HuggingChat: A New AI-Powered Chatbot
HuggingChat, the first open source alternative to ChatGPT, is now available for testing on Hugging Face. This AI assistant aims to provide users with a friendly, human-like personality and the ability to understand and respond to natural language queries. HuggingChat is capable of performing various tasks, such as drafting articles, solving coding problems, and answering questions, making it an attractive option for users seeking an AI-powered chatbot.
HuggingChat’s Impressive Features and Future Goals
With 30 billion parameters, HuggingChat is currently the best open source chat model according to Hugging Face. The AI startup plans to expose all chat models available on the Hub in the long term. HuggingChat is based on the latest LLaMa model developed by the OpenAssistant project, which has ambitious goals to go beyond ChatGPT. OpenAssistant aims to build the assistant of the future, capable of not only writing emails and cover letters but also performing meaningful work, using APIs, dynamically researching information, and much more. The project also envisions the AI assistant to be small, efficient, and customizable, allowing it to run on consumer hardware.
OpenAssistant and LAION: Fostering Machine Learning Research
OpenAssistant is managed under LAION, a non-profit organization that provides open datasets, tools, and models to foster machine learning research. LAION is responsible for the LAION-5B dataset, which forms the basis for the Stable Diffusion model. HuggingChat currently enforces a strict privacy model, ensuring that messages are only stored for user display and not shared for research or training purposes. Users are not authenticated or identified using cookies. However, this privacy model may change in the future to allow users to share their conversations with researchers.
Open Source Code and Customization Options
HuggingChat is the first AI-powered chat project that is open source. The code for the user interface can be found on GitHub, while the inference backend runs text-generation-inference on Hugging Face’s Inference API infrastructure. This setup allows the app to be deployed to a Hugging Face Space and customized in various ways, including swapping models, modifying the user interface, and changing the policy regarding storing user messages.
Leave a Reply