What is Chatty?
Chatty is your private AI that leverages WebGPU technology to run large language models (LLMs) natively in your browser. This means you can enjoy a feature-rich in-browser AI experience, all while keeping your data private.
Features
- In-browser privacy: All AI models run locally on your hardware with no server-side processing.
- Offline capability: After the initial model download, Chatty works without an internet connection.
- Chat history: Easily access and manage your past conversations.
- Open-source model support: Engage with popular models like Gemma, Llama2 3, and Mistral!
- Responsive design: Optimized for phone usage if WebGL is supported.
- Intuitive UI: Inspired by familiar AI interfaces for a seamless user experience.
- Markdown code highlighting: Code snippets returned as markdown are highlighted for clarity.
- Chat with files: Load non-binary files (including PDFs and code files) to pose questions about their content.
- Custom memory support: Tailor AI to provide personalized responses with custom memory settings.
- Export chat messages: Save your conversations in JSON or markdown formats.
- Voice input support: Use voice interactions with the AI models.
- Response regeneration: Quickly regenerate responses without retyping prompts.
- Light/Dark mode: Switch between light and dark themes for your comfort.
Installation Guide
If you’re eager to explore Chatty, it’s live on the official website. However, if you’re interested in setting it up locally, follow these steps:
- Clone the repository:
git clone https://github.com/addyosmanichatty - Change into the directory:
cd chatty - Install the necessary dependencies:
npm install - Start the development server:
npm run dev - Visit localhost:3000 to start chatting!
Understanding the Technology – An Analogy
Imagine Chatty as a small, self-contained library located within your home. Instead of needing to travel to a large city library to access books (which represent online databases), you can pick and choose from your own collection of knowledge without ever stepping out.
Furthermore, this library allows you to invite friends over (which parallels the chat history feature), and you can have discussions without anyone outside the home overhearing (comparable to in-browser privacy). The experience is streamlined and personalized, allowing you to interact with your books (AI models) and even have them respond to prompts tailored just for you! When you want a different discussion about a particular book, you can revisit the topic without repeatedly starting from scratch (response regeneration).
Troubleshooting
If you run into issues during installation or while running Chatty, here are a few troubleshooting steps:
- Ensure you have the latest version of Node.js (18+).
- Check that you’ve correctly installed all dependencies by running
npm install.
- For issues with the server starting, ensure that you are in the correct directory.
- In case your browser does not support WebGPU, consider switching to Chrome or Edge, or follow instructions to enable it in Firefox.
For more insights, updates, or to collaborate on AI development projects, stay connected with fxis.ai.
Final Thoughts
At fxis.ai, we believe that such advancements are crucial for the future of AI, as they enable more comprehensive and effective solutions. Our team is continually exploring new methodologies to push the envelope in artificial intelligence, ensuring that our clients benefit from the latest technological innovations.