How idvorkin/nlp works
Main Components
a) Core Processing Components:
• NLP Processing Engine (centered around files like think.py, gpt.py, gpt3.py)
• Language Model Integration (evidenced by OpenAI integration via openai_wrapper.py)
• Text Analysis Tools (shown by various notebooks and Python scripts)
b) Input/Output Systems:
• CLI Interface Layer (multiple CLI tools mentioned in README)
• Journal Processing System (referenced in igor_journal.py, journal.py)
• Voice/Audio Processing (transcribe.py, tts.py)
c) Bot Systems:
• Multiple Bot Implementations (bestie-bot.py, improv_bot.py)
• Conversation Management (convos directory)
d) Evaluation Framework:
• Testing Infrastructure (tests directory)
• Evaluation Tools (eval directory)
• Prompt Management System (prompts.py)
┌─ User Interface Layer ─────────────────────┐
• CLI Tools │ Notebooks │ Bot Interfaces │
• Application Layer ──────────────────────┐ │
• Journal Processing │ NLP Analysis │ │
• Conversation Management │ Bot Logic │ │
• Core Services Layer ────────────────────┤ │
• OpenAI Integration │ Language Processing │ │
• Text Analysis │ Voice Processing │ │
• ┘
Key Relationships to Highlight
• Data flow between journal entries and NLP processing
• How the bot systems interact with the core NLP services
• Demonstrate the evaluation framework's relationship with other components
• Integration points with external services (OpenAI, etc.)
Implementation Details
• Rectangles: Core services
• Hexagons: External services
• Rounded rectangles: User interfaces
Annotations for:
• File formats supported
• API endpoints
• Configuration points
• Testing boundaries
Additional Elements
• Version/environment information
• Dependencies and requirements
• Scaling/deployment considerations