| Before we start | |||
| Why this course is different | 00:01:00 | ||
| Prerequisites | 00:01:00 | ||
| Essential topics and terms (theory) | 00:04:00 | ||
| Why this course does not cover Open Source models like LLama2 | 00:01:00 | ||
| Optional: Install Visual Studio Code | 00:02:00 | ||
| Get the source files with Git from Github | 00:02:00 | ||
| Create OpenAI Account and create API Key | 00:02:00 | ||
| Preparation | |||
| Setup of a virtual environment | 00:03:00 | ||
| Setup OpenAI Api-Key as environment variable | 00:03:00 | ||
| Exploring the vanilla OpenAI package | 00:03:00 | ||
| LangChain Basics | |||
| LLM Basics | 00:07:00 | ||
| Prompting Basics | 00:02:00 | ||
| Theory: Prompt Engineering Basics | 00:02:00 | ||
| Few Shot Prompting | 00:05:00 | ||
| Chain of thought prompting | 00:02:00 | ||
| Pipeline-Prompts | 00:04:00 | ||
| Prompt Serialisation | 00:03:00 | ||
| Chains - From basic to advanced chains | |||
| Introduction to chains | 00:01:00 | ||
| Basic chains – the LLMChain | 00:03:00 | ||
| Response Schemas and Output Parsers | 00:06:00 | ||
| LLMChain with multiple inputs | 00:02:00 | ||
| Sequential Chains | 00:04:00 | ||
| Router Chains | 00:04:00 | ||
| Callbacks | |||
| Call backs | 00:05:00 | ||
| Memory | |||
| Memory basics – Conversation Buffer Memory | 00:04:00 | ||
| Conversation Summary Memory | 00:03:00 | ||
| EXERCISE: Use Memory to build a streamlit Chatbot | 00:01:00 | ||
| SOLUTION: Chatbot with Streamlit | 00:03:00 | ||
| OpenAI Function Calling | |||
| OpenAI Function Calling – Vanilla OpenAI Package | 00:08:00 | ||
| Function Calling with LangChain | 00:04:00 | ||
| Limits and issues of the langchain Implementation | 00:03:00 | ||
| Retrieval Augmented Generation (RAG) | |||
| RAG – Theory and building blocks | 00:03:00 | ||
| Loaders and Splitters | 00:04:00 | ||
| Embeddings – Theory and practice | 00:04:00 | ||
| Vector Stores and Retrievers | 00:07:00 | ||
| RAG Service with Fast API | 00:05:00 | ||
| Agents | |||
| Agents Basics – LLMs learn to use tools | 00:06:00 | ||
| Agents with a custom RAG-Tool | 00:07:00 | ||
| Chat Agents | 00:03:00 | ||
| Indexing API | |||
| Indexing API – keep your documents in sync | 00:02:00 | ||
| PREREQUISITE: Docker Installation | 00:01:00 | ||
| Setup of PgVector and Record Manager | 00:04:00 | ||
| Indexing Documents in practice | 00:06:00 | ||
| Document Retrieval with PgVector | 00:03:00 | ||
| LangSmith | |||
| Introduction to Lang Smith (User Interface and Hub) | 00:02:00 | ||
| Lang Smith Projects | 00:07:00 | ||
| Lang Smith Datasets and Evaluation | 00:13:00 | ||
| Microservice Architecture for LLM Applications | |||
| Introduction to Microservice Architecture | 00:04:00 | ||
| How our Chatbot works in a Microservice Architecture | 00:02:00 | ||
| Introduction to Docker | 00:05:00 | ||
| Introduction to Kubernetes | 00:02:00 | ||
| Deployment of the LLM Microservices to Kubernetes | 00:13:00 | ||
| LangChain Expression Language (LCEL) | |||
| Intro to Lang Chain Expression Language | 00:01:00 | ||
| LCEL Part 1 – Pipes and OpenAI Function Calling | 00:07:00 | ||
| LCEL – Part 2 – Vector Stores, Item Getter, Tools | 00:06:00 | ||
| LCEL – Part 3 – Arbitrary Functions, Runnable Interface, Fallbacks | 00:07:00 | ||