File size: 3,069 Bytes
986bcfb ac5a01b 76e3527 ac5a01b c3cbc12 ac5a01b 0c98c41 ac5a01b 0c98c41 ac5a01b 3a623e1 953a6f0 c3cbc12 3a623e1 953a6f0 c3cbc12 3a623e1 ac5a01b 3709c4b ac5a01b a052a6d | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 | ---
title: README
emoji: 📊
colorFrom: red
colorTo: green
sdk: static
pinned: false
---
<p align="center">
<img src="https://cdn-uploads.huggingface.co/production/uploads/6921fa6332f7fb129563d495/jemMkfi73Fck611ID18ts.png" width="128" alt="Nomi Logo">
</p>
<h1 align="center">Jally AI by Lazy Loop Studio</h1>
<p align="center">
Small, fast & indie‑built language models for real people, not just benchmarks.
</p>
---
## 🌱 Our Models
### Nomi
Nomi is a series of **compact, high‑quality language models** designed to run on everyday hardware while still feeling surprisingly capable.
Instead of relying on massive data centers, Nomi focuses on efficient 3B models that you can run locally, privately, and with minimal setup.
Our models are primarily optimized for **clean structure**, **solid Markdown**, and **useful code**, making them a great fit for developers, indie projects, and anyone who wants full control over their AI stack.
---
## 🧩 Indie Spirit
Nomi AI is a small, independent project driven by a passion for **small but strong models**, not by a corporate budget.
We experiment with modern base models, fine‑tune them on carefully selected datasets, and try to squeeze as much “brain” as possible out of just 3B parameters.
If you care about local, open, and hackable AI tools, you are in the right place.
---
## 🔧 What We Build
- 🧠 3B models that feel like larger models (e.g., Nomi‑1.0, Nomi‑1.1).
- 📝 Assistants focused on **structured reports**, **Markdown**, and **developer workflows**.
- 🌍 Models with strong **German/English support** for European developer setups.
Whenever possible, our models are released under the **Apache‑2.0 license**, so you can freely use, modify, and integrate them into your own projects.
---
## Our Models
- **Nomi‑1.0‑3B** – Base model with core instruction tuning and strong Markdown + German/English support (GGUF and .safetensors)
👉 https://huggingface.co/Jally/Nomi-1.0-3b
- **Nomi‑1.1‑3B** – Refined behavior, better formatting, and improved multilingual abilities, optimized for local inference
👉 https://huggingface.co/Jally/Nomi-1.1-GGUF
---
## 🗒️ Our Plan
Right now, we are fully focused on Nomi. Our goal is to develop the series up to Version 2.0 and make it competitive with much larger models:
- Nomi 1.0 → Basic instruction tuning
- Nomi 1.1 → Improved behavior plus some code and math
- Next step → Stronger coding capabilities
- After that → Stronger math and reasoning
- One step further → Agentic, tool‑using tasks that can stand alongside bigger models
We are also thinking about making vision and reasoning models
---
## 🤝 Join the Journey
We welcome feedback, benchmarks, pull requests, and experiments using Nomi in your own projects.
If you are building on laptops, mini PCs, or homelabs and want something fast and lightweight, give Nomi a try and tell us what you built with it.
Check out our homepage: [https://lazyloopstudio.netlify.app](https://lazyloopstudio.netlify.app)
|