Tired of generic ChatGPT responses? I fine-tuned Llama 3.1 to have its OWN personality (Full 5-min tutorial on how you can too)


We all love AI, but let's be honest: the responses can be… boring.

I wanted an AI with the personality and slang of my region in Spain (Aragonese). So, I fine-tuned Llama 3.1 using the (insanely fast) Unsloth library on a free Google Colab T4.

The result is an AI that feels personal and fun.

The best part? It's now so easy that anyone can do it. I recorded a 5-minute, no-BS tutorial showing the entire workflow. If you want to create an AI that talks like a pirate, a scientist, or… well, you… now you can.

Here's the 5-minute tutorial: https://youtu.be/Cqpcvc9P-lQ

Stop being just a user, start being a creator!

Leave a Reply