    .. math::
        Out = (input - label)^2

    If :attr:`reduction` is set to ``'mean'``, loss is calculated as:

    .. math::
        Out = \operatorname{mean}((input - label)^2)

    If :attr:`reduction` is set to ``'sum'``, loss is calculated as:

    .. math::
        Out = \operatorname{sum}((input - label)^2)

    Parameters:
        input (Tensor): Input tensor, the data type should be float32 or float64.
        label (Tensor): Label tensor, the data type should be float32 or float64.
        reduction (string, optional): The reduction method for the output,
            could be 'none' | 'mean' | 'sum'.
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned.
            If :attr:`reduction` is ``'sum'``, the reduced sum loss is returned.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned.
            Default is ``'mean'``.
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.


    Returns:
        Tensor: The tensor tensor storing the mean square error difference of input and label.

    Return type: Tensor.
997

998 999 1000
    Examples:

        .. code-block:: python
1001

1002 1003
            import paddle
            mse_loss = paddle.nn.loss.MSELoss()
1004 1005
            input = paddle.to_tensor(1.5)
            label = paddle.to_tensor(1.7)
1006
            output = mse_loss(input, label)
B
Bai Yifan 已提交
1007
            print(output)
1008 1009 1010 1011 1012 1013 1014 1015 1016
            # [0.04000002]

    """

    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "'reduction' in 'mse_loss' should be 'sum', 'mean' or 'none', "
            "but received {}.".format(reduction))

Z
zhiboniu 已提交
1017
    if not in_dynamic_mode():
1018 1019 1020 1021 1022 1023
        paddle.fluid.data_feeder.check_variable_and_dtype(
            input, 'input', ['float32', 'float64'], 'mse_loss')
        paddle.fluid.data_feeder.check_variable_and_dtype(
            label, 'label', ['float32', 'float64'], 'mse_loss')

    if reduction == 'none':
1024
        return paddle.square(paddle.subtract(input, label), name=name)
1025 1026
    elif reduction == 'mean':
        return paddle.mean(
1027
            paddle.square(paddle.subtract(input, label)), name=name)
1028
    else:
1029
        return paddle.sum(paddle.square(paddle.subtract(input, label)),
1030
                          name=name)
1031 1032


1033 1034 1035 1036 1037
def ctc_loss(log_probs,
             labels,
             input_lengths,
             label_lengths,
             blank=0,
1038
             reduction='mean',
H
Hui Zhang 已提交
1039
             norm_by_times=False):
1040 1041
    """

1042 1043 1044
    An operator integrating the open source Warp-CTC library (https://github.com/baidu-research/warp-ctc)
    to compute Connectionist Temporal Classification (CTC) loss.
    It can be aliased as softmax with CTC, since a native softmax activation
1045 1046 1047
    is interated to the Warp-CTC library to normalize values for each row of the input tensor.

    Parameters:
1048
        log_probs (Tensor): The unscaled probability sequence with padding, which is a 3-D Tensor. The tensor shape is [max_logit_length, batch_size, num_classes + 1], where max_logit_length is the longest length of input logit sequence. The data type should be float32 or float64.
1049 1050 1051 1052 1053
        labels (Tensor): The ground truth sequence with padding, which must be a 3-D Tensor. The tensor shape is [batch_size, max_label_length], where max_label_length is the longest length of label sequence. The data type must be int32.
        input_lengths (Tensor): The length for each input sequence, it should have shape [batch_size] and dtype int64.
        label_lengths (Tensor): The length for each label sequence, it should have shape [batch_size] and dtype int64.
        blank (int, optional): The blank label index of Connectionist Temporal Classification (CTC) loss, which is in the half-opened interval [0, num_classes + 1). The data type must be int32. Default is 0.
        reduction (string, optional): Indicate how to average the loss, the candicates are ``'none'`` | ``'mean'`` | ``'sum'``. If :attr:`reduction` is ``'mean'``, the output loss will be divided by the label_lengths, and then return the mean of quotient; If :attr:`reduction` is ``'sum'``, return the sum of loss; If :attr:`reduction` is ``'none'``, no reduction will be applied. Default is ``'mean'``.
1054
        norm_by_times (bool, default False) – Whether to normalize the gradients by the number of time-step, which is also the sequence’s length. There is no need to normalize the gradients if reduction mode is 'mean'.
H
Hui Zhang 已提交
1055

1056 1057
    Returns:
        Tensor, The Connectionist Temporal Classification (CTC) loss between ``log_probs`` and  ``labels``. If attr:`reduction` is ``'none'``, the shape of loss is [batch_size], otherwise, the shape of loss is [1]. Data type is the same as ``log_probs``.
1058

1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
    Examples:

        .. code-block:: python

            # declarative mode
            import paddle.nn.functional as F
            import numpy as np
            import paddle

            # length of the longest logit sequence
            max_seq_length = 4
            #length of the longest label sequence
            max_label_length = 3
            # number of logit sequences
            batch_size = 2
            # class num
            class_num = 3

            np.random.seed(1)
            log_probs = np.array([[[4.17021990e-01, 7.20324516e-01, 1.14374816e-04],
                                    [3.02332580e-01, 1.46755889e-01, 9.23385918e-02]],

                                    [[1.86260208e-01, 3.45560730e-01, 3.96767467e-01],
                                    [5.38816750e-01, 4.19194520e-01, 6.85219526e-01]],

                                    [[2.04452246e-01, 8.78117442e-01, 2.73875929e-02],
                                    [6.70467496e-01, 4.17304814e-01, 5.58689833e-01]],

                                    [[1.40386939e-01, 1.98101491e-01, 8.00744593e-01],
                                    [9.68261600e-01, 3.13424170e-01, 6.92322612e-01]],

                                    [[8.76389146e-01, 8.94606650e-01, 8.50442126e-02],
                                    [3.90547849e-02, 1.69830427e-01, 8.78142476e-01]]]).astype("float32")
            labels = np.array([[1, 2, 2],
                            [1, 2, 2]]).astype("int32")
            input_lengths = np.array([5, 5]).astype("int64")
            label_lengths = np.array([3, 3]).astype("int64")

1097 1098 1099 1100
            log_probs = paddle.to_tensor(log_probs)
            labels = paddle.to_tensor(labels)
            input_lengths = paddle.to_tensor(input_lengths)
            label_lengths = paddle.to_tensor(label_lengths)
1101

1102 1103 1104 1105
            loss = F.ctc_loss(log_probs, labels,
                input_lengths,
                label_lengths,
                blank=0,
1106
                reduction='none')
1107
            print(loss)  #[3.9179852 2.9076521]
1108

1109 1110 1111 1112 1113
            loss = F.ctc_loss(log_probs, labels,
                input_lengths,
                label_lengths,
                blank=0,
                reduction='mean')
1114
            print(loss)  #[1.1376063]
1115 1116 1117

    """

1118
    loss_out = fluid.layers.warpctc(log_probs, labels, blank, norm_by_times,
H
Hui Zhang 已提交
1119
                                    input_lengths, label_lengths)
1120

Z
zhiboniu 已提交
1121
    loss_out = paddle.squeeze(loss_out, [-1])
1122 1123
    assert reduction in ['mean', 'sum', 'none']
    if reduction == 'mean':
S
ShenLiang 已提交
1124
        loss_out = paddle.mean(loss_out / label_lengths)
1125 1126 1127 1128 1129
    elif reduction == 'sum':
        loss_out = paddle.sum(loss_out)
    return loss_out


1130 1131 1132 1133 1134 1135 1136 1137 1138
def margin_cross_entropy(logits,
                         label,
                         margin1=1.0,
                         margin2=0.5,
                         margin3=0.0,
                         scale=64.0,
                         group=None,
                         return_softmax=False,
                         reduction='mean'):
1139
    r"""
1140 1141
    .. math::

1142
        L=-\frac{1}{N}\sum^N_{i=1}\log\frac{e^{s(cos(m_{1}\theta_{y_i}+m_{2})-m_{3})}}{e^{s(cos(m_{1}\theta_{y_i}+m_{2})-m_{3})}+\sum^n_{j=1,j\neq y_i} e^{scos\theta_{y_i}}}
1143

1144
    where the :math:`\theta_{y_i}` is the angle between the feature :math:`x` and
1145 1146 1147 1148
    the representation of class :math:`i`. The details of ArcFace loss
    could be referred to https://arxiv.org/abs/1801.07698.

    .. hint::
1149 1150 1151 1152 1153 1154
        The API supports single GPU and multi GPU, and don't supports CPU.

        For data parallel mode, set ``group=False``.

        For model parallel mode, set ``group=None`` or the group instance return by paddle.distributed.new_group.
        And logits.shape[-1] can be different at each rank.
1155 1156

    Args:
G
Guoxia Wang 已提交
1157
        logits (Tensor): shape[N, local_num_classes], the output of the normalized X multiply the normalized W.
1158
                The logits is shard_logits when using model parallel.
G
Guoxia Wang 已提交
1159 1160 1161 1162 1163
        label (Tensor): shape[N] or shape[N, 1], the groud truth label.
        margin1 (float, optional): m1 of margin loss, default value is `1.0`.
        margin2 (float, optional): m2 of margin loss, default value is `0.5`.
        margin3 (float, optional): m3 of margin loss, default value is `0.0`.
        scale (float, optional): s of margin loss, default value is `64.0`.
1164 1165 1166
        group (Group, optional): The group instance return by paddle.distributed.new_group 
            or ``None`` for global default group or ``False`` for data parallel (do not communication cross ranks).
            Default is ``None``.
1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184
        return_softmax (bool, optional): Whether return softmax probability. Default value is `False`.
        reduction (str, optional): The candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
                    If :attr:`reduction` is ``'mean'``, return the average of loss;
                    If :attr:`reduction` is ``'sum'``, return the sum of loss;
                    If :attr:`reduction` is ``'none'``, no reduction will be applied.
                    Default value is `'mean'`.

    Returns:
        ``Tensor`` or Tuple of two ``Tensor`` : Return the cross entropy loss if \
            `return_softmax` is False, otherwise the tuple \
            (loss, softmax), softmax is shard_softmax when \
            using model parallel, otherwise softmax is in \
            the same shape with input logits. If ``reduction == None``, \
            the shape of loss is ``[N, 1]``, otherwise the shape is ``[1]``.

    Examples:

    .. code-block:: python
G
Guoxia Wang 已提交
1185
        :name: code-example1
1186 1187 1188 1189 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 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233

        # required: gpu
        # Single GPU
        import paddle
        m1 = 1.0
        m2 = 0.5
        m3 = 0.0
        s = 64.0
        batch_size = 2
        feature_length = 4
        num_classes = 4

        label = paddle.randint(low=0, high=num_classes, shape=[batch_size], dtype='int64')

        X = paddle.randn(
            shape=[batch_size, feature_length],
            dtype='float64')
        X_l2 = paddle.sqrt(paddle.sum(paddle.square(X), axis=1, keepdim=True))
        X = paddle.divide(X, X_l2)

        W = paddle.randn(
            shape=[feature_length, num_classes],
            dtype='float64')
        W_l2 = paddle.sqrt(paddle.sum(paddle.square(W), axis=0, keepdim=True))
        W = paddle.divide(W, W_l2)

        logits = paddle.matmul(X, W)
        loss, softmax = paddle.nn.functional.margin_cross_entropy(
            logits, label, margin1=m1, margin2=m2, margin3=m3, scale=s, return_softmax=True, reduction=None)

        print(logits)
        print(label)
        print(loss)
        print(softmax)
        
        #Tensor(shape=[2, 4], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[ 0.85204151, -0.55557678,  0.04994566,  0.71986042],
        #        [-0.20198586, -0.35270476, -0.55182702,  0.09749021]])
        #Tensor(shape=[2], dtype=int64, place=CUDAPlace(0), stop_gradient=True,
        #       [2, 3])
        #Tensor(shape=[2, 1], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[82.37059586],
        #        [12.13448420]])
        #Tensor(shape=[2, 4], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[0.99978819, 0.00000000, 0.00000000, 0.00021181],
        #        [0.99992995, 0.00006468, 0.00000000, 0.00000537]])

    .. code-block:: python
G
Guoxia Wang 已提交
1234
        :name: code-example2
1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324

        # required: distributed
        # Multi GPU, test_margin_cross_entropy.py
        import paddle
        import paddle.distributed as dist
        strategy = dist.fleet.DistributedStrategy()
        dist.fleet.init(is_collective=True, strategy=strategy)
        rank_id = dist.get_rank()
        m1 = 1.0
        m2 = 0.5
        m3 = 0.0
        s = 64.0
        batch_size = 2
        feature_length = 4
        num_class_per_card = [4, 8]
        num_classes = paddle.sum(paddle.to_tensor(num_class_per_card))

        label = paddle.randint(low=0, high=num_classes.item(), shape=[batch_size], dtype='int64')
        label_list = []
        dist.all_gather(label_list, label)
        label = paddle.concat(label_list, axis=0)

        X = paddle.randn(
            shape=[batch_size, feature_length],
            dtype='float64')
        X_list = []
        dist.all_gather(X_list, X)
        X = paddle.concat(X_list, axis=0)
        X_l2 = paddle.sqrt(paddle.sum(paddle.square(X), axis=1, keepdim=True))
        X = paddle.divide(X, X_l2)

        W = paddle.randn(
            shape=[feature_length, num_class_per_card[rank_id]],
            dtype='float64')
        W_l2 = paddle.sqrt(paddle.sum(paddle.square(W), axis=0, keepdim=True))
        W = paddle.divide(W, W_l2)

        logits = paddle.matmul(X, W)
        loss, softmax = paddle.nn.functional.margin_cross_entropy(
            logits, label, margin1=m1, margin2=m2, margin3=m3, scale=s, return_softmax=True, reduction=None)

        print(logits)
        print(label)
        print(loss)
        print(softmax)

        # python -m paddle.distributed.launch --gpus=0,1 test_margin_cross_entropy.py 
        ## for rank0 input
        #Tensor(shape=[4, 4], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[ 0.32888934,  0.02408748, -0.02763289,  0.18173063],
        #        [-0.52893978, -0.10623845, -0.21596515, -0.06432517],
        #        [-0.00536345, -0.03924667,  0.66735314, -0.28640926],
        #        [-0.09907366, -0.48534973, -0.10365338, -0.39472322]])
        #Tensor(shape=[4], dtype=int64, place=CUDAPlace(0), stop_gradient=True,
        #       [11, 1 , 10, 11])

        ## for rank1 input
        #Tensor(shape=[4, 8], dtype=float64, place=CUDAPlace(1), stop_gradient=True,
        #       [[ 0.68654754,  0.28137170,  0.69694954, -0.60923933, -0.57077653,  0.54576703, -0.38709028,  0.56028204],
        #        [-0.80360371, -0.03042448, -0.45107338,  0.49559349,  0.69998950, -0.45411693,  0.61927630, -0.82808600],
        #        [ 0.11457570, -0.34785879, -0.68819499, -0.26189226, -0.48241491, -0.67685711,  0.06510185,  0.49660849],
        #        [ 0.31604851,  0.52087884,  0.53124749, -0.86176582, -0.43426329,  0.34786144, -0.10850784,  0.51566383]])
        #Tensor(shape=[4], dtype=int64, place=CUDAPlace(1), stop_gradient=True,
        #       [11, 1 , 10, 11])

        ## for rank0 output
        #Tensor(shape=[4, 1], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[38.96608230],
        #        [81.28152394],
        #        [69.67229865],
        #        [31.74197251]])
        #Tensor(shape=[4, 4], dtype=float64, place=CUDAPlace(0), stop_gradient=True,
        #       [[0.00000000, 0.00000000, 0.00000000, 0.00000000],
        #        [0.00000000, 0.00000000, 0.00000000, 0.00000000],
        #        [0.00000000, 0.00000000, 0.99998205, 0.00000000],
        #        [0.00000000, 0.00000000, 0.00000000, 0.00000000]])
        ## for rank1 output
        #Tensor(shape=[4, 1], dtype=float64, place=CUDAPlace(1), stop_gradient=True,
        #       [[38.96608230],
        #        [81.28152394],
        #        [69.67229865],
        #        [31.74197251]])
        #Tensor(shape=[4, 8], dtype=float64, place=CUDAPlace(1), stop_gradient=True,
        #       [[0.33943993, 0.00000000, 0.66051859, 0.00000000, 0.00000000, 0.00004148, 0.00000000, 0.00000000],
        #        [0.00000000, 0.00000000, 0.00000000, 0.00000207, 0.99432097, 0.00000000, 0.00567696, 0.00000000],
        #        [0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00000000, 0.00001795],
        #        [0.00000069, 0.33993085, 0.66006319, 0.00000000, 0.00000000, 0.00000528, 0.00000000, 0.00000000]])
    """

    assert reduction in ['mean', 'sum', 'none', None]
1325 1326 1327 1328 1329 1330 1331
    if not (group == False or group is None or hasattr(group, 'is_member')):
        raise ValueError(
            'Expected group is False, None or instance of paddle.distributed.collective.Group \
             (got group: {})'.format(group))
        return

    if hasattr(group, 'is_member') and not group.is_member():
1332 1333
        return

1334
    ring_id = 0
1335 1336
    rank = 0
    nranks = 1
1337 1338 1339 1340 1341 1342 1343 1344
    if group != False:
        ring_id = 0 if group is None else group.id
        if core.is_compiled_with_dist():
            parallel_env = paddle.distributed.ParallelEnv()
            global_rank = parallel_env.rank
            rank = global_rank if group is None else group.get_group_rank(
                global_rank)
            nranks = parallel_env.world_size if group is None else group.nranks
1345 1346 1347 1348 1349

    input_dims = len(list(logits.shape))
    label_dims = len(list(label.shape))
    if input_dims - 1 != label_dims and input_dims != label_dims:
        raise ValueError(
1350
            'Expected input_dims - 1 = label_dims or input_dims == label_dims\
1351 1352 1353 1354
             (got nput_dims{}, label_dims{})'.format(input_dims, label_dims))
    if input_dims - 1 == label_dims:
        label = paddle.unsqueeze(label, axis=-1)

Z
zhiboniu 已提交
1355
    if in_dynamic_mode():
1356
        softmax, loss = _C_ops.margin_cross_entropy(
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
            logits, label, 'ring_id', ring_id, 'rank', rank, 'nranks', nranks,
            'margin1', margin1, 'margin2', margin2, 'margin3', margin3, 'scale',
            scale, 'return_softmax', return_softmax)
        if reduction == 'mean':
            loss = paddle.mean(loss)
        elif reduction == 'sum':
            loss = paddle.sum(loss)
        if not return_softmax:
            return loss
        else:
            return loss, softmax

    op_type = 'margin_cross_entropy'
    helper = LayerHelper(op_type, **locals())
    softmax = helper.create_variable_for_type_inference(dtype=logits.dtype)
    loss = helper.create_variable_for_type_inference(dtype=logits.dtype)

    check_variable_and_dtype(logits, 'logits',
                             ['float16', 'float32', 'float64'],
                             'margin_cross_entropy')
    check_variable_and_dtype(label, 'label', ['int32', 'int64'],
                             'margin_cross_entropy')

    helper.append_op(
        type=op_type,
        inputs={'Logits': logits,
                'Label': label},
        outputs={'Softmax': softmax,
                 'Loss': loss},
        attrs={
            'return_softmax': return_softmax,
            'ring_id': ring_id,
            'rank': rank,
            'nranks': nranks,
            'margin1': margin1,
            'margin2': margin2,
            'margin3': margin3,
            'scale': scale,
        })

    if reduction == 'mean':
        loss = paddle.mean(loss)
    elif reduction == 'sum':
        loss = paddle.sum(loss)

    if not return_softmax:
        return loss
    else:
        return loss, softmax


1408 1409 1410 1411 1412 1413 1414
@deprecated(
    since="2.0.0",
    update_to="paddle.nn.functional.cross_entropy",
    level=1,
    reason=(
        'Please notice that behavior of "paddle.nn.functional.softmax_with_cross_entropy" '
        'and "paddle.nn.functional.cross_entropy" is different.'))
1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426
def softmax_with_cross_entropy(logits,
                               label,
                               soft_label=False,
                               ignore_index=-100,
                               numeric_stable_mode=True,
                               return_softmax=False,
                               axis=-1):
    return fluid_softmax_with_cross_entropy(logits, label, soft_label,
                                            ignore_index, numeric_stable_mode,
                                            return_softmax, axis)


1427 1428 1429 1430
def cross_entropy(input,
                  label,
                  weight=None,
                  ignore_index=-100,
1431 1432 1433
                  reduction='mean',
                  soft_label=False,
                  axis=-1,
1434
                  use_softmax=True,
1435
                  name=None):
1436
    r"""
H
HydrogenSulfate 已提交
1437 1438 1439
    By default, this operator implements the cross entropy loss function with softmax. This function 
    combines the calculation of the softmax operation and the cross entropy loss function 
    to provide a more numerically stable computing. 
1440

1441
    This operator will calculate the cross entropy loss function without softmax when use_softmax=False.
1442

H
HydrogenSulfate 已提交
1443 1444
    By default, this operator will calculate the mean of the result, and you can also affect 
    the default behavior by using the reduction parameter. Please refer to the part of 
1445
    parameters for details.
1446

1447
    This operator can be used to calculate the softmax cross entropy loss with soft and hard labels.
H
HydrogenSulfate 已提交
1448
    Where, the hard labels mean the actual label value, 0, 1, 2, etc.  And the soft labels 
1449
    mean the probability of the actual label, 0.6, 0.8, 0.2, etc.
1450

1451
    The calculation of this operator includes the following two steps.
1452

1453
    - **1.softmax cross entropy**
1454

1455
        1. Hard label (each sample can only be assigned into one category)
1456

1457
        1.1. when use_softmax=True
1458

1459 1460
            .. math::
              \\loss_j=-\text{logits}_{label_j}+\log\left(\sum_{i=0}^{C}\exp(\text{logits}_i)\right) , j = 1,...,N
1461

1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502
            where, N is the number of samples and C is the number of categories.

        1.2. when use_softmax=False

            .. math::
              \\loss_j=-\log\left({P}_{label_j}\right) , j = 1,...,N

            where, N is the number of samples and C is the number of categories, P is input(the output of softmax).


        2. Soft label (each sample is assigned to multiple categories with a certain probability, and the probability sum is 1).

        2.1. when use_softmax=True

            .. math::
              \\loss_j=-\sum_{i=0}^{C}\text{label}_i\left(\text{logits}_i-\log\left(\sum_{i=0}^{C}\exp(\text{logits}_i)\right)\right) , j = 1,...,N

            where, N is the number of samples and C is the number of categories.

        2.2. when use_softmax=False

            .. math::
              \\loss_j=-\sum_{j=0}^{C}\left({label}_j*\log\left({P}_{label_j}\right)\right) , j = 1,...,N

            where, N is the number of samples and C is the number of categories, P is input(the output of softmax).




    - **2. Weight and reduction processing**

        1. Weight

            If the ``weight`` parameter is ``None`` , go to the next step directly.

            If the ``weight`` parameter is not ``None`` , the cross entropy of each sample is weighted by weight
            according to soft_label = False or True as follows.

            1.1. Hard labels (soft_label = False)

            .. math::
H
HydrogenSulfate 已提交
1503
                \\loss_j=loss_j*weight[label_j] 
1504

1505

1506 1507 1508 1509 1510 1511 1512
            1.2. Soft labels (soft_label = True)

             .. math::
                \\loss_j=loss_j*\sum_{i}\left(weight[label_i]*logits_i\right)

        2. reduction

H
HydrogenSulfate 已提交
1513
            2.1 if the ``reduction`` parameter is ``none`` 
1514 1515 1516

                Return the previous result directly

H
HydrogenSulfate 已提交
1517
            2.2 if the ``reduction`` parameter is ``sum`` 
1518 1519 1520 1521 1522 1523

                Return the sum of the previous results

            .. math::
               \\loss=\sum_{j}loss_j

H
HydrogenSulfate 已提交
1524 1525
            2.3 if the ``reduction`` parameter is ``mean`` , it will be processed according to 
            the ``weight`` parameter as follows. 
1526

H
HydrogenSulfate 已提交
1527
            2.3.1. If the  ``weight``  parameter is ``None`` 
1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540

                   Return the average value of the previous results

             .. math::
                \\loss=\sum_{j}loss_j/N

                  where, N is the number of samples and C is the number of categories.

            2.3.2. If the 'weight' parameter is not 'None', the weighted average value of the previous result will be returned

            1. Hard labels (soft_label = False)

             .. math::
H
HydrogenSulfate 已提交
1541
                \\loss=\sum_{j}loss_j/\sum_{j}weight[label_j] 
1542 1543 1544 1545 1546

            2. Soft labels (soft_label = True)

             .. math::
                \\loss=\sum_{j}loss_j/\sum_{j}\left(\sum_{i}weight[label_i]\right)
H
HydrogenSulfate 已提交
1547 1548
 
 
1549
    Parameters:
1550 1551 1552 1553

        - **input** (Tensor)

            Input tensor, the data type is float32, float64. Shape is
H
HydrogenSulfate 已提交
1554
	    :math:`[N_1, N_2, ..., N_k, C]`, where C is number of classes ,  ``k >= 1`` . 
1555

H
HydrogenSulfate 已提交
1556
            Note: 
1557

H
HydrogenSulfate 已提交
1558
                1. when use_softmax=True, it expects unscaled logits. This operator should not be used with the 
1559 1560 1561
                output of softmax operator, which will produce incorrect results.

                2. when use_softmax=False, it expects the output of softmax operator.
H
HydrogenSulfate 已提交
1562
 
1563 1564 1565 1566 1567 1568
        - **label** (Tensor)

            1. If soft_label=False, the shape is
            :math:`[N_1, N_2, ..., N_k]` or :math:`[N_1, N_2, ..., N_k, 1]`, k >= 1.
            the data type is int32, int64, float32, float64, where each value is [0, C-1].

H
HydrogenSulfate 已提交
1569
            2. If soft_label=True, the shape and data type should be same with ``input`` , 
1570 1571 1572 1573
            and the sum of the labels for each sample should be 1.

        - **weight** (Tensor, optional)

H
HydrogenSulfate 已提交
1574 1575
            a manual rescaling weight given to each class. 
            If given, has to be a Tensor of size C and the data type is float32, float64. 
1576 1577 1578 1579 1580
            Default is ``'None'`` .

        - **ignore_index** (int64, optional)

            Specifies a target value that is ignored
H
HydrogenSulfate 已提交
1581 1582
            and does not contribute to the loss. A negative value means that no label 
            value needs to be ignored. Only valid when soft_label = False.  
1583 1584 1585 1586 1587
            Default is ``-100`` .

        - **reduction** (str, optional)

            Indicate how to average the loss by batch_size,
1588 1589
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
H
Hui Zhang 已提交
1590
            If :attr:`size_average` is ``'sum'``, the reduced sum loss is returned.
1591 1592
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned.
            Default is ``'mean'``.
1593

1594 1595
        - **soft_label** (bool, optional)

H
HydrogenSulfate 已提交
1596
            Indicate whether label is soft. 
1597 1598 1599 1600
            Default is ``False``.

        - **axis** (int, optional)

H
HydrogenSulfate 已提交
1601 1602 1603
            The index of dimension to perform softmax calculations. 
            It should be in range :math:`[-1, rank - 1]`, where :math:`rank` is the 
            number of dimensions of input :attr:`input`. 
1604 1605 1606 1607 1608 1609 1610
            Default is ``-1`` .

        - **use_softmax** (bool, optional)

            Indicate whether compute softmax before cross_entropy.
            Default is ``True``.

Z
zhiboniu 已提交
1611
        - **name** (str, optional)
1612 1613 1614

            The name of the operator. Default is ``None`` .
            For more information, please refer to :ref:`api_guide_Name` .
1615 1616 1617

    Returns:

1618 1619
        Tensor. Return the softmax cross_entropy loss of ``input`` and ``label``.
        The data type is the same as input.
1620

1621
        If :attr:`reduction` is ``'mean'`` or ``'sum'`` , the dimension of return value is ``1``.
1622

1623
        If :attr:`reduction` is ``'none'``:
C
Chen Long 已提交
1624

H
HydrogenSulfate 已提交
1625
        1. If soft_label = False, the dimension of return value is the same with ``label`` . 
C
Chen Long 已提交
1626

H
HydrogenSulfate 已提交
1627
        2. if soft_label = True, the dimension of return value is :math:`[N_1, N_2, ..., N_k, 1]` . 
1628 1629 1630 1631 1632


     Example1(hard labels):

        .. code-block:: python
H
HydrogenSulfate 已提交
1633
            
1634 1635 1636 1637 1638
            import paddle
            paddle.seed(99999)
            N=100
            C=200
            reduction='mean'
H
HydrogenSulfate 已提交
1639
            input =  paddle.rand([N, C], dtype='float64')  
1640
            label =  paddle.randint(0, C, shape=[N], dtype='int64')
H
HydrogenSulfate 已提交
1641 1642
            weight = paddle.rand([C], dtype='float64') 
            
1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653
            cross_entropy_loss = paddle.nn.loss.CrossEntropyLoss(
                weight=weight, reduction=reduction)
            dy_ret = cross_entropy_loss(
                                       input,
                                       label)
            print(dy_ret.numpy()) #[5.41993642]


    Example2(soft labels):

        .. code-block:: python
H
HydrogenSulfate 已提交
1654
            
1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667
            import paddle
            paddle.seed(99999)
            axis = -1
            ignore_index = -100
            N = 4
            C = 3
            shape = [N, C]
            reduction='mean'
            weight = None
            logits = paddle.uniform(shape, dtype='float64', min=0.1, max=1.0)
            labels = paddle.uniform(shape, dtype='float64', min=0.1, max=1.0)
            labels /= paddle.sum(labels, axis=axis, keepdim=True)
            paddle_loss_mean = paddle.nn.functional.cross_entropy(
H
HydrogenSulfate 已提交
1668 1669 1670
                                                                  logits,  
                                                                  labels, 
                                                                  soft_label=True, 
1671 1672 1673 1674
                                                                  axis=axis,
                                                                  weight=weight,
                                                                  reduction=reduction)
            print(paddle_loss_mean.numpy()) #[1.12908343]
C
Chen Long 已提交
1675

1676 1677 1678 1679
    """

    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
1680 1681 1682
            "The value of 'reduction' in softmax_cross_entropy"
            "should be 'sum', 'mean' or 'none', but received %s, which is not allowed."
            % reduction)
1683 1684 1685 1686 1687 1688
    if ignore_index > 0 and soft_label == True:
        raise ValueError(
            "When soft_label == True, the value of 'ignore_index' in softmax_cross_entropy"
            "should be '-100', but received %s, which is not allowed." %
            ignore_index)

1689
    input_dims = len(list(input.shape))
1690 1691 1692
    if input_dims == 0:
        raise ValueError('The dimention of input should be larger than zero!')

1693 1694
    label_dims = len(list(label.shape))
    if input_dims - 1 != label_dims and input_dims != label_dims:
1695
        raise ValueError(
1696 1697 1698 1699
            'Expected nput_dims - 1 = label_dims or input_dims == label_dims\
             (got nput_dims{}, label_dims{})'.format(input_dims, label_dims))
    if input_dims - 1 == label_dims:
        label = paddle.unsqueeze(label, axis=axis)
Z
zhiboniu 已提交
1700
    if in_dynamic_mode():
H
HydrogenSulfate 已提交
1701
        if soft_label == False:
H
HydrogenSulfate 已提交
1702 1703
            valid_label = paddle.cast(
                label != ignore_index, dtype=label.dtype) * label
H
HydrogenSulfate 已提交
1704 1705 1706
            label_min = paddle.min(valid_label)
            label_max = paddle.max(valid_label)
            if label_min < 0:
1707 1708
                raise ValueError("Target {} is out of lower bound.".format(
                    label_min.item()))
H
HydrogenSulfate 已提交
1709
            if label_max >= input.shape[axis]:
1710 1711
                raise ValueError("Target {} is out of upper bound.".format(
                    label_max.item()))
F
fwenguang 已提交
1712
        if core.is_compiled_with_npu() or core.is_compiled_with_mlu():
1713 1714 1715 1716 1717 1718 1719 1720 1721
            _, _, out = _C_ops.softmax_with_cross_entropy(
                input, label, 'soft_label', soft_label, 'ignore_index',
                ignore_index, 'numeric_stable_mode', True, 'axis', axis,
                'use_softmax', use_softmax)
        else:
            _, out = _C_ops.softmax_with_cross_entropy(
                input, label, 'soft_label', soft_label, 'ignore_index',
                ignore_index, 'numeric_stable_mode', True, 'axis', axis,
                'use_softmax', use_softmax)
1722

1723
        if weight is not None:
1724

H
HydrogenSulfate 已提交
1725
            # trans weight from class to sample, shape:N or [N,H,W] for 1d and 2d cases.
1726 1727
            if soft_label == True:
                # chajchaj:
H
HydrogenSulfate 已提交
1728
                # weight's shape is C, where C is class num.
1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739
                # for 1d case: label's shape is [N,C], weight_gather's shape is N.
                # for 2d case: label's shape is [N,H,W,C], weight_gather's shape is [N,H,W].
                weight_gather = paddle.matmul(
                    x=paddle.cast(label, weight.dtype),
                    y=weight,
                    transpose_x=False,
                    transpose_y=True)
                out_shape = list(out.shape)
                weight_gather_reshape = reshape(weight_gather, shape=out_shape)
                out = paddle.cast(out, weight_gather_reshape.dtype)

W
wanghuancoder 已提交
1740
                out = _C_ops.elementwise_mul(out, weight_gather_reshape)
1741 1742

            else:
1743 1744 1745 1746 1747 1748 1749
                if input.shape[axis] != weight.shape[-1]:
                    raise ValueError(
                        "input's class_dimension({}) must equal to "
                        "weight's class_dimension({}) "
                        "when weight is provided" \
                            .format(input.shape[axis], weight.shape[-1]))

H
HydrogenSulfate 已提交
1750 1751
                ignore_weight_mask = paddle.cast((label != ignore_index),
                                                 out.dtype)
H
HydrogenSulfate 已提交
1752
                if ignore_weight_mask.ndim > 1 and ignore_weight_mask.shape[
1753
                        axis] == 1:
H
HydrogenSulfate 已提交
1754
                    # TODO: Temporarily use squeeze instead of squeeze_
H
HydrogenSulfate 已提交
1755 1756
                    ignore_weight_mask = paddle.squeeze(ignore_weight_mask,
                                                        axis)
H
HydrogenSulfate 已提交
1757
                if axis != -1 and axis != valid_label.ndim - 1:
1758
                    temp_perm = list(range(axis % valid_label.ndim)) \
1759
                                + list(range((axis % valid_label.ndim + 1), valid_label.ndim)) \
H
HydrogenSulfate 已提交
1760
                                + [axis % valid_label.ndim]
1761 1762 1763 1764
                    weight_gather = _C_ops.gather_nd(
                        weight, valid_label.transpose(temp_perm))
                else:
                    weight_gather = _C_ops.gather_nd(weight, valid_label)
H
HydrogenSulfate 已提交
1765 1766
                weight_gather = _C_ops.elementwise_mul(weight_gather,
                                                       ignore_weight_mask)
1767 1768 1769 1770
                input_shape = list(label.shape)
                weight_gather_reshape = reshape(
                    weight_gather, shape=input_shape)
                out = paddle.cast(out, weight_gather_reshape.dtype)
W
wanghuancoder 已提交
1771
                out = _C_ops.elementwise_mul(out, weight_gather_reshape)
1772

1773
        if reduction == "sum":
H
HydrogenSulfate 已提交
1774
            #   because of fluid_softmax_with_cross_entropy op's inner logic,
1775 1776
            #   in the out tensor of this op, the loss of sample with class_index==ignore_index is 0
            #   so, reduce_sum all directly is ok
W
wanghuancoder 已提交
1777
            return _C_ops.reduce_sum(out, 'reduce_all', True)
1778
        elif reduction == "mean":
H
HydrogenSulfate 已提交
1779 1780 1781 1782 1783 1784
            # 1. if weight==none,
            #     numerator: reduce_sum all loss directly is ok causeof fluid_softmax_with_cross_entropy's inner logic
            #     denominator: count sample num with class_index!=ignore_index
            # 2. else
            #     numerator: loss's weighted sum
            #     denominator: cal the sum of weight where the sample's class_index!=ignore_index
1785
            if ignore_index != -100:
W
wanghuancoder 已提交
1786
                out_sum = _C_ops.reduce_sum(out, 'reduce_all', True)
H
HydrogenSulfate 已提交
1787 1788 1789
                # for each label[i],set 1 or 0, according to ignore_index
                # mask[i]=0, if label[i]==ignore_index
                # mask[i]=1, otherwise
1790
                mask = (label != ignore_index)
1791
                if weight is None:
1792
                    mask = paddle.cast(mask, dtype=out_sum.dtype)
W
wanghuancoder 已提交
1793
                    count = _C_ops.reduce_sum(mask, 'reduce_all', True)
1794
                    ret = out_sum / (count + (count == 0.0))
1795 1796
                else:
                    mask = paddle.cast(mask, weight_gather_reshape.dtype)
W
wanghuancoder 已提交
1797
                    weight_ignored = _C_ops.elementwise_mul(
1798
                        mask, weight_gather_reshape)
W
wanghuancoder 已提交
1799 1800
                    weight_sum = _C_ops.reduce_sum(weight_ignored, 'reduce_all',
                                                   True)
1801
                    ret = out_sum / (weight_sum + (weight_sum == 0.0))
1802 1803
                return ret
            elif weight is not None:
W
wanghuancoder 已提交
1804 1805 1806
                out_sum = _C_ops.reduce_sum(out, 'reduce_all', True)
                total_weight = _C_ops.reduce_sum(weight_gather_reshape,
                                                 'reduce_all', True)
1807
                return out_sum / (total_weight + (total_weight == 0.0))
1808
            else:
W
wanghuancoder 已提交
1809
                return _C_ops.mean(out)
1810

1811
        else:
1812 1813
            if input_dims - 1 == label_dims:
                out = paddle.squeeze(out, axis=axis)
1814
            return out
1815

1816 1817 1818
    fluid.data_feeder.check_variable_and_dtype(
        input, 'input', ['float32', 'float64'], 'softmax_cross_entropy')
    fluid.data_feeder.check_variable_and_dtype(
1819 1820
        label, 'label',
        ['uint8', 'int8', 'int16', 'int32', 'int64', 'float32', 'float64'],
1821
        'softmax_cross_entropy')
1822 1823 1824 1825 1826
    attrs = {
        'soft_label': soft_label,
        'ignore_index': ignore_index,
        'numeric_stable_mode': True,
        'axis': axis,
1827
        'use_softmax': use_softmax
1828 1829 1830 1831
    }
    helper = LayerHelper('softmax_with_cross_entropy', **locals())
    softmax = helper.create_variable_for_type_inference(dtype=input.dtype)
    out = helper.create_variable_for_type_inference(dtype=input.dtype)
1832 1833 1834 1835 1836

    outputs = {'Softmax': softmax, 'Loss': out}
    if core.is_compiled_with_npu() or core.is_compiled_with_mlu():
        backprop = helper.create_variable_for_type_inference(dtype=input.dtype)
        outputs['Backprop'] = backprop
1837 1838 1839 1840
    helper.append_op(
        type='softmax_with_cross_entropy',
        inputs={'Logits': input,
                'Label': label},
1841
        outputs=outputs,
1842 1843
        attrs=attrs)

1844
    if weight is not None:
1845 1846 1847
        fluid.data_feeder.check_variable_and_dtype(
            weight, 'weight', ['float32', 'float64'], 'softmax_cross_entropy')
        weight_name = name if reduction == 'none' else None
1848 1849
        if soft_label == True:
            # chajchaj:
H
HydrogenSulfate 已提交
1850
            # trans weight from class to sample, shape:N or [N,H,W] for 1d and 2d cases.
1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863
            # weight's shape is C, where C is class num.
            # for 1d case: label's shape is [N,C], weight_gather's shape is N.
            # for 2d case: label's shape is [N,H,W,C], weight_gather's shape is [N,H,W].
            weight_gather = paddle.matmul(
                x=paddle.cast(label, weight.dtype),
                y=weight,
                transpose_x=False,
                transpose_y=True)

            out_shape = list(out.shape)
            weight_gather_reshape = reshape(weight_gather, shape=out_shape)
            out = paddle.cast(out, weight_gather_reshape.dtype)
        else:
1864 1865
            if input.shape[axis] != weight.shape[-1]:
                raise ValueError("input's class_dimension({}) must equal to "
1866 1867
                                 "weight's class_dimension({}) "
                                 "when weight is provided" \
1868
                                 .format(input.shape[axis], weight.shape[-1]))
H
HydrogenSulfate 已提交
1869

H
HydrogenSulfate 已提交
1870 1871 1872 1873 1874
            valid_label = paddle.multiply(
                paddle.cast(
                    label != ignore_index, dtype=label.dtype), label)
            ignore_weight_mask = paddle.cast((label != ignore_index),
                                             input.dtype)
H
HydrogenSulfate 已提交
1875
            if ignore_weight_mask.ndim > 1 and ignore_weight_mask.shape[
1876 1877
                    axis] == 1:
                ignore_weight_mask = paddle.squeeze(ignore_weight_mask, axis)
H
HydrogenSulfate 已提交
1878
            if axis != -1 and axis != valid_label.ndim - 1:
1879
                temp_perm = list(range(axis % valid_label.ndim)) \
H
HydrogenSulfate 已提交
1880
                            + list(range((axis % valid_label.ndim + 1), valid_label.ndim)) \
1881 1882 1883 1884 1885
                            + [axis % valid_label.ndim]
                weight_gather = paddle.gather_nd(
                    weight, paddle.transpose(valid_label, temp_perm))
            else:
                weight_gather = paddle.gather_nd(weight, valid_label)
H
HydrogenSulfate 已提交
1886 1887
            weight_gather = paddle.multiply(weight_gather, ignore_weight_mask)

1888 1889
            input_shape = list(label.shape)
            weight_gather_reshape = reshape(weight_gather, shape=input_shape)
1890
        out = paddle.multiply(out, weight_gather_reshape, name=weight_name)
1891

1892 1893 1894
    if reduction == "sum":
        return paddle.sum(out, name=name)
    elif reduction == "mean":
1895 1896
        if ignore_index != -100:
            out_sum = paddle.sum(out, name=name)
H
HydrogenSulfate 已提交
1897 1898 1899
            # for each label[i],set 1 or 0, according to ignore_index
            # mask[i]=0, if label[i]==ignore_index
            # mask[i]=1, otherwise
1900 1901 1902 1903
            mask = (label != ignore_index)
            if (weight is None):
                mask = paddle.cast(mask, dtype=out_sum.dtype)
                count = paddle.sum(mask, name=name)
1904
                ret = out_sum / (count + (count == 0.0))
1905 1906 1907 1908
            else:
                mask = paddle.cast(mask, weight_gather_reshape.dtype)
                weight_ignored = paddle.multiply(mask, weight_gather_reshape)
                weight_sum = paddle.sum(weight_ignored, name=name)
1909
                ret = out_sum / (weight_sum + (weight_sum == 0.0))
1910 1911
            return ret
        elif weight is not None:
1912 1913
            out_sum = paddle.sum(out, name=name)
            total_weight = paddle.sum(weight_gather_reshape)
1914
            return out_sum / (total_weight + (total_weight == 0.0))
1915 1916
        else:
            return paddle.mean(out, name=name)
1917

1918
    else:
1919 1920 1921
        if input_dims - 1 == label_dims:
            out = paddle.squeeze(out, axis=axis)

1922
        return out
1923 1924 1925 1926 1927 1928 1929 1930 1931


def sigmoid_focal_loss(logit,
                       label,
                       normalizer=None,
                       alpha=0.25,
                       gamma=2.0,
                       reduction='sum',
                       name=None):
1932
    r"""
1933 1934 1935 1936 1937 1938
    `Focal Loss <https://arxiv.org/abs/1708.02002>`_ is proposed to address the
    foreground-background class imbalance for classification tasks. It down-weights
    easily-classified examples and thus focuses training on hard examples. For example,
    it is used in one-stage object detection where the foreground-background class
    imbalance is extremely high.

H
HydrogenSulfate 已提交
1939
    This operator measures focal loss function as follows: 
1940 1941

    .. math::
1942
           Out = -Labels * alpha * {(1 - \sigma(Logit))}^{gamma}\log(\sigma(Logit)) - (1 - Labels) * (1 - alpha) * {\sigma(Logit)}^{gamma}\log(1 - \sigma(Logit))
1943

H
HydrogenSulfate 已提交
1944
    We know that :math:`\sigma(Logit) = \frac{1}{1 + \exp(-Logit)}`. 
1945 1946 1947 1948 1949

    Then, if :attr:`normalizer` is not None, this operator divides the
    normalizer tensor on the loss `Out`:

    .. math::
1950
           Out = \frac{Out}{normalizer}
1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970

    Finally, this operator applies reduce operation on the loss.
    If :attr:`reduction` set to ``'none'``, the operator will return the original loss `Out`.
    If :attr:`reduction` set to ``'mean'``, the reduced mean loss is :math:`Out = MEAN(Out)`.
    If :attr:`reduction` set to ``'sum'``, the reduced sum loss is :math:`Out = SUM(Out)`.

    Note that the target ``label`` is 0 for the negative class and is 1 for the positive class.

    Args:
        logit (Tensor): The input logit tensor. The shape is [N, *], where N is batch_size,
            `*` means any number of additional dimensions. The ``logit`` is usually the
            output of a convolution layer. Available dtype is float32, float64.
        label (Tensor): The target label tensor with the same shape as
            ``logit``. The target label whose value should be numbers between 0 and 1.
            Available dtype is float32, float64.
        normalizer (Tensor, optional): The number normalizes the focal loss. It has to be
            a 1-D Tensor whose shape is `[1, ]`. The data type is float32, float64.
            For object detection task, it is the the number of positive samples.
            If set to None, the focal loss will not be normalized. Default is None.
        alpha(int|float, optional): Hyper-parameter to balance the positive and negative example,
H
HydrogenSulfate 已提交
1971
            it should be between 0 and 1.  Default value is set to 0.25. 
1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995
        gamma(int|float, optional): Hyper-parameter to modulate the easy and hard examples.
            Default value is set to 2.0.
        reduction (str, optional): Indicate how to average the loss by batch_size,
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned;
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
            If :attr:`reduction` is ``'sum'``, the summed loss is returned.
            Default is ``'sum'``.
        name (str, optional): Name for the operation (optional, default is None).
            For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        Tensor, if :attr:`reduction` is ``'mean'`` or ``'sum'``, the out shape is :math:`[1]`, otherwise the shape is the same as ``logit``. The same dtype as ``logit`` tensor.

    Examples:

        .. code-block:: python

            import paddle

            logit = paddle.to_tensor([[0.97, 0.91, 0.03], [0.55, 0.43, 0.71]], dtype='float32')
            label = paddle.to_tensor([[1.0, 0.0, 0.0], [0.0, 1.0, 0.0]], dtype='float32')
            one = paddle.to_tensor([1.], dtype='float32')
            fg_label = paddle.greater_equal(label, one)
1996
            fg_num = paddle.sum(paddle.cast(fg_label, dtype='float32'))
1997
            output = paddle.nn.functional.sigmoid_focal_loss(logit, label, normalizer=fg_num)
1998
            print(output)  # [0.65782464]
1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

    """
    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "The value of 'reduction' in sigmoid_focal_loss "
            "should be 'sum', 'mean' or 'none', but received %s, which is not allowed."
            % reduction)

    if normalizer is not None:
        fluid.data_feeder.check_variable_and_dtype(normalizer, 'normalizer',
                                                   ['float32', 'float64'],
                                                   'sigmoid_focal_loss')
        normalizer_shape = list(normalizer.shape)
        normalizer_dims = len(normalizer_shape)
        if normalizer_dims > 1:
            raise ValueError(
                "Expected one dimension of normalizer in sigmoid_focal_loss but got {}.".
                format(normalizer_dims))

Z
zhiboniu 已提交
2018
    if in_dynamic_mode():
2019
        one = _varbase_creator(dtype=logit.dtype)
W
wanghuancoder 已提交
2020 2021 2022 2023 2024 2025 2026 2027 2028 2029
        _C_ops.fill_constant(one, 'value',
                             float(1.0), 'force_cpu', False, 'dtype', one.dtype,
                             'str_value', '1.0', 'shape', logit.shape)
        loss = _C_ops.sigmoid_cross_entropy_with_logits(logit, label)
        pred = _C_ops.sigmoid(logit)
        p_t = _C_ops.elementwise_add(
            _C_ops.elementwise_mul(pred, label),
            _C_ops.elementwise_mul(
                _C_ops.elementwise_sub(one, pred),
                _C_ops.elementwise_sub(one, label)))
2030 2031

        alpha = fluid.dygraph.base.to_variable([alpha], dtype=loss.dtype)
W
wanghuancoder 已提交
2032 2033 2034 2035 2036 2037
        alpha_t = _C_ops.elementwise_add(
            _C_ops.elementwise_mul(alpha, label),
            _C_ops.elementwise_mul(
                _C_ops.elementwise_sub(one, alpha),
                _C_ops.elementwise_sub(one, label)))
        loss = _C_ops.elementwise_mul(alpha_t, loss)
2038 2039

        gamma = fluid.dygraph.base.to_variable([gamma], dtype=loss.dtype)
W
wanghuancoder 已提交
2040 2041 2042
        gamma_t = _C_ops.elementwise_pow(
            _C_ops.elementwise_sub(one, p_t), gamma)
        loss = _C_ops.elementwise_mul(gamma_t, loss)
2043 2044

        if normalizer is not None:
W
wanghuancoder 已提交
2045
            loss = _C_ops.elementwise_div(loss, normalizer)
2046 2047

        if reduction == "sum":
W
wanghuancoder 已提交
2048
            return _C_ops.reduce_sum(loss, 'reduce_all', True)
2049
        elif reduction == "mean":
W
wanghuancoder 已提交
2050
            return _C_ops.mean(loss)
2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064

        return loss

    fluid.data_feeder.check_variable_and_dtype(
        logit, 'logit', ['float32', 'float64'], 'sigmoid_focal_loss')
    fluid.data_feeder.check_variable_and_dtype(
        label, 'label', ['float32', 'float64'], 'sigmoid_focal_loss')

    bce_name = None
    if reduction == 'none' and normalizer is None:
        bce_name = name
    loss = paddle.nn.functional.binary_cross_entropy_with_logits(
        logit, label, reduction='none', name=bce_name)

Z
zhiboniu 已提交
2065
    pred = paddle.nn.functional.sigmoid(logit)
2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083
    p_t = pred * label + (1 - pred) * (1 - label)

    alpha_t = alpha * label + (1 - alpha) * (1 - label)
    loss = paddle.multiply(alpha_t, loss)

    gamma_t = paddle.pow((1 - p_t), gamma)
    loss = paddle.multiply(gamma_t, loss)

    if normalizer is not None:
        normalizer_name = name if reduction == 'none' else None
        loss = paddle.divide(loss, normalizer, name=normalizer_name)

    if reduction == 'mean':
        loss = paddle.mean(loss, name=name)
    elif reduction == 'sum':
        loss = paddle.sum(loss, name=name)

    return loss
2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164


def hinge_embedding_loss(input, label, margin=1.0, reduction='mean', name=None):
    r"""
    This operator calculates hinge_embedding_loss. Measures the loss given an input tensor :math:`x` and a labels tensor :math:`y`(containing 1 or -1).
    This is usually used for measuring whether two inputs are similar or dissimilar, e.g. using the L1 pairwise distance as :math:`x`,
    and is typically used for learning nonlinear embeddings or semi-supervised learning.

    The loss function for :math:`n`-th sample in the mini-batch is

    .. math::
        l_n = \begin{cases}
            x_n, & \text{if}\; y_n = 1,\\
            \max \{0, \Delta - x_n\}, & \text{if}\; y_n = -1,
        \end{cases}

    and the total loss functions is

    .. math::
        \ell(x, y) = \begin{cases}
            \operatorname{mean}(L), & \text{if reduction} = \text{'mean';}\\
            \operatorname{sum}(L),  & \text{if reduction} = \text{'sum'.}
        \end{cases}

    where :math:`L = \{l_1,\dots,l_N\}^\top`.

    Parameters:
        input (Tensor): Input tensor, the data type is float32 or float64.
            the shape is [N, \*], N is batch size and `\*` means any number of additional dimensions, available dtype is float32, float64.
        label (Tensor): Label tensor containing 1 or -1, the data type is float32 or float64.
            The shape of label is the same as the shape of input.
        margin (float, optional): Specifies the hyperparameter margin to be used.
            The value determines how large the input need to be to calculate in
            hinge_embedding_loss. When label is -1, Input smaller than margin are minimized with hinge_embedding_loss.
            Default = 1.0
        reduction (str, optional): Indicate how to average the loss by batch_size.
            the candicates are ``'none'`` | ``'mean'`` | ``'sum'``.
            If :attr:`reduction` is ``'none'``, the unreduced loss is returned;
            If :attr:`reduction` is ``'mean'``, the reduced mean loss is returned;
            If :attr:`reduction` is ``'sum'``, the summed loss is returned.
            Default: ``'mean'``
        name (str, optional): Name for the operation (optional, default is None).
            For more information, please refer to :ref:`api_guide_Name`.

    Shape:

        input: N-D Tensor, the shape is [N, \*], N is batch size and `\*` means any number of additional dimensions, available dtype is float32, float64. The sum operationoperates over all the elements.

        label: N-D Tensor, same shape as the input. tensor elements should containing 1 or -1, the data type is float32 or float64.

        output: scalar. If :attr:`reduction` is ``'none'``, then same shape as the input.

    Returns:
        Tensor. The tensor variable storing the hinge_embedding_loss of input and label.

    Examples:
        .. code-block:: python

            import paddle
            import paddle.nn.functional as F

            input = paddle.to_tensor([[1, -2, 3], [0, -1, 2], [1, 0, 1]], dtype=paddle.float32)
            # label elements in {1., -1.}
            label = paddle.to_tensor([[-1, 1, -1], [1, 1, 1], [1, -1, 1]], dtype=paddle.float32)

            loss = F.hinge_embedding_loss(input, label, margin=1.0, reduction='none')
            print(loss)
            # Tensor([[0., -2., 0.],
            #         [0., -1., 2.],
            #         [1., 1., 1.]])

            loss = F.hinge_embedding_loss(input, label, margin=1.0, reduction='mean')
            print(loss)
            # Tensor([0.22222222])
    """

    if reduction not in ['sum', 'mean', 'none']:
        raise ValueError(
            "'reduction' in 'hinge_embedding_loss' should be 'sum', 'mean' or 'none', "
            "but received {}.".format(reduction))

Z
zhiboniu 已提交
2165
    if not in_dynamic_mode():
2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180
        check_variable_and_dtype(input, 'input', ['float32', 'float64'],
                                 'hinge_embedding_loss')
        check_variable_and_dtype(label, 'label', ['float32', 'float64'],
                                 'hinge_embedding_loss')

    zero_ = paddle.zeros([1], dtype=input.dtype)
    loss = paddle.where(label == 1., input, zero_) + \
           paddle.where(label == -1., paddle.nn.functional.relu(margin - input), zero_)

    if reduction == 'mean':
        return paddle.mean(loss, name=name)
    elif reduction == 'sum':
        return paddle.sum(loss, name=name)
    elif reduction == 'none':
        return loss