Fine-tune Llama 2 in Google Colab

This colab shows how to fine tune llama2 using a single T4 GPU with 15GB VRAM using LoRa.

The colab can be found at here.

A peak at the instruction fine tuning database: