Repository navigation
slides
marp: true theme: base style: | body { font-size: 40%; } class: invert
paginate: true backgroundColor: #000
backgroundImage: url('https://marp.app/assets/hero-background.svg')
If you are excited about AI and exploring ChatGPT and other Large Language Models (LLMs) and want to build AI applications, consider LangChain. LangChain is a python library that enables AI app development via composability.
We will explore the concepts behind LangChain in a demo heavy presentation that will spark your imagination to build AI agents, web apps and command line tools.
David Cobb is a freelance Azure teacher & consultant since 2020, and a Microsoft Certified Trainer (MCT) since 2002 when he began teaching SQL Server. He is an I.T. generalist, a fanatic about PowerShell and command line tools, and enjoys keeping up with cloud technology and teaching others with a casual hands-on approach.
Outline & Objectives
- ChatGPT has introduced the world to the possibilties of Large Language Models (LLMS)
- Schema
- Models
- Indexes
- Memory
- Chains
- Agents
The schema in LangChain refers to the structured representation of data and its organization. It defines the data types, fields, and relationships between different entities within the system.
Models in LangChain are pre-trained language models, such as OpenAI's GPT-3.5, that are utilized for various language-related tasks. These models have been trained on a vast amount of textual data and can generate human-like text, understand natural language, and perform language translation.
Indexes in LangChain provide efficient and fast access to the data stored within the system. They are used to optimize search operations and enable quick retrieval of information based on specific criteria.
The memory component in LangChain refers to the storage mechanism that holds the data and information processed by the system. It stores the input, output, and intermediate representations of language-related tasks.
Chains in LangChain represent the blockchain aspect of the platform. They are decentralized and distributed ledgers that record transactions, data, and smart contracts across the network. Chains ensure transparency, immutability, and security of the information stored within LangChain.
Agents in LangChain are intelligent entities that interact with the system and perform various tasks. They can be autonomous programs or human users that utilize LangChain's capabilities, such as language translation, content generation, or natural language understanding.