LLM Data Science Bootcamp

Foundations of LLMs and Python Basics

• Overview of large language models
• Real-world applications of LLMs
• Python essentials for LLMs
• Common data structures and algorithms
• Core python libraries for LLM development

Understanding Natural Language Processing

• Overview of foundational NLP techniques
• Hands-on development of bigram models
• Deriving embeddings from natural language
• Common language processing techniques such as tokenization

Transformers and Attention

• Introduction to LLM’s core technology; Transformers
• Dive into transformer architecture
• Overview of attention mechanism
• Next token predictions via autoregressive model
• Hands-on with Hugging Face Transformers library

LLM Development: Fine-tuning and Prompt Engineering

• Overview of LLM fine-tuning techniques
• Hands-on fine-tuning of pre-trained models
• Prompt engineering walkthroughs and best practices
• Introduction to Few-shot prompting
• Exploring LLM-provider cookbooks for optimal LLM results
• Pre-training vs. Supervised Fine-tuning (SFT) vs. Prompt Engineering
• Reinforcement Learning for tuning LLMs

Retrieval-Augmented Generation (RAG)

• Explore RAG and why it is the most popular LLM use-case
• Exploring document chunking, embedding, and retrieval
• Vector RAG vs. Graph RAG
• RAG for reducing LLM hallucination
• Create your own RAG application

Introduction to LLM Agents

• Overview of Agent frameworks and motivation
• LLMs improving from feedback
• Using tools with LLMs to perform complex tasks
• Create your own Agents
• Planning and Execution
• Reflection, Chain-of-Thought (CoT), and multi-hop

Advanced Topics for Production LLM Application

• Releasing and monitoring LLMs in production
• LLM evaluations and learning from feedback
• Pre-processing techniques (intent detection, routing, safety, retrieval)
• Post-processing techniques (attribution, hallucination)
• Practical exercises with advanced LLM techniques
• Kickstart your own LLM project

• Interactive Sessions: Engage in live discussions with experts.
• Video Lessons: Access curated video content.
• Hands-on Workshops: Apply concepts to real-world scenarios.
• Custom Projects: Build a portfolio of LLM applications.

• Proficiency in LLMs and Python
• Practical experience with LLM libraries
• A capstone project to showcase your skills

• Proficiency in LLMs and Python
• Practical experience with LLM libraries
• A capstone project to showcase your skills