Instructions to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf",
	filename="Prosodia_T0.6B-Instruct-Q4_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M

Use Docker

docker model run hf.co/Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M

LM Studio
Jan
Ollama
How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with Ollama:
```
ollama run hf.co/Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
```

Unsloth Studio

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf to start chatting

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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": "Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-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 Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M

Run Hermes

hermes

Docker Model Runner
How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with Docker Model Runner:
```
docker model run hf.co/Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M
```

Lemonade

How to use Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf:Q4_K_M

Run and chat with the model

lemonade run user.Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf-Q4_K_M

List all available models

lemonade list

Introduction

Prosodia is an organization dedicated to developing and transparently distributing open-source Portuguese language models. Prosodia T0.6B is a 0.6 billion parameter Small Language Model (SLM) trained on publicly available data and released for community use. Built on the Qwen3 architecture, it represents our commitment to accessible AI development.

This is the Instruct (instruction-tuned) variant of the model. This repository provides the model in Q4_K_M GGUF format under the filename Prosodia_T0.6B-Instruct-Q4_K_M.gguf.

Training

This model was developed through a focused one-month effort with substantially limited computational resources compared to industry leaders. Its primary purpose is to demonstrate that through intelligent design, transparency, and community collaboration, it is possible to create high-quality Brazilian and Portuguese language models without massive infrastructure.

The training regimen utilized approximately 20 billion tokens for the base model and just under 1 billion tokens for instruction tuning and subsequent refinements. While these volumes are far from ideal, they serve as a rapid proof-of-concept that establishes a foundation for future, more comprehensive development.

Inference

The model is fully compatible with standard LLM deployment platforms and can be used and distributed across frameworks such as HuggingFace, vLLM, GGUF, and similar tools.

You can download the GGUF artifact programmatically using:

from huggingface_hub import hf_hub_download

path = hf_hub_download(
    repo_id="Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf",
    filename="Prosodia_T0.6B-Instruct-Q4_K_M.gguf",
)
print(path)

Downloads last month: 1

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

Model tree for Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf

Base model

Qwen/Qwen3-0.6B-Base

Finetuned

Qwen/Qwen3-0.6B

Quantized

(309)

this model

Prosodia
/

Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf

Model tree for Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf

Datasets used to train Prosodia/Prosodia_T0.6B-Instruct-Q4_K_M-quantization-gguf