Skip to content

2.5 install doc #5866

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 27 commits into
base: develop
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
27 commits
Select commit Hold shift + click to select a range
76498bc
copy v2.3 to v2.4 (#5309)
pangyoki Sep 27, 2022
bf33f8a
[install doc] update 2.4.0rc0 install doc (#5312)
pangyoki Sep 28, 2022
bc1ea59
add tensorrt in docs (#5313)
JZZ-NOTE Sep 28, 2022
e7e61b6
add tensorrt in docs (#5315)
JZZ-NOTE Sep 28, 2022
ffa9552
Jzz docs 2.4rc (#5317)
JZZ-NOTE Sep 28, 2022
2784e11
modify details (#5319)
JZZ-NOTE Sep 28, 2022
4559ba9
update doc of win (#5321)
zhwesky2010 Sep 29, 2022
e2f2bb9
[install doc] fix 2.4.0rc0 install doc (#5320)
pangyoki Sep 29, 2022
54e60ae
优化2.4rc 安装文档 (#5329)
JZZ-NOTE Sep 29, 2022
4114092
Update hyperlink in Chinese Overview doc (#5307) (#5339)
caolonghao Oct 9, 2022
e04cf7d
Fix infrence_lib download link (#5356)
JZZ-NOTE Oct 13, 2022
7914d20
Add paddle.geometric docs (#5292) (#5344)
DesmonDay Oct 25, 2022
7ba8fe9
[cherry-pick2.4]docs fix (#5401)
sunzhongkai588 Nov 3, 2022
bcccdf4
fix doc in recompute (#5407)
sljlp Nov 7, 2022
084af18
[cherry-pick] Delete geometric api release 2.4 (#5435)
DesmonDay Nov 21, 2022
87c9ba5
2.4.0 docs update: del suffix rc0 and many slight modify. (#5420)
zhengqiwen1997 Nov 22, 2022
36815e3
del conda macOS 1.2.2 python version and Tables.md:cuda11.0 (#5449)
zhengqiwen1997 Nov 24, 2022
5b126b5
2.4.0 docs update (#5451)
zhengqiwen1997 Nov 25, 2022
b5f42af
fix create parameter link error (#5472) (#5473)
chenwhql Dec 2, 2022
7e6f0b0
add 2.4.0 release note (#5464) (#5483)
dingjiaweiww Dec 5, 2022
4c509cc
change 2.4.0 to 2.4.1 (#5491)
zhengqiwen1997 Dec 8, 2022
3713090
mac py37 install link modify (#5499)
zhengqiwen1997 Dec 9, 2022
dfc0e17
[MLU] add mlu docs for r2.4 (#5432)
ShawnNew Jan 4, 2023
070b96b
Add chinese doc of paddle sparse api (#5603)
zhwesky2010 Feb 3, 2023
10844b7
[cherry-pick] add audio doc(#5299 #5363 #5378 #5445#5609) (#5608)
SmileGoat Feb 23, 2023
4d1b25b
2.4.1 to 2.4.2 and macOS avx installation (#5630)
zhengqiwen1997 Feb 27, 2023
fb0c30e
[cherry-pick]Release note2.4.1 (#5617)
sunzhongkai588 Mar 1, 2023
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Add chinese doc of paddle sparse api (#5603)
  • Loading branch information
zhwesky2010 authored Feb 3, 2023
commit 070b96b8fc8ebc6dcefc9dbdf154e42bdb9de67f
54 changes: 0 additions & 54 deletions docs/api/paddle/incubate/sparse/Overview_cn.rst

This file was deleted.

2 changes: 1 addition & 1 deletion docs/api/paddle/kthvalue_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ kthvalue
:::::::::
- **x** (Tensor) - 一个输入的 N-D ``Tensor``,支持的数据类型:float32、float64、int32、int64。
- **k** (int,Tensor) - 需要沿轴查找的第 ``k`` 小,所对应的 ``k`` 值。
- **axis** (int,可选) - 指定对输入 Tensor 进行运算的轴,``axis`` 的有效范围是[-R, R),R 是输入 ``x`` 的 Rank, ``axis`` 为负时与 ``axis`` + R 等价。默认值为-1。
- **axis** (int,可选) - 指定对输入 Tensor 进行运算的轴,``axis`` 的有效范围是[-R, R),R 是输入 ``x`` 的 Rank, ``axis`` 为负时与 ``axis`` + R 等价。默认值为 None, 此时等当于-1。
- **keepdim** (bool,可选)- 是否保留指定的轴。如果是 True,维度会与输入 x 一致,对应所指定的轴的 size 为 1。否则,由于对应轴被展开,输出的维度会比输入小 1。默认值为 1。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

Expand Down
96 changes: 96 additions & 0 deletions docs/api/paddle/sparse/Overview_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,96 @@
.. _cn_overview_paddle_sparse:

paddle.sparse
---------------------

paddle.sparse 目录包含飞桨框架支持稀疏数据存储和计算相关的 API。具体如下:

- :ref:`稀疏 Tensor 创建 <about_sparse_tensor>`
- :ref:`稀疏 Tensor 运算 <about_sparse_math>`
- :ref:`稀疏组网类 <about_sparse_nn>`
- :ref:`稀疏组网类的函数式 API <about_sparse_nn_functional>`

.. _about_sparse_tensor:

稀疏 Tensor 创建
::::::::::::::::::::

.. csv-table::
:header: "API 名称", "API 功能"

" :ref:`paddle.sparse.sparse_coo_tensor <cn_api_paddle_sparse_coo_tensor>` ", "创建一个 COO 格式的 SparseTensor"
" :ref:`paddle.sparse.sparse_csr_tensor <cn_api_paddle_sparse_csr_tensor>` ", "创建一个 CSR 格式的 SparseTensor"

.. _about_sparse_math:

稀疏 Tensor 运算
::::::::::::::::::::

.. csv-table::
:header: "API 名称", "API 功能"

" :ref:`paddle.sparse.sin <cn_api_paddle_sparse_sin>` ", "对稀疏 Tensor 逐元素求正弦"
" :ref:`paddle.sparse.tan <cn_api_paddle_sparse_tan>` ", "对稀疏 Tensor 逐元素求正切"
" :ref:`paddle.sparse.asin <cn_api_paddle_sparse_asin>` ", "对稀疏 Tensor 逐元素求反正弦"
" :ref:`paddle.sparse.atan <cn_api_paddle_sparse_atan>` ", "对稀疏 Tensor 逐元素求反正切"
" :ref:`paddle.sparse.sinh <cn_api_paddle_sparse_sinh>` ", "对稀疏 Tensor 逐元素求双曲正弦"
" :ref:`paddle.sparse.tanh <cn_api_paddle_sparse_tanh>` ", "对稀疏 Tensor 逐元素求双曲正切"
" :ref:`paddle.sparse.asinh <cn_api_paddle_sparse_asinh>` ", "对稀疏 Tensor 逐元素求反双曲正弦"
" :ref:`paddle.sparse.atanh <cn_api_paddle_sparse_atanh>` ", "对稀疏 Tensor 逐元素求反双曲正切"
" :ref:`paddle.sparse.sqrt <cn_api_paddle_sparse_sqrt>` ", "对稀疏 Tensor 逐元素求算数平方根"
" :ref:`paddle.sparse.square <cn_api_paddle_sparse_square>` ", "对稀疏 Tensor 逐元素求平方"
" :ref:`paddle.sparse.log1p <cn_api_paddle_sparse_log1p>` ", "对稀疏 Tensor 逐元素计算 ln(x+1)"
" :ref:`paddle.sparse.abs <cn_api_paddle_sparse_abs>` ", "对稀疏 Tensor 逐元素求绝对值"
" :ref:`paddle.sparse.pow <cn_api_paddle_sparse_pow>` ", "对稀疏 Tensor 逐元素计算 x 的 y 次幂"
" :ref:`paddle.sparse.cast <cn_api_paddle_sparse_cast>` ", "对稀疏 Tensor 逐元素转换类型"
" :ref:`paddle.sparse.neg <cn_api_paddle_sparse_neg>` ", "对稀疏 Tensor 逐元素计算相反数"
" :ref:`paddle.sparse.deg2rad <cn_api_paddle_sparse_deg2rad>` ", "对稀疏 Tensor 逐元素从度转换为弧度"
" :ref:`paddle.sparse.rad2deg <cn_api_paddle_sparse_rad2deg>` ", "对稀疏 Tensor 逐元素从弧度转换为度"
" :ref:`paddle.sparse.expm1 <cn_api_paddle_sparse_expm1>` ", "对稀疏 Tensor 逐元素进行以自然数 e 为底的指数运算并减 1"
" :ref:`paddle.sparse.mv <cn_api_paddle_sparse_mv>` ", "稀疏矩阵乘向量,第一个参数为稀疏矩阵,第二个参数为稠密向量"
" :ref:`paddle.sparse.matmul <cn_api_paddle_sparse_matmul>` ", "稀疏矩阵乘,第一个参数为稀疏矩阵,第二个参数为稠密矩阵或者稀疏矩阵"
" :ref:`paddle.sparse.addmm <cn_api_paddle_sparse_addmm>` ", "稀疏矩阵乘与加法的组合运算"
" :ref:`paddle.sparse.masked_matmul <cn_api_paddle_sparse_masked_matmul>` ", "稀疏矩阵乘,第一、二个参数均为稠密矩阵,返回值为稀疏矩阵"
" :ref:`paddle.sparse.add <cn_api_paddle_sparse_add>` ", "对稀疏 Tensor 逐元素相加"
" :ref:`paddle.sparse.subtract <cn_api_paddle_sparse_subtract>` ", "对稀疏 Tensor 逐元素相减"
" :ref:`paddle.sparse.multiply <cn_api_paddle_sparse_multiply>` ", "对稀疏 Tensor 逐元素相乘"
" :ref:`paddle.sparse.divide <cn_api_paddle_sparse_divide>` ", "对稀疏 Tensor 逐元素相除"
" :ref:`paddle.sparse.is_same_shape <cn_api_paddle_sparse_is_same_shape>` ", "判断两个稀疏 Tensor 或稠密 Tensor 的 shape 是否一致"
" :ref:`paddle.sparse.reshape <cn_api_paddle_sparse_reshape>` ", "改变一个 SparseTensor 的形状"
" :ref:`paddle.sparse.coalesce<cn_api_paddle_sparse_coalesce>` ", "对 SparseCooTensor 进行排序并合并"
" :ref:`paddle.sparse.transpose <cn_api_paddle_sparse_transpose>` ", "在不改变数据的情况下改变 ``x`` 的维度顺序, 支持 COO 格式的多维 SparseTensor 以及 COO 格式的 2 维和 3 维 SparseTensor"

.. _about_sparse_nn:

稀疏组网类
::::::::::::::::::::

.. csv-table::
:header: "API 名称", "API 功能"

" :ref:`paddle.sparse.nn.ReLU <cn_api_paddle_sparse_nn_ReLU>` ", "激活层"
" :ref:`paddle.sparse.nn.ReLU6 <cn_api_paddle_sparse_nn_ReLU6>` ", "激活层"
" :ref:`paddle.sparse.nn.LeakyReLU <cn_api_paddle_sparse_nn_LeakyReLU>` ", "激活层"
" :ref:`paddle.sparse.nn.Softmax <cn_api_paddle_sparse_nn_Softmax>` ", "激活层"
" :ref:`paddle.sparse.nn.Conv3D <cn_api_paddle_sparse_nn_Conv3D>` ", "三维卷积层"
" :ref:`paddle.sparse.nn.SubmConv3D <cn_api_paddle_sparse_nn_SubmConv3D>` ", "子流形三维卷积层"
" :ref:`paddle.sparse.nn.BatchNorm<cn_api_paddle_sparse_nn_BatchNorm>` ", " Batch Normalization 层"
" :ref:`paddle.sparse.nn.SyncBatchNorm<cn_api_paddle_sparse_nn_SyncBatchNorm>` ", " Synchronized Batch Normalization 层"
" :ref:`paddle.sparse.nn.MaxPool3D<cn_api_paddle_sparse_nn_MaxPool3D>` ", "三维最大池化层"

.. _about_sparse_nn_functional:

稀疏组网类函数式 API
::::::::::::::::::::

.. csv-table::
:header: "API 名称", "API 功能"

" :ref:`paddle.sparse.nn.functional.relu <cn_api_paddle_sparse_nn_functional_relu>` ", "激活函数"
" :ref:`paddle.sparse.nn.functional.relu6 <cn_api_paddle_sparse_nn_functional_relu6>` ", "激活函数"
" :ref:`paddle.sparse.nn.functional.leaky_relu <cn_api_paddle_sparse_nn_functional_leaky_relu>` ", "激活函数"
" :ref:`paddle.sparse.nn.functional.softmax <cn_api_paddle_sparse_nn_functional_softmax>` ", "激活函数"
" :ref:`paddle.sparse.nn.functional.attention <cn_api_paddle_sparse_nn_functional_attention>` ", "稀疏 attention 函数"
" :ref:`paddle.sparse.nn.functional.conv3d <cn_api_paddle_sparse_nn_functional_conv3d>` ", "三维卷积函数"
" :ref:`paddle.sparse.nn.functional.subm_conv3d <cn_api_paddle_sparse_nn_functional_subm_conv3d>` ", "子流形三维卷积函数"
" :ref:`paddle.sparse.nn.functional.max_pool3d <cn_api_paddle_sparse_nn_functional_max_pool3d>` ", "三维最大池化函数"
29 changes: 29 additions & 0 deletions docs/api/paddle/sparse/abs_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
.. _cn_api_paddle_sparse_abs:

abs
-------------------------------

.. py:function:: paddle.sparse.abs(x, name=None)


逐元素计算输入 :attr:`x` 的绝对值,要求 输入 :attr:`x` 为 `SparseCooTensor` 或 `SparseCsrTensor` 。

数学公式:

.. math::
out = |x|

参数
:::::::::
- **x** (SparseTensor) - 输入的稀疏 Tensor,可以为 Coo 或 Csr 格式,数据类型为 float32、float64。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
多维稀疏 Tensor, 数据类型和稀疏格式与 :attr:`x` 相同 。


代码示例
:::::::::

COPY-FROM: paddle.sparse.abs
Original file line number Diff line number Diff line change
@@ -1,10 +1,9 @@
.. _cn_api_paddle_incubate_sparse_add:
.. _cn_api_paddle_sparse_add:

add
-------------------------------

.. py:function:: paddle.incubate.sparse.add(x, y, name=None)

.. py:function:: paddle.sparse.add(x, y, name=None)


输入 :attr:`x` 与输入 :attr:`y` 逐元素相加,并将各个位置的输出元素保存到返回结果中。
Expand All @@ -14,10 +13,10 @@ add
等式为:

.. math::
Out = X + Y
out = x + y

- :math:`X`:多维稀疏 Tensor。
- :math:`Y`:多维稀疏 Tensor。
- :math:`x`:多维稀疏 Tensor。
- :math:`y`:多维稀疏 Tensor。

参数
:::::::::
Expand All @@ -33,4 +32,4 @@ add
代码示例
:::::::::

COPY-FROM: paddle.incubate.sparse.add
COPY-FROM: paddle.sparse.add
49 changes: 49 additions & 0 deletions docs/api/paddle/sparse/addmm_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,49 @@
.. _cn_api_paddle_sparse_addmm:

addmm
-------------------------------

.. py:function:: paddle.sparse.addmm(input, x, y, beta=1.0, alpha=1.0, name=None)

.. note::
该 API 从 `CUDA 11.0` 开始支持。

对输入 :attr:`x` 与输入 :attr:`y` 求稀疏矩阵乘法,并将 `input` 加到计算结果上。

数学公式:

.. math::
out = alpha * x * y + beta * input

输入、输出的格式对应关系如下:

.. note::

input[SparseCsrTensor] + x[SparseCsrTensor] @ y[SparseCsrTensor] -> out[SparseCsrTensor]

input[DenseTensor] + x[SparseCsrTensor] @ y[DenseTensor] -> out[DenseTensor]

input[SparseCooTensor] + x[SparseCooTensor] @ y[SparseCooTensor] -> out[SparseCooTensor]

input[DenseTensor] + x[SparseCooTensor] @ y[DenseTensor] -> out[DenseTensor]

该 API 支持反向传播,`input` 、 `x` 、 `y` 的维度相同且>=2D,不支持自动广播。

参数
:::::::::
- **input** (SparseTensor|DenseTensor) - 输入 Tensor,可以为 Coo 或 Csr 格式 或 DenseTensor。数据类型为 float32、float64。
- **x** (SparseTensor) - 输入 Tensor,可以为 Coo 或 Csr 格式。数据类型为 float32、float64。
- **y** (SparseTensor|DenseTensor) - 输入 Tensor,可以为 Coo 或 Csr 格式 或 DenseTensor。数据类型为 float32、float64。
- **beta** (float, 可选) - `input` 的系数。默认:1.0。
- **alpha** (float, 可选) - `x * y` 的系数。默认:1.0。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
SparseTensor|DenseTensor: 其 Tensor 类型、dtype、shape 与 `input` 相同。


代码示例
:::::::::

COPY-FROM: paddle.sparse.addmm
29 changes: 29 additions & 0 deletions docs/api/paddle/sparse/asin_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
.. _cn_api_paddle_sparse_asin:

asin
-------------------------------

.. py:function:: paddle.sparse.asin(x, name=None)


逐元素计算输入 :attr:`x` 的反正弦,要求 输入 :attr:`x` 为 `SparseCooTensor` 或 `SparseCsrTensor` 。

数学公式:

.. math::
out = asin(x)

参数
:::::::::
- **x** (SparseTensor) - 输入的稀疏 Tensor,可以为 Coo 或 Csr 格式,数据类型为 float32、float64。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
多维稀疏 Tensor, 数据类型和稀疏格式与 :attr:`x` 相同 。


代码示例
:::::::::

COPY-FROM: paddle.sparse.asin
29 changes: 29 additions & 0 deletions docs/api/paddle/sparse/asinh_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
.. _cn_api_paddle_sparse_asinh:

asinh
-------------------------------

.. py:function:: paddle.sparse.asinh(x, name=None)


逐元素计算输入 :attr:`x` 的反双曲正弦,要求 输入 :attr:`x` 为 `SparseCooTensor` 或 `SparseCsrTensor` 。

数学公式:

.. math::
out = asinh(x)

参数
:::::::::
- **x** (SparseTensor) - 输入的稀疏 Tensor,可以为 Coo 或 Csr 格式,数据类型为 float32、float64。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
多维稀疏 Tensor, 数据类型和稀疏格式与 :attr:`x` 相同 。


代码示例
:::::::::

COPY-FROM: paddle.sparse.asinh
29 changes: 29 additions & 0 deletions docs/api/paddle/sparse/atan_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
.. _cn_api_paddle_sparse_atan:

atan
-------------------------------

.. py:function:: paddle.sparse.atan(x, name=None)


逐元素计算输入 :attr:`x` 的反正切,要求 输入 :attr:`x` 为 `SparseCooTensor` 或 `SparseCsrTensor` 。

数学公式:

.. math::
out = atan(x)

参数
:::::::::
- **x** (SparseTensor) - 输入的稀疏 Tensor,可以为 Coo 或 Csr 格式,数据类型为 float32、float64。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
多维稀疏 Tensor, 数据类型和稀疏格式与 :attr:`x` 相同 。


代码示例
:::::::::

COPY-FROM: paddle.sparse.atan
29 changes: 29 additions & 0 deletions docs/api/paddle/sparse/atanh_cn.rst
Original file line number Diff line number Diff line change
@@ -0,0 +1,29 @@
.. _cn_api_paddle_sparse_atanh:

atanh
-------------------------------

.. py:function:: paddle.sparse.atanh(x, name=None)


逐元素计算输入 :attr:`x` 的反双曲正切,要求 输入 :attr:`x` 为 `SparseCooTensor` 或 `SparseCsrTensor` 。

数学公式:

.. math::
out = atanh(x)

参数
:::::::::
- **x** (SparseTensor) - 输入的稀疏 Tensor,可以为 Coo 或 Csr 格式,数据类型为 float32、float64。
- **name** (str,可选) - 具体用法请参见 :ref:`api_guide_Name`,一般无需设置,默认值为 None。

返回
:::::::::
多维稀疏 Tensor, 数据类型和稀疏格式与 :attr:`x` 相同 。


代码示例
:::::::::

COPY-FROM: paddle.sparse.atanh
Loading