代码拉取完成，页面将自动刷新
同步操作将从 MindSpore/mindquantum 强制同步，此操作会覆盖自 Fork 仓库以来所做的任何修改，且无法恢复！！！
确定后同步将在后台操作，完成时将刷新页面，请耐心等待。
It's required to sign CLA before your first code submission to MindSpore Quantum community.
For individual contributor, please refer to ICLA online document for the detailed information.
With the gradual failure of Moore's Law, it is more and more difficult to improve the integrated circuit board process and chip performance, and the computing power has reached a bottleneck. The architecture of classical computers is close to the limit, and this new technology field has gradually attracted the attention of the public at a time quantum computing when the computing power requirements of large models are becoming more and more prominent.
Quantum is a physical concept, which means the minimum and indivisible basic unit. The minimum unit is quantum; the indivisible concept is called quantization. It can be used to describe the characteristics of particles in the micro-physical world, such as atoms and electrons. "Quantum" comes from the Latin quantum, which translates as "a certain amount of matter".
Quantum computing is a cutting-edge technology, which uses the laws of quantum mechanics to control quantum information units for computing.。 In short, manipulating quantum bits is used in computers. At present, people use mobile phones, computers and other media to read this article. The underlying software is 0 and 1 data streams, and the corresponding hardware control is high and low levels. Quantum computing manipulates quantum bits, microscopic particles with quantum state characteristics.
Quantum computing is widely used in cryptography analysis, transportation problems (combinatorial optimization mathematical problems), materials and chemical engineering, biomedicine (chemical molecular energy simulation), artificial intelligence (quantum-machine learning), new semiconductors (quantum chips) and other fields.
quantum state $$ |\varphi>=\alpha |0>+\beta |1> $$
Install MindSpore
Please install MindSpore version 1.4.0 and above according to MindSpore website Installation Guide.
MindSpore is a new open source deep learning training/inference framework that could be used for mobile, edge and cloud scenarios. MindSpore is designed to provide development experience with friendly design and efficient execution for the data scientists and algorithmic engineers, native support for Ascend AI processor, and software hardware co-optimization.
pip install mindquantum
cd ~
git clone https://gitee.com/mindspore/mindquantum.git
Compile MindQuantum
Make Linux system sure that CMake > = 3.18.3 is installed, and then run the following command:
cd ~/mindquantum
bash build.sh --gitee
Here --gitee let the script download the third-party dependencies from the gitee code hosting platform. If you need to compile the GPU version, please first install CUDA 11.x and the corresponding graphics driver, and then use the --gpu parameters to execute the following compilation instructions:
cd ~/mindquantum
bash build.sh --gitee --gpu
Make Windows system sure that MinGW-W64 and CMake > = 3.18.3 are installed, and then run the following command:
cd ~/mindquantum
build.bat /Gitee
Make Mac system sure that OpenMP and CMake > = 3.18.3 are installed, and then run the following command:
cd ~/mindquantum
bash build.sh --gitee
Install the compiled whl package
Enter the output directory and install the compiled mindquantum whl package through the pip command.
Execute the following command. If no error No module named 'mindquantum' is reported, the installation is successful.
python -c 'import mindquantum'
MindQuantum provides Compile the package and Local compilation two methods
1.Compile the package. If the user needs to adapt to different system environments and python versions, he can compile the source code into a package and then install it. For detailed process, please refer to the above Source code installation
2.Local compilation. If the user adds or modifies part of the code and needs to quickly verify whether it is effective, he can use local compilation to compile the C + + code into a *.so file, and then use Python to call and set the environment variables. There is no need to recompile the package, and then uninstall and reinstall it, which is convenient for debugging and verification.
Download the source code from the code warehouse
cd ~
git clone https://gitee.com/mindspore/mindquantum.git
Linux System, depending on CMake > = 3.18.3, compiled natively.
--gitee The parameter specifies that the script downloads third-party dependencies from the gitee code hosting platform; export the command adds the mindquantum source code path to the PYTHONPATH environment variable.
cd ~/mindquantum
bash build_locally.sh --gitee
export PYTHONPATH=`pwd`$PYTHONPATH
Windows system, depending on MinGW-W64 and CMake > = 3.18.3, compiled natively.
cd ~/mindquantum
build_locally.bat /Gitee -G 'MinGW Makefiles'
set PYTHONPATH=%cd%;%PYTHONPATH%
Mac System, depending on OpenMP and CMake > = 3.18.3, compiled natively.
cd ~/mindquantum
bash build_locally.sh --gitee
export PYTHONPATH=`pwd`$PYTHONPATH
System Dependent Component Installation
GCC-GNU Compiler Suite, Installation Instruction
1.Linux and MacOS installation
1.1 Linux/Ubuntu/Centos systems and MacOS systems will come with versions gcc that meet the MindQuantum compilation.
1.2 Use the command to install GCC
# Ubuntu install
apt-get update
apt-get install gcc
apt-get install build-essential
# Centos install
yum install gcc gcc-c++
# output gcc version,verify successful installation
gcc --version
2.Window installation
2.1 Window's C/C + + compiler suite generally recommends Mingw-w64, which can be used to compile and run Window applications and DLL files.
2.Enter the Mingw-w64 installation package page and download the offline installation package. The recommended x86_64-posix-seh version here corresponds to the x64 architecture.
2.3. Double-click locally to open the installation package and install it according to the prompt. (Note that the environment variable is written)
2.4 After installation, open the cmd terminal and enter the command
C:\Users\xx> gcc --version
gcc (x86_64-win32-seh-rev0, Built by MinGW-W64 project) 8.1.0
Copyright (C) 2018 Free Software Foundation, Inc.
Encounter bug, can send issue, or Google Baidu, reference csnd blog.
CMake-Cross-platform open source builder, installation instructions
1.Enter the CMake official website and select the appropriate version of the installation package according to the system.
Windows system recommend cmake-3.2*-windows-x86_64.msi ,
Macos system recommend cmake-3.2*-macos-universal.dmg ,
Linux system recommendation cmake-3.2*-linux-x86_64.sh.
2.Double-click locally to open the CMake installation package, install it according to the prompt, note that CMake needs to be added to the user variable, and finally click Finish to complete the installation.
3.Open the terminal, output the command in the command line, and output the version number to indicate success.
cmake --version
>>> cmake version 3.24.2
The mindquantum is mainly developed in C + + and python, the core computing unit is implemented in C/C + +, and the upper interface and peripheral modules are implemented in Python
The development process is mainly divided into two categories, developing new functions and fixing bugs:
Develop new features, enter the issue page of MindQuantum, submit the issue, write new feature descriptions, categories, implementation methods, etc. Talk to the MindQuantum development team to confirm if this feature is necessary. Start the implementation locally, write the code, implement the function, write the corresponding test cases, and the corresponding documentation. Submit PR, merge into the main branch after the code passes the review, and complete the development of new functions.
Fix the bug, go to the issue page of MindQuantum, read the unclosed issues, and claim the issue to solve the problem. Or if you encounter a bug when using MindQuantum, you are welcome to submit an issue to help improve the MindQuantum function module.
mindquantum to a quantum circuit including a H-gate, an RX gate, and an ry gate, and obtain a quantum statefrom mindquantum import *
import numpy as np
encoder = Circuit().h(0).rx({'a0': 2}, 0).ry('a1', 1)
print(encoder)
print(encoder.get_qs(pr={'a0': np.pi / 2, 'a1': np.pi / 2}, ket=True))
Running the above code will output the quantum circuit and the final state.
┏━━━┓ ┏━━━━━━━━━━┓
q0: ──┨ H ┠─┨ RX(2*a0) ┠───
┗━━━┛ ┗━━━━━━━━━━┛
┏━━━━━━━━┓
q1: ──┨ RY(a1) ┠───────────
┗━━━━━━━━┛
-1/2j¦00⟩
-1/2j¦01⟩
-1/2j¦10⟩
-1/2j¦11⟩
ccsrc core computing module, using C/C + + implementationcmake cmake compiles configuration informationdocs Mind Quantum API Documentationmindquantum Mind Quantum quantum computing module, implemented in Python
mindquantum.dtype MindQuantum data type simulation.mindquantum.core Core features of MindQuantum (eDSL).
gata quantum gate module providing different quantum gates.circuit Quantum circuit module, It can easily build quantum circuits that meet the requirements, including parameterized quantum circuits.operators MindQuantum Operator libraryparameterresolver The parameter parser module is used for declaring the used parameters.mindquantum.simulator A quantum simulator for simulating the evolution of a quantum system.mindquantum.framework Quantum neural network operators and cell.mindquantum.algorithm Quantum algorithm.
compiler Quantum circuit compiling modulelibrary Common algorithm modulenisq NISQ algorithmerror_mitigation Error mitigation modulemapping Bit Mapping Modulemindquantum.device Mind Quantum hardware module.mindquantum.io Input/output module for MindQuantum.mindquantum.engine MindQuantum Engine Module.mindquantum.utils utility.mindquantum_config Item Configuration Informationscripts Compiling a Dependent Tools Update Scripttests MindQuantum unit tests, based on Pytest, written in Pythonthird_party Third-party open source packages on which MindQuantum compilestutorials Mind Quantum tutorial, which can be run directly with jupyter. Readable in Official document of MindSpore.MindQuantum writes unit test cases based on pytest. It is recommended that developers write corresponding unit test cases after implementing a new function or module to ensure normal functions
Instructions for writing documentation, MindQuantum has two main types of documentation:
MindQuanntum/docs base. If you're interested in adding new developer documentation, read this page on the wiki to learn about best practices and write comments after you've written the code. The API is collated by extracting comments from the code.Please follow this style to make MindSpore easy to review, maintain and develop.
Coding guidelines
The Python coding style suggested by Python PEP 8 Coding Style and C++ coding style suggested by Google C++ Coding Guidelines are used in MindSpore community. The CppLint, CppCheck, CMakeLint, CodeSpell, Lizard, ShellCheck and PyLint are used to check the format of codes, installing these plugins in your IDE is recommended.
Unittest guidelines
The Python unittest style suggested by pytest and C++ unittest style suggested by Googletest Primer are used in MindSpore community. The design intent of a testcase should be reflected by its name of comment.
Refactoring guidelines
We encourage developers to refactor our code to eliminate the code smell. All codes should conform to needs to the coding style and testing style, and refactoring codes are no exception. Lizard threshold for nloc (lines of code without comments) is 100 and for cnc (cyclomatic complexity number) is 20, when you receive a Lizard warning, you have to refactor the code you want to merge.
Document guidelines
We use MarkdownLint to check the format of markdown documents. MindSpore CI modifies the following rules based on the default configuration.
For details, please refer to RULES.
Fork MindSpore repository
Before submitting code to MindSpore project, please make sure that this project have been forked to your own repository. It means that there will be parallel development between MindSpore repository and your own repository, so be careful to avoid the inconsistency between them.
Clone the remote repository
If you want to download the code to the local machine, git is the best way:
# For GitHub
git clone https://github.com/{insert_your_forked_repo}/mindspore.git
git remote add upstream https://github.com/mindspore-ai/mindspore.git
# For Gitee
git clone https://gitee.com/{insert_your_forked_repo}/mindspore.git
git remote add upstream https://gitee.com/mindspore/mindspore.git
Develop code locally
To avoid inconsistency between multiple branches, checking out to a new branch is SUGGESTED:
git checkout -b {new_branch_name} origin/master
Taking the master branch as an example, MindSpore may create version branches and downstream development branches as needed, please fix bugs upstream first. Then you can change the code arbitrarily.
Push the code to the remote repository
After updating the code, you should push the update in the formal way:
git add .
git status # Check the update status
git commit -m "Your commit title"
git commit -s --amend #Add the concrete description of your commit
git push origin {new_branch_name}
Pull a request to MindSpore repository
In the last step, your need to pull a compare request between your new branch and MindSpore master branch. After finishing the pull request, the Jenkins CI will be automatically set up for building test. Your pull request should be merged into the upstream master branch as soon as possible to reduce the risk of merging.
A great way to contribute to the project is to send a detailed report when you encounter an issue. We always appreciate a well-written, thorough bug report, and will thank you for it!
When reporting issues, refer to this format:
Issues advisory:
PRs advisory:
In the development process, it is recommended to use the pre-push function to perform local code self-check. Code scanning similar to the Code Check stage of CI access control can be performed locally to improve the success rate of access control. For usage, see Pre-push Quick Guide.
After the development is completed, suggest vscode or pycharm format functions to standardize the code style.
此处可能存在不合适展示的内容，页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容，可点击提交进行申诉，我们将尽快为您处理。