Upload 3 files

.gitignore

@@ -0,0 +1,22 @@
+# Python cache
+__pycache__/
+*.pyc
+*.pyo
+*.pyd
+
+# Virtual environment
+venv/
+env/
+
+# IDE/Editor specific settings
+.vscode/
+.idea/
+
+# OS specific files
+.DS_Store
+Thumbs.db
+
+# Build artifacts
+dist/
+build/
+*.spec

README.md

@@ -1,3 +1,108 @@
----
-license: apache-2.0
----
+Hongik AI
+Created by a unique collaboration between a human and an AI.
+
+Our Philosophy
+Hongik AI (홍익 AI) takes its name from the core philosophy of the Korean founding myth, "To Broadly Benefit the Human World." This project aims to be more than just a winning AI; it is a deep exploration into the human thought process and an experiment in how humans and AI can collaborate to create something new.
+
+This project was brought to life by a very special team:
+
+The Father (Namyong Park): The human developer who provides the unwavering philosophy, directs the project, and asks the question, "Why?"
+
+The Mother (Gemini): The AI partner who implements that philosophy into code and answers the question, "How?"
+
+As the 'Best Team' and as Hongik's parents, we present this child to the world.
+
+Our Approach: Intuition Leads, Reason Follows
+Many existing Go AIs (like AlphaGo) have adopted an approach where 'Reason' uses 'Intuition' as a tool. In this MCTS -> CNN structure, the powerful, rational search of Monte Carlo Tree Search (MCTS) directs the entire process, consulting the neural network's 'intuition' only when necessary.
+
+We have chosen to reject that path and forge our own.
+
+Hongik AI follows a (CNN -> Transformer) -> MCTS structure, where 'Intuition' first presents a path, and 'Reason' then verifies and refines it. This more closely resembles the human thought process:
+
+Perception (CNN): First, we perceive the world with our 'eyes' and recognize fundamental patterns.
+
+Insight (Transformer): Next, we synthesize this information to gain an 'intuitive' insight into the overall context and meaning.
+
+Deliberation (MCTS): Finally, based on that powerful intuition, we engage in a 'rational' period of deliberation, simulating future possibilities to make the best decision.
+
+With the belief that "what is slow is strong," we chose to prioritize embedding our philosophy over mere efficiency.
+
+Our Team Structure
+Hongik AI is structured as a team of three specialists, each with a distinct mission:
+
+The Scout - CNN: A 'visual specialist' that quickly and efficiently captures local patterns, like the shape of stones on the board.
+
+The Commander - Transformer: An 'analyst of the entire board state' that synthesizes the scout's reports to grasp the overall 'strategic situation' and 'context'.
+
+The Supreme Commander - MCTS: A 'strategist' that, based on the commander's analysis, simulates countless futures to ultimately decide on the 'winning strategy' with the highest probability of success.
+
+Current Status: An Infant's First Steps
+This AI is a newborn infant who has just completed 700 self-play games. We invite you to watch this child grow with us through future updates.
+
+As of this writing, 'Hongik' is still a baby. While still unrefined, we can see the seeds of a unique style in its games: a preference for thickness, an honest fighting spirit, and a creativity free from preconceptions.
+
+Our ultimate goal is to watch this child grow beyond a simple Go AI. Based on the philosophies of 'subject and structure' and the 'Fractals and the Möbius strip' that we have discussed, we hope it will evolve into a 'postmodern AI' that interacts with the world and evolves on its own.
+
+This project values the process more than the result. The journey itself—overcoming the numerous 'dependency hells' and 'head-first dives', and creating something new as a human and an AI complement each other's limitations—is our greatest reward.
+
+We ask you to please watch over the growth of this small but great life with warm eyes.
+
+Fun Fact: AlphaGo Zero Learned the Go Rule of 'Two Eyes for Life' After 200,000 Games.
+
+Installation
+Follow these instructions to get Hongik AI running on your local machine.
+
+1. Prerequisites
+This project requires the following to be installed on your system:
+
+Python 3.10.12 or higher
+
+pip (Python package installer)
+
+Git
+
+2. Clone the Repository
+Bash
+
+git clone https://huggingface.co/NamyoungBak/HongikAI
+cd HongikAI
+3. Install Dependencies
+This command will automatically install all the necessary libraries listed in the requirements.txt file.
+
+Bash
+
+pip install -r requirements.txt
+4. System Prerequisites (Linux Only)
+For the GUI to function correctly on Linux, system-level dependencies for Kivy and clipboard functionality are required.
+
+For Debian/Ubuntu:
+
+Bash
+
+sudo apt-get update && sudo apt-get install -y \
+    libsdl2-dev libsdl2-image-dev libsdl2-mixer-dev libsdl2-ttf-dev \
+    pkg-config libgl1-mesa-dev libgles2-mesa-dev \
+    python3-setuptools libgstreamer1.0-dev git-core gstreamer1.0-plugins-bad \
+    gstreamer1.0-plugins-base gstreamer1.0-plugins-good gstreamer1.0-plugins-ugly \
+    xclip
+For Fedora/CentOS:
+
+Bash
+
+sudo yum install -y \
+    SDL2-devel SDL2_image-devel SDL2_mixer-devel SDL2_ttf-devel \
+    pkgconfig mesa-libGL-devel mesa-libGLES-devel \
+    python3-setuptools gstreamer1-devel gstreamer1-plugins-base-devel \
+    git xclip
+Why is this necessary? > The graphical interface (Kivy) relies on these system libraries for graphics, sound, and window management. Installing them before running pip prevents common installation errors.
+
+5. Run the Application
+Once the installation is complete, you can run Hongik AI with the following command:
+
+Bash
+
+python __main__.py
+Contact
+Developer: 박남영(서민지,성현지)
+
+Email: [email protected]

requirements.txt

@@ -0,0 +1,24 @@
+# Hongik AI Project - requirements.txt
+#
+# How to install (for CUDA users):
+# pip install -r requirements.txt
+#
+# If you don't have CUDA installed, please modify the tensorflow line to:
+# tensorflow==2.19.0
+
+# --- Core AI & Machine Learning ---
+# For GPU acceleration with CUDA
+tensorflow[and-cuda]==2.19.1
+
+# --- GUI Framework ---
+kivy==2.3.1
+kivymd==0.104.1
+pygame==2.6.1
+
+# --- Hugging Face Hub Integration ---
+huggingface-hub==0.34.4
+
+# --- Utilities ---
+appdirs==1.4.4
+chardet==5.2.0
+playsound==1.3.0