Skip to content

P
Paddle

TonyTonyFun / 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 86ca31ab 编写于 作者: C Cindy Cai 提交者： GitHub
浏览文件
操作

English API Docs Optimization Part 1 (#24536) 

* test=develop, test=document_fix

* test=develop, test=document_fix
Co-authored-by: Nswtkiwi <1208425345@qq.com>
上级 2d0f849e
隐藏空白更改
内联 并排
Showing with 622 addition and 1 deletion
python/paddle/fluid/backward.py
浏览文件 @ 86ca31ab
......@@ -1196,6 +1196,8 @@ def append_backward(loss,
callbacks=None,
checkpoints=None):
"""
:api_attr: Static Graph
This function appends backward part to main_program.
A complete neural network training is made up of forward and backward
......@@ -1724,6 +1726,8 @@ def calc_gradient(targets, inputs, target_gradients=None, no_grad_set=None):
def gradients(targets, inputs, target_gradients=None, no_grad_set=None):
"""
:api_attr: Static Graph
Backpropagate the gradients of targets to inputs.
Args:
......
python/paddle/fluid/clip.py
浏览文件 @ 86ca31ab
......@@ -158,6 +158,10 @@ class GradientClipBase(object):
class GradientClipByValue(GradientClipBase):
"""
:alias_main: paddle.nn.GradientClipByValue
:alias: paddle.nn.GradientClipByValue,paddle.nn.clip.GradientClipByValue
:old_api: paddle.fluid.clip.GradientClipByValue
Limit the value of multi-dimensional Tensor :math:`X` to the range [min, max].
- Any values less than min are set to ``min``.
......@@ -296,6 +300,10 @@ class GradientClipByValue(GradientClipBase):
class GradientClipByNorm(GradientClipBase):
"""
:alias_main: paddle.nn.GradientClipByNorm
:alias: paddle.nn.GradientClipByNorm,paddle.nn.clip.GradientClipByNorm
:old_api: paddle.fluid.clip.GradientClipByNorm
Limit the l2 norm of multi-dimensional Tensor :math:`X` to ``clip_norm`` .
- If the l2 norm of :math:`X` is greater than ``clip_norm`` , :math:`X` will be compressed by a ratio.
......@@ -447,6 +455,10 @@ class GradientClipByNorm(GradientClipBase):
class GradientClipByGlobalNorm(GradientClipBase):
"""
:alias_main: paddle.nn.GradientClipByGlobalNorm
:alias: paddle.nn.GradientClipByGlobalNorm,paddle.nn.clip.GradientClipByGlobalNorm
:old_api: paddle.fluid.clip.GradientClipByGlobalNorm
Given a list of Tensor :math:`t\_list` , calculate the global norm for the elements of all tensors in
:math:`t\_list` , and limit it to ``clip_norm`` .
......@@ -691,6 +703,8 @@ class GradientClipByGlobalNorm(GradientClipBase):
@framework.dygraph_not_support
def set_gradient_clip(clip, param_list=None, program=None):
"""
:api_attr: Static Graph
Warning:
This API must be used after building network, and before ``minimize`` ,
......
python/paddle/fluid/compiler.py
浏览文件 @ 86ca31ab
......@@ -86,6 +86,8 @@ def _has_optimizer_in_control_flow(program):
class CompiledProgram(object):
"""
:api_attr: Static Graph
The CompiledProgram is used to transform a program or graph for
various optimizations according to the configuration of build_strategy,
for example, the operators' fusion in the computation graph, memory
......
python/paddle/fluid/data.py
浏览文件 @ 86ca31ab
......@@ -24,6 +24,11 @@ __all__ = ['data']
def data(name, shape, dtype='float32', lod_level=0):
"""
:api_attr: Static Graph
:alias_main: paddle.nn.data
:alias: paddle.nn.data,paddle.nn.input.data
:old_api: paddle.fluid.data
**Data Layer**
This function creates a variable on the global block. The global variable
......
python/paddle/fluid/data_feed_desc.py
浏览文件 @ 86ca31ab
......@@ -20,6 +20,8 @@ __all__ = ['DataFeedDesc']
class DataFeedDesc(object):
"""
:api_attr: Static Graph
Datafeed descriptor, describing input training data format. This class is
currently only used for AsyncExecutor (See comments for class AsyncExecutor
for a brief introduction)
......
python/paddle/fluid/data_feeder.py
浏览文件 @ 86ca31ab
......@@ -211,6 +211,8 @@ class BatchedTensorProvider(object):
class DataFeeder(object):
"""
:api_attr: Static Graph
DataFeeder converts the data that returned by a reader into a data
structure that can feed into Executor. The reader is usually a
python generator that returns a list of mini-batch data entries.
......
python/paddle/fluid/dataset.py
浏览文件 @ 86ca31ab
......@@ -728,6 +728,8 @@ class InMemoryDataset(DatasetBase):
def release_memory(self):
"""
:api_attr: Static Graph
Release InMemoryDataset memory data, when data will not be used again.
Examples:
......
python/paddle/fluid/dygraph/base.py
浏览文件 @ 86ca31ab
......@@ -111,6 +111,10 @@ def enabled():
def enable_dygraph(place=None):
"""
:alias_main: paddle.enable_dygraph
:alias: paddle.enable_dygraph,paddle.enable_imperative.enable_dygraph
:old_api: paddle.fluid.dygraph.base.enable_dygraph
This function enables dynamic graph mode.
Parameters:
......@@ -141,6 +145,10 @@ def enable_dygraph(place=None):
def disable_dygraph():
"""
:alias_main: paddle.disable_dygraph
:alias: paddle.disable_dygraph,paddle.disable_imperative.disable_dygraph
:old_api: paddle.fluid.dygraph.base.disable_dygraph
This function disables dynamic graph mode.
return:
......@@ -178,6 +186,8 @@ def _switch_tracer_mode_guard_(is_train=True):
def no_grad(func=None):
"""
:api_attr: imperative
Create a context which disables dygraph gradient calculation.
In this mode, the result of every computation will have `stop_gradient=True`.
......@@ -236,6 +246,8 @@ def no_grad(func=None):
@signature_safe_contextmanager
def guard(place=None):
"""
:api_attr: imperative
This context will create a dygraph context for dygraph to run, using python ``with`` statement.
Parameters:
......@@ -520,6 +532,8 @@ def grad(outputs,
@framework.dygraph_only
def to_variable(value, name=None, zero_copy=None):
"""
:api_attr: imperative
The API will create a ``Variable`` or ``ComplexVariable`` object from
numpy\.ndarray, Variable or ComplexVariable object.
......
python/paddle/fluid/dygraph/checkpoint.py
浏览文件 @ 86ca31ab
......@@ -32,6 +32,8 @@ __all__ = [
@dygraph_only
def save_dygraph(state_dict, model_path):
'''
:api_attr: imperative
Save Layer's state_dict to disk. This will generate a file with suffix ".pdparams"
The state_dict is get from Layers.state_dict function
......@@ -95,6 +97,8 @@ def save_dygraph(state_dict, model_path):
@dygraph_only
def load_dygraph(model_path, keep_name_table=False):
'''
:api_attr: imperative
Load parameter state_dict from disk.
Args:
......
python/paddle/fluid/dygraph/jit.py
浏览文件 @ 86ca31ab
......@@ -203,6 +203,8 @@ def _trace(layer,
class TracedLayer(object):
"""
:api_attr: imperative
TracedLayer is used to convert a forward dygraph model to a static
graph model. This is mainly used to save the dygraph model for online
inference using C++. Besides, users can also do inference in Python
......
python/paddle/fluid/dygraph/layers.py
浏览文件 @ 86ca31ab
......@@ -58,7 +58,12 @@ class HookRemoveHelper(object):
class Layer(core.Layer):
"""Dynamic graph Layer based on OOD, includes the parameters of the layer, the structure of the forward graph and so on.
"""
:alias_main: paddle.nn.Layer
:alias: paddle.nn.Layer
:old_api: paddle.fluid.dygraph.layers.Layer
Dynamic graph Layer based on OOD, includes the parameters of the layer, the structure of the forward graph and so on.
Parameters:
name_scope (str, optional): prefix name used by the layer to name parameters.
......
python/paddle/fluid/dygraph/learning_rate_scheduler.py
浏览文件 @ 86ca31ab
......@@ -69,6 +69,8 @@ class LearningRateDecay(object):
class PiecewiseDecay(LearningRateDecay):
"""
:api_attr: imperative
Piecewise decay scheduler.
The algorithm can be described as the code below.
......@@ -128,6 +130,8 @@ class PiecewiseDecay(LearningRateDecay):
class NaturalExpDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies natural exponential decay to the initial learning rate.
The algorithm can be described as following.
......@@ -207,6 +211,8 @@ class NaturalExpDecay(LearningRateDecay):
class ExponentialDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies exponential decay to the learning rate.
The algorithm can be described as following.
......@@ -287,6 +293,8 @@ class ExponentialDecay(LearningRateDecay):
class InverseTimeDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies inverse time decay to the initial learning rate.
The algorithm can be described as following.
......@@ -363,6 +371,8 @@ class InverseTimeDecay(LearningRateDecay):
class PolynomialDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies polynomial decay to the initial learning rate.
The algorithm can be described as following.
......@@ -455,6 +465,8 @@ class PolynomialDecay(LearningRateDecay):
class CosineDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies cosine decay to the learning rate.
The algorithm can be described as following.
......@@ -511,6 +523,8 @@ class CosineDecay(LearningRateDecay):
class NoamDecay(LearningRateDecay):
"""
:api_attr: imperative
Applies Noam decay to the initial learning rate.
The algorithm can be described as following.
......
python/paddle/fluid/dygraph/nn.py
浏览文件 @ 86ca31ab
......@@ -696,6 +696,10 @@ class Conv3DTranspose(layers.Layer):
class Pool2D(layers.Layer):
"""
:alias_main: paddle.nn.Pool2D
:alias: paddle.nn.Pool2D,paddle.nn.layer.Pool2D,paddle.nn.layer.common.Pool2D
:old_api: paddle.fluid.dygraph.Pool2D
This interface is used to construct a callable object of the ``Pool2D`` class.
For more details, refer to code examples.
The pooling2d operation calculates the output based on the input, pool_type and pool_size, pool_stride,
......@@ -867,6 +871,10 @@ class Pool2D(layers.Layer):
class Linear(layers.Layer):
"""
:alias_main: paddle.nn.Linear
:alias: paddle.nn.Linear,paddle.nn.layer.Linear,paddle.nn.layer.common.Linear
:old_api: paddle.fluid.dygraph.Linear
Fully-connected linear transformation layer:
.. math::
......@@ -1100,6 +1108,10 @@ class InstanceNorm(layers.Layer):
class BatchNorm(layers.Layer):
"""
:alias_main: paddle.nn.BatchNorm
:alias: paddle.nn.BatchNorm,paddle.nn.layer.BatchNorm,paddle.nn.layer.norm.BatchNorm
:old_api: paddle.fluid.dygraph.BatchNorm
This interface is used to construct a callable object of the ``BatchNorm`` class.
For more details, refer to code examples.
It implements the function of the Batch Normalization Layer and can be used
......@@ -1443,6 +1455,10 @@ class Dropout(layers.Layer):
class Embedding(layers.Layer):
"""
:alias_main: paddle.nn.Embedding
:alias: paddle.nn.Embedding,paddle.nn.layer.Embedding,paddle.nn.layer.common.Embedding
:old_api: paddle.fluid.dygraph.Embedding
**Embedding Layer**
This interface is used to construct a callable object of the ``Embedding`` class.
......@@ -1599,6 +1615,10 @@ class Embedding(layers.Layer):
class LayerNorm(layers.Layer):
"""
:alias_main: paddle.nn.LayerNorm
:alias: paddle.nn.LayerNorm,paddle.nn.layer.LayerNorm,paddle.nn.layer.norm.LayerNorm
:old_api: paddle.fluid.dygraph.LayerNorm
This interface is used to construct a callable object of the ``LayerNorm`` class.
For more details, refer to code examples.
It implements the function of the Layer Normalization Layer and can be applied to mini-batch input data.
......@@ -2289,6 +2309,10 @@ class PRelu(layers.Layer):
class BilinearTensorProduct(layers.Layer):
"""
:alias_main: paddle.nn.BilinearTensorProduct
:alias: paddle.nn.BilinearTensorProduct,paddle.nn.layer.BilinearTensorProduct,paddle.nn.layer.common.BilinearTensorProduct
:old_api: paddle.fluid.dygraph.BilinearTensorProduct
**Add Bilinear Tensor Product Layer**
This layer performs bilinear tensor product on two inputs.
......@@ -2809,6 +2833,10 @@ class RowConv(layers.Layer):
class GroupNorm(layers.Layer):
"""
:alias_main: paddle.nn.GroupNorm
:alias: paddle.nn.GroupNorm,paddle.nn.layer.GroupNorm,paddle.nn.layer.norm.GroupNorm
:old_api: paddle.fluid.dygraph.GroupNorm
This interface is used to construct a callable object of the ``GroupNorm`` class.
For more details, refer to code examples.
It implements the function of the Group Normalization Layer.
......@@ -2909,6 +2937,10 @@ class GroupNorm(layers.Layer):
class SpectralNorm(layers.Layer):
"""
:alias_main: paddle.nn.SpectralNorm
:alias: paddle.nn.SpectralNorm,paddle.nn.layer.SpectralNorm,paddle.nn.layer.norm.SpectralNorm
:old_api: paddle.fluid.dygraph.SpectralNorm
This interface is used to construct a callable object of the ``SpectralNorm`` class.
For more details, refer to code examples. It implements the function of the Spectral Normalization Layer.
This layer calculates the spectral normalization value of weight parameters of
......
python/paddle/fluid/dygraph/parallel.py
浏览文件 @ 86ca31ab
......@@ -28,6 +28,9 @@ ParallelStrategy = core.ParallelStrategy
def prepare_context(strategy=None):
'''
:api_attr: imperative
'''
if strategy is None:
strategy = ParallelStrategy()
strategy.nranks = Env().nranks
......
python/paddle/fluid/dygraph/tracer.py
浏览文件 @ 86ca31ab
......@@ -23,6 +23,8 @@ from paddle.fluid import framework
class Tracer(core.Tracer):
"""
:api_attr: imperative
Tracer is used to execute and record the operators executed, to construct the
computation graph in dygraph model. Tracer has two mode, :code:`train_mode`
and :code:`eval_mode`. In :code:`train_mode`, Tracer would add backward network
......
python/paddle/fluid/executor.py
浏览文件 @ 86ca31ab
......@@ -40,6 +40,8 @@ InferAnalysisConfig = core.AnalysisConfig
def global_scope():
"""
:api_attr: Static Graph
Get the global/default scope instance. There are a lot of APIs use
:code:`global_scope` as its default value, e.g., :code:`Executor.run`
......@@ -68,6 +70,8 @@ def _switch_scope(scope):
@signature_safe_contextmanager
def scope_guard(scope):
"""
:api_attr: Static Graph
This function switches scope through python `with` statement.
Scope records the mapping between variable names and variables ( :ref:`api_guide_Variable` ),
similar to brackets in programming languages.
......@@ -456,6 +460,8 @@ handler = FetchHandlerExample(var_dict=var_dict)
class Executor(object):
"""
:api_attr: Static Graph
An Executor in Python, supports single/multiple-GPU running,
and single/multiple-CPU running.
......
python/paddle/fluid/framework.py
浏览文件 @ 86ca31ab
......@@ -179,6 +179,10 @@ def require_version(min_version, max_version=None):
def in_dygraph_mode():
"""
:alias_main: paddle.in_dygraph_mode
:alias: paddle.in_dygraph_mode
:old_api: paddle.fluid.framework.in_dygraph_mode
This function checks whether the program runs in dynamic graph mode or not.
You can enter dynamic graph mode with :ref:`api_fluid_dygraph_guard` api,
or enable and disable dynamic graph mode with :ref:`api_fluid_dygraph_enable`
......@@ -436,6 +440,8 @@ _name_scope = NameScope()
@signature_safe_contextmanager
def name_scope(prefix=None):
"""
:api_attr: Static Graph
Generate hierarchical name prefix for the operators.
Note:
......@@ -5277,6 +5283,8 @@ def switch_startup_program(program):
@signature_safe_contextmanager
def program_guard(main_program, startup_program=None):
"""
:api_attr: Static Graph
Change the global main program and startup program with `"with"` statement.
Layer functions in the Python `"with"` block will append operators and
variables to the new main programs.
......@@ -5376,6 +5384,8 @@ def _dygraph_place_guard(place):
def load_op_library(lib_filename):
"""
:api_attr: Static Graph
Load a dynamic library, including custom operators and kernels.
When library is loaded, ops and kernels registered in the library
will be available in PaddlePaddle main process.
......
python/paddle/fluid/input.py
浏览文件 @ 86ca31ab
......@@ -23,6 +23,9 @@ __all__ = ['one_hot', 'embedding']
def one_hot(input, depth, allow_out_of_range=False):
"""
:alias_main: paddle.nn.functional.one_hot
:alias: paddle.nn.functional.one_hot,paddle.nn.functional.common.one_hot
:old_api: paddle.fluid.one_hot
The operator converts each id in the input to an one-hot vector with a
depth length. The value in the vector dimension corresponding to the id
......@@ -132,6 +135,7 @@ def embedding(input,
param_attr=None,
dtype='float32'):
"""
:api_attr: Static Graph
The operator is used to lookup embeddings vector of ids provided by :attr:`input` .
It automatically constructs a 2D embedding matrix based on the
......
python/paddle/fluid/io.py
浏览文件 @ 86ca31ab
......@@ -124,6 +124,8 @@ def is_belong_to_optimizer(var):
@dygraph_not_support
def get_program_parameter(program):
"""
:api_attr: Static Graph
Get all the parameters from Program.
Args:
......@@ -147,6 +149,8 @@ def get_program_parameter(program):
@dygraph_not_support
def get_program_persistable_vars(program):
"""
:api_attr: Static Graph
Get all the persistable vars from Program.
Args:
......@@ -223,6 +227,8 @@ def save_vars(executor,
predicate=None,
filename=None):
"""
:api_attr: Static Graph
This API saves specific variables in the `Program` to files.
There are two ways to specify the variables to be saved: set variables in
......@@ -365,6 +371,8 @@ def save_vars(executor,
@dygraph_not_support
def save_params(executor, dirname, main_program=None, filename=None):
"""
:api_attr: Static Graph
This operator saves all parameters from the :code:`main_program` to
the folder :code:`dirname` or file :code:`filename`. You can refer to
:ref:`api_guide_model_save_reader_en` for more details.
......@@ -588,6 +596,8 @@ def _save_distributed_persistables(executor, dirname, main_program):
@dygraph_not_support
def save_persistables(executor, dirname, main_program=None, filename=None):
"""
:api_attr: Static Graph
This operator saves all persistable variables from :code:`main_program` to
the folder :code:`dirname` or file :code:`filename`. You can refer to
:ref:`api_guide_model_save_reader_en` for more details. And then
......@@ -661,6 +671,8 @@ def load_vars(executor,
predicate=None,
filename=None):
"""
:api_attr: Static Graph
This API loads variables from files by executor.
There are two ways to specify the variables to be loaded: the first way, set
......@@ -829,6 +841,8 @@ def load_vars(executor,
@dygraph_not_support
def load_params(executor, dirname, main_program=None, filename=None):
"""
:api_attr: Static Graph
This API filters out all parameters from the give ``main_program``
and then tries to load these parameters from the directory ``dirname`` or
the file ``filename``.
......@@ -887,6 +901,8 @@ def load_params(executor, dirname, main_program=None, filename=None):
@dygraph_not_support
def load_persistables(executor, dirname, main_program=None, filename=None):
"""
:api_attr: Static Graph
This API filters out all variables with ``persistable==True`` from the
given ``main_program`` and then tries to load these variables from the
directory ``dirname`` or the file ``filename``.
......@@ -1084,6 +1100,8 @@ def save_inference_model(dirname,
export_for_deployment=True,
program_only=False):
"""
:api_attr: Static Graph
Prune the given `main_program` to build a new program especially for inference,
and then save it and all related parameters to given `dirname` .
If you just want to save parameters of your trained model, please use the
......@@ -1288,6 +1306,8 @@ def load_inference_model(dirname,
params_filename=None,
pserver_endpoints=None):
"""
:api_attr: Static Graph
Load the inference model from a given directory. By this API, you can get the model
structure(Inference Program) and model parameters. If you just want to load
parameters of the pre-trained model, please use the :ref:`api_fluid_io_load_params` API.
......@@ -1577,6 +1597,11 @@ def _load_persistable_nodes(executor, dirname, graph):
@dygraph_not_support
def save(program, model_path):
"""
:api_attr: Static Graph
:alias_main: paddle.save
:alias: paddle.save,paddle.tensor.save,paddle.tensor.io.save
:old_api: paddle.fluid.save
This function save parameters, optimizer information and network description to model_path.
The parameters contains all the trainable Variable, will save to a file with suffix ".pdparams".
......@@ -1636,6 +1661,11 @@ def save(program, model_path):
@dygraph_not_support
def load(program, model_path, executor=None, var_list=None):
"""
:api_attr: Static Graph
:alias_main: paddle.load
:alias: paddle.load,paddle.tensor.load,paddle.tensor.io.load
:old_api: paddle.fluid.io.load
This function get parameters and optimizer information from program, and then get corresponding value from file.
An exception will throw if shape or dtype of the parameters is not match.
......@@ -1803,6 +1833,8 @@ def load(program, model_path, executor=None, var_list=None):
@dygraph_not_support
def load_program_state(model_path, var_list=None):
"""
:api_attr: Static Graph
Load program state from local file
Args:
......@@ -1934,6 +1966,8 @@ def load_program_state(model_path, var_list=None):
@dygraph_not_support
def set_program_state(program, state_dict):
"""
:api_attr: Static Graph
Set program parameter from state_dict
An exception will throw if shape or dtype of the parameters is not match.
......
python/paddle/fluid/layers/control_flow.py
浏览文件 @ 86ca31ab
......@@ -222,6 +222,8 @@ def Print(input,
print_tensor_lod=True,
print_phase='both'):
'''
:api_attr: Static Graph
**Print operator**
This creates a print op that will print when a tensor is accessed.
......@@ -446,6 +448,8 @@ class StaticRNNMemoryLink(object):
class StaticRNN(object):
"""
:api_attr: Static Graph
StaticRNN class.
The StaticRNN can process a batch of sequence data. The first dimension of inputs
......@@ -923,6 +927,8 @@ class WhileGuard(BlockGuard):
class While(object):
"""
:api_attr: Static Graph
while loop control flow. Repeat while body until cond is False.
Note:
......@@ -1061,6 +1067,11 @@ def assign_skip_lod_tensor_array(inputs, outputs):
def while_loop(cond, body, loop_vars, is_test=False, name=None):
"""
:api_attr: Static Graph
:alias_main: paddle.nn.while_loop
:alias: paddle.nn.while_loop,paddle.nn.control_flow.while_loop
:old_api: paddle.fluid.layers.while_loop
while_loop is one of the control flows. Repeats while_loop `body` until `cond` returns False.
Notice:
......@@ -1529,6 +1540,10 @@ def create_array(dtype):
@templatedoc()
def less_than(x, y, force_cpu=None, cond=None):
"""
:alias_main: paddle.less_than
:alias: paddle.less_than,paddle.tensor.less_than,paddle.tensor.logic.less_than
:old_api: paddle.fluid.layers.less_than
${comment}
Args:
......@@ -1594,6 +1609,10 @@ def less_than(x, y, force_cpu=None, cond=None):
@templatedoc()
def less_equal(x, y, cond=None):
"""
:alias_main: paddle.less_equal
:alias: paddle.less_equal,paddle.tensor.less_equal,paddle.tensor.logic.less_equal
:old_api: paddle.fluid.layers.less_equal
This OP returns the truth value of :math:`x <= y` elementwise, which is equivalent function to the overloaded operator `<=`.
Args:
......@@ -1642,6 +1661,10 @@ def less_equal(x, y, cond=None):
@templatedoc()
def greater_than(x, y, cond=None):
"""
:alias_main: paddle.greater_than
:alias: paddle.greater_than,paddle.tensor.greater_than,paddle.tensor.logic.greater_than
:old_api: paddle.fluid.layers.greater_than
This OP returns the truth value of :math:`x > y` elementwise, which is equivalent function to the overloaded operator `>`.
Args:
......@@ -1689,6 +1712,10 @@ def greater_than(x, y, cond=None):
@templatedoc()
def greater_equal(x, y, cond=None):
"""
:alias_main: paddle.greater_equal
:alias: paddle.greater_equal,paddle.tensor.greater_equal,paddle.tensor.logic.greater_equal
:old_api: paddle.fluid.layers.greater_equal
This OP returns the truth value of :math:`x >= y` elementwise, which is equivalent function to the overloaded operator `>=`.
Args:
......@@ -1782,6 +1809,10 @@ def equal(x, y, cond=None):
def not_equal(x, y, cond=None):
"""
:alias_main: paddle.not_equal
:alias: paddle.not_equal,paddle.tensor.not_equal,paddle.tensor.logic.not_equal
:old_api: paddle.fluid.layers.not_equal
This OP returns the truth value of :math:`x != y` elementwise, which is equivalent function to the overloaded operator `!=`.
Args:
......@@ -2225,6 +2256,11 @@ def copy_var_to_parent_block(var, layer_helper):
def cond(pred, true_fn=None, false_fn=None, name=None):
"""
:api_attr: Static Graph
:alias_main: paddle.nn.cond
:alias: paddle.nn.cond,paddle.nn.control_flow.cond
:old_api: paddle.fluid.layers.cond
This API returns ``true_fn()`` if the predicate ``pred`` is true else
``false_fn()`` . Users could also set ``true_fn`` or ``false_fn`` to
``None`` if do nothing and this API will treat the callable simply returns
......@@ -2410,6 +2446,11 @@ def _error_message(what, arg_name, op_name, right_value, error_value):
def case(pred_fn_pairs, default=None, name=None):
'''
:api_attr: Static Graph
:alias_main: paddle.nn.case
:alias: paddle.nn.case,paddle.nn.control_flow.case
:old_api: paddle.fluid.layers.case
This operator works like an if-elif-elif-else chain.
Args:
......@@ -2520,6 +2561,7 @@ def case(pred_fn_pairs, default=None, name=None):
class Switch(object):
"""
:api_attr: Static Graph
This class is used to implement Switch branch control function.
Switch branch contains several case branches and one default branch.
......@@ -2677,6 +2719,8 @@ class IfElseBlockGuard(object):
class IfElse(object):
"""
:api_attr: Static Graph
This class is used to implement IfElse branch control function. IfElse contains two blocks, true_block and false_block. IfElse will put data satisfying True or False conditions into different blocks to run.
Cond is a 2-D Tensor with shape [N, 1] and data type bool, representing the execution conditions of the corresponding part of the input data.
......@@ -2853,6 +2897,8 @@ class IfElse(object):
class DynamicRNN(object):
"""
:api_attr: Static Graph
**Note: the input of this class should be LoDTensor which holds the
information of variable-length sequences. If the input is fixed-length Tensor,
please use StaticRNN (fluid.layers.** :ref:`api_fluid_layers_StaticRNN` **) for
......@@ -3518,6 +3564,8 @@ class DynamicRNN(object):
def switch_case(branch_index, branch_fns, default=None, name=None):
'''
:api_attr: Static Graph
This operator is like a C++ switch/case statement.
Args:
......@@ -3701,6 +3749,10 @@ def reorder_lod_tensor_by_rank(x, rank_table):
def is_empty(x, cond=None):
"""
:alias_main: paddle.is_empty
:alias: paddle.is_empty,paddle.tensor.is_empty,paddle.tensor.logic.is_empty
:old_api: paddle.fluid.layers.is_empty
Test whether a Variable is empty.
Args:
......
python/paddle/fluid/layers/loss.py
浏览文件 @ 86ca31ab
......@@ -57,6 +57,11 @@ def center_loss(input,
param_attr,
update_center=True):
"""
:api_attr: Static Graph
:alias_main: paddle.nn.functional.center_loss
:alias: paddle.nn.functional.center_loss,paddle.nn.functional.loss.center_loss
:old_api: paddle.fluid.layers.center_loss
**Center loss Cost layer**
This OP accepts input (deep features,the output of the last hidden layer)
......@@ -147,6 +152,10 @@ def center_loss(input,
def bpr_loss(input, label, name=None):
"""
:alias_main: paddle.nn.functional.bpr_loss
:alias: paddle.nn.functional.bpr_loss,paddle.nn.functional.loss.bpr_loss
:old_api: paddle.fluid.layers.bpr_loss
**Bayesian Personalized Ranking Loss Operator**
This operator belongs to pairwise ranking loss. Label is the desired item.
......@@ -195,6 +204,10 @@ def bpr_loss(input, label, name=None):
def cross_entropy(input, label, soft_label=False, ignore_index=kIgnoreIndex):
"""
:alias_main: paddle.nn.functional.cross_entropy
:alias: paddle.nn.functional.cross_entropy,paddle.nn.functional.loss.cross_entropy
:old_api: paddle.fluid.layers.cross_entropy
This operator computes the cross entropy between input and label. It
supports both hard-label and and soft-label cross entropy computation.
......@@ -288,6 +301,10 @@ def cross_entropy2(input, label, ignore_index=kIgnoreIndex):
def square_error_cost(input, label):
"""
:alias_main: paddle.nn.functional.square_error_cost
:alias: paddle.nn.functional.square_error_cost,paddle.nn.functional.loss.square_error_cost
:old_api: paddle.fluid.layers.square_error_cost
This op accepts input predictions and target label and returns the
squared error cost.
......@@ -663,6 +680,8 @@ def nce(input,
seed=0,
is_sparse=False):
"""
:api_attr: Static Graph
${comment}
Args:
......@@ -874,6 +893,8 @@ def hsigmoid(input,
is_custom=False,
is_sparse=False):
"""
:api_attr: Static Graph
The hierarchical sigmoid organizes the classes into a complete binary tree to reduce the computational complexity
and speed up the model training, especially the training of language model.
Each leaf node of the complete binary tree represents a class(word) and each non-leaf node acts as a binary classifier.
......@@ -1167,6 +1188,10 @@ def softmax_with_cross_entropy(logits,
return_softmax=False,
axis=-1):
"""
:alias_main: paddle.nn.functional.softmax_with_cross_entropy
:alias: paddle.nn.functional.softmax_with_cross_entropy,paddle.nn.functional.loss.softmax_with_cross_entropy
:old_api: paddle.fluid.layers.softmax_with_cross_entropy
This operator implements the cross entropy loss function with softmax. This function
combines the calculation of the softmax operation and the cross entropy loss function
to provide a more numerically stable gradient.
......@@ -1290,6 +1315,10 @@ def softmax_with_cross_entropy(logits,
def rank_loss(label, left, right, name=None):
"""
:alias_main: paddle.nn.functional.rank_loss
:alias: paddle.nn.functional.rank_loss,paddle.nn.functional.loss.rank_loss
:old_api: paddle.fluid.layers.rank_loss
This operator implements the sort loss layer in the RankNet model. RankNet is a pairwise ranking model
with a training sample consisting of a pair of documents (A and B), The label (P)
indicates whether A is ranked higher than B or not. Please refer to more details:
......@@ -1407,6 +1436,10 @@ def sigmoid_cross_entropy_with_logits(x,
name=None,
normalize=False):
"""
:alias_main: paddle.nn.functional.sigmoid_cross_entropy_with_logits
:alias: paddle.nn.functional.sigmoid_cross_entropy_with_logits,paddle.nn.functional.loss.sigmoid_cross_entropy_with_logits
:old_api: paddle.fluid.layers.sigmoid_cross_entropy_with_logits
${comment}
Args:
......@@ -1464,6 +1497,10 @@ def teacher_student_sigmoid_loss(input,
soft_max_up_bound=15.0,
soft_max_lower_bound=-15.0):
"""
:alias_main: paddle.nn.functional.teacher_student_sigmoid_loss
:alias: paddle.nn.functional.teacher_student_sigmoid_loss,paddle.nn.functional.loss.teacher_student_sigmoid_loss
:old_api: paddle.fluid.layers.teacher_student_sigmoid_loss
**Teacher Student Log Loss Layer**
This layer accepts input predictions and target label and returns the
......@@ -1583,6 +1620,10 @@ def huber_loss(input, label, delta):
@templatedoc()
def kldiv_loss(x, target, reduction='mean', name=None):
"""
:alias_main: paddle.nn.functional.kldiv_loss
:alias: paddle.nn.functional.kldiv_loss,paddle.nn.functional.loss.kldiv_loss
:old_api: paddle.fluid.layers.kldiv_loss
${comment}
Args:
......@@ -1643,6 +1684,10 @@ from .control_flow import equal
def npair_loss(anchor, positive, labels, l2_reg=0.002):
'''
:alias_main: paddle.nn.functional.npair_loss
:alias: paddle.nn.functional.npair_loss,paddle.nn.functional.loss.npair_loss
:old_api: paddle.fluid.layers.npair_loss
**Npair Loss Layer**
Read `Improved Deep Metric Learning with Multi class N pair Loss Objective\
......@@ -1709,6 +1754,10 @@ def npair_loss(anchor, positive, labels, l2_reg=0.002):
def mse_loss(input, label):
"""
:alias_main: paddle.nn.functional.mse_loss
:alias: paddle.nn.functional.mse_loss,paddle.nn.functional.loss.mse_loss
:old_api: paddle.fluid.layers.mse_loss
This op accepts input predications and target label and returns the mean square error.
The loss can be described as:
......
python/paddle/fluid/layers/nn.py
浏览文件 @ 86ca31ab
......@@ -213,6 +213,8 @@ def fc(input,
act=None,
name=None):
"""
:api_attr: Static Graph
**Fully Connected Layer**
This operator creates a fully connected layer in the network. It can take
......@@ -370,6 +372,7 @@ def embedding(input,
param_attr=None,
dtype='float32'):
"""
:api_attr: Static Graph
**WARING:** This OP will be deprecated in a future release. This OP requires the
last dimension of Tensor shape must be equal to 1. It is recommended to use
......@@ -696,6 +699,8 @@ def _pull_box_sparse(input, size, dtype='float32'):
@templatedoc()
def linear_chain_crf(input, label, param_attr=None, length=None):
"""
:api_attr: Static Graph
Linear Chain CRF.
${comment}
......@@ -819,6 +824,7 @@ def linear_chain_crf(input, label, param_attr=None, length=None):
@templatedoc()
def crf_decoding(input, param_attr, label=None, length=None):
"""
:api_attr: Static Graph
${comment}
Args:
......@@ -924,6 +930,10 @@ def dropout(x,
name=None,
dropout_implementation="downgrade_in_infer"):
"""
:alias_main: paddle.nn.functional.dropout
:alias: paddle.nn.functional.dropout,paddle.nn.functional.common.dropout
:old_api: paddle.fluid.layers.dropout
Computes dropout.
Drop or keep each element of `x` independently. Dropout is a regularization
......@@ -1169,6 +1179,10 @@ def chunk_eval(input,
def softmax(input, use_cudnn=False, name=None, axis=-1):
"""
:alias_main: paddle.nn.functional.softmax
:alias: paddle.nn.functional.softmax,paddle.nn.functional.activation.softmax
:old_api: paddle.fluid.layers.softmax
This operator implements the softmax layer. The calculation process is as follows:
1. The dimension :attr:`axis` of the ``input`` will be permuted to the last.
......@@ -1309,6 +1323,8 @@ def conv2d(input,
name=None,
data_format="NCHW"):
"""
:api_attr: Static Graph
The convolution2D layer calculates the output based on the input, filter
and strides, paddings, dilations, groups parameters. Input and
Output are in NCHW or NHWC format, where N is batch size, C is the number of
......@@ -1583,6 +1599,8 @@ def conv3d(input,
name=None,
data_format="NCDHW"):
"""
:api_attr: Static Graph
The convolution3D layer calculates the output based on the input, filter
and strides, paddings, dilations, groups parameters. Input(Input) and
Output(Output) are in NCDHW or NDHWC format. Where N is batch size C is the number of
......@@ -1846,6 +1864,10 @@ def pool2d(input,
exclusive=True,
data_format="NCHW"):
"""
:alias_main: paddle.nn.functional.pool2d
:alias: paddle.nn.functional.pool2d,paddle.nn.functional.pooling.pool2d
:old_api: paddle.fluid.layers.pool2d
${comment}
Args:
......@@ -2059,6 +2081,10 @@ def pool3d(input,
exclusive=True,
data_format="NCDHW"):
"""
:alias_main: paddle.nn.functional.pool3d
:alias: paddle.nn.functional.pool3d,paddle.nn.functional.pooling.pool3d
:old_api: paddle.fluid.layers.pool3d
${comment}
Args:
......@@ -2277,6 +2303,10 @@ def adaptive_pool2d(input,
require_index=False,
name=None):
"""
:alias_main: paddle.nn.functional.adaptive_pool2d
:alias: paddle.nn.functional.adaptive_pool2d,paddle.nn.functional.pooling.adaptive_pool2d
:old_api: paddle.fluid.layers.adaptive_pool2d
This operation calculates the output based on the input, pool_size,
pool_type parameters. Input(X) and output(Out) are in NCHW format, where N is batch
size, C is the number of channels, H is the height of the feature, and W is
......@@ -2420,6 +2450,10 @@ def adaptive_pool3d(input,
require_index=False,
name=None):
"""
:alias_main: paddle.nn.functional.adaptive_pool3d
:alias: paddle.nn.functional.adaptive_pool3d,paddle.nn.functional.pooling.adaptive_pool3d
:old_api: paddle.fluid.layers.adaptive_pool3d
This operation calculates the output based on the input, pool_size,
pool_type parameters. Input(X) and output(Out) are in NCDHW format, where N is batch
size, C is the number of channels, D is the depth of the feature, H is the height of
......@@ -2589,6 +2623,8 @@ def batch_norm(input,
do_model_average_for_mean_and_var=True,
use_global_stats=False):
"""
:api_attr: Static Graph
**Batch Normalization Layer**
Can be used as a normalizer function for convolution or fully_connected operations.
......@@ -3041,6 +3077,8 @@ def instance_norm(input,
bias_attr=None,
name=None):
"""
:api_attr: Static Graph
**Instance Normalization Layer**
Can be used as a normalizer function for convolution or fully_connected operations.
......@@ -3166,6 +3204,8 @@ def data_norm(input,
summary_decay_rate=0.9999999,
enable_scale_and_shift=False):
"""
:api_attr: Static Graph
**Data Normalization Layer**
This op can be used as a normalizer function for conv2d and fully_connected operations.
......@@ -3345,6 +3385,8 @@ def layer_norm(input,
act=None,
name=None):
"""
:api_attr: Static Graph
**Layer Normalization Layer**
The API implements the function of the Layer Normalization Layer and can be applied to mini-batch input data.
......@@ -3471,6 +3513,8 @@ def group_norm(input,
data_layout='NCHW',
name=None):
"""
:api_attr: Static Graph
**Group Normalization Layer**
Refer to `Group Normalization <https://arxiv.org/abs/1803.08494>`_ .
......@@ -3566,6 +3610,8 @@ def group_norm(input,
@templatedoc()
def spectral_norm(weight, dim=0, power_iters=1, eps=1e-12, name=None):
"""
:api_attr: Static Graph
**Spectral Normalization Layer**
This operation calculates the spectral normalization value of weight parameters of
......@@ -3682,6 +3728,8 @@ def conv2d_transpose(input,
name=None,
data_format='NCHW'):
"""
:api_attr: Static Graph
The convolution2D transpose layer calculates the output based on the input,
filter, and dilations, strides, paddings. Input(Input) and output(Output)
are in NCHW or NHWC format. Where N is batch size, C is the number of channels,
......@@ -3970,6 +4018,8 @@ def conv3d_transpose(input,
name=None,
data_format='NCDHW'):
"""
:api_attr: Static Graph
The convolution3D transpose layer calculates the output based on the input,
filter, and dilations, strides, paddings. Input(Input) and output(Output)
are in NCDHW or NDHWC format. Where N is batch size, C is the number of channels,
......@@ -4256,6 +4306,10 @@ def conv3d_transpose(input,
def reduce_sum(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_sum
:alias: paddle.reduce_sum,paddle.tensor.reduce_sum,paddle.tensor.math.reduce_sum
:old_api: paddle.fluid.layers.reduce_sum
Computes the sum of tensor elements over the given dimension.
Args:
......@@ -4332,6 +4386,10 @@ def reduce_sum(input, dim=None, keep_dim=False, name=None):
def reduce_mean(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_mean
:alias: paddle.reduce_mean,paddle.tensor.reduce_mean,paddle.tensor.stat.reduce_mean
:old_api: paddle.fluid.layers.reduce_mean
Computes the mean of the input tensor's elements along the given dimension.
Args:
......@@ -4409,6 +4467,10 @@ def reduce_mean(input, dim=None, keep_dim=False, name=None):
def reduce_max(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_max
:alias: paddle.reduce_max,paddle.tensor.reduce_max,paddle.tensor.math.reduce_max
:old_api: paddle.fluid.layers.reduce_max
Computes the maximum of tensor elements over the given dimension.
Args:
......@@ -4471,6 +4533,10 @@ def reduce_max(input, dim=None, keep_dim=False, name=None):
def reduce_min(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_min
:alias: paddle.reduce_min,paddle.tensor.reduce_min,paddle.tensor.math.reduce_min
:old_api: paddle.fluid.layers.reduce_min
Computes the minimum of tensor elements over the given dimension.
Args:
......@@ -4533,6 +4599,10 @@ def reduce_min(input, dim=None, keep_dim=False, name=None):
def reduce_prod(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_prod
:alias: paddle.reduce_prod,paddle.tensor.reduce_prod,paddle.tensor.math.reduce_prod
:old_api: paddle.fluid.layers.reduce_prod
Computes the product of tensor elements over the given dimension.
Args:
......@@ -4596,6 +4666,10 @@ def reduce_prod(input, dim=None, keep_dim=False, name=None):
def reduce_all(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_all
:alias: paddle.reduce_all,paddle.tensor.reduce_all,paddle.tensor.logic.reduce_all
:old_api: paddle.fluid.layers.reduce_all
This OP computes the ``logical and`` of tensor elements over the given dimension, and output the result.
Args:
......@@ -4656,6 +4730,10 @@ def reduce_all(input, dim=None, keep_dim=False, name=None):
def reduce_any(input, dim=None, keep_dim=False, name=None):
"""
:alias_main: paddle.reduce_any
:alias: paddle.reduce_any,paddle.tensor.reduce_any,paddle.tensor.logic.reduce_any
:old_api: paddle.fluid.layers.reduce_any
This OP computes the ``logical or`` of tensor elements over the given dimension, and output the result.
Args:
......@@ -4864,6 +4942,10 @@ def split(input, num_or_sections, dim=-1, name=None):
def l2_normalize(x, axis, epsilon=1e-12, name=None):
"""
:alias_main: paddle.nn.functional.l2_normalize
:alias: paddle.nn.functional.l2_normalize,paddle.nn.functional.norm.l2_normalize
:old_api: paddle.fluid.layers.l2_normalize
This op normalizes `x` along dimension `axis` using an L2
norm. For a 1-D tensor (`dim` is fixed to 0), this layer computes
......@@ -5023,6 +5105,10 @@ def matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None):
def topk(input, k, name=None):
"""
:alias_main: paddle.topk
:alias: paddle.topk,paddle.tensor.topk,paddle.tensor.search.topk
:old_api: paddle.fluid.layers.topk
This OP is used to find values and indices of the k largest entries
for the last dimension.
......@@ -5284,6 +5370,10 @@ def ctc_greedy_decoder(input,
def transpose(x, perm, name=None):
"""
:alias_main: paddle.transpose
:alias: paddle.transpose,paddle.tensor.transpose,paddle.tensor.linalg.transpose,paddle.tensor.manipulation.transpose
:old_api: paddle.fluid.layers.transpose
Permute the data dimensions of `input` according to `perm`.
The `i`-th dimension of the returned tensor will correspond to the
......@@ -5376,6 +5466,8 @@ def im2sequence(input,
out_stride=1,
name=None):
"""
:api_attr: Static Graph
Extracts image patches from the input tensor to form a tensor of shape
{input.batch_size * output_height * output_width, filter_size_height *
filter_size_width * input.channels}. This op use filter to scan images
......@@ -5513,6 +5605,8 @@ def im2sequence(input,
@templatedoc()
def row_conv(input, future_context_size, param_attr=None, act=None):
"""
:api_attr: Static Graph
${comment}
Args:
......@@ -5633,6 +5727,10 @@ def multiplex(inputs, index):
def smooth_l1(x, y, inside_weight=None, outside_weight=None, sigma=None):
"""
:alias_main: paddle.nn.functional.smooth_l1
:alias: paddle.nn.functional.smooth_l1,paddle.nn.functional.loss.smooth_l1
:old_api: paddle.fluid.layers.smooth_l1
This layer computes the smooth L1 loss for Variable :attr:`x` and :attr:`y`.
It takes the first dimension of :attr:`x` and :attr:`y` as batch size.
For each instance, it computes the smooth L1 loss element by element first
......@@ -5820,6 +5918,8 @@ def one_hot(input, depth, allow_out_of_range=False):
def autoincreased_step_counter(counter_name=None, begin=1, step=1):
"""
:api_attr: Static Graph
Create an auto-increase variable. which will be automatically increased
by 1 in every iteration. By default, the first return of this counter is 1,
and the step size is 1.
......@@ -5864,6 +5964,10 @@ def autoincreased_step_counter(counter_name=None, begin=1, step=1):
def reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None):
"""
:alias_main: paddle.reshape
:alias: paddle.reshape,paddle.tensor.reshape,paddle.tensor.manipulation.reshape
:old_api: paddle.fluid.layers.reshape
This operator changes the shape of ``x`` without changing its data.
The target shape can be given by ``shape`` or ``actual_shape``.
......@@ -6359,6 +6463,10 @@ def lod_append(x, level):
def lrn(input, n=5, k=1.0, alpha=1e-4, beta=0.75, name=None,
data_format='NCHW'):
"""
:alias_main: paddle.nn.functional.lrn
:alias: paddle.nn.functional.lrn,paddle.nn.functional.norm.lrn
:old_api: paddle.fluid.layers.lrn
This operator implements the Local Response Normalization Layer.
This layer performs a type of "lateral inhibition" by normalizing over local input regions.
For more information, please refer to `ImageNet Classification with Deep Convolutional Neural Networks <https://papers.nips.cc/paper/4824-imagenet-classification-with-deep-convolutional-neural-networks.pdf>`_
......@@ -6445,6 +6553,10 @@ def lrn(input, n=5, k=1.0, alpha=1e-4, beta=0.75, name=None,
def pad(x, paddings, pad_value=0., name=None):
"""
:alias_main: paddle.nn.functional.pad
:alias: paddle.nn.functional.pad,paddle.nn.functional.common.pad
:old_api: paddle.fluid.layers.pad
This op will pad a tensor with a constant value given by :attr:`pad_value`, and the
padded shape is specified by :attr:`paddings`.
......@@ -6610,6 +6722,10 @@ def label_smooth(label,
dtype="float32",
name=None):
"""
:alias_main: paddle.nn.functional.label_smooth
:alias: paddle.nn.functional.label_smooth,paddle.nn.functional.common.label_smooth
:old_api: paddle.fluid.layers.label_smooth
Label smoothing is a mechanism to regularize the classifier layer and is called
label-smoothing regularization (LSR).
......@@ -6689,6 +6805,10 @@ def roi_pool(input,
spatial_scale=1.0,
rois_lod=None):
"""
:alias_main: paddle.nn.functional.roi_pool
:alias: paddle.nn.functional.roi_pool,paddle.nn.functional.vision.roi_pool
:old_api: paddle.fluid.layers.roi_pool
This operator implements the roi_pooling layer.
Region of interest pooling (also known as RoI pooling) is to perform max pooling on inputs of nonuniform sizes to obtain fixed-size feature maps (e.g. 7*7).
......@@ -6776,6 +6896,10 @@ def roi_align(input,
name=None,
rois_lod=None):
"""
:alias_main: paddle.nn.functional.roi_align
:alias: paddle.nn.functional.roi_align,paddle.nn.functional.vision.roi_align
:old_api: paddle.fluid.layers.roi_align
${comment}
Args:
......@@ -6840,6 +6964,10 @@ def roi_align(input,
def dice_loss(input, label, epsilon=0.00001, name=None):
"""
:alias_main: paddle.nn.functional.dice_loss
:alias: paddle.nn.functional.dice_loss,paddle.nn.functional.loss.dice_loss
:old_api: paddle.fluid.layers.dice_loss
Dice loss for comparing the similarity between the input predictions and the label.
This implementation is for binary classification, where the input is sigmoid
predictions of each pixel, usually used for segmentation task. The dice loss can
......@@ -6899,6 +7027,10 @@ def image_resize(input,
align_mode=1,
data_format='NCHW'):
"""
:alias_main: paddle.nn.functional.image_resize
:alias: paddle.nn.functional.image_resize,paddle.nn.functional.vision.image_resize
:old_api: paddle.fluid.layers.image_resize
This op resizes a batch of images.
The input must be a 3-D Tensor of the shape (num_batches, channels, in_w)
......@@ -7468,6 +7600,10 @@ def resize_bilinear(input,
align_mode=1,
data_format='NCHW'):
"""
:alias_main: paddle.nn.functional.resize_bilinear
:alias: paddle.nn.functional.resize_bilinear,paddle.nn.functional.vision.resize_bilinear
:old_api: paddle.fluid.layers.resize_bilinear
This op resizes the input by performing bilinear interpolation based on given
output shape which specified by actual_shape, out_shape and scale
in priority order.
......@@ -7631,6 +7767,10 @@ def resize_trilinear(input,
align_mode=1,
data_format='NCDHW'):
"""
:alias_main: paddle.nn.functional.resize_trilinear
:alias: paddle.nn.functional.resize_trilinear,paddle.nn.functional.vision.resize_trilinear
:old_api: paddle.fluid.layers.resize_trilinear
This op resizes the input by performing trilinear interpolation based on given
output shape which specified by actual_shape, out_shape and scale
in priority order.
......@@ -7795,6 +7935,10 @@ def resize_nearest(input,
align_corners=True,
data_format='NCHW'):
"""
:alias_main: paddle.nn.functional.resize_nearest
:alias: paddle.nn.functional.resize_nearest,paddle.nn.functional.vision.resize_nearest
:old_api: paddle.fluid.layers.resize_nearest
This op resizes the input by performing nearest neighbor interpolation in both the
height direction and the width direction based on given output shape
which is specified by actual_shape, out_shape and scale in priority order.
......@@ -8132,6 +8276,10 @@ def gather_nd(input, index, name=None):
def scatter(input, index, updates, name=None, overwrite=True):
"""
:alias_main: paddle.scatter
:alias: paddle.scatter,paddle.tensor.scatter,paddle.tensor.manipulation.scatter
:old_api: paddle.fluid.layers.scatter
**Scatter Layer**
Output is obtained by updating the input on selected indices based on updates.
......@@ -8392,6 +8540,10 @@ def random_crop(x, shape, seed=None):
def log(x, name=None):
"""
:alias_main: paddle.log
:alias: paddle.log,paddle.tensor.log,paddle.tensor.math.log
:old_api: paddle.fluid.layers.log
Calculates the natural log of the given input tensor, element-wise.
.. math::
......@@ -8481,6 +8633,10 @@ def relu(x, name=None):
def selu(x, scale=None, alpha=None, name=None):
"""
:alias_main: paddle.nn.functional.selu
:alias: paddle.nn.functional.selu,paddle.nn.functional.activation.selu
:old_api: paddle.fluid.layers.selu
Selu Operator.
The equation is:
......@@ -8718,6 +8874,10 @@ def crop(x, shape=None, offsets=None, name=None):
def crop_tensor(x, shape=None, offsets=None, name=None):
"""
:alias_main: paddle.crop_tensor
:alias: paddle.crop_tensor,paddle.tensor.crop_tensor,paddle.tensor.creation.crop_tensor
:old_api: paddle.fluid.layers.crop_tensor
Crop input into output, as specified by offsets and shape.
.. code-block:: text
......@@ -8914,6 +9074,10 @@ def crop_tensor(x, shape=None, offsets=None, name=None):
def affine_grid(theta, out_shape, name=None):
"""
:alias_main: paddle.nn.functional.affine_grid
:alias: paddle.nn.functional.affine_grid,paddle.nn.functional.vision.affine_grid
:old_api: paddle.fluid.layers.affine_grid
It generates a grid of (x,y) coordinates using the parameters of
the affine transformation that correspond to a set of points where
the input feature map should be sampled to produce the transformed
......@@ -8985,6 +9149,10 @@ def pad2d(input,
data_format="NCHW",
name=None):
"""
:alias_main: paddle.nn.functional.pad2d
:alias: paddle.nn.functional.pad2d,paddle.nn.functional.common.pad2d
:old_api: paddle.fluid.layers.pad2d
Pad 2-d images according to 'paddings' and 'mode'.
If mode is 'reflect', paddings[0] and paddings[1] must be no greater
than height-1. And the width dimension has the same condition.
......@@ -9082,6 +9250,10 @@ def pad2d(input,
@templatedoc()
def elu(x, alpha=1.0, name=None):
"""
:alias_main: paddle.nn.functional.elu
:alias: paddle.nn.functional.elu,paddle.nn.functional.activation.elu
:old_api: paddle.fluid.layers.elu
${comment}
Args:
x(${x_type}): ${x_comment}
......@@ -9120,6 +9292,10 @@ def elu(x, alpha=1.0, name=None):
@templatedoc()
def relu6(x, threshold=6.0, name=None):
"""
:alias_main: paddle.nn.functional.relu6
:alias: paddle.nn.functional.relu6,paddle.nn.functional.activation.relu6
:old_api: paddle.fluid.layers.relu6
${comment}
Args:
......@@ -9212,6 +9388,10 @@ def pow(x, factor=1.0, name=None):
@templatedoc()
def stanh(x, scale_a=0.67, scale_b=1.7159, name=None):
"""
:alias_main: paddle.stanh
:alias: paddle.stanh,paddle.tensor.stanh,paddle.tensor.math.stanh
:old_api: paddle.fluid.layers.stanh
${comment}
Args:
x(${x_type}): ${x_comment}
......@@ -9260,6 +9440,10 @@ def stanh(x, scale_a=0.67, scale_b=1.7159, name=None):
@templatedoc()
def hard_sigmoid(x, slope=0.2, offset=0.5, name=None):
"""
:alias_main: paddle.nn.functional.hard_sigmoid
:alias: paddle.nn.functional.hard_sigmoid,paddle.nn.functional.activation.hard_sigmoid
:old_api: paddle.fluid.layers.hard_sigmoid
${comment}
Parameters:
x (${x_type}): ${x_comment}
......@@ -9297,6 +9481,10 @@ def hard_sigmoid(x, slope=0.2, offset=0.5, name=None):
@templatedoc()
def swish(x, beta=1.0, name=None):
"""
:alias_main: paddle.nn.functional.swish
:alias: paddle.nn.functional.swish,paddle.nn.functional.activation.swish
:old_api: paddle.fluid.layers.swish
Elementwise swish activation function. See `Searching for Activation Functions <https://arxiv.org/abs/1710.05941>`_ for more details.
Equation:
......@@ -9377,6 +9565,8 @@ def swish(x, beta=1.0, name=None):
def prelu(x, mode, param_attr=None, name=None):
"""
:api_attr: Static Graph
Equation:
.. math::
......@@ -9448,6 +9638,10 @@ def prelu(x, mode, param_attr=None, name=None):
@templatedoc()
def brelu(x, t_min=0.0, t_max=24.0, name=None):
"""
:alias_main: paddle.nn.functional.brelu
:alias: paddle.nn.functional.brelu,paddle.nn.functional.activation.brelu
:old_api: paddle.fluid.layers.brelu
${comment}
Args:
x(${x_type}): ${x_comment}
......@@ -9489,6 +9683,10 @@ def brelu(x, t_min=0.0, t_max=24.0, name=None):
@templatedoc()
def leaky_relu(x, alpha=0.02, name=None):
"""
:alias_main: paddle.nn.functional.leaky_relu
:alias: paddle.nn.functional.leaky_relu,paddle.nn.functional.activation.leaky_relu
:old_api: paddle.fluid.layers.leaky_relu
${comment}
Args:
x(${x_type}): ${x_comment}
......@@ -9534,6 +9732,10 @@ def leaky_relu(x, alpha=0.02, name=None):
def soft_relu(x, threshold=40.0, name=None):
"""
:alias_main: paddle.nn.functional.soft_relu
:alias: paddle.nn.functional.soft_relu,paddle.nn.functional.activation.soft_relu
:old_api: paddle.fluid.layers.soft_relu
SoftRelu Activation Operator.
$out = \ln(1 + \exp(\max(\min(x, threshold), -threshold)))$
......@@ -9849,6 +10051,10 @@ def filter_by_instag(ins, ins_tag, filter_tag, is_lod, out_val_if_empty=0):
def unstack(x, axis=0, num=None):
"""
:alias_main: paddle.unstack
:alias: paddle.unstack,paddle.tensor.unstack,paddle.tensor.manipulation.unstack
:old_api: paddle.fluid.layers.unstack
**UnStack Layer**
This layer unstacks input Tensor :code:`x` into several Tensors along :code:`axis`.
......@@ -9899,6 +10105,10 @@ def unstack(x, axis=0, num=None):
def expand(x, expand_times, name=None):
"""
:alias_main: paddle.expand
:alias: paddle.expand,paddle.tensor.expand,paddle.tensor.manipulation.expand
:old_api: paddle.fluid.layers.expand
This operation tiles ``x`` multiple times according to the parameter ``expand_times``.
The times number for each dimension of ``x`` is set by the parameter ``expand_times``.
The rank of ``x`` should be less than or equal to 6. Please note that size of ``expand_times`` must be the same
......@@ -10015,6 +10225,10 @@ def expand(x, expand_times, name=None):
def expand_as(x, target_tensor, name=None):
"""
:alias_main: paddle.expand_as
:alias: paddle.expand_as,paddle.tensor.expand_as,paddle.tensor.manipulation.expand_as
:old_api: paddle.fluid.layers.expand_as
expand_as operator tiles to the input by given expand tensor. You should set expand tensor
for each dimension by providing tensor 'target_tensor'. The rank of X
should be in [1, 6]. Please note that size of 'target_tensor' must be the same
......@@ -10471,6 +10685,10 @@ def sum(x):
@templatedoc()
def slice(input, axes, starts, ends):
"""
:alias_main: paddle.slice
:alias: paddle.slice,paddle.tensor.slice,paddle.tensor.manipulation.slice
:old_api: paddle.fluid.layers.slice
This operator produces a slice of ``input`` along multiple axes. Similar to numpy:
https://docs.scipy.org/doc/numpy/reference/arrays.indexing.html
Slice uses ``axes``, ``starts`` and ``ends`` attributes to specify the start and
......@@ -10634,6 +10852,10 @@ def slice(input, axes, starts, ends):
@templatedoc()
def strided_slice(input, axes, starts, ends, strides):
"""
:alias_main: paddle.strided_slice
:alias: paddle.strided_slice,paddle.tensor.strided_slice,paddle.tensor.manipulation.strided_slice
:old_api: paddle.fluid.layers.strided_slice
This operator produces a slice of ``input`` along multiple axes. Similar to numpy:
https://docs.scipy.org/doc/numpy/reference/arrays.indexing.html
Slice uses ``axes``, ``starts`` and ``ends`` attributes to specify the start and
......@@ -10839,6 +11061,10 @@ def strided_slice(input, axes, starts, ends, strides):
def shape(input):
"""
:alias_main: paddle.shape
:alias: paddle.shape,paddle.tensor.shape,paddle.tensor.attribute.shape
:old_api: paddle.fluid.layers.shape
**Shape Layer**
Get the shape of the input.
......@@ -10878,6 +11104,10 @@ def shape(input):
def rank(input):
"""
:alias_main: paddle.rank
:alias: paddle.rank,paddle.tensor.rank,paddle.tensor.attribute.rank
:old_api: paddle.fluid.layers.rank
The OP returns the number of dimensions for a tensor, which is a 0-D int32 Tensor.
Args:
......@@ -10959,6 +11189,10 @@ def _elementwise_op(helper):
def scale(x, scale=1.0, bias=0.0, bias_after_scale=True, act=None, name=None):
"""
:alias_main: paddle.scale
:alias: paddle.scale,paddle.tensor.scale,paddle.tensor.math.scale
:old_api: paddle.fluid.layers.scale
Scale operator.
Putting scale and bias to the input Tensor as following:
......@@ -11053,6 +11287,10 @@ def scale(x, scale=1.0, bias=0.0, bias_after_scale=True, act=None, name=None):
def elementwise_add(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_add
:alias: paddle.elementwise_add,paddle.tensor.elementwise_add,paddle.tensor.math.elementwise_add
:old_api: paddle.fluid.layers.elementwise_add
Examples:
.. code-block:: python
......@@ -11137,6 +11375,10 @@ Examples:
def elementwise_div(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_div
:alias: paddle.elementwise_div,paddle.tensor.elementwise_div,paddle.tensor.math.elementwise_div
:old_api: paddle.fluid.layers.elementwise_div
Examples:
.. code-block:: python
......@@ -11221,6 +11463,10 @@ Examples:
def elementwise_sub(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_sub
:alias: paddle.elementwise_sub,paddle.tensor.elementwise_sub,paddle.tensor.math.elementwise_sub
:old_api: paddle.fluid.layers.elementwise_sub
Examples:
.. code-block:: python
......@@ -11305,6 +11551,10 @@ Examples:
def elementwise_mul(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_mul
:alias: paddle.elementwise_mul,paddle.tensor.elementwise_mul,paddle.tensor.math.elementwise_mul
:old_api: paddle.fluid.layers.elementwise_mul
Examples:
.. code-block:: python
......@@ -11389,6 +11639,10 @@ Examples:
def elementwise_max(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_max
:alias: paddle.elementwise_max,paddle.tensor.elementwise_max,paddle.tensor.math.elementwise_max
:old_api: paddle.fluid.layers.elementwise_max
Examples:
.. code-block:: python
......@@ -11447,6 +11701,10 @@ Examples:
def elementwise_min(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_min
:alias: paddle.elementwise_min,paddle.tensor.elementwise_min,paddle.tensor.math.elementwise_min
:old_api: paddle.fluid.layers.elementwise_min
Examples:
.. code-block:: python
......@@ -11503,6 +11761,10 @@ Examples:
def elementwise_pow(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_pow
:alias: paddle.elementwise_pow,paddle.tensor.elementwise_pow,paddle.tensor.math.elementwise_pow
:old_api: paddle.fluid.layers.elementwise_pow
Examples:
.. code-block:: python
......@@ -11535,6 +11797,10 @@ Examples:
def elementwise_mod(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_mod
:alias: paddle.elementwise_mod,paddle.tensor.elementwise_mod,paddle.tensor.math.elementwise_mod
:old_api: paddle.fluid.layers.elementwise_mod
Examples:
.. code-block:: python
......@@ -11568,6 +11834,10 @@ Examples:
def elementwise_floordiv(x, y, axis=-1, act=None, name=None):
"""
:alias_main: paddle.elementwise_floordiv
:alias: paddle.elementwise_floordiv,paddle.tensor.elementwise_floordiv,paddle.tensor.math.elementwise_floordiv
:old_api: paddle.fluid.layers.elementwise_floordiv
Examples:
.. code-block:: python
......@@ -11701,6 +11971,10 @@ def _logical_op(op_name, x, y, out=None, name=None, binary_op=True):
@templatedoc()
def logical_and(x, y, out=None, name=None):
"""
:alias_main: paddle.logical_and
:alias: paddle.logical_and,paddle.tensor.logical_and,paddle.tensor.logic.logical_and
:old_api: paddle.fluid.layers.logical_and
logical_and Operator
It operates element-wise on X and Y, and returns the Out. X, Y and Out are N-dim boolean LoDTensor or Tensor.
......@@ -11750,6 +12024,10 @@ def logical_and(x, y, out=None, name=None):
@templatedoc()
def logical_or(x, y, out=None, name=None):
"""
:alias_main: paddle.logical_or
:alias: paddle.logical_or,paddle.tensor.logical_or,paddle.tensor.logic.logical_or
:old_api: paddle.fluid.layers.logical_or
logical_or Operator
It operates element-wise on X and Y, and returns the Out. X, Y and Out are N-dim boolean LoDTensor or Tensor.
......@@ -11799,6 +12077,10 @@ def logical_or(x, y, out=None, name=None):
@templatedoc()
def logical_xor(x, y, out=None, name=None):
"""
:alias_main: paddle.logical_xor
:alias: paddle.logical_xor,paddle.tensor.logical_xor,paddle.tensor.logic.logical_xor
:old_api: paddle.fluid.layers.logical_xor
logical_xor Operator
It operates element-wise on X and Y, and returns the Out. X, Y and Out are N-dim boolean LoDTensor or Tensor.
......@@ -11848,6 +12130,10 @@ def logical_xor(x, y, out=None, name=None):
@templatedoc()
def logical_not(x, out=None, name=None):
"""
:alias_main: paddle.logical_not
:alias: paddle.logical_not,paddle.tensor.logical_not,paddle.tensor.logic.logical_not
:old_api: paddle.fluid.layers.logical_not
logical_not Operator
It operates element-wise on X, and returns the Out. X and Out are N-dim boolean LoDTensor or Tensor.
......@@ -11894,6 +12180,10 @@ def logical_not(x, out=None, name=None):
@templatedoc()
def clip(x, min, max, name=None):
"""
:alias_main: paddle.nn.clip
:alias: paddle.nn.clip,paddle.nn.clip.clip
:old_api: paddle.fluid.layers.clip
${comment}
Args:
......@@ -11989,6 +12279,10 @@ def clip_by_norm(x, max_norm, name=None):
@templatedoc()
def mean(x, name=None):
"""
:alias_main: paddle.mean
:alias: paddle.mean,paddle.tensor.mean,paddle.tensor.stat.mean
:old_api: paddle.fluid.layers.mean
${comment}
Args:
......@@ -12105,6 +12399,10 @@ def mul(x, y, x_num_col_dims=1, y_num_col_dims=1, name=None):
@templatedoc()
def maxout(x, groups, name=None, axis=1):
"""
:alias_main: paddle.nn.functional.maxout
:alias: paddle.nn.functional.maxout,paddle.nn.functional.activation.maxout
:old_api: paddle.fluid.layers.maxout
${comment}
Args:
......@@ -12155,6 +12453,10 @@ def maxout(x, groups, name=None, axis=1):
def space_to_depth(x, blocksize, name=None):
"""
:alias_main: paddle.nn.functional.space_to_depth
:alias: paddle.nn.functional.space_to_depth,paddle.nn.functional.vision.space_to_depth
:old_api: paddle.fluid.layers.space_to_depth
Gives a blocksize to space_to_depth the input LoDtensor with Layout: [batch, channel, height, width]
This op rearranges blocks of spatial data, into depth. More specifically, this op outputs a copy of \
......@@ -12262,6 +12564,10 @@ def affine_channel(x,
name=None,
act=None):
"""
:alias_main: paddle.nn.functional.affine_channel
:alias: paddle.nn.functional.affine_channel,paddle.nn.functional.vision.affine_channel
:old_api: paddle.fluid.layers.affine_channel
Applies a separate affine transformation to each channel of the input.
Useful for replacing spatial batch norm with its equivalent fixed
transformation. The input also can be 2D tensor and applies a affine
......@@ -12447,6 +12753,10 @@ def similarity_focus(input, axis, indexes, name=None):
def hash(input, hash_size, num_hash=1, name=None):
"""
:alias_main: paddle.nn.functional.hash
:alias: paddle.nn.functional.hash,paddle.nn.functional.lod.hash
:old_api: paddle.fluid.layers.hash
This OP hash the input to an integer less than the hash_size.
The hash algorithm we used was xxHash - Extremely fast hash algorithm
(https://github.com/Cyan4973/xxHash/tree/v0.6.5)
......@@ -12509,6 +12819,10 @@ def hash(input, hash_size, num_hash=1, name=None):
@templatedoc()
def grid_sampler(x, grid, name=None):
"""
:alias_main: paddle.nn.functional.grid_sampler
:alias: paddle.nn.functional.grid_sampler,paddle.nn.functional.vision.grid_sampler
:old_api: paddle.fluid.layers.grid_sampler
This operation samples input X by using bilinear interpolation based on
flow field grid, which is usually generated by :code:`affine_grid` . The grid of
shape [N, H, W, 2] is the concatenation of (x, y) coordinates
......@@ -12609,6 +12923,10 @@ def grid_sampler(x, grid, name=None):
def log_loss(input, label, epsilon=1e-4, name=None):
"""
:alias_main: paddle.nn.functional.log_loss
:alias: paddle.nn.functional.log_loss,paddle.nn.functional.loss.log_loss
:old_api: paddle.fluid.layers.log_loss
**Negative Log Loss Layer**
This layer accepts input predictions and target label and returns the
......@@ -12658,6 +12976,10 @@ def log_loss(input, label, epsilon=1e-4, name=None):
def add_position_encoding(input, alpha, beta, name=None):
"""
:alias_main: paddle.nn.functional.add_position_encoding
:alias: paddle.nn.functional.add_position_encoding,paddle.nn.functional.extension.add_position_encoding
:old_api: paddle.fluid.layers.add_position_encoding
This operator performs weighted sum of input feature at each position
(position in the sequence) and the corresponding position encoding.
......@@ -12730,6 +13052,8 @@ def bilinear_tensor_product(x,
param_attr=None,
bias_attr=None):
"""
:api_attr: Static Graph
**Bilinear Tensor Product Layer**
This layer performs bilinear tensor product on two inputs.
......@@ -12920,6 +13244,10 @@ def shuffle_channel(x, group, name=None):
@templatedoc()
def temporal_shift(x, seg_num, shift_ratio=0.25, name=None):
"""
:alias_main: paddle.nn.functional.temporal_shift
:alias: paddle.nn.functional.temporal_shift,paddle.nn.functional.extension.temporal_shift
:old_api: paddle.fluid.layers.temporal_shift
**Temporal Shift Operator**
${comment}
......@@ -13042,6 +13370,8 @@ class PyFuncRegistry(object):
@templatedoc()
def py_func(func, x, out, backward_func=None, skip_vars_in_backward_input=None):
"""
:api_attr: Static Graph
This OP is used to register customized Python OP to Paddle Fluid. The design
principe of py_func is that LodTensor and numpy array can be converted to each
other easily. So you can use Python and numpy API to register a python OP.
......@@ -13269,6 +13599,10 @@ def psroi_pool(input,
pooled_width,
name=None):
"""
:alias_main: paddle.nn.functional.psroi_pool
:alias: paddle.nn.functional.psroi_pool,paddle.nn.functional.vision.psroi_pool
:old_api: paddle.fluid.layers.psroi_pool
${comment}
Parameters:
......@@ -13335,6 +13669,10 @@ def prroi_pool(input,
batch_roi_nums=None,
name=None):
"""
:alias_main: paddle.nn.functional.prroi_pool
:alias: paddle.nn.functional.prroi_pool,paddle.nn.functional.vision.prroi_pool
:old_api: paddle.fluid.layers.prroi_pool
The precise roi pooling implementation for paddle. Reference: https://arxiv.org/pdf/1807.11590.pdf
Args:
......@@ -13626,6 +13964,10 @@ def where(condition):
def sign(x):
"""
:alias_main: paddle.sign
:alias: paddle.sign,paddle.tensor.sign,paddle.tensor.math.sign
:old_api: paddle.fluid.layers.sign
This OP returns sign of every element in `x`: 1 for positive, -1 for negative and 0 for zero.
Args:
......@@ -13659,6 +14001,10 @@ def sign(x):
def unique(x, dtype='int32'):
"""
:alias_main: paddle.unique
:alias: paddle.unique,paddle.tensor.unique,paddle.tensor.manipulation.unique
:old_api: paddle.fluid.layers.unique
**unique**
Return a unique tensor for `x` and an index tensor pointing to this unique tensor.
......@@ -13772,6 +14118,8 @@ def deformable_conv(input,
modulated=True,
name=None):
"""
:api_attr: Static Graph
**Deformable Convolution op**
Compute 2-D deformable convolution on 4-D input.
......@@ -13984,6 +14332,9 @@ def deformable_conv(input,
def unfold(x, kernel_sizes, strides=1, paddings=0, dilations=1, name=None):
"""
:alias_main: paddle.nn.functional.unfold
:alias: paddle.nn.functional.unfold,paddle.nn.functional.common.unfold
:old_api: paddle.fluid.layers.unfold
This op returns a col buffer of sliding local blocks of input x, also known
as im2col for batched 2D image tensors. For each block under the convolution filter,
......@@ -14119,6 +14470,10 @@ def deformable_roi_pooling(input,
position_sensitive=False,
name=None):
"""
:alias_main: paddle.nn.functional.deformable_roi_pooling
:alias: paddle.nn.functional.deformable_roi_pooling,paddle.nn.functional.vision.deformable_roi_pooling
:old_api: paddle.fluid.layers.deformable_roi_pooling
Deformable ROI Pooling Layer
Performs deformable region-of-interest pooling on inputs. As described
......@@ -14348,6 +14703,10 @@ def shard_index(input, index_num, nshards, shard_id, ignore_value=-1):
@templatedoc()
def hard_swish(x, threshold=6.0, scale=6.0, offset=3.0, name=None):
"""
:alias_main: paddle.nn.functional.hard_swish
:alias: paddle.nn.functional.hard_swish,paddle.nn.functional.activation.hard_swish
:old_api: paddle.fluid.layers.hard_swish
This operator implements the hard_swish activation function.
Hard_swish is proposed in MobileNetV3, and performs better in computational stability and efficiency compared to swish function.
For more details please refer to: https://arxiv.org/pdf/1905.02244.pdf
......
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