# henrywoo/pyllama

LLaMA: Open and Efficient Foundation Language Models

Repository: https://github.com/henrywoo/pyllama
Canonical: https://ross.abutalabs.com/products/pyllama
Language: Python
License: GPL-3.0
License Family: copyleft
Last push: 2023-11-08T06:48:00+00:00

## Health v2 (maintenance only)
Score: 21/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 8, longevity 91
- inputs: {"age_days": 1282, "days_push": 1029, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2778, forks 297 (observed 2026-08-28T04:07:21.500588+00:00)

## What it is
A Python library that is a modified version of Meta's LLaMA implementation, making it easy to download model weights and run LLaMA models on consumer-grade GPUs with as little as 4GB of memory via quantization. It includes CLI tools for downloading checkpoints and quantizing models at 2/3/4/8-bit precision.

## Use cases
- run llama models on a 4gb gpu
- download llama model checkpoints from the command line
- quantize llama to 4-bit
- run large language models locally on consumer hardware
- load llama weights with hugging face transformers

## When to choose
- you want to run original LLaMA models on a low-VRAM consumer GPU
- you need a simple CLI to download and quantize LLaMA checkpoints

## When to avoid
- you need the latest LLaMA 2 or 3 features
- you want a maintained library with active updates
- you prefer full-precision inference on datacenter GPUs

## Facets
- artifact type: library
- maturity: maintenance
- function: llm-inference, machine-learning
- domain: large-language-models, machine-learning, deep-learning
- platform: python, windows
- tags: llama, quantization, gptq, consumer-gpu, model-download, huggingface, gpu, linux, macos

## Member repositories
- henrywoo/pyllama (main) score 21

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:21.500588+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 2026-08-30T08:16:29.956609+00:00, confidence not recorded.
  - readme: https://github.com/henrywoo/pyllama (fetched 2026-08-28T04:07:21.500588+00:00, sha 2eaf9a333d0d)
- Data as of 2026-08-30T08:39:29.467469+00:00.