Skip to content

P
Paddle

Crayon鑫 / Paddle
与 Fork 源项目一致

Fork自 PaddlePaddle / Paddle

通知 1
Star 1
Fork 0
P
Paddle
未验证 提交 f5c6dd6d 编写于 作者: S swtkiwi 提交者： GitHub
浏览文件
操作

test=develop (#24522)

上级 a9520db2
隐藏空白更改
内联 并排
Showing with 566 addition and 5 deletion
python/paddle/fluid/layers/detection.py
浏览文件 @ f5c6dd6d
...@@ -454,6 +454,10 @@ def rpn_target_assign(bbox_pred, ...@@ -454,6 +454,10 @@ def rpn_target_assign(bbox_pred,
def sigmoid_focal_loss(x, label, fg_num, gamma=2.0, alpha=0.25): def sigmoid_focal_loss(x, label, fg_num, gamma=2.0, alpha=0.25):
""" """
:alias_main: paddle.nn.functional.sigmoid_focal_loss
:alias: paddle.nn.functional.sigmoid_focal_loss,paddle.nn.functional.loss.sigmoid_focal_loss
:old_api: paddle.fluid.layers.sigmoid_focal_loss
**Sigmoid Focal Loss Operator.** **Sigmoid Focal Loss Operator.**
`Focal Loss <https://arxiv.org/abs/1708.02002>`_ is used to address the foreground-background `Focal Loss <https://arxiv.org/abs/1708.02002>`_ is used to address the foreground-background
...@@ -550,6 +554,10 @@ def detection_output(loc, ...@@ -550,6 +554,10 @@ def detection_output(loc,
nms_eta=1.0, nms_eta=1.0,
return_index=False): return_index=False):
""" """
:alias_main: paddle.nn.functional.detection_output
:alias: paddle.nn.functional.detection_output,paddle.nn.functional.vision.detection_output
:old_api: paddle.fluid.layers.detection_output
Given the regression locations, classification confidences and prior boxes, Given the regression locations, classification confidences and prior boxes,
calculate the detection outputs by performing following steps: calculate the detection outputs by performing following steps:
...@@ -679,6 +687,10 @@ def detection_output(loc, ...@@ -679,6 +687,10 @@ def detection_output(loc,
@templatedoc() @templatedoc()
def iou_similarity(x, y, box_normalized=True, name=None): def iou_similarity(x, y, box_normalized=True, name=None):
""" """
:alias_main: paddle.nn.functional.iou_similarity
:alias: paddle.nn.functional.iou_similarity,paddle.nn.functional.loss.iou_similarity
:old_api: paddle.fluid.layers.iou_similarity
${comment} ${comment}
Args: Args:
...@@ -735,6 +747,10 @@ def box_coder(prior_box, ...@@ -735,6 +747,10 @@ def box_coder(prior_box,
name=None, name=None,
axis=0): axis=0):
""" """
:alias_main: paddle.nn.functional.box_coder
:alias: paddle.nn.functional.box_coder,paddle.nn.functional.vision.box_coder
:old_api: paddle.fluid.layers.box_coder
**Box Coder Layer** **Box Coder Layer**
Encode/Decode the target bounding box with the priorbox information. Encode/Decode the target bounding box with the priorbox information.
...@@ -922,6 +938,10 @@ def yolov3_loss(x, ...@@ -922,6 +938,10 @@ def yolov3_loss(x,
name=None, name=None,
scale_x_y=1.): scale_x_y=1.):
""" """
:alias_main: paddle.nn.functional.yolov3_loss
:alias: paddle.nn.functional.yolov3_loss,paddle.nn.functional.vision.yolov3_loss
:old_api: paddle.fluid.layers.yolov3_loss
${comment} ${comment}
Args: Args:
...@@ -1045,6 +1065,10 @@ def yolo_box(x, ...@@ -1045,6 +1065,10 @@ def yolo_box(x,
name=None, name=None,
scale_x_y=1.): scale_x_y=1.):
""" """
:alias_main: paddle.nn.functional.yolo_box
:alias: paddle.nn.functional.yolo_box,paddle.nn.functional.vision.yolo_box
:old_api: paddle.fluid.layers.yolo_box
${comment} ${comment}
Args: Args:
...@@ -1220,6 +1244,10 @@ def bipartite_match(dist_matrix, ...@@ -1220,6 +1244,10 @@ def bipartite_match(dist_matrix,
dist_threshold=None, dist_threshold=None,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.bipartite_match
:alias: paddle.nn.functional.bipartite_match,paddle.nn.functional.vision.bipartite_match
:old_api: paddle.fluid.layers.bipartite_match
This operator implements a greedy bipartite matching algorithm, which is This operator implements a greedy bipartite matching algorithm, which is
used to obtain the matching with the maximum distance based on the input used to obtain the matching with the maximum distance based on the input
distance matrix. For input 2D matrix, the bipartite matching algorithm can distance matrix. For input 2D matrix, the bipartite matching algorithm can
...@@ -1310,6 +1338,10 @@ def target_assign(input, ...@@ -1310,6 +1338,10 @@ def target_assign(input,
mismatch_value=None, mismatch_value=None,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.target_assign
:alias: paddle.nn.functional.target_assign,paddle.nn.functional.extension.target_assign
:old_api: paddle.fluid.layers.target_assign
This operator can be, for given the target bounding boxes or labels, This operator can be, for given the target bounding boxes or labels,
to assign classification and regression targets to each prediction as well as to assign classification and regression targets to each prediction as well as
weights to prediction. The weights is used to specify which prediction would weights to prediction. The weights is used to specify which prediction would
...@@ -1424,6 +1456,10 @@ def ssd_loss(location, ...@@ -1424,6 +1456,10 @@ def ssd_loss(location,
normalize=True, normalize=True,
sample_size=None): sample_size=None):
""" """
:alias_main: paddle.nn.functional.ssd_loss
:alias: paddle.nn.functional.ssd_loss,paddle.nn.functional.loss.ssd_loss
:old_api: paddle.fluid.layers.ssd_loss
**Multi-box loss layer for object detection algorithm of SSD** **Multi-box loss layer for object detection algorithm of SSD**
This layer is to compute detection loss for SSD given the location offset This layer is to compute detection loss for SSD given the location offset
...@@ -1667,6 +1703,10 @@ def prior_box(input, ...@@ -1667,6 +1703,10 @@ def prior_box(input,
name=None, name=None,
min_max_aspect_ratios_order=False): min_max_aspect_ratios_order=False):
""" """
:alias_main: paddle.nn.functional.prior_box
:alias: paddle.nn.functional.prior_box,paddle.nn.functional.vision.prior_box
:old_api: paddle.fluid.layers.prior_box
This op generates prior boxes for SSD(Single Shot MultiBox Detector) algorithm. This op generates prior boxes for SSD(Single Shot MultiBox Detector) algorithm.
Each position of the input produce N prior boxes, N is determined by Each position of the input produce N prior boxes, N is determined by
the count of min_sizes, max_sizes and aspect_ratios, The size of the the count of min_sizes, max_sizes and aspect_ratios, The size of the
...@@ -1824,6 +1864,10 @@ def density_prior_box(input, ...@@ -1824,6 +1864,10 @@ def density_prior_box(input,
flatten_to_2d=False, flatten_to_2d=False,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.density_prior_box
:alias: paddle.nn.functional.density_prior_box,paddle.nn.functional.vision.density_prior_box
:old_api: paddle.fluid.layers.density_prior_box
This op generates density prior boxes for SSD(Single Shot MultiBox Detector) This op generates density prior boxes for SSD(Single Shot MultiBox Detector)
algorithm. Each position of the input produce N prior boxes, N is algorithm. Each position of the input produce N prior boxes, N is
...@@ -2012,6 +2056,8 @@ def multi_box_head(inputs, ...@@ -2012,6 +2056,8 @@ def multi_box_head(inputs,
name=None, name=None,
min_max_aspect_ratios_order=False): min_max_aspect_ratios_order=False):
""" """
:api_attr: Static Graph
Base on SSD ((Single Shot MultiBox Detector) algorithm, generate prior boxes, Base on SSD ((Single Shot MultiBox Detector) algorithm, generate prior boxes,
regression location and classification confidence on multiple input feature regression location and classification confidence on multiple input feature
maps, then output the concatenate results. The details of this algorithm, maps, then output the concatenate results. The details of this algorithm,
...@@ -2287,6 +2333,10 @@ def anchor_generator(input, ...@@ -2287,6 +2333,10 @@ def anchor_generator(input,
offset=0.5, offset=0.5,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.anchor_generator
:alias: paddle.nn.functional.anchor_generator,paddle.nn.functional.vision.anchor_generator
:old_api: paddle.fluid.layers.anchor_generator
**Anchor generator operator** **Anchor generator operator**
Generate anchors for Faster RCNN algorithm. Generate anchors for Faster RCNN algorithm.
...@@ -2488,6 +2538,10 @@ def generate_proposal_labels(rpn_rois, ...@@ -2488,6 +2538,10 @@ def generate_proposal_labels(rpn_rois,
is_cls_agnostic=False, is_cls_agnostic=False,
is_cascade_rcnn=False): is_cascade_rcnn=False):
""" """
:alias_main: paddle.nn.functional.generate_proposal_labels
:alias: paddle.nn.functional.generate_proposal_labels,paddle.nn.functional.vision.generate_proposal_labels
:old_api: paddle.fluid.layers.generate_proposal_labels
**Generate Proposal Labels of Faster-RCNN** **Generate Proposal Labels of Faster-RCNN**
This operator can be, for given the GenerateProposalOp output bounding boxes and groundtruth, This operator can be, for given the GenerateProposalOp output bounding boxes and groundtruth,
...@@ -2602,6 +2656,10 @@ def generate_proposal_labels(rpn_rois, ...@@ -2602,6 +2656,10 @@ def generate_proposal_labels(rpn_rois,
def generate_mask_labels(im_info, gt_classes, is_crowd, gt_segms, rois, def generate_mask_labels(im_info, gt_classes, is_crowd, gt_segms, rois,
labels_int32, num_classes, resolution): labels_int32, num_classes, resolution):
""" """
:alias_main: paddle.nn.functional.generate_mask_labels
:alias: paddle.nn.functional.generate_mask_labels,paddle.nn.functional.vision.generate_mask_labels
:old_api: paddle.fluid.layers.generate_mask_labels
**Generate Mask Labels for Mask-RCNN** **Generate Mask Labels for Mask-RCNN**
This operator can be, for given the RoIs and corresponding labels, This operator can be, for given the RoIs and corresponding labels,
...@@ -2757,6 +2815,10 @@ def generate_proposals(scores, ...@@ -2757,6 +2815,10 @@ def generate_proposals(scores,
name=None, name=None,
return_rois_num=False): return_rois_num=False):
""" """
:alias_main: paddle.nn.functional.generate_proposals
:alias: paddle.nn.functional.generate_proposals,paddle.nn.functional.vision.generate_proposals
:old_api: paddle.fluid.layers.generate_proposals
**Generate proposal Faster-RCNN** **Generate proposal Faster-RCNN**
This operation proposes RoIs according to each box with their This operation proposes RoIs according to each box with their
...@@ -2867,6 +2929,10 @@ def generate_proposals(scores, ...@@ -2867,6 +2929,10 @@ def generate_proposals(scores,
def box_clip(input, im_info, name=None): def box_clip(input, im_info, name=None):
""" """
:alias_main: paddle.nn.functional.box_clip
:alias: paddle.nn.functional.box_clip,paddle.nn.functional.vision.box_clip
:old_api: paddle.fluid.layers.box_clip
Clip the box into the size given by im_info Clip the box into the size given by im_info
For each input box, The formula is given as follows: For each input box, The formula is given as follows:
...@@ -3092,6 +3158,10 @@ def multiclass_nms(bboxes, ...@@ -3092,6 +3158,10 @@ def multiclass_nms(bboxes,
background_label=0, background_label=0,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.multiclass_nms
:alias: paddle.nn.functional.multiclass_nms,paddle.nn.functional.extension.multiclass_nms
:old_api: paddle.fluid.layers.multiclass_nms
**Multiclass NMS** **Multiclass NMS**
This operator is to do multi-class non maximum suppression (NMS) on This operator is to do multi-class non maximum suppression (NMS) on
...@@ -3369,6 +3439,10 @@ def distribute_fpn_proposals(fpn_rois, ...@@ -3369,6 +3439,10 @@ def distribute_fpn_proposals(fpn_rois,
refer_scale, refer_scale,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.distribute_fpn_proposals
:alias: paddle.nn.functional.distribute_fpn_proposals,paddle.nn.functional.vision.distribute_fpn_proposals
:old_api: paddle.fluid.layers.distribute_fpn_proposals
**This op only takes LoDTensor as input.** In Feature Pyramid Networks **This op only takes LoDTensor as input.** In Feature Pyramid Networks
(FPN) models, it is needed to distribute all proposals into different FPN (FPN) models, it is needed to distribute all proposals into different FPN
level, with respect to scale of the proposals, the referring scale and the level, with respect to scale of the proposals, the referring scale and the
...@@ -3454,6 +3528,10 @@ def box_decoder_and_assign(prior_box, ...@@ -3454,6 +3528,10 @@ def box_decoder_and_assign(prior_box,
box_clip, box_clip,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.box_decoder_and_assign
:alias: paddle.nn.functional.box_decoder_and_assign,paddle.nn.functional.vision.box_decoder_and_assign
:old_api: paddle.fluid.layers.box_decoder_and_assign
${comment} ${comment}
Args: Args:
prior_box(${prior_box_type}): ${prior_box_comment} prior_box(${prior_box_type}): ${prior_box_comment}
...@@ -3525,6 +3603,10 @@ def collect_fpn_proposals(multi_rois, ...@@ -3525,6 +3603,10 @@ def collect_fpn_proposals(multi_rois,
post_nms_top_n, post_nms_top_n,
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.collect_fpn_proposals
:alias: paddle.nn.functional.collect_fpn_proposals,paddle.nn.functional.vision.collect_fpn_proposals
:old_api: paddle.fluid.layers.collect_fpn_proposals
**This OP only supports LoDTensor as input**. Concat multi-level RoIs **This OP only supports LoDTensor as input**. Concat multi-level RoIs
(Region of Interest) and select N RoIs with respect to multi_scores. (Region of Interest) and select N RoIs with respect to multi_scores.
This operation performs the following steps: This operation performs the following steps:
......
python/paddle/fluid/layers/io.py
浏览文件 @ f5c6dd6d
...@@ -557,6 +557,8 @@ def py_reader(capacity, ...@@ -557,6 +557,8 @@ def py_reader(capacity,
name=None, name=None,
use_double_buffer=True): use_double_buffer=True):
""" """
:api_attr: Static Graph
Create a Python reader for data feeding in Python Create a Python reader for data feeding in Python
This operator returns a Reader Variable. This operator returns a Reader Variable.
...@@ -724,6 +726,8 @@ def create_py_reader_by_data(capacity, ...@@ -724,6 +726,8 @@ def create_py_reader_by_data(capacity,
name=None, name=None,
use_double_buffer=True): use_double_buffer=True):
""" """
:api_attr: Static Graph
The OP creates a Python reader for data feeding in Python, it is similar The OP creates a Python reader for data feeding in Python, it is similar
to :ref:`api_fluid_layers_py_reader` except that it can read data from to :ref:`api_fluid_layers_py_reader` except that it can read data from
the list of feed variables. the list of feed variables.
...@@ -861,6 +865,8 @@ def double_buffer(reader, place=None, name=None): ...@@ -861,6 +865,8 @@ def double_buffer(reader, place=None, name=None):
def read_file(reader): def read_file(reader):
""" """
:api_attr: Static Graph
Execute the given reader and get data via it. Execute the given reader and get data via it.
A reader is also a Variable. It can be a raw reader generated by A reader is also a Variable. It can be a raw reader generated by
......
python/paddle/fluid/layers/learning_rate_scheduler.py
浏览文件 @ f5c6dd6d
...@@ -52,6 +52,10 @@ def _decay_step_counter(begin=0): ...@@ -52,6 +52,10 @@ def _decay_step_counter(begin=0):
def noam_decay(d_model, warmup_steps, learning_rate=1.0): def noam_decay(d_model, warmup_steps, learning_rate=1.0):
""" """
:alias_main: paddle.nn.functional.noam_decay
:alias: paddle.nn.functional.noam_decay,paddle.nn.functional.learning_rate.noam_decay
:old_api: paddle.fluid.layers.noam_decay
Noam decay method. The numpy implementation of noam decay as follows. Noam decay method. The numpy implementation of noam decay as follows.
.. code-block:: python .. code-block:: python
...@@ -111,6 +115,10 @@ def noam_decay(d_model, warmup_steps, learning_rate=1.0): ...@@ -111,6 +115,10 @@ def noam_decay(d_model, warmup_steps, learning_rate=1.0):
def exponential_decay(learning_rate, decay_steps, decay_rate, staircase=False): def exponential_decay(learning_rate, decay_steps, decay_rate, staircase=False):
""" """
:alias_main: paddle.nn.functional.exponential_decay
:alias: paddle.nn.functional.exponential_decay,paddle.nn.functional.learning_rate.exponential_decay
:old_api: paddle.fluid.layers.exponential_decay
Applies exponential decay to the learning rate. Applies exponential decay to the learning rate.
When training a model, it is often recommended to lower the learning rate as the When training a model, it is often recommended to lower the learning rate as the
...@@ -167,7 +175,12 @@ def exponential_decay(learning_rate, decay_steps, decay_rate, staircase=False): ...@@ -167,7 +175,12 @@ def exponential_decay(learning_rate, decay_steps, decay_rate, staircase=False):
def natural_exp_decay(learning_rate, decay_steps, decay_rate, staircase=False): def natural_exp_decay(learning_rate, decay_steps, decay_rate, staircase=False):
"""Applies natural exponential decay to the initial learning rate. """
:alias_main: paddle.nn.functional.natural_exp_decay
:alias: paddle.nn.functional.natural_exp_decay,paddle.nn.functional.learning_rate.natural_exp_decay
:old_api: paddle.fluid.layers.natural_exp_decay
Applies natural exponential decay to the initial learning rate.
When training a model, it is often recommended to lower the learning rate as the When training a model, it is often recommended to lower the learning rate as the
training progresses. By using this function, the learning rate will be decayed by training progresses. By using this function, the learning rate will be decayed by
...@@ -224,6 +237,10 @@ def natural_exp_decay(learning_rate, decay_steps, decay_rate, staircase=False): ...@@ -224,6 +237,10 @@ def natural_exp_decay(learning_rate, decay_steps, decay_rate, staircase=False):
def inverse_time_decay(learning_rate, decay_steps, decay_rate, staircase=False): def inverse_time_decay(learning_rate, decay_steps, decay_rate, staircase=False):
""" """
:alias_main: paddle.nn.functional.inverse_time_decay
:alias: paddle.nn.functional.inverse_time_decay,paddle.nn.functional.learning_rate.inverse_time_decay
:old_api: paddle.fluid.layers.inverse_time_decay
Applies inverse time decay to the initial learning rate. Applies inverse time decay to the initial learning rate.
When training a model, it is often recommended to lower the learning rate as the When training a model, it is often recommended to lower the learning rate as the
...@@ -285,6 +302,10 @@ def polynomial_decay(learning_rate, ...@@ -285,6 +302,10 @@ def polynomial_decay(learning_rate,
power=1.0, power=1.0,
cycle=False): cycle=False):
""" """
:alias_main: paddle.nn.functional.polynomial_decay
:alias: paddle.nn.functional.polynomial_decay,paddle.nn.functional.learning_rate.polynomial_decay
:old_api: paddle.fluid.layers.polynomial_decay
2
Applies polynomial decay to the initial learning rate. Applies polynomial decay to the initial learning rate.
.. code-block:: text .. code-block:: text
...@@ -349,7 +370,12 @@ def polynomial_decay(learning_rate, ...@@ -349,7 +370,12 @@ def polynomial_decay(learning_rate,
def piecewise_decay(boundaries, values): def piecewise_decay(boundaries, values):
"""Applies piecewise decay to the initial learning rate. """
:alias_main: paddle.nn.functional.piecewise_decay
:alias: paddle.nn.functional.piecewise_decay,paddle.nn.functional.learning_rate.piecewise_decay
:old_api: paddle.fluid.layers.piecewise_decay
Applies piecewise decay to the initial learning rate.
The algorithm can be described as the code below. The algorithm can be described as the code below.
...@@ -424,6 +450,10 @@ def piecewise_decay(boundaries, values): ...@@ -424,6 +450,10 @@ def piecewise_decay(boundaries, values):
def cosine_decay(learning_rate, step_each_epoch, epochs): def cosine_decay(learning_rate, step_each_epoch, epochs):
""" """
:alias_main: paddle.nn.functional.cosine_decay
:alias: paddle.nn.functional.cosine_decay,paddle.nn.functional.learning_rate.cosine_decay
:old_api: paddle.fluid.layers.cosine_decay
Applies cosine decay to the learning rate. Applies cosine decay to the learning rate.
when training a model, it is often recommended to lower the learning rate as the when training a model, it is often recommended to lower the learning rate as the
...@@ -469,6 +499,10 @@ def cosine_decay(learning_rate, step_each_epoch, epochs): ...@@ -469,6 +499,10 @@ def cosine_decay(learning_rate, step_each_epoch, epochs):
def linear_lr_warmup(learning_rate, warmup_steps, start_lr, end_lr): def linear_lr_warmup(learning_rate, warmup_steps, start_lr, end_lr):
""" """
:alias_main: paddle.nn.functional.linear_lr_warmup
:alias: paddle.nn.functional.linear_lr_warmup,paddle.nn.functional.learning_rate.linear_lr_warmup
:old_api: paddle.fluid.layers.linear_lr_warmup
This operator use the linear learning rate warm up strategy to adjust the learning rate preliminarily before the normal learning rate scheduling. This operator use the linear learning rate warm up strategy to adjust the learning rate preliminarily before the normal learning rate scheduling.
For more information, please refer to `Bag of Tricks for Image Classification with Convolutional Neural Networks <https://arxiv.org/abs/1812.01187>`_ For more information, please refer to `Bag of Tricks for Image Classification with Convolutional Neural Networks <https://arxiv.org/abs/1812.01187>`_
......
python/paddle/fluid/layers/ops.py
浏览文件 @ f5c6dd6d
...@@ -80,6 +80,10 @@ def softshrink(x, alpha=None): ...@@ -80,6 +80,10 @@ def softshrink(x, alpha=None):
softshrink.__doc__ = """ softshrink.__doc__ = """
:alias_main: paddle.nn.functional.softshrink
:alias: paddle.nn.functional.softshrink,paddle.nn.functional.activation.softshrink
:old_api: paddle.fluid.layers.softshrink
:strong:`Softshrink Activation Operator` :strong:`Softshrink Activation Operator`
.. math:: .. math::
...@@ -123,6 +127,10 @@ def hard_shrink(x, threshold=None): ...@@ -123,6 +127,10 @@ def hard_shrink(x, threshold=None):
hard_shrink.__doc__ = _hard_shrink_.__doc__ + """ hard_shrink.__doc__ = _hard_shrink_.__doc__ + """
:alias_main: paddle.nn.functional.hard_shrink
:alias: paddle.nn.functional.hard_shrink,paddle.nn.functional.activation.hard_shrink
:old_api: paddle.fluid.layers.hard_shrink
Examples: Examples:
>>> import paddle.fluid as fluid >>> import paddle.fluid as fluid
...@@ -146,6 +154,10 @@ def cumsum(x, axis=None, exclusive=None, reverse=None): ...@@ -146,6 +154,10 @@ def cumsum(x, axis=None, exclusive=None, reverse=None):
cumsum.__doc__ = """ cumsum.__doc__ = """
:alias_main: paddle.cumsum
:alias: paddle.cumsum,paddle.tensor.cumsum,paddle.tensor.math.cumsum
:old_api: paddle.fluid.layers.cumsum
The cumulative sum of the elements along a given axis. By default, the first element of the result is the same of the first element of the input. If exlusive is true, the first element of the result is 0. The cumulative sum of the elements along a given axis. By default, the first element of the result is the same of the first element of the input. If exlusive is true, the first element of the result is 0.
Args: Args:
...@@ -184,6 +196,10 @@ def thresholded_relu(x, threshold=None): ...@@ -184,6 +196,10 @@ def thresholded_relu(x, threshold=None):
thresholded_relu.__doc__ = """ thresholded_relu.__doc__ = """
:alias_main: paddle.nn.functional.thresholded_relu
:alias: paddle.nn.functional.thresholded_relu,paddle.nn.functional.activation.thresholded_relu
:old_api: paddle.fluid.layers.thresholded_relu
:strong:`Thresholded ReLU Activation Operator` :strong:`Thresholded ReLU Activation Operator`
Equation: Equation:
...@@ -266,6 +282,10 @@ def gelu(x, approximate=False): ...@@ -266,6 +282,10 @@ def gelu(x, approximate=False):
gelu.__doc__ = """ gelu.__doc__ = """
:alias_main: paddle.nn.functional.gelu
:alias: paddle.nn.functional.gelu,paddle.nn.functional.activation.gelu
:old_api: paddle.fluid.layers.gelu
:strong:`GeLU Activation Operator` :strong:`GeLU Activation Operator`
For more details, see [Gaussian Error Linear Units](https://arxiv.org/abs/1606.08415). For more details, see [Gaussian Error Linear Units](https://arxiv.org/abs/1606.08415).
...@@ -350,6 +370,10 @@ def erf(x): ...@@ -350,6 +370,10 @@ def erf(x):
erf.__doc__ = """ erf.__doc__ = """
:alias_main: paddle.erf
:alias: paddle.erf,paddle.tensor.erf,paddle.tensor.math.erf,paddle.nn.functional.erf,paddle.nn.functional.activation.erf
:old_api: paddle.fluid.layers.erf
:strong:`Erf Operator` :strong:`Erf Operator`
For more details, see [Error function](https://en.wikipedia.org/wiki/Error_function). For more details, see [Error function](https://en.wikipedia.org/wiki/Error_function).
......
python/paddle/fluid/layers/rnn.py
浏览文件 @ f5c6dd6d
...@@ -57,6 +57,8 @@ __all__ = [ ...@@ -57,6 +57,8 @@ __all__ = [
class RNNCell(object): class RNNCell(object):
""" """
:api_attr: Static Graph
RNNCell is the base class for abstraction representing the calculations RNNCell is the base class for abstraction representing the calculations
mapping the input and state to the output and new state. It is suitable to mapping the input and state to the output and new state. It is suitable to
and mostly used in RNN. and mostly used in RNN.
...@@ -221,6 +223,8 @@ class RNNCell(object): ...@@ -221,6 +223,8 @@ class RNNCell(object):
class GRUCell(RNNCell): class GRUCell(RNNCell):
""" """
:api_attr: Static Graph
Gated Recurrent Unit cell. It is a wrapper for Gated Recurrent Unit cell. It is a wrapper for
`fluid.contrib.layers.rnn_impl.BasicGRUUnit` to make it adapt to RNNCell. `fluid.contrib.layers.rnn_impl.BasicGRUUnit` to make it adapt to RNNCell.
...@@ -317,6 +321,8 @@ class GRUCell(RNNCell): ...@@ -317,6 +321,8 @@ class GRUCell(RNNCell):
class LSTMCell(RNNCell): class LSTMCell(RNNCell):
""" """
:api_attr: Static Graph
Long-Short Term Memory cell. It is a wrapper for Long-Short Term Memory cell. It is a wrapper for
`fluid.contrib.layers.rnn_impl.BasicLSTMUnit` to make it adapt to RNNCell. `fluid.contrib.layers.rnn_impl.BasicLSTMUnit` to make it adapt to RNNCell.
...@@ -431,6 +437,8 @@ def rnn(cell, ...@@ -431,6 +437,8 @@ def rnn(cell,
is_reverse=False, is_reverse=False,
**kwargs): **kwargs):
""" """
:api_attr: Static Graph
rnn creates a recurrent neural network specified by RNNCell `cell`, rnn creates a recurrent neural network specified by RNNCell `cell`,
which performs :code:`cell.call()` repeatedly until reaches to the maximum which performs :code:`cell.call()` repeatedly until reaches to the maximum
length of `inputs`. length of `inputs`.
...@@ -575,6 +583,8 @@ def rnn(cell, ...@@ -575,6 +583,8 @@ def rnn(cell,
class Decoder(object): class Decoder(object):
""" """
:api_attr: Static Graph
Decoder is the base class for any decoder instance used in `dynamic_decode`. Decoder is the base class for any decoder instance used in `dynamic_decode`.
It provides interface for output generation for one time step, which can be It provides interface for output generation for one time step, which can be
used to generate sequences. used to generate sequences.
...@@ -686,6 +696,8 @@ class Decoder(object): ...@@ -686,6 +696,8 @@ class Decoder(object):
class BeamSearchDecoder(Decoder): class BeamSearchDecoder(Decoder):
""" """
:api_attr: Static Graph
Decoder with beam search decoding strategy. It wraps a cell to get probabilities, Decoder with beam search decoding strategy. It wraps a cell to get probabilities,
and follows a beam search step to calculate scores and select candidate and follows a beam search step to calculate scores and select candidate
token ids for each decoding step. token ids for each decoding step.
...@@ -1153,6 +1165,8 @@ def dynamic_decode(decoder, ...@@ -1153,6 +1165,8 @@ def dynamic_decode(decoder,
return_length=False, return_length=False,
**kwargs): **kwargs):
""" """
:api_attr: Static Graph
Dynamic decoding performs :code:`decoder.step()` repeatedly until the returned Dynamic decoding performs :code:`decoder.step()` repeatedly until the returned
Tensor indicating finished status contains all True values or the number of Tensor indicating finished status contains all True values or the number of
decoding step reaches to :attr:`max_step_num`. decoding step reaches to :attr:`max_step_num`.
...@@ -1975,6 +1989,8 @@ def dynamic_lstm(input, ...@@ -1975,6 +1989,8 @@ def dynamic_lstm(input,
dtype='float32', dtype='float32',
name=None): name=None):
""" """
:api_attr: Static Graph
**Note**: **Note**:
1. This OP only supports LoDTensor as inputs. If you need to deal with Tensor, please use :ref:`api_fluid_layers_lstm` . 1. This OP only supports LoDTensor as inputs. If you need to deal with Tensor, please use :ref:`api_fluid_layers_lstm` .
2. In order to improve efficiency, users must first map the input of dimension [T, hidden_size] to input of [T, 4 * hidden_size], and then pass it to this OP. 2. In order to improve efficiency, users must first map the input of dimension [T, hidden_size] to input of [T, 4 * hidden_size], and then pass it to this OP.
...@@ -2145,6 +2161,8 @@ def lstm(input, ...@@ -2145,6 +2161,8 @@ def lstm(input,
default_initializer=None, default_initializer=None,
seed=-1): seed=-1):
""" """
:api_attr: Static Graph
**Note**: **Note**:
This OP only supports running on GPU devices. This OP only supports running on GPU devices.
...@@ -2330,6 +2348,8 @@ def dynamic_lstmp(input, ...@@ -2330,6 +2348,8 @@ def dynamic_lstmp(input,
cell_clip=None, cell_clip=None,
proj_clip=None): proj_clip=None):
""" """
:api_attr: Static Graph
**Note**: **Note**:
1. In order to improve efficiency, users must first map the input of dimension [T, hidden_size] to input of [T, 4 * hidden_size], and then pass it to this OP. 1. In order to improve efficiency, users must first map the input of dimension [T, hidden_size] to input of [T, 4 * hidden_size], and then pass it to this OP.
...@@ -2539,6 +2559,8 @@ def dynamic_gru(input, ...@@ -2539,6 +2559,8 @@ def dynamic_gru(input,
h_0=None, h_0=None,
origin_mode=False): origin_mode=False):
""" """
:api_attr: Static Graph
**Note: The input type of this must be LoDTensor. If the input type to be **Note: The input type of this must be LoDTensor. If the input type to be
processed is Tensor, use** :ref:`api_fluid_layers_StaticRNN` . processed is Tensor, use** :ref:`api_fluid_layers_StaticRNN` .
...@@ -2691,6 +2713,8 @@ def gru_unit(input, ...@@ -2691,6 +2713,8 @@ def gru_unit(input,
gate_activation='sigmoid', gate_activation='sigmoid',
origin_mode=False): origin_mode=False):
""" """
:api_attr: Static Graph
Gated Recurrent Unit (GRU) RNN cell. This operator performs GRU calculations for Gated Recurrent Unit (GRU) RNN cell. This operator performs GRU calculations for
one time step and it supports these two modes: one time step and it supports these two modes:
...@@ -2847,6 +2871,10 @@ def beam_search(pre_ids, ...@@ -2847,6 +2871,10 @@ def beam_search(pre_ids,
name=None, name=None,
return_parent_idx=False): return_parent_idx=False):
""" """
:alias_main: paddle.nn.beam_search
:alias: paddle.nn.beam_search,paddle.nn.decode.beam_search
:old_api: paddle.fluid.layers.beam_search
Beam search is a classical algorithm for selecting candidate words in a Beam search is a classical algorithm for selecting candidate words in a
machine translation task. machine translation task.
...@@ -2988,6 +3016,10 @@ def beam_search(pre_ids, ...@@ -2988,6 +3016,10 @@ def beam_search(pre_ids,
def beam_search_decode(ids, scores, beam_size, end_id, name=None): def beam_search_decode(ids, scores, beam_size, end_id, name=None):
""" """
:alias_main: paddle.nn.beam_search_decode
:alias: paddle.nn.beam_search_decode,paddle.nn.decode.beam_search_decode
:old_api: paddle.fluid.layers.beam_search_decode
This operator is used after beam search has completed. It constructs the This operator is used after beam search has completed. It constructs the
full predicted sequences for each sample by walking back along the search full predicted sequences for each sample by walking back along the search
paths stored in lod of ``ids`` . The result sequences are stored in a paths stored in lod of ``ids`` . The result sequences are stored in a
...@@ -3067,6 +3099,8 @@ def lstm_unit(x_t, ...@@ -3067,6 +3099,8 @@ def lstm_unit(x_t,
bias_attr=None, bias_attr=None,
name=None): name=None):
""" """
:api_attr: Static Graph
Long-Short Term Memory (LSTM) RNN cell. This operator performs LSTM calculations for Long-Short Term Memory (LSTM) RNN cell. This operator performs LSTM calculations for
one time step, whose implementation is based on calculations described in `RECURRENT one time step, whose implementation is based on calculations described in `RECURRENT
NEURAL NETWORK REGULARIZATION <http://arxiv.org/abs/1409.2329>`_ . NEURAL NETWORK REGULARIZATION <http://arxiv.org/abs/1409.2329>`_ .
......
python/paddle/fluid/layers/sequence_lod.py
浏览文件 @ f5c6dd6d
...@@ -52,6 +52,8 @@ def sequence_conv(input, ...@@ -52,6 +52,8 @@ def sequence_conv(input,
act=None, act=None,
name=None): name=None):
""" """
:api_attr: Static Graph
**Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use conv2d Op.(fluid.layers.** :ref:`api_fluid_layers_conv2d` ). **Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use conv2d Op.(fluid.layers.** :ref:`api_fluid_layers_conv2d` ).
This operator receives input sequences with variable length and other convolutional This operator receives input sequences with variable length and other convolutional
...@@ -174,6 +176,8 @@ def sequence_conv(input, ...@@ -174,6 +176,8 @@ def sequence_conv(input,
def sequence_softmax(input, use_cudnn=False, name=None): def sequence_softmax(input, use_cudnn=False, name=None):
""" """
:api_attr: Static Graph
**Note**: **Note**:
**The input type of the OP must be LoDTensor. For Tensor, use:** :ref:`api_fluid_layers_softmax` **The input type of the OP must be LoDTensor. For Tensor, use:** :ref:`api_fluid_layers_softmax`
...@@ -256,6 +260,8 @@ def sequence_softmax(input, use_cudnn=False, name=None): ...@@ -256,6 +260,8 @@ def sequence_softmax(input, use_cudnn=False, name=None):
def sequence_pool(input, pool_type, is_test=False, pad_value=0.0): def sequence_pool(input, pool_type, is_test=False, pad_value=0.0):
""" """
:api_attr: Static Graph
**Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use pool2d Op.(fluid.layers.** :ref:`api_fluid_layers_pool2d` ). **Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use pool2d Op.(fluid.layers.** :ref:`api_fluid_layers_pool2d` ).
This operator only supports LoDTensor as input. It will apply specified pooling This operator only supports LoDTensor as input. It will apply specified pooling
...@@ -368,6 +374,8 @@ def sequence_pool(input, pool_type, is_test=False, pad_value=0.0): ...@@ -368,6 +374,8 @@ def sequence_pool(input, pool_type, is_test=False, pad_value=0.0):
@templatedoc() @templatedoc()
def sequence_concat(input, name=None): def sequence_concat(input, name=None):
""" """
:api_attr: Static Graph
**Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use concat Op.(fluid.layers.** :ref:`api_fluid_layers_concat` ). **Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use concat Op.(fluid.layers.** :ref:`api_fluid_layers_concat` ).
This operator only supports LoDTensor as input. It concatenates the multiple LoDTensor from input by the LoD information, This operator only supports LoDTensor as input. It concatenates the multiple LoDTensor from input by the LoD information,
...@@ -427,6 +435,8 @@ def sequence_concat(input, name=None): ...@@ -427,6 +435,8 @@ def sequence_concat(input, name=None):
def sequence_first_step(input): def sequence_first_step(input):
""" """
:api_attr: Static Graph
This operator only supports LoDTensor as input. Given the input LoDTensor, it will This operator only supports LoDTensor as input. Given the input LoDTensor, it will
select first time-step feature of each sequence as output. select first time-step feature of each sequence as output.
...@@ -479,6 +489,8 @@ def sequence_first_step(input): ...@@ -479,6 +489,8 @@ def sequence_first_step(input):
def sequence_last_step(input): def sequence_last_step(input):
""" """
:api_attr: Static Graph
This operator only supports LoDTensor as input. Given the input LoDTensor, it will This operator only supports LoDTensor as input. Given the input LoDTensor, it will
select last time-step feature of each sequence as output. select last time-step feature of each sequence as output.
...@@ -532,6 +544,8 @@ def sequence_last_step(input): ...@@ -532,6 +544,8 @@ def sequence_last_step(input):
def sequence_slice(input, offset, length, name=None): def sequence_slice(input, offset, length, name=None):
""" """
:api_attr: Static Graph
**Sequence Slice Layer** **Sequence Slice Layer**
The layer crops a subsequence from given sequence with given start The layer crops a subsequence from given sequence with given start
...@@ -617,7 +631,10 @@ def sequence_slice(input, offset, length, name=None): ...@@ -617,7 +631,10 @@ def sequence_slice(input, offset, length, name=None):
def sequence_expand(x, y, ref_level=-1, name=None): def sequence_expand(x, y, ref_level=-1, name=None):
"""Sequence Expand Layer. This layer will expand the input variable ``x`` \ """
:api_attr: Static Graph
Sequence Expand Layer. This layer will expand the input variable ``x`` \
according to specified level ``ref_level`` lod of ``y``. Please note that \ according to specified level ``ref_level`` lod of ``y``. Please note that \
the lod level of ``x`` is at most 1. If the lod level of ``x`` is 1, than \ the lod level of ``x`` is at most 1. If the lod level of ``x`` is 1, than \
the size of lod of ``x`` must be equal to the length of ``ref_level`` lod \ the size of lod of ``x`` must be equal to the length of ``ref_level`` lod \
...@@ -750,7 +767,10 @@ def sequence_expand(x, y, ref_level=-1, name=None): ...@@ -750,7 +767,10 @@ def sequence_expand(x, y, ref_level=-1, name=None):
def sequence_expand_as(x, y, name=None): def sequence_expand_as(x, y, name=None):
"""Sequence Expand As Layer. This OP will expand the input variable ``x`` \ """
:api_attr: Static Graph
Sequence Expand As Layer. This OP will expand the input variable ``x`` \
according to the zeroth level lod of ``y``. Current implementation requires \ according to the zeroth level lod of ``y``. Current implementation requires \
the level number of ``y``'s lod must be 1, and the first dimension of \ the level number of ``y``'s lod must be 1, and the first dimension of \
``x`` should be equal to the size of ``y``'s zeroth level lod, thus \ ``x`` should be equal to the size of ``y``'s zeroth level lod, thus \
...@@ -865,6 +885,8 @@ def sequence_expand_as(x, y, name=None): ...@@ -865,6 +885,8 @@ def sequence_expand_as(x, y, name=None):
def sequence_pad(x, pad_value, maxlen=None, name=None): def sequence_pad(x, pad_value, maxlen=None, name=None):
""" """
:api_attr: Static Graph
This layer padding the sequences in a same batch to a common length (according \ This layer padding the sequences in a same batch to a common length (according \
to ``maxlen``). The padding value is defined by ``pad_value``, and will be \ to ``maxlen``). The padding value is defined by ``pad_value``, and will be \
appended to the tail of sequences. The result is a Python tuple ``(Out, Length)``: \ appended to the tail of sequences. The result is a Python tuple ``(Out, Length)``: \
...@@ -977,6 +999,8 @@ def sequence_pad(x, pad_value, maxlen=None, name=None): ...@@ -977,6 +999,8 @@ def sequence_pad(x, pad_value, maxlen=None, name=None):
def sequence_unpad(x, length, name=None): def sequence_unpad(x, length, name=None):
""" """
:api_attr: Static Graph
**Note**: **Note**:
**The input of the OP is Tensor and the output is LoDTensor. For padding operation, See:** :ref:`api_fluid_layers_sequence_pad` **The input of the OP is Tensor and the output is LoDTensor. For padding operation, See:** :ref:`api_fluid_layers_sequence_pad`
...@@ -1050,6 +1074,8 @@ def sequence_unpad(x, length, name=None): ...@@ -1050,6 +1074,8 @@ def sequence_unpad(x, length, name=None):
def sequence_reshape(input, new_dim): def sequence_reshape(input, new_dim):
""" """
:api_attr: Static Graph
**Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use reshape Op.(fluid.layers.** :ref:`api_fluid_layers_reshape` ). **Notes: The Op only receives LoDTensor as input. If your input is Tensor, please use reshape Op.(fluid.layers.** :ref:`api_fluid_layers_reshape` ).
This operator only supports LoDTensor as input. Given :attr:`new_dim` , This operator only supports LoDTensor as input. Given :attr:`new_dim` ,
...@@ -1110,6 +1136,8 @@ def sequence_reshape(input, new_dim): ...@@ -1110,6 +1136,8 @@ def sequence_reshape(input, new_dim):
def sequence_scatter(input, index, updates, name=None): def sequence_scatter(input, index, updates, name=None):
""" """
:api_attr: Static Graph
**Note**: **Note**:
**The index and updates parameters of the OP must be LoDTensor.** **The index and updates parameters of the OP must be LoDTensor.**
...@@ -1198,6 +1226,8 @@ def sequence_scatter(input, index, updates, name=None): ...@@ -1198,6 +1226,8 @@ def sequence_scatter(input, index, updates, name=None):
def sequence_enumerate(input, win_size, pad_value=0, name=None): def sequence_enumerate(input, win_size, pad_value=0, name=None):
""" """
:api_attr: Static Graph
Generate a new sequence for the input index sequence with \ Generate a new sequence for the input index sequence with \
shape ``[d_1, win_size]``, which enumerates all the \ shape ``[d_1, win_size]``, which enumerates all the \
sub-sequences with length ``win_size`` of the input with \ sub-sequences with length ``win_size`` of the input with \
......
python/paddle/fluid/layers/tensor.py
浏览文件 @ f5c6dd6d
...@@ -75,6 +75,8 @@ def create_parameter(shape, ...@@ -75,6 +75,8 @@ def create_parameter(shape,
is_bias=False, is_bias=False,
default_initializer=None): default_initializer=None):
""" """
:api_attr: Static Graph
This function creates a parameter. The parameter is a learnable variable, which can have This function creates a parameter. The parameter is a learnable variable, which can have
gradient, and can be optimized. gradient, and can be optimized.
...@@ -195,6 +197,10 @@ def create_global_var(shape, ...@@ -195,6 +197,10 @@ def create_global_var(shape,
def cast(x, dtype): def cast(x, dtype):
""" """
:alias_main: paddle.cast
:alias: paddle.cast,paddle.tensor.cast,paddle.tensor.manipulation.cast
:old_api: paddle.fluid.layers.cast
This OP takes in the Variable :attr:`x` with :attr:`x.dtype` and casts it This OP takes in the Variable :attr:`x` with :attr:`x.dtype` and casts it
to the output with :attr:`dtype`. It's meaningless if the output dtype to the output with :attr:`dtype`. It's meaningless if the output dtype
equals the input dtype, but it's fine if you do so. equals the input dtype, but it's fine if you do so.
...@@ -257,6 +263,10 @@ def cast(x, dtype): ...@@ -257,6 +263,10 @@ def cast(x, dtype):
def concat(input, axis=0, name=None): def concat(input, axis=0, name=None):
""" """
:alias_main: paddle.concat
:alias: paddle.concat,paddle.tensor.concat,paddle.tensor.manipulation.concat
:old_api: paddle.fluid.layers.concat
**Concat** **Concat**
This OP concatenates the input along the axis. This OP concatenates the input along the axis.
...@@ -535,6 +545,10 @@ def sums(input, out=None): ...@@ -535,6 +545,10 @@ def sums(input, out=None):
def assign(input, output=None): def assign(input, output=None):
""" """
:alias_main: paddle.nn.functional.assign
:alias: paddle.nn.functional.assign,paddle.nn.functional.common.assign
:old_api: paddle.fluid.layers.assign
The OP copies the :attr:`input` to the :attr:`output`. The OP copies the :attr:`input` to the :attr:`output`.
Parameters: Parameters:
...@@ -607,6 +621,10 @@ def assign(input, output=None): ...@@ -607,6 +621,10 @@ def assign(input, output=None):
def fill_constant(shape, dtype, value, force_cpu=False, out=None): def fill_constant(shape, dtype, value, force_cpu=False, out=None):
""" """
:alias_main: paddle.fill_constant
:alias: paddle.fill_constant,paddle.tensor.fill_constant,paddle.tensor.creation.fill_constant
:old_api: paddle.fluid.layers.fill_constant
This OP creates a Tensor with specified `shape` and `dtype`, and This OP creates a Tensor with specified `shape` and `dtype`, and
initializes it with a constant specified by `value`. initializes it with a constant specified by `value`.
...@@ -787,6 +805,10 @@ def fill_constant_batch_size_like(input, ...@@ -787,6 +805,10 @@ def fill_constant_batch_size_like(input,
def argmin(x, axis=0): def argmin(x, axis=0):
""" """
:alias_main: paddle.argmin
:alias: paddle.argmin,paddle.tensor.argmin,paddle.tensor.search.argmin
:old_api: paddle.fluid.layers.argmin
**argmin** **argmin**
This OP computes the indices of the min elements of the input tensor's This OP computes the indices of the min elements of the input tensor's
...@@ -913,6 +935,10 @@ def argmax(x, axis=0): ...@@ -913,6 +935,10 @@ def argmax(x, axis=0):
def argsort(input, axis=-1, descending=False, name=None): def argsort(input, axis=-1, descending=False, name=None):
""" """
:alias_main: paddle.argsort
:alias: paddle.argsort,paddle.tensor.argsort,paddle.tensor.search.argsort
:old_api: paddle.fluid.layers.argsort
This OP sorts the input along the given axis, and returns sorted output This OP sorts the input along the given axis, and returns sorted output
data Varibale and its corresponding index Variable with the same shape as data Varibale and its corresponding index Variable with the same shape as
:attr:`input`. :attr:`input`.
...@@ -1061,6 +1087,10 @@ def zeros(shape, dtype, force_cpu=False): ...@@ -1061,6 +1087,10 @@ def zeros(shape, dtype, force_cpu=False):
def reverse(x, axis): def reverse(x, axis):
""" """
:alias_main: paddle.reverse
:alias: paddle.reverse,paddle.tensor.reverse,paddle.tensor.manipulation.reverse
:old_api: paddle.fluid.layers.reverse
The OP reverses the tensor :attr:`x` along the given :attr:`axis`. The OP reverses the tensor :attr:`x` along the given :attr:`axis`.
Parameters: Parameters:
...@@ -1171,6 +1201,10 @@ def load_combine(out, file_path): ...@@ -1171,6 +1201,10 @@ def load_combine(out, file_path):
def has_inf(x): def has_inf(x):
""" """
:alias_main: paddle.has_inf
:alias: paddle.has_inf,paddle.tensor.has_inf,paddle.tensor.search.has_inf
:old_api: paddle.fluid.layers.has_inf
Test if any of x contains an infinity number Test if any of x contains an infinity number
Args: Args:
...@@ -1196,6 +1230,10 @@ def has_inf(x): ...@@ -1196,6 +1230,10 @@ def has_inf(x):
def has_nan(x): def has_nan(x):
""" """
:alias_main: paddle.has_nan
:alias: paddle.has_nan,paddle.tensor.has_nan,paddle.tensor.search.has_nan
:old_api: paddle.fluid.layers.has_nan
Test if any of x contains a NAN Test if any of x contains a NAN
Args: Args:
...@@ -1221,6 +1259,10 @@ def has_nan(x): ...@@ -1221,6 +1259,10 @@ def has_nan(x):
def isfinite(x): def isfinite(x):
""" """
:alias_main: paddle.isfinite
:alias: paddle.isfinite,paddle.tensor.isfinite,paddle.tensor.logic.isfinite
:old_api: paddle.fluid.layers.isfinite
Test if any of x contains an infinity/NAN number. If all the elements are finite, Test if any of x contains an infinity/NAN number. If all the elements are finite,
returns true, else false. returns true, else false.
...@@ -1418,6 +1460,10 @@ def zeros_like(x, out=None): ...@@ -1418,6 +1460,10 @@ def zeros_like(x, out=None):
def diag(diagonal): def diag(diagonal):
""" """
:alias_main: paddle.diag
:alias: paddle.diag,paddle.tensor.diag,paddle.tensor.creation.diag
:old_api: paddle.fluid.layers.diag
This OP creates a square matrix which has diagonal values specified by input :attr:`diagonal`. This OP creates a square matrix which has diagonal values specified by input :attr:`diagonal`.
Args: Args:
...@@ -1461,6 +1507,10 @@ def diag(diagonal): ...@@ -1461,6 +1507,10 @@ def diag(diagonal):
def eye(num_rows, num_columns=None, batch_shape=None, dtype='float32'): def eye(num_rows, num_columns=None, batch_shape=None, dtype='float32'):
""" """
:alias_main: paddle.eye
:alias: paddle.eye,paddle.tensor.eye,paddle.tensor.creation.eye
:old_api: paddle.fluid.layers.eye
**eye** **eye**
This function constructs an identity tensor, or a batch of tensor. This function constructs an identity tensor, or a batch of tensor.
......
python/paddle/fluid/layers/utils.py
浏览文件 @ f5c6dd6d
...@@ -112,6 +112,10 @@ def _yield_flat_nest(nest): ...@@ -112,6 +112,10 @@ def _yield_flat_nest(nest):
def flatten(nest): def flatten(nest):
""" """
:alias_main: paddle.flatten
:alias: paddle.flatten,paddle.tensor.flatten,paddle.tensor.manipulation.flatten
:old_api: paddle.fluid.layers.flatten
Traverse all entries in the nested structure and put them into an list. Traverse all entries in the nested structure and put them into an list.
""" """
if is_sequence(nest): if is_sequence(nest):
......
python/paddle/fluid/lod_tensor.py
浏览文件 @ f5c6dd6d
...@@ -114,6 +114,8 @@ def create_lod_tensor(data, recursive_seq_lens, place): ...@@ -114,6 +114,8 @@ def create_lod_tensor(data, recursive_seq_lens, place):
def create_random_int_lodtensor(recursive_seq_lens, base_shape, place, low, def create_random_int_lodtensor(recursive_seq_lens, base_shape, place, low,
high): high):
""" """
:api_attr: Static Graph
Create a LoDTensor containing random integers. Create a LoDTensor containing random integers.
The implementation is as follows: The implementation is as follows:
......
python/paddle/fluid/nets.py
浏览文件 @ f5c6dd6d
...@@ -43,6 +43,8 @@ def simple_img_conv_pool(input, ...@@ -43,6 +43,8 @@ def simple_img_conv_pool(input,
act=None, act=None,
use_cudnn=True): use_cudnn=True):
""" """
:api_attr: Static Graph
The simple_img_conv_pool api is composed of :ref:`api_fluid_layers_conv2d` and :ref:`api_fluid_layers_pool2d` . The simple_img_conv_pool api is composed of :ref:`api_fluid_layers_conv2d` and :ref:`api_fluid_layers_pool2d` .
Args: Args:
...@@ -149,6 +151,8 @@ def img_conv_group(input, ...@@ -149,6 +151,8 @@ def img_conv_group(input,
pool_type="max", pool_type="max",
use_cudnn=True): use_cudnn=True):
""" """
:api_attr: Static Graph
The Image Convolution Group is composed of Convolution2d, BatchNorm, DropOut, The Image Convolution Group is composed of Convolution2d, BatchNorm, DropOut,
and Pool2d. According to the input arguments, img_conv_group will do serials of and Pool2d. According to the input arguments, img_conv_group will do serials of
computation for Input using Convolution2d, BatchNorm, DropOut, and pass the last computation for Input using Convolution2d, BatchNorm, DropOut, and pass the last
...@@ -257,6 +261,8 @@ def sequence_conv_pool(input, ...@@ -257,6 +261,8 @@ def sequence_conv_pool(input,
pool_type="max", pool_type="max",
bias_attr=None): bias_attr=None):
""" """
:api_attr: Static Graph
**This api takes input as an LoDTensor. If input is a Tensor, please use** **This api takes input as an LoDTensor. If input is a Tensor, please use**
:ref:`api_fluid_nets_simple_img_conv_pool` **instead** :ref:`api_fluid_nets_simple_img_conv_pool` **instead**
...@@ -321,6 +327,8 @@ def sequence_conv_pool(input, ...@@ -321,6 +327,8 @@ def sequence_conv_pool(input,
def glu(input, dim=-1): def glu(input, dim=-1):
""" """
:api_attr: Static Graph
The Gated Linear Units(GLU) composed by :ref:`api_fluid_layers_split` , The Gated Linear Units(GLU) composed by :ref:`api_fluid_layers_split` ,
:ref:`api_fluid_layers_sigmoid` and :ref:`api_fluid_layers_elementwise_mul` . :ref:`api_fluid_layers_sigmoid` and :ref:`api_fluid_layers_elementwise_mul` .
Specifically, GLU will plit the input into two equal-sized parts, Specifically, GLU will plit the input into two equal-sized parts,
...@@ -367,6 +375,8 @@ def scaled_dot_product_attention(queries, ...@@ -367,6 +375,8 @@ def scaled_dot_product_attention(queries,
num_heads=1, num_heads=1,
dropout_rate=0.): dropout_rate=0.):
""" """
:api_attr: Static Graph
This interface Multi-Head Attention using scaled dot product. This interface Multi-Head Attention using scaled dot product.
Attention mechanism can be seen as mapping a query and a set of key-value Attention mechanism can be seen as mapping a query and a set of key-value
pairs to an output. Multi-Head Attention performs attention using multi-head pairs to an output. Multi-Head Attention performs attention using multi-head
......
python/paddle/fluid/optimizer.py
浏览文件 @ f5c6dd6d
...@@ -1070,6 +1070,8 @@ class MomentumOptimizer(Optimizer): ...@@ -1070,6 +1070,8 @@ class MomentumOptimizer(Optimizer):
class DGCMomentumOptimizer(Optimizer): class DGCMomentumOptimizer(Optimizer):
""" """
:api_attr: Static Graph
DGC (Deep Gradient Compression) Momentum Optimizer. Original paper is https://arxiv.org/abs/1712.01887 DGC (Deep Gradient Compression) Momentum Optimizer. Original paper is https://arxiv.org/abs/1712.01887
DGC reduces the communication bandwidth by sending only the important gradients (sparse update):\ DGC reduces the communication bandwidth by sending only the important gradients (sparse update):\
...@@ -2994,6 +2996,8 @@ Lamb = LambOptimizer ...@@ -2994,6 +2996,8 @@ Lamb = LambOptimizer
class ModelAverage(Optimizer): class ModelAverage(Optimizer):
""" """
:api_attr: Static Graph
The ModelAverage optimizer accumulates specific continuous historical parameters The ModelAverage optimizer accumulates specific continuous historical parameters
during training. The accumulated historical range can be controlled by the passed during training. The accumulated historical range can be controlled by the passed
``average_window_rate`` argument. The averaged ``Parameter`` are used in the prediction, ``average_window_rate`` argument. The averaged ``Parameter`` are used in the prediction,
...@@ -3301,6 +3305,8 @@ class ModelAverage(Optimizer): ...@@ -3301,6 +3305,8 @@ class ModelAverage(Optimizer):
class ExponentialMovingAverage(object): class ExponentialMovingAverage(object):
""" """
:api_attr: Static Graph
Compute the moving average of parameters with exponential decay. Compute the moving average of parameters with exponential decay.
Given a parameter :math:`\\theta`, its exponential moving average (EMA) Given a parameter :math:`\\theta`, its exponential moving average (EMA)
will be will be
...@@ -3549,6 +3555,8 @@ class ExponentialMovingAverage(object): ...@@ -3549,6 +3555,8 @@ class ExponentialMovingAverage(object):
class PipelineOptimizer(object): class PipelineOptimizer(object):
""" """
:api_attr: Static Graph
Pipeline Optimizer Pipeline Optimizer
Train with pipeline mode. The program will be split by cut_list. Train with pipeline mode. The program will be split by cut_list.
...@@ -3849,6 +3857,8 @@ class PipelineOptimizer(object): ...@@ -3849,6 +3857,8 @@ class PipelineOptimizer(object):
class RecomputeOptimizer(Optimizer): class RecomputeOptimizer(Optimizer):
""" """
:api_attr: Static Graph
Recompute Optimizer Wrapper Recompute Optimizer Wrapper
Normally, a training step contains three sub-steps: first, run forward Normally, a training step contains three sub-steps: first, run forward
...@@ -3921,6 +3931,8 @@ class RecomputeOptimizer(Optimizer): ...@@ -3921,6 +3931,8 @@ class RecomputeOptimizer(Optimizer):
def load(self, stat_dict): def load(self, stat_dict):
""" """
:api_attr: Static Graph
load function is not supported by Recompute Optimizer for now. load function is not supported by Recompute Optimizer for now.
:return: None :return: None
...@@ -4137,6 +4149,8 @@ class RecomputeOptimizer(Optimizer): ...@@ -4137,6 +4149,8 @@ class RecomputeOptimizer(Optimizer):
class LookaheadOptimizer(object): class LookaheadOptimizer(object):
""" """
:api_attr: Static Graph
This implements the Lookahead optimizer of the This implements the Lookahead optimizer of the
paper : https://arxiv.org/abs/1907.08610. paper : https://arxiv.org/abs/1907.08610.
......
python/paddle/fluid/parallel_executor.py
浏览文件 @ f5c6dd6d
...@@ -28,6 +28,8 @@ BuildStrategy = core.ParallelExecutor.BuildStrategy ...@@ -28,6 +28,8 @@ BuildStrategy = core.ParallelExecutor.BuildStrategy
class ParallelExecutor(object): class ParallelExecutor(object):
""" """
:api_attr: Static Graph
The ParallelExecutor is an upgraded version of :code:`fluid.Executor` that supports multi-node model The ParallelExecutor is an upgraded version of :code:`fluid.Executor` that supports multi-node model
training and testing based on the data-parallel mode. In data-parallel mode, training and testing based on the data-parallel mode. In data-parallel mode,
ParallelExecutor will broadcast the parameters from Node0 to other nodes during ParallelExecutor will broadcast the parameters from Node0 to other nodes during
......
python/paddle/fluid/param_attr.py
浏览文件 @ f5c6dd6d
...@@ -202,6 +202,8 @@ class ParamAttr(object): ...@@ -202,6 +202,8 @@ class ParamAttr(object):
class WeightNormParamAttr(ParamAttr): class WeightNormParamAttr(ParamAttr):
""" """
:api_attr: Static Graph
Parameter of weight Norm. Weight Norm is a reparameterization of the weight vectors Parameter of weight Norm. Weight Norm is a reparameterization of the weight vectors
in a neural network that decouples the magnitude of those weight vectors from in a neural network that decouples the magnitude of those weight vectors from
their direction. Weight Norm has been implemented as discussed in this their direction. Weight Norm has been implemented as discussed in this
......
python/paddle/fluid/transpiler/distribute_transpiler.py
浏览文件 @ f5c6dd6d
...@@ -140,6 +140,8 @@ def slice_variable(var_list, slice_count, min_block_size): ...@@ -140,6 +140,8 @@ def slice_variable(var_list, slice_count, min_block_size):
class DistributeTranspilerConfig(object): class DistributeTranspilerConfig(object):
""" """
:api_attr: Static Graph
A configuration class that provide support for transpiler distributed jobs. A configuration class that provide support for transpiler distributed jobs.
Some important parameters are explained as follows: Some important parameters are explained as follows:
...@@ -253,6 +255,8 @@ class ServerRuntimeConfig(object): ...@@ -253,6 +255,8 @@ class ServerRuntimeConfig(object):
class DistributeTranspiler(object): class DistributeTranspiler(object):
""" """
:api_attr: Static Graph
**DistributeTranspiler** **DistributeTranspiler**
Convert the fluid program to distributed data-parallelism programs. Convert the fluid program to distributed data-parallelism programs.
......
python/paddle/fluid/transpiler/memory_optimization_transpiler.py
浏览文件 @ f5c6dd6d
...@@ -21,6 +21,8 @@ def memory_optimize(input_program, ...@@ -21,6 +21,8 @@ def memory_optimize(input_program,
level=0, level=0,
skip_grads=True): skip_grads=True):
""" """
:api_attr: Static Graph
This API is deprecated since 1.6. Please do not use it. The better This API is deprecated since 1.6. Please do not use it. The better
memory optimization strategies are enabled by default. memory optimization strategies are enabled by default.
""" """
...@@ -41,6 +43,8 @@ def memory_optimize(input_program, ...@@ -41,6 +43,8 @@ def memory_optimize(input_program,
def release_memory(input_program, skip_opt_set=None): def release_memory(input_program, skip_opt_set=None):
""" """
:api_attr: Static Graph
This API is deprecated since 1.6. Please do not use it. The better This API is deprecated since 1.6. Please do not use it. The better
memory optimization strategies are enabled by default. memory optimization strategies are enabled by default.
""" """
......
python/paddle/fluid/transpiler/ps_dispatcher.py
浏览文件 @ f5c6dd6d
...@@ -48,6 +48,8 @@ class PSDispatcher(object): ...@@ -48,6 +48,8 @@ class PSDispatcher(object):
class HashName(PSDispatcher): class HashName(PSDispatcher):
""" """
:api_attr: Static Graph
Hash variable names to several endpoints using python Hash variable names to several endpoints using python
"hash()" function. "hash()" function.
...@@ -88,6 +90,8 @@ class HashName(PSDispatcher): ...@@ -88,6 +90,8 @@ class HashName(PSDispatcher):
class RoundRobin(PSDispatcher): class RoundRobin(PSDispatcher):
""" """
:api_attr: Static Graph
Distribute variables to several endpoints using Distribute variables to several endpoints using
RondRobin<https://en.wikipedia.org/wiki/Round-robin_scheduling> method. RondRobin<https://en.wikipedia.org/wiki/Round-robin_scheduling> method.
......
python/paddle/framework/random.py
浏览文件 @ f5c6dd6d
...@@ -20,6 +20,9 @@ __all__ = ['manual_seed'] ...@@ -20,6 +20,9 @@ __all__ = ['manual_seed']
def manual_seed(seed): def manual_seed(seed):
""" """
:alias_main: paddle.manual_seed
:alias: paddle.manual_seed,paddle.framework.random.manual_seed
Set global manual seed for program Set global manual seed for program
Args: Args:
......
python/paddle/nn/functional/activation.py
浏览文件 @ f5c6dd6d
...@@ -78,6 +78,9 @@ def hsigmoid(input, ...@@ -78,6 +78,9 @@ def hsigmoid(input,
path_code=None, path_code=None,
is_sparse=False): is_sparse=False):
""" """
:alias_main: paddle.nn.functional.hsigmoid
:alias: paddle.nn.functional.hsigmoid,paddle.nn.functional.activation.hsigmoid
The hierarchical sigmoid organizes the classes into a complete binary tree to reduce the computational complexity 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. 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. Each leaf node of the complete binary tree represents a class(word) and each non-leaf node acts as a binary classifier.
...@@ -191,6 +194,9 @@ def hsigmoid(input, ...@@ -191,6 +194,9 @@ def hsigmoid(input,
def relu(input, inplace=False, name=None): def relu(input, inplace=False, name=None):
""" """
:alias_main: paddle.nn.functional.relu
:alias: paddle.nn.functional.relu,paddle.nn.functional.activation.relu
ReLU Activation. ReLU Activation.
.. math: .. math:
...@@ -240,6 +246,9 @@ def relu(input, inplace=False, name=None): ...@@ -240,6 +246,9 @@ def relu(input, inplace=False, name=None):
def sigmoid(input, inplace=False, name=None): def sigmoid(input, inplace=False, name=None):
""" """
:alias_main: paddle.nn.functional.sigmoid
:alias: paddle.nn.functional.sigmoid,paddle.nn.functional.activation.sigmoid
Sigmoid Activation. Sigmoid Activation.
.. math: .. math:
...@@ -298,6 +307,9 @@ def sigmoid(input, inplace=False, name=None): ...@@ -298,6 +307,9 @@ def sigmoid(input, inplace=False, name=None):
def log_softmax(input, axis=None, dtype=None, name=None): def log_softmax(input, axis=None, dtype=None, name=None):
""" """
:alias_main: paddle.nn.functional.log_softmax
:alias: paddle.nn.functional.log_softmax,paddle.nn.functional.activation.log_softmax
This operator implements the log_softmax layer. The calculation process is as follows: This operator implements the log_softmax layer. The calculation process is as follows:
.. math:: .. math::
......
python/paddle/nn/functional/common.py
浏览文件 @ f5c6dd6d
...@@ -51,6 +51,9 @@ def interpolate(input, ...@@ -51,6 +51,9 @@ def interpolate(input,
align_mode=1, align_mode=1,
data_format='NCHW'): data_format='NCHW'):
""" """
:alias_main: paddle.nn.functional.interpolate
:alias: paddle.nn.functional.interpolate,paddle.nn.functional.common.interpolate
This op resizes a batch of images. This op resizes a batch of images.
The input must be a 4-D Tensor of the shape (num_batches, channels, in_h, in_w) The input must be a 4-D Tensor of the shape (num_batches, channels, in_h, in_w)
or (num_batches, in_h, in_w, channels), or a 5-D Tensor of the shape or (num_batches, in_h, in_w, channels), or a 5-D Tensor of the shape
......
python/paddle/nn/functional/conv.py
浏览文件 @ f5c6dd6d
...@@ -99,6 +99,9 @@ def conv2d(input, ...@@ -99,6 +99,9 @@ def conv2d(input,
data_format="NCHW", data_format="NCHW",
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.conv2d
:alias: paddle.nn.functional.conv2d,paddle.nn.functional.conv.conv2d
The convolution2D layer calculates the output based on the input, filter The convolution2D layer calculates the output based on the input, filter
and strides, paddings, dilations, groups parameters. Input and 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 Output are in NCHW or NHWC format, where N is batch size, C is the number of
...@@ -333,6 +336,9 @@ def conv2d_transpose(input, ...@@ -333,6 +336,9 @@ def conv2d_transpose(input,
data_format='NCHW', data_format='NCHW',
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.conv2d_transpose
:alias: paddle.nn.functional.conv2d_transpose,paddle.nn.functional.conv.conv2d_transpose
The convolution2D transpose layer calculates the output based on the input, The convolution2D transpose layer calculates the output based on the input,
filter, and dilations, strides, paddings. Input(Input) and output(Output) 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, are in NCHW or NHWC format. Where N is batch size, C is the number of channels,
...@@ -572,6 +578,9 @@ def conv3d(input, ...@@ -572,6 +578,9 @@ def conv3d(input,
data_format="NCDHW", data_format="NCDHW",
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.conv3d
:alias: paddle.nn.functional.conv3d,paddle.nn.functional.conv.conv3d
The convolution3D layer calculates the output based on the input, filter The convolution3D layer calculates the output based on the input, filter
and strides, paddings, dilations, groups parameters. Input(Input) and 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 Output(Output) are in NCDHW or NDHWC format. Where N is batch size C is the number of
...@@ -786,6 +795,9 @@ def conv3d_transpose(input, ...@@ -786,6 +795,9 @@ def conv3d_transpose(input,
data_format='NCDHW', data_format='NCDHW',
name=None): name=None):
""" """
:alias_main: paddle.nn.functional.conv3d_transpose
:alias: paddle.nn.functional.conv3d_transpose,paddle.nn.functional.conv.conv3d_transpose
The convolution3D transpose layer calculates the output based on the input, The convolution3D transpose layer calculates the output based on the input,
filter, and dilations, strides, paddings. Input(Input) and output(Output) 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, are in NCDHW or NDHWC format. Where N is batch size, C is the number of channels,
......
python/paddle/nn/functional/extension.py
浏览文件 @ f5c6dd6d
...@@ -48,6 +48,9 @@ from ...fluid.layers.layer_function_generator import templatedoc ...@@ -48,6 +48,9 @@ from ...fluid.layers.layer_function_generator import templatedoc
def diag_embed(input, offset=0, dim1=-2, dim2=-1): def diag_embed(input, offset=0, dim1=-2, dim2=-1):
""" """
:alias_main: paddle.nn.functional.diag_embed
:alias: paddle.nn.functional.diag_embed,paddle.nn.functional.extension.diag_embed
This OP creates a tensor whose diagonals of certain 2D planes (specified by dim1 and dim2) This OP creates a tensor whose diagonals of certain 2D planes (specified by dim1 and dim2)
are filled by ``input``. By default, a 2D plane formed by the last two dimensions are filled by ``input``. By default, a 2D plane formed by the last two dimensions
of the returned tensor will be selected. of the returned tensor will be selected.
...@@ -165,6 +168,9 @@ def diag_embed(input, offset=0, dim1=-2, dim2=-1): ...@@ -165,6 +168,9 @@ def diag_embed(input, offset=0, dim1=-2, dim2=-1):
@templatedoc() @templatedoc()
def row_conv(input, weight, act=None): def row_conv(input, weight, act=None):
""" """
:alias_main: paddle.nn.functional.row_conv
:alias: paddle.nn.functional.row_conv,paddle.nn.functional.extension.row_conv
${comment} ${comment}
Args: Args:
......
python/paddle/nn/layer/activation.py
浏览文件 @ f5c6dd6d
...@@ -31,6 +31,8 @@ from .. import functional ...@@ -31,6 +31,8 @@ from .. import functional
class HSigmoid(layers.Layer): class HSigmoid(layers.Layer):
""" """
:alias_main: paddle.nn.HSigmoid
:alias: paddle.nn.HSigmoid,paddle.nn.layer.HSigmoid,paddle.nn.layer.activation.HSigmoid
Hierarchical Sigmoid Layer. Hierarchical Sigmoid Layer.
...@@ -165,6 +167,9 @@ class HSigmoid(layers.Layer): ...@@ -165,6 +167,9 @@ class HSigmoid(layers.Layer):
class ReLU(layers.Layer): class ReLU(layers.Layer):
""" """
:alias_main: paddle.nn.ReLU
:alias: paddle.nn.ReLU,paddle.nn.layer.ReLU,paddle.nn.layer.activation.ReLU
ReLU Activation. ReLU Activation.
.. math: .. math:
...@@ -204,6 +209,9 @@ class ReLU(layers.Layer): ...@@ -204,6 +209,9 @@ class ReLU(layers.Layer):
class Sigmoid(layers.Layer): class Sigmoid(layers.Layer):
""" """
:alias_main: paddle.nn.Sigmoid
:alias: paddle.nn.Sigmoid,paddle.nn.layer.Sigmoid,paddle.nn.layer.activation.Sigmoid
Sigmoid Activation. Sigmoid Activation.
.. math: .. math:
...@@ -246,6 +254,9 @@ class Sigmoid(layers.Layer): ...@@ -246,6 +254,9 @@ class Sigmoid(layers.Layer):
class LogSoftmax(layers.Layer): class LogSoftmax(layers.Layer):
""" """
:alias_main: paddle.nn.LogSoftmax
:alias: paddle.nn.LogSoftmax,paddle.nn.layer.LogSoftmax,paddle.nn.layer.activation.LogSoftmax
This operator implements the log_softmax layer. The calculation process is as follows: This operator implements the log_softmax layer. The calculation process is as follows:
.. math:: .. math::
......
python/paddle/nn/layer/conv.py
浏览文件 @ f5c6dd6d
...@@ -40,6 +40,9 @@ def _get_default_param_initializer(num_channels, filter_size): ...@@ -40,6 +40,9 @@ def _get_default_param_initializer(num_channels, filter_size):
class Conv2D(layers.Layer): class Conv2D(layers.Layer):
""" """
:alias_main: paddle.nn.Conv2D
:alias: paddle.nn.Conv2D,paddle.nn.layer.Conv2D,paddle.nn.layer.conv.Conv2D
This interface is used to construct a callable object of the ``Conv2D`` class. This interface is used to construct a callable object of the ``Conv2D`` class.
For more details, refer to code examples. For more details, refer to code examples.
The convolution2D layer calculates the output based on the input, filter The convolution2D layer calculates the output based on the input, filter
...@@ -235,6 +238,9 @@ class Conv2D(layers.Layer): ...@@ -235,6 +238,9 @@ class Conv2D(layers.Layer):
class Conv2DTranspose(layers.Layer): class Conv2DTranspose(layers.Layer):
""" """
:alias_main: paddle.nn.Conv2DTranspose
:alias: paddle.nn.Conv2DTranspose,paddle.nn.layer.Conv2DTranspose,paddle.nn.layer.conv.Conv2DTranspose
This interface is used to construct a callable object of the ``Conv2DTranspose`` class. This interface is used to construct a callable object of the ``Conv2DTranspose`` class.
For more details, refer to code examples. For more details, refer to code examples.
The convolution2D transpose layer calculates the output based on the input, The convolution2D transpose layer calculates the output based on the input,
...@@ -431,6 +437,9 @@ class Conv2DTranspose(layers.Layer): ...@@ -431,6 +437,9 @@ class Conv2DTranspose(layers.Layer):
class Conv3D(layers.Layer): class Conv3D(layers.Layer):
""" """
:alias_main: paddle.nn.Conv3D
:alias: paddle.nn.Conv3D,paddle.nn.layer.Conv3D,paddle.nn.layer.conv.Conv3D
**Convlution3D Layer** **Convlution3D Layer**
The convolution3D layer calculates the output based on the input, filter The convolution3D layer calculates the output based on the input, filter
...@@ -621,6 +630,9 @@ class Conv3D(layers.Layer): ...@@ -621,6 +630,9 @@ class Conv3D(layers.Layer):
class Conv3DTranspose(layers.Layer): class Conv3DTranspose(layers.Layer):
""" """
:alias_main: paddle.nn.Conv3DTranspose
:alias: paddle.nn.Conv3DTranspose,paddle.nn.layer.Conv3DTranspose,paddle.nn.layer.conv.Conv3DTranspose
**Convlution3D transpose layer** **Convlution3D transpose layer**
The convolution3D transpose layer calculates the output based on the input, The convolution3D transpose layer calculates the output based on the input,
......
python/paddle/nn/layer/extension.py
浏览文件 @ f5c6dd6d
...@@ -20,6 +20,9 @@ from .. import functional as F ...@@ -20,6 +20,9 @@ from .. import functional as F
class RowConv(layers.Layer): class RowConv(layers.Layer):
""" """
:alias_main: paddle.nn.RowConv
:alias: paddle.nn.RowConv,paddle.nn.layer.RowConv,paddle.nn.layer.extension.RowConv
**Row-convolution operator** **Row-convolution operator**
The row convolution is called lookahead convolution. This operator was The row convolution is called lookahead convolution. This operator was
......
python/paddle/nn/layer/loss.py
浏览文件 @ f5c6dd6d
...@@ -28,6 +28,9 @@ __all__ = [ ...@@ -28,6 +28,9 @@ __all__ = [
class CrossEntropyLoss(fluid.dygraph.Layer): class CrossEntropyLoss(fluid.dygraph.Layer):
""" """
:alias_main: paddle.nn.CrossEntropyLoss
:alias: paddle.nn.CrossEntropyLoss,paddle.nn.layer.CrossEntropyLoss,paddle.nn.layer.loss.CrossEntropyLoss
This operator implements the cross entropy loss function. This OP combines ``LogSoftmax``, This operator implements the cross entropy loss function. This OP combines ``LogSoftmax``,
and ``NLLLoss`` together. and ``NLLLoss`` together.
...@@ -143,6 +146,9 @@ class CrossEntropyLoss(fluid.dygraph.Layer): ...@@ -143,6 +146,9 @@ class CrossEntropyLoss(fluid.dygraph.Layer):
class MSELoss(fluid.dygraph.layers.Layer): class MSELoss(fluid.dygraph.layers.Layer):
""" """
:alias_main: paddle.nn.MSELoss
:alias: paddle.nn.MSELoss,paddle.nn.layer.MSELoss,paddle.nn.layer.loss.MSELoss
**Mean Square Error Loss** **Mean Square Error Loss**
Computes the mean square error (squared L2 norm) of given input and label. Computes the mean square error (squared L2 norm) of given input and label.
...@@ -244,6 +250,9 @@ class MSELoss(fluid.dygraph.layers.Layer): ...@@ -244,6 +250,9 @@ class MSELoss(fluid.dygraph.layers.Layer):
class L1Loss(fluid.dygraph.Layer): class L1Loss(fluid.dygraph.Layer):
""" """
:alias_main: paddle.nn.L1Loss
:alias: paddle.nn.L1Loss,paddle.nn.layer.L1Loss,paddle.nn.layer.loss.L1Loss
This interface is used to construct a callable object of the ``L1Loss`` class. This interface is used to construct a callable object of the ``L1Loss`` class.
The L1Loss layer calculates the L1 Loss of input predictions and target The L1Loss layer calculates the L1 Loss of input predictions and target
labels as follows. labels as follows.
...@@ -331,6 +340,9 @@ class L1Loss(fluid.dygraph.Layer): ...@@ -331,6 +340,9 @@ class L1Loss(fluid.dygraph.Layer):
class BCELoss(fluid.dygraph.Layer): class BCELoss(fluid.dygraph.Layer):
""" """
:alias_main: paddle.nn.BCELoss
:alias: paddle.nn.BCELoss,paddle.nn.layer.BCELoss,paddle.nn.layer.loss.BCELoss
This interface is used to construct a callable object of the ``BCELoss`` class. This interface is used to construct a callable object of the ``BCELoss`` class.
The BCELoss layer measures the binary_cross_entropy loss between input predictions The BCELoss layer measures the binary_cross_entropy loss between input predictions
and target labels. The binary_cross_entropy loss can be described as: and target labels. The binary_cross_entropy loss can be described as:
...@@ -456,6 +468,9 @@ class BCELoss(fluid.dygraph.Layer): ...@@ -456,6 +468,9 @@ class BCELoss(fluid.dygraph.Layer):
class NLLLoss(fluid.dygraph.Layer): class NLLLoss(fluid.dygraph.Layer):
""" """
:alias_main: paddle.nn.NLLLoss
:alias: paddle.nn.NLLLoss,paddle.nn.layer.NLLLoss,paddle.nn.layer.loss.NLLLoss
This op accepts input and target label and returns negative log likelihood This op accepts input and target label and returns negative log likelihood
cross error. It is useful to train a classification problem with C classes. cross error. It is useful to train a classification problem with C classes.
......
python/paddle/tensor/creation.py
浏览文件 @ f5c6dd6d
...@@ -60,6 +60,9 @@ def full_like(input, ...@@ -60,6 +60,9 @@ def full_like(input,
stop_gradient=True, stop_gradient=True,
name=None): name=None):
""" """
:alias_main: paddle.full_like
:alias: paddle.full_like,paddle.tensor.full_like,paddle.tensor.creation.full_like
**full_like** **full_like**
This function creates a tensor filled with `fill_value` which has identical shape and dtype This function creates a tensor filled with `fill_value` which has identical shape and dtype
with `input`. with `input`.
...@@ -118,6 +121,9 @@ def full_like(input, ...@@ -118,6 +121,9 @@ def full_like(input,
def linspace(start, stop, num, dtype, out=None, device=None, name=None): def linspace(start, stop, num, dtype, out=None, device=None, name=None):
""" """
:alias_main: paddle.linspace
:alias: paddle.linspace,paddle.tensor.linspace,paddle.tensor.creation.linspace
This OP return fixed number of evenly spaced values within a given interval. This OP return fixed number of evenly spaced values within a given interval.
**NOTICE**: The output of this OP has no gradient. **NOTICE**: The output of this OP has no gradient.
...@@ -204,6 +210,9 @@ def linspace(start, stop, num, dtype, out=None, device=None, name=None): ...@@ -204,6 +210,9 @@ def linspace(start, stop, num, dtype, out=None, device=None, name=None):
def ones(shape, dtype=None, out=None, device=None): def ones(shape, dtype=None, out=None, device=None):
""" """
:alias_main: paddle.ones
:alias: paddle.ones,paddle.tensor.ones,paddle.tensor.creation.ones
The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 1. The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 1.
Args: Args:
...@@ -243,6 +252,9 @@ def ones(shape, dtype=None, out=None, device=None): ...@@ -243,6 +252,9 @@ def ones(shape, dtype=None, out=None, device=None):
def ones_like(input, dtype=None, device=None, name=None): def ones_like(input, dtype=None, device=None, name=None):
""" """
:alias_main: paddle.ones_like
:alias: paddle.ones_like,paddle.tensor.ones_like,paddle.tensor.creation.ones_like
This function creates a ones tensor which has identical shape and dtype This function creates a ones tensor which has identical shape and dtype
with `input`. with `input`.
...@@ -310,6 +322,9 @@ def ones_like(input, dtype=None, device=None, name=None): ...@@ -310,6 +322,9 @@ def ones_like(input, dtype=None, device=None, name=None):
def zeros(shape, dtype, out=None, device=None): def zeros(shape, dtype, out=None, device=None):
""" """
:alias_main: paddle.zeros
:alias: paddle.zeros,paddle.tensor.zeros,paddle.tensor.creation.zeros
The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 0. The OP creates a tensor of specified :attr:`shape` and :attr:`dtype`, and fills it with 0.
Args: Args:
...@@ -349,6 +364,9 @@ def zeros(shape, dtype, out=None, device=None): ...@@ -349,6 +364,9 @@ def zeros(shape, dtype, out=None, device=None):
def zeros_like(input, dtype=None, device=None, name=None): def zeros_like(input, dtype=None, device=None, name=None):
""" """
:alias_main: paddle.zeros_like
:alias: paddle.zeros_like,paddle.tensor.zeros_like,paddle.tensor.creation.zeros_like
This function creates a zeros tensor which has identical shape and dtype This function creates a zeros tensor which has identical shape and dtype
with `input`. with `input`.
...@@ -485,6 +503,9 @@ def full(shape, ...@@ -485,6 +503,9 @@ def full(shape,
stop_gradient=True, stop_gradient=True,
name=None): name=None):
""" """
:alias_main: paddle.full
:alias: paddle.full,paddle.tensor.full,paddle.tensor.creation.full
This Op return a Tensor with the `fill_value` which size is same as `shape` This Op return a Tensor with the `fill_value` which size is same as `shape`
Args: Args:
...@@ -562,6 +583,9 @@ def full(shape, ...@@ -562,6 +583,9 @@ def full(shape,
def arange(start, end, step=1, dtype=None, name=None): def arange(start, end, step=1, dtype=None, name=None):
""" """
:alias_main: paddle.arange
:alias: paddle.arange,paddle.tensor.arange,paddle.tensor.creation.arange
Return evenly spaced values within a given interval. Return evenly spaced values within a given interval.
Values are generated within the half-open interval [start, stop) (in other words, Values are generated within the half-open interval [start, stop) (in other words,
...@@ -666,6 +690,9 @@ def _tril_triu_op(helper): ...@@ -666,6 +690,9 @@ def _tril_triu_op(helper):
def tril(input, diagonal=0, name=None): def tril(input, diagonal=0, name=None):
""" """
:alias_main: paddle.tril
:alias: paddle.tril,paddle.tensor.tril,paddle.tensor.creation.tril
This op returns the lower triangular part of a matrix (2-D tensor) or batch This op returns the lower triangular part of a matrix (2-D tensor) or batch
of matrices :attr:`input`, the other elements of the result tensor are set of matrices :attr:`input`, the other elements of the result tensor are set
to 0. The lower triangular part of the matrix is defined as the elements to 0. The lower triangular part of the matrix is defined as the elements
...@@ -740,6 +767,9 @@ def tril(input, diagonal=0, name=None): ...@@ -740,6 +767,9 @@ def tril(input, diagonal=0, name=None):
def triu(input, diagonal=0, name=None): def triu(input, diagonal=0, name=None):
""" """
:alias_main: paddle.triu
:alias: paddle.triu,paddle.tensor.triu,paddle.tensor.creation.triu
This op returns the upper triangular part of a matrix (2-D tensor) or batch of matrices This op returns the upper triangular part of a matrix (2-D tensor) or batch of matrices
:attr:`input`, the other elements of the result tensor are set to 0. :attr:`input`, the other elements of the result tensor are set to 0.
The upper triangular part of the matrix is defined as the elements on and The upper triangular part of the matrix is defined as the elements on and
...@@ -814,6 +844,9 @@ def triu(input, diagonal=0, name=None): ...@@ -814,6 +844,9 @@ def triu(input, diagonal=0, name=None):
def meshgrid(input, name=None): def meshgrid(input, name=None):
""" """
:alias_main: paddle.meshgrid
:alias: paddle.meshgrid,paddle.tensor.meshgrid,paddle.tensor.creation.meshgrid
This op takes a list of N tensors as input, each of which is 1-dimensional This op takes a list of N tensors as input, each of which is 1-dimensional
vector, and creates N-dimensional grids. vector, and creates N-dimensional grids.
......
python/paddle/tensor/linalg.py
浏览文件 @ f5c6dd6d
...@@ -36,6 +36,9 @@ __all__ = [ ...@@ -36,6 +36,9 @@ __all__ = [
def matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None): def matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None):
""" """
:alias_main: paddle.matmul
:alias: paddle.matmul,paddle.tensor.matmul,paddle.tensor.linalg.matmul
Applies matrix multiplication to two tensors. Applies matrix multiplication to two tensors.
Currently, the input tensors' rank can be any, but when the rank of any Currently, the input tensors' rank can be any, but when the rank of any
...@@ -169,6 +172,9 @@ def matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None): ...@@ -169,6 +172,9 @@ def matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None):
def norm(input, p='fro', axis=None, keepdim=False, out=None, name=None): def norm(input, p='fro', axis=None, keepdim=False, out=None, name=None):
""" """
:alias_main: paddle.norm
:alias: paddle.norm,paddle.tensor.norm,paddle.tensor.linalg.norm
Returns the matrix norm (Frobenius) or vector norm (the 1-norm, the Euclidean Returns the matrix norm (Frobenius) or vector norm (the 1-norm, the Euclidean
or 2-norm, and in general the p-norm for p > 0) of a given tensor. or 2-norm, and in general the p-norm for p > 0) of a given tensor.
...@@ -344,6 +350,9 @@ def norm(input, p='fro', axis=None, keepdim=False, out=None, name=None): ...@@ -344,6 +350,9 @@ def norm(input, p='fro', axis=None, keepdim=False, out=None, name=None):
def dist(x, y, p=2): def dist(x, y, p=2):
""" """
:alias_main: paddle.dist
:alias: paddle.dist,paddle.tensor.dist,paddle.tensor.linalg.dist
This OP returns the p-norm of (x - y). It is not a norm in a strict sense, only as a measure This OP returns the p-norm of (x - y). It is not a norm in a strict sense, only as a measure
of distance. The shapes of x and y must be broadcastable. The definition is as follows, for of distance. The shapes of x and y must be broadcastable. The definition is as follows, for
details, please refer to the `numpy's broadcasting <https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html>`_: details, please refer to the `numpy's broadcasting <https://docs.scipy.org/doc/numpy/user/basics.broadcasting.html>`_:
...@@ -442,6 +451,9 @@ def dist(x, y, p=2): ...@@ -442,6 +451,9 @@ def dist(x, y, p=2):
def dot(x, y, name=None): def dot(x, y, name=None):
""" """
:alias_main: paddle.dot
:alias: paddle.dot,paddle.tensor.dot,paddle.tensor.linalg.dot
This operator calculates inner product for vectors. This operator calculates inner product for vectors.
.. note:: .. note::
...@@ -498,6 +510,9 @@ def dot(x, y, name=None): ...@@ -498,6 +510,9 @@ def dot(x, y, name=None):
def t(input, name=None): def t(input, name=None):
""" """
:alias_main: paddle.t
:alias: paddle.t,paddle.tensor.t,paddle.tensor.linalg.t
Transpose <=2-D tensor. Transpose <=2-D tensor.
0-D and 1-D tensors are returned as it is and 2-D tensor is equal to 0-D and 1-D tensors are returned as it is and 2-D tensor is equal to
the fluid.layers.transpose function which perm dimensions set 0 and 1. the fluid.layers.transpose function which perm dimensions set 0 and 1.
...@@ -569,6 +584,9 @@ def t(input, name=None): ...@@ -569,6 +584,9 @@ def t(input, name=None):
def cross(input, other, dim=None): def cross(input, other, dim=None):
""" """
:alias_main: paddle.cross
:alias: paddle.cross,paddle.tensor.cross,paddle.tensor.linalg.cross
Returns the cross product of vectors in dimension `dim` of the `input` and `other` tensor. Returns the cross product of vectors in dimension `dim` of the `input` and `other` tensor.
Inputs must have the same shape, and the size of their dim-th dimension should be equla to 3. Inputs must have the same shape, and the size of their dim-th dimension should be equla to 3.
If `dim` is not given, it defaults to the first dimension found with the size 3. If `dim` is not given, it defaults to the first dimension found with the size 3.
...@@ -631,6 +649,9 @@ def cross(input, other, dim=None): ...@@ -631,6 +649,9 @@ def cross(input, other, dim=None):
def cholesky(x, upper=False): def cholesky(x, upper=False):
""" """
:alias_main: paddle.cholesky
:alias: paddle.cholesky,paddle.tensor.cholesky,paddle.tensor.linalg.cholesky
Computes the Cholesky decomposition of one symmetric positive-definite Computes the Cholesky decomposition of one symmetric positive-definite
matrix or batches of symmetric positive-definite matrice. matrix or batches of symmetric positive-definite matrice.
...@@ -684,6 +705,9 @@ def cholesky(x, upper=False): ...@@ -684,6 +705,9 @@ def cholesky(x, upper=False):
def bmm(x, y, name=None): def bmm(x, y, name=None):
""" """
:alias_main: paddle.bmm
:alias: paddle.bmm,paddle.tensor.bmm,paddle.tensor.linalg.bmm
Applies batched matrix multiplication to two tensors. Applies batched matrix multiplication to two tensors.
Both of the two input tensors must be three-dementional and share the same batch size. Both of the two input tensors must be three-dementional and share the same batch size.
......
python/paddle/tensor/logic.py
浏览文件 @ f5c6dd6d
...@@ -54,6 +54,9 @@ __all__ = [ ...@@ -54,6 +54,9 @@ __all__ = [
def equal(x, y, axis=-1, name=None): def equal(x, y, axis=-1, name=None):
""" """
:alias_main: paddle.equal
:alias: paddle.equal,paddle.tensor.equal,paddle.tensor.logic.equal
This OP returns the truth value of :math:`x == y`. True if two inputs have the same elements, False otherwise. This OP returns the truth value of :math:`x == y`. True if two inputs have the same elements, False otherwise.
**NOTICE**: The output of this OP has no gradient, and this OP supports broadcasting by :attr:`axis`. **NOTICE**: The output of this OP has no gradient, and this OP supports broadcasting by :attr:`axis`.
...@@ -122,6 +125,9 @@ def equal(x, y, axis=-1, name=None): ...@@ -122,6 +125,9 @@ def equal(x, y, axis=-1, name=None):
@templatedoc() @templatedoc()
def allclose(input, other, rtol=1e-05, atol=1e-08, equal_nan=False, name=None): def allclose(input, other, rtol=1e-05, atol=1e-08, equal_nan=False, name=None):
""" """
:alias_main: paddle.allclose
:alias: paddle.allclose,paddle.tensor.allclose,paddle.tensor.logic.allclose
${comment} ${comment}
Args: Args:
...@@ -204,6 +210,9 @@ def allclose(input, other, rtol=1e-05, atol=1e-08, equal_nan=False, name=None): ...@@ -204,6 +210,9 @@ def allclose(input, other, rtol=1e-05, atol=1e-08, equal_nan=False, name=None):
def elementwise_equal(x, y, name=None): def elementwise_equal(x, y, name=None):
""" """
:alias_main: paddle.elementwise_equal
:alias: paddle.elementwise_equal,paddle.tensor.elementwise_equal,paddle.tensor.logic.elementwise_equal
This layer returns the truth value of :math:`x == y` elementwise. This layer returns the truth value of :math:`x == y` elementwise.
Args: Args:
......
python/paddle/tensor/manipulation.py
浏览文件 @ f5c6dd6d
...@@ -68,6 +68,9 @@ __all__ = [ ...@@ -68,6 +68,9 @@ __all__ = [
def flip(input, dims, name=None): def flip(input, dims, name=None):
""" """
:alias_main: paddle.flip
:alias: paddle.flip,paddle.tensor.flip,paddle.tensor.manipulation.flip
Reverse the order of a n-D tensor along given axis in dims. Reverse the order of a n-D tensor along given axis in dims.
...@@ -118,6 +121,9 @@ def flip(input, dims, name=None): ...@@ -118,6 +121,9 @@ def flip(input, dims, name=None):
def roll(input, shifts, dims=None): def roll(input, shifts, dims=None):
""" """
:alias_main: paddle.roll
:alias: paddle.roll,paddle.tensor.roll,paddle.tensor.manipulation.roll
Roll the `input` tensor along the given dimension(s). Elements that are shifted beyond Roll the `input` tensor along the given dimension(s). Elements that are shifted beyond
the last position are re-introduced at the first position. If a dimension is not specified, the last position are re-introduced at the first position. If a dimension is not specified,
the tensor will be flattened before rolling and then restored to the original shape. the tensor will be flattened before rolling and then restored to the original shape.
...@@ -189,6 +195,9 @@ def roll(input, shifts, dims=None): ...@@ -189,6 +195,9 @@ def roll(input, shifts, dims=None):
def stack(x, axis=0, out=None, name=None): def stack(x, axis=0, out=None, name=None):
""" """
:alias_main: paddle.stack
:alias: paddle.stack,paddle.tensor.stack,paddle.tensor.manipulation.stack
This OP stacks all the inputs :code:`x` along axis. This OP stacks all the inputs :code:`x` along axis.
...@@ -299,6 +308,9 @@ def stack(x, axis=0, out=None, name=None): ...@@ -299,6 +308,9 @@ def stack(x, axis=0, out=None, name=None):
def split(input, num_or_sections, dim=-1, name=None): def split(input, num_or_sections, dim=-1, name=None):
""" """
:alias_main: paddle.split
:alias: paddle.split,paddle.tensor.split,paddle.tensor.manipulation.split
Split the input tensor into multiple sub-Tensors. Split the input tensor into multiple sub-Tensors.
Args: Args:
input (Variable): The input variable which is an N-D Tensor or LoDTensor, data type being float32, float64, int32 or int64. input (Variable): The input variable which is an N-D Tensor or LoDTensor, data type being float32, float64, int32 or int64.
...@@ -435,6 +447,9 @@ def split(input, num_or_sections, dim=-1, name=None): ...@@ -435,6 +447,9 @@ def split(input, num_or_sections, dim=-1, name=None):
def squeeze(input, axes, out=None, name=None): def squeeze(input, axes, out=None, name=None):
""" """
:alias_main: paddle.squeeze
:alias: paddle.squeeze,paddle.tensor.squeeze,paddle.tensor.manipulation.squeeze
This OP will squeeze single-dimensional entries of input tensor's shape. If axes is provided, will This OP will squeeze single-dimensional entries of input tensor's shape. If axes is provided, will
remove the dims by axes, the dims selected by axes should be one. If not provide axes, all dims equal remove the dims by axes, the dims selected by axes should be one. If not provide axes, all dims equal
to one will be deleted. to one will be deleted.
...@@ -511,6 +526,9 @@ def squeeze(input, axes, out=None, name=None): ...@@ -511,6 +526,9 @@ def squeeze(input, axes, out=None, name=None):
def unsqueeze(input, axes, out=None, name=None): def unsqueeze(input, axes, out=None, name=None):
""" """
:alias_main: paddle.unsqueeze
:alias: paddle.unsqueeze,paddle.tensor.unsqueeze,paddle.tensor.manipulation.unsqueeze
Insert single-dimensional entries to the shape of a Tensor. Takes one Insert single-dimensional entries to the shape of a Tensor. Takes one
required argument axes, a list of dimensions that will be inserted. required argument axes, a list of dimensions that will be inserted.
Dimension indices in axes are as seen in the output tensor. Dimension indices in axes are as seen in the output tensor.
...@@ -591,6 +609,9 @@ def unsqueeze(input, axes, out=None, name=None): ...@@ -591,6 +609,9 @@ def unsqueeze(input, axes, out=None, name=None):
def gather(input, index, overwrite=True): def gather(input, index, overwrite=True):
""" """
:alias_main: paddle.gather
:alias: paddle.gather,paddle.tensor.gather,paddle.tensor.manipulation.gather
**Gather Layer** **Gather Layer**
Output is obtained by gathering entries of the outer-most dimension Output is obtained by gathering entries of the outer-most dimension
...@@ -662,6 +683,9 @@ def gather(input, index, overwrite=True): ...@@ -662,6 +683,9 @@ def gather(input, index, overwrite=True):
def unbind(input, axis=0): def unbind(input, axis=0):
""" """
:alias_main: paddle.tensor.unbind
:alias: paddle.tensor.unbind,paddle.tensor.manipulation.unbind
Removes a tensor dimension, then split the input tensor into multiple sub-Tensors. Removes a tensor dimension, then split the input tensor into multiple sub-Tensors.
Args: Args:
input (Variable): The input variable which is an N-D Tensor, data type being float32, float64, int32 or int64. input (Variable): The input variable which is an N-D Tensor, data type being float32, float64, int32 or int64.
......
python/paddle/tensor/math.py
浏览文件 @ f5c6dd6d
...@@ -186,6 +186,9 @@ Examples: ...@@ -186,6 +186,9 @@ Examples:
@templatedoc() @templatedoc()
def pow(input, exponent, out=None, name=None): def pow(input, exponent, out=None, name=None):
""" """
:alias_main: paddle.pow
:alias: paddle.pow,paddle.tensor.pow,paddle.tensor.math.pow
This is Pow Activation Operator. This is Pow Activation Operator.
:math:`out = input^{exponent}` :math:`out = input^{exponent}`
...@@ -254,6 +257,9 @@ def pow(input, exponent, out=None, name=None): ...@@ -254,6 +257,9 @@ def pow(input, exponent, out=None, name=None):
def mul(x, y, x_num_col_dims=1, y_num_col_dims=1, out=None, name=None): def mul(x, y, x_num_col_dims=1, y_num_col_dims=1, out=None, name=None):
""" """
:alias_main: paddle.mul
:alias: paddle.mul,paddle.tensor.mul,paddle.tensor.math.mul
Mul Operator. Mul Operator.
This operator is used to perform matrix multiplication for input $x$ and $y$. This operator is used to perform matrix multiplication for input $x$ and $y$.
The equation is: The equation is:
...@@ -400,6 +406,9 @@ def _elementwise_op(helper): ...@@ -400,6 +406,9 @@ def _elementwise_op(helper):
def add(x, y, alpha=1, out=None, name=None): def add(x, y, alpha=1, out=None, name=None):
""" """
:alias_main: paddle.add
:alias: paddle.add,paddle.tensor.add,paddle.tensor.math.add
Examples: Examples:
.. code-block:: python .. code-block:: python
...@@ -542,6 +551,9 @@ Examples: ...@@ -542,6 +551,9 @@ Examples:
def div(x, y, out=None, name=None): def div(x, y, out=None, name=None):
""" """
:alias_main: paddle.div
:alias: paddle.div,paddle.tensor.div,paddle.tensor.math.div
Examples: Examples:
.. code-block:: python .. code-block:: python
...@@ -692,6 +704,9 @@ for func in [ ...@@ -692,6 +704,9 @@ for func in [
def sum(input, dim=None, dtype=None, keep_dim=False, name=None): def sum(input, dim=None, dtype=None, keep_dim=False, name=None):
""" """
:alias_main: paddle.sum
:alias: paddle.sum,paddle.tensor.sum,paddle.tensor.math.sum
Computes the sum of tensor elements over the given dimension. Computes the sum of tensor elements over the given dimension.
Args: Args:
...@@ -794,6 +809,9 @@ def sum(input, dim=None, dtype=None, keep_dim=False, name=None): ...@@ -794,6 +809,9 @@ def sum(input, dim=None, dtype=None, keep_dim=False, name=None):
@templatedoc(op_type="sum") @templatedoc(op_type="sum")
def elementwise_sum(inputs, name=None): def elementwise_sum(inputs, name=None):
""" """
:alias_main: paddle.elementwise_sum
:alias: paddle.elementwise_sum,paddle.tensor.elementwise_sum,paddle.tensor.math.elementwise_sum
${comment} ${comment}
Case 1: Case 1:
...@@ -889,6 +907,9 @@ def elementwise_sum(inputs, name=None): ...@@ -889,6 +907,9 @@ def elementwise_sum(inputs, name=None):
def mm(input, mat2, out=None, name=None): def mm(input, mat2, out=None, name=None):
""" """
:alias_main: paddle.mm
:alias: paddle.mm,paddle.tensor.mm,paddle.tensor.math.mm
Applies matrix multiplication to two tensors. Applies matrix multiplication to two tensors.
Currently, the input tensors' rank can be any, but when the rank of any Currently, the input tensors' rank can be any, but when the rank of any
...@@ -991,6 +1012,9 @@ def mm(input, mat2, out=None, name=None): ...@@ -991,6 +1012,9 @@ def mm(input, mat2, out=None, name=None):
def addmm(input, x, y, alpha=1.0, beta=1.0, name=None): def addmm(input, x, y, alpha=1.0, beta=1.0, name=None):
""" """
:alias_main: paddle.addmm
:alias: paddle.addmm,paddle.tensor.addmm,paddle.tensor.math.addmm
**addmm** **addmm**
This operator is used to perform matrix multiplication for input $x$ and $y$. This operator is used to perform matrix multiplication for input $x$ and $y$.
...@@ -1057,6 +1081,9 @@ def addmm(input, x, y, alpha=1.0, beta=1.0, name=None): ...@@ -1057,6 +1081,9 @@ def addmm(input, x, y, alpha=1.0, beta=1.0, name=None):
def logsumexp(x, dim=None, keepdim=False, out=None, name=None): def logsumexp(x, dim=None, keepdim=False, out=None, name=None):
""" """
:alias_main: paddle.logsumexp
:alias: paddle.logsumexp,paddle.tensor.logsumexp,paddle.tensor.math.logsumexp
This operator calculates the log of the sum of exponentials of the input Tensor. This operator calculates the log of the sum of exponentials of the input Tensor.
.. math:: .. math::
...@@ -1125,6 +1152,9 @@ def logsumexp(x, dim=None, keepdim=False, out=None, name=None): ...@@ -1125,6 +1152,9 @@ def logsumexp(x, dim=None, keepdim=False, out=None, name=None):
def inverse(input, out=None, name=None): def inverse(input, out=None, name=None):
""" """
:alias_main: paddle.inverse
:alias: paddle.inverse,paddle.tensor.inverse,paddle.tensor.math.inverse
Takes the inverse of the square matrix. A square matrix is a matrix with Takes the inverse of the square matrix. A square matrix is a matrix with
the same number of rows and columns. The input can be a square matrix the same number of rows and columns. The input can be a square matrix
(2-D Tensor) or batches of square matrices. (2-D Tensor) or batches of square matrices.
...@@ -1197,6 +1227,9 @@ def inverse(input, out=None, name=None): ...@@ -1197,6 +1227,9 @@ def inverse(input, out=None, name=None):
def max(input, dim=None, keep_dim=False, out=None, name=None): def max(input, dim=None, keep_dim=False, out=None, name=None):
""" """
:alias_main: paddle.max
:alias: paddle.max,paddle.tensor.max,paddle.tensor.math.max
Computes the maximum of tensor elements over the given dimension. Computes the maximum of tensor elements over the given dimension.
Args: Args:
...@@ -1274,6 +1307,9 @@ def max(input, dim=None, keep_dim=False, out=None, name=None): ...@@ -1274,6 +1307,9 @@ def max(input, dim=None, keep_dim=False, out=None, name=None):
def min(input, dim=None, keep_dim=False, out=None, name=None): def min(input, dim=None, keep_dim=False, out=None, name=None):
""" """
:alias_main: paddle.min
:alias: paddle.min,paddle.tensor.min,paddle.tensor.math.min
Computes the minimum of tensor elements over the given dimension. Computes the minimum of tensor elements over the given dimension.
Args: Args:
...@@ -1350,6 +1386,9 @@ def min(input, dim=None, keep_dim=False, out=None, name=None): ...@@ -1350,6 +1386,9 @@ def min(input, dim=None, keep_dim=False, out=None, name=None):
def log1p(x, out=None, name=None): def log1p(x, out=None, name=None):
""" """
:alias_main: paddle.log1p
:alias: paddle.log1p,paddle.tensor.log1p,paddle.tensor.math.log1p
Calculates the natural log of the given input tensor, element-wise. Calculates the natural log of the given input tensor, element-wise.
.. math:: .. math::
Out = \\ln(x+1) Out = \\ln(x+1)
...@@ -1393,6 +1432,9 @@ def log1p(x, out=None, name=None): ...@@ -1393,6 +1432,9 @@ def log1p(x, out=None, name=None):
def addcmul(input, tensor1, tensor2, value=1.0, out=None, name=None): def addcmul(input, tensor1, tensor2, value=1.0, out=None, name=None):
""" """
:alias_main: paddle.addcmul
:alias: paddle.addcmul,paddle.tensor.addcmul,paddle.tensor.math.addcmul
Calculate the element-wise multiplication of tensor1 and tensor2, Calculate the element-wise multiplication of tensor1 and tensor2,
then multiply the result by value, and add it to input. The shape of input, then multiply the result by value, and add it to input. The shape of input,
tensor1, tensor2 should be broadcastable. tensor1, tensor2 should be broadcastable.
...@@ -1439,6 +1481,9 @@ def addcmul(input, tensor1, tensor2, value=1.0, out=None, name=None): ...@@ -1439,6 +1481,9 @@ def addcmul(input, tensor1, tensor2, value=1.0, out=None, name=None):
def clamp(input, min=None, max=None, output=None, name=None): def clamp(input, min=None, max=None, output=None, name=None):
""" """
:alias_main: paddle.clamp
:alias: paddle.clamp,paddle.tensor.clamp,paddle.tensor.math.clamp
**clampe layer** **clampe layer**
This operator clamps all elements in input into the range [ min, max ] and return This operator clamps all elements in input into the range [ min, max ] and return
...@@ -1524,6 +1569,9 @@ def clamp(input, min=None, max=None, output=None, name=None): ...@@ -1524,6 +1569,9 @@ def clamp(input, min=None, max=None, output=None, name=None):
def trace(input, offset=0, dim1=0, dim2=1, out=None, name=None): def trace(input, offset=0, dim1=0, dim2=1, out=None, name=None):
""" """
:alias_main: paddle.trace
:alias: paddle.trace,paddle.tensor.trace,paddle.tensor.math.trace
This OP computes the sum along diagonals of the input tensor. This OP computes the sum along diagonals of the input tensor.
If ``input`` is 2D, returns the sum of diagonal. If ``input`` is 2D, returns the sum of diagonal.
...@@ -1617,7 +1665,11 @@ def trace(input, offset=0, dim1=0, dim2=1, out=None, name=None): ...@@ -1617,7 +1665,11 @@ def trace(input, offset=0, dim1=0, dim2=1, out=None, name=None):
@templatedoc(op_type="kron") @templatedoc(op_type="kron")
def kron(x, y, out=None, name=None): def kron(x, y, out=None, name=None):
"""${comment} """
:alias_main: paddle.kron
:alias: paddle.kron,paddle.tensor.kron,paddle.tensor.math.kron
${comment}
Args: Args:
x (Variable): the fist operand of kron op, data type: float16, float32, x (Variable): the fist operand of kron op, data type: float16, float32,
......
python/paddle/tensor/random.py
浏览文件 @ f5c6dd6d
...@@ -47,6 +47,9 @@ def randint(low, ...@@ -47,6 +47,9 @@ def randint(low,
seed=0, seed=0,
name=None): name=None):
""" """
:alias_main: paddle.randint
:alias: paddle.randint,paddle.tensor.randint,paddle.tensor.random.randint
This function returns a Tensor filled with random integers from the "discrete uniform" distribution of the This function returns a Tensor filled with random integers from the "discrete uniform" distribution of the
specified data type in the interval [low, high). If high is None (the default), then results are from [0, low). specified data type in the interval [low, high). If high is None (the default), then results are from [0, low).
...@@ -209,6 +212,9 @@ def randn(shape, ...@@ -209,6 +212,9 @@ def randn(shape,
stop_gradient=True, stop_gradient=True,
name=None): name=None):
""" """
:alias_main: paddle.randn
:alias: paddle.randn,paddle.tensor.randn,paddle.tensor.random.randn
This function returns a tensor filled with random numbers from a normal This function returns a tensor filled with random numbers from a normal
distribution with mean 0 and variance 1 (also called the standard normal distribution with mean 0 and variance 1 (also called the standard normal
distribution). distribution).
...@@ -317,6 +323,9 @@ def randperm(n, ...@@ -317,6 +323,9 @@ def randperm(n,
stop_gradient=True, stop_gradient=True,
seed=0): seed=0):
""" """
:alias_main: paddle.randperm
:alias: paddle.randperm,paddle.tensor.randperm,paddle.tensor.random.randperm
${comment} ${comment}
Args: Args:
...@@ -398,6 +407,9 @@ def randperm(n, ...@@ -398,6 +407,9 @@ def randperm(n,
def rand(shape, out=None, dtype=None, device=None, stop_gradient=True): def rand(shape, out=None, dtype=None, device=None, stop_gradient=True):
""" """
:alias_main: paddle.rand
:alias: paddle.rand,paddle.tensor.rand,paddle.tensor.random.rand
This OP initializes a variable with random values sampled from a This OP initializes a variable with random values sampled from a
uniform distribution in the range [0, 1). uniform distribution in the range [0, 1).
......
python/paddle/tensor/search.py
浏览文件 @ f5c6dd6d
...@@ -44,6 +44,9 @@ from paddle.common_ops_import import * ...@@ -44,6 +44,9 @@ from paddle.common_ops_import import *
def argmax(input, axis=None, dtype=None, out=None, keepdims=False, name=None): def argmax(input, axis=None, dtype=None, out=None, keepdims=False, name=None):
""" """
:alias_main: paddle.argmax
:alias: paddle.argmax,paddle.tensor.argmax,paddle.tensor.search.argmax
This OP computes the indices of the max elements of the input tensor's This OP computes the indices of the max elements of the input tensor's
element along the provided axis. element along the provided axis.
...@@ -134,6 +137,9 @@ def argmax(input, axis=None, dtype=None, out=None, keepdims=False, name=None): ...@@ -134,6 +137,9 @@ def argmax(input, axis=None, dtype=None, out=None, keepdims=False, name=None):
def index_select(input, index, dim=0): def index_select(input, index, dim=0):
""" """
:alias_main: paddle.index_select
:alias: paddle.index_select,paddle.tensor.index_select,paddle.tensor.search.index_select
Returns a new tensor which indexes the `input` tensor along dimension `dim` using Returns a new tensor which indexes the `input` tensor along dimension `dim` using
the entries in `index` which is a Tensor. The returned tensor has the same number the entries in `index` which is a Tensor. The returned tensor has the same number
of dimensions as the original `input` tensor. The dim-th dimension has the same of dimensions as the original `input` tensor. The dim-th dimension has the same
...@@ -195,6 +201,9 @@ def index_select(input, index, dim=0): ...@@ -195,6 +201,9 @@ def index_select(input, index, dim=0):
def nonzero(input, as_tuple=False): def nonzero(input, as_tuple=False):
""" """
:alias_main: paddle.nonzero
:alias: paddle.nonzero,paddle.tensor.nonzero,paddle.tensor.search.nonzero
Return a tensor containing the indices of all non-zero elements of the `input` Return a tensor containing the indices of all non-zero elements of the `input`
tensor. If as_tuple is True, return a tuple of 1-D tensors, one for each dimension tensor. If as_tuple is True, return a tuple of 1-D tensors, one for each dimension
in `input`, each containing the indices (in that dimension) of all non-zero elements in `input`, each containing the indices (in that dimension) of all non-zero elements
...@@ -279,6 +288,9 @@ def nonzero(input, as_tuple=False): ...@@ -279,6 +288,9 @@ def nonzero(input, as_tuple=False):
def sort(input, axis=-1, descending=False, out=None, name=None): def sort(input, axis=-1, descending=False, out=None, name=None):
""" """
:alias_main: paddle.sort
:alias: paddle.sort,paddle.tensor.sort,paddle.tensor.search.sort
This OP sorts the input along the given axis, and returns sorted output This OP sorts the input along the given axis, and returns sorted output
data Varibale and its corresponding index Variable with the same shape as data Varibale and its corresponding index Variable with the same shape as
:attr:`input`. :attr:`input`.
...@@ -368,6 +380,9 @@ def sort(input, axis=-1, descending=False, out=None, name=None): ...@@ -368,6 +380,9 @@ def sort(input, axis=-1, descending=False, out=None, name=None):
def where(condition, x, y, name=None): def where(condition, x, y, name=None):
""" """
:alias_main: paddle.where
:alias: paddle.where,paddle.tensor.where,paddle.tensor.search.where
Return a tensor of elements selected from either $x$ or $y$, depending on $condition$. Return a tensor of elements selected from either $x$ or $y$, depending on $condition$.
.. math:: .. math::
...@@ -443,6 +458,9 @@ def where(condition, x, y, name=None): ...@@ -443,6 +458,9 @@ def where(condition, x, y, name=None):
def index_sample(x, index): def index_sample(x, index):
""" """
:alias_main: paddle.index_sample
:alias: paddle.index_sample,paddle.tensor.index_sample,paddle.tensor.search.index_sample
**IndexSample Layer** **IndexSample Layer**
IndexSample OP returns the element of the specified location of X, IndexSample OP returns the element of the specified location of X,
......
python/paddle/tensor/stat.py
浏览文件 @ f5c6dd6d
...@@ -28,6 +28,9 @@ from ..fluid.data_feeder import convert_dtype, check_variable_and_dtype, check_t ...@@ -28,6 +28,9 @@ from ..fluid.data_feeder import convert_dtype, check_variable_and_dtype, check_t
def var(input, axis=None, keepdim=False, unbiased=True, out=None, name=None): def var(input, axis=None, keepdim=False, unbiased=True, out=None, name=None):
""" """
:alias_main: paddle.var
:alias: paddle.var,paddle.tensor.var,paddle.tensor.stat.var
Computes the variance of the input Variable's elements along the specified Computes the variance of the input Variable's elements along the specified
axis. axis.
...@@ -104,6 +107,9 @@ def var(input, axis=None, keepdim=False, unbiased=True, out=None, name=None): ...@@ -104,6 +107,9 @@ def var(input, axis=None, keepdim=False, unbiased=True, out=None, name=None):
def std(input, axis=None, keepdim=False, unbiased=True, out=None, name=None): def std(input, axis=None, keepdim=False, unbiased=True, out=None, name=None):
""" """
:alias_main: paddle.std
:alias: paddle.std,paddle.tensor.std,paddle.tensor.stat.std
Computes the standard-deviation of the input Variable's elements along the specified Computes the standard-deviation of the input Variable's elements along the specified
axis. axis.
......
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