Instructions to use CompiwerAI/Mtrini-SVL-1.0-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries
PEFT
How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with PEFT:
```
Task type is invalid.
```
Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use CompiwerAI/Mtrini-SVL-1.0-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 CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

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 CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

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 CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

Use Docker

docker model run hf.co/CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

LM Studio
Jan

vLLM

How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "CompiwerAI/Mtrini-SVL-1.0-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": "CompiwerAI/Mtrini-SVL-1.0-GGUF",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

Ollama
How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with Ollama:
```
ollama run hf.co/CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
```
Unsloth Desktop

How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

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": "CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with Docker Model Runner:
```
docker model run hf.co/CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0
```

Lemonade

How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

Run and chat with the model

lemonade run user.Mtrini-SVL-1.0-GGUF-Q8_0

List all available models

lemonade list

Hermes Agent

How to use CompiwerAI/Mtrini-SVL-1.0-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 CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

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 CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use CompiwerAI/Mtrini-SVL-1.0-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0

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 "CompiwerAI/Mtrini-SVL-1.0-GGUF:Q8_0" \
  --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"

🧠 Mtrini-SVL-1.0 (GGUF)

Mtrini (متربّي) comes from Darija — Moroccan Arabic — and means "trained" or "well-fit." 💪 This is the easy-to-run version of that trained model: one file, no Python required. 🚀

🤔 What is this, in plain English?

This is a single downloadable file containing a complete AI model that can:

💻 Code — write functions, fix bugs, explain code
🧮 Do math — solve problems and show its reasoning
👀 See images — describe pictures, read screenshots, answer visual questions

It's the merged + compressed version of our fine-tuned model, so you can run it locally with a tool called llama.cpp — no GPU required if you have enough RAM (though a GPU makes it faster). ⚡

📖 What's in the name?

Mtrini (متربّي) in Darija means trained or well-fit. The entire story of this model is training: a strong foundation model, further trained on coding and math data, then merged and quantized so anyone can run it.

🗣️ Say it like: "em-tree-nee."

✨ Why you'll like this release

✅ Feature	📝 Detail
🗜️ Single file	Everything bundled — no extra download for vision
🖥️ Local	Runs on your own machine, private and free
🧩 Q8_0 quality	Near-original quality at half the size
💬 Easy chat	Works with the standard Qwen3-VL chat format

🚀 Quick start (2 steps)

Step 1 — Get llama.cpp

Download the llama.cpp tools from github.com/ggml-org/llama.cpp/releases and build them (or download a prebuilt binary if your OS has one).

Step 2 — Run the model

💬 Chat (text only):

llama-cli -m Mtrini-SVL-1.0-Q8_0.gguf -p "Write a Python function that checks if a number is prime."

🖼️ Chat with an image:

llama-mtmd-cli \
  -m Mtrini-SVL-1.0-Q8_0.gguf \
  -p "Describe this image." \
  --image path/to/your/image.jpg

🎉 That's it. The vision parts are already inside this one file — no extra --mmproj needed.

📦 File details

Property	Value
🏗️ Base model	`Qwen/Qwen3-VL-8B-Instruct` (8.81B params)
🧩 Source adapter	CompiwerAI/Mtrini-SVL-1.0
📚 Training data	Coding (17k) + Math (220k)
🗜️ Format	GGUF `qwen3vl`, Q8_0 quantization
💾 Size	~8.7 GB
🔢 Tensors	399

🧪 What is quantization? (quick version)

The original model is ~17.6 GB in full precision. Quantization compresses the numbers so the model still works great but takes half the space. Q8_0 is one of the higher-quality compression levels — a great balance of size and smarts. 🧠⚖️

⚠️ Keep in mind

This is a fine-tune trained on a focused coding/math mix.
It inherits the strengths and limitations of the base model.
🧪 Test on your own use case before production.

📜 License

Apache 2.0 — same as the base model. ✅

Downloads last month: -

GGUF

Model size

8B params

Architecture

qwen3vl

Hardware compatibility

8-bit

Model tree for CompiwerAI/Mtrini-SVL-1.0-GGUF

Base model

Qwen/Qwen3-VL-8B-Instruct

Quantized

(107)

this model