|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/ml-model-deployment-with-fastapi-and-streamlit/
课程评论：没有评论
课程名称：使用 FastAPI 和 Streamlit 部署 ML 模型 课程概述：为什么要部署模型？机器学习模型不仅仅是展示用的，它们旨在在现实世界中应用。通过部署您的模型，您可以使用户能够利用其洞察力和预测。部署模型的几个理由包括： 1. 用户交互：机器学习不仅是构建复杂的模型，更是为用户提供价值。通过部署模型，用户可以通过用户友好的接口（如 API 或 Web 应用）与洞察进行交互，从而让您的工作更易获取并产生影响。 2. 复杂应用：一些机器学习模型旨在用于复杂应用，如 AI 语音助手、视频推荐和天气预报。通过部署模型，这些应用可以利用其预测能力，为用户提供有价值的服务。 3. 可扩展性和效率：将模型部署在服务器或云平台上可以确保可扩展性和效率，从而同时处理大量请求，保证模型可以服务多个用户而不影响性能。 4. 实时预测：通过部署模型，可以实现实时预测。这对于需要即时响应的应用至关重要，如欺诈检测系统和股票交易平台。部署模型使其能够即时作出预测，为用户提供最新的见解。 5. 持续改进：部署模型并非一次性任务，它涉及持续监控与改进。通过在实际场景中跟踪模型的表现，您可以识别改进空间并做出必要调整，以提高其准确性和有效性。 本课程包含八个部分，帮助您全面理解模型部署，包括最佳实践、不同部署方法和各种用例的考虑。每个章节旨在实现特定目标，帮助您掌握成功部署机器学习模型所需的知识与技能。 课程内容简介： 1. **介绍**：深入探讨模型部署的重要性和多样化策略，提供 FastAPI 和 Streamlit 的简要概述。 2. **使用 FastAPI 构建 API**：涵盖 API 的构建、参数处理、数据接收及用户友好界面的创建，确保高质量和可维护性。 3. **将 ML 模型作为 API 实现 FastAPI**：通过构建天气预测 API，学习如何用 FastAPI 创建 ML 模型 API。 4. **使用 Streamlit 构建 Web 应用程序**：介绍 Streamlit 应用的基本组件及性能优化技巧。 5. **将 FastAPI 与 Streamlit 集成**：构建天气预测模型 API 的用户界面，并将其与 Streamlit 集成。 6. **部署**：展示如何使用 Render 和 Streamlit Cloud 部署模型 API 和 Streamlit 应用。 7. **WhatsApp AI 文本到图像聊天机器人**：利用 FastAPI 和外部工具（如 Vonage 和 DALL-E API）构建 WhatsApp 聊天机器人项目。 8. **顶点项目**：展示在本课程中学习的技能，通过构建支持房地产中介预测房价的完整应用程序来巩固所学知识。
Why Deploy Models?Machine learning models aren't just for show; they're meant to be used in the real world. By deploying your model, you can make it accessible to users who can benefit from its insights and predictions. There are several reasons why you should deploy your model:1. User Interaction:Machine learning is not just about building complex models; it's about providing value to users. Deploying your model allows users to interact with your insights through user-friendly interfaces like APIs or web apps, making your work more accessible and impactful.2. Complex Applications:Some machine learning models are designed to be used in complex applications such as AI voice assistants, video recommendations, and weather forecasting. Deploying your model enables these applications to leverage their predictions and provide valuable services to users.3. Scalability and Efficiency:Deploying your model on a server or cloud platform ensures scalability and efficiency. This allows you to handle a large number of requests simultaneously, ensuring that your model can serve multiple users without compromising performance.4. Real-Time Predictions:By deploying your model, you can enable real-time predictions. This is crucial for applications that require immediate responses, such as fraud detection systems or stock trading platforms. Deploying your model allows it to make predictions on the fly, providing users with up-to-date insights.5. Continuous Improvement:Deploying your model is not a one-time task. It involves continuous monitoring and improvement. By tracking your model's performance in real-world scenarios, you can identify areas for improvement and make the necessary adjustments to enhance its accuracy and effectiveness.As you work through this course, consisting of eight sections, you will gain a comprehensive understanding of model deployment, including best practices, different deployment methods, and considerations for various use cases.Each chapter is designed to achieve specific aims and objectives, equipping you with the knowledge and skills necessary to successfully deploy your machine learning models.IntroductionIn this section, you will delve into the concept of model deployment, exploring its significance and the diverse strategies employed in the process. We will provide a concise overview of FastAPI and Streamlit, shedding light on their distinct purposes and how they contribute to model deployment.Building APIs with FastAPIThis section dives deep into building APIs with FastAPI, covering essential concepts like handling various parameters, receiving data inputs, and crafting user-friendly interfaces. You'll learn how to create APIs that accept and process diverse data from various sources while ensuring their quality and maintainability through effective testing practices. By the end, you'll be equipped to build robust and scalable APIs, confident in their ability to meet modern web requirements.ML Models as an API with FastAPIYou will learn how to use FastAPI to create ML model APIs by building a weather model forecast API. You will build a forecast model from scratch and use it for prediction on an API endpoint.Building Web Applications with StreamlitIn this section, you will be introduced to the basic components of a Streamlit application, including inputs, widgets, and layouts. You will also learn about caching and session management, two important features for building high-performance web applications.Integrating FastAPI with StreamlitIn this section, you will build the user interface for the weather forecasting model API built in the previous section and also integrate the API with Streamlit.DeploymentThis section will show you how to deploy your Model API and Streamlit application using Render and Streamlit Cloud.WhatsApp AI Text-to-Image ChatbotThis project will show you how to use your existing FastAPI knowledge with external tools such as Vonage and the DALL-E API to build a WhatsApp chatbot.Capstone ProjectYou will showcase the skills you have learned in this course by building a full application that allows real estate agencies to predict house prices using various features.
