Llama 2 by Meta AI

Llama 2 by Meta AI

Meta has released LLaMA 2 (Large Language Model Meta AI), a foundational large language model designed for open science and AI research.

LLaMA's smaller, efficient models allow broader research access, especially for those with limited infrastructure, fostering democratization in this rapidly evolving field. These models are ideal for various tasks due to their training on a vast set of unlabeled data. However, challenges such as bias and toxicity remain, necessitating further research.

Access to LLaMA is granted under a noncommercial license for specific research purposes, aiming to encourage responsible AI development and use within the AI community.

Changes on Llama 2

Llama 2 was trained with 40% more data than Llama 1, and has double the context length.

Llama 2 details
Llama 2 pretrained models are trained on 2 trillion tokens, and have double the context length than Llama 1. Its fine-tuned models have been trained on over 1 million human annotations.


Llama 2 outperforms other open source language models on many external benchmarks, including reasoning, coding, proficiency, and knowledge tests.

More model details

Llama Chat

Llama 2 was pretrained on publicly available online data sources.
The fine-tuned model, Llama Chat, leverages publicly available instruction datasets and over 1 million human annotations.

Code Llama

Code Llama is a code generation model built on Llama 2, trained on 500B tokens of code. It supports common programming languages being used today, including Python, C++, Java, PHP, Typescript (Javascript), C#, and Bash.

Download the model

More details here:

Llama 2 - Meta AI
Llama 2 — The next generation of our open source large language model, available for free for research and commercial use.