Refactor and Enhance Kistmat_AI Model

[mentatbot[bot]]

This pull request addresses the refactoring and enhancement of the Kistmat_AI model as outlined in the issue. The changes include:

1. Code Reorganization:

   • Created a logical folder structure (src/, tests/, docs/).
   • Moved existing classes and functions to appropriate files.
   • Created a main entry point (src/main.py) for initializing and running the model.
   • Updated import statements across all files.

2. Implementation of Tests:

   • Added unit tests for key components (ExternalMemory, Kistmat_AI, SymbolicReasoner).
   • Implemented integration tests to verify the complete training flow.
   • Added specific tests for symbolic reasoning, knowledge transfer, symbolic consistency, long-term memory, and concept generalization.

3. Symbolic Reasoning Enhancement:

   • Expanded the SymbolicReasoner class with additional rules and symbols.
   • Integrated symbolic reasoning more deeply into the training process.
   • Implemented a symbolic loss function to ensure symbolic consistency.

4. Proximal Policy Optimization (PPO) Integration:

   • Created a PPOAgent class to handle training with PPO.
   • Modified the training function to incorporate PPO.
   • Adjusted hyperparameters for optimal performance with PPO.

5. New Memory Systems:

   • Developed a MemorySystem class integrating various memory components (Formulative, Conceptual, Short-Term, Long-Term, Inference).
   • Integrated the new memory system into the Kistmat_AI model.
   • Updated training and evaluation functions to utilize the new memory system.

6. Bug Fixes:

   • Identified and corrected issues during epoch transitions.
   • Implemented control mechanisms to ensure model stability during training.

7. Documentation:

   • Updated existing documentation to reflect changes.
   • Added new documentation for PPO and memory systems.
   • Created a detailed README with setup and usage instructions.

8. Performance Optimization:

   • Conducted performance analysis to identify bottlenecks.
   • Optimized costly operations in terms of time and resources.

Acceptance Criteria:

• All tests pass successfully.
• Significant improvement in symbolic reasoning.
• Correct implementation and performance improvement with PPO.
• New memory systems are integrated and functional.
• No errors during epoch transitions.
• Code is well-documented and follows Python best practices.

Closes #1

Thanks for using MentatBot. Give comments a :+1: or :-1: to help me improve!