Skip to content

P
Paddle

wmsofts / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 0
Fork 0
P
Paddle

体验新版 GitCode，发现更多精彩内容 >>

未验证 提交 91f4d1ce 编写于 作者: U ustiniankw 提交者： GitHub
浏览文件
操作

Fixdocs (#47986) 

* list112-122, test=document_fix

* precommitfix, test=document_fix

* list112-127, test=document_fix

* fix_ResNetBasicBlock, test=document_fix

* pre-commit_resnet, test=document_fix

* refix, test=document

* refix, test=document_fix
上级 94c6ec86
隐藏空白更改
内联 并排
Showing with 235 addition and 162 deletion
python/paddle/fft.py
浏览文件 @ 91f4d1ce
......@@ -626,6 +626,7 @@ def ifftn(x, s=None, axes=None, norm="backward", name=None):
def rfftn(x, s=None, axes=None, norm="backward", name=None):
"""
The N dimensional FFT for real input.
This function computes the N-dimensional discrete Fourier Transform over
......@@ -659,9 +660,9 @@ def rfftn(x, s=None, axes=None, norm="backward", name=None):
three operations are shown below:
- "backward": The factor of forward direction and backward direction are ``1``
and ``1/n`` respectively;
and ``1/n`` respectively;
- "forward": The factor of forward direction and backward direction are ``1/n``
and ``1`` respectively;
and ``1`` respectively;
- "ortho": The factor of forward direction and backword direction are both ``1/sqrt(n)``.
Where ``n`` is the multiplication of each element in ``s`` .
......@@ -670,36 +671,35 @@ def rfftn(x, s=None, axes=None, norm="backward", name=None):
refer to :ref:`api_guide_Name` .
Returns:
out(Tensor): complex tensor
out(Tensor), complex tensor
Examples:
.. code-block:: python
.. code-block:: python
import paddle
import paddle
# default, all axis will be used to exec fft
x = paddle.ones((2, 3, 4))
print(paddle.fft.rfftn(x))
# Tensor(shape=[2, 3, 3], dtype=complex64, place=CUDAPlace(0), stop_gradient=True,
# [[[(24+0j), 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]],
#
# [[0j , 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]]])
# use axes(2, 0)
print(paddle.fft.rfftn(x, axes=(2, 0)))
# Tensor(shape=[2, 3, 3], dtype=complex64, place=CUDAPlace(0), stop_gradient=True,
# [[[(8+0j), 0j , 0j ],
# [(8+0j), 0j , 0j ],
# [(8+0j), 0j , 0j ]],
#
# [[0j , 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]]])
# default, all axis will be used to exec fft
x = paddle.ones((2, 3, 4))
print(paddle.fft.rfftn(x))
# Tensor(shape=[2, 3, 3], dtype=complex64, place=CUDAPlace(0), stop_gradient=True,
# [[[(24+0j), 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]],
#
# [[0j , 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]]])
# use axes(2, 0)
print(paddle.fft.rfftn(x, axes=(2, 0)))
# Tensor(shape=[2, 3, 3], dtype=complex64, place=CUDAPlace(0), stop_gradient=True,
# [[[(8+0j), 0j , 0j ],
# [(8+0j), 0j , 0j ],
# [(8+0j), 0j , 0j ]],
#
# [[0j , 0j , 0j ],
# [0j , 0j , 0j ],
# [0j , 0j , 0j ]]])
"""
return fftn_r2c(x, s, axes, norm, forward=True, onesided=True, name=name)
......
python/paddle/fluid/contrib/sparsity/supported_layer_list.py
浏览文件 @ 91f4d1ce
......@@ -82,15 +82,17 @@ supported_layers_and_prune_func_map = {}
def add_supported_layer(layer, pruning_func=None):
r"""
Add supported layers and its corresponding pruning function.
Args:
name (string|Layer): The name or type of layer, needed to support. If layer is `Layer` then
it would be turn to string internally. ASP would use this name to match parameter's name and call
its the corresponding pruning function.
it would be turn to string internally. ASP would use this name to match parameter's name and call
its the corresponding pruning function.
pruning_func (function, optional): a function type which receives five argument (weight_nparray,
m, n, func_name, param_name), weight_nparray is a nparray of weight, param_name is the name of weight,
m, n, and func_name, please see `prune_model` for details.
m, n, func_name, param_name), weight_nparray is a nparray of weight, param_name is the name of weight,
m, n, and func_name, please see `prune_model` for details.
"""
name = None
if isinstance(layer, str):
......
python/paddle/fluid/contrib/sparsity/utils.py
浏览文件 @ 91f4d1ce
......@@ -92,20 +92,25 @@ class CheckMethod(Enum):
def calculate_density(x):
r"""
Return the density of the input tensor.
Args:
x (nparray): The input tensor.
Returns:
float: The density of :attr:`x`.
float, The density of :attr:`x`.
Examples:
.. code-block:: python
import paddle
import numpy as np
x = np.array([[0, 1, 3, 0],
import paddle
import numpy as np
x = np.array([[0, 1, 3, 0],
[1, 1, 0, 1]])
paddle.incubate.asp.calculate_density(x) # 0.625
paddle.incubate.asp.calculate_density(x) # 0.625
"""
x_flattened = x.flatten()
return float(np.nonzero(x_flattened)[0].size) / x_flattened.size
......
python/paddle/fluid/dygraph/layers.py
浏览文件 @ 91f4d1ce
......@@ -177,13 +177,14 @@ class Layer:
def train(self):
"""
Sets this Layer and all its sublayers to training mode.
This only effects certain modules like `Dropout` and `BatchNorm`.
Returns:
None
Example::
Examples:
.. code-block:: python
import paddle
......@@ -260,6 +261,7 @@ class Layer:
def apply(self, fn):
"""
Applies ``fn`` recursively to every sublayer (as returned by ``.sublayers()``)
as well as self. Typical use includes initializing the parameters of a model.
......@@ -267,7 +269,7 @@ class Layer:
fn (function): a function to be applied to each sublayer
Returns:
Layer: self
Layer, self
Example::
.. code-block:: python
......@@ -287,6 +289,7 @@ class Layer:
net.apply(init_weights)
print(net.state_dict())
"""
for layer in self.children():
layer.apply(fn)
......@@ -296,10 +299,12 @@ class Layer:
return self
def full_name(self):
"""Full name for this layer, composed by name_scope + "/" + MyLayer.__class__.__name__
"""
Full name for this layer, composed by name_scope + "/" + MyLayer.__class__.__name__
Returns:
str: full name of this layer.
str, full name of this layer.
Example::
.. code-block:: python
......@@ -321,7 +326,9 @@ class Layer:
return self._full_name
def register_forward_post_hook(self, hook):
"""Register a forward post-hook for Layer. The hook will be called after `forward` function has been computed.
"""
Register a forward post-hook for Layer. The hook will be called after `forward` function has been computed.
It should have the following form, `input` and `output` of the `hook` is `input` and `output` of the `Layer` respectively.
User can use forward post-hook to change the output of the Layer or perform information statistics tasks on the Layer.
......@@ -332,7 +339,7 @@ class Layer:
hook(function): a function registered as a forward post-hook
Returns:
HookRemoveHelper: a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .
HookRemoveHelper, a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .
Examples:
.. code-block:: python
......@@ -364,13 +371,16 @@ class Layer:
# hook change the linear's output to output * 2, so out0 is equal to out1 * 2.
assert (out0.numpy() == (out1.numpy()) * 2).any()
"""
hook_remove_helper = HookRemoveHelper(self._forward_post_hooks)
self._forward_post_hooks[hook_remove_helper._hook_id] = hook
return hook_remove_helper
def register_forward_pre_hook(self, hook):
"""Register a forward pre-hook for Layer. The hook will be called before `forward` function has been computed.
"""
Register a forward pre-hook for Layer. The hook will be called before `forward` function has been computed.
It should have the following form, `input` of the `hook` is `input` of the `Layer`,
hook can either return a tuple or a single modified value in the hook. We will wrap the value into a tuple if
......@@ -383,7 +393,7 @@ class Layer:
hook(function): a function registered as a forward pre-hook
Returns:
HookRemoveHelper: a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .
HookRemoveHelper, a HookRemoveHelper object that can be used to remove the added hook by calling `hook_remove_helper.remove()` .
Examples:
.. code-block:: python
......@@ -581,18 +591,20 @@ class Layer:
)
def parameters(self, include_sublayers=True):
"""Returns a list of all Parameters from current layer and its sub-layers.
"""
Returns a list of all Parameters from current layer and its sub-layers.
Returns:
list of Tensor : a list of Parameters.
list of Tensor, a list of Parameters.
Examples:
.. code-block:: python
import paddle
import paddle
linear = paddle.nn.Linear(1,1)
print(linear.parameters()) # print linear_0.w_0 and linear_0.b_0
linear = paddle.nn.Linear(1,1)
print(linear.parameters()) # print linear_0.w_0 and linear_0.b_0
"""
ret = [
......@@ -604,7 +616,9 @@ class Layer:
return ret
def children(self):
"""Returns an iterator over immediate children layers.
"""
Returns an iterator over immediate children layers.
Yields:
Layer: a child layer
......@@ -654,13 +668,15 @@ class Layer:
yield name, layer
def sublayers(self, include_self=False):
"""Returns a list of sub layers.
"""
Returns a list of sub layers.
Parameters:
include_self(bool, optional): Whether return self as sublayers. Default: False
Returns:
list of Layer : a list of sub layers.
list of Layer, a list of sub layers.
Examples:
.. code-block:: python
......@@ -839,13 +855,14 @@ class Layer:
def buffers(self, include_sublayers=True):
"""
Returns a list of all buffers from current layer and its sub-layers.
Parameters:
include_sublayers(bool, optional): Whether include the buffers of sublayers. If True, also include the buffers from sublayers. Default: True
Returns:
list of Tensor : a list of buffers.
list of Tensor, a list of buffers.
Examples:
.. code-block:: python
......@@ -1020,7 +1037,9 @@ class Layer:
raise ValueError("Layer shouldn't implement backward")
def add_sublayer(self, name, sublayer):
"""Adds a sub Layer instance.
"""
Adds a sub Layer instance.
Added sublayer can be accessed by self.name
......@@ -1028,7 +1047,7 @@ class Layer:
name(str): name of this sublayer.
sublayer(Layer): an instance of Layer.
Returns:
Layer: the sublayer passed in.
Layer, the sublayer passed in.
Examples:
.. code-block:: python
......@@ -1055,6 +1074,7 @@ class Layer:
model = MySequential(fc1, fc2)
for prefix, layer in model.named_sublayers():
print(prefix, layer)
"""
assert isinstance(sublayer, Layer) or sublayer is None
......@@ -1070,7 +1090,7 @@ class Layer:
name(str): name of this sublayer.
parameter(Parameter): an instance of Parameter.
Returns:
Parameter: the parameter passed in.
Parameter, the parameter passed in.
Examples:
.. code-block:: python
......@@ -1503,6 +1523,7 @@ class Layer:
use_hook=True,
):
'''
Get all parameters and buffers of current layer and its sub-layers. And set them into a dict
Parameters:
......@@ -1511,7 +1532,7 @@ class Layer:
use_hook(bool, optional) : If true, the operations contained in _state_dict_hooks will be appended to the destination. Default: True
Retruns:
dict: a dict contains all the parameters and persistable buffers.
dict, a dict contains all the parameters and persistable buffers.
Examples:
.. code-block:: python
......
python/paddle/fluid/framework.py
浏览文件 @ 91f4d1ce
......@@ -1352,12 +1352,13 @@ class ParameterMetaClass(VariableMetaClass):
class Variable(metaclass=VariableMetaClass):
"""
**Notes**:
**The constructor of Variable should not be invoked directly.**
**In Static Graph Mode: Please use** `Block.create_var` **to create a Static variable which has no data until being feed.**
Notes:
The constructor of Variable should not be invoked directly.
In Static Graph Mode: Please use ** `Block.create_var` ** to create a Static variable which has no data until being feed.
**In Dygraph Mode: Please use** :ref:`api_fluid_dygraph_to_variable` **to create a dygraph variable with real data**
In Dygraph Mode: Please use ** :ref:`api_fluid_dygraph_to_variable` ** to create a dygraph variable with real data.
In Fluid, every input and output of an OP is a variable. In most
cases, variables are used for holding different kinds of data or training
......@@ -1513,12 +1514,13 @@ class Variable(metaclass=VariableMetaClass):
def detach(self):
"""
Returns a new Variable, detached from the current graph.
It will share data with origin Variable and without tensor copy.
In addition, the detached Variable doesn't provide gradient propagation.
Returns:
( :ref:`api_guide_Variable_en` | dtype is same as current Variable): The detached Variable.
( :ref:`api_guide_Variable_en` | dtype is same as current Variable), The detached Variable.
Examples:
.. code-block:: python
......@@ -1532,6 +1534,7 @@ class Variable(metaclass=VariableMetaClass):
# create a detached Variable
y = x.detach()
"""
assert (
......@@ -2081,6 +2084,7 @@ class Variable(metaclass=VariableMetaClass):
@property
def T(self):
"""
Permute current Variable with its dimensions reversed.
If `n` is the dimensions of `x` , `x.T` is equivalent to `x.transpose([n-1, n-2, ..., 0])`.
......@@ -2099,6 +2103,7 @@ class Variable(metaclass=VariableMetaClass):
x_T_np = exe.run(paddle.static.default_main_program(), fetch_list=[x_T])[0]
print(x_T_np.shape)
# (5, 3, 2)
"""
if len(self.shape) == 1:
return self
......@@ -2137,7 +2142,7 @@ class Variable(metaclass=VariableMetaClass):
as ``out = assign(tensor)`` .
Returns:
Variable: The cloned Variable.
Variable, The cloned Variable.
Examples:
.. code-block:: python
......@@ -2167,6 +2172,7 @@ class Variable(metaclass=VariableMetaClass):
def _set_error_clip(self, error_clip):
"""
Set the error_clip.
Args:
......@@ -2174,11 +2180,13 @@ class Variable(metaclass=VariableMetaClass):
Returns:
None
"""
self.error_clip = error_clip
def _set_info(self, key, value):
"""
Set key-value information for this variable.
Args:
......@@ -2187,6 +2195,7 @@ class Variable(metaclass=VariableMetaClass):
Returns:
None
"""
if not hasattr(self, "_info"):
self._info = {}
......@@ -2194,6 +2203,7 @@ class Variable(metaclass=VariableMetaClass):
def _get_info(self, key):
"""
Get the information of this variable corresponding to key.
Args:
......@@ -2201,6 +2211,7 @@ class Variable(metaclass=VariableMetaClass):
Returns:
object
"""
if hasattr(self, "_info") and key in self._info:
return self._info[key]
......@@ -2208,7 +2219,9 @@ class Variable(metaclass=VariableMetaClass):
def _slice_indices(self, slice, length):
"""
Reference implementation for the slice.indices method.
"""
# Compute step and length as integers.
step = 1 if slice.step is None else slice.step
......@@ -2379,7 +2392,7 @@ class Variable(metaclass=VariableMetaClass):
Default: None
Returns:
Tensor: the value in given scope.
Tensor, the value in given scope.
Examples:
.. code-block:: python
......@@ -2434,6 +2447,7 @@ class Variable(metaclass=VariableMetaClass):
def set_value(self, value, scope=None):
'''
Set the value to the tensor in given scope.
Args:
......@@ -2473,6 +2487,7 @@ class Variable(metaclass=VariableMetaClass):
if var.persistable:
t_load = paddle.load(path+var.name+'.pdtensor')
var.set_value(t_load)
'''
# The 'framework' is a low-level module, and 'executor'
......@@ -2543,10 +2558,11 @@ class Variable(metaclass=VariableMetaClass):
def size(self):
"""
Returns the number of elements for current Variable, which is a int64 Variable with shape [1]
Returns:
Variable: the number of elements for current Variable
Variable, the number of elements for current Variable
Examples:
.. code-block:: python
......@@ -2560,6 +2576,7 @@ class Variable(metaclass=VariableMetaClass):
# get the number of elements of the Variable
y = x.size()
"""
output = self.block.create_var(
......@@ -2574,23 +2591,27 @@ class Variable(metaclass=VariableMetaClass):
def _set_attr(self, name, val):
"""
Set the value of attribute by attribute's name.
Args:
name(str): the attribute name.
val(int|str|list): the value of the attribute.
"""
self._update_desc_attr(name, val)
def _has_attr(self, name):
"""
Whether this Variable has the attribute with the name `name` or not.
Args:
name(str): the attribute name.
Returns:
bool: True if has this attribute.
bool, True if has this attribute.
"""
return self.desc.has_attr(name)
......@@ -2620,7 +2641,7 @@ class Variable(metaclass=VariableMetaClass):
name(str): the attribute name.
Returns:
int|str|list: The attribute value. The return value
int|str|list, The attribute value. The return value
can be any valid attribute type.
"""
return self.desc.attr(name)
......@@ -3193,14 +3214,16 @@ class Operator:
def input(self, name):
r"""
Get the input arguments according to the input parameter name.
Args:
name(str): The input parameter name.
Returns:
list: return the list of argument names that associated with \
list, return the list of argument names that associated with \
the specific parameter name.
"""
return self.desc.input(name)
......
python/paddle/fluid/layers/metric_op.py
浏览文件 @ 91f4d1ce
......@@ -37,22 +37,29 @@ __all__ = ['accuracy', 'auc']
def accuracy(input, label, k=1, correct=None, total=None):
"""
accuracy layer.
Refer to the https://en.wikipedia.org/wiki/Precision_and_recall
This function computes the accuracy using the input and label.
If the correct label occurs in top k predictions, then correct will increment by one.
Note: the dtype of accuracy is determined by input. the input and label dtype can be different.
Note:
the dtype of accuracy is determined by input. the input and label dtype can be different.
Args:
input(Tensor): The input of accuracy layer, which is the predictions of network. A Tensor with type float32,float64.
The shape is ``[sample_number, class_dim]`` .
label(Tensor): The label of dataset. Tensor with type int32,int64. The shape is ``[sample_number, 1]`` .
k(int): The top k predictions for each class will be checked. Data type is int64 or int32.
correct(Tensor): The correct predictions count. A Tensor with type int64 or int32.
total(Tensor): The total entries count. A tensor with type int64 or int32.
k(int, optional): The top k predictions for each class will be checked. Data type is int64 or int32. Default is 1.
correct(Tensor, optional): The correct predictions count. A Tensor with type int64 or int32. Default is None.
total(Tensor, optional): The total entries count. A tensor with type int64 or int32. Default is None.
Returns:
Tensor: The correct rate. A Tensor with type float32.
Tensor, The correct rate. A Tensor with type float32.
Examples:
.. code-block:: python
import numpy as np
import paddle
import paddle.static as static
......@@ -72,6 +79,7 @@ def accuracy(input, label, k=1, correct=None, total=None):
fetch_list=[result[0]])
print(output)
#[array([0.], dtype=float32)]
"""
if _non_static_mode():
if correct is None:
......
python/paddle/fluid/layers/nn.py
浏览文件 @ 91f4d1ce
......@@ -9057,6 +9057,7 @@ def pow(x, factor=1.0, name=None):
@deprecated(since="2.0.0", update_to="paddle.static.nn.prelu")
def prelu(x, mode, param_attr=None, data_format="NCHW", name=None):
r"""
prelu activation.
.. math::
......@@ -9071,26 +9072,20 @@ def prelu(x, mode, param_attr=None, data_format="NCHW", name=None):
element: All elements do not share alpha. Each element has its own alpha.
Parameters:
x (Tensor): The input Tensor or LoDTensor with data type float32.
mode (str): The mode for weight sharing.
param_attr (ParamAttr|None, optional): The parameter attribute for the learnable \
weight (alpha), it can be create by ParamAttr. None by default. \
For detailed information, please refer to :ref:`api_fluid_ParamAttr`.
name (str, optional): Name for the operation (optional, default is None). \
For more information, please refer to :ref:`api_guide_Name`.
param_attr (ParamAttr|None, optional): The parameter attribute for the learnable
weight (alpha), it can be create by ParamAttr. None by default.
For detailed information, please refer to :ref:`api_fluid_ParamAttr`.
data_format(str, optional): Data format that specifies the layout of input.
It may be "NC", "NCL", "NCHW", "NCDHW", "NLC", "NHWC" or "NDHWC". Default: "NCHW".
name (str, optional): Name for the operation (optional, default is None).
For more information, please refer to :ref:`api_guide_Name`.
Returns:
Tensor: A tensor with the same shape and data type as x.
Tensor, A tensor with the same shape and data type as x.
Examples:
.. code-block:: python
import paddle
......
python/paddle/incubate/nn/functional/fused_transformer.py
浏览文件 @ 91f4d1ce
......@@ -284,9 +284,11 @@ def fused_bias_dropout_residual_layer_norm(
name=None,
):
r"""
The fused_bias_dropout_residual_layer_norm operator. The pseudo code is as follows:
.. code-block:: python
y = layer_norm(residual + dropout(bias + x))
Parameters:
......@@ -315,10 +317,9 @@ def fused_bias_dropout_residual_layer_norm(
name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
Returns:
Tensor: The output Tensor, the data type and shape is same as `x`.
Tensor, The output Tensor, the data type and shape is same as `x`.
Examples:
.. code-block:: python
# required: gpu
......@@ -336,6 +337,7 @@ def fused_bias_dropout_residual_layer_norm(
x, residual, bias)
# [2, 4, 128]
print(output.shape)
"""
seed = None
if mode not in ('downscale_in_infer', 'upscale_in_train'):
......
python/paddle/incubate/nn/layer/fused_transformer.py
浏览文件 @ 91f4d1ce
......@@ -705,6 +705,7 @@ class FusedFeedForward(Layer):
class FusedTransformerEncoderLayer(Layer):
"""
FusedTransformerEncoderLayer is composed of two sub-layers which are self (multi-head)
attention and feedforward network. Before and after each sub-layer, pre-process
and post-precess would be applied on the input and output accordingly. If
......@@ -746,7 +747,6 @@ class FusedTransformerEncoderLayer(Layer):
Examples:
.. code-block:: python
# required: gpu
......@@ -759,6 +759,7 @@ class FusedTransformerEncoderLayer(Layer):
attn_mask = paddle.rand((2, 2, 4, 4))
encoder_layer = FusedTransformerEncoderLayer(128, 2, 512)
enc_output = encoder_layer(enc_input, attn_mask) # [2, 4, 128]
"""
def __init__(
......@@ -835,7 +836,9 @@ class FusedTransformerEncoderLayer(Layer):
def forward(self, src, src_mask=None, cache=None):
"""
Applies a Transformer encoder layer on the input.
Parameters:
src (Tensor): The input of Transformer encoder layer. It is
a tensor with shape `[batch_size, sequence_length, d_model]`.
......@@ -851,17 +854,19 @@ class FusedTransformerEncoderLayer(Layer):
`-INF` values and the others have 0 values. It can be None when
nothing wanted or needed to be prevented attention to. Default None.
cache (Tensor, optional): It is an instance of `MultiHeadAttention.Cache`.
See `TransformerEncoderLayer.gen_cache` for more details. It is
See :ref:`api_paddle_nn_TransformerEncoderLayer`.gen_cache for more details. It is
only used for inference and should be None for training. Default
None.
Returns:
Tensor|tuple: It is a tensor that has the same shape and data type \
Tensor|tuple, It is a tensor that has the same shape and data type \
as `enc_input`, representing the output of Transformer encoder \
layer. Or a tuple if `cache` is not None, except for encoder \
layer output, the tuple includes the new cache which is same \
as input `cache` argument but `incremental_cache` has an \
incremental length. See `MultiHeadAttention.gen_cache` and \
`MultiHeadAttention.forward` for more details.
"""
src_mask = _convert_attention_mask(src_mask, src.dtype)
if cache is None:
......
python/paddle/incubate/operators/graph_khop_sampler.py
浏览文件 @ 91f4d1ce
......@@ -28,6 +28,7 @@ def graph_khop_sampler(
name=None,
):
"""
Graph Khop Sampler API.
This API is mainly used in Graph Learning domain, and the main purpose is to
......@@ -50,38 +51,36 @@ def graph_khop_sampler(
sample_sizes (list|tuple): The number of neighbors and number of layers we want
to sample. The data type should be int, and the shape
should only have one dimension.
sorted_eids (Tensor): The sorted edge ids, should not be None when `return_eids`
sorted_eids (Tensor, optional): The sorted edge ids, should not be None when `return_eids`
is True. The shape should be [num_edges, 1], and the data
type should be the same with `row`.
return_eids (bool): Whether to return the id of the sample edges. Default is False.
type should be the same with `row`. Default is None.
return_eids (bool, optional): Whether to return the id of the sample edges. 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:
edge_src (Tensor): The src index of the output edges, also means the first column of
the edges. The shape is [num_sample_edges, 1] currently.
edge_dst (Tensor): The dst index of the output edges, also means the second column
of the edges. The shape is [num_sample_edges, 1] currently.
sample_index (Tensor): The original id of the input nodes and sampled neighbor nodes.
reindex_nodes (Tensor): The reindex id of the input nodes.
edge_eids (Tensor): Return the id of the sample edges if `return_eids` is True.
- edge_src (Tensor), The src index of the output edges, also means the first column of
the edges. The shape is [num_sample_edges, 1] currently.
- edge_dst (Tensor), The dst index of the output edges, also means the second column
of the edges. The shape is [num_sample_edges, 1] currently.
- sample_index (Tensor), The original id of the input nodes and sampled neighbor nodes.
- reindex_nodes (Tensor), The reindex id of the input nodes.
- edge_eids (Tensor), Return the id of the sample edges if `return_eids` is True.
Examples:
.. code-block:: python
import paddle
import paddle
row = [3, 7, 0, 9, 1, 4, 2, 9, 3, 9, 1, 9, 7]
colptr = [0, 2, 4, 5, 6, 7, 9, 11, 11, 13, 13]
nodes = [0, 8, 1, 2]
sample_sizes = [2, 2]
row = paddle.to_tensor(row, dtype="int64")
colptr = paddle.to_tensor(colptr, dtype="int64")
nodes = paddle.to_tensor(nodes, dtype="int64")
row = [3, 7, 0, 9, 1, 4, 2, 9, 3, 9, 1, 9, 7]
colptr = [0, 2, 4, 5, 6, 7, 9, 11, 11, 13, 13]
nodes = [0, 8, 1, 2]
sample_sizes = [2, 2]
row = paddle.to_tensor(row, dtype="int64")
colptr = paddle.to_tensor(colptr, dtype="int64")
nodes = paddle.to_tensor(nodes, dtype="int64")
edge_src, edge_dst, sample_index, reindex_nodes = \
paddle.incubate.graph_khop_sampler(row, colptr, nodes, sample_sizes, False)
edge_src, edge_dst, sample_index, reindex_nodes = paddle.incubate.graph_khop_sampler(row, colptr, nodes, sample_sizes, False)
"""
......
python/paddle/incubate/operators/graph_reindex.py
浏览文件 @ 91f4d1ce
......@@ -35,6 +35,7 @@ def graph_reindex(
name=None,
):
"""
Graph Reindex API.
This API is mainly used in Graph Learning domain, which should be used
......@@ -42,11 +43,11 @@ def graph_reindex(
is to reindex the ids information of the input nodes, and return the
corresponding graph edges after reindex.
**Notes**:
Notes:
The number in x should be unique, otherwise it would cause potential errors.
Besides, we also support multi-edge-types neighbors reindexing. If we have different
edge_type neighbors for x, we should concatenate all the neighbors and count of x.
We will reindex all the nodes from 0.
Besides, we also support multi-edge-types neighbors reindexing. If we have different
edge_type neighbors for x, we should concatenate all the neighbors and count of x.
We will reindex all the nodes from 0.
Take input nodes x = [0, 1, 2] as an example.
If we have neighbors = [8, 9, 0, 4, 7, 6, 7], and count = [2, 3, 2],
......@@ -60,53 +61,52 @@ def graph_reindex(
should be the same with `x`.
count (Tensor): The neighbor count of the input nodes `x`. And the
data type should be int32.
value_buffer (Tensor|None): Value buffer for hashtable. The data type should
be int32, and should be filled with -1.
index_buffer (Tensor|None): Index buffer for hashtable. The data type should
be int32, and should be filled with -1.
flag_buffer_hashtable (bool): Whether to use buffer for hashtable to speed up.
value_buffer (Tensor, optional): Value buffer for hashtable. The data type should
be int32, and should be filled with -1. Default is None.
index_buffer (Tensor, optional): Index buffer for hashtable. The data type should
be int32, and should be filled with -1. Default is None.
flag_buffer_hashtable (bool, optional): Whether to use buffer for hashtable to speed up.
Default is False. Only useful for gpu version currently.
name (str, optional): Name for the operation (optional, default is None).
For more information, please refer to :ref:`api_guide_Name`.
Returns:
reindex_src (Tensor): The source node index of graph edges after reindex.
reindex_dst (Tensor): The destination node index of graph edges after reindex.
out_nodes (Tensor): The index of unique input nodes and neighbors before reindex,
where we put the input nodes `x` in the front, and put neighbor
nodes in the back.
- reindex_src (Tensor), The source node index of graph edges after reindex.
- reindex_dst (Tensor), The destination node index of graph edges after reindex.
- out_nodes (Tensor), The index of unique input nodes and neighbors before reindex,
where we put the input nodes `x` in the front, and put neighbor
nodes in the back.
Examples:
.. code-block:: python
import paddle
x = [0, 1, 2]
neighbors_e1 = [8, 9, 0, 4, 7, 6, 7]
count_e1 = [2, 3, 2]
x = paddle.to_tensor(x, dtype="int64")
neighbors_e1 = paddle.to_tensor(neighbors_e1, dtype="int64")
count_e1 = paddle.to_tensor(count_e1, dtype="int32")
reindex_src, reindex_dst, out_nodes = \
paddle.incubate.graph_reindex(x, neighbors_e1, count_e1)
# reindex_src: [3, 4, 0, 5, 6, 7, 6]
# reindex_dst: [0, 0, 1, 1, 1, 2, 2]
# out_nodes: [0, 1, 2, 8, 9, 4, 7, 6]
neighbors_e2 = [0, 2, 3, 5, 1]
count_e2 = [1, 3, 1]
neighbors_e2 = paddle.to_tensor(neighbors_e2, dtype="int64")
count_e2 = paddle.to_tensor(count_e2, dtype="int32")
neighbors = paddle.concat([neighbors_e1, neighbors_e2])
count = paddle.concat([count_e1, count_e2])
reindex_src, reindex_dst, out_nodes = \
paddle.incubate.graph_reindex(x, neighbors, count)
# reindex_src: [3, 4, 0, 5, 6, 7, 6, 0, 2, 8, 9, 1]
# reindex_dst: [0, 0, 1, 1, 1, 2, 2, 0, 1, 1, 1, 2]
# out_nodes: [0, 1, 2, 8, 9, 4, 7, 6, 3, 5]
import paddle
x = [0, 1, 2]
neighbors_e1 = [8, 9, 0, 4, 7, 6, 7]
count_e1 = [2, 3, 2]
x = paddle.to_tensor(x, dtype="int64")
neighbors_e1 = paddle.to_tensor(neighbors_e1, dtype="int64")
count_e1 = paddle.to_tensor(count_e1, dtype="int32")
reindex_src, reindex_dst, out_nodes = \
paddle.incubate.graph_reindex(x, neighbors_e1, count_e1)
# reindex_src: [3, 4, 0, 5, 6, 7, 6]
# reindex_dst: [0, 0, 1, 1, 1, 2, 2]
# out_nodes: [0, 1, 2, 8, 9, 4, 7, 6]
neighbors_e2 = [0, 2, 3, 5, 1]
count_e2 = [1, 3, 1]
neighbors_e2 = paddle.to_tensor(neighbors_e2, dtype="int64")
count_e2 = paddle.to_tensor(count_e2, dtype="int32")
neighbors = paddle.concat([neighbors_e1, neighbors_e2])
count = paddle.concat([count_e1, count_e2])
reindex_src, reindex_dst, out_nodes = \
paddle.incubate.graph_reindex(x, neighbors, count)
# reindex_src: [3, 4, 0, 5, 6, 7, 6, 0, 2, 8, 9, 1]
# reindex_dst: [0, 0, 1, 1, 1, 2, 2, 0, 1, 1, 1, 2]
# out_nodes: [0, 1, 2, 8, 9, 4, 7, 6, 3, 5]
"""
if flag_buffer_hashtable:
......
python/paddle/incubate/xpu/resnet_block.py
浏览文件 @ 91f4d1ce
......@@ -325,6 +325,7 @@ def resnet_basic_block(
class ResNetBasicBlock(Layer):
r"""
ResNetBasicBlock is designed for optimize the performence of the basic unit of ssd resnet block.
If has_shortcut = True, it can calculate 3 Conv2D, 3 BatchNorm and 2 ReLU in one time.
If has_shortcut = False, it can calculate 2 Conv2D, 2 BatchNorm and 2 ReLU in one time. In this
......@@ -362,14 +363,14 @@ class ResNetBasicBlock(Layer):
and variance are also used during train period. Default: False.
is_test (bool, optional): A flag indicating whether it is in
test phrase or not. Default: False.
filter_attr (ParamAttr|None): The parameter attribute for learnable parameters/weights
filter_attr (ParamAttr, optional): The parameter attribute for learnable parameters/weights
of conv2d. If it is set to None or one attribute of ParamAttr, conv2d
will create ParamAttr as param_attr. Default: None.
scale_attr (ParamAttr|None): The parameter attribute for Parameter `scale`
scale_attr (ParamAttr, optional): The parameter attribute for Parameter `scale`
of batch_norm. If it is set to None or one attribute of ParamAttr, batch_norm will create ParamAttr
as param_attr, the name of scale can be set in ParamAttr. If the Initializer of the param_attr is not set,
the parameter is initialized with Xavier. Default: None.
bias_attr (ParamAttr|None): The parameter attribute for the bias of batch_norm.
bias_attr (ParamAttr, optional): The parameter attribute for the bias of batch_norm.
If it is set to None or one attribute of ParamAttr, batch_norm
will create ParamAttr as bias_attr, the name of bias can be set in ParamAttr.
If the Initializer of the bias_attr is not set, the bias is initialized zero.
......@@ -396,7 +397,6 @@ class ResNetBasicBlock(Layer):
Examples:
.. code-block:: python
# required: xpu
......@@ -426,6 +426,7 @@ class ResNetBasicBlock(Layer):
out = resnet_basic_block.forward(x)
print(out.shape) # [2, 8, 16, 16]
"""
def __init__(
......
python/paddle/signal.py
浏览文件 @ 91f4d1ce
......@@ -259,6 +259,7 @@ def stft(
name=None,
):
r"""
Short-time Fourier transform (STFT).
The STFT computes the discrete Fourier transforms (DFT) of short overlapping
......@@ -271,9 +272,12 @@ def stft(
Where:
- :math:`t`: The :math:`t`-th input window.
- :math:`\omega`: Frequency :math:`0 \leq \omega < \text{n\_fft}` for `onesided=False`,
or :math:`0 \leq \omega < \lfloor \text{n\_fft} / 2 \rfloor + 1` for `onesided=True`.
or :math:`0 \leq \omega < \lfloor \text{n\_fft} / 2 \rfloor + 1` for `onesided=True`.
- :math:`N`: Value of `n_fft`.
- :math:`H`: Value of `hop_length`.
Args:
......@@ -300,9 +304,9 @@ def stft(
to set this property. For more information, please refer to :ref:`api_guide_Name`.
Returns:
The complex STFT output tensor with shape `[..., n_fft//2 + 1, num_frames]`(
real-valued input and `onesided` is `True`) or `[..., n_fft, num_frames]`(
`onesided` is `False`)
The complex STFT output tensor with shape `[..., n_fft//2 + 1, num_frames]`
(real-valued input and `onesided` is `True`) or `[..., n_fft, num_frames]`
(`onesided` is `False`)
Examples:
.. code-block:: python
......@@ -319,6 +323,7 @@ def stft(
x = paddle.randn([8, 48000], dtype=paddle.float64) + \
paddle.randn([8, 48000], dtype=paddle.float64)*1j # [8, 48000] complex128
y1 = stft(x, n_fft=512, center=False, onesided=False) # [8, 512, 372]
"""
check_variable_and_dtype(
x, 'x', ['float32', 'float64', 'complex64', 'complex128'], 'stft'
......
python/paddle/sparse/nn/layer/activation.py
浏览文件 @ 91f4d1ce
......@@ -20,6 +20,7 @@ __all__ = []
class ReLU(Layer):
"""
Sparse ReLU Activation, requiring x to be a SparseCooTensor or SparseCsrTensor.
.. math::
......@@ -44,6 +45,7 @@ class ReLU(Layer):
relu = paddle.sparse.nn.ReLU()
out = relu(sparse_x)
# [0., 0., 1.]
"""
def __init__(self, name=None):
......@@ -60,6 +62,7 @@ class ReLU(Layer):
class Softmax(Layer):
r"""
Sparse Softmax Activation, requiring x to be a SparseCooTensor or SparseCsrTensor.
Note:
......@@ -129,6 +132,7 @@ class Softmax(Layer):
class ReLU6(Layer):
"""
Sparse ReLU6 Activation, requiring x to be a SparseCooTensor or SparseCsrTensor.
.. math::
......@@ -152,6 +156,7 @@ class ReLU6(Layer):
sparse_x = dense_x.to_sparse_coo(1)
relu6 = paddle.sparse.nn.ReLU6()
out = relu6(sparse_x)
"""
def __init__(self, name=None):
......@@ -168,6 +173,7 @@ class ReLU6(Layer):
class LeakyReLU(Layer):
r"""
Sparse Leaky ReLU Activation, requiring x to be a SparseCooTensor or SparseCsrTensor.
.. math::
......@@ -199,6 +205,7 @@ class LeakyReLU(Layer):
sparse_x = dense_x.to_sparse_coo(1)
leaky_relu = paddle.sparse.nn.LeakyReLU(0.5)
out = leaky_relu(sparse_x)
"""
def __init__(self, negative_slope=0.01, name=None):
......
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或