Generative AI with LLMs: Project with Python and FastAPI

Connect llms to Python by setting up Python and VS Code, creating a virtual environment, and exploring OpenAI models like GPT-4 and Whisper to build a Python-based AI application.

Verify that Python is installed and pip is ready. Learn pip commands to install, list, and uninstall packages, and create a virtual environment to isolate dependencies; discuss requirements files later.

Create and activate a Python virtual environment in your project folder using PowerShell as an administrator, then install OpenAI with pip inside the environment.

Activate the script in Python by adjusting PowerShell remote policy, running PowerShell as administrator, and answering yes to enable virtual environments for script execution.

Explore how to build your first python application with OpenAI's llms by using the OpenAI library, chat completion, and audio transcription to interact with GPT models.

Create and securely save your OpenAI API key, then use it to call OpenAI services from Python while learning to manage organizations, projects, and API types.

Build AI apps and services using the OpenAI API in Python, exploring chat completion, audio transcription, text-to-speech, and multiple LLM models.

Explore how FastAPI turns Python code into a web app with asynchronous capability. Build routes using get, post, put, delete; connect front end and back end with the fetch API.

Turn your machine into a web server with uvicorn and FastAPI, then host a first site using Python, http, and websockets, with hands-on practice and Azure deployment notes.

Create your first website using FastAPI in Python, define a hello world route with an async function on an app object, and run it via uvicorn.

Define and organize FastAPI routers for custom, sales, and employees, then consolidate them in main.py and include their endpoints as category-based routes to build a structured website.

Create and organize routes in FastAPI using routers, define list and create endpoints for customers, employees, and sales, and assemble them in the main app with include_router, exposing docs.

Build a Python and FastAPI app that leverages OpenAI chat completion and image generation, uses embeddings to turn data into vectors with a vector database, and excludes fine-tuning.

Build a chat bot app using Python and FastAPI by structuring text, vision, and audio services and implementing a text generation service with OpenAI completion.

Create a Python environment variable in a virtual environment to store the OpenAI API key once, load it with load_dotenv and os, and use it across files.

Finalize a chat bot app by creating a post route for text chat, wiring the OpenAI completion with GPT four turbo, and testing the flow via a front-end call.

Create a text moderation endpoint in a FastAPI project using OpenAI's moderation service to flag violence or discriminatory content, with a post route at /api/text/moderation and fetch API testing.

Create a vision image generation route using OpenAI, load environment variables, and return a generated image URL via a post endpoint at api/vision/image_generation with prompt, size, and quality.

Learn how files are transferred over http and how text becomes audio via an OpenAI text-to-speech service. Compare saving audio in RAM versus hard drive and memory management.

Learn to build a text-to-speech service with OpenAI in a Python FastAPI app, turning user text into audio in memory for an avatar that speaks aloud.

Explore in-memory data streams with the io module and BytesIO, saving audio bytes in ram and streaming them via FastAPI's stream response as mp3 audio.

Demonstrate building a transcription app by sending audio as in-memory bytes to OpenAI whisper, using the SDK or HTTP with multipart data, and understanding the HTTP request flow.

Develop a fastapi endpoint that accepts audio uploads via multipart form data, reads files in memory, and uses whisper to transcribe audio to text.