Welcome to spladtool’s documentation!

spladtool is a Python library designed to demonstrate the concept simple automatic differentiation. It is the abbreviation for Simple PyTorch-like Auto Differentiation Toolkit. spladtool offers an accessible interface similar to PyTorch, and offers very simple utils to perform gradient-based optimization.

spladtool is written in pure Python with the support of NumPy, and is protected by MIT license.

Check out the Usage section for further information, including how to Installation the project.

Note

This project is only for educational purposes.

Contents

• Concepts in our library
  • Jacobian Matrix and Gradients
  • Forward Mode
  • Reverse Mode
• Usage
  • Installation
  • Live example
  • Forward Mode
  • Reverse Mode
  • Build Model and Optimize
• APIs
• Development Designs
  • Data Structures
  • Functional APIs
  • Supported Operations
  • Python Typing
• Test Suite
• Future Extensions
• Broader Impact
• Software Inclusivity