Hands-on Gen AI: From Neural Nets to Vector DBs (Virtual)

Workshop Date: Sep 13th 2024 

Topic: Hands-on Gen AI: From Neural Nets to Vector DBs

Global Big Data Conference is organizing full day Virtual Workshop (7hrs) on Sep 13th 2024 - Gen AI - The complete guide – Hands-on Gen AI with  neural nets, LLMs, RAG, vector DBs, and more! (full day)

Get hands-on learning to understand and utilize Generative AI from the ground. Work with key AI techniques and implement simple neural nets, vector databases, large language models, retrieval augmented generation and more - all in one single day session!

Generative AI is everywhere these days. But there are so many parts of it and so much to understand that it can be overwhelming and confusing for anyone not already immersed in it. In this full-day workshop, open-source author, trainer, and technologist Brent Laster will explain the concepts and working of Generative AI from the ground up. You’ll learn about core concepts like neural networks all the way through to working with Large Language Models (LLM), Retrieval Augmented Generation (RAG) and AI Agents. Along the way we’ll explain integrated concepts like embeddings, vector databases and the current ecosystem around LLMs including sites like HuggingFace and frameworks like LangChain. And, for the key concepts, you’ll be

doing hands-on labs using Python and a pre-configured environment to internalize the learning.

Agenda

Section 1: A quick introduction to AI

-------------------------------------

Content: In this section, we’ll cover what AI means, and the primary types of applications that AI

is currently used for. We’ll discuss machine learning and how it is being used. We’ll dive in to

learning about neural networks and understand how they function and how they form the

building blocks for larger AI functionality using layers.

Lab 1: Creating a simple neural network: In this lab, we’ll construct a simple neural network with

Python and see how it functions to do basic processing and learning.

Section 2: How AI represents unstructured data

----------------------------------------------

Content: In this section, we’ll start to see how models using neural networks can represent large

amounts of unstructured data. We’ll learn about mechanics like tokenizing, generating

embeddings and using vectors to represent relationships.

Lab 2: Understanding embeddings and vectors: In this lab, we’ll use some simple Python

constructs to illustrate how embeddings and vectors allow LLMs and NLP applications to

capture and represent data relationships.

Section 3: Training and Storing relationships

---------------------------------------------

Content: In this section, we’ll introduce Vector databases – a type of database specifically

designed to store data along with many dimensions of how it related to other data. We’ll cover

what these databases are, how they work, and discuss some commonly used ones. We’ll also

lightly touch on some of the mathematics involved.

Lab 3: Storing data with Vectors – In this lab, we’ll work with some simple vector storage

mechanisms and explore the relationships. We’ll also look at how to use a vector database in a

simple Python application.

Section 4: Models

-----------------

Content: In this section, we’ll discuss how the training and datastores work together to create

models. We’ll discuss what the various parameters of models mean and look at some

commonly available models. We’ll also look at different ways of running and interacting with

models. We’ll survey some common ways to run models locally vs calling out to public

instances. We’ll look at several popular applications . We’ll also

Lab 4: Working with LLMs – In this lab, we’ll use a tool called LMStudio to easily survey and

load a model and interact with it through a visual interface.

Section 5: Community and Frameworks.

------------------------------------

Content: in this section, we’ll discuss the model ecosystem and community by looking at

HuggingFace which is a collaboration platform that offers a collection of models and other

related pieces supported by various users and groups. We’ll also explore LangChain , a

framework for developing applications powered by LLMs.

Lab 5: Using Langchain - In this lab, we’ll use LangChain and Ollama to implement a simple

interface to an LLM.

Section 6: Fine-tuning and RAG

------------------------------

Content: In this section, we’ll discuss the issues with LLMs using out-dated or limited training

data and the two options to help: fine-tuning and RAG. We’ll look at the difference between the

two and discuss the complexities and advantages and limitations of each. We’ll also explore the

overall model for using RAG.

Lab 6: Implementing a simple RAG example – In this lab, we’ll use llamaindex to help

implement a RAG example where we can use our own local PDF files to augment the LLM and

see how to be able to query our documents just like the LLM.

Section 7: Agents

-----------------

Content: In this section, we’ll discuss what an AI agent is and how it is intended to operate. We’ll

look at use cases and some example implementations.

Lab 7: Creating an agent – in this lab, we’ll build a simple agent to do a particular task.

Section 8: Advanced AI workflows

---------------------------------

Content: In this section, we’ll look at how we can combine agents together to solve more

complex problems.

Lab 8: Leveraging multiple agents – In this lab, we’ll use our agent from the last section and

combine it with additional constructs to solve a higher-order problem.

Section 9: The future

---------------------

Content: In this section, we’ll discuss some of the additional ideas and implementations for

taking advantage of LLMs and increasing their utility and capabilities.

Section 10: Wrap-up

Speaker Bio

I'm Brent Laster - a global trainer and book author, DevOps director at a top technology firm, and founder and president of Tech Skills Transformations LLC. I've been working with and presenting at NFJS events for ten years now and it is always exciting and interesting.

Through my decades in programming and management,I've always tried to make time to learn and develop both technical and leadership skills and share them with others Regardless of the topic or technology, my belief is that there is no substitute for the excitement and sense of potential that come from providing others with the knowledge they need to help them accomplish their goals.

In my spare time, I hang out with my wife Anne-Marie, 4 children and a small dog in Cary, North Carolina and design trainings and write books.

NOTE: Agenda and speakers subject to change without notice

Refund Policy

No refunds will be given for cancellations.

Please note: Ticket prices are subject to increase or decrease, at the discretion of Global Big Data Conference, before and/or after you have made your purchase, and do not entitle the purchaser to a partial refund or credit.

Terms & Conditions In order to obtain a high-quality audience at the Conferences, Global Big Data Conference reserves the right to revoke any purchased tickets from an attendee without explanation.

If you have any questions concerning the event, please do not hesitate to contact [email protected] or Call 408-400-3769

Know what’s Happening Next — before everyone else does.

Get App

Host or Publisher Global Big Data Conference