您现在访问的是微软AZURE全球版技术文档网站,若需要访问由世纪互联运营的MICROSOFT AZURE中国区技术文档网站,请访问 https://docs.azure.cn.

使用 Azure 实验室服务设置专注于自然语言处理中的深度学习的实验室Set up a lab focused on deep learning in natural language processing using Azure Lab Services

本文介绍如何使用 Azure 实验室服务来设置一个专注于自然语言处理 (NLP) 中的深度学习的实验室。This article shows you how to set up a lab focused on deep learning in natural language processing (NLP) using Azure Lab Services. 自然语言处理 (NLP) 是某种形式的人工智能 (AI),可在计算机中实现翻译、语音识别和其他语言理解功能。Natural language processing (NLP) is a form of artificial intelligence (AI) that enables computers with translation, speech recognition, and other language understanding capabilities.

使用 NLP 类的学生可以通过 Linux 虚拟机 (VM) 了解如何应用神经网络算法,以开发深度学习模型用于分析人类手写语言。Students taking an NLP class get a Linux virtual machine (VM) to learn how to apply neural network algorithms to develop deep learning models that are used for analyzing written human language.

实验室配置Lab configuration

若要设置此实验室,你需要具有一个 Azure 订阅才能开始。To set up this lab, you need an Azure subscription to get started. 如果没有 Azure 订阅,请在开始之前创建一个免费帐户If you don't have an Azure subscription, create a free account before you begin. 拥有 Azure 订阅后,可以在 Azure 实验室服务中创建新的实验室帐户,也可以使用现有的实验室帐户。Once you have an Azure subscription, you can either create a new lab account in Azure Lab Services or use an existing lab account. 请参阅以下教程,了解如何创建新的实验室帐户:有关设置实验室帐户的教程See the following tutorial for creating a new lab account: Tutorial to Setup a Lab Account.

创建实验室帐户后,在实验室帐户中启用以下设置:After you create the lab account, enable following settings in the lab account:

实验室帐户设置Lab account setting InstructionsInstructions
市场映像Marketplace images 启用用于 Linux (Ubuntu) 的 Data Science Virtual Machine 映像,以便在实验室帐户中使用。Enable the Data Science Virtual Machine for Linux (Ubuntu) image for use within your lab account. 有关说明,请参阅以下文章:指定可供实验室创建者使用的市场映像See the following article for instructions: Specify marketplace images available to lab creators.

按照此教程创建新的实验室并应用以下设置:Follow this tutorial to create a new lab and apply the following settings:

实验室设置Lab settings 值/说明Value/instructions
虚拟机 (VM) 大小Virtual machine (VM) size **小型 GPU (计算) **。Small GPU (Compute). 此大小最适用于计算密集型和网络密集型应用程序(如人工智能和深度学习)。This size is best suited for compute-intensive and network-intensive applications like Artificial Intelligence and Deep Learning.
VM 映像VM image 用于 Linux (Ubuntu)的 Data Science Virtual MachineData Science Virtual Machine for Linux (Ubuntu). 此映像提供机器学习和数据科学深度学习框架和工具。This image provides deep learning frameworks and tools for machine learning and data science. 若要查看此映像上安装的工具的完整列表,请参阅以下文章:DSVM 中包含哪些组件?To view the full list of installed tools on this image, see the following article: What’s included on the DSVM?.
启用远程桌面连接Enable remote desktop connection

数据科学映像已配置为使用 X2Go,使教师和学生能够使用 GUI 远程桌面进行连接。The Data Science image is already configured to use X2Go so that teachers and students can connect using a GUI remote desktop. X2Go 不要求启用“启用远程桌面连接”设置。X2Go does not require the Enable remote desktop connection setting to be enabled. 仅当你要改用 RDP 时,才需要启用此设置。This setting only needs to be enabled if you choose to instead use RDP.

重要说明:尽管我们建议对数据科学映像使用 X2Go,但如果你要改用 RDP,则首次连接到 Linux VM 时需要使用 SSH,并安装 RDP 和 GUI 包。Important: Although we recommend using X2Go with the Data Science image, if you choose to instead use RDP, you will need to connect to the Linux VM using SSH the first time and install the RDP and GUI packages. 以后,你/学生可以使用 RDP 连接到 Linux VM。Then, you/students can connect to the Linux VM using RDP later. 有关详细信息,请参阅为 Linux VM 启用图形远程桌面For more information, see Enable graphical remote desktop for Linux VMs.

用于 Linux 的 Data Science Virtual Machine 映像提供此类型课程所需的深度学习框架和工具。The Data Science Virtual Machine for Linux image provides the necessary deep learning frameworks and tools required for this type of class. 因此,在创建模板虚拟机后,无需进一步自定义。As a result, after the template machine creation, you don't need to customize it further. 可以将其发布供学生使用。It can be published for students to use. 选择模板页上的“发布”按钮,将模板发布到实验室。Select the Publish button on template page to publish the template to the lab.

成本Cost

如果要估算此实验室的成本,可以使用以下示例:If you would like to estimate the cost of this lab, you can use the following example:

如果一个班级有 25 个学生,计划的课堂时间为 20 小时且家庭作业或任务时数配额为 10 小时,则实验室的价格为 - 25 个学生 * (20 + 10) 个小时 * 139 个实验室单位 * 0.01 美元/小时 = 1042.5 美元For a class of 25 students with 20 hours of scheduled class time and 10 hours of quota for homework or assignments, the price for the lab would be - 25 students * (20 + 10) hours * 139 Lab Units * 0.01 USD per hour = 1042.5 USD

有关定价的更多详细信息,请参阅 Azure 实验室服务定价Further more details on pricing, see Azure Lab Services Pricing.

结束语Conclusion

本文已指导你完成为自然语言处理课程创建实验室的步骤。This article walked you through the steps to create a lab for natural language processing class. 对于其他深度学习课程,可以使用类似的设置。You can use a similar setup for other deep learning classes.

后续步骤Next steps

接下来的步骤是设置任何实验室时通用的:Next steps are common to setting up any lab: