metadata

title: LaMini-LM-API
emoji: 🚀
colorFrom: blue
colorTo: green
sdk: docker
app_file: main.py
pinned: false

LaMini-LM API

This is a FastAPI-based API for text generation using the MBZUAI/LaMini-GPT-774M model from the LaMini-LM series. It features a web interface for easy interaction and a REST API for programmatic access.

Features

Web Interface: User-friendly UI accessible at the root URL
REST API: Programmatic access via /api/generate endpoint
Model: MBZUAI/LaMini-GPT-774M for high-quality text generation
Configurable Parameters: Control max length, temperature, and top-p sampling

Installation

Clone the repository:

git clone <your-repo-url>
cd lamini-lm-api

Set up a virtual environment and install dependencies:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate
pip install -r requirements.txt

Run the API locally:

uvicorn main:app --host 0.0.0.0 --port 7860

Usage

Web Interface

Simply visit the root URL to access the interactive web interface where you can:

Enter text instructions
Adjust generation parameters (max length, temperature, top-p)
Generate text with a single click

API Endpoints

Generate Text: POST /api/generate

Request Body (JSON):

{
  "instruction": "Tell me about camels",
  "max_length": 100,
  "temperature": 1.0,
  "top_p": 0.9
}

Response:

{
  "generated_text": "Camels are remarkable desert animals known for their ability to survive in harsh conditions. These fascinating creatures have evolved unique adaptations including their iconic humps, which store fat rather than water. Camels can go for extended periods without drinking water, making them invaluable companions for desert travelers throughout history."
}

Health Check: GET /api/health

{
  "status": "healthy"
}

API Info: GET /api

{
  "message": "Welcome to the LaMini-LM API. Use POST /generate to generate text."
}

Parameter Constraints

instruction: Required, non-empty string
max_length: 10-500 tokens (default: 100)
temperature: 0.1-2.0 (default: 1.0)
top_p: 0.1-1.0 (default: 0.9)

Deployment

This API is designed to be deployed on Hugging Face Spaces using Docker. The Dockerfile handles all dependencies and model loading automatically.

License

The LaMini-GPT-774M model is licensed under CC BY-NC 4.0 (non-commercial use only). Ensure compliance when using this API.

Contributing

This project is a community contribution. If you're from MBZUAI, feel free to adopt this Hugging Face Space! Contact for details.