# Creating a Project

[![View Source On Gitee](https://mindspore-website.obs.cn-north-4.myhuaweicloud.com/website-images/master/resource/_static/logo_source_en.svg)](https://gitee.com/mindspore/docs/blob/master/docs/devtoolkit/docs/source_en/mindspore_project_wizard.md)

## Background

This function is implemented based on the [conda](https://conda.io). Conda is a package management and environment management system. It is one of the recommended installation modes for MindSpore.

## Functions

* It creates a conda environment or selects an existing conda environment, and installs the MindSpore binary package in the conda environment.
* It deploys the best practice template. In addition to testing whether the environment is successfully installed, it also provides a tutorial for MindSpore beginners.
* When the network condition is good, the environment can be installed within 10 minutes and you can experience MindSpore immediately. It reduces up to 80% environment configuration time for beginners.

## Procedure

1. Choose **File** > **New Project**.

    ![img](images/clip_image002.jpg)

2. Select **MindSpore**.

    ![img](images/clip_image004.jpg)

3. Download and install Miniconda. ***If conda has been installed, skip this step.***

    3.1 Click **Install Miniconda Automatically**.

      ![img](images/clip_image006.jpg)

    3.2 Select an installation folder. **You are advised to use the default path to install conda.**

      ![img](images/clip_image008.jpg)

    3.3 Click **Install**.

      ![img](images/clip_image010.jpg)

      ![img](images/clip_image012.jpg)

    3.4 Wait for the installation to complete.

      ![img](images/clip_image014.jpg)

    3.5 Restart PyCharm as prompted or restart PyCharm later. ***Note: The following steps can be performed only after PyCharm is restarted.***

      ![img](images/clip_image015.jpg)

4. If **Conda executable** is not automatically filled, select the path of the installed conda.

    ![img](images/clip_image016.jpg)

5. Create a conda environment or select an existing conda environment.

    * Create a conda environment. **You are advised to use the default path to create the conda environment. Due to PyCharm restrictions on Linux, you can only select the default directory.**

      ![img](images/clip_image018.jpg)

    * Select an existing conda environment in PyCharm.

      ![img](images/clip_image019.jpg)

6. Select a hardware environment and a MindSpore best practice template.

    6.1 Select a hardware environment.

      ![img](images/clip_image020.jpg)

    6.2 Select a best practice template. The best practice template provides some sample projects for beginners to get familiar with MindSpore. The best practice template can be run directly.

      ![img](images/clip_image021.jpg)

7. Click **Create** to create a project and wait until MindSpore is successfully downloaded and installed.

    7.1 Click **Create** to create a MindSpore project.

      ![img](images/clip_image022.jpg)

    7.2 The conda environment is being created.

      ![img](images/clip_image023.jpg)

    7.3 MindSpore is being configured through conda.

      ![img](images/clip_image024.jpg)

8. Wait till the MindSpore project is created.

    ![img](images/clip_image025.jpg)

9. Check whether the MindSpore project is successfully created.

    * Click **Terminal**, enter **python -c "import mindspore;mindspore.run_check()"**, and check the output.  If the version number shown in the following figure is displayed, the MindSpore environment is available.

      ![img](images/clip_image026.jpg)

    * If you select a best practice template, you can run the best practice to test the MindSpore environment.

      ![img](images/clip_image027.jpg)

      ![img](images/clip_image028.jpg)

      ![img](images/clip_image029.jpg)