rishal-hurbans/Grokking-Artificial-Intelligence-Algorithms

The official code repository supporting the book, Grokking Artificial Intelligence Algorithms

What it solves

The repository gives you working Python code that implements the AI algorithms described in the book Grokking AI Algorithms. Instead of just reading about an algorithm, you can run it yourself, modify it, and see how it behaves. It's meant to fill the gap between understanding the concept and actually coding it.

How it works

Each chapter of the book has a corresponding directory (e.g., ch03-intelligent_search/informed_search) containing runnable Python scripts. You set up a Python environment (3.9+), install the dependencies in requirements.txt, and then run any example script directly. Some of the heavier demos in chapters 11 and 12 can optionally use a GPU with PyTorch. There's also an interactive notebook available on Google Colab that lets you explore the code in your browser without local setup.

Who it's for

It's aimed at readers of the book who want to see the algorithms implemented in code, or who want a programming-focused look at how these AI algorithms work. The author notes that the examples are easier to understand if you've read the book, but they may still be useful to others.

Highlights

  • Official companion code for the book Grokking AI Algorithms.
  • Runnable examples for each chapter, with clear directory structure.
  • Minimal setup: Python 3.9+, pip install -r requirements.txt.
  • Interactive Colab notebook available for browser-based exploration.
  • Optional GPU support for the more demanding demos in later chapters.

Related

  • Project
    chi2liu/ABC-GRPO

    An asymmetric and adaptive refinement of the GRPO reinforcement learning algorithm that uses four independent clipping boundaries to prevent entropy collapse and improve generalization in LLM reasoning tasks.

  • Project
    ServiceNow/PipelineRL

    A scalable asynchronous reinforcement learning framework for LLM agents that uses in-flight weight updates to maximize GPU utilization while maintaining on-policy data freshness.

  • Project
    nftechie/doomfly

    A simulation that connects a biologically reconstructed fruit fly connectome to a ViZDoom arena to test if biological neural wiring can be trained for survival using dopamine-gated memory rules.

  • Project
    rllm-team/rllm

    rLLM is a PyTorch library that provides a modular framework for Relational Table Learning, allowing users to combine LLMs, GNNs, and TNNs to analyze complex, interconnected tabular data.