Welcome to the Horizon project! This application features a chatbot and a scientific text summarization tool, built using Python Django and leveraging the Gemma model by Google. The project is containerized using Docker for ease of deployment.
Horizon is a web application that provides a chatbot interface and a large text summarization feature for scientific texts. The application utilizes the Gemma 2b-it model, an open-source model developed by Google, for advanced language processing tasks. The backend is built using Django with ASGI, and the application is designed to run on CUDA-compatible GPUs for efficient inference.
- Python Django: Web framework for building the application.
- ASGI Daphne: Asynchronous Server Gateway Interface for handling asynchronous requests.
- Gemma 2b-it Model: A 2 billion parameter model by Google for natural language processing.
- PyTorch: Deep learning framework used with CUDA for model inference.
- Docker: Containerization platform for easy deployment.
- SQLite: Default database for testing purposes.
- CUDA-compatible GPU with NVIDIA drivers (version > 12).
- Docker installed on your machine.
-
Pull the Docker Image:
docker pull saikiranappidi/inference:latest
-
Run the Docker Container:
docker run --gpus all -p 8000:8000 saikiranappidi/inference:latest
Ensure you have administrative privileges to execute the above command.
- Main URL: Horizon Website
- Docker Hub Container: Docker Hub
- Gemma Model: Gemma by Google
- GitHub Repository: InferenceAPI Repository
- CUDA Downloads: CUDA Download Link
- Django: Django Website
- Testing UI: UI Testing Link
- UI GitHub Repository: Chat-UI
We are actively developing and experimenting with various technologies and open-source LLMs. Contributions are welcome! If you are interested in contributing to the project, please fork the repository and submit a pull request with your proposed changes.
This project is licensed under the MIT License. See the LICENSE file for details.