Rewriting Code with PyTorch and Optimizing with Kan Network

[BUG423]

Issue Description:
I am currently in the process of rewriting some sections of our codebase using PyTorch. This decision comes after observing numerous discussions regarding code execution issues, particularly given that this is an older project with many deprecated tools.

Key Points:
Rewriting with PyTorch: I am transitioning the code from TensorFlow to PyTorch to leverage its current support and capabilities.
Optimization with Kan Network: I am introducing the Kan network for optimization. This approach has already been successfully tested.
Code Update: I am in the process of cleaning up and organizing the updated code for a seamless integration and will upload it as soon as possible.
Next Steps:
Complete the remaining sections of the code rewrite.
Ensure thorough testing of the new implementation with PyTorch and the Kan network.
Document the changes and update the repository with the new code.
I will keep you all posted with my progress and aim to have everything uploaded shortly. Your feedback and suggestions are always welcome!