Instructions to use selfnify/linxrn-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 selfnify/linxrn-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 selfnify/linxrn-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf selfnify/linxrn-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 selfnify/linxrn-gguf:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf selfnify/linxrn-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 selfnify/linxrn-gguf:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf selfnify/linxrn-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 selfnify/linxrn-gguf:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf selfnify/linxrn-gguf:Q4_K_M

Use Docker

docker model run hf.co/selfnify/linxrn-gguf:Q4_K_M

LM Studio
Jan
Ollama
How to use selfnify/linxrn-gguf with Ollama:
```
ollama run hf.co/selfnify/linxrn-gguf:Q4_K_M
```
Unsloth Desktop

How to use selfnify/linxrn-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf selfnify/linxrn-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": "selfnify/linxrn-gguf:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use selfnify/linxrn-gguf with Docker Model Runner:
```
docker model run hf.co/selfnify/linxrn-gguf:Q4_K_M
```

Lemonade

How to use selfnify/linxrn-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull selfnify/linxrn-gguf:Q4_K_M

Run and chat with the model

lemonade run user.linxrn-gguf-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use selfnify/linxrn-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 selfnify/linxrn-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 selfnify/linxrn-gguf:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use selfnify/linxrn-gguf with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf selfnify/linxrn-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 "selfnify/linxrn-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"

Linxrn (GGUF)

Linxrn is a small Linux-only debugging and help assistant. It is fine-tuned from Qwen3-0.6B, it thinks inside <think> tags before answering, and it is built to stay on Linux topics.

This is an early prototype. It was trained on 27 hand-written examples as a proof that the pipeline works end to end. It will ramble, make guesses, and sometimes give weak advice. It is not ready for production use or for anything where a wrong command could cause damage.

Files

File	Description
`linxrn-Q4_K_M.gguf`	Quantized model, about 397 MB, runs on CPU with under 1 GB of RAM
`Modelfile`	Ollama config that points at the GGUF and sets the system prompt

Quick start with Ollama

Install Ollama from https://ollama.com
Download linxrn-Q4_K_M.gguf and Modelfile into the same folder.
Create and run the model:

ollama create linxrn -f Modelfile
ollama run linxrn

Try a question like apt says E: Unable to locate package htop.

Intended use

Explaining common Linux errors and what they usually mean
Suggesting diagnostic commands, such as checking logs, service status, or package lists
Giving step-by-step troubleshooting for apt, systemd, permissions, networking, disks, and similar topics

Limitations

Trained on only 27 examples, so coverage is very narrow
Can produce long, repetitive reasoning and sometimes irrelevant steps
Can give incorrect or outdated commands. Check every command before running it, especially anything involving rm, dd, mkfs, fdisk, or boot settings
Does not know your system. Always share your distro, kernel version, and the full error text when asking for help

Training details

Base model: Qwen/Qwen3-0.6B
Method: LoRA fine-tuning (r=16, alpha=32) with TRL SFTTrainer, 2 epochs, learning rate 2e-4
Data: 27 hand-written rows of Linux troubleshooting questions with a thinking trace and an answer
Export: merged to fp16, converted to GGUF with llama.cpp, quantized to Q4_K_M
Hardware: single Tesla T4 GPU on Google Colab

Roadmap

Build a larger, cleaned Linux Q&A dataset with reasoning traces
Retrain and compare against this prototype
Add evaluation on a held-out set of Linux problems

License

Apache 2.0, matching the base model.

Downloads last month: 53

GGUF

Model size

0.6B params

Architecture

qwen3

Hardware compatibility

4-bit

Inference Providers NEW

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