Instructions to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M

Use Docker

docker model run hf.co/LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF with Ollama:
```
ollama run hf.co/LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
```
Unsloth Desktop

How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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": "LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF with Docker Model Runner:
```
docker model run hf.co/LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M
```

Lemonade

How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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 "LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-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"

Qwen 3.6 27B RYS 2XL MTP Q4_K_M (GGUF)

This is a modified version of the Qwen 3.6 27B model, utilizing the RYS (Repeat Your Self) technique and retaining MTP heads.

What is RYS?

The RYS (Repeat Your Self) technique, discovered by David Ng, enhances the reasoning capabilities of Large Language Models by duplicating specific "reasoning" layers in the middle of the transformer stack. This increases the depth of the model's computation for semantic and logic-heavy tasks without requiring additional training.

Model Details

Architecture: Qwen 3.6 27B
RYS Configuration: Physical 2X duplication of layers (26, 34).
Variant: 2XL (16 additional layers, bringing the total depth to 80 layers).
Format: GGUF (Quantized to Q4_K_M).
Tokenizer: Full Qwen 3.5/3.6 201-language support.

Performance

By repeating layers 26 through 34 twice, the model spends more time processing the internal semantic representation of a prompt. This is particularly effective for:

Mathematical reasoning
Complex logic puzzles
Large-scale coding tasks

Usage

This GGUF model is compatible with any tool that uses llama.cpp, such as:

Prompt Format

This model uses the standard Qwen Chat template:

<|im_start|>system
You are a helpful assistant.<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant


CREDITS
Base Model: The Qwen Team at Alibaba Cloud.
RYS Technique: David Ng (dnhkng).
Quantization: Processed on a Mac Studio

Downloads last month: 10

GGUF

Model size

33B params

Architecture

qwen35

Hardware compatibility

4-bit

Inference Providers NEW

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

Model tree for LogicBombaklot/Qwen3.6-27B-RYS-2XL-MTP-Q4_K_M-GGUF

Base model

Qwen/Qwen3.6-27B

Quantized

(714)

this model