AI-434: Closed- and Open-source LLM Application Developer
(GPT, Google Gemini, Llama, Mistral)
Course description: Artificial intelligence has become an extremely important area for IT professionals and engineers in the past 10-20 years with the scientific breakthroughs and practical applications of deep learning and more recently of generative AI systems, especially with its Large Language Model (LLM) variant such as OpenAI’s ChatGPT, Google’s Gemini and many open-source models. Due to its importance and impact on every aspect of our lives, understanding the concepts, functionalities and practical usage of AI systems is quickly becoming essential for all IT and other technical professionals as well as for managers with technical background.
This training focuses on LLM concepts as well as GPT, Gemini and Open-source LLM prompt engineering and application development, and teaches participants the following topics:
- Introduction to LLM based applications
- The Foundation Technologies of LLMs (Neural Networks, Tokenizer, Transformer)
- The 3-phase training process of LLMs (pre-training, fine-tuning, RLHF)
- Using LLMs via APIs
- Prompt engineering
- Retriever Augmented Generation (RAG)
- Creating LLM chains with LangChain
- LLM Agents
- Fast Web Interface Prototyping for LLMs (Gradio)
- Debugging and Evaluating LLM-based apps (Langsmith)
- Fine-tuning Open-source LLM models
- Basics of AI Threats and LLM Security (optional)
Besides gaining a basic understanding of the concepts of prompt engineering, students will also do extensive lab exercises using the Python APIs of the GPT as well as popular and powerful Open-source LLMs to see how these concepts work in practice.
Course Length: 32 training hours
Structure: 50% theory, 50% hands on lab exercises
Target audience: Software developers and other IT and technical professionals as well as managers with technical background who want to understand the concepts, techniques, and best practices of prompt engineering and application development with the Python APIs of GPT, Gemini as well as the popular and powerful Llama and Mistral open-source LLMs.
Prerequisites: Basic understanding of AI concepts. basic Python programming skills.