Generative Ai: Build And Deploy A Llm Chatbot

Build and deploy a LLM-based chatbot to answer questions using your private dataset.

What you'll learn

Build and deploy a LLM chatbot to answer questions using your private dataset.

Deep dive into LLMs and their use cases.

Understand the differences between fine tuning vs RAG methodology.

Extra: Learn how to programmatically transcribe videos and audio as part of the data collection step!

End to end chatbot build: includes all the components of a RAG pipeline and explains how to improve accuracy for each.

Requirements

Mid to Senior level understanding of Python and Data Science concepts. Feel free to try this out as a beginner as well, and ask questions along the way!

Description

Build and deploy a LLM-based chatbot to answer questions using your private dataset!To build, we will use Anthropic’s Claude 2 LLM on Amazon Bedrock and Langchain, with RAG implementation. I’ll also show you the code for using other LLMs (OpenAI’s ChatGPT, GPT-4, etc.), in place of Claude 2, as well as other embedding models in place of Amazon Titan Text Embeddings. To deploy, we will use Gradio. This is an end to end build of a chatbot solution that can be used within your organization. Use this Generative AI solution to improve things across your organization like; enhance customer support, streamline information retrieval, aid in the training and onboarding of new employees, promote data-driven decision making, customize insights for clients and customers, and enable efficient knowledge sharing.I also cover programmatic audio/ video transcription within the data collection step. This can be applied to other use cases within your organization, outside of the chatbot. E.g. transcribe audio/video content to build content recommendation models, etc.This course is best for those with mid to senior level Python and Data Science understanding. For more beginner levels, feel free to dive in and ask questions along the way.  Hopefully you all enjoy this course and have fun with this project!

Overview

Section 1: Introduction

Lecture 1 Introduction

Section 2: Generative AI, LLMs, and Benefits of Our Project

Lecture 2 Generative AI, LLMs, and Benefits of our Project

Section 3: Fine Tuning vs Using RAG Methodology

Lecture 3 Fine Tuning vs Using RAG Methodology

Section 4: Environment Setup

Lecture 4 Set up your Environment

Section 5: Build Pt 1: Dataset Collection and Audio/Video Transcription

Lecture 5 Dataset Collection and Audio/Video Transcription

Section 6: Build Pt 2: Generate Embeddings and Create Vectorstore

Lecture 6 Generate Embeddings and Create Vectorstore

Section 7: Build Pt 3: LLM Component

Lecture 7 LLM Component

Section 8: Build Pt 4: Deployment, Summary and Concluding Thoughts

Lecture 8 Deployment, Summary and Concluding Thoughts

For those with mid to senior level Python/Data Science experience. Feel free to try this out as a beginner as well, and ask questions along the way!