Quickstart – Get started using GPT⑶5-Turbo and GPT⑷ with Azure OpenAI Service –
I. 甚么是Azure OpenAI服务
A. Azure OpenAI服务概述
Azure OpenAI服务是由微软和OpenAI共同开发的高级语言AI服务，旨在为企业用户提供强大的语言模型和AI解决方案。
-
Azure OpenAI服务的开发背景和目标
Azure OpenAI服务由微软和OpenAI共同开发，旨在为企业用户提供高级语言AI服务。
- OpenAI支持的模型包括OpenAIGPT⑷、GPT⑶、Codex和DALL-E，能够满足多种语言生成和处理需求。
- Azure OpenAI服务结合了Azure云计算平台的安全性和企业前景，为客户提供可靠的AI解决方案。
-
Azure OpenAI服务的功能和用处
Azure OpenAI服务提供强大的语言模型，可用于内容生成、摘要生成、语义分析等多种任务。
- 客户可以对语言模型进行细化调剂，以满足个性化需求。
- 通过REST API访问和Azure OpenAI Studio界面，方便客户开发和使用AI模型。
II. 怎样使用Azure OpenAI服务
A. 添加Azure OpenAI服务的API到Azure API Management实例中
在Azure API Management实例中添加Azure OpenAI服务的API，可以方便地使用OpenAI的语言模型。
-
登录Azure门户并进入API Management实例
打开Azure门户，进入API Management实例。
-
在API Management实例中添加API
在左边菜单选择”API > + 添加API”，填写API的相关信息，包括API名称、描写、URL等。
-
添加并配置API的操作
肯定API的访问权限和其他设置，完成API的添加和配置。
B. 配置OpenAI API密钥和终结点
需要在代码中初始化OpenAI SDK，并设置API密钥和终结点。
-
在代码中初始化OpenAI SDK
导入openai包，并设置API密钥和终结点。
import openai
openai.api_key = ‘your-API-key’
openai.api_base = ‘https://your-api-endpoint’ -
根据自己的API密钥和终结点信息进行替换
在初始化OpenAI SDK时，根据自己的API密钥和终结点信息，将代码中的相关字段进行替换。
III. Azure OpenAI服务的利用场景和优势
A. 利用场景
Azure OpenAI服务的利用场景非常广泛，可以满足区别领域的语言处理需求。
-
内容生成
利用OpenAI的语言模型生成文章、评论、描写等内容，帮助企业自动生成大量的文本数据。
-
摘要生成
通过语言模型生成文章或文本的摘要，提供核心信息，帮助用户快速了解文本内容。
-
语义分析
使用语言模型理解和解释文本的含义，进行情感分析、主题标签等功能，帮助用户深入理解文本。
-
个性化调剂
OpenAI模型支持对语言模型进行细化调剂，以适应具体需求，给用户提供个性化的语言处理功能。
B. 优势
Azure OpenAI服务的优势在于强大的语言模型、个性化定制、快速便捷的API访问和结合Azure平台优势。
-
强大的语言模型
Azure OpenAI服务提供行业领先的语言模型，可以满足各种语言处理任务，如内容生成、摘要生成和语义分析等。
-
个性化定制
通过细化调剂，客户可以根据自己的需求对语言模型进行定制，使其更符合特定的业务场景。
-
快速、便捷的API访问
Azure OpenAI服务通过REST API和Azure OpenAI Studio界面提供快速、便捷的访问方式，使开发者能够轻松使用OpenAI模型。
-
结合Azure平台优势
Azure OpenAI服务结合了Azure云计算平台的安全性和企业前景，为客户提供可靠的AI解决方案，帮助企业实现业务上的创新和改进。
openai api on azure的进一步展开说明
### Table of Contents
1. Quickstart: Get started using GPT⑶5-Turbo and GPT⑷ with Azure OpenAI Service
2. Article
3. 08/31/2023
4. 5 contributors’ Feedback
## Quickstart: Get started using GPT⑶5-Turbo and GPT⑷ with Azure OpenAI Service
Are you interested in using Azure OpenAI? This article provides a step-by-step guide on how to get started with Azure OpenAI and use the GPT⑶5-Turbo and GPT⑷ models.
### Prerequisites
Before you begin, you will need the following:
– An Azure subscription: If you don’t have one, you can create a free subscription.
– Access to Azure OpenAI: Currently, access to Azure OpenAI is granted only by application. You can apply for access by completing the form at [https://aka.ms/oai/access](https://aka.ms/oai/access). If you encounter any issues, you can open an issue on the [Azure OpenAI GitHub repository](https://github.com/Azure/oai).
To deploy the models, you will need an Azure OpenAI Service resource with either the GPT⑶5-Turbo or GPT⑷ models deployed. For more information on model deployment, refer to the resource deployment guide. Please note that GPT⑷ models are currently only available by request. Existing Azure OpenAI customers can apply for access to these models by filling out [this form](https://aka.ms/oai/access).
If you encounter any issues with the prerequisites, you can navigate to Azure OpenAI Studio for further guidance.
### Navigate to Azure OpenAI Studio
To access Azure OpenAI Studio, go to [https://oai.azure.com/](https://oai.azure.com/) and sign in using your credentials. Make sure your credentials have access to your OpenAI resource. During the sign-in process, select the appropriate directory, Azure subscription, and Azure OpenAI resource.
### Chat Playground
Once you’re in the Azure OpenAI Studio landing page, you can start exploring OpenAI capabilities with the Chat playground. The Chat playground allows you to experiment with the capabilities of Azure OpenAI using a no-code approach.
### Assistant Setup
In the Chat playground, you can use the Assistant setup dropdown to select pre-loaded System message examples. These examples give instructions to the model on how it should behave and provide context for generating responses. You can modify the assistant’s personality, specify what it should or shouldn’t answer, and define response formatting.
You can also use the Add few-shot examples feature to provide conversational examples for in-context learning by the model.
While using the Chat playground, you can view code samples in Python, curl, and json by selecting the View code option. These code samples are pre-populated based on your current chat session and settings selections. You can utilize this code to build an application that performs the same task as the playground.
### Chat Session
In the Chat playground, you can enter text and send it as a query to the completions API. The results of the API call will be displayed in the text box. To clear the chat history, you can select the Clear chat button.
### Settings
The Settings section allows you to configure various parameters for your chat session. Some of the key settings are:
– Deployments: Specify the name of the deployment associated with a specific model.
– Temperature: This parameter controls the randomness of the model’s responses. Lower temperature values result in more deterministic responses, while higher values yield more creative and unexpected responses.
– Max length (tokens): Set a limit on the number of tokens in the model’s response. The API supports a maximum of 4096 tokens, including the prompt, examples, message history, and the model’s response.
– Top probabilities: Similar to temperature, this parameter affects the randomness of the model’s token selection. Lower values narrow down the token selection to likelier tokens, while higher values allow the model to choose from both high and low likelihood tokens.
– Multi-turn conversations: Select the number of past messages to include in each new API request. This provides context to the model for generating responses.
– Stop sequences: Specify certain phrases that make the model end its response at a desired point.
– Show panels: Customize the display of panels in the Chat playground.
### Start a Chat Session
To start a chat session, select the Xbox customer support agent option from the Assistant setup dropdown. You will be prompted to update the system message. In the chat session pane, enter the question “I’m interested in buying a new Xbox” and select Send.
You will receive a response from the assistant. You can then ask follow-up questions and explore the capabilities of the model. The View code option allows you to view the code behind the conversation so far.
### Understanding the Prompt Structure
The prompt structure in the Chat playground consists of different tokens that serve specific purposes. The GPT⑶5-Turbo model uses special tokens to separate different parts of the prompt. The <|im_start|> and
openai api on azure的常见问答Q&A
问题1：Azure OpenAI服务是甚么？
答案：Azure OpenAI服务是微软推出的一个服务，旨在为开发者提供先进的语言模型和编码功能。它可让开发者使用GPT⑶、GPT⑷、Codex和DALL-E等强大的语言AI模型来生成文本、总结内容、进行语义分析等。Azure OpenAI服务通过提供REST API和Web界面的两种访问方式，使用户能够轻松地在Azure平台上使用这些功能。
- Azure OpenAI服务可以为用户提供高级语言AI，包括生成文本、总结内容、语义分析等功能。
- 用户可使用GPT⑶、GPT⑷、Codex和DALL-E等先进的语言模型。
- Azure OpenAI服务提供REST API和Web界面两种访问方式。
问题2：如何申请和使用Azure OpenAI服务？
答案：申请和使用Azure OpenAI服务非常简单。以下是申请和使用Azure OpenAI服务的步骤：
- 访问Azure门户并登录。
- 进入您的API Management实例。
- 在左边菜单当选择“API > + 添加 API”。
- 根据您的资源进行更改API的终结点。
- 配置OpenAI API密钥。
- 根据得到的密钥和终结点初始化SDK。
- 安装Azure OpenAI SDK。
- 使用SDK进行开发和测试。
通过这些简单的步骤，您就能够申请和使用Azure OpenAI服务，从而享遭到其中强大的语言AI功能。
问题3：Azure OpenAI服务提供了哪些功能？
答案：Azure OpenAI服务提供了多种功能，包括但不限于以下内容：
- 文本生成：使用先进的语言模型生成各种类型的文本。
- 内容总结：根据一段文本生成其摘要或总结。
- 语义分析：分析文本的意义和语境，进行情感分析、关键词提取等。
- 编码能力：将文本转换为向量表示，用于各种机器学习任务。
- 图象生成：使用DALL-E模型根据描写生成图象。
- 自然语言处理：处理和理解自然语言的能力，进行词法分析、语法分析等。
Azure OpenAI服务的这些功能可以帮助开发者更轻松地完成各种语言相关的任务，并提高工作效力。
