Instructions to use SupraLabs/Supra-Title-350M-exp-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SupraLabs/Supra-Title-350M-exp-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="SupraLabs/Supra-Title-350M-exp-GGUF",
	filename="LiquidAI_LFM2.5-350M-Base_1781204855.BF16.gguf",
)

llm.create_chat_completion(
	messages = "\"Today is a sunny day and I will get some ice cream.\""
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use SupraLabs/Supra-Title-350M-exp-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama-cli -hf SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M

Use Docker

docker model run hf.co/SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M

LM Studio
Jan
Ollama
How to use SupraLabs/Supra-Title-350M-exp-GGUF with Ollama:
```
ollama run hf.co/SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
```

Unsloth Studio

How to use SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for SupraLabs/Supra-Title-350M-exp-GGUF to start chatting

How to use SupraLabs/Supra-Title-350M-exp-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf SupraLabs/Supra-Title-350M-exp-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": "SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-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 SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use SupraLabs/Supra-Title-350M-exp-GGUF with Docker Model Runner:
```
docker model run hf.co/SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M
```

Lemonade

How to use SupraLabs/Supra-Title-350M-exp-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull SupraLabs/Supra-Title-350M-exp-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Supra-Title-350M-exp-GGUF-Q4_K_M

List all available models

lemonade list

Supra For Chat Titles (experimental)

What is Supra Title?

Supra Title is a compact title generation model currently in alpha testing.

Many AI platforms automatically generate titles for conversations using larger language models. While this works well, it often requires running models that were designed for far more than just title generation.

Supra Title takes a different approach. Instead of relying on a general-purpose model, it is being developed specifically for generating short, accurate, and context-aware conversation titles.

The goal of this project is to provide an open-source title generation model that is lightweight, fast, and easy to integrate into applications, while maintaining strong title quality across a wide range of conversations.

Why Supra Title?

Supra Title is a compact, purpose-built model designed specifically for chat title generation.

Many AI platforms rely on large, general-purpose language models to generate conversation titles solely through prompting. While this approach can produce acceptable results, it often introduces unnecessary computational overhead, increased latency, and higher operating costs.

Supra Title takes a different approach. By focusing exclusively on title generation, it delivers fast, consistent, and efficient results while requiring only a fraction of the resources used by larger models. This makes it particularly well-suited for real-time applications, local AI deployments, and large-scale chat platforms where speed and efficiency matter.

Roadmap for Supra Title

The current experimental release of Supra Title is available in GGUF format, allowing the community to evaluate its performance, efficiency, and real-world usability.

Our team at SupraLabs is actively developing the next generation of the model. Current efforts include expanding our supervised fine-tuning (SFT) dataset to approximately 300,000 high-quality samples, improving title accuracy, and exploring preference optimization techniques to further enhance output quality.

A preview version of the full model is already under development and will be made available for public testing as it reaches our quality standards. While we are committed to open releases, publishing the final weights may take additional time as we continue to optimize the model for both speed and accuracy.

Our goal is to build a lightweight, purpose-built title generation model that delivers reliable, high-quality results without the overhead of large general-purpose language models.

How to use Supra Title

Here at SupraLabs, speed, accuracy, and simplicity for the developer are crucial! We made sure our training approach performs with less effort at runtime!

As of now, to generate text, the user can send text to the assistant (without a system prompt! Not having a system prompt is crucial, as it is not trained with one!) Here is a visualization of optimal inference:

User prompts an AI →
Backend sends the message to Supra title simultaneously →
*Chat title generated*

Examples of samples

User

bruh my wifi keeps disconnecting every 10 minutes 😭

Assistant

WiFi Issues

User

what's the easiest way to make fluffy pancakes?

Assistant

Fluffy Pancakes

User

i accidentally sent a message to the wrong person and now i'm dying inside

Assistant

Wrong Message

User

can someone explain taxes to me like i'm five because none of this makes sense

Assistant

Understanding Taxes

User

I've been trying to learn guitar for a few months now, but I still struggle with switching between chords smoothly. My fingers feel awkward and I keep losing rhythm whenever I change positions. Is this normal, and are there any exercises that could help me improve faster?

Assistant

Guitar Practice Tips

User

My friend and I are arguing about whether humanity would have progressed faster if the internet had been invented fifty years earlier. We started talking about education, scientific collaboration, economic growth, and even how culture might have evolved differently. Now I'm curious what the strongest arguments are on both sides.

Assistant

Earlier Internet Debate

Downloads last month: 694

GGUF

Model size

0.4B params

Architecture

lfm2

Hardware compatibility

2-bit

3-bit

4-bit

5-bit

6-bit

8-bit

16-bit

Model tree for SupraLabs/Supra-Title-350M-exp-GGUF

Base model

LiquidAI/LFM2.5-350M-Base

Finetuned

LiquidAI/LFM2.5-350M

Quantized

(34)

this model

Space using SupraLabs/Supra-Title-350M-exp-GGUF 1

Collection including SupraLabs/Supra-Title-350M-exp-GGUF

Supra-Title

Collection

All Supra-Title models • 1 item • Updated 1 day ago • 1