README.md

---
title: README
emoji: 📚
colorFrom: red
colorTo: purple
pinned: true
---

# Welcome to ThingsAI! 🤗

**Building efficient, bilingual AI models that run anywhere.** 🇮🇹 🇬🇧

---

## 🤖 Our Models

### [Quark-135M](https://huggingface.co/ThingAI/Quark-135m-Bilingual)
A lightweight bilingual (Italian + English) language model with **135M parameters**. Features GQA, SwiGLU, RMSNorm, and RoPE. Trained on 50B+ tokens of curated Italian and English data.

### Quark-270M *(training now)*
Our most powerful small model — **270M parameters** with 32 layers, 768 hidden dimensions, and 65K vocabulary. Currently in active training on 10B+ tokens with planned scaling to 135B tokens.

### [Quark-Mod](https://huggingface.co/ThingAI/Quark-Mod)
A multi-label moderation model covering **9 categories** for safe AI deployment: toxic, severe_toxic, obscene, threat, insult, identity_hate, cyberbullying, hate_speech, offensive.

---

## 🎯 What We Focus On

- ⚡ **Small, efficient architectures** — GQA, weight tying, deep-thin design
- 🌍 **Bilingual training** — Italian + English from scratch
- 🔓 **Open-source everything** — weights, code, datasets
- 💻 **Real-world deployment** — runs on consumer hardware

---

## 📂 Resources

| | |
|---|---|
| 📚 [Quark-135M-Bilingual](https://huggingface.co/ThingAI/Quark-135m-Bilingual) | Our flagship bilingual model |
| 🛡️ [Quark-Moderation-Multi](https://huggingface.co/ThingsAI/Quark-Mod) | Multi-label content moderation |
| 📝 [HuggingFace Community](https://huggingface.co/ThingsAI) | All our models and datasets |
| 💻 [GitHub](https://github.com/overcastlab) | Code, scripts, and tools |

---

<p align="center">
  <sub>Made with ❤️ by ThingsAI · <a href="https://things-ai.org">Website</a> · <a href="https://github.com/overcastlab">GitHub</a></sub>
</p>