---
title: "Install new Python packages on SQL Server | Microsoft Docs"
ms.date: "01/04/2018"
ms.reviewer: 
ms.suite: sql
ms.prod: machine-learning-services
ms.prod_service: machine-learning-services
ms.component: python
ms.technology: 
  - "r-services"
ms.tgt_pltfrm: ""
ms.topic: "article"
ms.assetid: 21456462-e58a-44c3-9d3a-68b4263575d7
caps.latest.revision: 1
author: "jeannt"
ms.author: "jeannt"
manager: "cgronlund"
ms.workload: "On Demand"
---
# Install new Python packages on SQL Server

This article describes how to install new Python packages on an instance of SQL Server 2017.

It also describes how to list packages that are installed in the current environment.

## Prerequisites

The process for installing new packages is simple. Some additional steps are required if the server does not have an internet connection.

+ You must have installed Machine Learning Services (In-Database) with the Python language option. For instructions, see [Set up Python Machine Learning Services](setup-python-machine-learning-services.md).

+ For each server instance, you must install a separate copy of the package. Packages cannot be shared across instances.

+ Determine whether the package you intend to use will work with Python 3.5 and in the Windows environment. In general, there are few limitations on the packages that you can import and use in the SQL Server environment. Possible problems include packages that use networking functionality that is blocked on the server or by the firewall, or packages with dependencies that cannot be installed on a Windows computer.

## Add a new Python package (online install)

For this example, we assume that you want to install a new package directly on the SQL Server computer, which is connected to the internet.

The package installed in this example is [CNTK](https://docs.microsoft.com/cognitive-toolkit/CNTK-FAQ), a framework for deep learning from Microsoft that supports customization, training, and sharing of different types of neural networks.

> [!TIP]
> For additional information, see this blog, which describes how to use Jupyter notebooks, and how to use **revoscalepy** with Python Tools for Visual Studio or PyCharm: [Getting Started with Python Web Services using Machine Learning Server](https://blogs.msdn.microsoft.com/mlserver/2017/12/13/getting-started-with-python-web-services-using-machine-learning-server/)

### Step 1. Download the Windows version of the Python package

+ If you are installing Python packages on a server with no internet access, you must download the WHL file to a different computer and then copy it to a local folder.

+ SQL Server 2017 uses Python 3.5. Be sure to get the Windows version of the package, and a version that is compatible with Python 3.5.

This example used a download link from this page, which supports downloads for multiple platforms and for multiple Python versions: [Installing CNTK for Python on Windows](https://docs.microsoft.com/cognitive-toolkit/setup-windows-python?tabs=cntkpy21)

### Step 2. Open a Python command prompt

Locate the default Python library location used by SQL Server. If you have installed multiple instances, be sure to locate the PYTHON_SERVICE folder for the instance where you want to add the package.

For example, if Machine Learning Services has been installed using defaults, and machine learning is enabled on the default instance, the path would be as follows:

    `C:\Program Files\Microsoft SQL  Server\MSSQL14.MSSQLSERVER\PYTHON_SERVICES`

Open the Python command prompt associated with the instance.

### Step 3. Install the package using pip

From the command line, you can use PIP.exe to install new packages. You can find the **pip** installer in the `Scripts` subfolder.

For example, the following command installs the version of CNTK that is supported for Windows and Python 3.5.

    `pip install https://cntk.ai/PythonWheel/CPU-Only/cntk-2.1-cp35-cp35m-win_amd64.whl`

If you get an error that **pip** is not recognized as an internal or external command, you can add the path of the Python executable and the Python scripts folder to the PATH variable in Windows.

The full path of the **Scripts** folder in a default installation is as follows:

    `C:\Program Files\Microsoft SQL Server\MSSQL14.MSSQLSERVER\PYTHON_SERVICES\Scripts`

As installation progresses, you can see status messages in the command prompt window:

```python
pip install https://cntk.ai/PythonWheel/CPU-Only/cntk-2.1-cp35-cp35m-win_amd64.whl
Collecting cntk==2.1 from https://cntk.ai/PythonWheel/CPU-Only/cntk-2.1-cp35-cp35m-win_amd64.whl
  Downloading https://cntk.ai/PythonWheel/CPU-Only/cntk-2.1-cp35-cp35m-win_amd64.whl (34.1MB)
    100% |################################| 34.1MB 13kB/s
...
Installing collected packages: cntk
Successfully installed cntk-2.1
```

When installation is complete, you can immediately begin using the package as described in the next step.

For examples of deep learning using CNTK, see these tutorials: [Python API for CNTK](https://cntk.ai/pythondocs/tutorials.html)

### Step 4. Load the package or its functions as part of your script

After the package has been installed, you can load the package in your script by using the standard `import <package_name>` syntax in the first lines of the script:

```python
import numpy as np
import cntk as cntk
cntk._version_
```

If you are running Python from a stored procedure, the same requirement applies: first load any packages you need in the first lines of the script.

For example, the following stored procedure call loads TensorFlow (which should already be installed), and uses a few simple functions to return messages:

```sql
EXECUTE sp_execute_external script @language = N'Python',
, @script = N'
import os
import tensorflow as tf
hello = tf.constant("hello")
sess = tf.Session()
print(sess.run(hello))'
```

##  How to view installed packages using conda

There are different ways that you can get a list of installed packages. The easiest is to use the **conda** package manager, which is included with the Anaconda Python environment added by SQL Server setup.

To view Python packages that have been installed in the current environment, run this command from the command prompt windows:

```python
conda list
```

For more information about **conda** and how you can use it to create and manage multiple Python environments, see [Managing environments with conda](https://conda.io/docs/user-guide/tasks/manage-environments.html).