Instructions to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="magiccodingman/Qwen3.6-27B-MagicQuant-GGUF",
	filename="Qwen3.6-27B-LM-IQ2_M.gguf",
)

llm.create_chat_completion(
	messages = [
		{
			"role": "user",
			"content": "What is the capital of France?"
		}
	]
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_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 magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
./llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_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 magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Use Docker

docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

LM Studio
Jan

vLLM

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "magiccodingman/Qwen3.6-27B-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/Qwen3.6-27B-MagicQuant-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Ollama
How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Ollama:
```
ollama run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
```

Unsloth Studio new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-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 magiccodingman/Qwen3.6-27B-MagicQuant-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 magiccodingman/Qwen3.6-27B-MagicQuant-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for magiccodingman/Qwen3.6-27B-MagicQuant-GGUF to start chatting

Pi new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_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": "magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_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 magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Run Hermes

hermes

Docker Model Runner
How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Docker Model Runner:
```
docker model run hf.co/magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M
```

Lemonade

How to use magiccodingman/Qwen3.6-27B-MagicQuant-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull magiccodingman/Qwen3.6-27B-MagicQuant-GGUF:IQ2_M

Run and chat with the model

lemonade run user.Qwen3.6-27B-MagicQuant-GGUF-IQ2_M

List all available models

lemonade list

Qwen3.6-27B-MagicQuant-GGUF / magicquant-manifest /magicquant.replacements.json

magiccodingman

Add files using upload-large-folder tool

60fadc3 verified 14 days ago

raw

history blame contribute delete

15.5 kB

	[
	{
	"reasonCode": "SPACING_COLLAPSE",
	"reasonDescription": "Two candidates were too close in practical output space; the stronger one was kept.",
	"rawReason": "meaningful spacing collapse; size gap below 564,472,013 bytes",
	"removed": {
	"key": "0:1:2:1:1:1:1:2:0:0",
	"fileName": "Qwen3.6-27B-MQ-Q6_K.gguf",
	"displayName": "Qwen3.6-27B-MQ-Q6_K",
	"shortName": "MQ-Q8_0",
	"provider": "MagicQuant",
	"quantFamily": "Q8_0",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.003061,
	"ppl": 5.682076,
	"pplDeltaPercent": -0.1076614746316901,
	"sizeBytes": 26940999360,
	"sizeGB": 26.94099936,
	"sizeGiB": 25.09076088666916
	},
	"winner": {
	"key": "0:1:1:1:1:1:1:2:0:0",
	"fileName": "Qwen3.6-27B-MQ-Q6_K_1.gguf",
	"displayName": "MQ-Q6_K_1",
	"shortName": "MQ-Q6_K_1",
	"provider": "MagicQuant",
	"quantFamily": "Q6_K",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.002845,
	"ppl": 5.680246,
	"pplDeltaPercent": -0.13983333919341434,
	"sizeBytes": 27248916160,
	"sizeGB": 27.24891616,
	"sizeGiB": 25.37753075361252
	},
	"deltas": {
	"kld": 0.00021600000000000005,
	"sizeBytes": -307916800,
	"sizeGB": -0.3079168,
	"sizeGiB": -0.2867698669433594,
	"removedPplDeltaPercent": -0.1076614746316901,
	"winnerPplDeltaPercent": -0.13983333919341434,
	"pplDeltaPercentImprovement": 0.032171864561724245
	}
	},
	{
	"reasonCode": "STRICT_DOMINANCE",
	"reasonDescription": "The winner was no larger and had lower real KLD than the removed anchor.",
	"rawReason": "strict hybrid dominance: best accepted actual KLD at same-or-smaller real size",
	"removed": {
	"key": "0:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-Q8_0.gguf",
	"displayName": "Qwen3.6-27B-LM-Q8_0",
	"shortName": "LM-Q8_0",
	"provider": "llama.cpp",
	"quantFamily": "Q8_0",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.003768,
	"ppl": 5.677782,
	"pplDeltaPercent": -0.18315108470167157,
	"sizeBytes": 28595762880,
	"sizeGB": 28.59576288,
	"sizeGiB": 26.63187950849533
	},
	"winner": {
	"key": "0:1:2:1:1:1:1:2:0:0",
	"fileName": "Qwen3.6-27B-MQ-Q6_K.gguf",
	"displayName": "Qwen3.6-27B-MQ-Q6_K",
	"shortName": "MQ-Q8_0",
	"provider": "MagicQuant",
	"quantFamily": "Q8_0",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.003061,
	"ppl": 5.682076,
	"pplDeltaPercent": -0.1076614746316901,
	"sizeBytes": 26940999360,
	"sizeGB": 26.94099936,
	"sizeGiB": 25.09076088666916
	},
	"deltas": {
	"kld": 0.0007070000000000002,
	"sizeBytes": 1654763520,
	"sizeGB": 1.65476352,
	"sizeGiB": 1.5411186218261719,
	"removedPplDeltaPercent": -0.18315108470167157,
	"winnerPplDeltaPercent": -0.1076614746316901,
	"pplDeltaPercentImprovement": -0.07548961006998148
	}
	},
	{
	"reasonCode": "STRICT_DOMINANCE",
	"reasonDescription": "The winner was no larger and had lower real KLD than the removed anchor.",
	"rawReason": "strict hybrid dominance: best accepted actual KLD at same-or-smaller real size",
	"removed": {
	"key": "1:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-Q6_K.gguf",
	"displayName": "Qwen3.6-27B-LM-Q6_K",
	"shortName": "LM-Q6_K",
	"provider": "llama.cpp",
	"quantFamily": "Q6_K",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.007249,
	"ppl": 5.706878,
	"pplDeltaPercent": 0.32836398157588564,
	"sizeBytes": 22082528960,
	"sizeGB": 22.08252896,
	"sizeGiB": 20.565957725048065
	},
	"winner": {
	"key": "0:6:2:7:1:1:108:3:0:0",
	"fileName": "Qwen3.6-27B-MQ-Q5_K_S_1.gguf",
	"displayName": "MQ-Q5_K_S_1",
	"shortName": "MQ-Q5_K_S_1",
	"provider": "MagicQuant",
	"quantFamily": "Q5_K_S",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.006477,
	"ppl": 5.681127,
	"pplDeltaPercent": -0.12434513554375913,
	"sizeBytes": 21900871360,
	"sizeGB": 21.90087136,
	"sizeGiB": 20.396775901317596
	},
	"deltas": {
	"kld": 0.0007720000000000001,
	"sizeBytes": 181657600,
	"sizeGB": 0.1816576,
	"sizeGiB": 0.16918182373046875,
	"removedPplDeltaPercent": 0.32836398157588564,
	"winnerPplDeltaPercent": -0.12434513554375913,
	"pplDeltaPercentImprovement": 0.4527091171196448
	}
	},
	{
	"reasonCode": "STRICT_DOMINANCE",
	"reasonDescription": "The winner was no larger and had lower real KLD than the removed anchor.",
	"rawReason": "smart baseline-tuning strict dominance fallback: real benchmark validated lower KLD at same-or-smaller size",
	"removed": {
	"key": "105:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-UD-Q4_K_XL.gguf",
	"displayName": "Qwen3.6-27B-UD-Q4_K_XL",
	"shortName": "UD-Q4_K_XL",
	"provider": "Unsloth",
	"quantFamily": "UD-Q4_K_XL",
	"isHybrid": false,
	"isExternalPureBaseline": true,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.023521,
	"ppl": 5.760769,
	"pplDeltaPercent": 1.2757814422840206,
	"sizeBytes": 17612564160,
	"sizeGB": 17.61256416,
	"sizeGiB": 16.402978599071503
	},
	"winner": {
	"key": "105:6:106:7:14:106:106:106:0:0",
	"fileName": "Qwen3.6-27B-MQ-IQ4_NL_1.gguf",
	"displayName": "MQ-IQ4_NL_1",
	"shortName": "MQ-IQ4_NL_1",
	"provider": "MagicQuant",
	"quantFamily": "IQ4_NL",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.019687,
	"ppl": 5.727233,
	"pplDeltaPercent": 0.6862100488731036,
	"sizeBytes": 17588438720,
	"sizeGB": 17.58843872,
	"sizeGiB": 16.38051003217697
	},
	"deltas": {
	"kld": 0.0038340000000000006,
	"sizeBytes": 24125440,
	"sizeGB": 0.024125439999999998,
	"sizeGiB": 0.02246856689453125,
	"removedPplDeltaPercent": 1.2757814422840206,
	"winnerPplDeltaPercent": 0.6862100488731036,
	"pplDeltaPercentImprovement": 0.589571393410917
	}
	},
	{
	"reasonCode": "NEAR_BASELINE_PREMIUM",
	"reasonDescription": "The winner used only the configured near-baseline size premium and beat the real linear KLD trade line.",
	"rawReason": "smart baseline-tuning near-baseline fallback within \u002B1% size premium",
	"removed": {
	"key": "104:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-UD-Q3_K_XL.gguf",
	"displayName": "Qwen3.6-27B-UD-Q3_K_XL",
	"shortName": "UD-Q3_K_XL",
	"provider": "Unsloth",
	"quantFamily": "UD-Q3_K_XL",
	"isHybrid": false,
	"isExternalPureBaseline": true,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.044843,
	"ppl": 5.733645,
	"pplDeltaPercent": 0.7989346366161519,
	"sizeBytes": 14473430720,
	"sizeGB": 14.47343072,
	"sizeGiB": 13.479432761669159
	},
	"winner": {
	"key": "104:6:105:105:14:105:105:105:0:0",
	"fileName": "Qwen3.6-27B-MQ-IQ3_M_1.gguf",
	"displayName": "MQ-IQ3_M_1",
	"shortName": "MQ-IQ3_M_1",
	"provider": "MagicQuant",
	"quantFamily": "IQ3_M",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.043802,
	"ppl": 5.751239,
	"pplDeltaPercent": 1.1082416230090335,
	"sizeBytes": 14492272320,
	"sizeGB": 14.49227232,
	"sizeGiB": 13.496980369091034
	},
	"deltas": {
	"kld": 0.0010410000000000003,
	"sizeBytes": -18841600,
	"sizeGB": -0.0188416,
	"sizeGiB": -0.017547607421875,
	"removedPplDeltaPercent": 0.7989346366161519,
	"winnerPplDeltaPercent": 1.1082416230090335,
	"pplDeltaPercentImprovement": -0.30930698639288157
	}
	},
	{
	"reasonCode": "SPACING_COLLAPSE",
	"reasonDescription": "Two candidates were too close in practical output space; the stronger one was kept.",
	"rawReason": "meaningful spacing collapse; size gap below 564,472,013 bytes",
	"removed": {
	"key": "104:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-UD-Q3_K_XL.gguf",
	"displayName": "Qwen3.6-27B-UD-Q3_K_XL",
	"shortName": "UD-Q3_K_XL",
	"provider": "Unsloth",
	"quantFamily": "UD-Q3_K_XL",
	"isHybrid": false,
	"isExternalPureBaseline": true,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.044843,
	"ppl": 5.733645,
	"pplDeltaPercent": 0.7989346366161519,
	"sizeBytes": 14473430720,
	"sizeGB": 14.47343072,
	"sizeGiB": 13.479432761669159
	},
	"winner": {
	"key": "104:6:105:105:14:105:105:105:0:0",
	"fileName": "Qwen3.6-27B-MQ-IQ3_M_1.gguf",
	"displayName": "MQ-IQ3_M_1",
	"shortName": "MQ-IQ3_M_1",
	"provider": "MagicQuant",
	"quantFamily": "IQ3_M",
	"isHybrid": true,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": true,
	"kld": 0.043802,
	"ppl": 5.751239,
	"pplDeltaPercent": 1.1082416230090335,
	"sizeBytes": 14492272320,
	"sizeGB": 14.49227232,
	"sizeGiB": 13.496980369091034
	},
	"deltas": {
	"kld": 0.0010410000000000003,
	"sizeBytes": -18841600,
	"sizeGB": -0.0188416,
	"sizeGiB": -0.017547607421875,
	"removedPplDeltaPercent": 0.7989346366161519,
	"winnerPplDeltaPercent": 1.1082416230090335,
	"pplDeltaPercentImprovement": -0.30930698639288157
	}
	},
	{
	"reasonCode": "SPACING_COLLAPSE",
	"reasonDescription": "Two candidates were too close in practical output space; the stronger one was kept.",
	"rawReason": "meaningful spacing collapse; size gap below 564,472,013 bytes",
	"removed": {
	"key": "8:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-IQ3_XS.gguf",
	"displayName": "Qwen3.6-27B-LM-IQ3_XS",
	"shortName": "LM-IQ3_XS",
	"provider": "llama.cpp",
	"quantFamily": "IQ3_XS",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.071307,
	"ppl": 5.76996,
	"pplDeltaPercent": 1.4373615555008623,
	"sizeBytes": 11967129280,
	"sizeGB": 11.96712928,
	"sizeGiB": 11.145257651805878
	},
	"winner": {
	"key": "7:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-IQ3_S.gguf",
	"displayName": "LM-IQ3_S",
	"shortName": "LM-IQ3_S",
	"provider": "llama.cpp",
	"quantFamily": "IQ3_S",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.064393,
	"ppl": 5.795439,
	"pplDeltaPercent": 1.8852888435708988,
	"sizeBytes": 12419327680,
	"sizeGB": 12.41932768,
	"sizeGiB": 11.566400229930878
	},
	"deltas": {
	"kld": 0.00691399999999999,
	"sizeBytes": -452198400,
	"sizeGB": -0.45219840000000006,
	"sizeGiB": -0.421142578125,
	"removedPplDeltaPercent": 1.4373615555008623,
	"winnerPplDeltaPercent": 1.8852888435708988,
	"pplDeltaPercentImprovement": -0.44792728807003646
	}
	},
	{
	"reasonCode": "SPACING_COLLAPSE",
	"reasonDescription": "Two candidates were too close in practical output space; the stronger one was kept.",
	"rawReason": "meaningful spacing collapse; size gap below 564,472,013 bytes",
	"removed": {
	"key": "100:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-UD-IQ2_M.gguf",
	"displayName": "Qwen3.6-27B-UD-IQ2_M",
	"shortName": "UD-IQ2_M",
	"provider": "Unsloth",
	"quantFamily": "UD-IQ2_M",
	"isHybrid": false,
	"isExternalPureBaseline": true,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.133316,
	"ppl": 6.250984,
	"pplDeltaPercent": 9.893885587707882,
	"sizeBytes": 10846136000,
	"sizeGB": 10.846136,
	"sizeGiB": 10.101251304149628
	},
	"winner": {
	"key": "9:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-IQ3_XXS.gguf",
	"displayName": "LM-IQ3_XXS",
	"shortName": "LM-IQ3_XXS",
	"provider": "llama.cpp",
	"quantFamily": "IQ3_XXS",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.093578,
	"ppl": 5.939434,
	"pplDeltaPercent": 4.416757497978273,
	"sizeBytes": 11186370240,
	"sizeGB": 11.18637024,
	"sizeGiB": 10.418119132518768
	},
	"deltas": {
	"kld": 0.039737999999999996,
	"sizeBytes": -340234240,
	"sizeGB": -0.34023424,
	"sizeGiB": -0.3168678283691406,
	"removedPplDeltaPercent": 9.893885587707882,
	"winnerPplDeltaPercent": 4.416757497978273,
	"pplDeltaPercentImprovement": 5.477128089729608
	}
	},
	{
	"reasonCode": "SPACING_COLLAPSE",
	"reasonDescription": "Two candidates were too close in practical output space; the stronger one was kept.",
	"rawReason": "meaningful spacing collapse; size gap below 564,472,013 bytes",
	"removed": {
	"key": "11:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-IQ2_XS.gguf",
	"displayName": "Qwen3.6-27B-LM-IQ2_XS",
	"shortName": "LM-IQ2_XS",
	"provider": "llama.cpp",
	"quantFamily": "IQ2_XS",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.226206,
	"ppl": 6.678807,
	"pplDeltaPercent": 17.41512253436939,
	"sizeBytes": 9090590400,
	"sizeGB": 9.090590400000002,
	"sizeGiB": 8.466272056102753
	},
	"winner": {
	"key": "10:0:0:0:0:0:0:0:0:0",
	"fileName": "Qwen3.6-27B-LM-IQ2_S.gguf",
	"displayName": "LM-IQ2_S",
	"shortName": "LM-IQ2_S",
	"provider": "llama.cpp",
	"quantFamily": "IQ2_S",
	"isHybrid": false,
	"isExternalPureBaseline": false,
	"isExternalRebuiltBaseline": false,
	"isMaterializedTensorMapped": false,
	"kld": 0.210251,
	"ppl": 6.500917,
	"pplDeltaPercent": 14.287771175415775,
	"sizeBytes": 9362912960,
	"sizeGB": 9.362912960000001,
	"sizeGiB": 8.71989220380783
	},
	"deltas": {
	"kld": 0.015954999999999997,
	"sizeBytes": -272322560,
	"sizeGB": -0.27232256,
	"sizeGiB": -0.2536201477050781,
	"removedPplDeltaPercent": 17.41512253436939,
	"winnerPplDeltaPercent": 14.287771175415775,
	"pplDeltaPercentImprovement": 3.127351358953616
	}
	}
	]