Instructions to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b",
	filename="Hugston-code-rl-Qwen3-4B-Instruct-2507-SFT-30b-F32-f32.gguf",
)

llm.create_chat_completion(
	messages = "\"I like you. I love you\""
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b: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 Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b: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 Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M

Use Docker

docker model run hf.co/Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M

LM Studio
Jan
Ollama
How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with Ollama:
```
ollama run hf.co/Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
```

Unsloth Studio new

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b 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 Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b 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 Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b to start chatting

Pi new

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b: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": "Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b: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 Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with Docker Model Runner:
```
docker model run hf.co/Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M
```

Lemonade

How to use Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b:Q4_K_M

Run and chat with the model

lemonade run user.Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b-Q4_K_M

List all available models

lemonade list

Trilogix1/Hugston_code-rl-Qwen3-4B-Instruct-2507-SFT-30b pipeline_tag: text-generation tags:

Qwen3 Instruct

Coder 4B

Hugston

Original weights at: https://huggingface.co/code-rl/Qwen3-4B-Instruct-2507-SFT-30b

This model is converted and quantized version by Hugston Team created with Quanta (see Github to know more about it). This is a real, proof-of-concept and implementation on how to convert and quantize a .safetensor llm model in GGUF.

Quantization was performed using an automatic and faster method, which leads to less time and faster results.

This model was made possible by: https://Hugston.com

You can use the model with HugstonOne Enterprise Edition

Tested in general and coding tasks. Loaded with 262000 tokens ctx and feed with 150kb code as input, and gave back 230kb code output or ~ 60000 tokens at once. The code had 5 errors and certainly is not a 0-shot in long coding. It is working with 2-3 tries, which makes it very impressive for it´s size and considering being an instruct model.

Watch HugstonOne coding and preview in action:

https://vimeo.com/1121493834?share=copy&fl=sv&fe=ci

-Download App HugstonOne at Hugston.com or at https://github.com/Mainframework

-Download model from https://hugston.com/explore?folder=llm_models or Huggingface

-If you already have the Llm Model downloaded chose it by clicking pick model in HugstonOne -Then click Load model in Cli or Server

-For multimodal use you need a VL/multimodal LLM model with the Mmproj file in the same folder. -Select model and select mmproj.

-Note: if the mmproj is inside the same folder with other models non multimodal, the non model will not load unless the mmproj is moved from folder.

Downloads last month: 57

GGUF

Model size

4B params

Architecture

qwen3

Hardware compatibility

4-bit

5-bit

6-bit

8-bit

16-bit

32-bit