# openlm-research/open_llama

OpenLLaMA, a permissively licensed open source reproduction of Meta AI’s LLaMA 7B trained on the RedPajama dataset

Repository: https://github.com/openlm-research/open_llama
Canonical: https://ross.abutalabs.com/products/open_llama
License: Apache-2.0
License Family: permissive
Last push: 2023-07-16T13:42:13+00:00

## Health v2 (maintenance only)
Score: 30/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 87
- inputs: {"age_days": 1223, "days_push": 1144, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 7528, forks 404 (observed 2026-08-28T04:10:00.506766+00:00)

## What it is
OpenLLaMA is a permissively licensed (Apache-2.0) open reproduction of Meta AI's LLaMA, releasing 3B, 7B, and 13B pretrained model weights trained on the RedPajama dataset. It provides PyTorch weights for Hugging Face transformers and JAX weights for EasyLM, usable as drop-in replacements for LLaMA.

## Use cases
- download permissively licensed llama weights
- use an open source alternative to meta llama
- fine-tune a 7b open source language model
- run llama weights without meta's restrictive license
- compare open llama reproduction against original llama benchmarks
- get jax weights for easylm training

## When to choose
- you need LLaMA-style model weights under a permissive Apache-2.0 license
- you want drop-in replacements for LLaMA in Hugging Face transformers or EasyLM
- you need small-to-mid size open LLMs (3B/7B/13B) for research or fine-tuning

## When to avoid
- you need code generation performance out of the box (v1 tokenizer struggles with code tasks)
- you want the newest open models with instruction tuning or chat capabilities
- you lack the GPU resources to run multi-billion-parameter models

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-training, llm-inference, machine-learning
- domain: large-language-models, deep-learning, artificial-intelligence
- platform: python, cross-platform
- tags: open-llama, llama-reproduction, model-weights, huggingface, jax, pytorch, redpajama, pretrained-models, gpu

## Member repositories
- openlm-research/open_llama (main) score 30

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:00.506766+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:37:27.659139+00:00, confidence not recorded.
  - readme: https://github.com/openlm-research/open_llama (fetched 2026-08-28T04:10:00.506766+00:00, sha 4b80acbce07d)
- Data as of 2026-08-30T08:39:29.467469+00:00.