---
title: ThingAI
type: org
tags:
- slm
- llm
- pytorch
- bash
- shell
- code
- tokenizer
- rust
---

# ThingsAI

Building efficient, specialist Small Language Models that run on consumer hardware. Zero telemetry. Open weights. Everything from tokenizer to training script is public.

## Models

* **Dwarf-15M** 
  A 15.54M parameter shell/bash specialist. 12 layers, d_model=320, GQA 5Q/1KV, SwiGLU, RMSNorm, RoPE. Custom 8202-token vocabulary via DwarfGoToken.  1390:1 token-to-parameter ratio across 11 datasets spanning raw shell, Python, C, instruction pairs, and English web text. Target use case: CLI tool that translates natural language into bash commands with user review before execution.

* **Quark-270M**
  Our largest model. 252M effective parameters, 32 layers, d_model=768, GQA 12Q/4KV, 65K bilingual vocabulary (Italian + English). Trained on curated multilingual data. Available as Base and Instruct variants.

* **Quark-135M**
  Bilingual (Italian + English) general-purpose model. 135M parameters, 30 layers, 9 attention heads (3 KV, GQA), SwiGLU, RMSNorm, RoPE θ=10k. Trained on 15B+ tokens. Published benchmarks: HellaSwag 31.37%, ARC-Easy 41.46%, PIQA 61.26%.

## Links

* Models and tokenizers: [HuggingFace](https://huggingface.co/ThingAI)
* Script & Tool: [GitHub](https://github.com/overcastlab)
* Website: [things-ai.org](https://things-ai.org)
* GoToken: [crates.io](https://crates.io/crates/gotoken) · [PyPI](https://pypi.org/project/gotoken/)