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"

Qwen3.8-27B-MXFP4-MagicQuant-GGUF / magicquant-manifest /experiments /UD-IQ3_S-Unsloth /results.json

magiccodingman

Add files using upload-large-folder tool

dc44e1f verified about 1 month ago

Raw History Blame Contribute Delete

4.01 kB

	{
	"schemaVersion": 1,
	"createdUtc": "2026-08-26T08:30:23.975078+00:00",
	"experiment": "UD-IQ3_S-Unsloth",
	"status": "complete",
	"models": {
	"native": {
	"path": "<REPO_ROOT>/Qwen3.8-27B-Quark-AWQ-MXFP4-native.gguf",
	"bytes": 18892854880,
	"decimalGB": 18.89285488,
	"GiB": 17.595342248678207,
	"sha256": "1a37c48570811215fa1a9d2a493211293ca375d17287771e85d28642946e88fc"
	},
	"candidate": {
	"path": "<REPO_ROOT>/Qwen3.8-27B-Quark-MXFP4-UD-IQ3_S-Unsloth.gguf",
	"bytes": 11988330400,
	"decimalGB": 11.9883304,
	"GiB": 11.16500273346901,
	"sha256": "e165d77ffd43985ebee7763e6a9642cdc8e6168ffed5017d954f6c2fc059f719"
	},
	"savings": {
	"bytes": 6904524480,
	"decimalGB": 6.90452448,
	"GiB": 6.430339515209198,
	"percentOfNative": 36.545691605926315
	}
	},
	"tensorAudit": {
	"verificationPassed": true,
	"tensorCount": 866,
	"changedTensorCount": 282,
	"unchangedPayloadsVerifiedByteExact": 584,
	"nativeMxfp4PreservedCount": 224,
	"nativeMxfp4TensorCount": 496,
	"effectiveTypeCounts": {
	"F32": 360,
	"IQ1_S": 2,
	"IQ2_S": 21,
	"IQ2_XS": 12,
	"IQ2_XXS": 12,
	"IQ3_S": 127,
	"IQ3_XXS": 77,
	"MXFP4": 224,
	"Q2_K": 6,
	"Q3_K": 16,
	"Q5_K": 1,
	"Q6_K": 6,
	"Q8_0": 2
	}
	},
	"benchmark": {
	"domain": "general",
	"tokenTarget": 32768,
	"chunks": 15,
	"context": 2048,
	"threads": 4,
	"gpuLayers": 0,
	"gpuVisible": false,
	"nativeStandalonePpl": 5.8033,
	"nativeStandalonePplError": 0.11146,
	"paired": {
	"candidatePpl": 5.951552,
	"basePpl": 5.801511,
	"pplDifference": 0.150041,
	"pplRatio": 1.025862,
	"logPplRatio": 0.025534,
	"correlationPercent": 98.94
	},
	"kld": {
	"mean": 0.042218,
	"standardError": 0.000861,
	"maximum": 5.45832,
	"p99_9": 1.217797,
	"p99": 0.400938,
	"p95": 0.151289,
	"p90": 0.092036,
	"median": 0.018492
	},
	"tokenProbability": {
	"meanDeltaPercent": -0.711,
	"rmsDeltaPercent": 6.061,
	"sameTopPercent": 91.398
	},
	"scope": {
	"languageLogits": true,
	"visionProjectorEvaluated": false,
	"mtpTensorsEvaluated": false,
	"note": "llama-perplexity logged block 64 MTP tensors as unused; ordinary KLD primarily measures the changed token embedding and output tensors"
	}
	},
	"imatrixApplicability": {
	"supplied": true,
	"label": "Unsloth",
	"sha256": "0ee5b10bd0c2fa2127c6f4b43dbfe1efd71e383b63217af9dade1de36599f1c1",
	"entries": 496,
	"changedTensorsWithEntries": 272,
	"changedTensorCount": 282,
	"note": "Coverage is inferred from llama.cpp's per-conversion missing-weight messages; tensors present in the imatrix are supplied to the selected quantizer."
	},
	"artifacts": {
	"corpus": {
	"path": "<REPO_ROOT>/Experiments/MQ-IQ4_XS_1-Generic/benchmark/_ppl_corpora/ppl_corpus_general.txt",
	"bytes": 131514,
	"sha256": "5d38d98dce15f54e9a1a926187b6058e65cd8b3dd9b5cc2729b0a5fd249228b0"
	},
	"referenceLogits": {
	"path": "<REPO_ROOT>/Experiments/MQ-IQ4_XS_1-Generic/benchmark/native-reference/logits/kld_logits_general.bin",
	"bytes": 7621186460,
	"sha256": "604db2df7253cba18a6c0a569ed819ee3065d4ac20cd94ae72f4329e7bba8516"
	},
	"nativeLog": {
	"path": "<REPO_ROOT>/Experiments/MQ-IQ4_XS_1-Generic/benchmark/native-reference/perplexity_general.log",
	"sha256": "bade3445565a354969f7515045507b854bee83f6695eec7490b70eab9c6eebc2"
	},
	"candidateLog": {
	"path": "<REPO_ROOT>/Experiments/UD-IQ3_S-Unsloth/benchmark/candidate/perplexity_general.log",
	"sha256": "2c5839e28b216eddcb10f658398d76977cd605d13dc866c116b5ecbc849b3771"
	},
	"quantizationLog": {
	"path": "<REPO_ROOT>/Experiments/UD-IQ3_S-Unsloth/quantization.log",
	"sha256": "3a8478c77f23bc9adf6c570658134c4ad68ebbae55b5b49fde2f99e960f6a3f7"
	}
	}
	}