chore: import upstream snapshot with attribution

paddle/scripts/CMakeLists.txt

@@ -0,0 +1,13 @@
+configure_file(submit_local.sh.in paddle @ONLY)
+
+install(
+  FILES ${CMAKE_CURRENT_BINARY_DIR}/paddle
+  DESTINATION bin
+  PERMISSIONS
+    OWNER_EXECUTE
+    OWNER_WRITE
+    OWNER_READ
+    GROUP_EXECUTE
+    GROUP_READ
+    WORLD_EXECUTE
+    WORLD_READ)

paddle/scripts/README.md

@@ -0,0 +1,175 @@
+# Building PaddlePaddle
+
+## Goals
+
+We want to make the building procedures:
+
+1. Static, can reproduce easily.
+1. Generate python `whl` packages that can be widely use cross many distributions.
+1. Build different binaries per release to satisfy different environments:
+    - Binaries for different CUDA and CUDNN versions, like CUDA 7.5, 8.0, 9.0
+    - Binaries containing only capi
+    - Binaries for python with wide unicode support or not.
+1. Build docker images with PaddlePaddle pre-installed, so that we can run
+PaddlePaddle applications directly in docker or on Kubernetes clusters.
+
+To achieve this, we maintain a dockerhub repo:https://hub.docker.com/r/paddlepaddle/paddle
+which provides pre-built environment images to build PaddlePaddle and generate corresponding `whl`
+binaries.(**We strongly recommend building paddlepaddle in our pre-specified Docker environment.**)
+
+## Development Workflow
+
+Here we describe how the workflow goes on.  We start from considering our daily development environment.
+
+Developers work on a computer, which is usually a laptop or desktop:
+
+<img src="doc/paddle-development-environment.png" width=500 />
+
+or, they might rely on a more sophisticated box (like with GPUs):
+
+<img src="doc/paddle-development-environment-gpu.png" width=500 />
+
+A principle here is that source code lies on the development computer (host) so that editors like Eclipse can parse the source code to support auto-completion.
+
+## Build With Docker
+
+### Build Environments
+
+The latest pre-built build environment images are:
+
+| Image | Tag |
+| ----- | --- |
+| paddlepaddle/paddle | latest-dev |
+
+### Start Build
+
+```bash
+git clone https://github.com/PaddlePaddle/Paddle.git
+cd Paddle
+./paddle/scripts/paddle_docker_build.sh build
+```
+
+After the build finishes, you can get output `whl` package under
+`build/python/dist`.
+
+This command will download the most recent dev image from docker hub, start a container in the backend and then run the build script `/paddle/paddle/scripts/paddle_build.sh build` in the container.
+The container mounts the source directory on the host into `/paddle`.
+When it writes to `/paddle/build` in the container, it writes to `$PWD/build` on the host indeed.
+
+### Build Options
+
+Users can specify the following Docker build arguments with either "ON" or "OFF" value:
+
+| Option | Default | Description |
+| ------ | -------- | ----------- |
+| `WITH_GPU` | OFF | Generates NVIDIA CUDA GPU code and relies on CUDA libraries. |
+| `WITH_AVX` | OFF | Set to "ON" to enable AVX support. |
+| `WITH_TESTING` | OFF | Build unit tests binaries. |
+| `WITH_MKL` | ON | Build with [Intel® oneMKL](https://www.intel.com/content/www/us/en/developer/tools/oneapi/onemkl.html) and [Intel® oneDNN](https://github.com/uxlfoundation/oneDNN) support. |
+| `WITH_PYTHON` | ON | Build with python support. Turn this off if build is only for capi. |
+| `WITH_STYLE_CHECK` | ON | Check the code style when building. |
+| `PYTHON_ABI` | "" | Build for different python ABI support, can be cp27-cp27m or cp27-cp27mu |
+| `RUN_TEST` | OFF | Run unit test immediately after the build. |
+
+## Docker Images
+
+You can get the latest PaddlePaddle docker images by
+`docker pull paddlepaddle/paddle:<version>` or build one by yourself.
+
+### Official Docker Releases
+
+Official docker images at
+[here](https://hub.docker.com/r/paddlepaddle/paddle/tags/),
+you can choose either latest or images with a release tag like `0.10.0`,
+Currently available tags are:
+
+|   Tag  | Description |
+| ------ | --------------------- |
+| latest | latest CPU only image |
+| latest-gpu | latest binary with GPU support |
+| 0.10.0 | release 0.10.0 CPU only binary image |
+| 0.10.0-gpu | release 0.10.0 with GPU support |
+
+### Build Your Own Image
+
+Build PaddlePaddle docker images are quite simple since PaddlePaddle can
+be installed by just running `pip install`. A sample `Dockerfile` is:
+
+```dockerfile
+FROM nvidia/cuda:7.5-cudnn5-runtime-centos6
+RUN yum install -y centos-release-SCL
+RUN yum install -y python27
+# This whl package is generated by previous build steps.
+ADD python/dist/paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl /
+RUN pip install /paddlepaddle-0.10.0-cp27-cp27mu-linux_x86_64.whl && rm -f /*.whl
+```
+
+Then build the image by running `docker build -t [REPO]/paddle:[TAG] .` under
+the directory containing your own `Dockerfile`.
+
+We also release a script and Dockerfile for building PaddlePaddle docker images
+across different cuda versions. To build these docker images, run:
+
+```bash
+bash ./build_docker_images.sh
+docker build -t [REPO]/paddle:tag -f [generated_docker_file] .
+```
+
+- NOTE: note that you can choose different base images for your environment, you can find all the versions [here](https://hub.docker.com/r/nvidia/cuda/).
+
+### Use Docker Images
+
+Suppose that you have written an application program `train.py` using
+PaddlePaddle, we can test and run it using docker:
+
+```bash
+docker run --rm -it -v $PWD:/work paddlepaddle/paddle /work/a.py
+```
+
+But this works only if all dependencies of `train.py` are in the production image. If this is not the case, we need to build a new Docker image from the production image and with more dependencies installs.
+
+### Run PaddlePaddle Book In Docker
+
+Our [book repo](https://github.com/paddlepaddle/book) also provide a docker
+image to start a jupiter notebook inside docker so that you can run this book
+using docker:
+
+```bash
+docker run -d -p 8888:8888 paddlepaddle/book
+```
+
+Please refer to https://github.com/paddlepaddle/book if you want to build this
+docker image by your self.
+
+### Run Distributed Applications
+
+In our [API design doc](https://github.com/PaddlePaddle/Paddle/blob/develop/doc/design/api.md#distributed-training), we proposed an API that starts a distributed training job on a cluster.  This API need to build a PaddlePaddle application into a Docker image as above and calls kubectl to run it on the cluster.  This API might need to generate a Dockerfile look like above and call `docker build`.
+
+Of course, we can manually build an application image and launch the job using the kubectl tool:
+
+```bash
+docker build -f some/Dockerfile -t myapp .
+docker tag myapp me/myapp
+docker push
+kubectl ...
+```
+
+
+## More Options
+
+### Build Without Docker
+
+Follow the *Dockerfile* in the paddlepaddle repo to set up your local dev environment and run:
+
+```bash
+./paddle/scripts/paddle_build.sh build
+```
+
+### Additional Tasks
+
+You can get the help menu for the build scripts by running with no options:
+
+```bash
+./paddle/scripts/paddle_build.sh
+or ./paddle/scripts/paddle_docker_build.sh
+```

paddle/scripts/compile_requirements.txt

@@ -0,0 +1,7 @@
+setuptools==57.4.0 ; python_version <= '3.11' and sys_platform == 'win32'
+setuptools==69.0.2 ; python_version > '3.11' and sys_platform == 'win32'
+wheel==0.45.1 ; sys_platform == 'win32'
+pyyaml ; sys_platform == 'win32'
+wget ; sys_platform == 'win32'
+jinja2 ; sys_platform == 'linux'
+pybind11_stubgen ; sys_platform == 'linux'

paddle/scripts/docker/root/.bashrc

@@ -0,0 +1,40 @@
+# Locales
+
+export LC_ALL=en_US.UTF-8
+export LANG=en_US.UTF-8
+export LANGUAGE=en_US.UTF-8
+
+# Aliases
+
+alias rm='rm -i'
+alias cp='cp -i'
+alias mv='mv -i'
+
+alias ls='ls -hFG'
+alias l='ls -lF'
+alias ll='ls -alF'
+alias lt='ls -ltrF'
+alias ll='ls -alF'
+alias lls='ls -alSrF'
+alias llt='ls -altrF'
+
+# Colorize directory listing
+
+alias ls="ls -ph --color=auto"
+
+# Colorize grep
+
+if echo hello|grep --color=auto l >/dev/null 2>&1; then
+  export GREP_OPTIONS="--color=auto" GREP_COLOR="1;31"
+fi
+
+# Shell
+
+export CLICOLOR="1"
+
+YELLOW="\[\033[1;33m\]"
+NO_COLOUR="\[\033[0m\]"
+GREEN="\[\033[1;32m\]"
+WHITE="\[\033[1;37m\]"
+
+export PS1="\[\033[1;33m\]λ $WHITE\h $GREEN\w $NO_COLOUR"

paddle/scripts/docker/root/.gitconfig

@@ -0,0 +1,43 @@
+[user]
+  name =
+  email =
+
+[alias]
+  st = status --branch --short
+  ci = commit
+  br = branch
+  co = checkout
+  df = diff
+  l = log --pretty=format:\"%h %ad | %s%d [%an]\" --graph --date=short
+  ll = log --stat
+
+[merge]
+  tool = vimdiff
+
+[core]
+  excludesfile = ~/.gitignore
+  editor = vim
+
+[color]
+  branch = auto
+  diff = auto
+  status = auto
+
+[color "branch"]
+  current = yellow reverse
+  local = yellow
+  remote = green
+
+[color "diff"]
+  meta = yellow bold
+  frag = magenta bold
+  old = red bold
+  new = green bold
+
+[color "status"]
+  added = yellow
+  changed = green
+  untracked = cyan
+
+[push]
+  default = matching

paddle/scripts/get_target_size.py

@@ -0,0 +1,51 @@
+# Copyright (c) 2024 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import subprocess
+import sys
+
+
+class HandleTarget:
+    def __init__(self) -> None:
+        argv = sys.argv
+        len_ = len(argv)
+        fromPath = './' if len_ <= 1 else argv[1]
+        cmd = f"find {fromPath} -name '*.o' -not -path './third_party/*' | xargs du -sch > log"
+        subprocess.run(cmd, shell=True)
+
+    def calcuSize(self, item):
+        size = float(item[:-1])
+        res = size * 1024 if item.find('G') >= 0 else size
+        return size / 1024 if item.find('K') >= 0 else res
+
+    def getSize(self):
+        ctx = self.getDatas()
+        sum = 0
+        for item in ctx:
+            if item.find('total') >= 0:
+                item = item.split('\t')[0]
+                sum += self.calcuSize(item)
+
+        print(f"Total size is {sum} M (without third_party)")
+
+    def getDatas(self):
+        with open('log', 'r') as file:
+            ctx = file.read()
+        ctx = ctx.split('\n')
+        return ctx
+
+
+if __name__ == '__main__':
+    handler = HandleTarget()
+    handler.getSize()

paddle/scripts/installation_validate.py

@@ -0,0 +1,18 @@
+# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import paddle as pd
+
+pd.utils.run_check()
+print(pd.__version__)

paddle/scripts/musl_build/Dockerfile

@@ -0,0 +1,71 @@
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+ARG PYTHON_VERSION=3.7
+
+FROM python:${PYTHON_VERSION}-alpine3.11
+
+USER root
+
+WORKDIR /root
+
+VOLUME /root/.ccache
+
+VOLUME /root/.cache
+
+RUN apk update
+
+RUN apk add --no-cache \
+    g++ gfortran make cmake patchelf git ccache
+
+ARG package
+
+RUN if [ "$package" ]; then \
+        set -e; \
+        pkgs=$(echo "$package" | base64 -d -); \
+        echo ">>> decode package:"; \
+        echo "$pkgs"; \
+        for nm in $pkgs; do \
+            echo ">>> install package: $nm"; \
+            apk add --no-cache --force-overwrite "$nm"; \
+        done; \
+    fi
+
+ARG requirement
+ARG requirement_ut
+ARG pip_index
+
+RUN if [ "$requirement" ]; then \
+        set -e; \
+        echo "$requirement" | base64 -d - > "requirement.txt"; \
+        echo ">>> decode requirement:"; \
+        cat "requirement.txt"; \
+        echo ">>> install python requirement:"; \
+        PIP_ARGS="--timeout 300"; \
+        if [ "$pip_index" ]; then \
+            PIP_DOMAIN=$(echo "$pip_index" | awk -F/ '{print $3}'); \
+            PIP_ARGS="$PIP_ARGS -i $pip_index --trusted-host $PIP_DOMAIN"; \
+            echo ">>> pip index: $pip_index"; \
+        fi; \
+        pip3 install $PIP_ARGS -r "requirement.txt"; \
+        rm -f "requirement.txt"; \
+        if [ "$requirement_ut" ]; then \
+            echo "$requirement_ut" | base64 -d - > "requirement_ut.txt"; \
+            cat "requirement_ut.txt"; \
+            pip3 install $PIP_ARGS -r "requirement_ut.txt"; \
+            rm -f "requirement_ut.txt"; \
+        fi; \
+    fi
+
+
+ENTRYPOINT [ "/bin/sh" ]

paddle/scripts/musl_build/README.md

@@ -0,0 +1,121 @@
+Paddle for Linux-musl Usage Guide
+===========================================
+
+# Introduction
+Paddle can be built for linux-musl such as alpine, and be used in libos-liked SGX TEE environment. Currently supported commercial product TEE Scone, and community maintanced TEE Occlum. We also working on to support open source TEE Graphene.
+
+
+# Build Automatically
+1. clone paddle source from github
+
+```bash
+git clone https://github.com/PaddlePaddle/Paddle.git
+```
+
+2. setup build directory
+
+```bash
+# enter paddle directory
+cd  ./Paddle
+
+# create and enter building directory
+mkdir -p build && cd build
+```
+
+3. build docker for compiling. use environment HTTP_PROXY/HTTPS_PROXY for proxy setup.
+
+```bash
+# setup proxy address, when the speed of internet is not good.
+# export HTTP_PROXY='http://127.0.0.1:8080'
+# export HTTPS_PROXY='https://127.0.0.1:8080'
+
+# invoke build script
+../paddle/scripts/musl_build/build_docker.sh
+```
+
+4. compile paddle in previous built docker. proxy setup method is same as previous step.
+
+
+```bash
+# setup proxy address, when the speed of internet is not good.
+# export HTTP_PROXY='http://127.0.0.1:8080'
+# export HTTPS_PROXY='https://127.0.0.1:8080'
+
+# invoke build paddle script
+# all arguments, such as -j8 optional, is past to make procedure.
+../paddle/scripts/musl_build/build_paddle.sh -j8
+
+# find output wheel package
+# output wheel packages will save to "./output" directory.
+ls ./output/*.whl
+```
+
+# Build Manually
+
+1. start up the building docker, and enter the shell in the container
+```bash
+# checkout paddle source code
+git clone https://github.com/PaddlePaddle/Paddle.git
+
+# enter paddle directory
+cd ./Paddle
+
+# build docker image
+../paddle/scripts/musl_build/build_docker.sh
+
+# enter the container interactive shell
+BUILD_MAN=1 ../paddle/scripts/musl_build/build_paddle.sh
+```
+
+2. type commands and compile source manually
+```sh
+# compile paddle by commands
+# paddle is mount to /paddle directory
+# working directory is /root
+mkdir build && cd build
+
+# install python requirement
+pip install -r /paddle/python/requirements.txt
+
+# configure project with cmake
+cmake -DWITH_MUSL=ON -DWITH_CRYPTO=OFF -DWITH_MKL=OFF -DWITH_GPU=OFF /paddle
+
+# run the make to build project.
+# the argument -j8 is optional to accelerate compiling.
+make -j8
+```
+
+# Scripts
+1. **build_docker.sh**
+   compiling docker building script. it use alpine linux 3.10 as musl linux build environment. it will try to install all the compiling tools, development packages, and python requirements for paddle musl compiling.
+
+    environment variables:
+   - PYTHON_VERSION: the version of python used for image building, default=3.8.
+   - WITH_PRUNE_DAYS: prune old docker images, with days limitation.
+   - WITH_REBUILD: force to rebuild the image, default=0.
+   - WITH_REQUIREMENT: build with the python requirements, default=1.
+   - WITH_UT_REQUIREMENT: build with the unit test requirements, default=0.
+   - WITH_PIP_INDEX: use custom pip index when pip install packages.
+   - ONLY_NAME: only print the docker name, and exit.
+   - HTTP_PROXY: use http proxy.
+   - HTTPS_PROXY: use https proxy.
+
+2. **build_paddle.sh** automatically or manually paddle building script. it will mount the root directory of paddle source to /paddle, and run compile procedure in /root/build directory. the output wheel package will save to the ./output directory relative to working directory.
+
+    environment variables:
+    - BUILD_MAN: build the paddle manually, default=0.
+    - WITH_TEST: build with unittest, and run unittest check, default=0.
+    - WITH_PRUNE_CONTAINER: remove the container after building, default=1.
+    - CTEST_*: CTEST flags used for unit test.
+    - FLAGS_*: build flags used for paddle building.
+    - HTTP_PROXY: use http proxy.
+    - HTTPS_PROXY: use https proxy.
+
+# Files
+- **build_docker.sh**: docker building script
+- **build_paddle.sh**: paddle building script
+- **build_inside.sh**: build_paddle.sh will invoke this script inside the docker for compiling.
+- **config.sh**: build config script for configure compiling option setting.
+- **Dockerfile**: build docker definition file.
+- **package.txt**: build required develop packages for alpine linux.
+- **README.md**: this file.

paddle/scripts/musl_build/config.sh

@@ -0,0 +1,41 @@
+#!/bin/bash
+
+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+CUR_DIR=$(dirname "${BASH_SOURCE[0]}")
+CUR_DIR=$(realpath "$CUR_DIR")
+
+# shellcheck disable=2034
+PADDLE_DIR=$(realpath "$CUR_DIR/../../../")
+MOUNT_DIR="/paddle"
+
+BUILD_DOCKERFILE="$CUR_DIR/Dockerfile"
+PACKAGE_REQ="$CUR_DIR/package.txt"
+
+PYTHON_REQ="python/requirements.txt"
+UNITTEST_REQ="python/unittest_py/requirements.txt"
+
+function chksum(){
+    cat $* | md5sum - | cut -b-8
+}
+
+# shellcheck disable=2034
+BUILD_NAME="paddle-musl-build"
+BUILD_TAG=$(chksum "$BUILD_DOCKERFILE" "$PACKAGE_REQ")
+BUILD_IMAGE="$BUILD_NAME:$BUILD_TAG"
+BUILD_CONTAINER="$BUILD_NAME-$(date +%Y%m%d-%H%M%S)"
+
+CCACHE_DIR="${CCACHE_DIR-${HOME}/.paddle-musl/ccache}"
+CACHE_DIR="${CACHE_DIR-${HOME}/.paddle-musl/cache}"

paddle/scripts/musl_build/package.txt

@@ -0,0 +1,9 @@
+linux-headers
+freetype-dev
+libjpeg-turbo-dev
+zlib-dev
+lapack-dev
+openblas-dev
+openssl-dev
+libuv-dev
+graphviz

paddle/scripts/paddle_docker_build.sh

@@ -0,0 +1,71 @@
+#!/usr/bin/env bash
+
+# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+function start_build_docker() {
+    docker pull $IMG
+
+    apt_mirror='s#http://archive.ubuntu.com/ubuntu#mirror://mirrors.ubuntu.com/mirrors.txt#g'
+    DOCKER_ENV=$(cat <<EOL
+        -e FLAGS_fraction_of_gpu_memory_to_use=0.15 \
+        -e CTEST_OUTPUT_ON_FAILURE=1 \
+        -e CTEST_PARALLEL_LEVEL=1 \
+        -e APT_MIRROR=${apt_mirror} \
+        -e WITH_GPU=ON \
+        -e CUDA_ARCH_NAME=Auto \
+        -e WITH_AVX=ON \
+        -e WITH_TESTING=ON \
+        -e WITH_COVERAGE=ON \
+        -e COVERALLS_UPLOAD=ON \
+        -e WITH_DEB=OFF \
+        -e CMAKE_BUILD_TYPE=RelWithDebInfo \
+        -e PADDLE_FRACTION_GPU_MEMORY_TO_USE=0.15 \
+        -e CUDA_VISIBLE_DEVICES=0,1 \
+        -e WITH_DISTRIBUTE=ON \
+        -e RUN_TEST=ON
+EOL
+    )
+
+    DOCKER_CMD="nvidia-docker"
+    if ! [ -x "$(command -v ${DOCKER_CMD})" ]; then
+        DOCKER_CMD="docker"
+    fi
+    if [ ! -d "${HOME}/.ccache" ]; then
+        mkdir ${HOME}/.ccache
+    fi
+    set -ex
+    ${DOCKER_CMD} run -it \
+        ${DOCKER_ENV} \
+        -e SCRIPT_NAME=$0 \
+        -e CONTENT_DEC_PASSWD=$CONTENT_DEC_PASSWD \
+        -e TRAVIS_BRANCH=$TRAVIS_BRANCH \
+        -e TRAVIS_PULL_REQUEST=$TRAVIS_PULL_REQUEST \
+        -v $PADDLE_ROOT:/paddle \
+        -v ${HOME}/.ccache:/root/.ccache \
+        -w /paddle \
+        $IMG \
+        paddle/scripts/paddle_build.sh $@
+    set +x
+}
+
+function main() {
+    DOCKER_REPO="paddlepaddle/paddle"
+    VERSION="latest-dev"
+    PADDLE_ROOT="$( cd "$( dirname "${BASH_SOURCE[0]}")/../../" && pwd )"
+    IMG=${DOCKER_REPO}:${VERSION}
+    start_build_docker $@
+}
+
+main $@

paddle/scripts/submit_local.sh.in

@@ -0,0 +1,169 @@
+#!/bin/bash
+
+function version(){
+        echo "PaddlePaddle @PADDLE_VERSION@, compiled with"
+        echo "    with_avx: @WITH_AVX@"
+        echo "    with_gpu: @WITH_GPU@"
+        echo "    with_mkl: @WITH_MKL@"
+        echo "    with_mkldnn: @WITH_ONEDNN@"
+        echo "    with_python: @WITH_PYTHON@"
+}
+
+function ver2num() {
+  set -e
+  # convert version to number.
+  if [ -z "$1" ]; then # empty argument
+    printf "%03d%03d%03d%03d%03d" 0
+  else
+    local VERN=$(echo $1 | sed 's#v##g' | sed 's#\.# #g' \
+        | sed 's#a# 0 #g' | sed 's#b# 1 #g' | sed 's#rc# 2 #g')
+    if [ `echo $VERN | wc -w` -eq 3 ] ; then
+      printf "%03d%03d%03d%03d%03d" $VERN 999 999
+    else
+      printf "%03d%03d%03d%03d%03d" $VERN
+    fi
+  fi
+  set +e
+}
+
+function cpu_config() {
+  # auto set KMP_AFFINITY and OMP_DYNAMIC from Hyper Threading Status
+  # only when MKL enabled
+  if [ "@WITH_MKL@" == "OFF" ]; then
+    return 0
+  fi
+  platform="`uname -s`"
+  ht=0
+  if [ $platform == "Linux" ]; then
+    ht=`lscpu |grep "per core"|awk -F':' '{print $2}'|xargs`
+  elif [ $platform == "Darwin" ]; then
+    if [ `sysctl -n hw.physicalcpu` -eq `sysctl -n hw.logicalcpu` ]; then
+      # HT is OFF
+      ht=1
+    fi
+  else
+    return 0
+  fi
+  if [ $ht -eq 1 ]; then # HT is OFF
+    if [ -z "$KMP_AFFINITY" ]; then
+      export KMP_AFFINITY="granularity=fine,compact,0,0"
+    fi
+    if [ -z "$OMP_DYNAMIC" ]; then
+      export OMP_DYNAMIC="FALSE"
+    fi
+  else # HT is ON
+    if [ -z "$KMP_AFFINITY" ]; then
+      export KMP_AFFINITY="granularity=fine,compact,1,0"
+    fi
+    if [ -z "$OMP_DYNAMIC" ]; then
+      export OMP_DYNAMIC="True"
+    fi
+  fi
+}
+
+function threads_config() {
+  # auto set OMP_NUM_THREADS and MKL_NUM_THREADS
+  # according to trainer_count and total processors
+  # only when MKL enabled
+  # auto set OPENBLAS_NUM_THREADS when do not use MKL
+  platform="`uname -s`"
+  processors=0
+  if [ $platform == "Linux" ]; then
+    processors=`grep "processor" /proc/cpuinfo|sort -u|wc -l`
+  elif [ $platform == "Darwin" ]; then
+    processors=`sysctl -n hw.logicalcpu`
+  else
+    return 0
+  fi
+  trainers=`grep -Eo 'trainer_count.[0-9]+' <<< "$@" |grep -Eo '[0-9]+'|xargs`
+  if [ -z $trainers ]; then
+    trainers=1
+  fi
+  threads=$((processors / trainers))
+  if [ $threads -eq 0 ]; then
+    threads=1
+  fi
+  if [ "@WITH_MKL@" == "ON" ]; then
+    if [ -z "$OMP_NUM_THREADS" ]; then
+      export OMP_NUM_THREADS=$threads
+    fi
+    if [ -z "$MKL_NUM_THREADS" ]; then
+      export MKL_NUM_THREADS=$threads
+    fi
+  else
+    if [ -z "$OPENBLAS_NUM_THREADS" ]; then
+      export OPENBLAS_NUM_THREADS=$threads
+    fi
+    if [ $threads -gt 1 ] && [ -z "$OPENBLAS_MAIN_FREE" ]; then
+      export OPENBLAS_MAIN_FREE=1
+    fi
+  fi
+
+}
+
+PADDLE_CONF_HOME="$HOME/.config/paddle"
+mkdir -p ${PADDLE_CONF_HOME}
+
+if [ -z "${PADDLE_NO_STAT+x}" ]; then
+    SERVER_VER=`curl -m 5 -X POST --data content="{ \"version\": \"@PADDLE_VERSION@\" }"\
+        -b ${PADDLE_CONF_HOME}/paddle.cookie \
+        -c ${PADDLE_CONF_HOME}/paddle.cookie \
+        http://api.paddlepaddle.org/version 2>/dev/null`
+    if [ $? -eq 0 ] && [ "$(ver2num @PADDLE_VERSION@)" -lt  $(ver2num $SERVER_VER) ]; then
+      echo "Paddle release a new version ${SERVER_VER}, you can get the install package in http://www.paddlepaddle.org"
+    fi
+fi
+
+PADDLE_BIN_PATH="$( cd "$( dirname "${BASH_SOURCE[0]}" )" && pwd )"
+
+if [ ! -z "${DEBUGGER}" ]; then
+    echo "Using debug command ${DEBUGGER}"
+fi
+
+CUDNN_LIB_PATH="@CUDNN_LIB_PATH@"
+
+if [ ! -z "${CUDNN_LIB_PATH}" ]; then
+    export LD_LIBRARY_PATH=${CUDNN_LIB_PATH}:${LD_LIBRARY_PATH}
+fi
+
+export PYTHONPATH=${PWD}:${PYTHONPATH}
+
+
+# Check python lib installed or not.
+pip --help > /dev/null
+if [ $? -ne 0 ]; then
+    echo "pip should be installed to run paddle."
+    exit 1
+fi
+
+if [ "@WITH_GPU@" == "ON" ]; then
+    PADDLE_NAME="paddlepaddle-gpu"
+else
+    PADDLE_NAME="paddlepaddle"
+fi
+
+INSTALLED_VERSION=`pip freeze 2>/dev/null | grep "^${PADDLE_NAME}==" | sed 's/.*==//g'`
+
+if [ -z "${INSTALLED_VERSION}" ]; then
+   INSTALLED_VERSION="0.0.0"  # not installed
+fi
+cat <<EOF | python -
+from distutils.version import LooseVersion
+import sys
+if LooseVersion("${INSTALLED_VERSION}") < LooseVersion("@PADDLE_VERSION@"):
+  sys.exit(1)
+else:
+  sys.exit(0)
+EOF
+
+cpu_config
+# echo $KMP_AFFINITY $OMP_DYNAMIC
+
+case "$1" in
+    "version")
+        version
+        ;;
+    *)
+        version
+        ;;
+ esac

paddle/scripts/windows_build/build.bat

@@ -0,0 +1,187 @@
+@ECHO OFF
+SETLOCAL
+set source_path=%1
+set PYTHON_DIR=%2
+set WITH_GPU=%3
+set WITH_MKL=%4
+set ON_INFER=%5
+set PADDLE_VERSION=%6
+set BATDIR=%7
+set CUDA_DIR=%8
+
+set RETRY_TIMES=3
+
+set CUDA_DIR_WIN=%CUDA_DIR:/=\%
+set PATH=%CUDA_DIR_WIN%\nvvm\bin\;%CUDA_DIR_WIN%\bin;%PATH%
+
+
+for /f "tokens=1,2,* delims=\\" %%a in ("%PYTHON_DIR%") do (
+	set c1=%%a
+	set c2=%%b
+)
+set PYTHONV=%c2%
+
+echo %CUDA_DIR% | findstr 10.0 > NULL
+if %errorlevel% == 0 (set PADDLE_VERSION=%PADDLE_VERSION%
+set CUDAV=v10.0)
+echo %CUDA_DIR% | findstr 9.2 > NULL
+if %errorlevel% == 0 (set PADDLE_VERSION=%PADDLE_VERSION%.post97
+set CUDAV=v9.2)
+echo %CUDA_DIR% | findstr 9.0 > NULL
+if %errorlevel% == 0 (set PADDLE_VERSION=%PADDLE_VERSION%.post97
+set CUDAV=v9.0)
+echo %CUDA_DIR% | findstr 8.0 > NULL
+if %errorlevel% == 0 (set PADDLE_VERSION=%PADDLE_VERSION%.post87
+set CUDAV=v8.0)
+set PLAT=GPU
+if "%WITH_GPU%"=="OFF" (
+    set PLAT=CPU
+    set CUDAV=CPU
+)
+
+if "%WITH_MKL%"=="ON" (
+    set BLAS=MKL
+) else (
+    set BLAS=OPEN
+)
+
+if "%ON_INFER%"=="ON" (
+    goto :INFERENCE_LIBRARY
+)
+
+echo "begin to do build noavx ..."
+
+set "dst_path=%source_path%\build_%PYTHONV%_%PLAT%_%BLAS%_%CUDAV%_noavx"
+
+if exist %dst_path% rmdir /q /s %dst_path%
+mkdir %dst_path%
+
+cd /d %dst_path%
+echo Current directory : %cd%
+
+call:rest_env
+
+echo cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DWITH_AVX=OFF -DPYTHON_INCLUDE_DIR=%PYTHON_DIR%\include\ -DPYTHON_LIBRARY=%PYTHON_DIR%\libs\ -DPYTHON_EXECUTABLE=%PYTHON_DIR%\python.exe -DCMAKE_BUILD_TYPE=Release -DWITH_TESTING=OFF -DWITH_PYTHON=ON -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR% -DCUDA_ARCH_NAME=All
+cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DWITH_AVX=OFF -DPYTHON_INCLUDE_DIR=%PYTHON_DIR%\include\ -DPYTHON_LIBRARY=%PYTHON_DIR%\libs\ -DPYTHON_EXECUTABLE=%PYTHON_DIR%\python.exe -DCMAKE_BUILD_TYPE=Release -DWITH_TESTING=OFF -DWITH_PYTHON=ON -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR% -DCUDA_ARCH_NAME=All
+
+set  MSBUILDDISABLENODEREUSE=1
+
+set BUILD_TYPE=NO_AVX
+call:Build
+
+REM -------------------------------------------------------------------------
+
+echo "begin to do build avx ..."
+set "dst_path=%source_path%\build_%PYTHONV%_%PLAT%_%BLAS%_%CUDAV%"
+
+if exist %dst_path% rmdir /q /s %dst_path%
+mkdir %dst_path%
+
+cd /d %dst_path%
+echo Current directory : %cd%
+
+call:rest_env
+
+echo cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DWITH_AVX=ON -DPYTHON_INCLUDE_DIR=%PYTHON_DIR%\include\ -DPYTHON_LIBRARY=%PYTHON_DIR%\libs\ -DPYTHON_EXECUTABLE=%PYTHON_DIR%\python.exe -DCMAKE_BUILD_TYPE=Release -DWITH_TESTING=OFF -DWITH_PYTHON=ON -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR% -DCUDA_ARCH_NAME=All
+cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DWITH_AVX=ON -DPYTHON_INCLUDE_DIR=%PYTHON_DIR%\include\ -DPYTHON_LIBRARY=%PYTHON_DIR%\libs\ -DPYTHON_EXECUTABLE=%PYTHON_DIR%\python.exe -DCMAKE_BUILD_TYPE=Release -DWITH_TESTING=OFF -DWITH_PYTHON=ON -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR% -DCUDA_ARCH_NAME=All
+
+set  MSBUILDDISABLENODEREUSE=1
+
+set BUILD_TYPE=AVX
+call:Build
+
+echo BUILD WHL PACKAGE COMPLETE
+goto :END
+REM -------------------------------------------------------------------------
+
+:INFERENCE_LIBRARY
+
+echo "begin to do build inference library ..."
+set "dst_path=%source_path%\build_INFERENCE_LIBRARY_%PLAT%_%BLAS%_%CUDAV%"
+
+if exist %dst_path% rmdir /q /s %dst_path%
+mkdir %dst_path%
+
+cd /d %dst_path%
+echo Current directory : %cd%
+
+call:rest_env
+
+echo cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DCMAKE_BUILD_TYPE=Release -DWITH_PYTHON=OFF -DON_INFER=ON -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR% -DCUDA_ARCH_NAME=All
+cmake %dst_path%\..\Paddle -G "Visual Studio 15 2017 Win64" -T host=x64 -DWITH_GPU=%WITH_GPU% -DWITH_MKL=%WITH_MKL% -DCMAKE_BUILD_TYPE=Release -DWITH_PYTHON=OFF -DON_INFER=ON  -DCUDA_TOOLKIT_ROOT_DIR=%CUDA_DIR%  -DCUDA_ARCH_NAME=All
+
+set  MSBUILDDISABLENODEREUSE=1
+
+set BUILD_TYPE=INFERENCE LIBRARY
+call:Build
+
+echo PACKAGE INFERENCE LIBRARY
+
+mkdir inference_dist
+%PYTHON_DIR%\python.exe -c "import shutil;shutil.make_archive('inference_dist/paddle_inference_install_dir', 'zip', root_dir='paddle_inference_install_dir')"
+%PYTHON_DIR%\python.exe -c "import shutil;shutil.make_archive('inference_dist/paddle_install_dir', 'zip', root_dir='paddle_install_dir')"
+
+echo BUILD INFERENCE LIBRARY COMPLETE
+goto :END
+
+
+:Rest_env
+echo "Reset Build Environment ..."
+taskkill /f /im cmake.exe   2>NUL
+taskkill /f /im msbuild.exe 2>NUL
+taskkill /f /im git.exe 2>NUL
+taskkill /f /im cl.exe 2>NUL
+taskkill /f /im lib.exe 2>NUL
+taskkill /f /im git-remote-https.exe 2>NUL
+taskkill /f /im vctip.exe 2>NUL
+goto:eof
+
+:Build
+set build_times=1
+:build_thirdparty
+
+echo Build %BUILD_TYPE% Third Party Libraries, Round : %build_times%
+
+echo msbuild /m /p:Configuration=Release third_party.vcxproj ^>^> build_thirdparty_%build_times%.log
+msbuild /m /p:Configuration=Release third_party.vcxproj >> build_thirdparty_%build_times%.log
+
+IF %ERRORLEVEL% NEQ 0 (
+    echo Build %BUILD_TYPE% Third Party Libraries, Round : %build_times% Failed!
+    set /a build_times=%build_times%+1
+
+    if %build_times% GTR %RETRY_TIMES% (
+      goto :FAILURE
+  ) else (
+      goto :build_thirdparty
+  )
+)
+
+set build_times=1
+:build_paddle
+
+echo Build %BUILD_TYPE% Paddle Solutions, Round : %build_times%
+
+echo msbuild /m /p:Configuration=Release paddle.sln ^>^> build_%build_times%.log
+msbuild /m /p:Configuration=Release paddle.sln >> build_%build_times%.log
+
+IF %ERRORLEVEL% NEQ 0 (
+    echo Build %BUILD_TYPE% Paddle Solutions, Round : %build_times% Failed!
+    set /a build_times=%build_times%+1
+
+    if %build_times% GTR %RETRY_TIMES% (
+      goto :FAILURE
+  ) else (
+      goto :build_paddle
+  )
+)
+goto:eof
+
+
+:FAILURE
+echo BUILD FAILED
+exit /b 1
+
+:END
+echo BUILD SUCCESSFULLY
+
+ENDLOCAL

paddle/scripts/windows_build/config.ini

@@ -0,0 +1,19 @@
+############# the scripts for bulk publish package on windows platform ################
+
+# Usage: please use # to comment, use space or ; to separate multiple variables
+# Generally, you only need to change CUDA_PATH and PYTHON_PATH to publish package
+
+PADDLE_VERSION=2.0.0-alpha0
+
+BRANCH=v2.0.0-alpha0
+
+http_proxy=#please edit your proxy#
+https_proxy=#please edit your proxy#
+
+# Just for example, please set by your windows environment
+vcvarsall_dir="C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Auxiliary\Build\vcvarsall.bat"
+PYTHON3_PATH=C:\Python37
+
+CUDA_PATH="C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v10.0" "C:/Program Files/NVIDIA GPU Computing Toolkit/CUDA/v9.0"
+
+PYTHON_PATH=C:\Python37

paddle/scripts/windows_build/series_build.bat

@@ -0,0 +1,63 @@
+@ECHO OFF
+
+for /f "eol=# delims== tokens=1,2" %%i in (config.ini) do (
+    set %%i=%%j
+)
+
+set "source_path=%1"
+
+if "%source_path%"=="" (
+    set /p source_path="Please input the dst path :  =======>"
+)
+
+if not exist %source_path% mkdir %source_path%
+cd /d %source_path%
+if %errorlevel% NEQ 0 GOTO END
+
+
+echo "begin to download the source code from https://github.com/paddlepaddle/paddle"
+git clone https://github.com/PaddlePaddle/Paddle
+cd Paddle
+git checkout %BRANCH%
+rem sed -i "s/@PADDLE_VERSION@/%PADDLE_VERSION%/g" paddle\fluid\framework\commit.h.in
+rem sed -i "s/add_definitions(-DPADDLE_VERSION=\${PADDLE_VERSION})//g" cmake\version.cmake
+echo "download done!!!"
+
+cd ..
+set start_path=%~dp0
+echo %start_path%
+
+
+echo Init Visual Studio Env
+call %vcvarsall_dir% amd64
+
+rem source_path PYTHON_DIR WITH_GPU WITH_MKL ON_INFER PADDLE_VERSION BATDIR CUDA_DIR
+
+
+rem ==============build GPU============
+for /D %%i in ( %CUDA_PATH% ) do (
+    for /D %%j in ( %PYTHON_PATH% ) do (
+        call %start_path%build.bat %source_path% %%j ON ON OFF %PADDLE_VERSION% %start_path% %%i
+        call %start_path%build.bat %source_path% %%j ON OFF OFF %PADDLE_VERSION% %start_path% %%i
+    )
+    rem ===build inference library===
+    call %start_path%build.bat %source_path% %PYTHON3_PATH% ON ON ON %PADDLE_VERSION% %start_path% %release_dir% %%i
+    call %start_path%build.bat %source_path% %PYTHON3_PATH% ON OFF ON %PADDLE_VERSION% %start_path% %release_dir% %%i
+)
+
+
+rem ==============build CPU=============
+
+for /D %%j in ( %PYTHON_PATH% ) do (
+    call %start_path%build.bat %source_path% %%j OFF ON OFF %PADDLE_VERSION% %start_path% NEEDLESS
+    call %start_path%build.bat %source_path% %%j OFF OFF OFF %PADDLE_VERSION% %start_path% NEEDLESS
+)
+rem ===build inference library===
+call %start_path%build.bat %source_path% %PYTHON3_PATH% OFF ON ON %PADDLE_VERSION% %start_path% %release_dir% NEEDLESS
+call %start_path%build.bat %source_path% %PYTHON3_PATH% OFF OFF ON %PADDLE_VERSION% %start_path% %release_dir% NEEDLESS
+
+:END
+rem reset environment variable
+for /f "eol=# delims== tokens=1,2" %%i in (config.ini) do (
+    set %%i=
+)