Introduction

I started writing this book in April 2022. At the time, I had implemented Mu-Zero and was thinking of writing a book using it. After coding the model and completing the training, I looked at the working examples and pondered for a moment. Was I really understanding it correctly? What was lacking? After a while, I found out the reason.

Deep learning and AI are the result of countless failures and trials in a long history of technology. In addition to basic math and physics knowledge, I realized that I lacked a deep understanding of the many alternatives that had failed in the history of technology and why they survived. About a year and a half ago, I started studying math and physics again and rewriting the book from scratch.

Now, I want to share with you the parts I struggled with, albeit insufficiently, and make it public on the internet. The content of the book will be continuously updated as my studies progress. It was written in Korean, and translations into other languages were made with the help of LLM. I also plan to expand translations into other languages continuously.

As I spent a long time working on one subject, there were times when I felt tired. Whenever that happened, my sisters, brothers, and nieces and nephews gave me great strength. Above all, my parents were the driving force behind my ability to complete the book.

Thank you.

March 12, 2025, Park Seon-yong

Environment Setup

1. Using CoLab

Currently, each chapter has a launch button in the top right corner that can be run inside CoLab. It is best to study while running code cells in the CoLab environment.

2. Viewing the Book in HTML Format Locally

After downloading the repository locally, run the following command to view the book in HTML format:

git clone https://github.com/Quantum-Intelligence-Frontier/dldna.git
cd dldna
python -m http.server

Running this command will start a local web server, and you can view the HTML version of the book through a web browser. By default, it can be accessed at http://localhost:8000.

3. Running Jupyter Notebook Locally (Conda Environment)

To run Jupyter Notebook locally and install necessary packages, follow these steps. Using a Conda virtual environment is recommended.

  1. Setting Up the Conda Environment (Anaconda or Miniconda Installation Required)

    • If Anaconda or Miniconda is not installed, install it first. Download the installation file from the Anaconda download page or Miniconda download page and install it.

    • Create a new Conda virtual environment. dldna_env can be changed to the desired environment name. It is recommended to use a Python version compatible with the repository (e.g., 3.9, 3.10, 3.11).

      conda create -n dldna_env python=3.10

    • Activate the created virtual environment.

      conda activate dldna_env

  2. Downloading and Moving to the Repository

    git clone https://github.com/Quantum-Intelligence-Frontier/dldna.git
cd dldna

  3. Installing Necessary Packages

    The requirements.txt file in the dldna directory contains a list of necessary packages. Install these packages using pip.

    pip install -r requirements.txt
    • If an error occurs while running pip install -r requirements.txt, run it from the root folder of the repository, not inside the dldna folder.

  4. Running Jupyter Notebook

    jupyter notebook

    or

    jupyter lab

    Jupyter Notebook or Jupyter Lab will open in a web browser. You can run the notebooks by opening the .ipynb files inside the dldna folder.