README.md

---
title: README
emoji: 😻
colorFrom: yellow
colorTo: blue
sdk: static
pinned: false
short_description: Small, local models distilled from frontier teachers
---

<p align="center">
  <img src="http://serve.thelabsource.com/u/FhQgYP.gif" width="720" alt="Advanced Data Intelligence">
</p>

<h1 align="center">Advanced Data Intelligence</h1>

<p align="center">
  <strong>Small, local, open models — distilled from frontier teachers.</strong>
</p>

<p align="center">
  ADI is a line of compact language models built at <a href="https://thelabsource.com">theLAB</a> (<em>Learning. Algorithms. Breakthroughs.</em>). Each model is a <strong>knowledge distillation</strong>: a strong frontier "teacher" generates high-quality answers across thousands of prompts, and a small "student" model is fine-tuned to imitate them — producing a model that reasons and responds like something much larger, while staying small enough to run on a single consumer GPU.
</p>

<p align="center">
  Every model here is built end-to-end on theLAB hardware — no cloud training — then quantized to GGUF and shipped ready to run in <a href="https://ollama.com">Ollama</a> or any llama.cpp-based runtime.
</p>

<p align="center">
  <strong>Links:</strong>
  <a href="https://advanced-data-intelligence.com">Website</a> ·
  <a href="https://thelabsource.com">theLAB</a> ·
  <a href="https://www.youtube.com/@AdvancedDataIntelligence">YouTube — Advanced Data Intelligence</a> ·
  <a href="https://www.youtube.com/@adi_onlin3">YouTube — ADI Online</a>
</p>

<p align="center">
  <img src="https://serve.thelabsource.com/u/PjCf8w.png" width="700" alt="">
</p>

<p align="center">
  <img src="https://serve.thelabsource.com/u/oLHfvT.png" width="900" alt="New here? Start with one of these — adi-qwen3.5-4b, adi-qwen2.5-coder-7b, adi-qwen3-8b">
</p>

### 🐱 adi-qwen3.5-4b-glm5.2-general

General-purpose local assistant. Qwen3.5-4B distilled from **glm-5.2**. Reasons and explains like a frontier model on general topics. Native tool-calling, 262K context, ~2.7 GB.

```bash
ollama run hf.co/AdvancedDataIntelligence/adi-qwen3.5-4b-glm5.2-general-GGUF:Q4_K_M
```

### 🐱 adi-qwen3-8b-glm5.2-general

General-purpose local assistant. Qwen3-8B distilled from **glm-5.2**. Reasons and explains like a frontier model on general topics, with more headroom than the 4B. Native tool-calling, 128K context, ~5 GB.

```bash
ollama run hf.co/AdvancedDataIntelligence/adi-qwen3-8b-glm5.2-general-GGUF:Q4_K_M
```

### 🐱 adi-qwen3.5-9b-glm5.2-general

General-purpose local assistant. Qwen3.5-9B distilled from **glm-5.2**. The most capable general student in the line — more parametric headroom for nuanced reasoning while still fitting a single consumer GPU. Native tool-calling, 262K context, ~5.6 GB.

```bash
ollama run hf.co/AdvancedDataIntelligence/adi-qwen3.5-9b-glm5.2-general-GGUF:Q4_K_M
```

### 🐱 adi-qwen2.5-coder-7b-kimi2.7-code

Local coding assistant. Qwen2.5-Coder-7B distilled from **kimi-k2.7-code**. Writes, explains, and debugs code with frontier-style quality. Native tool-calling, 128K context, ~4.4 GB.

```bash
ollama run hf.co/AdvancedDataIntelligence/adi-qwen2.5-coder-7b-kimi2.7-code-GGUF:Q4_K_M
```

<p align="center">
  <img src="https://serve.thelabsource.com/u/uSv2Lp.png" width="760" alt="ADI model lineup — size on disk">
</p>

---

<h2 align="center">Browse the whole line</h2>

<p align="center">
  <strong><a href="https://huggingface.co/spaces/AdvancedDataIntelligence/adi-models-lab">ADI Models Lab</a></strong> — the full lineup in one place. Pick a student from the rail (Qwen3.5 4B, Qwen3.5 9B, Qwen3 8B, Coder 7B, and the <code>hey-adi</code> wakeword), read its teacher, context, and size at a glance, then copy a ready-to-paste run command. Includes the live in-browser demo — no install to try, no sign-in to copy.
</p>

<p align="center">
  <a href="https://huggingface.co/spaces/AdvancedDataIntelligence/adi-models-lab">
    <img src="http://serve.thelabsource.com/u/c8cTr3.gif" alt="ADI Models Lab — pick a student, copy a command, run offline" width="800">
  </a>
</p>

<p align="center">
  <em>Pick a student. Copy a command. Run offline.</em><br>
  <a href="https://huggingface.co/spaces/AdvancedDataIntelligence/adi-models-lab">▶ Open ADI Models Lab</a>
</p>

---

<h2 align="center">Try it live</h2>

<p align="center">
  A hosted demo is available as a Hugging Face Space — chat with the model directly in your browser, no install required.
</p>

<p align="center">
  <a href="https://huggingface.co/spaces/AdvancedDataIntelligence/adi-qwen3.5-4b-glm5.2-general-demo">
    <img src="https://serve.thelabsource.com/u/4Kb3iS.gif" alt="adi-qwen3.5-4b-glm5.2-general live demo" width="800">
  </a>
</p>

<p align="center">
  <a href="https://huggingface.co/spaces/AdvancedDataIntelligence/adi-qwen3.5-4b-glm5.2-general-demo">▶ Launch the demo</a>
</p>

---

## How to run

**Ollama (recommended).** Pull and run any model directly from this org — no manual download needed. Ollama fetches the GGUF from Hugging Face on first run:

```bash
ollama run hf.co/AdvancedDataIntelligence/adi-qwen3-8b-glm5.2-general-GGUF:Q4_K_M
```

Swap `:Q4_K_M` for another quant tag if a model ships multiple. To pull without running:

```bash
ollama pull hf.co/AdvancedDataIntelligence/adi-qwen3-8b-glm5.2-general-GGUF:Q4_K_M
```

**Manual download (llama.cpp or offline).** Grab the raw GGUF with the Hugging Face CLI:

```bash
huggingface-cli download AdvancedDataIntelligence/adi-qwen3-8b-glm5.2-general-GGUF adi-qwen3-8b-glm5.2-q4_k_m.gguf --local-dir .
```

Then point any llama.cpp-based runtime at the downloaded file.

<p align="center">
  <img src="https://serve.thelabsource.com/u/T5Kdlg.png" width="700" alt="">
</p>

## The approach

- **Distillation, not retraining.** We transfer a teacher's reasoning style and answer quality into a small student — not net-new facts. For raw recall, pair these with retrieval (RAG).
- **Local-first.** Every model runs fully offline on consumer hardware. No API, no data leaving the machine.
- **Open.** Apache-2.0 where the base license allows, with full training details on each model card so the work is reproducible.

<p align="center">
  <img src="https://serve.thelabsource.com/u/ih5dUC.png" width="760" alt="The ADI distillation pipeline">
</p>

---

## Coming next

In the pipeline, distilled the same way and headed here soon:

- **adi-qwen2.5-14b-glm5.2-general** — a larger general student with more parametric headroom.
- **adi-gemma3-12b-glm5.2-general** — a Gemma-based general distill, broadening the lineup beyond Qwen.

Follow the org to catch them on release.

## Naming

Models follow the pattern `adi-<base>-<size>-<teacher>-<purpose>` — so the name tells you the student base, its size, the teacher it learned from, and what it's tuned for.

<p align="center">
  <img src="https://serve.thelabsource.com/u/yZylMt.gif" width="560" alt="ADI">
</p>

<p align="center">
  <img src="https://serve.thelabsource.com/u/O8Cq1i.gif" width="560" alt="ADI">
</p>

<p align="center">
  <em>Built at <a href="https://thelabsource.com">theLAB</a> — Learning. Algorithms. Breakthroughs.</em>
</p>