Jumari Advanced Virtual Intelligence System
Overview Jarvis-Chat is a Flask-based web application that utilizes various AI models to provide users with a conversational interface for answering questions and completing tasks. It features basic memory for follow-up questions, customizable system prompts, and allows you to select from a variety of AI models. Designed for ease of use, it requires minimal setup to get started. Powering a local ChatGPT-like environment. The system is designed to be user-friendly, accurate, and informative, with a focus on providing relevant and contextual responses.
Features
- Rudimentary Memory: Jarvis has a basic memory that allows it to remember previous conversations and respond accordingly.
- Customizable System Prompt: You can customize the system prompt to fit your specific needs.
- Choose AI Model: Select from a variety of AI models to use with the Groq, Cohere, Mistral, Cerebras, Sambanova API.
- Session Reset: Easily reset the session to start a new conversation.
- Markdown Rendering: Supports rendering of Markdown syntax for formatted text.
- Code Highlighting: Highlights code snippets with syntax highlighting for better readability.
- Code Copying: Allows users to easily copy generated code with a single click.
- Web Search Function: You can perform a web search to get up-to-date information.
- PDF Document Support: Upload PDF documents to JARVIS, which can then be used to provide context-specific responses.
Getting Started
- Clone the repository or download the code.
- cd to jarvis-chat
- Install the required dependencies using
pip install -r requirements.txt
- Set up your API keys and credentials
- Run the application using:
python app.py
Open a web browser and navigate to http://localhost:4001
Example Use Cases
- General Knowledge: Ask JARVIS about general knowledge topics, such as history, science, or entertainment.
- PDF Document Analysis: Upload a PDF document and ask JARVIS to analyze its contents.
- Web Search: Ask JARVIS to search the web for information on a particular topic.
Web search Function
To perform a web search, you can type the word "search" or "cari," and Jarvis will use the top 2 search results as context to provide a response.
You can also get information from a URL. Jarvis will add the content from the URL as context before providing a response.
Contributing
We welcome contributions to JARVIS! If you're interested in adding new features or improving existing ones, please submit a pull request.
License
JARVIS is licensed under the MIT License.