Chatbot and LLM Bootcamp
Conversational Agents using Modern Natural Language Processing
ChatGPT fundamentally changed the world of Artificial Intelligence (AI). ChatGPT is a Chatbot driven by the GPT3.5/GPT4 Large Language Models (LLMs) from OpenAI. These models are neural networks with billions of parameters, trained on vast amounts of text from the Internet and other sources and provide human-level intelligence on a wide variety of tasks. ChatGPT makes them available through an interactive, conversational prompt. Users can ask it to do anything. ChatGPT power users quickly sought to solve new and unexpected problems with ChatGPT and introduced the world to the field of Prompt Engineering.
Prompt Engineering is the science and art of using libraries like Langchain which is an application framework for conversational agents along with document processing libraries to ingest documents, language models and embeddings to encode text on the semantic language of an LLM, vector databases and search engines to provide an agent with memory and prompt formatting modules to provide an LLM with the context to bring its knowledge to bear on a problem.
Chatbots and LLMs are highly disruptive AI technologies. While the importance of Chatbots like ChatGPT is evident to anyone that uses them (try ChatGPT now if you haven’t) and considers possible use cases. The most obvious Chatbot use case is customer service, which has seen widespread adoption. Increasingly sophisticated Chatbot customer service representatives are rapidly replacing human jobs, resulting in significant savings. Customer service agents are just the beginning.
This trend will continue, driving mass unemployment across a swath of industries and job functions. If you don’t adapt, your company’s business model will be disrupted and left behind.
The consequences of not embracing these technologies can be severe. Sending your technical staff to this course is a way to plant seeds of innovation that will spread across your organization at an accelerated rate.
What You will Learn
In this course we will introduce students to building conversational agents using Python, Graphics Processing Units (GPUs), natural language processing, representation learning, language models and text embeddings, document parsing and vector search, prompt templates and large language models combined through the Langchain Chatbot Framework to conduct Prompt Engineering and Prompt Learning to solve Machine Learning problems via conversational agents driven by Large Language Models (LLMs).
Students will leave the course with working code for their own Chatbot project using cutting-edge technology that solves real-world problems. This project and its code can serve as the foundation for additional work in students’ employers’ business domain to create custom Chatbots for their companies.
The course isn’t limited to building your own Chatbots – Students will use LLMs to perform several different Machine-Learning tasks. Students will learn how to solve AI problems using LLMs without building Chatbots. Chatbots, LLMs and Prompt Learning are powerful tools, regardless of the form your application takes.
Students will use their basic Python data skills to learn to:
- Build their own Chatbots driven by LLMs
- Grasp the theory and fundamentals behind modern LLM technologies
- Build Chatbots using the popular Langchain framework
- Interact with remote LLM APIs like ChatGPT
- Use open-source LLMs locally for privacy
- Fine-tuning open-source LLMs like LLaMa for your business domain
- Entity Resolution (ER): de-duplicate records using Prompt Engineering
- Utilize vector database and search engines for semantic search
- Add memory to Chatbots to add context to chat sessions
Timing: from 2 pm to 6 pm Italian time