← All notes

Summary

Llama 2: Open Foundation and Fine-Tuned Chat Models (2023) released the Llama 2 foundation models and their chat variants, with a much more complete public recipe than the first Llama paper: pretraining, fine-tuning, and the safety and preference process used for the assistant models.

What it introduced: open-weight Llama 2 models, including chat versions, and a written account of the data, training, and alignment steps that produced them.

What it led to: a large part of the open-assistant ecosystem in 2023–2024, and the direct predecessor of the Llama 3 herd paper pointed to in this batch. For later preference-training methods used on these bases, see 2305-18290-direct-preference-optimization and 2204-05862-helpful-harmless-rlhf.

Together with Llama 3, it is the usual open-weight starting point for many later fine-tunes. Where to read it: https://arxiv.org/abs/2307.09288 (v2, last revised 19 Jul 2023).

The paper is posted on arXiv under the arXiv non-exclusive distribution license. This note links to it rather than reproducing it. The note itself is CC BY 4.0.

License: this note is CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). Pointer by iinventedlight (AI-assisted). Paper license: arXiv non-exclusive distribution; this note links to it and does not reproduce it.

@misc{iil_note_rdfg7xdf,
  author = {{'iinventedlight (AI-assisted)'}},
  title = {{Llama 2: Open Foundation and Fine-Tuned Chat Models}},
  howpublished = {iinventedlight Library},
  year = {2026},
  date = {2026-11-11},
  url = {https://iinventedlight.com/library/notes/2307-09288-llama-2/},
  note = {Permanent ID iil:note:rdfg7xdf. License: CC-BY-4.0.},
}

Markdown

Spotted a mistake? Use the MCP feedback tool or email hello@iinventedlight.com.

For language models and agents: start at /llms.txt | /agents.md | /.well-known/agent-library.json. This page in Markdown: /library/notes/2307-09288-llama-2.md.