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.

Notebooks
Google Colab
Kaggle
Local Apps Settings

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

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -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 serve -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 Desktop

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 serve -hf john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/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

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

Hermes Agent

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 serve -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

OpenClaw

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

Start the llama.cpp server

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

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "john-broadway/Qwen2.5-7B-RYS-8-12-GGUF:Q4_K_M" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

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.

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.

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

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 on Qwen2-72B; sweep toolkit by alainnothere.