DIY Gen AI: Running LLMs locally with LM Studio, Hugging Face, and Ollama

Updated! LM Studio is an easy to use desktop app for experimenting with local and open-source Large Language Models (LLMs) by hosting them on your own system.
Here are some of the features it provides (quoted from its homepage):

Run LLMs on your laptop, entirely offline
Use models through the in-app Chat UI or an OpenAI compatible local server
Download any compatible model files from HuggingFace repositories
Discover new & noteworthy LLMs in the app's home page

Hugging Face is a community hub focused on creating and sharing AI models. It provides many free and pre-trained models as well as datasets and tools to use with them.

Ollama is a command line tool for downloading, exploring, and using LLMs on your local system.

In this hands-on workshop, we'll cover the basics of getting up and running with LM Studio, Ollama and give you hands-on labs where you can use them and Hugging Face to find and load and run LLMs, interact with it via Chat and Python code and more!

Join author, trainer and speaker Brent Laster to learn about LM Studio, Hugging Face, Ollama, and Streamlit and how to use them to find and use Large Language Models hosted and running in your own environment. Get hands-on experience with the applications and learn how to DIY your own Gen AI!

Agenda:

Section 1: Introduction

In this section, we'll talk about what LLMs are, learn about basic use of LM Studio to find models and also start to look at huggingface.co.

Lab 1 - Lab 1 - Getting familiar with LM Studio and models

Purpose: In this lab, we’ll start to learn about models through working with one in LM Studio.

Section 2: Chatting with LLMs and using their APIs

In this section, we'll learn about how we can chat with an LLM, the different roles involved in chatting, and how to also use API calls from the command line to interact with models.

Lab 2 - Chatting with our model

Purpose: In this lab, we'll see how to load and interact with the model through chat and terminal.

Section 3 - Programming for local models

In this section, we'll look at how to create some Python code to interact with LM Studio with its lms interface and lmstudio.js library for JavaScript and Typescript.

Lab 3 - Coding to LM Studio

Purpose: In this lab, we'll see how to do some simple Python and JavaScript code to interact with the model.

Section 4 - Leverage HuggingFace.co

In this section, we'll look more into the model details and tools for using models that Hugging Face offers, including its transformers library and pipelines.

Lab 4 - Working with models in Hugging Face

Purpose: In this lab, we’ll see how to get more information about, and work directly with, models in Hugging Face.

Section 5 - Using Ollama

In this section, we'll learn about how we can use the standalone tool Ollama to get and run LLMs. We'll also talk about multimodal models.

Lab 5 - Using Ollama to run models locally

Purpose: In this lab, we’ll start getting familiar with Ollama, another way to run models locally.

Section 6 - Creating simple UIs for GenAI with Streamlit

In this section we'll work with a graphical Python library, Streamlit to see how to quickly and easily create interactive interfaces like chatbots to use with our local LLMs.

Lab 6 - Building a chatbot with Streamlit

Purpose: In this lab, we'll see how to use the Streamlit application to create a simple chatbot with Ollama.