Instructions to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
# Run inference directly in the terminal:
llama cli -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
# Run inference directly in the terminal:
llama cli -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
# Run inference directly in the terminal:
./llama-cli -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
# Run inference directly in the terminal:
./build/bin/llama-cli -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

Use Docker

docker model run hf.co/magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

LM Studio
Jan

vLLM

How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-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": "magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-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/magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

Ollama
How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with Ollama:
```
ollama run hf.co/magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
```
Unsloth Desktop

How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

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": "magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with Docker Model Runner:
```
docker model run hf.co/magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S
```

Lemonade

How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

Run and chat with the model

lemonade run user.Qwen3.8-27B-MXFP4-MagicQuant-GGUF-UD-Q4_K_S

List all available models

lemonade list

Hermes Agent

How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

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 magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S

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 "magiccodingman/Qwen3.8-27B-MXFP4-MagicQuant-GGUF:UD-Q4_K_S" \
  --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"

MagicQuant software

by ItzPingCat - opened 8 days ago

Discussion

ItzPingCat

8 days ago

When you release it, will it allow us to quantize any model by using eg, unsloth as a template?
So if I have an abliterated version of a model and unsloth has published their quant (eg, IQ3) for that model, can I easily quantize my model to the same quality?

magiccodingman

Owner 7 days ago

Yes, that's exactly what it does. The entire system has a YAML to allow you to target unsloth models or any other, on top of llama.cpp and so on. And it will then run and built the model for you :)

ItzPingCat

7 days ago

Thank you! Will I need to provide an IMAT calibration dataset? Or does it somehow autofit it.

magiccodingman

Owner 6 days ago

Yes, imatrix is wanted for this for sure, but it's pretty easy. You can bring your own or easy mode you can point to unsloths gguf imatrix. I actually plan on releasing the code I'm hoping in the next few months. I honestly just need to clean it up. The issue right now is I've done so much research and it has evolved so much. It's kind of unwieldy right now. It needs some cleaning before I open source it and am comfortable maintaining it moving forward.

ItzPingCat

5 days ago

I'd love you to release what you have already.

I'm sure others will too.

Even if it isn't production ready, we can fix it ourselves.

magiccodingman

Owner 1 day ago

I'm going to spend some time this weekend finally cleaning it up and will try getting it released :) I would love for others to use it. Honestly the biggest hold back for me is just how terribly sloppy it is right now. It went through so many prototype phases, changed so often, I mean the main path it follows is called, "evolution". because my initial theory and experiment was an evolutionary search, which doesn't even exist anymore lol.

So, your messages have pushed me to finally finish cleaning and sharing it. The current wiki will become where I both host the wiki and the code base. So, I'll be updating it there :)

ItzPingCat

about 6 hours ago

Yayyyyy!!!!!!

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment