common.py 67.8 KB
Newer Older
1 2 3 4 5 6 7 8 9 10 11 12 13 14
#   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.

X
xiaoting 已提交
15
import warnings
16 17
import paddle
from ...fluid.framework import in_dygraph_mode, default_main_program
X
xiaoting 已提交
18
from paddle.fluid.layer_helper import LayerHelper
L
littletomatodonkey 已提交
19
from paddle.fluid.layers.tensor import Variable, fill_constant, zeros, concat
20 21
from ...fluid.layers import core
from ...fluid import dygraph_utils
22
# TODO: define the common functions to build a neural network  
23 24
# from ...fluid import one_hot  #DEFINE_ALIAS
# from ...fluid.layers import pad2d  #DEFINE_ALIAS
25
from ...fluid.layers import unfold  #DEFINE_ALIAS
L
littletomatodonkey 已提交
26 27
from ...fluid.layers import squeeze  #DEFINE_ALIAS
from ...fluid.layers import unsqueeze  #DEFINE_ALIAS
Y
Yang Zhang 已提交
28 29 30
from ...tensor import clip
from ...tensor import sum
from ...tensor import sqrt
31 32 33 34
from ...tensor import sum  #DEFINE_ALIAS
from ...tensor import sqrt  #DEFINE_ALIAS
from ...fluid.data_feeder import check_variable_and_dtype, check_dtype
from ...fluid.framework import Variable, in_dygraph_mode, _varbase_creator
X
xiaoting 已提交
35

36
#from ...fluid.layers import fc  #DEFINE_ALIAS
37
# from ...fluid.layers import pad_constant_like  #DEFINE_ALIAS
38 39
from ...fluid.framework import in_dygraph_mode
from ...fluid import core, dygraph_utils
40 41
from ...fluid import core, layers
from ...fluid.data_feeder import check_variable_and_dtype
42

43 44
__all__ = [
    'dropout',
45 46
    'dropout2d',
    'dropout3d',
47
    'alpha_dropout',
48 49 50
    #       'embedding',
    #       'fc',
    'label_smooth',
51
    'linear',
52 53 54
    'pad',
    'unfold',
    #       'bilinear_tensor_product',
L
littletomatodonkey 已提交
55
    'interpolate',
X
xiaoting 已提交
56
    'upsample',
57
    'bilinear',
L
littletomatodonkey 已提交
58
    'cosine_similarity',
59
]
X
xiaoting 已提交
60 61


X
xiaoting 已提交
62
def interpolate(x,
63 64 65 66
                size=None,
                scale_factor=None,
                mode='nearest',
                align_corners=False,
X
xiaoting 已提交
67
                align_mode=0,
68 69
                data_format='NCHW',
                name=None):
X
xiaoting 已提交
70
    """
S
swtkiwi 已提交
71

X
xiaoting 已提交
72
    This op resizes a batch of images.
73 74
    The input must be a 3-D Tensor of the shape (num_batches, channels, in_w)
    or 4-D (num_batches, channels, in_h, in_w), or a 5-D Tensor of the shape
X
xiaoting 已提交
75
    (num_batches, channels, in_d, in_h, in_w) or (num_batches, in_d, in_h, in_w, channels),
76 77
    Where in_w is width of the input tensor, in_h is the height of the input tensor,
    in_d is the depth of the intput tensor.
X
xiaoting 已提交
78
    and the resizing only applies on the three dimensions(depth, height and width).
X
xiaoting 已提交
79

X
xiaoting 已提交
80
    Supporting resample methods:
81 82 83 84 85
        'linear' : Linear interpolation
        'bilinear' : Bilinear interpolation
        'trilinear' : Trilinear interpolation
        'nearest' : Nearest neighbor interpolation
        'bicubic' : Bicubic interpolation
86
        'area': Area interpolation
87 88 89 90

    Linear interpolation is the method of using a line connecting two known quantities 
    to determine the value of an unknown quantity between the two known quantities. 
    
X
xiaoting 已提交
91 92 93 94 95 96 97 98 99 100 101 102 103 104
    Nearest neighbor interpolation is to perform nearest neighbor interpolation
    in both the 3rd dimension(in height direction) and the 4th dimension(in width
    direction) on input tensor.

    Bilinear interpolation is an extension of linear interpolation for
    interpolating functions of two variables (e.g. H-direction and
    W-direction in this op) on a rectilinear 2D grid. The key idea is
    to perform linear interpolation first in one direction, and then
    again in the other direction.

    Trilinear interpolation is an extension of linear interpolation for
    interpolating functions of three variables (e.g. D-direction,
    H-direction and W-direction in this op) on a rectilinear 3D grid.
    The linear interpolation is performed on three directions.
X
xiaoting 已提交
105
    align_corners and align_mode are optional parameters,the calculation method
X
xiaoting 已提交
106 107 108 109 110 111 112
    of interpolation can be selected by them.

    Bicubic interpolation is an extension of cubic interpolation for interpolating
    data points on a two-dimensional regular grid. The interpolated surface is
    smoother than corresponding surfaces obtained by bilinear interpolation or
    nearest-neighbor interpolation.

113 114 115 116 117 118
    Area interpolation is to perform area interpolation
    in both the 3rd dimension(in height direction) , the 4th dimension(in width
    direction) and the 5th dimension(in depth direction) on input tensor. Set to 
    area will directly call `paddle.nn.functional.adaptive_avg_pool1d` or 
    `paddle.nn.functional.adaptive_avg_pool2d` or `paddle.nn.functional.adaptive_avg_pool3d`.

X
xiaoting 已提交
119 120 121 122
    Example:

    .. code-block:: text

123
        For scale_factor:
X
xiaoting 已提交
124 125 126 127 128
            if align_corners = True && out_size > 1 :
              scale_factor = (in_size-1.0)/(out_size-1.0)
            else:
              scale_factor = float(in_size/out_size)

129 130 131 132 133 134 135 136 137 138 139
        Linear interpolation:
            if:
                align_corners = False , align_mode = 0
                input : (N,C,W_in)
                output: (N,C,W_out) where:
                W_out = (W_{in}+0.5) * scale_{factor} - 0.5
            else:
                input : (N,C,W_in)
                output: (N,C,W_out) where:
                W_out = W_{in} * scale_{factor}
        
X
xiaoting 已提交
140
        Nearest neighbor interpolation:
X
xiaoting 已提交
141

X
xiaoting 已提交
142 143 144 145 146
              align_corners = False
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = floor (H_{in} * scale_{factor})
              W_out = floor (W_{in} * scale_{factor})
147

X
xiaoting 已提交
148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
        Bilinear interpolation:
          if:
              align_corners = False , align_mode = 0
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}

        Bicubic interpolation:
          if:
              align_corners = False
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}

        Trilinear interpolation:
          if:
              align_corners = False , align_mode = 0
              input : (N,C,D_in,H_in,W_in)
              output: (N,C,D_out,H_out,W_out) where:
              D_out = (D_{in}+0.5) * scale_{factor} - 0.5
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,D_in,H_in,W_in)
              output: (N,C,D_out,H_out,W_out) where:
              D_out = D_{in} * scale_{factor}
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}

189 190 191
    For details of linear interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Linear_interpolation.
    
X
xiaoting 已提交
192 193
    For details of nearest neighbor interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Nearest-neighbor_interpolation.
194
    
X
xiaoting 已提交
195 196
    For details of bilinear interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Bilinear_interpolation.
197
    
X
xiaoting 已提交
198 199
    For details of trilinear interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Trilinear_interpolation.
200
    
X
xiaoting 已提交
201 202
    For details of bicubic interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Bicubic_interpolation
203
    
X
xiaoting 已提交
204
    Parameters:
X
xiaoting 已提交
205
        x (Tensor): 3-D, 4-D or 5-D Tensor, its data type is float32, float64, or uint8,
X
xiaoting 已提交
206
                          its data format is specified by :attr:`data_format`.
X
xiaoting 已提交
207
        size (list|tuple|Tensor|None): Output shape of image resize
208 209
             layer, the shape is (out_w, ) when input is a 3-D Tensor, the shape is (out_h, out_w) 
             when input is a 4-D Tensor and is (out_d, out_h, out_w) when input is a 5-D Tensor. 
X
xiaoting 已提交
210 211
             Default: None. If a list, each element can be an integer or a Tensor of shape: [1].
             If a Tensor, its dimensions size should be a 1.
212 213 214
        scale_factor (float|Tensor|list|tuple|None): The multiplier for the input height or width. At
             least one of :attr:`size` or :attr:`scale_factor` must be set.
             And :attr:`size` has a higher priority than :attr:`scale_factor`.Has to match input size if it is either a list or a tuple or a Tensor.
X
xiaoting 已提交
215
             Default: None.
216
        mode (str): The resample method. It supports 'linear', 'area', 'nearest', 'bilinear',
217
                       'bicubic' and 'trilinear' currently. Default: 'nearest'
X
xiaoting 已提交
218 219
        align_corners(bool) :  An optional bool, If True, the centers of the 4 corner pixels of the
                               input and output tensors are aligned, preserving the values at the
X
xiaoting 已提交
220
                               corner pixels.This only has an effect when 'linear', 'bilinear', 'bicubic' or 'trilinear'.
221 222 223 224
                               Default: False
        align_mode(int)  :  An optional for linear/bilinear/trilinear interpolation. Refer to the formula in the example above,
                            it can be \'0\' for src_idx = scale_factor*(dst_indx+0.5)-0.5 , can be \'1\' for
                            src_idx = scale_factor*dst_index.
X
xiaoting 已提交
225
        data_format (str, optional): Specify the data format of the input, and the data format of the output
226
            will be consistent with that of the input. An optional string from:`NCW`, `NWC`,  `"NCHW"`, `"NHWC"`, `"NCDHW"`,
X
xiaoting 已提交
227 228 229
            `"NDHWC"`. The default is `"NCHW"`. When it is `"NCHW"`, the data is stored in the order of:
            `[batch_size, input_channels, input_height, input_width]`. When it is `"NCHW"`, the data is stored
            in the order of: `[batch_size, input_channels, input_depth, input_height, input_width]`.
230 231 232
        name(str, optional): The default value is None.
                             Normally there is no need for user to set this property.
                             For more information, please refer to :ref:`api_guide_Name`
X
xiaoting 已提交
233
    Returns:
234
        A 3-D Tensor of the shape (num_batches, channels, out_w) or (num_batches, out_w, channels),
X
xiaoting 已提交
235 236 237
        A 4-D Tensor of the shape (num_batches, channels, out_h, out_w) or (num_batches, out_h, out_w, channels),
        or 5-D Tensor of the shape (num_batches, channels, out_d, out_h, out_w) or (num_batches, out_d, out_h, out_w, channels).
    Raises:
X
xiaoting 已提交
238
        TypeError: size should be a list or tuple or Tensor.
239
        ValueError: The 'mode' of image_resize can only be 'linear', 'bilinear',
240
                    'trilinear', 'bicubic', 'area' or 'nearest' currently.
241 242 243 244 245 246 247 248
        ValueError: 'linear' only support 3-D tensor.
        ValueError: 'bilinear', 'bicubic' and 'nearest' only support 4-D tensor.
        ValueError: 'trilinear' only support 5-D tensor.
        ValueError: One of size and scale_factor must not be None.
        ValueError: size length should be 1 for input 3-D tensor.
        ValueError: size length should be 2 for input 4-D tensor.
        ValueError: size length should be 3 for input 5-D tensor.
        ValueError: scale_factor should be greater than zero.
X
xiaoting 已提交
249 250
        TypeError: align_corners should be a bool value
        ValueError: align_mode can only be '0' or '1'
251 252
        ValueError: data_format can only be 'NCW', 'NWC', 'NCHW', 'NHWC', 'NCDHW' or 'NDHWC'.

X
xiaoting 已提交
253 254 255 256 257
    Examples:
        .. code-block:: python

	    import paddle
	    import numpy as np
X
xiaoting 已提交
258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275
            import paddle.nn.functional as F
            
            # given out size
            input_data = np.random.rand(2,3,6,10).astype("float32")
            x = paddle.to_tensor(input_data)
            output_1 = F.interpolate(x=x, size=[12,12])
    	    print(output_1.shape)
	    # [2L, 3L, 12L, 12L]
            
            # given scale
            output_2 = F.interpolate(x=x, scale_factor=[2,1])
            print(output_2.shape)
            # [2L, 3L, 12L, 10L]
            
            # bilinear interp
            output_3 = F.interpolate(x=x, scale_factor=[2,1], mode="bilinear")
            print(output_2.shape)
            # [2L, 3L, 12L, 10L]
X
xiaoting 已提交
276
    """
277 278 279 280 281 282 283 284 285 286
    data_format = data_format.upper()
    resample = mode.upper()
    resample_type = mode.lower()

    resample_methods = [
        'LINEAR',
        'BILINEAR',
        'TRILINEAR',
        'NEAREST',
        'BICUBIC',
287
        'AREA',
288
    ]
X
xiaoting 已提交
289 290
    if resample not in resample_methods:
        raise ValueError(
291
            "The 'resample' of image_resize can only be 'area', 'linear', 'bilinear', 'trilinear', "
292
            " 'bicubic' or 'nearest' currently.")
X
xiaoting 已提交
293

X
xiaoting 已提交
294
    if resample in ['LINEAR'] and len(x.shape) != 3:
295
        raise ValueError("'linear' only support 3-D tensor.")
296

X
xiaoting 已提交
297
    if resample in ['BILINEAR', 'NEAREST', 'BICUBIC'] and len(x.shape) != 4:
X
xiaoting 已提交
298
        raise ValueError(
299
            "'bilinear', 'bicubic' and 'nearest' only support 4-D tensor.")
X
xiaoting 已提交
300
    if resample == 'TRILINEAR' and len(x.shape) != 5:
301 302 303 304
        raise ValueError("'trilinear'only support 5-D tensor.")

    if size is None and scale_factor is None:
        raise ValueError("One of size and scale_factor must not be None.")
X
xiaoting 已提交
305 306 307

    if not isinstance(align_corners, bool):
        raise TypeError("Attr align_corners should be a bool value")
308

X
xiaoting 已提交
309 310
    if align_mode != 0 and align_mode != 1:
        raise ValueError("align_mode can only be 0 or 1")
X
xiaoting 已提交
311 312 313 314
    if align_corners != 0 and resample == 'NEAREST':
        raise ValueError(
            "align_corners option can only be set with the interpolating modes: linear | bilinear | bicubic | trilinear"
        )
315 316 317 318 319 320 321 322 323

    if resample == 'AREA' and len(x.shape) == 3:
        return paddle.nn.functional.adaptive_avg_pool1d(x, size)

    if resample == 'AREA' and len(x.shape) == 4:
        return paddle.nn.functional.adaptive_avg_pool2d(x, size)
    if resample == 'AREA' and len(x.shape) == 5:
        return paddle.nn.functional.adaptive_avg_pool3d(x, size)

X
xiaoting 已提交
324
    helper = LayerHelper('{}_interp_v2'.format(resample_type), **locals())
325
    dtype = helper.input_dtype(input_param_name='x')
X
xiaoting 已提交
326
    if len(x.shape) == 3 and data_format not in ['NCW', 'NWC']:
327 328
        raise ValueError(
            "Got wrong value for param `data_format`: " + data_format +
329
            " received but only `NCW` or `NWC` supported for 3-D input.")
X
xiaoting 已提交
330
    elif len(x.shape) == 4 and data_format not in ['NCHW', 'NHWC']:
X
xiaoting 已提交
331 332 333
        raise ValueError(
            "Got wrong value for param `data_format`: " + data_format +
            " received but only `NCHW` or `NHWC` supported for 4-D input.")
X
xiaoting 已提交
334
    elif len(x.shape) == 5 and data_format not in ['NCDHW', 'NDHWC']:
X
xiaoting 已提交
335 336 337 338 339 340 341
        raise ValueError(
            "Got wrong value for param `data_format`: " + data_format +
            " received but only `NCDHW` or `NDHWC` supported for 5-D input.")

    def _is_list_or_turple_(data):
        return (isinstance(data, list) or isinstance(data, tuple))

342
    if data_format == 'NCHW' or data_format == 'NCDHW' or data_format == 'NCW':
X
xiaoting 已提交
343
        data_layout = 'NCHW'
344
    if data_format == 'NHWC' or data_format == 'NDHWC' or data_format == 'NWC':
X
xiaoting 已提交
345 346
        data_layout = 'NHWC'

X
xiaoting 已提交
347 348 349 350
    if resample == 'NEAREST':
        align_corners = False

    inputs = {"X": x}
X
xiaoting 已提交
351 352 353 354 355 356 357 358 359 360
    attrs = {
        "out_d": -1,
        "out_h": -1,
        "out_w": -1,
        "interp_method": resample_type,
        "align_corners": align_corners,
        "align_mode": align_mode,
        "data_layout": data_layout
    }

361 362
    out_shape = size
    scale = scale_factor
363 364
    if out_shape is not None and scale is not None:
        raise ValueError("Only one of size or scale_factor should be defined.")
X
xiaoting 已提交
365
    if out_shape is not None:
366 367

        if isinstance(out_shape, Variable) and not in_dygraph_mode():
X
xiaoting 已提交
368 369
            out_shape.stop_gradient = True
            inputs['OutSize'] = out_shape
370

X
xiaoting 已提交
371
        else:
372 373 374 375 376 377
            if in_dygraph_mode():
                if isinstance(out_shape, Variable):
                    out_shape = list(out_shape.numpy())
                for i, dim in enumerate(out_shape):
                    if isinstance(dim, Variable):
                        out_shape[i] = dim.numpy()[0]
X
xiaoting 已提交
378
            if not (_is_list_or_turple_(out_shape)):
379
                raise TypeError("size should be a list or tuple or Variable.")
X
xiaoting 已提交
380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
            # Validate the shape
            contain_var = False
            for dim_idx, dim_size in enumerate(out_shape):
                if isinstance(dim_size, Variable):
                    contain_var = True
                    continue
                assert dim_size > 0, (
                    "Each dimension size given in out_shape must be greater than 0."
                )

            if contain_var:
                new_size_tensor = []
                size_list = []
                for dim in out_shape:
                    if isinstance(dim, Variable):
                        dim.stop_gradient = True
                        new_size_tensor.append(dim)
                        size_list.append(-1)
                    else:
                        assert (isinstance(dim, int))
                        temp_out = helper.create_variable_for_type_inference(
                            'int32')
                        fill_constant(
                            [1], 'int32', dim, force_cpu=True, out=temp_out)
                        new_size_tensor.append(temp_out)
                        size_list.append(dim)
                inputs['SizeTensor'] = new_size_tensor

X
xiaoting 已提交
408
            if len(x.shape) == 3:
409 410
                if len(out_shape) != 1:
                    raise ValueError(
411
                        "size length should be 2 for input 3-D tensor")
412 413 414 415 416
                if contain_var:
                    attrs['out_w'] = size_list[0]
                else:
                    out_shape = list(map(int, out_shape))
                    attrs['out_w'] = out_shape[0]
X
xiaoting 已提交
417
            if len(x.shape) == 4:
X
xiaoting 已提交
418
                if len(out_shape) != 2:
419
                    raise ValueError("size length should be 2 for "
X
xiaoting 已提交
420 421 422 423 424 425 426 427
                                     "input 4-D tensor.")
                if contain_var:
                    attrs['out_h'] = size_list[0]
                    attrs['out_w'] = size_list[1]
                else:
                    out_shape = list(map(int, out_shape))
                    attrs['out_h'] = out_shape[0]
                    attrs['out_w'] = out_shape[1]
X
xiaoting 已提交
428
            if len(x.shape) == 5:
X
xiaoting 已提交
429
                if len(out_shape) != 3:
430
                    raise ValueError("size length should be 3 for "
X
xiaoting 已提交
431 432 433 434 435 436 437 438 439 440 441 442
                                     "input 5-D tensor.")
                if contain_var:
                    attrs['out_d'] = size_list[0]
                    attrs['out_h'] = size_list[1]
                    attrs['out_w'] = size_list[2]
                else:
                    out_shape = list(map(int, out_shape))
                    attrs['out_d'] = out_shape[0]
                    attrs['out_h'] = out_shape[1]
                    attrs['out_w'] = out_shape[2]

    else:
443 444
        if in_dygraph_mode() and isinstance(scale, Variable):
            scale = list(scale.numpy())
X
xiaoting 已提交
445 446 447 448 449 450
        if isinstance(scale, Variable):
            scale.stop_gradient = True
            inputs["Scale"] = scale
        elif isinstance(scale, float) or isinstance(scale, int):
            if scale <= 0:
                raise ValueError("Attr(scale) should be greater than zero.")
X
xiaoting 已提交
451 452 453 454
            scale_list = []
            for i in range(len(x.shape) - 2):
                scale_list.append(scale)
            attrs['scale'] = list(map(float, scale_list))
X
xiaoting 已提交
455
        elif isinstance(scale, list) or isinstance(scale, tuple):
X
xiaoting 已提交
456 457 458 459 460 461 462 463
            if len(scale) != len(x.shape) - 2:
                raise ValueError("scale_shape length should be {} for "
                                 "input {}-D tensor.".format(
                                     len(x.shape) - 2, len(x.shape)))
            for value in scale:
                if value <= 0:
                    raise ValueError("Attr(scale) should be greater than zero.")
            attrs['scale'] = list(map(float, scale))
X
xiaoting 已提交
464 465
        else:
            raise TypeError(
466 467
                "Attr(scale)'s type should be float, int, list, tuple, or Tensor."
            )
X
xiaoting 已提交
468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486

    if in_dygraph_mode():
        attr_list = []
        for k, v in attrs.items():
            attr_list.append(k)
            attr_list.append(v)
        dy_attr = tuple(attr_list)

        if resample_type == "linear":
            out = core.ops.linear_interp_v2(x, *dy_attr)
        if resample_type == "bilinear":
            out = core.ops.bilinear_interp_v2(x, *dy_attr)
        if resample_type == "trilinear":
            out = core.ops.trilinear_interp_v2(x, *dy_attr)
        if resample_type == "nearest":
            out = core.ops.nearest_interp_v2(x, *dy_attr)
        if resample_type == "bicubic":
            out = core.ops.bicubic_interp_v2(x, *dy_attr)
        return out
X
xiaoting 已提交
487 488
    out = helper.create_variable_for_type_inference(dtype)
    helper.append_op(
X
xiaoting 已提交
489
        type='{}_interp_v2'.format(resample_type),
X
xiaoting 已提交
490 491 492 493
        inputs=inputs,
        outputs={"Out": out},
        attrs=attrs)
    return out
L
littletomatodonkey 已提交
494 495


X
xiaoting 已提交
496 497 498 499 500 501 502 503 504 505
def upsample(x,
             size=None,
             scale_factor=None,
             mode='nearest',
             align_corners=False,
             align_mode=0,
             data_format='NCHW',
             name=None):
    """
    This op resizes a batch of images.
506

X
xiaoting 已提交
507 508 509
    The input must be a 3-D Tensor of the shape (num_batches, channels, in_w)
    or 4-D (num_batches, channels, in_h, in_w), or a 5-D Tensor of the shape
    (num_batches, channels, in_d, in_h, in_w) or (num_batches, in_d, in_h, in_w, channels),
510 511
    Where in_w is width of the input tensor, in_h is the height of the input tensor,
    in_d is the depth of the intput tensor.
X
xiaoting 已提交
512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535
    and the resizing only applies on the three dimensions(depth, height and width).

    Supporting resample methods:
        'linear' : Linear interpolation
        'bilinear' : Bilinear interpolation
        'trilinear' : Trilinear interpolation
        'nearest' : Nearest neighbor interpolation
        'bicubic' : Bicubic interpolation
    Linear interpolation is the method of using a line connecting two known quantities 
    to determine the value of an unknown quantity between the two known quantities. 
    
    Nearest neighbor interpolation is to perform nearest neighbor interpolation
    in both the 3rd dimension(in height direction) and the 4th dimension(in width
    direction) on input tensor.
    Bilinear interpolation is an extension of linear interpolation for
    interpolating functions of two variables (e.g. H-direction and
    W-direction in this op) on a rectilinear 2D grid. The key idea is
    to perform linear interpolation first in one direction, and then
    again in the other direction.
    
    Bicubic interpolation is an extension of cubic interpolation for interpolating
    data points on a two-dimensional regular grid. The interpolated surface is
    smoother than corresponding surfaces obtained by bilinear interpolation or
    nearest-neighbor interpolation.
536

X
xiaoting 已提交
537 538 539
    Trilinear interpolation is an extension of linear interpolation for
    interpolating functions of three variables (e.g. D-direction,
    H-direction and W-direction in this op) on a rectilinear 3D grid.
540

X
xiaoting 已提交
541 542 543
    The linear interpolation is performed on three directions.
    align_corners and align_mode are optional parameters,the calculation method
    of interpolation can be selected by them.
544 545 546 547 548 549 550

    Area interpolation is to perform area interpolation
    in both the 3rd dimension(in height direction) , the 4th dimension(in width
    direction) and the 5th dimension(in depth direction) on input tensor. Set to
    area will directly call `paddle.nn.functional.adaptive_avg_pool1d` or
    `paddle.nn.functional.adaptive_avg_pool2d` or `paddle.nn.functional.adaptive_avg_pool3d`.

X
xiaoting 已提交
551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640
    Example:
    .. code-block:: text
        For scale_factor:
            if align_corners = True && out_size > 1 :
              scale_factor = (in_size-1.0)/(out_size-1.0)
            else:
              scale_factor = float(in_size/out_size)
        Linear interpolation:
            if:
                align_corners = False , align_mode = 0
                input : (N,C,W_in)
                output: (N,C,W_out) where:
                W_out = (W_{in}+0.5) * scale_{factor} - 0.5
            else:
                input : (N,C,W_in)
                output: (N,C,W_out) where:
                W_out = W_{in} * scale_{factor}
        Nearest neighbor interpolation:
          if:
              align_corners = False
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = floor (H_{in} * scale_{factor})
              W_out = floor (W_{in} * scale_{factor})
          else:
              align_corners = True
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = round(H_{in} * scale_{factor})
              W_out = round(W_{in} * scale_{factor})
        
        Bilinear interpolation:
          if:
              align_corners = False , align_mode = 0
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}
        Bicubic interpolation:
          if:
              align_corners = False
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,H_in,W_in)
              output: (N,C,H_out,W_out) where:
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}
        Trilinear interpolation:
          if:
              align_corners = False , align_mode = 0
              input : (N,C,D_in,H_in,W_in)
              output: (N,C,D_out,H_out,W_out) where:
              D_out = (D_{in}+0.5) * scale_{factor} - 0.5
              H_out = (H_{in}+0.5) * scale_{factor} - 0.5
              W_out = (W_{in}+0.5) * scale_{factor} - 0.5
          else:
              input : (N,C,D_in,H_in,W_in)
              output: (N,C,D_out,H_out,W_out) where:
              D_out = D_{in} * scale_{factor}
              H_out = H_{in} * scale_{factor}
              W_out = W_{in} * scale_{factor}
    https://en.wikipedia.org/wiki/Linear_interpolation.
    For details of linear interpolation, please refer to Wikipedia:
    
    For details of nearest neighbor interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Nearest-neighbor_interpolation.
    
    For details of bilinear interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Bilinear_interpolation.
    
    For details of bicubic interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Bicubic_interpolation
    
    For details of trilinear interpolation, please refer to Wikipedia:
    https://en.wikipedia.org/wiki/Trilinear_interpolation.
    
    Parameters:
        x (Tensor): 3-D, 4-D or 5-D Tensor, its data type is float32, float64, or uint8,
                          its data format is specified by :attr:`data_format`.
        size (list|tuple|Tensor|None): Output shape of image resize
             layer, the shape is (out_w, ) when input is a 3-D Tensor, the shape is (out_h, out_w) 
             when input is a 4-D Tensor and is (out_d, out_h, out_w) when input is a 5-D Tensor. 
X
xiaoting 已提交
641 642
             Default: None. If a list, each element can be an integer or a Tensor of shape: [1].
             If a Tensor , its dimensions size should be a 1.
643 644 645 646
        scale_factor (float|Tensor|list|tuple|None): The multiplier for the input height or width. At
             least one of :attr:`size` or :attr:`scale_factor` must be set.
             And :attr:`size` has a higher priority than :attr:`scale_factor`.Has to match input size if 
             it is either a list or a tuple or a Tensor.
X
xiaoting 已提交
647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689
             Default: None.
        mode (str): The resample method. It supports 'linear', 'nearest', 'bilinear',
                       'bicubic' and 'trilinear' currently. Default: 'nearest'
        align_corners(bool) :  An optional bool, If True, the centers of the 4 corner pixels of the
                               input and output tensors are aligned, preserving the values at the
                               corner pixels.
                               Default: False
        align_mode(int)  :  An optional for linear/bilinear/trilinear interpolation. Refer to the formula in the example above,
                            it can be \'0\' for src_idx = scale_factor*(dst_indx+0.5)-0.5 , can be \'1\' for
                            src_idx = scale_factor*dst_index.
        data_format (str, optional): Specify the data format of the input, and the data format of the output
            will be consistent with that of the input. An optional string from:`NCW`, `NWC`, `"NCHW"`, `"NHWC"`, `"NCDHW"`,
            `"NDHWC"`. The default is `"NCHW"`. When it is `"NCHW"`, the data is stored in the order of:
            `[batch_size, input_channels, input_height, input_width]`. When it is `"NCHW"`, the data is stored
            in the order of: `[batch_size, input_channels, input_depth, input_height, input_width]`.
        name(str, optional): The default value is None.
                             Normally there is no need for user to set this property.
                             For more information, please refer to :ref:`api_guide_Name`
    Returns:
        A 3-D Tensor of the shape (num_batches, channels, out_w) or (num_batches, out_w, channels),
        A 4-D Tensor of the shape (num_batches, channels, out_h, out_w) or (num_batches, out_h, out_w, channels),
        or 5-D Tensor of the shape (num_batches, channels, out_d, out_h, out_w) or (num_batches, out_d, out_h, out_w, channels).
    Raises:
        TypeError: size should be a list or tuple or Tensor.
        ValueError: The 'mode' of image_resize can only be 'linear', 'bilinear',
                    'trilinear', 'bicubic', or 'nearest' currently.
        ValueError: 'linear' only support 3-D tensor.
        ValueError: 'bilinear', 'bicubic' and 'nearest' only support 4-D tensor.
        ValueError: 'trilinear' only support 5-D tensor.
        ValueError: One of size and scale_factor must not be None.
        ValueError: size length should be 1 for input 3-D tensor.
        ValueError: size length should be 2 for input 4-D tensor.
        ValueError: size length should be 3 for input 5-D tensor.
        ValueError: scale_factor should be greater than zero.
        TypeError: align_corners should be a bool value
        ValueError: align_mode can only be '0' or '1'
        ValueError: data_format can only be 'NCW', 'NWC', 'NCHW', 'NHWC', 'NCDHW' or 'NDHWC'.
        Examples:
        .. code-block:: python
            import paddle
            import numpy as np
            import paddle.nn.functional as F

X
xiaoting 已提交
690
            input_data = np.random.rand(2,3,6,10).astype("float32")
X
xiaoting 已提交
691
            input = paddle.to_tensor(input_data)
X
xiaoting 已提交
692
            output = F.upsample(x=input, size=[12,12])
X
xiaoting 已提交
693 694 695 696 697 698 699 700
            print(output.shape)
            # [2L, 3L, 12L, 12L]

    """
    return interpolate(x, size, scale_factor, mode, align_corners, align_mode,
                       data_format)


701 702 703 704
def bilinear(x1, x2, weight, bias=None, name=None):
    """

    This layer performs bilinear on two inputs.
705
    See :ref:`api_nn_Bilinear` for details and output shape.
706 707 708 709 710 711 712 713 714 715

    Parameters:
       x1 (Tensor): the first input tensor, it's data type should be float32, float64.
       x2 (Tensor): the second input tensor, it's data type should be float32, float64.
       weight (Parameter): The learnable weights of this layer, shape is [out_features, in1_features, in2_features].
       bias (Parameter, optional): The learnable bias(Bias) of this layer, shape is [1, out_features]. If it is set to None, no bias will be added to the output units. The default value is None.
       name (str, optional): The default value is None. Normally there is no need for user
           to set this property. For more information, please refer to :ref:`api_guide_Name`. Default: None.

    Returns:
716
       Tensor: A 2-D Tensor of shape [batch_size, out_features].
717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752

    Examples:
       .. code-block:: python

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

        x1 = numpy.random.random((5, 5)).astype('float32')
        x2 = numpy.random.random((5, 4)).astype('float32')
        w = numpy.random.random((1000, 5, 4)).astype('float32')
        b = numpy.random.random((1, 1000)).astype('float32')

        result = F.bilinear(paddle.to_tensor(x1), paddle.to_tensor(x2), paddle.to_tensor(w), paddle.to_tensor(b))           # result shape [5, 1000]

    """

    if in_dygraph_mode():
        return core.ops.bilinear_tensor_product(x1, x2, weight, bias)

    check_variable_and_dtype(x1, 'x1', ['float32', 'float64'], 'bilinear')
    check_variable_and_dtype(x2, 'x2', ['float32', 'float64'], 'bilinear')

    inputs = {"X": x1, "Y": x2, "Weight": weight}
    if bias is not None:
        inputs["Bias"] = bias

    helper = LayerHelper("bilinear", **locals())
    out = helper.create_variable_for_type_inference(dtype=x1.dtype)

    helper.append_op(
        type="bilinear_tensor_product", inputs=inputs, outputs={"Out": out})

    return out


753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769
def dropout(x,
            p=0.5,
            axis=None,
            training=True,
            mode="upscale_in_train",
            name=None):
    """
    Dropout is a regularization technique for reducing overfitting by preventing
    neuron co-adaption during training. The dropout operator randomly sets the
    outputs of some units to zero, while upscale others according to the given
    dropout probability.

    Args:
        x (Tensor): The input tensor. The data type is float32 or float64.
        p (float | int): Probability of setting units to zero. Default 0.5.
        axis (int | list): The axis along which the dropout is performed. Default None.
        training (bool): A flag indicating whether it is in train phrase or not. Default True.
770
        mode(str): ['upscale_in_train'(default) | 'downscale_in_infer'].
771 772 773 774 775 776 777 778 779 780

                           1. upscale_in_train(default), upscale the output at training time

                              - train: out = input * mask / ( 1.0 - dropout_prob )
                              - inference: out = input

                           2. downscale_in_infer, downscale the output at inference

                              - train: out = input * mask
                              - inference: out = input * (1.0 - dropout_prob)
781
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
782 783 784 785

    Returns:
        A Tensor representing the dropout, has same shape and data type as `x` .

786

787 788
    Examples:
        We use ``p=0.5`` in the following description for simplicity.
789

790
        1. When ``axis=None`` , this is commonly used dropout, which dropout each element of x randomly.
791 792 793

        ..  code-block:: text

794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818
            Let's see a simple case when x is a 2d tensor with shape 2*3:
            [[1 2 3]
             [4 5 6]]
            we generate mask with the same shape as x, which is 2*3. The value of mask is
            sampled from a Bernoulli distribution randomly. For example, we may get such mask:
            [[0 1 0]
             [1 0 1]]
            So the output is obtained from elementwise multiply of x and mask:
            [[0 2 0]
             [4 0 6]]
            Using default setting, i.e. ``mode='upscale_in_train'`` ,
            if in training phase, the final upscale output is:
            [[0 4 0 ]
             [8 0 12]]
            if in test phase, the output is the same as input:
            [[1 2 3]
             [4 5 6]]
            we can also set ``mode='downscale_in_infer'`` , then
            if in training phase, the final output is:
            [[0 2 0]
             [4 0 6]]
            if in test phase, the scale output is:
            [[0.5 1.  1.5]
             [2.  2.5 3. ]]

819 820


821
        2. When ``axis!=None`` , this is useful for dropping whole channels from an image or sequence.
822 823 824

        ..  code-block:: text

825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852
            Let's see the simple case when x is a 2d tensor with shape 2*3 again:
            [[1 2 3]
             [4 5 6]]
            (1) If ``axis=0`` , this means the dropout is only performed in axis `0` .
                we generate mask with the shape 2*1. Only in axis `0` the value is randomly selected.
                For example, we may get such mask:
                [[1]
                 [0]]
                The output is obtained from elementwise multiply of x and mask. Doing that the mask will be
                broadcast from 2*1 to 2*3:
                [[1 1 1]
                 [0 0 0]]
                and the result after elementwise multiply is:
                [[1 2 3]
                 [0 0 0]]
                then we can do upscale or downscale according to the setting of other arguments.
            (2) If ``axis=1`` , this means the dropout is only performed in axis `1` .
                we generate mask with the shape 1*3. Only in axis `1` the value is randomly selected.
                For example, we may get such mask:
                [[1 0 1]]
                Doing elementwise multiply the mask will be broadcast from 1*3 to 2*3:
                [[1 0 1]
                 [1 0 1]]
                and the result after elementwise multiply is:
                [[1 0 3]
                 [4 0 6]]
            (3) What about ``axis=[0, 1]`` ? This means the dropout is performed in all axes of x,
                which is the same case as default setting ``axis=None`` .
853
            (4) You may note that logically `axis=None` means the dropout is performed in none axis of x,
854 855 856 857 858 859 860 861 862 863
                We generate mask with the shape 1*1. Whole input is randomly selected or dropped.
                For example, we may get such mask:
                [[0]]
                Doing elementwise multiply the mask will be broadcast from 1*1 to 2*3:
                [[0 0 0]
                 [0 0 0]]
                and the result after elementwise multiply is:
                [[0 0 0]
                 [0 0 0]]
                Actually this is not what we want because all elements may set to zero~
864 865 866

        When x is a 4d tensor with shape `NCHW`, we can set ``axis=[0,1]`` and the dropout will be performed in channel `N` and `C`, `H` and `W` is tied, i.e. paddle.nn.dropout(x, p, axis=[0,1]) . Please refer to ``paddle.nn.functional.dropout2d`` for more details.
        Similarly, when x is a 5d tensor with shape `NCDHW`, we can set ``axis=[0,1]`` to perform dropout3d. Please refer to ``paddle.nn.functional.dropout3d`` for more details.
867 868

        .. code-block:: python
869

870 871 872 873 874 875 876 877 878 879
            import paddle
            import numpy as np

            x = np.array([[1,2,3], [4,5,6]]).astype('float32')
            x = paddle.to_tensor(x)
            y_train = paddle.nn.functional.dropout(x, 0.5)
            y_test = paddle.nn.functional.dropout(x, 0.5, training=False) 
            y_0 = paddle.nn.functional.dropout(x, axis=0)
            y_1 = paddle.nn.functional.dropout(x, axis=1)
            y_01 = paddle.nn.functional.dropout(x, axis=[0,1])
880 881 882 883 884 885
            print(x)
            print(y_train)
            print(y_test)
            print(y_0)
            print(y_1)
            print(y_01)
886 887

    """
888 889 890 891
    # fast return for p == 0
    if p == 0:
        return x

892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958
    if not isinstance(p, (float, int)):
        raise TypeError("p argument should be a number")
    if p < 0 or p > 1:
        raise ValueError("p argument should between 0 and 1")
    if mode not in ('downscale_in_infer', 'upscale_in_train'):
        raise ValueError(
            "mode argument should be 'downscale_in_infer' or 'upscale_in_train'")
    if axis and not isinstance(axis, (int, list)):
        raise TypeError("datatype of axis argument should be int or list")

    if axis == None:  # commonly used dropout
        seed = None
        mode = 'downgrade_in_infer' if mode == 'downscale_in_infer' else mode  #semantic transfer

        def get_attrs(prog, dropout_prob, is_test, seed):
            if (seed is None or seed == 0) and prog.random_seed != 0:
                seed = prog.random_seed
            attrs = {
                'dropout_prob': dropout_prob,
                'is_test': is_test,
                'fix_seed': seed is not None,
                'seed': seed if seed is not None else 0,
                'dropout_implementation': mode,
            }
            return attrs

        if in_dygraph_mode():
            if default_main_program().random_seed != 0:
                seed = default_main_program().random_seed
            out, mask = core.ops.dropout(
                x, 'dropout_prob', p, 'is_test', not training, 'fix_seed',
                seed is not None, 'seed', seed
                if seed is not None else 0, 'dropout_implementation', mode)
            return out

        helper = LayerHelper('dropout', **locals())
        check_variable_and_dtype(x, 'x', ['float16', 'float32', 'float64'],
                                 'dropout')

        out = helper.create_variable_for_type_inference(dtype=x.dtype)
        mask = helper.create_variable_for_type_inference(
            dtype=core.VarDesc.VarType.UINT8, stop_gradient=True)

        attrs = get_attrs(helper.main_program, p, not training, seed)

        helper.append_op(
            type='dropout',
            inputs={'X': [x]},
            outputs={'Out': [out],
                     'Mask': [mask]},
            attrs=attrs)
        return out
    else:  #sometimes called dropout_nd #TODO: optimize with c++
        if not in_dygraph_mode():
            check_variable_and_dtype(x, 'x', ['float32', 'float64'], 'dropout')
        dtype = x.dtype
        keep_prob = 1 - p
        if training:
            if p == 1.:
                return layers.scale(x, scale=0.)

            scale_input = layers.scale(
                x, scale=1 / keep_prob) if mode == 'upscale_in_train' else x

            #get mask shape
            input_shape = x.shape
            drop_axes = [axis] if isinstance(axis, int) else axis
959 960
            if min(drop_axes) < 0 or max(drop_axes) > len(input_shape) - 1:
                raise ValueError("axis value should be greater than or equal to 0 and less than dimensions of x:{}, but get axis value:{} " \
961 962 963
                                 .format(len(input_shape), max(drop_axes)))
            if len(drop_axes) > len(input_shape):
                raise ValueError(
964
                    "length of axis should not be greater than dimensions of x:{}, but get length of axis: {}".
965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998
                    format(len(input_shape), len(drop_axes)))
            mask_shape = [1] * len(input_shape)
            for i in drop_axes:
                mask_shape[i] = input_shape[i]

            #get mask
            random_tensor = layers.uniform_random(
                mask_shape, dtype='float32', min=0., max=1.0)
            p = layers.fill_constant(shape=[1], dtype='float32', value=p)
            keep_mask = layers.greater_equal(random_tensor, p)

            scale_input = layers.cast(scale_input, dtype)
            keep_mask = layers.cast(keep_mask, dtype)
            ret = paddle.multiply(scale_input, keep_mask, name=name)
            return ret
        else:  # test
            ret = layers.scale(
                x, scale=keep_prob) if mode == 'downscale_in_infer' else x
            return ret


def dropout2d(x, p=0.5, training=True, data_format='NCHW', name=None):
    """
    Randomly zero out entire channels (in the batched input 4d tensor with the shape `NCHW` ,
    a channel is a 2D feature map with the shape `HW` ). Each channel will be zeroed out independently
    on every forward call with probability `p` using samples from a Bernoulli distribution.

    See ``paddle.nn.functional.dropout`` for more details.

    Args:
        x (Tensor):  The input is 4-D Tensor with shape [N, C, H, W] or [N, H, W, C].
                     The data type is float32 or float64.
        p (float): Probability of setting units to zero. Default 0.5.
        training (bool): A flag indicating whether it is in train phrase or not. Default True.
999
        data_format (str, optional): Specify the data format of the input, and the data format of the output will be consistent with that of the input. An optional string from `NCHW` or `NHWC` . The default is `NCHW` . When it is `NCHW` , the data is stored in the order of: [batch_size, input_channels, input_height, input_width].
1000 1001 1002 1003 1004
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        A Tensor representing the dropout2d, has same shape and data type as `x` .

1005

1006 1007
    Examples:
        .. code-block:: python
1008

1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053
            import paddle
            import numpy as np

            x = np.random.random(size=(2, 3, 4, 5)).astype('float32')
            x = paddle.to_tensor(x)
            y_train = paddle.nn.functional.dropout2d(x)  #train
            y_test = paddle.nn.functional.dropout2d(x, training=False) #test
            for i in range(2):
                for j in range(3):
                    print(x.numpy()[i,j,:,:])
                    print(y_train.numpy()[i,j,:,:]) # may all 0
                    print(y_test.numpy()[i,j,:,:])
    """
    input_shape = x.shape
    if len(input_shape) != 4:
        raise ValueError("dimensions of x should be 4, but received {} != 4"\
        .format(len(input_shape)))

    if data_format not in ["NCHW", "NHWC"]:
        raise ValueError(
            "Attr(data_format) should be 'NCHW' or 'NHWC'. Received "
            "Attr(data_format): %s." % str(data_format))

    return dropout(
        x,
        p=p,
        axis=[0, 1] if data_format == 'NCHW' else [0, 3],
        training=training,
        mode="upscale_in_train",
        name=name)


def dropout3d(x, p=0.5, training=True, data_format='NCDHW', name=None):
    """
    Randomly zero out entire channels (in the batched input 5d tensor with the shape `NCDHW` ,
    a channel is a 3D feature map with the shape `DHW` ). Each channel will be zeroed out independently
    on every forward call with probability `p` using samples from a Bernoulli distribution.

    See ``paddle.nn.functional.dropout`` for more details.

    Args:
        x (Tensor):  The input is 5-D Tensor with shape [N, C, D, H, W] or [N, D, H, W, C].
                     The data type is float32 or float64.
        p (float): Probability of setting units to zero. Default 0.5.
        training (bool): A flag indicating whether it is in train phrase or not. Default True.
1054
        data_format (str, optional): Specify the data format of the input, and the data format of the output will be consistent with that of the input. An optional string from ``NCDHW`` or ``NDHWC``. The default is ``NCDHW`` . When it is ``NCDHW`` , the data is stored in the order of: [batch_size, input_channels, input_depth, input_height, input_width].
1055 1056 1057 1058 1059
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        A Tensor representing the dropout3d, has same shape and data type with `x` .

1060

1061 1062
    Examples:
        .. code-block:: python
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
            import paddle
            import numpy as np

            x = np.random.random(size=(2, 3, 4, 5, 6)).astype('float32')
            x = paddle.to_tensor(x)
            y_train = paddle.nn.functional.dropout3d(x)  #train
            y_test = paddle.nn.functional.dropout3d(x, training=False) #test
            print(x.numpy()[0,0,:,:,:])
            print(y_train.numpy()[0,0,:,:,:]) # may all 0
            print(y_test.numpy()[0,0,:,:,:])
    """

    input_shape = x.shape
    if len(input_shape) != 5:
        raise ValueError("dimensions of x should be 5, but received {} != 5" \
        .format(len(input_shape)))

    if data_format not in ["NCDHW", "NDHWC"]:
        raise ValueError(
            "Attr(data_format) should be 'NCDHW' or 'NDHWC'. Received "
            "Attr(data_format): %s." % str(data_format))

    return dropout(
        x,
        p=p,
        axis=[0, 1] if data_format == 'NCDHW' else [0, 4],
        training=training,
        mode="upscale_in_train",
        name=name)


1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112
def alpha_dropout(x, p=0.5, training=True, name=None):
    """
    Alpha Dropout is a type of Dropout that maintains the self-normalizing property.
    For an input with zero mean and unit standard deviation, the output of Alpha Dropout
    maintains the original mean and standard deviation of the input.
    Alpha Dropout fits well to SELU activate function by randomly setting activations to the negative saturation value.

    Args:
        x (Tensor): The input tensor. The data type is float32 or float64.
        p (float | int): Probability of setting units to zero. Default 0.5.
        training (bool): A flag indicating whether it is in train phrase or not. Default True.
        name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.

    Returns:
        A Tensor representing the dropout, has same shape and data type as `x`.

    Examples:
        .. code-block:: python
1113

1114 1115 1116 1117 1118 1119 1120
            import paddle
            import numpy as np

            x = np.array([[-1, 1], [-1, 1]]).astype('float32')
            x = paddle.to_tensor(x)
            y_train = paddle.nn.functional.alpha_dropout(x, 0.5)
            y_test = paddle.nn.functional.alpha_dropout(x, 0.5, training=False)
1121 1122
            print(x)
            print(y_train)
1123
            # [[-0.10721093, 1.6655989 ], [-0.7791938, -0.7791938]] (randomly)
1124
            print(y_test)
1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135
    """
    if not isinstance(p, (float, int)):
        raise TypeError("p argument should be a float or int")
    if p < 0 or p > 1:
        raise ValueError("p argument should between 0 and 1")

    if not in_dygraph_mode():
        check_variable_and_dtype(x, 'x', ['float32', 'float64'],
                                 'alpha_dropout')

    if training:
1136 1137
        if p == 1:
            return layers.scale(x, scale=0.)
1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170
        #get transformation params
        alpha = 1.6732632423543772848170429916717
        scale = 1.0507009873554804934193349852946
        alpha_p = -alpha * scale
        a = ((1 - p) * (1 + p * alpha_p**2))**-0.5
        b = -a * alpha_p * p

        dtype = x.dtype
        input_shape = x.shape

        #get mask
        random_tensor = layers.uniform_random(
            input_shape, dtype='float32', min=0., max=1.0)
        p = layers.fill_constant(shape=[1], dtype='float32', value=p)
        keep_mask = layers.greater_equal(random_tensor, p)
        keep_mask = layers.cast(keep_mask, dtype)
        drop_mask = layers.elementwise_sub(
            layers.fill_constant(
                shape=input_shape, dtype=dtype, value=1.),
            keep_mask)

        #apply mask
        b = layers.fill_constant(shape=[1], dtype=dtype, value=b)
        y = layers.elementwise_add(
            paddle.multiply(x, keep_mask),
            layers.scale(
                drop_mask, scale=alpha_p))
        res = layers.elementwise_add(layers.scale(y, scale=a), b, name=name)
        return res
    else:  # test
        return x


L
littletomatodonkey 已提交
1171 1172 1173
def pad(x, pad, mode='constant', value=0, data_format="NCHW", name=None):
    """
    Pad tensor according to 'pad' and 'mode'.
L
littletomatodonkey 已提交
1174 1175 1176
    If mode is 'constant' and length of pad is twice as length of x dimension,
    then the padding will be started from the first dimension and moved back onto x
    according to 'pad' and 'value'.
L
littletomatodonkey 已提交
1177 1178 1179 1180 1181 1182 1183 1184 1185 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 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244
    If mode is 'reflect', pad[0] and pad[1] must be no greater
    than width-1. The height and depth dimension has the same condition.

    Parameters:
        x (Tensor): The input tensor with data type float32/double/int32/int64_t.
        pad (Tensor | List[int32]): The padding size with data type int32. [len(padding)/2] dimensions
            of input will be padded. 1. If input dimension is 3, then the pad has the form (pad_left,
            pad_right). 2. If the input dimension is 4, then the pad has the form (pad_left, pad_right, 
            pad_top, pad_bottom). 3. If the input dimension is 5, then the pad has the form 
            (pad_left, pad_right, pad_top, pad_bottom, pad_front, pad_back).
            
        mode (str): Four modes: 'constant' (default), 'reflect', 'replicate', 'circular'.
            When in 'constant' mode, this op uses a constant value to pad the input tensor.
            When in 'reflect' mode, uses reflection of the input boundaries to pad the input tensor.
            When in 'replicate' mode, uses input boundaries to pad the input tensor.
            When in 'circular' mode, uses circular input to pad the input tensor.
            Default is 'constant'
        value (float32): The value to fill the padded areas in 'constant' mode . Default is 0.0
        data_format (str): An string from: "NCL", "NLC", NHWC", "NCHW", "NCDHW", "NDHWC". Specify the data format of
           the input data.
           Default is  "NCHW"
        name (str, optional) : The default value is None.  Normally there is no need for
            user to set this property.  For more information, please refer to :ref:`api_guide_Name`.
                    
    Returns: a Tensor padded according to pad and mode and data type is same as input.
    Return Type: Tensor

    Examples:
        .. code-block:: text

            x = [[[[[1., 2., 3.],
                    [4., 5., 6.]]]]]

            Case 0:
                pad = [2, 2, 1, 1, 0, 0],
                mode = 'constant'
                value = 0
                Out = [[[[[0. 0. 0. 0. 0. 0. 0.]
                          [0. 0. 1. 2. 3. 0. 0.]
                          [0. 0. 4. 5. 6. 0. 0.]
                          [0. 0. 0. 0. 0. 0. 0.]]]]]

            Case 1:
                pad = [2, 2, 1, 1, 0, 0],
                mode = 'reflect'
                Out = [[[[[6. 5. 4. 5. 6. 5. 4.]
                          [3. 2. 1. 2. 3. 2. 1.]
                          [6. 5. 4. 5. 6. 5. 4.]
                          [3. 2. 1. 2. 3. 2. 1.]]]]]

            Case 2:
                pad = [2, 2, 1, 1, 0, 0],
                mode = 'replicate'
                Out = [[[[[1. 1. 1. 2. 3. 3. 3.]
                          [1. 1. 1. 2. 3. 3. 3.]
                          [4. 4. 4. 5. 6. 6. 6.]
                          [4. 4. 4. 5. 6. 6. 6.]]]]]

            Case 3:
                pad = [2, 2, 1, 1, 0, 0],
                mode = 'circular'
                Out = [[[[[5. 6. 4. 5. 6. 4. 5.]
                          [2. 3. 1. 2. 3. 1. 2.]
                          [5. 6. 4. 5. 6. 4. 5.]
                          [2. 3. 1. 2. 3. 1. 2.]]]]]

    Code Examples:
        .. code-block:: python
L
littletomatodonkey 已提交
1245

L
littletomatodonkey 已提交
1246 1247 1248 1249 1250 1251
            import numpy as np
            import paddle
            import paddle.nn.functional as F
            
            # example 1
            x_shape = (1, 1, 3)
L
littletomatodonkey 已提交
1252 1253 1254
            x = paddle.arange(np.prod(x_shape), dtype="float32").reshape(x_shape) + 1
            y = F.pad(x, [2, 3], value=1, mode='constant', data_format="NCL")
            print(y)
L
littletomatodonkey 已提交
1255
            # [[[1. 1. 1. 2. 3. 1. 1. 1.]]]
1256
            
L
littletomatodonkey 已提交
1257 1258
            # example 2
            x_shape = (1, 1, 2, 3)
L
littletomatodonkey 已提交
1259 1260 1261
            x = paddle.arange(np.prod(x_shape), dtype="float32").reshape(x_shape) + 1
            y = F.pad(x, [1, 2, 1, 1], value=1, mode='circular')
            print(y)
L
littletomatodonkey 已提交
1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276
            # [[[[6. 4. 5. 6. 4. 5.]
            #    [3. 1. 2. 3. 1. 2.]
            #    [6. 4. 5. 6. 4. 5.]
            #    [3. 1. 2. 3. 1. 2.]]]]
    """
    assert mode in ['reflect', 'replicate', 'constant', 'circular'], \
            "mode should be one of constant, reflect, replicate, circular, but got {}.".format(mode)

    data_format = data_format.upper()
    assert data_format in ["NCL", "NCHW", "NCDHW", "NLC", "NHWC", "NDHWC"], \
        "data_format should be in one of [NCL, NCHW, NCDHW, NLC, NHWC, NDHWC], " \
        "but got {}".format(data_format)

    x_dim = len(x.shape)

L
littletomatodonkey 已提交
1277 1278 1279
    if mode == "constant" and isinstance(pad, list) and len(pad) == x_dim * 2:
        return layers.pad(x, pad, pad_value=value)

1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292
    assert x_dim in [
        3, 4, 5
    ], "input tesor dimension must be in [3, 4, 5] but got {}".format(x_dim)

    supported_format_map = {
        3: ["NCL", "NLC"],
        4: ["NCHW", "NHWC"],
        5: ["NCDHW", "NDHWC"],
    }
    assert data_format in supported_format_map[x_dim], \
    "input tensor dimension is {}, it's data format should be in {} but got {}".format(
        x_dim, supported_format_map[x_dim], data_format)

L
littletomatodonkey 已提交
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 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364
    unsqueezed_dim = []

    if isinstance(pad, Variable):
        if data_format in ["NCL", "NCHW", "NCDHW"]:
            data_format = "NCDHW"
            if x_dim == 3:
                pad = concat([zeros((4, ), dtype="int32"), pad], axis=0)
                unsqueezed_dim = [3, 4]
                x = unsqueeze(x, axes=unsqueezed_dim)
            elif x_dim == 4:
                pad = concat([pad, zeros((2, ), dtype="int32")], axis=0)
                unsqueezed_dim = [2]
                x = unsqueeze(x, axes=unsqueezed_dim)
        elif data_format in ["NLC", "NHWC", "NDHWC"]:
            data_format = "NDHWC"
            if x_dim == 3:
                pad = concat([zeros((4, ), dtype="int32"), pad], axis=0)
                unsqueezed_dim = [2, 3]
                x = unsqueeze(x, axes=unsqueezed_dim)
            elif x_dim == 4:
                pad = concat([pad, zeros((2, ), dtype="int32")], axis=0)
                unsqueezed_dim = [1]
                x = unsqueeze(x, axes=unsqueezed_dim)
    else:
        if data_format in ["NCL", "NCHW", "NCDHW"]:
            data_format = "NCDHW"
            if x_dim == 3:
                pad = [0, 0, 0, 0] + pad
                unsqueezed_dim = [3, 4]
                x = unsqueeze(x, axes=unsqueezed_dim)
            elif x_dim == 4:
                pad = pad + [0, 0]
                unsqueezed_dim = [2]
                x = unsqueeze(x, axes=unsqueezed_dim)
        elif data_format in ["NLC", "NHWC", "NDHWC"]:
            data_format = "NDHWC"
            if x_dim == 3:
                pad = [0, 0, 0, 0] + pad
                unsqueezed_dim = [2, 3]
                x = unsqueeze(x, axes=unsqueezed_dim)
            elif x_dim == 4:
                pad = pad + [0, 0]
                unsqueezed_dim = [1]
                x = unsqueeze(x, axes=unsqueezed_dim)

    if in_dygraph_mode():
        if isinstance(pad, Variable):
            pad = pad.numpy()
        out = core.ops.pad3d(x, "paddings", pad, "mode", mode, "value", value,
                             "data_format", data_format, "name", name)
    else:
        attrs = {'mode': mode, 'value': value, 'data_format': data_format}
        inputs = {'X': [x]}
        if isinstance(pad, Variable):
            inputs['Paddings'] = [pad]
            attrs['paddings'] = []
        else:
            attrs['paddings'] = pad

        helper = LayerHelper('pad3d', **locals())

        dtype = helper.input_dtype(input_param_name='input')
        out = helper.create_variable_for_type_inference(dtype)
        helper.append_op(
            type='pad3d', inputs=inputs, outputs={"Out": out}, attrs=attrs)

    if len(unsqueezed_dim) != 0:
        out = squeeze(out, axes=unsqueezed_dim)

    return out


Y
Yang Zhang 已提交
1365
def cosine_similarity(x1, x2, axis=1, eps=1e-8):
L
littletomatodonkey 已提交
1366
    """
Y
Yang Zhang 已提交
1367
    Compute cosine similarity between x1 and x2 along axis.
L
littletomatodonkey 已提交
1368 1369 1370 1371

    Parameters:
        x1 (Tensor): First input. float32/double.
        x2 (Tensor): Second input. float32/double.
Y
Yang Zhang 已提交
1372
        axis (int): Dimension of vectors to compute cosine similarity. Default is 1.
L
littletomatodonkey 已提交
1373 1374
        eps(float): Small value to avoid division by zero. Default is 1e-8.
                    
Y
Yang Zhang 已提交
1375
    Returns: a Tensor representing cosine similarity between x1 and x2 along axis.
L
littletomatodonkey 已提交
1376 1377 1378 1379
    Return Type: Tensor

    Examples:
        .. code-block:: text
1380

L
littletomatodonkey 已提交
1381 1382 1383 1384 1385 1386 1387 1388 1389
            Case 0:
                x1 = [[0.8024077  0.9927354  0.27238318 0.8344984 ]
                     [0.48949873 0.5797396  0.65444374 0.66510963]
                     [0.1031398  0.9614342  0.08365563 0.6796464 ]
                     [0.10760343 0.7461209  0.7726148  0.5801006 ]]
                x2 = [[0.62913156 0.1536727  0.9847992  0.04591406]
                     [0.9098952  0.15715368 0.8671125  0.3156102 ]
                     [0.4427798  0.54136837 0.5276275  0.32394758]
                     [0.3769419  0.8535014  0.48041078 0.9256797 ]]
Y
Yang Zhang 已提交
1390
                axis = 1
L
littletomatodonkey 已提交
1391 1392 1393 1394 1395
                eps = 1e-8
                Out: [0.5275037  0.8368967  0.75037485 0.9245899]

    Code Examples:
        .. code-block:: python
1396

L
littletomatodonkey 已提交
1397 1398 1399 1400 1401 1402 1403 1404 1405
            import paddle
            import paddle.nn as nn
            import numpy as np

            np.random.seed(0)
            x1 = np.random.rand(2,3)
            x2 = np.random.rand(2,3)
            x1 = paddle.to_tensor(x1)
            x2 = paddle.to_tensor(x2)
Y
Yang Zhang 已提交
1406
            result = paddle.nn.functional.cosine_similarity(x1, x2, axis=0)
L
littletomatodonkey 已提交
1407
            print(result)
L
littletomatodonkey 已提交
1408 1409 1410
            # [0.99806249 0.9817672  0.94987036]
            
    """
1411 1412 1413
    w12 = sum(paddle.multiply(x1, x2), axis=axis)
    w1 = sum(paddle.multiply(x1, x1), axis=axis)
    w2 = sum(paddle.multiply(x2, x2), axis=axis)
Y
Yang Zhang 已提交
1414
    n12 = sqrt(clip(w1 * w2, min=eps * eps))
L
littletomatodonkey 已提交
1415 1416
    cos_sim = w12 / n12
    return cos_sim
1417 1418 1419


def linear(x, weight, bias=None, name=None):
1420
    r"""
1421

1422 1423
    Fully-connected linear transformation operator. For each input :math:`X` ,
    the equation is:
1424 1425 1426

    .. math::

1427
        Out = XW + b
1428

1429
    where :math:`W` is the weight and :math:`b` is the bias.
1430

1431 1432 1433 1434 1435 1436 1437
    If the weight is a 2-D tensor of shape :math:`[in\_features, out\_features]` ,
    input should be a multi-dimensional tensor of shape
    :math:`[batch\_size, *, in\_features]` , where :math:`*` means any number of
    additional dimensions. The linear operator multiplies input tensor with
    weight and produces an output tensor of shape :math:`[batch\_size, *, out\_features]` , 
    If :math:`bias` is not None, the bias should be a 1-D tensor of shape
    :math:`[out\_features]` and will be added to the output.
1438

1439 1440 1441 1442 1443 1444 1445
    Parameters:
        x (Tensor): Input tensor. The data type should be float16, float32 or float64.
        weight (Tensor): Weight tensor. The data type should be float16, float32 or float64.
        bias (Tensor, optional): Bias tensor. The data type should be float16, float32 or float64.
                                 If it is set to None, no bias will be added to the output units.
        name (str, optional): Normally there is no need for user to set this parameter.
                              For detailed information, please refer to :ref:`api_guide_Name` .
1446 1447

    Returns:
1448 1449
        Tensor, the shape is :math:`[batch\_size, *, out\_features]` and the
        data type is the same with input :math:`x` .
1450 1451 1452 1453 1454 1455

    Examples:
        .. code-block:: python
          
          import paddle
          
1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468
          x = paddle.randn((3, 2), dtype="float32")
          # x: [[-0.32342386 -1.200079  ]
          #     [ 0.7979031  -0.90978354]
          #     [ 0.40597573  1.8095392 ]]
          weight = paddle.full(shape=[2, 4], fill_value="0.5", dtype="float32", name="weight")
          # weight: [[0.5 0.5 0.5 0.5]
          #          [0.5 0.5 0.5 0.5]]
          bias = paddle.ones(shape=[4], dtype="float32", name="bias")
          # bias: [1. 1. 1. 1.]
          y = paddle.nn.functional.linear(x, weight, bias)
          # y: [[0.23824859 0.23824859 0.23824859 0.23824859]
          #     [0.9440598  0.9440598  0.9440598  0.9440598 ]
          #     [2.1077576  2.1077576  2.1077576  2.1077576 ]]
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 1503
    """
    if in_dygraph_mode():
        pre_bias = _varbase_creator(dtype=x.dtype)
        core.ops.matmul(x, weight, pre_bias, 'transpose_X', False,
                        'transpose_Y', False, "alpha", 1)
        return dygraph_utils._append_bias_in_dygraph(
            pre_bias, bias, axis=len(x.shape) - 1)
    else:
        helper = LayerHelper('linear', **locals())
        dtype = x.dtype

        check_variable_and_dtype(x, 'x', ['float16', 'float32', 'float64'],
                                 'linear')
        check_dtype(dtype, 'dtype', ['float16', 'float32', 'float64'], 'linear')

        inputs = {'X': [x], 'Y': [weight]}
        attrs = {
            'transpose_X': False,
            'transpose_Y': False,
            'alpha': 1,
        }
        tmp = helper.create_variable_for_type_inference(dtype)
        helper.append_op(
            type='matmul', inputs=inputs, outputs={'Out': tmp}, attrs=attrs)
        if bias is not None:
            res = helper.create_variable_for_type_inference(dtype)
            helper.append_op(
                type='elementwise_add',
                inputs={'X': [tmp],
                        'Y': [bias]},
                outputs={'Out': [res]},
                attrs={'axis': len(x.shape) - 1})
        else:
            res = tmp
        return res
1504 1505 1506


def label_smooth(label, prior_dist=None, epsilon=0.1, name=None):
1507
    r"""
1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
    Label smoothing is a mechanism to regularize the classifier layer and is called
    label-smoothing regularization (LSR).

    Label smoothing is proposed to encourage the model to be less confident,
    since optimizing the log-likelihood of the correct label directly may
    cause overfitting and reduce the ability of the model to adapt. Label
    smoothing replaces the ground-truth label :math:`y` with the weighted sum
    of itself and some fixed distribution :math:`\mu`. For class :math:`k`,
    i.e.

    .. math::

        \\tilde{y_k} = (1 - \epsilon) * y_k + \epsilon * \mu_k,

    where :math:`1 - \epsilon` and :math:`\epsilon` are the weights
    respectively, and :math:`\\tilde{y}_k` is the smoothed label. Usually
    uniform distribution is used for :math:`\mu`.

    See more details about label smoothing in https://arxiv.org/abs/1512.00567.

    Parameters:
        label(Tensor): The input variable containing the label data. The
                        label data should use one-hot representation. It's
                        a multidimensional tensor with a shape of
                        :math:`[N_1, ..., Depth]`, where Depth is class number. The dtype can be "float32" and "float64".
        prior_dist(Tensor, optional): The prior distribution to be used to smooth
                        labels. If not provided, an uniform distribution
                        is used. It's a multidimensional tensor with a shape of
                        :math:`[1, class\_num]` . The default value is None.
        epsilon(float, optional): The weight used to mix up the original ground-truth
                        distribution and the fixed distribution. The default value is
                        0.1.
        name(str, optional): The default value is None. Normally there is no need for user
                        to set this property. For more information, please refer to
                        :ref:`api_guide_Name`.

    Returns:
        Tensor: The tensor containing the smoothed labels.

    Examples:
        .. code-block:: python

            import paddle
            import numpy as np
            
            x_data = np.array([[[0, 1, 0],
                                [ 1,  0, 1]]]).astype("float32")
            print(x_data.shape)
            paddle.disable_static()
            x = paddle.to_tensor(x_data, stop_gradient=False)
            output = paddle.nn.functional.label_smooth(x)
1559
            print(output)
1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583
            
            #[[[0.03333334 0.93333334 0.03333334]
            #  [0.93333334 0.03333334 0.93333334]]]
    """
    if epsilon > 1. or epsilon < 0.:
        raise ValueError("The value of epsilon must be between 0 and 1.")

    if in_dygraph_mode():
        return core.ops.label_smooth(label, prior_dist, 'epsilon',
                                     float(epsilon))

    check_variable_and_dtype(label, 'label', ['float32', 'float64'],
                             'label_smooth')

    helper = LayerHelper("label_smooth", **locals())
    label.stop_gradient = True
    smooth_label = helper.create_variable_for_type_inference(label.dtype)
    helper.append_op(
        type="label_smooth",
        inputs={"X": label,
                "PriorDist": prior_dist} if prior_dist else {"X": label},
        outputs={"Out": smooth_label},
        attrs={"epsilon": float(epsilon)})
    return smooth_label