Instructions to use Chungulus/Arete-27B-Q4_K_M-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 Chungulus/Arete-27B-Q4_K_M-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 Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Chungulus/Arete-27B-Q4_K_M-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 Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama cli -hf Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Use Docker

docker model run hf.co/Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Llama new
LM Studio
Jan

vLLM

How to use Chungulus/Arete-27B-Q4_K_M-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Chungulus/Arete-27B-Q4_K_M-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Chungulus/Arete-27B-Q4_K_M-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Ollama
How to use Chungulus/Arete-27B-Q4_K_M-GGUF with Ollama:
```
ollama run hf.co/Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M
```
Unsloth Desktop

How to use Chungulus/Arete-27B-Q4_K_M-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Chungulus/Arete-27B-Q4_K_M-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": "Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Chungulus/Arete-27B-Q4_K_M-GGUF with Docker Model Runner:
```
docker model run hf.co/Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M
```

Lemonade

How to use Chungulus/Arete-27B-Q4_K_M-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Arete-27B-Q4_K_M-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use Chungulus/Arete-27B-Q4_K_M-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 Chungulus/Arete-27B-Q4_K_M-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 Chungulus/Arete-27B-Q4_K_M-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Chungulus/Arete-27B-Q4_K_M-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Chungulus/Arete-27B-Q4_K_M-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 "Chungulus/Arete-27B-Q4_K_M-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"

Arete-27B - Q4_K_M

Q4_K_M GGUF of Nous Research's original BF16 Arete checkpoint. Arete is Nous Research's supervised fine-tune of Qwen3.8-27B. This is an independent Chungulus quantization, not an official Nous or Qwen release.

The original source has 27,356,728,560 total parameters, including its vision encoder. This main text GGUF contains 26,895,998,464 parameters. The full-precision mmproj-Arete-27B-BF16.gguf exported from the same source vision weights is included in every repository. MTP is disabled and absent from the source; it is not exported or synthesized.

Every format was quantized independently from the original BF16 GGUF export. No low-bit checkpoint was requantized, and no calibration dataset or importance matrix was used. All 18 original source shards were fully SHA256-verified before conversion; local tensor headers independently confirmed the complete parameter count. Sampled original BF16 text tensors differ from the pinned parent checkpoint.

Use

Use a recent llama.cpp runtime supporting Qwen3.5/Qwen3.8. The tested build is commit 6414d5b0c9e5895473036c90c2bd305ce4ceccc4 on Apple Silicon. Disable MTP/speculative prediction in frontends that attempt to enable it.

hf download Chungulus/Arete-27B-Q4_K_M-GGUF --local-dir ./Arete-27B-Q4_K_M
llama-completion -m ./Arete-27B-Q4_K_M/Arete-27B-Q4_K_M.gguf -cnv -st --jinja -c 8192 -ngl 99 -p 'Explain the difference between a list and a tuple in Python.'
llama-mtmd-cli -m ./Arete-27B-Q4_K_M/Arete-27B-Q4_K_M.gguf --mmproj ./Arete-27B-Q4_K_M/mmproj-Arete-27B-BF16.gguf --image input.png --image-min-tokens 1024 -c 8192 -ngl 99 -p 'Describe the central object.'

Actual validation and limits

This format passed real native-runtime arithmetic, JSON and factual checks, plus red/blue image-understanding fixtures with opposing central-square/corner-circle colors. Settings: temperature 0, seed 42, 8192 context, at most 128 generated tokens, thinking disabled, 1024 image tokens for vision. The actual GGUF type, every output tensor name/shape and quantized parameter fraction were inspected. See validation.json and quantizatia-manifest.json for exact results, pins, digests and provenance.

These are short functional checks. Political/value-alignment behavior, the creator's evaluation claims, benchmark accuracy, tool use, video and full-context quality were not evaluated. Source training claims and scores are not measured quantized-model results. The full-precision projector was preserved; no claim of unchanged model quality is made.

License and attribution

Apache 2.0, with original LICENSE included. Original model: Nous Research; base model: Alibaba/Qwen. Retain applicable license and attribution notices when redistributing. The GGUF export and Q4_K_M quantization are Chungulus modifications.