Skip to content

Latest commit

 

History

History
136 lines (103 loc) · 3.34 KB

README.md

File metadata and controls

136 lines (103 loc) · 3.34 KB

Ideas Identification

A sophisticated text analysis tool powered by Llama 3.1-8b through Groq's API, built with Chainlit. This application helps analyze text content by identifying core ideas, relationships, analogies, and generating insights.

Features

  • 🔍 Comprehensive text analysis following a structured framework:
    • Core Ideas identification (Main, Supporting, Context, Counterpoints)
    • Relationship mapping between ideas
    • Analogy analysis
    • Insight generation
  • 🚀 Real-time streaming responses
  • 💬 Persistent chat history within sessions
  • 🎯 Focused analysis prompts
  • ⚡ High-performance inference using Groq's API
  • 🛡️ Secure environment variable handling

Analysis Framework

The tool analyzes text using the following structured approach:

  1. Core Ideas

    • Main Ideas: Central propositions or themes
    • Supporting Ideas: Details, examples, or evidence
    • Contextual Elements: Background and framing devices
    • Counterpoints: Risks, challenges, or opposing views

  2. Relationships

    • Causal relationships
    • Contrast/Comparison
    • Sequential progression
    • Hierarchical connections
    • Associative links

  3. Analogies

    • Comparative elements
    • Supporting concepts
    • Implications and risks

  4. Updated Insights

    • Evolution of ideas
    • Key takeaways
    • Trade-offs and considerations

Prerequisites

Installation

  1. Clone the repository:
git clone https://github.com/quietrags/ideas_identification.git
cd ideas_identification
  1. Install dependencies:
pip install -r requirements.txt
  1. Set up your environment variables:
cp .env.example .env

Then edit .env and add your Groq API key:

GROQ_API_KEY=your_groq_api_key_here

Usage

  1. Start the application:
chainlit run app.py -w
  1. Open your browser and navigate to:
http://localhost:8000
  1. Paste any text you want to analyze into the chat interface
  2. The system will provide a structured analysis following the framework

Example Analysis

Input any text, such as articles, papers, or documents, and receive a structured analysis that includes:

1. Core Ideas
   - Main Ideas: [Key themes and central arguments]
   - Supporting Ideas: [Evidence and examples]
   - Context: [Background information]
   - Counterpoints: [Challenges and opposing views]

2. Relationships
   - [Identified connections between ideas]
   - [Cause-effect relationships]
   - [Comparisons and contrasts]

3. Analogies
   - [Analysis of comparative elements]
   - [Implications and insights]

4. Updated Insights
   - [Key takeaways]
   - [Synthesized understanding]
   - [Practical implications]

Configuration

The application can be configured through:

  • .env: Environment variables including API keys
  • prompt.txt: System prompt defining the analysis framework
  • app.py: Model parameters and application settings

Model Parameters

Current configuration:

  • Model: llama3-8b-8192
  • Temperature: 0.7
  • Max Tokens: 2048
  • Top P: 1.0

License

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

Acknowledgments