PerPlexity Local

<<<<<<< HEAD A local web search and AI response generation tool that uses Ollama for text generation.

Overview

PerPlexity Local is a command-line application that allows you to search the web and get AI-generated responses based on the search results. It uses DuckDuckGo for web searches and Ollama for text generation.

Features

• Web search using DuckDuckGo
• YouTube video transcript extraction and summarization
• AI-powered response generation using Ollama
• Conversation history preservation
• Streaming responses for better user experience

Requirements

• Python 3.9+
• Ollama installed and running locally with a compatible model

Installation

1. Clone the repository
2. Install dependencies:

   pip install -e .
   

3. Make sure Ollama is running with your preferred model (default: dolphin-phi:latest)

Usage

Run the application using one of the following methods:

# Using the run script
python run.py

# Using the module
python -m perplexity_local

# If installed via pip
perplexity-local
=======
A local implementation of a Perplexity-like search and answer tool that uses DuckDuckGo for web searches and Ollama for local LLM-powered answer synthesis.

## Features

- **Web Search**: Uses DuckDuckGo to search the web for relevant information
- **Content Extraction**: Intelligently extracts and processes content from web pages
- **YouTube Support**: Extracts and processes transcripts from YouTube videos
- **Local LLM Integration**: Uses Ollama for local, private answer generation
- **Conversation History**: Maintains context across multiple queries
- **Streaming Responses**: Displays AI responses in real-time as they're generated

## Requirements

- Python 3.6+
- Ollama installed and running locally
- Internet connection for web searches

## Dependencies

requests beautifulsoup4 duckduckgo_search youtube-transcript-api (optional, for YouTube transcript extraction)

## Installation

1. Clone the repository:
```bash
git clone https://github.com/yourusername/PerPlexity_Local.git
cd PerPlexity_Local

2. Create a virtual environment (recommended):

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install the required packages:

pip install requests beautifulsoup4 duckduckgo_search

4. For YouTube transcript support (optional):

pip install youtube-transcript-api

5. Install and start Ollama:
   • Follow the instructions at Ollama's official website to install
   • Pull the model you want to use: ollama pull dolphin-phi

Configuration

The main configuration options are at the top of main.py:

• MAX_RESULTS: Number of search results to process (default: 5)
• REQUEST_TIMEOUT: Timeout for web requests in seconds (default: 4)
• MAX_LEN_PER_SOURCE: Maximum text length to extract from each source (default: 20000)
• OLLAMA_API_URL: URL for the Ollama API (default: "http://localhost:11434/api/generate")
• OLLAMA_MODEL: The Ollama model to use (default: "dolphin-phi:latest")
• OLLAMA_REQUEST_TIMEOUT: Timeout for Ollama requests in seconds (default: 120)
• MAX_HISTORY_TURNS: Number of conversation turns to remember (default: 3)
• HISTORY_ENABLED: Whether to use conversation history (default: True)

Usage

Run the program:

python main.py
>>>>>>> 78986bf3b7c36dc59543c1de9904efa742d7efe6

Commands

<<<<<<< HEAD

• Enter your query to search the web and get an AI-generated response
• Type exit to quit the application
• Type clear history to clear conversation history
• Type toggle history to enable/disable conversation history

Project Structure

perplexity_local/
├── config/           # Configuration settings
├── core/             # Core functionality modules
├── utils/            # Utility functions
└── __main__.py       # Entry point

Configuration

You can modify the configuration settings in perplexity_local/config/settings.py:

• MAX_RESULTS: Maximum number of search results to fetch (default: 5)
• OLLAMA_MODEL: The Ollama model to use (default: dolphin-phi:latest)
• MAX_HISTORY_TURNS: Number of previous turns to keep in context (default: 3)
• HISTORY_ENABLED: Toggle for enabling/disabling history (default: True)

License

MIT

• Enter your query at the prompt
• Type exit to quit the program
• Type clear history to clear conversation history
• Type toggle history to enable/disable conversation history

YouTube Support

You can include a YouTube URL in your query to get a summary of the video:

Query: Summarize this video https://www.youtube.com/watch?v=dQw4w9WgXcQ

How It Works

1. Search: The program searches the web using DuckDuckGo or processes a YouTube URL
2. Extract: It extracts and processes content from the search results
3. Synthesize: The extracted content is sent to Ollama for answer generation
4. Display: The answer is displayed in real-time as it's generated

Limitations

• Depends on DuckDuckGo's search API which may change
• YouTube transcript extraction requires the video to have captions
• Limited by the capabilities of the local LLM model used with Ollama
• Web content extraction may not work perfectly on all websites

License

MIT License

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

78986bf3b7c36dc59543c1de9904efa742d7efe6