Advaced Practical GenAI - English version

Disclaimer: The audio narration in this course is AI-generated, based on human-written scripts and human-designed slides. The use of AI narration is to improve clarity for learners, while all instructional content remains instructor-created.

This is Part 3 of the Practical GenAI Sequel.

The goal of this sequel is to prepare you to become a professional GenAI engineer or developer. We’ll start from the foundations of LLMs and GenAI and progress to building fully working, production-ready applications.

The sequel follows a hands-on approach. Every concept is taught through code-based examples, with final projects built step by step in Python, Google Colab, and deployed using Streamlit.

By the end of the full sequel, you will have built a diverse range of applications, including a ChatGPT clone, MidJourney-style image generator, Chat with Your Data app, YouTube Assistant, Ask YouTube Video app, Study Mate, Recommender system, Image Description app with GPT-V, Image Generation apps with DALL·E and Stable Diffusion, Video Commentator with Whisper, and more.

In this part, you will work with different kinds of LLMs—both open-source and proprietary. You’ll get hands-on exposure to:

• GPT models by OpenAI

• LLaMA models by Meta

• Gemini and Bard by Google

• Orca by Microsoft

• Mixtral by Mistral AI

• …and other emerging models.

You’ll learn how to use pre-trained models, and also how to fine-tune them on your own data.

We’ll dive into Hugging Face for model fine-tuning, leveraging Parameter-Efficient Fine-Tuning (PEFT) methods. You’ll work with Low-Rank Adaptation (LoRA) to train models efficiently.

You’ll also learn how to deploy models in the cloud, or host them privately when working with sensitive company data.

Finally, you’ll explore knowledge distillation, where existing pre-trained models act as “teachers” to help you train your own optimized custom models.

Disclaimer: The audio narration in this course is AI-generated, based on human-written scripts and human-designed slides. The use of AI narration is to improve clarity for learners, while all instructional content remains instructor-created.