README.md

metadata

title: 🤖 Agentic Browser
emoji: 🤖
colorFrom: blue
colorTo: purple
sdk: docker
app_port: 8501
pinned: false
license: mit

🤖 Agentic Browser

A powerful, privacy-focused AI assistant that runs locally on your machine. This application allows you to interact with various open-source language models directly in your browser, with support for both lightweight and more powerful models.

🌟 Features

• Local AI Models: Run models entirely on your hardware
• Multiple Models: Choose between different models based on your needs
• Privacy-Focused: Your data never leaves your machine
• Easy to Use: Simple and intuitive chat interface
• Customizable: Adjust parameters like temperature for different response styles

🚀 Getting Started

Prerequisites

• Python 3.10 or higher
• Git
• At least 8GB RAM (16GB+ recommended for larger models)
• At least 10GB free disk space for models

Installation

1. Clone the repository:

   git clone https://huggingface.co/spaces/anu151105/agentic-browser
   cd agentic-browser
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Download models (this may take a while):

   python scripts/download_models.py --model tiny-llama
   

   For the full experience with all models:

   python scripts/download_models.py --model all
   

Running the Application

1. Start the Streamlit app:

   streamlit run src/streamlit_app.py
   

2. Open your browser to the URL shown in the terminal (usually http://localhost:8501)

3. Select a model from the sidebar and click "Load Model"

4. Start chatting!

🤖 Available Models

• TinyLlama (1.1B): Fast and lightweight, runs well on most hardware

  • Great for quick responses
  • Lower resource requirements
  • Good for testing and development

• Mistral-7B (7B): More powerful but requires more resources

  • Better understanding and generation
  • Requires more RAM and VRAM
  • Slower response times

⚙️ Configuration

You can customize the behavior of the application by setting environment variables in a .env file:

# Model settings
MODEL_DEVICE=auto  # 'auto', 'cuda', 'cpu'
MODEL_CACHE_DIR=./models

# Text generation settings
DEFAULT_TEMPERATURE=0.7
MAX_TOKENS=1024

# UI settings
THEME=light  # 'light' or 'dark'

🛠️ Development

Adding New Models

1. Edit config/model_config.py to add your model configuration
2. Update the DEFAULT_MODELS dictionary with your model details
3. The model will be available in the UI after restarting the app

Running Tests

pytest tests/

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• Hugging Face for the amazing Transformers library
• Streamlit for the awesome web framework
• The open-source AI community for making these models available emoji: 🚀 colorFrom: red colorTo: red sdk: docker app_port: 8501 tags:
• streamlit pinned: false short_description: An autonomus browser agent for browser based tasks license: mit

---

Welcome to Streamlit!

Edit /src/streamlit_app.py to customize this app to your heart's desire. :heart:

If you have any questions, checkout our documentation and community forums.