# tekaratzas/RustGPT

An transformer based LLM. Written completely in Rust

Repository: https://github.com/tekaratzas/RustGPT
Canonical: https://ross.abutalabs.com/products/rustgpt
Language: Rust
License: MIT
License Family: permissive
Last push: 2025-10-10T19:47:38+00:00

## Health v2 (maintenance only)
Score: 38/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 46, release rhythm 35, longevity 25
- inputs: {"age_days": 354, "days_push": 327, "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 3157, forks 267 (observed 2026-08-28T04:07:46.446710+00:00)

## What it is
A transformer-based large language model implemented entirely in pure Rust with no external ML frameworks, using only ndarray for matrix operations. It is an educational toy project demonstrating pre-training, instruction tuning, and full backpropagation from scratch, not a production-grade LLM.

## Use cases
- learn how transformers work under the hood
- implement an LLM from scratch in Rust
- study backpropagation and self-attention implementations
- run a tiny chat model without ML frameworks
- teach deep learning concepts with readable code

## When to choose
- you want to understand LLM internals with clean, modular Rust code
- you need a dependency-light educational transformer implementation
- you prefer Rust over Python for learning ML fundamentals

## When to avoid
- you need a production-grade or competitive LLM
- you want GPU-accelerated large-scale training
- you need a mature ecosystem of model weights and tooling

## Facets
- artifact type: library
- maturity: active
- function: llm-training, llm-inference, machine-learning, deep-learning, chatbot
- domain: large-language-models, deep-learning, machine-learning, education, tutorials
- platform: rust, cross-platform
- tags: transformer, from-scratch, no-ml-frameworks, ndarray, backpropagation, self-attention, adam-optimizer, toy-project, educational

## Member repositories
- tekaratzas/RustGPT (main) score 38

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:46.446710+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 2026-08-30T07:25:37.270414+00:00, confidence not recorded.
  - readme: https://github.com/tekaratzas/RustGPT (fetched 2026-08-28T04:07:46.446710+00:00, sha 30b326c15c72)
- Data as of 2026-08-30T08:39:29.467469+00:00.