Instructions to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="john-broadway/Qwen2.5-7B-RYS-8-12-GGUF",
	filename="Qwen2.5-7B-RYS-8-12-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Use Docker

docker model run hf.co/john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Ollama:
```
ollama run hf.co/john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
```

Unsloth Studio

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for john-broadway/Qwen2.5-7B-RYS-8-12-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for john-broadway/Qwen2.5-7B-RYS-8-12-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for john-broadway/Qwen2.5-7B-RYS-8-12-GGUF to start chatting

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Docker Model Runner:
```
docker model run hf.co/john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M
```

Lemonade

How to use john-broadway/Qwen2.5-7B-RYS-8-12-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen2.5-7B-RYS-8-12-GGUF-Q4_K_M

List all available models

lemonade list

Qwen2.5-7B-RYS-8-12-GGUF

File size: 2,507 Bytes

ef8d7f8

---
license: apache-2.0
base_model: Qwen/Qwen2.5-7B-Instruct
tags:
  - rys
  - layer-duplication
  - reasoning-circuits
  - gguf
  - sovereign-collection-v1
---

# Qwen2.5-7B-RYS-8-12

Qwen2.5-7B-Instruct with layers 8-12 duplicated. The math circuit runs twice on every forward pass.

28 base layers → 32 after duplication. No training, no merging, no weight changes.

**Math +10% (0.5653 → 0.6645). EQ +1.0. Reasoning held at 94.12%.**

## Results

| Metric | Baseline | RYS (8,12) | Delta |
|--------|----------|------------|-------|
| Math | 0.5653 | 0.6645 | +10% |
| EQ | 89.69 | 90.66 | +0.97 |
| Reasoning | 94.12% | 94.12% | 0.00 |

**The math specialist.** The first sub-14B model the method was extended to. David Ng demonstrated layer-duplication on Qwen2-72B; we found it works just as cleanly here, 10× smaller. Of 51 swept configurations, (8,12) is the one that gets the math circuit alone — duplicate that 4-layer block and nothing else trades down. The original GitHub-issue writeup describing exactly this result lives in the v2 corpus: [`docs/github-issue-draft.md`](https://huggingface.co/datasets/john-broadway/rys-sovereign-collection-v2/blob/main/github-issue-draft.md).

## Usage

```
llama-server -m Qwen2.5-7B-RYS-8-12-Q4_K_M.gguf -ngl 99
```

## Full sweep data

51 configurations tested. Full sweep data in the v2 corpus dataset. For vLLM-serving with AWQ quantization, see companion repo [`john-broadway/Qwen2.5-7B-RYS-8-12-AWQ`](https://huggingface.co/john-broadway/Qwen2.5-7B-RYS-8-12-AWQ).

Part of the RYS Sovereign Collection v1.

---

## Where this sits in the Sovereign Collection

**v1 — Qwen2.5 cross-scale + Qwen3-32B headline.** Four sizes from 0.5B to 32B; RYS works at every scale, with the lift size and dimension shifting by baseline:

- 0.5B → EQ specialist
- 1.5B → balanced daily driver
- 7B → math specialist via (8,12)
- 32B → the headline "Big Boy"

**v2 — cross-architecture extension.** 21 model variants across 10 architecture families. Headline: weak baselines lift more, in their weakest dimension. → [`john-broadway/rys-sovereign-collection-v2`](https://huggingface.co/datasets/john-broadway/rys-sovereign-collection-v2)


**Credit**

John Broadway, with collaboration from Claude (Opus 4.6 in April 2026 build; Opus 4.7 in May 2026 analysis and publication). Original RYS method by [David Ng](https://dnhkng.github.io/posts/rys/) on Qwen2-72B; sweep toolkit by [alainnothere](https://github.com/alainnothere/llm-circuit-finder).