# samuel-vitorino/lm.rs

Minimal LLM inference in Rust

Repository: https://github.com/samuel-vitorino/lm.rs
Canonical: https://ross.abutalabs.com/products/lmrs
Language: Rust
License: MIT
License Family: permissive
Last push: 2024-10-24T03:24:35+00:00

## Health v2 (maintenance only)
Score: 23/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 55
- inputs: {"age_days": 781, "days_push": 678, "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 1033, forks 43 (observed 2026-08-28T04:03:18.341002+00:00)

## What it is
A minimal Rust implementation for running LLM inference locally on CPU without ML libraries, supporting Gemma 2, Llama 3.2, and Phi 3.5 (including vision) models with quantization. Inspired by Karpathy's llama2.c and llm.c.

## Use cases
- run LLM inference locally on CPU
- run language models in Rust without ML frameworks
- run quantized Llama or Gemma models on a local machine
- run multimodal vision-language models on CPU
- learn how LLM inference works from minimal code

## When to choose
- you want dependency-free CPU-only LLM inference in Rust
- you want a small, readable codebase to study inference internals
- you need to run small quantized models (1B-9B) on modest hardware

## When to avoid
- you need GPU acceleration or maximum throughput
- you need broad model support or production-grade features
- you need a battle-tested inference server for production workloads

## Facets
- artifact type: library
- maturity: active
- function: llm-inference, machine-learning
- domain: large-language-models, machine-learning, developer-tools
- platform: rust, windows, cli
- tags: cpu-inference, quantization, minimal-implementation, multimodal, gemma, llama, phi-3, linux, macos

## Member repositories
- samuel-vitorino/lm.rs (main) score 23

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:18.341002+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 2026-08-30T07:06:41.487277+00:00, confidence not recorded.
  - readme: https://github.com/samuel-vitorino/lm.rs (fetched 2026-08-28T04:03:18.341002+00:00, sha 68934e29e285)
- Data as of 2026-08-30T08:39:29.467469+00:00.