Instructions to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf",
	filename="edgeguard-cypher-qwen3-4b-v0.8-feedback-validated.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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M

Use Docker

docker model run hf.co/ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M

LM Studio
Jan
Ollama
How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with Ollama:
```
ollama run hf.co/ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
```

Unsloth Studio

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf to start chatting

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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": "ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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 "ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-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"

Docker Model Runner
How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with Docker Model Runner:
```
docker model run hf.co/ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M
```

Lemonade

How to use ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf:Q4_K_M

Run and chat with the model

lemonade run user.edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf-Q4_K_M

List all available models

lemonade list

EdgeGuard Cypher Qwen3 4B v0.8 Feedback-Validated GGUF

Private GGUF export for EdgeGuard direct text-to-Cypher edge-node runtime use. This preview continues EGM-026 v0.8 feedback-validated LoRA.

Artifact

File: edgeguard-cypher-qwen3-4b-v0.8-feedback-validated.Q4_K_M.gguf
Format: GGUF
Quantization: Q4_K_M
SHA256: 1da636448aafe9f2ff1826008e5f6f10ccd33488172c386f84783b0f5b485f77
Size bytes: 2497278816

Source

Base model: Qwen/Qwen3-4B-Instruct-2507
Base revision: cdbee75f17c01a7cc42f958dc650907174af0554
Source adapter SHA256: 9dced541868ccb6a8977591b8e79bc1ce4824f2cacebb0e369366ac529addf8d
Training target: assistant output is one valid read-only Cypher query string only.

Method

Merged EGM-026 QLoRA adapter into pinned Qwen/Qwen3-4B-Instruct-2507 revision cdbee75f17c01a7cc42f958dc650907174af0554, converted to BF16 GGUF, then quantized to Q4_K_M for edge-node.

Evaluation Summary

Best raw EdgeGuard Cypher model so far on the frozen v0.8 bundle: validation subgraph accepted 99/100, sealed-test subgraph accepted 98/100, probe subgraph accepted 197/200, scalar regressions 0. Generated-live hard success is 173/180 = 96.1% when no-graph rows are non-blocking; no-graph rows are 13.

Promotion Status

Private edge-node GGUF publication. No-graph live results are accepted as empty/coverage results; hard live failures remain 4 planner failures and 3 execution failures. Use behind the EdgeGuard schema/read-only validator.

Runtime

The intended consumer is edge-node LLM_INFERENCE_API through LlamaCppBaseServingProcess, using either Hugging Face MODEL_NAME + MODEL_FILENAME or a mounted local MODEL_PATH.

Recommended edge-node values:

AI_ENGINE=edgeguard_qwen_4b
MODEL_NAME=ratio1/edgeguard-cypher-qwen3-4b-v0.8-feedback-validated-gguf
MODEL_FILENAME=edgeguard-cypher-qwen3-4b-v0.8-feedback-validated.Q4_K_M.gguf

This model must be used behind the EdgeGuard schema/read-only validator. Do not execute raw model output directly against Neo4j.

Downloads last month: 73

GGUF

Model size

4B params

Architecture

qwen3

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support