Model Quantization: Run LLMs Locally with Less RAM
Chapter 1

Breaking free from the cloud

Chapter 2

The silicon inside your machine

Chapter 3

The open source brain trust

Chapter 4

Living with a local assistant

YOUR GOAL

Master of Your private home AI

Ch 3 · The open source brain trust
007skaty
Open-source software lets you run powerful private AI offline on your home computer.

007skaty A desktop computer screen displays secure code running a private local AI model.

Oct 5

007skaty

007skaty Model Quantization

Squeeze the bits, free the RAM, run it all at home!

Oct 5

007skaty

007skaty The community hubs

Sep 4

007skaty
Open-source code lets you run a powerful, private AI model right on your laptop.

007skaty A personal laptop screen displays open-source code running a private model.

Oct 5

007skaty
Global models for global voices

007skaty Open-source groups worldwide are building diverse AI models to ensure local technology represents different languages, cultures, and values.

Sep 4

007skaty
Little Brain Trust
BETA

007skaty Sort incoming data into local drives to learn how offline computers run private AI.

Sep 4

007skaty
Fun Fact

Model squeezing: Quantization can compress an AI model to the point where a high-school level digital tutor can easily fit onto a standard 8-gigabyte USB thumb drive.

Hide sticker

Sep 4

007skaty
Glossary
hugging face (noun)
An emoji depicting a smiling face with open hands, typically used to express warmth, support, or a hug. Founded in 2016, this collaborative platform hosts thousands of open-source models available for local offline processing.

Sep 4

007skaty
Gossip

Who should control the future of AI?

Sep 4

007skaty
Quiz

What unit of measurement is used to describe the size of an AI model's brain, such as 7B or 70B?

Sep 4

007skaty

007skaty Model Quantization

Squeeze those sixteen bits down to four, now we can run it all on our own!

Oct 5

007skaty

007skaty The leak that started it all

Sep 4

007skaty
On March 3, 2023, an anonymous leak sparked the global movement for private local AI.

007skaty The anonymous 4chan post by llamanon containing the leaked LLaMA model torrent link.

Oct 5

007skaty
Quiz

Which company's leaked LLaMA model in 2023 accidentally triggered the open-source local AI boom?

Sep 4

007skaty

007skaty Global models for global voices

Sep 4

007skaty
Open source developers use quantization to shrink local AI by 90% without losing intelligence.

007skaty A diagram shows a massive AI model squeezing into a tiny local device.

Oct 5

007skaty
Quiz

Which online platform acts as the primary global community hub for sharing open-source AI models?

Sep 4

007skaty

007skaty Quantization and squeezing models

Sep 4

007skaty
Quantization and squeezing models

007skaty Quantization is a clever mathematical trick that compresses giant AI models so they can fit into a computer's standard memory.

Sep 4

007skaty
Quiz

What does 'open-source' mean when referring to local AI software models?

Sep 4