Remember the good old days when we had to actually type out emails ourselves, or God forbid, write our own content? Then ChatGPT-3 came, the miraculous AI that saved us from our misery by doing all the hard work. OpenAI really outdid themselves, creating a model so advanced that it could almost pass for a human... well, a really smart, socially awkward human. ChatGPT-3 became the Swiss Army knife of text generation, handling everything from customer service to writing mediocre blog posts for wannabe influencers. Thanks to ChatGPT-3, businesses could pretend they cared about customer service without having to hire actual humans. Ah, progress!

The Launch of Llama3.1: A New Chapter in Open-Source AI

Just a few days ago, Meta (yes, the company formerly known as Facebook) launched Llama3.1, their latest and most ambitious open-source AI model yet. And it’s huge!!! like, 405 billion parameters huge, and that too open source!

While we could go on and on about how the colossal 405B model could out-chat ChatGPT in a battle of wits, let's talk about the real unsung heroes here: the smaller models. The 8B model, for instance, is like the underdog in a sports movie, small but mighty, and capable of running on your very own system. Yes, you heard that right—you don't need a NASA-grade supercomputer to harness its power. These 4-5GB sized models pack more data and context about the world than your average pub quiz champion, making cutting-edge AI as accessible as your local coffee shop's Wi-Fi. So, while the 405B model might be hogging the spotlight, it’s these lean, mean, 8B machines that are quietly revolutionizing the game. And guess what? We're about to try and install one of these marvels on our local system. Buckle up!

Testing Llama3.1 Locally: A Hands-On Guide

Here’s a step-by-step guide to getting started with Llama3.1 on your local machine:

Step 1: System Requirements

I am running this model on a Macbook M3 Pro, but technically you should be able to run it (maybe a bit slow) on any PC with 8GBs of RAM.

Step 2: Installation

First of all, download and install Ollama on your machine. Ollama is basically a platform designed to make it easier for developers to use large language models (LLMs) in their applications.

Step 3: Download and Run the Llama 3.1 model

Once you have Ollama downloaded and running on your system, just open the terminal window and type the following command

ollama run llama3.1

And that's it!

This would download the 8b version of llama 3.1 (this might take some time since the model size is close to 5gb). Once downloaded it would initiate a chatbot in the terminal which you can then chat with even without any external access to the internet!

Conclusion

This is just the beginning of my journey with Llama3.1. I'll be diving deeper into its capabilities, building custom bots, and exploring what it can truly do. I'll document every step, sharing insights and discoveries as I gain a deeper understanding of large language models. Join me as I uncover the full potential of Llama3.1 and delve into the fascinating world of LLMs. Stay tuned for more updates!