Instructions to use Briko/Marco-Nano-Instruct-APEX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Briko/Marco-Nano-Instruct-APEX with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Briko/Marco-Nano-Instruct-APEX",
	filename="Marco-Nano-Instruct-APEX-I-Balanced.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 Briko/Marco-Nano-Instruct-APEX with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Briko/Marco-Nano-Instruct-APEX
# Run inference directly in the terminal:
llama-cli -hf Briko/Marco-Nano-Instruct-APEX

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Briko/Marco-Nano-Instruct-APEX
# Run inference directly in the terminal:
llama-cli -hf Briko/Marco-Nano-Instruct-APEX

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 Briko/Marco-Nano-Instruct-APEX
# Run inference directly in the terminal:
./llama-cli -hf Briko/Marco-Nano-Instruct-APEX

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 Briko/Marco-Nano-Instruct-APEX
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Briko/Marco-Nano-Instruct-APEX

Use Docker

docker model run hf.co/Briko/Marco-Nano-Instruct-APEX

LM Studio
Jan

vLLM

How to use Briko/Marco-Nano-Instruct-APEX with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Briko/Marco-Nano-Instruct-APEX"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Briko/Marco-Nano-Instruct-APEX",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Briko/Marco-Nano-Instruct-APEX

Ollama
How to use Briko/Marco-Nano-Instruct-APEX with Ollama:
```
ollama run hf.co/Briko/Marco-Nano-Instruct-APEX
```

Unsloth Studio

How to use Briko/Marco-Nano-Instruct-APEX 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 Briko/Marco-Nano-Instruct-APEX 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 Briko/Marco-Nano-Instruct-APEX to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Briko/Marco-Nano-Instruct-APEX to start chatting

How to use Briko/Marco-Nano-Instruct-APEX with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Briko/Marco-Nano-Instruct-APEX

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": "Briko/Marco-Nano-Instruct-APEX"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Briko/Marco-Nano-Instruct-APEX with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Briko/Marco-Nano-Instruct-APEX

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 Briko/Marco-Nano-Instruct-APEX

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use Briko/Marco-Nano-Instruct-APEX with Docker Model Runner:
```
docker model run hf.co/Briko/Marco-Nano-Instruct-APEX
```

Lemonade

How to use Briko/Marco-Nano-Instruct-APEX with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Briko/Marco-Nano-Instruct-APEX

Run and chat with the model

lemonade run user.Marco-Nano-Instruct-APEX-{{QUANT_TAG}}

List all available models

lemonade list

Marco-Nano-Instruct-APEX

File size: 1,453 Bytes

1f43176
 
ecee4da
 
 
 
 
 
 
 
 
 
 
 
 
1f43176
ecee4da

---
license: apache-2.0
language:
- en
- zh
tags:
- GGUF
- llama.cpp
- apex
- quantized
- Mixture of Experts
base_model:
- AIDC-AI/Marco-Nano-Instruct
- mradermacher/Marco-Nano-Instruct-GGUF
pipeline_tag: text-generation
---

# Marco-Nano-Instruct-APEX APEX Quantized (GGUF)

This repository contains APEX-quantized GGUF files for [AIDC-AI's Marco-Nano-Instruct](https://huggingface.co/AIDC-AI/Marco-Nano-Instruct).

The quantization was performed using the [mudler/apex-quant](https://github.com/mudler/apex-quant) project, focusing on maximizing quality-to-size ratio using importance matrix (imatrix) guided quantization.

## 📥 Source & Credits

- **Base Model**: [AIDC-AI's Marco-Nano-Instruct](https://huggingface.co/AIDC-AI/Marco-Nano-Instruct).
- **F16 GGUF & Imatrix**: The F16 source model and the importance matrix file used for quantization were sourced from [mradermacher's GGUF repository](https://huggingface.co/mradermacher/Marco-Nano-Instruct-i1-GGUF). 

> **Special thanks to [@mradermacher](https://huggingface.co/mradermacher) for providing the high-quality imatrix file!**

## ⚠️ For technical validation only

- Severe accuracy loss due to quantization; outputs may contain hallucinations, gibberish, or fail basic tasks.
- Suitable **only** for researching quantization noise, debugging conversion scripts, or comparing compression artifacts.
- No post-training calibration, fine-tuning, or recovery techniques were applied.