Becoming an AI Engineer with LLM Application Development

Install and configure Python to build the back end and front end of a generative AI application, using Python.org downloads across Windows, Mac, and Linux.

Set up VSCode for AI development, edit code, debug, and manage repositories with essential extensions. Log in with GitHub, explore file manager, source control, copilot chat, testing, and Docker extensions.

Explore how GitHub hosts a git-based version control system, enables code collaboration and social coding through a web interface, and demonstrates signing up, signing in, and navigating repositories.

Explore Hugging Face as a machine learning model repository and deployment platform, learn to sign up, generate an access token, and leverage models, datasets, spaces, and docs.

Sign up for the OpenAI platform, log in with email, Gmail, or Microsoft account, and access organization, dashboards, and API keys to interact with models and fine-tune via the playground.

Explore HTTP and RESTful API concepts, including the client-server model, HTTP methods (get, post, put, delete), and how RESTful design maps to HTTP for simple, scalable, stateless APIs.

Explore the six-step API development and deployment workflow, from requirement analysis to deployment, covering API design, development, unit and functional testing, documentation, and container-based deployment.

Architect your Python API backend with FastAPI using an object oriented, modular approach that encapsulates components into modules for structured unit tested deployment and ci/cd powered development.

Demonstrates architecting a FastAPI back-end for an AI chatbot, including project setup with git, a dockerized API, middleware, componentized GPT chatbot, and environment-based API keys.

Discover fastapi as a python api backend with openapi docs and type checking for ai application development. The session covers credentials, calls to openai and hugging face, and response formatting.

Containerize your API with Docker to create a lightweight, isolated runtime that unifies development and deployment. Build images from a Dockerfile, install dependencies, and run the API.

Deploy your api on Hugging Face spaces by creating a public space with a docker-based template. Clone the repository, add app and dockerfile, push updates, and launch a public endpoint.

gradio, an open-source python library, lets you build user interfaces for machine learning models and prototype chatbots and image demos; learn to wire api calls and manage tokens for deployment.

Demonstrates deploying a Gradio application on Hugging Face spaces, setting up a front-end space, uploading files, debugging back-end and front-end, and testing a chatbot with llama and GPT versions.

DevOps unites development and operations through practices, culture, and people, enabling faster delivery and reliability via continuous integration and continuous deployment pipelines.

Explore software testing as a critical development phase, evaluating performance, functionality, and security to identify bugs and flaws. Define levels from unit and integration tests to system and end-to-end testing.

Explore software testing frameworks and pytest as an easy way to write unit tests in Python and validate applications.

Demonstrates writing Python code for a generative AI app, using VS Code, testing with FastAPI test client and pytest, validating a running server, endpoints, and model token generation.

Demonstrates building an automated test workflow with GitHub Actions for your app, including CI/CD tests on an Ubuntu runner with Python 3.9 and a Docker image, using Hugging Face tokens.

Discover domain annotations to provide context when data falls outside pre-trained data set. Compare three annotation approaches: in-context learning, RAC retrieval algorithm generation, and supervised fine-tuning on your own data.

Architect a fast api backend in Python for rapid development and production performance. Structure code with a modular, object-oriented design, supported by unit tests and ci/cd pipelines.

Set up the development environment, deploy a backend API to Hugging Face Space with a chat interface, and automate tests and pipelines with GitHub Actions.

Apply skills from this course by working backwards from real problems to ChatGPT-like solutions, while doing hands-on projects to test and learn from feedback.