Instructions to use TirGun/Qwen3.5-4B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use TirGun/Qwen3.5-4B-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="TirGun/Qwen3.5-4B-GGUF",
	filename="qwen3.5-4b-Q4_K_M.gguf",
)

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

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use TirGun/Qwen3.5-4B-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf TirGun/Qwen3.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf TirGun/Qwen3.5-4B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf TirGun/Qwen3.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf TirGun/Qwen3.5-4B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/TirGun/Qwen3.5-4B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use TirGun/Qwen3.5-4B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "TirGun/Qwen3.5-4B-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": "TirGun/Qwen3.5-4B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/TirGun/Qwen3.5-4B-GGUF:Q4_K_M

Ollama
How to use TirGun/Qwen3.5-4B-GGUF with Ollama:
```
ollama run hf.co/TirGun/Qwen3.5-4B-GGUF:Q4_K_M
```

Unsloth Studio

How to use TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for TirGun/Qwen3.5-4B-GGUF to start chatting

How to use TirGun/Qwen3.5-4B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf TirGun/Qwen3.5-4B-GGUF:Q4_K_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": "TirGun/Qwen3.5-4B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-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 TirGun/Qwen3.5-4B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use TirGun/Qwen3.5-4B-GGUF with Docker Model Runner:
```
docker model run hf.co/TirGun/Qwen3.5-4B-GGUF:Q4_K_M
```

Lemonade

How to use TirGun/Qwen3.5-4B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull TirGun/Qwen3.5-4B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.5-4B-GGUF-Q4_K_M

List all available models

lemonade list

Qwen 3.5 4B GGUF

Description

This repository contains GGUF weights for the Qwen/Qwen3.5-4B model. The files were converted and quantized using llama.cpp.

Provided Files

Q6_K: High quality, recommended for best performance if you have enough VRAM.
Q5_K_M: Balanced quality and speed.
Q4_K_M: Optimal for most users, fast and lightweight.

Usage

You can run these models using llama.cpp or any GGUF-compatible software like LM Studio, Ollama, or KoboldCPP.

Example command for llama-cli:

./llama-cli -m qwen3.5-4b-Q4_K_M.gguf -ngl 32

Example PowerShell command for llama-cli:

.\llama-cli.exe -m qwen3.5-4b-Q4_K_M.gguf -ngl 32 -fa 0 --no-mmap --reasoning off

Parameter Quick Reference (CLI Flags)

When running this model via llama-cli, you can use the following flags to optimize performance:

Flash Attention (`-fa`)

An optimization technique for the attention mechanism.

-fa 1: Enable. Significantly speeds up processing for long contexts (requires model and hardware support).
-fa 0: Disable. More stable, but slower when dealing with large contexts.

Memory Mapping (`--no-mmap`)

Controls how the model file is loaded into the system.

Without this flag: The model uses mmap (memory-mapped files) by default. It provides faster loading but may occasionally conflict with specific systems or GPU drivers.
With --no-mmap: The model is fully read into system RAM. This is more reliable for troubleshooting but results in slower startup times and higher RAM consumption.

Reasoning Process (`--reasoning`)

Controls the output of the model's internal "thinking" (for models trained with reasoning capabilities like Qwen 3.5).

--reasoning on: Allows the model to display its internal thought process (usually enclosed within <thought> tags).
--reasoning off: Disables the thought process output, forcing the model to provide a direct answer immediately.

Downloads last month: 711

GGUF

Model size

4B params

Architecture

qwen35

Hardware compatibility

4-bit

5-bit

6-bit

Model tree for TirGun/Qwen3.5-4B-GGUF

Base model

Qwen/Qwen3.5-4B-Base

Finetuned

Qwen/Qwen3.5-4B

Quantized

(241)

this model