{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## Начало крайней лабораторной в этом семестре, эх..." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Что необходимо сделать:\n", "Развернуть и запустить проект по реализации обучения с подкреплением для игры \"Крестики-нолики\". Перевести проект на библиотеку gymnasium и современную версию Python. Реализовать агента для игры \"Крестики-нолики\" в виде отдельного класса (по примеру из лекции). Переписать основной цикл обучения для работы с отдельным классом агента (по примеру из лекции). Выполнить тестирование новой версии программы." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Наши крестики-нолики: https://github.com/nczempin/gym-tic-tac-toe" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Перевод проекта на библиотеку gymnasium\n", "Gymnasium — это открытая библиотека Python, которая предоставляет стандартизированные среды для разработки и тестирования алгоритмов обучения с подкреплением (Reinforcement Learning, RL). Ранее была известна как OpenAI Gym (до 2022 года), но теперь развивается под новым именем.\n", "\n", "Библиотека позволяет разработчикам RL-агентов взаимодействовать с различными симуляциями — от простых игровых задач до сложных физических моделей. Gymnasium упрощает процесс тестирования и сравнения алгоритмов благодаря унифицированному интерфейсу.\n", "\n", "**Основные возможности Gymnasium:**\n", "1. Унифицированный API для RL-сред:\n", "\n", "- Gymnasium предлагает стандартный интерфейс для взаимодействия с RL-средами, включающий методы reset(), step(action), и другие.\n", "- Это позволяет легко переключаться между различными средами без изменения кода агента.\n", "2. Разнообразие встроенных сред:\n", "\n", "- Библиотека включает множество готовых симуляций, от простых (например, CartPole, MountainCar) до сложных (например, робототехника, Atari-игры).\n", "- Среды подразделяются на категории: контроль, игры, физика, робототехника и др.\n", "3. Поддержка классических задач RL:\n", "\n", "- Среды для изучения классических задач, таких как балансировка маятника, решение головоломок, управление роботами и т.д.\n", "4. Гибкость создания пользовательских сред:\n", "\n", "- Gymnasium позволяет разработчикам создавать собственные симуляции, соответствующие API.\n", "5. Совместимость с различными RL-библиотеками:\n", "\n", "- Gymnasium интегрируется с популярными RL-фреймворками, такими как Stable-Baselines3, Ray RLlib, TensorFlow Agents, PyTorch RL и другими.\n", "6. Визуализация:\n", "\n", "- Среды могут визуализироваться, что упрощает отладку и демонстрацию работы алгоритмов.\n", "\n", "**Основные функции Gymnasium:**\n", "- Инициализация среды (gymnasium.make()): позволяет создавать экземпляр среды по её имени.\n", "- Сброс среды (reset()): возвращает начальное состояние среды и другую информацию.\n", "- Выполнение действия (step(action)): передаёт действие агенту и возвращает результат.\n", "- Закрытие среды (close()): очищает ресурсы, связанные со средой.\n", "- Режим рендеринга (render()): позволяет визуализировать работу среды." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "import gymnasium as gym\n", "from gymnasium import spaces\n", "\n", "class TicTacToeEnv(gym.Env):\n", " metadata = {'render.modes': ['human']}\n", " \n", " symbols = ['O', ' ', 'X']\n", "\n", " def __init__(self):\n", " super().__init__()\n", " self.action_space = spaces.Discrete(9) # Дискретное пространство действий (0–8), соответствующее номерам клеток на игровом поле (от 0 до 8).\n", " self.observation_space = spaces.Discrete(9 * 3 * 2) # Дискретное пространство состояний. Расчёт: 9 клеток × 3 состояния (пустая, крестик, нолик) × 2 игрока (на чей ход).\n", " self.reset()\n", "\n", " def step(self, action):\n", " done = False\n", " reward = 0\n", "\n", " p, square = action # p - игрок (1 или -1), square - номер клетки\n", "\n", " board = self.state['board']\n", " proposed = board[square] \n", " om = self.state['on_move'] \n", " if proposed != 0: # Клетка уже занята\n", " print(f\"Незаконный ход: Квадрат {square} уже занят.\")\n", " done = True\n", " reward = -1 * om \n", " if p != om: # Не тот игрок на ходу\n", " print(f\"Незаконный ход: игрок {p} не находится в движении\")\n", " done = True\n", " reward = -1 * om\n", " else:\n", " board[square] = p\n", " self.state['on_move'] = -p\n", "\n", " for i in range(3):\n", " # Горизонтали и вертикали\n", " if (board[i * 3] == p and board[i * 3 + 1] == p and board[i * 3 + 2] == p) or \\\n", " (board[i] == p and board[i + 3] == p and board[i + 6] == p):\n", " reward = p\n", " done = True\n", " break\n", "\n", " # Диагонали\n", " if (board[0] == p and board[4] == p and board[8] == p) or \\\n", " (board[2] == p and board[4] == p and board[6] == p):\n", " reward = p\n", " done = True\n", " \n", " return self.state, reward, done, {}\n", "\n", " def reset(self):\n", " self.state = {}\n", " self.state['board'] = [0, 0, 0, 0, 0, 0, 0, 0, 0] \n", " self.state['on_move'] = 1 \n", " return self.state, {}\n", "\n", " def render(self, close=False):\n", " if close:\n", " return\n", " print(\"on move: \" , self.symbols[self.state['on_move']+1])\n", " for i in range (9):\n", " print (self.symbols[self.state['board'][i]+1], end=\" \")\n", " if ((i % 3) == 2):\n", " print()\n", "\n", " def move_generator(self):\n", " moves = []\n", " for i in range(9):\n", " if self.state['board'][i] == 0:\n", " p = self.state['on_move']\n", " m = [p, i]\n", " moves.append(m)\n", " return moves" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Реализация агента\n", "В контексте обучения с подкреплением (Reinforcement Learning, RL) агент — это программный компонент, который взаимодействует с окружающей средой (environment) с целью научиться выбирать оптимальные действия для достижения своей цели.\n", "\n", "**Основные аспекты агента:**\n", "1. Что делает агент?\n", "\n", "- Агент принимает решения: в каждом состоянии среды он выбирает действие (например, на какую клетку походить в крестиках-ноликах или как двигаться в игре).\n", "- Он учится выбирать действия, которые максимизируют его \"награду\" (reward), получаемую от среды.\n", "2. Как агент учится?\n", "\n", "- Агент улучшает свою стратегию на основе опыта взаимодействия со средой.\n", "- Этот процесс осуществляется с использованием алгоритмов обучения с подкреплением (например, Q-learning, глубокое Q-обучение или методы политики).\n", "\n", "Ключевые элементы агента:\n", "1. Наблюдение (Observation): Агент воспринимает текущее состояние среды. Это может быть, например, игровое поле, показания датчиков или изображение. Наблюдение агент получает в виде данных от среды.\n", "2. Действие (Action): В каждом шаге агент выбирает действие из допустимого набора действий. Этот набор зависит от правил среды (например, куда можно походить в игре).\n", "3. Политика (Policy): Политика — это стратегия, которой следует агент, чтобы выбирать действия.\n", "Она может быть:\n", "- Детерминированной: одно и то же состояние всегда приводит к одному и тому же действию.\n", "- Стохастической: для одного состояния агент выбирает действие с определённой вероятностью.\n", "4. Награда (Reward): После выполнения действия агент получает награду — числовое значение, отражающее, насколько его действие было \"хорошим\".\n", "Награды помогают агенту оценивать свои действия и формировать полезное поведение.\n", "5. Функция ценности (Value Function):\n", "Это внутренняя оценка агента, которая отражает, насколько \"хорошим\" или \"перспективным\" является определённое состояние.\n", "Она позволяет агенту предсказывать долгосрочные последствия своих действий.\n", "6. Алгоритм обучения:\n", "Агент использует алгоритмы для обновления своей политики и функции ценности, основываясь на полученных наградах.\n", "Примеры: метод Q-learning, методы на основе градиента политики (Policy Gradient), алгоритмы Actor-Critic и т. д.\n" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "import random\n", "\n", "class Agent:\n", " def __init__(self, symbol):\n", " self.symbol = symbol # Символ игрока (1 - X, -1 - O)\n", " \n", " def get_action(self, moves):\n", " return random.choice(moves) # Выбираем случайный ход из доступных" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### Основной цикл обучения" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " O X \n", " \n", " \n", "on move: O\n", " O X \n", " X \n", " \n", "on move: X\n", "O O X \n", " X \n", " \n", "on move: O\n", "O O X \n", " X X \n", " \n", "on move: X\n", "O O X \n", " X X \n", " O \n", "on move: O\n", "O O X \n", " X X \n", "X O \n", "Episode 1, Total Reward: 1\n", "Average Reward: 1.0\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", "O \n", "on move: O\n", "X X \n", " \n", "O \n", "on move: X\n", "X X \n", " \n", "O O \n", "on move: O\n", "X X \n", " X \n", "O O \n", "on move: X\n", "X X O \n", " X \n", "O O \n", "on move: O\n", "X X O \n", " X X \n", "O O \n", "on move: X\n", "X X O \n", "O X X \n", "O O \n", "on move: O\n", "X X O \n", "O X X \n", "O O X \n", "Episode 2, Total Reward: 1\n", "Average Reward: 1.0\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", " X \n", " \n", " X O \n", "on move: X\n", " X \n", " \n", "O X O \n", "on move: O\n", "X X \n", " \n", "O X O \n", "on move: X\n", "X X \n", " O \n", "O X O \n", "on move: O\n", "X X \n", " O X \n", "O X O \n", "on move: X\n", "X X \n", "O O X \n", "O X O \n", "on move: O\n", "X X X \n", "O O X \n", "O X O \n", "Episode 3, Total Reward: 1\n", "Average Reward: 1.0\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", "O \n", " X \n", " \n", "on move: O\n", "O \n", " X \n", " X \n", "on move: X\n", "O \n", " X O \n", " X \n", "on move: O\n", "O X \n", " X O \n", " X \n", "on move: X\n", "O X \n", " X O \n", " O X \n", "on move: O\n", "O X X \n", " X O \n", " O X \n", "on move: X\n", "O X X \n", " X O \n", "O O X \n", "on move: O\n", "O X X \n", "X X O \n", "O O X \n", "Episode 4, Total Reward: 0\n", "Average Reward: 0.75\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " O \n", " \n", "X \n", "on move: O\n", " O \n", "X \n", "X \n", "on move: X\n", " O \n", "X \n", "X O \n", "on move: O\n", " O \n", "X \n", "X O X \n", "on move: X\n", " O O \n", "X \n", "X O X \n", "on move: O\n", "X O O \n", "X \n", "X O X \n", "Episode 5, Total Reward: 1\n", "Average Reward: 0.8\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " \n", "X O \n", " \n", "on move: O\n", " \n", "X X O \n", " \n", "on move: X\n", " \n", "X X O \n", "O \n", "on move: O\n", " \n", "X X O \n", "O X \n", "on move: X\n", " O \n", "X X O \n", "O X \n", "on move: O\n", " X O \n", "X X O \n", "O X \n", "on move: X\n", " X O \n", "X X O \n", "O O X \n", "on move: O\n", "X X O \n", "X X O \n", "O O X \n", "Episode 6, Total Reward: 1\n", "Average Reward: 0.8333333333333334\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " O \n", " X \n", "on move: O\n", " \n", "X O \n", " X \n", "on move: X\n", " O \n", "X O \n", " X \n", "on move: O\n", "X O \n", "X O \n", " X \n", "on move: X\n", "X O \n", "X O \n", " O X \n", "Episode 7, Total Reward: -1\n", "Average Reward: 0.5714285714285714\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", " O \n", "on move: O\n", " \n", "X X \n", " O \n", "on move: X\n", " \n", "X X O \n", " O \n", "on move: O\n", " X \n", "X X O \n", " O \n", "on move: X\n", " O X \n", "X X O \n", " O \n", "on move: O\n", " O X \n", "X X O \n", " X O \n", "on move: X\n", " O X \n", "X X O \n", "O X O \n", "on move: O\n", "X O X \n", "X X O \n", "O X O \n", "Episode 8, Total Reward: 0\n", "Average Reward: 0.5\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", " X O \n", "on move: O\n", "X \n", " \n", " X O \n", "on move: X\n", "X \n", " O \n", " X O \n", "on move: O\n", "X X \n", " O \n", " X O \n", "on move: X\n", "X X \n", " O \n", "O X O \n", "on move: O\n", "X X \n", "X O \n", "O X O \n", "on move: X\n", "X O X \n", "X O \n", "O X O \n", "on move: O\n", "X O X \n", "X X O \n", "O X O \n", "Episode 9, Total Reward: 0\n", "Average Reward: 0.4444444444444444\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", " X O \n", "on move: O\n", " \n", " \n", "X X O \n", "on move: X\n", " \n", " O \n", "X X O \n", "on move: O\n", " \n", " O X \n", "X X O \n", "on move: X\n", "O \n", " O X \n", "X X O \n", "Episode 10, Total Reward: -1\n", "Average Reward: 0.3\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " O \n", "X X \n", " \n", "on move: X\n", " O \n", "X X \n", "O \n", "on move: O\n", " O \n", "X X X \n", "O \n", "Episode 11, Total Reward: 1\n", "Average Reward: 0.36363636363636365\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " \n", "X O \n", " \n", "on move: O\n", " \n", "X O \n", " X \n", "on move: X\n", " O \n", "X O \n", " X \n", "on move: O\n", " O \n", "X O \n", "X X \n", "on move: X\n", " O \n", "X O \n", "X O X \n", "on move: O\n", " X O \n", "X O \n", "X O X \n", "on move: X\n", "O X O \n", "X O \n", "X O X \n", "on move: O\n", "O X O \n", "X O X \n", "X O X \n", "Episode 12, Total Reward: 0\n", "Average Reward: 0.3333333333333333\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " O X \n", " \n", "on move: O\n", " X \n", " O X \n", " \n", "on move: X\n", " X \n", " O X \n", " O \n", "on move: O\n", " X \n", " O X \n", "X O \n", "on move: X\n", " X \n", " O X \n", "X O O \n", "on move: O\n", " X X \n", " O X \n", "X O O \n", "on move: X\n", " X X \n", "O O X \n", "X O O \n", "on move: O\n", "X X X \n", "O O X \n", "X O O \n", "Episode 13, Total Reward: 1\n", "Average Reward: 0.38461538461538464\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " \n", "X O \n", " \n", "on move: O\n", " \n", "X O \n", " X \n", "on move: X\n", " \n", "X O \n", " O X \n", "on move: O\n", " X \n", "X O \n", " O X \n", "on move: X\n", "O X \n", "X O \n", " O X \n", "on move: O\n", "O X \n", "X X O \n", " O X \n", "on move: X\n", "O X \n", "X X O \n", "O O X \n", "on move: O\n", "O X X \n", "X X O \n", "O O X \n", "Episode 14, Total Reward: 0\n", "Average Reward: 0.35714285714285715\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " O \n", " X \n", "on move: O\n", " X \n", " O \n", " X \n", "on move: X\n", " X \n", "O O \n", " X \n", "on move: O\n", " X \n", "O O \n", " X X \n", "on move: X\n", " O X \n", "O O \n", " X X \n", "on move: O\n", " O X \n", "O X O \n", " X X \n", "on move: X\n", " O X \n", "O X O \n", "O X X \n", "on move: O\n", "X O X \n", "O X O \n", "O X X \n", "Episode 15, Total Reward: 1\n", "Average Reward: 0.4\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " \n", "X O \n", "on move: O\n", " X \n", " \n", "X O \n", "on move: X\n", " X O \n", " \n", "X O \n", "on move: O\n", " X O \n", " \n", "X X O \n", "on move: X\n", "O X O \n", " \n", "X X O \n", "on move: O\n", "O X O \n", "X \n", "X X O \n", "on move: X\n", "O X O \n", "X O \n", "X X O \n", "Episode 16, Total Reward: -1\n", "Average Reward: 0.3125\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", " O \n", "on move: O\n", " X \n", " X \n", " O \n", "on move: X\n", " X \n", " O X \n", " O \n", "on move: O\n", "X X \n", " O X \n", " O \n", "on move: X\n", "X X \n", "O O X \n", " O \n", "on move: O\n", "X X X \n", "O O X \n", " O \n", "Episode 17, Total Reward: 1\n", "Average Reward: 0.35294117647058826\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", "O X \n", " \n", "on move: O\n", " \n", "O X \n", " X \n", "on move: X\n", " O \n", "O X \n", " X \n", "on move: O\n", "X O \n", "O X \n", " X \n", "Episode 18, Total Reward: 1\n", "Average Reward: 0.3888888888888889\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", " O X \n", "on move: O\n", "X \n", " \n", " O X \n", "on move: X\n", "X \n", " O \n", " O X \n", "on move: O\n", "X X \n", " O \n", " O X \n", "on move: X\n", "X O X \n", " O \n", " O X \n", "Episode 19, Total Reward: -1\n", "Average Reward: 0.3157894736842105\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", "O \n", " X \n", "on move: O\n", " \n", "O \n", "X X \n", "on move: X\n", " O \n", "O \n", "X X \n", "on move: O\n", " O \n", "O X \n", "X X \n", "on move: X\n", " O \n", "O X \n", "X X O \n", "on move: O\n", " X O \n", "O X \n", "X X O \n", "Episode 20, Total Reward: 1\n", "Average Reward: 0.35\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", "O \n", "on move: O\n", "X X \n", " \n", "O \n", "on move: X\n", "X X \n", " \n", "O O \n", "on move: O\n", "X X \n", " X \n", "O O \n", "on move: X\n", "X O X \n", " X \n", "O O \n", "on move: O\n", "X O X \n", "X X \n", "O O \n", "on move: X\n", "X O X \n", "X O X \n", "O O \n", "Episode 21, Total Reward: -1\n", "Average Reward: 0.2857142857142857\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " O \n", "X \n", "on move: O\n", "X \n", " O \n", "X \n", "on move: X\n", "X \n", " O O \n", "X \n", "on move: O\n", "X X \n", " O O \n", "X \n", "on move: X\n", "X X \n", " O O \n", "X O \n", "on move: O\n", "X X \n", "X O O \n", "X O \n", "Episode 22, Total Reward: 1\n", "Average Reward: 0.3181818181818182\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", "O \n", " \n", "on move: O\n", " X \n", "O \n", "X \n", "on move: X\n", " X O \n", "O \n", "X \n", "on move: O\n", "X X O \n", "O \n", "X \n", "on move: X\n", "X X O \n", "O \n", "X O \n", "on move: O\n", "X X O \n", "O X \n", "X O \n", "on move: X\n", "X X O \n", "O O X \n", "X O \n", "on move: O\n", "X X O \n", "O O X \n", "X X O \n", "Episode 23, Total Reward: 0\n", "Average Reward: 0.30434782608695654\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", " O \n", "on move: O\n", "X \n", " X \n", " O \n", "on move: X\n", "X \n", " X \n", " O O \n", "on move: O\n", "X X \n", " X \n", " O O \n", "on move: X\n", "X X \n", " X \n", "O O O \n", "Episode 24, Total Reward: -1\n", "Average Reward: 0.25\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " \n", "X O \n", "on move: O\n", " \n", " X \n", "X O \n", "on move: X\n", " O \n", " X \n", "X O \n", "on move: O\n", " O \n", "X X \n", "X O \n", "on move: X\n", "O O \n", "X X \n", "X O \n", "on move: O\n", "O O \n", "X X X \n", "X O \n", "Episode 25, Total Reward: 1\n", "Average Reward: 0.28\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " \n", "X O \n", "on move: O\n", " \n", " \n", "X O X \n", "on move: X\n", " \n", " O \n", "X O X \n", "on move: O\n", " X \n", " O \n", "X O X \n", "on move: X\n", " X \n", " O O \n", "X O X \n", "on move: O\n", " X X \n", " O O \n", "X O X \n", "on move: X\n", "O X X \n", " O O \n", "X O X \n", "on move: O\n", "O X X \n", "X O O \n", "X O X \n", "Episode 26, Total Reward: 0\n", "Average Reward: 0.2692307692307692\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", "O \n", "on move: O\n", " \n", "X X \n", "O \n", "on move: X\n", " \n", "X X \n", "O O \n", "on move: O\n", " \n", "X X \n", "O O X \n", "on move: X\n", " O \n", "X X \n", "O O X \n", "on move: O\n", "X O \n", "X X \n", "O O X \n", "Episode 27, Total Reward: 1\n", "Average Reward: 0.2962962962962963\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", "O X \n", "on move: O\n", " X \n", " \n", "O X \n", "on move: X\n", " X O \n", " \n", "O X \n", "on move: O\n", " X O \n", " X \n", "O X \n", "on move: X\n", " X O \n", " X \n", "O O X \n", "on move: O\n", " X O \n", "X X \n", "O O X \n", "on move: X\n", "O X O \n", "X X \n", "O O X \n", "on move: O\n", "O X O \n", "X X X \n", "O O X \n", "Episode 28, Total Reward: 1\n", "Average Reward: 0.32142857142857145\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " \n", " X X \n", "on move: X\n", " O O \n", " \n", " X X \n", "on move: O\n", "X O O \n", " \n", " X X \n", "on move: X\n", "X O O \n", " O \n", " X X \n", "on move: O\n", "X O O \n", " O X \n", " X X \n", "on move: X\n", "X O O \n", " O X \n", "O X X \n", "Episode 29, Total Reward: -1\n", "Average Reward: 0.27586206896551724\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " O \n", " X X \n", " \n", "on move: X\n", " O \n", "O X X \n", " \n", "on move: O\n", " O \n", "O X X \n", " X \n", "on move: X\n", "O O \n", "O X X \n", " X \n", "on move: O\n", "O O \n", "O X X \n", "X X \n", "on move: X\n", "O O O \n", "O X X \n", "X X \n", "Episode 30, Total Reward: -1\n", "Average Reward: 0.23333333333333334\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", " O \n", "on move: O\n", " \n", " X \n", "X O \n", "on move: X\n", " O \n", " X \n", "X O \n", "on move: O\n", " O \n", "X X \n", "X O \n", "on move: X\n", " O \n", "X X \n", "X O O \n", "on move: O\n", "X O \n", "X X \n", "X O O \n", "Episode 31, Total Reward: 1\n", "Average Reward: 0.25806451612903225\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X O \n", " \n", " \n", "on move: O\n", "X O \n", " \n", "X \n", "on move: X\n", "X O O \n", " \n", "X \n", "on move: O\n", "X O O \n", " X \n", "X \n", "on move: X\n", "X O O \n", " O X \n", "X \n", "on move: O\n", "X O O \n", " O X \n", "X X \n", "on move: X\n", "X O O \n", "O O X \n", "X X \n", "on move: O\n", "X O O \n", "O O X \n", "X X X \n", "Episode 32, Total Reward: 1\n", "Average Reward: 0.28125\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", "O \n", "X \n", " \n", "on move: O\n", "O \n", "X \n", "X \n", "on move: X\n", "O \n", "X O \n", "X \n", "on move: O\n", "O \n", "X O \n", "X X \n", "on move: X\n", "O O \n", "X O \n", "X X \n", "on move: O\n", "O X O \n", "X O \n", "X X \n", "on move: X\n", "O X O \n", "X O \n", "X X O \n", "Episode 33, Total Reward: -1\n", "Average Reward: 0.24242424242424243\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " X \n", " X \n", "on move: X\n", " O \n", " X O \n", " X \n", "on move: O\n", " O X \n", " X O \n", " X \n", "on move: X\n", " O X \n", "O X O \n", " X \n", "on move: O\n", " O X \n", "O X O \n", "X X \n", "Episode 34, Total Reward: 1\n", "Average Reward: 0.2647058823529412\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " O \n", " \n", "on move: O\n", " X \n", " O \n", " X \n", "on move: X\n", " X \n", "O O \n", " X \n", "on move: O\n", " X X \n", "O O \n", " X \n", "on move: X\n", " X X \n", "O O \n", " O X \n", "on move: O\n", "X X X \n", "O O \n", " O X \n", "Episode 35, Total Reward: 1\n", "Average Reward: 0.2857142857142857\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " O \n", " \n", "on move: O\n", " X X \n", " O \n", " \n", "on move: X\n", " X X \n", " O O \n", " \n", "on move: O\n", " X X \n", "X O O \n", " \n", "on move: X\n", "O X X \n", "X O O \n", " \n", "on move: O\n", "O X X \n", "X O O \n", " X \n", "on move: X\n", "O X X \n", "X O O \n", " X O \n", "Episode 36, Total Reward: -1\n", "Average Reward: 0.25\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " X \n", " O \n", "on move: O\n", " \n", " X \n", "X O \n", "on move: X\n", " \n", " X O \n", "X O \n", "on move: O\n", "X \n", " X O \n", "X O \n", "on move: X\n", "X \n", "O X O \n", "X O \n", "on move: O\n", "X X \n", "O X O \n", "X O \n", "Episode 37, Total Reward: 1\n", "Average Reward: 0.2702702702702703\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " O X \n", " X \n", " \n", "on move: X\n", " O X \n", " O X \n", " \n", "on move: O\n", " O X \n", " O X \n", " X \n", "on move: X\n", " O X \n", " O X \n", "O X \n", "on move: O\n", " O X \n", " O X \n", "O X X \n", "Episode 38, Total Reward: 1\n", "Average Reward: 0.2894736842105263\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", "O X \n", " \n", "on move: O\n", "X \n", "O X \n", " \n", "on move: X\n", "X \n", "O X \n", "O \n", "on move: O\n", "X X \n", "O X \n", "O \n", "on move: X\n", "X X \n", "O X \n", "O O \n", "on move: O\n", "X X \n", "O X \n", "O X O \n", "on move: X\n", "X X \n", "O O X \n", "O X O \n", "on move: O\n", "X X X \n", "O O X \n", "O X O \n", "Episode 39, Total Reward: 1\n", "Average Reward: 0.3076923076923077\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", "O \n", " \n", "on move: O\n", " X \n", "O \n", " X \n", "on move: X\n", " O X \n", "O \n", " X \n", "on move: O\n", " O X \n", "O \n", " X X \n", "on move: X\n", "O O X \n", "O \n", " X X \n", "on move: O\n", "O O X \n", "O X \n", " X X \n", "Episode 40, Total Reward: 1\n", "Average Reward: 0.325\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " O \n", " X \n", "on move: O\n", " \n", "X O \n", " X \n", "on move: X\n", " \n", "X O \n", " O X \n", "on move: O\n", " \n", "X O \n", "X O X \n", "on move: X\n", "O \n", "X O \n", "X O X \n", "on move: O\n", "O X \n", "X O \n", "X O X \n", "on move: X\n", "O X \n", "X O O \n", "X O X \n", "on move: O\n", "O X X \n", "X O O \n", "X O X \n", "Episode 41, Total Reward: 0\n", "Average Reward: 0.3170731707317073\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " X O \n", " X \n", " \n", "on move: X\n", " X O \n", " X \n", "O \n", "on move: O\n", " X O \n", " X X \n", "O \n", "on move: X\n", " X O \n", "O X X \n", "O \n", "on move: O\n", " X O \n", "O X X \n", "O X \n", "Episode 42, Total Reward: 1\n", "Average Reward: 0.3333333333333333\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " O \n", " \n", "X \n", "on move: O\n", " X O \n", " \n", "X \n", "on move: X\n", " X O \n", " \n", "X O \n", "on move: O\n", "X X O \n", " \n", "X O \n", "on move: X\n", "X X O \n", " O \n", "X O \n", "Episode 43, Total Reward: -1\n", "Average Reward: 0.3023255813953488\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", "X \n", " X \n", "on move: X\n", "O O \n", "X \n", " X \n", "on move: O\n", "O O \n", "X \n", " X X \n", "on move: X\n", "O O \n", "X O \n", " X X \n", "on move: O\n", "O O X \n", "X O \n", " X X \n", "on move: X\n", "O O X \n", "X O \n", "O X X \n", "on move: O\n", "O O X \n", "X O X \n", "O X X \n", "Episode 44, Total Reward: 1\n", "Average Reward: 0.3181818181818182\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " X \n", " X \n", "on move: X\n", " O O \n", " X \n", " X \n", "on move: O\n", " O O \n", " X \n", " X X \n", "on move: X\n", " O O \n", " O X \n", " X X \n", "on move: O\n", "X O O \n", " O X \n", " X X \n", "on move: X\n", "X O O \n", "O O X \n", " X X \n", "on move: O\n", "X O O \n", "O O X \n", "X X X \n", "Episode 45, Total Reward: 1\n", "Average Reward: 0.3333333333333333\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", " O \n", "on move: O\n", "X X \n", " \n", " O \n", "on move: X\n", "X X \n", "O \n", " O \n", "on move: O\n", "X X X \n", "O \n", " O \n", "Episode 46, Total Reward: 1\n", "Average Reward: 0.34782608695652173\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", "O \n", " \n", " X \n", "on move: O\n", "O \n", " X \n", " X \n", "on move: X\n", "O O \n", " X \n", " X \n", "on move: O\n", "O O \n", " X \n", "X X \n", "on move: X\n", "O O \n", " X O \n", "X X \n", "on move: O\n", "O O \n", " X O \n", "X X X \n", "Episode 47, Total Reward: 1\n", "Average Reward: 0.3617021276595745\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", "O \n", " \n", " X \n", "on move: O\n", "O \n", " X \n", " X \n", "on move: X\n", "O \n", "O X \n", " X \n", "on move: O\n", "O \n", "O X \n", " X X \n", "on move: X\n", "O \n", "O X \n", "O X X \n", "Episode 48, Total Reward: -1\n", "Average Reward: 0.3333333333333333\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " O \n", " X X \n", " \n", "on move: X\n", " O \n", " X X \n", "O \n", "on move: O\n", " O \n", " X X \n", "O X \n", "on move: X\n", " O \n", "O X X \n", "O X \n", "on move: O\n", " O \n", "O X X \n", "O X X \n", "on move: X\n", "O O \n", "O X X \n", "O X X \n", "Episode 49, Total Reward: -1\n", "Average Reward: 0.30612244897959184\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", "O \n", " X \n", " \n", "on move: O\n", "O \n", " X \n", " X \n", "on move: X\n", "O O \n", " X \n", " X \n", "on move: O\n", "O O \n", " X \n", " X X \n", "on move: X\n", "O O \n", " X \n", "O X X \n", "on move: O\n", "O O \n", "X X \n", "O X X \n", "on move: X\n", "O O O \n", "X X \n", "O X X \n", "Episode 50, Total Reward: -1\n", "Average Reward: 0.28\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " X \n", " X \n", "on move: X\n", " O \n", " X \n", "O X \n", "on move: O\n", " O \n", " X X \n", "O X \n", "on move: X\n", " O \n", "O X X \n", "O X \n", "on move: O\n", " O X \n", "O X X \n", "O X \n", "on move: X\n", "O O X \n", "O X X \n", "O X \n", "Episode 51, Total Reward: -1\n", "Average Reward: 0.2549019607843137\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", "O X \n", " \n", " \n", "on move: O\n", "O X \n", " X \n", " \n", "on move: X\n", "O X \n", " O X \n", " \n", "on move: O\n", "O X \n", " O X \n", " X \n", "on move: X\n", "O X O \n", " O X \n", " X \n", "on move: O\n", "O X O \n", "X O X \n", " X \n", "on move: X\n", "O X O \n", "X O X \n", "O X \n", "Episode 52, Total Reward: -1\n", "Average Reward: 0.23076923076923078\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " O X \n", " \n", " \n", "on move: O\n", "X O X \n", " \n", " \n", "on move: X\n", "X O X \n", " \n", " O \n", "on move: O\n", "X O X \n", " X \n", " O \n", "on move: X\n", "X O X \n", " X \n", " O O \n", "on move: O\n", "X O X \n", " X X \n", " O O \n", "on move: X\n", "X O X \n", " X X \n", "O O O \n", "Episode 53, Total Reward: -1\n", "Average Reward: 0.20754716981132076\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", "O \n", "on move: O\n", "X \n", " \n", "O X \n", "on move: X\n", "X \n", "O \n", "O X \n", "on move: O\n", "X X \n", "O \n", "O X \n", "on move: X\n", "X X \n", "O O \n", "O X \n", "on move: O\n", "X X \n", "O O X \n", "O X \n", "on move: X\n", "X X O \n", "O O X \n", "O X \n", "Episode 54, Total Reward: -1\n", "Average Reward: 0.18518518518518517\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " O X \n", " \n", "on move: O\n", " X \n", " O X \n", " \n", "on move: X\n", " X \n", " O X \n", "O \n", "on move: O\n", " X \n", " O X \n", "O X \n", "on move: X\n", "O X \n", " O X \n", "O X \n", "on move: O\n", "O X \n", "X O X \n", "O X \n", "on move: X\n", "O X O \n", "X O X \n", "O X \n", "Episode 55, Total Reward: -1\n", "Average Reward: 0.16363636363636364\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X O \n", " \n", " \n", "on move: O\n", " X O \n", " \n", "X \n", "on move: X\n", " X O \n", " O \n", "X \n", "on move: O\n", " X O \n", " O \n", "X X \n", "on move: X\n", " X O \n", " O \n", "X X O \n", "on move: O\n", " X O \n", " O X \n", "X X O \n", "on move: X\n", " X O \n", "O O X \n", "X X O \n", "on move: O\n", "X X O \n", "O O X \n", "X X O \n", "Episode 56, Total Reward: 0\n", "Average Reward: 0.16071428571428573\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", "O \n", "on move: O\n", " X \n", "X \n", "O \n", "on move: X\n", " X \n", "X O \n", "O \n", "on move: O\n", "X X \n", "X O \n", "O \n", "on move: X\n", "X O X \n", "X O \n", "O \n", "on move: O\n", "X O X \n", "X O \n", "O X \n", "on move: X\n", "X O X \n", "X O O \n", "O X \n", "on move: O\n", "X O X \n", "X O O \n", "O X X \n", "Episode 57, Total Reward: 0\n", "Average Reward: 0.15789473684210525\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " O \n", "X \n", "on move: O\n", " X \n", " O \n", "X \n", "on move: X\n", " X \n", "O O \n", "X \n", "on move: O\n", "X X \n", "O O \n", "X \n", "on move: X\n", "X O X \n", "O O \n", "X \n", "on move: O\n", "X O X \n", "O O \n", "X X \n", "on move: X\n", "X O X \n", "O O \n", "X O X \n", "on move: O\n", "X O X \n", "O X O \n", "X O X \n", "Episode 58, Total Reward: 1\n", "Average Reward: 0.1724137931034483\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " O X \n", " \n", "on move: O\n", " \n", " O X \n", " X \n", "on move: X\n", "O \n", " O X \n", " X \n", "on move: O\n", "O \n", " O X \n", " X X \n", "on move: X\n", "O O \n", " O X \n", " X X \n", "on move: O\n", "O O \n", " O X \n", "X X X \n", "Episode 59, Total Reward: 1\n", "Average Reward: 0.1864406779661017\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", "X \n", " X \n", "on move: X\n", " O \n", "X \n", " X O \n", "on move: O\n", " O \n", "X X \n", " X O \n", "on move: X\n", " O \n", "X X \n", "O X O \n", "on move: O\n", " O \n", "X X X \n", "O X O \n", "Episode 60, Total Reward: 1\n", "Average Reward: 0.2\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", " O \n", "on move: O\n", "X \n", " \n", "X O \n", "on move: X\n", "X \n", " O \n", "X O \n", "on move: O\n", "X \n", " O \n", "X X O \n", "on move: X\n", "X O \n", " O \n", "X X O \n", "on move: O\n", "X O \n", "X O \n", "X X O \n", "Episode 61, Total Reward: 1\n", "Average Reward: 0.21311475409836064\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " O \n", "X \n", "on move: O\n", "X \n", " O \n", "X \n", "on move: X\n", "X \n", " O \n", "X O \n", "on move: O\n", "X \n", "X O \n", "X O \n", "Episode 62, Total Reward: 1\n", "Average Reward: 0.22580645161290322\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " O \n", " \n", "on move: O\n", "X \n", " O \n", " X \n", "on move: X\n", "X \n", "O O \n", " X \n", "on move: O\n", "X \n", "O O \n", "X X \n", "on move: X\n", "X \n", "O O \n", "X X O \n", "on move: O\n", "X X \n", "O O \n", "X X O \n", "on move: X\n", "X X \n", "O O O \n", "X X O \n", "Episode 63, Total Reward: -1\n", "Average Reward: 0.20634920634920634\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " O \n", "X \n", " \n", "on move: O\n", "X O \n", "X \n", " \n", "on move: X\n", "X O \n", "X \n", " O \n", "on move: O\n", "X O \n", "X X \n", " O \n", "on move: X\n", "X O \n", "X X \n", "O O \n", "on move: O\n", "X O \n", "X X \n", "O X O \n", "on move: X\n", "X O O \n", "X X \n", "O X O \n", "on move: O\n", "X O O \n", "X X X \n", "O X O \n", "Episode 64, Total Reward: 1\n", "Average Reward: 0.21875\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " O \n", " \n", "X \n", "on move: O\n", "X O \n", " \n", "X \n", "on move: X\n", "X O \n", "O \n", "X \n", "on move: O\n", "X O \n", "O \n", "X X \n", "on move: X\n", "X O O \n", "O \n", "X X \n", "on move: O\n", "X O O \n", "O X \n", "X X \n", "on move: X\n", "X O O \n", "O X \n", "X X O \n", "on move: O\n", "X O O \n", "O X X \n", "X X O \n", "Episode 65, Total Reward: 0\n", "Average Reward: 0.2153846153846154\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " \n", "X X \n", "on move: X\n", "O O \n", " \n", "X X \n", "on move: O\n", "O O \n", " X \n", "X X \n", "on move: X\n", "O O O \n", " X \n", "X X \n", "Episode 66, Total Reward: -1\n", "Average Reward: 0.19696969696969696\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", " O X \n", " \n", "on move: O\n", " X \n", " O X \n", " \n", "on move: X\n", " X \n", " O X \n", "O \n", "on move: O\n", " X X \n", " O X \n", "O \n", "on move: X\n", " X X \n", "O O X \n", "O \n", "on move: O\n", " X X \n", "O O X \n", "O X \n", "on move: X\n", "O X X \n", "O O X \n", "O X \n", "Episode 67, Total Reward: -1\n", "Average Reward: 0.1791044776119403\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", " X \n", " \n", " O X \n", "on move: X\n", " X \n", " O \n", " O X \n", "on move: O\n", "X X \n", " O \n", " O X \n", "on move: X\n", "X X \n", " O \n", "O O X \n", "on move: O\n", "X X \n", " X O \n", "O O X \n", "Episode 68, Total Reward: 1\n", "Average Reward: 0.19117647058823528\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", "O X \n", "on move: O\n", " X \n", " \n", "O X \n", "on move: X\n", " O X \n", " \n", "O X \n", "on move: O\n", " O X \n", " \n", "O X X \n", "on move: X\n", " O X \n", " O \n", "O X X \n", "on move: O\n", " O X \n", "X O \n", "O X X \n", "on move: X\n", "O O X \n", "X O \n", "O X X \n", "on move: O\n", "O O X \n", "X X O \n", "O X X \n", "Episode 69, Total Reward: 0\n", "Average Reward: 0.18840579710144928\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", "O \n", " \n", "on move: O\n", " X \n", "O \n", " X \n", "on move: X\n", "O X \n", "O \n", " X \n", "on move: O\n", "O X \n", "O X \n", " X \n", "on move: X\n", "O O X \n", "O X \n", " X \n", "on move: O\n", "O O X \n", "O X \n", "X X \n", "on move: X\n", "O O X \n", "O X \n", "X X O \n", "on move: O\n", "O O X \n", "O X X \n", "X X O \n", "Episode 70, Total Reward: 1\n", "Average Reward: 0.2\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", "O \n", " \n", "on move: O\n", " X \n", "O \n", " X \n", "on move: X\n", " X \n", "O O \n", " X \n", "on move: O\n", " X \n", "O O X \n", " X \n", "on move: X\n", " X O \n", "O O X \n", " X \n", "on move: O\n", " X O \n", "O O X \n", " X X \n", "on move: X\n", " X O \n", "O O X \n", "O X X \n", "Episode 71, Total Reward: -1\n", "Average Reward: 0.18309859154929578\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " O \n", " X \n", " \n", "on move: O\n", " O \n", " X \n", "X \n", "on move: X\n", "O O \n", " X \n", "X \n", "on move: O\n", "O O \n", " X \n", "X X \n", "on move: X\n", "O O O \n", " X \n", "X X \n", "Episode 72, Total Reward: -1\n", "Average Reward: 0.16666666666666666\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X O \n", " \n", " \n", "on move: O\n", "X O \n", " \n", "X \n", "on move: X\n", "X O \n", " O \n", "X \n", "on move: O\n", "X O \n", " X O \n", "X \n", "on move: X\n", "X O \n", " X O \n", "X O \n", "on move: O\n", "X O \n", "X X O \n", "X O \n", "Episode 73, Total Reward: 1\n", "Average Reward: 0.1780821917808219\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", "O \n", "X \n", "on move: O\n", " X \n", "O \n", "X \n", "on move: X\n", " X O \n", "O \n", "X \n", "on move: O\n", "X X O \n", "O \n", "X \n", "on move: X\n", "X X O \n", "O \n", "X O \n", "on move: O\n", "X X O \n", "O X \n", "X O \n", "on move: X\n", "X X O \n", "O X \n", "X O O \n", "on move: O\n", "X X O \n", "O X X \n", "X O O \n", "Episode 74, Total Reward: 0\n", "Average Reward: 0.17567567567567569\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", " O \n", "on move: O\n", "X \n", " X \n", " O \n", "on move: X\n", "X \n", " O X \n", " O \n", "on move: O\n", "X X \n", " O X \n", " O \n", "on move: X\n", "X X \n", "O O X \n", " O \n", "on move: O\n", "X X X \n", "O O X \n", " O \n", "Episode 75, Total Reward: 1\n", "Average Reward: 0.18666666666666668\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " \n", " \n", "X O \n", "on move: O\n", " X \n", " \n", "X O \n", "on move: X\n", " X \n", " \n", "X O O \n", "on move: O\n", "X X \n", " \n", "X O O \n", "on move: X\n", "X X \n", " O \n", "X O O \n", "on move: O\n", "X X X \n", " O \n", "X O O \n", "Episode 76, Total Reward: 1\n", "Average Reward: 0.19736842105263158\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", "O X \n", " \n", " \n", "on move: O\n", "O X \n", " X \n", " \n", "on move: X\n", "O X \n", " X \n", " O \n", "on move: O\n", "O X \n", "X X \n", " O \n", "on move: X\n", "O X \n", "X X \n", " O O \n", "on move: O\n", "O X X \n", "X X \n", " O O \n", "on move: X\n", "O X X \n", "X X \n", "O O O \n", "Episode 77, Total Reward: -1\n", "Average Reward: 0.18181818181818182\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " O \n", "X \n", " \n", "on move: O\n", " O \n", "X \n", " X \n", "on move: X\n", " O \n", "X \n", " O X \n", "on move: O\n", " O \n", "X X \n", " O X \n", "on move: X\n", " O \n", "X X \n", "O O X \n", "on move: O\n", "X O \n", "X X \n", "O O X \n", "on move: X\n", "X O \n", "X O X \n", "O O X \n", "Episode 78, Total Reward: -1\n", "Average Reward: 0.16666666666666666\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", "X X \n", " \n", " O \n", "on move: X\n", "X X O \n", " \n", " O \n", "on move: O\n", "X X O \n", " X \n", " O \n", "on move: X\n", "X X O \n", " X \n", " O O \n", "on move: O\n", "X X O \n", "X X \n", " O O \n", "on move: X\n", "X X O \n", "X X O \n", " O O \n", "Episode 79, Total Reward: -1\n", "Average Reward: 0.1518987341772152\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " \n", " O \n", "on move: O\n", "X X \n", " \n", " O \n", "on move: X\n", "X X \n", " O \n", " O \n", "on move: O\n", "X X X \n", " O \n", " O \n", "Episode 80, Total Reward: 1\n", "Average Reward: 0.1625\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X O \n", " \n", " \n", "on move: O\n", " X O \n", "X \n", " \n", "on move: X\n", " X O \n", "X O \n", " \n", "on move: O\n", " X O \n", "X X O \n", " \n", "on move: X\n", " X O \n", "X X O \n", " O \n", "on move: O\n", " X O \n", "X X O \n", " O X \n", "on move: X\n", "O X O \n", "X X O \n", " O X \n", "on move: O\n", "O X O \n", "X X O \n", "X O X \n", "Episode 81, Total Reward: 0\n", "Average Reward: 0.16049382716049382\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", " \n", "O X \n", " \n", "on move: O\n", "X \n", "O X \n", " \n", "on move: X\n", "X \n", "O X \n", " O \n", "on move: O\n", "X \n", "O X \n", " X O \n", "on move: X\n", "X \n", "O X \n", "O X O \n", "on move: O\n", "X X \n", "O X \n", "O X O \n", "on move: X\n", "X X \n", "O X O \n", "O X O \n", "on move: O\n", "X X X \n", "O X O \n", "O X O \n", "Episode 82, Total Reward: 1\n", "Average Reward: 0.17073170731707318\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", " X \n", " \n", "X O \n", "on move: X\n", " X \n", " \n", "X O O \n", "on move: O\n", "X X \n", " \n", "X O O \n", "on move: X\n", "X X \n", " O \n", "X O O \n", "on move: O\n", "X X \n", " O X \n", "X O O \n", "on move: X\n", "X X O \n", " O X \n", "X O O \n", "on move: O\n", "X X O \n", "X O X \n", "X O O \n", "Episode 83, Total Reward: 1\n", "Average Reward: 0.18072289156626506\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X O \n", " \n", " \n", "on move: O\n", "X O \n", " X \n", " \n", "on move: X\n", "X O O \n", " X \n", " \n", "on move: O\n", "X O O \n", " X \n", " X \n", "on move: X\n", "X O O \n", " X \n", " O X \n", "on move: O\n", "X O O \n", "X X \n", " O X \n", "on move: X\n", "X O O \n", "X X \n", "O O X \n", "on move: O\n", "X O O \n", "X X X \n", "O O X \n", "Episode 84, Total Reward: 1\n", "Average Reward: 0.19047619047619047\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", "O X \n", "on move: O\n", " X \n", " \n", "O X \n", "on move: X\n", " X \n", " \n", "O O X \n", "on move: O\n", " X \n", "X \n", "O O X \n", "on move: X\n", " X \n", "X O \n", "O O X \n", "on move: O\n", " X \n", "X O X \n", "O O X \n", "Episode 85, Total Reward: 1\n", "Average Reward: 0.2\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", " O \n", " \n", "on move: O\n", "X X \n", " O \n", " \n", "on move: X\n", "X X \n", " O \n", "O \n", "on move: O\n", "X X \n", " X O \n", "O \n", "on move: X\n", "X X \n", " X O \n", "O O \n", "on move: O\n", "X X X \n", " X O \n", "O O \n", "Episode 86, Total Reward: 1\n", "Average Reward: 0.20930232558139536\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X O \n", " \n", " \n", "on move: O\n", "X O \n", " \n", " X \n", "on move: X\n", "X O O \n", " \n", " X \n", "on move: O\n", "X O O \n", " \n", " X X \n", "on move: X\n", "X O O \n", " \n", "O X X \n", "on move: O\n", "X O O \n", "X \n", "O X X \n", "on move: X\n", "X O O \n", "X O \n", "O X X \n", "on move: O\n", "X O O \n", "X X O \n", "O X X \n", "Episode 87, Total Reward: 1\n", "Average Reward: 0.21839080459770116\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", " X \n", " \n", " X O \n", "on move: X\n", " X \n", " \n", "O X O \n", "on move: O\n", " X X \n", " \n", "O X O \n", "on move: X\n", " X X \n", "O \n", "O X O \n", "on move: O\n", " X X \n", "O X \n", "O X O \n", "Episode 88, Total Reward: 1\n", "Average Reward: 0.22727272727272727\n", "on move: O\n", " \n", " \n", "X \n", "on move: X\n", " O \n", " \n", "X \n", "on move: O\n", " O \n", "X \n", "X \n", "on move: X\n", " O \n", "X O \n", "X \n", "on move: O\n", " O \n", "X O \n", "X X \n", "on move: X\n", " O O \n", "X O \n", "X X \n", "on move: O\n", " O O \n", "X O X \n", "X X \n", "on move: X\n", "O O O \n", "X O X \n", "X X \n", "Episode 89, Total Reward: -1\n", "Average Reward: 0.21348314606741572\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", "O \n", " \n", "on move: O\n", " X X \n", "O \n", " \n", "on move: X\n", " X X \n", "O \n", " O \n", "on move: O\n", "X X X \n", "O \n", " O \n", "Episode 90, Total Reward: 1\n", "Average Reward: 0.2222222222222222\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", "X \n", " X \n", "on move: X\n", " O \n", "X \n", "O X \n", "on move: O\n", " O \n", "X X \n", "O X \n", "on move: X\n", " O \n", "X O X \n", "O X \n", "on move: O\n", " O \n", "X O X \n", "O X X \n", "on move: X\n", " O O \n", "X O X \n", "O X X \n", "Episode 91, Total Reward: -1\n", "Average Reward: 0.2087912087912088\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " O \n", " \n", "on move: O\n", " X \n", "X O \n", " \n", "on move: X\n", " X O \n", "X O \n", " \n", "on move: O\n", " X O \n", "X O \n", " X \n", "on move: X\n", "O X O \n", "X O \n", " X \n", "on move: O\n", "O X O \n", "X O \n", "X X \n", "on move: X\n", "O X O \n", "X O \n", "X X O \n", "Episode 92, Total Reward: -1\n", "Average Reward: 0.1956521739130435\n", "on move: O\n", "X \n", " \n", " \n", "on move: X\n", "X \n", "O \n", " \n", "on move: O\n", "X \n", "O X \n", " \n", "on move: X\n", "X O \n", "O X \n", " \n", "on move: O\n", "X O \n", "O X \n", "X \n", "on move: X\n", "X O O \n", "O X \n", "X \n", "on move: O\n", "X O O \n", "O X \n", "X X \n", "on move: X\n", "X O O \n", "O O X \n", "X X \n", "on move: O\n", "X O O \n", "O O X \n", "X X X \n", "Episode 93, Total Reward: 1\n", "Average Reward: 0.20430107526881722\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " O \n", " X \n", "on move: O\n", " X \n", " O \n", " X \n", "on move: X\n", "O X \n", " O \n", " X \n", "on move: O\n", "O X \n", " O \n", " X X \n", "on move: X\n", "O X \n", " O \n", "O X X \n", "on move: O\n", "O X \n", " X O \n", "O X X \n", "Episode 94, Total Reward: 1\n", "Average Reward: 0.2127659574468085\n", "on move: O\n", " \n", "X \n", " \n", "on move: X\n", " O \n", "X \n", " \n", "on move: O\n", " O \n", "X \n", "X \n", "on move: X\n", " O \n", "X \n", "X O \n", "on move: O\n", " O \n", "X X \n", "X O \n", "on move: X\n", " O \n", "X X \n", "X O O \n", "on move: O\n", "X O \n", "X X \n", "X O O \n", "Episode 95, Total Reward: 1\n", "Average Reward: 0.22105263157894736\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", " X O \n", "on move: O\n", " X \n", " \n", " X O \n", "on move: X\n", " X \n", "O \n", " X O \n", "on move: O\n", " X \n", "O \n", "X X O \n", "on move: X\n", " X O \n", "O \n", "X X O \n", "on move: O\n", " X O \n", "O X \n", "X X O \n", "Episode 96, Total Reward: 1\n", "Average Reward: 0.22916666666666666\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " \n", " \n", "O X \n", "on move: O\n", "X \n", " \n", "O X \n", "on move: X\n", "X \n", "O \n", "O X \n", "on move: O\n", "X \n", "O X \n", "O X \n", "on move: X\n", "X \n", "O O X \n", "O X \n", "on move: O\n", "X X \n", "O O X \n", "O X \n", "on move: X\n", "X X \n", "O O X \n", "O O X \n", "on move: O\n", "X X X \n", "O O X \n", "O O X \n", "Episode 97, Total Reward: 1\n", "Average Reward: 0.23711340206185566\n", "on move: O\n", " \n", " X \n", " \n", "on move: X\n", "O \n", " X \n", " \n", "on move: O\n", "O X \n", " X \n", " \n", "on move: X\n", "O X \n", " X \n", " O \n", "on move: O\n", "O X \n", "X X \n", " O \n", "on move: X\n", "O X \n", "X X \n", "O O \n", "on move: O\n", "O X \n", "X X \n", "O X O \n", "on move: X\n", "O O X \n", "X X \n", "O X O \n", "on move: O\n", "O O X \n", "X X X \n", "O X O \n", "Episode 98, Total Reward: 1\n", "Average Reward: 0.24489795918367346\n", "on move: O\n", " \n", " \n", " X \n", "on move: X\n", " O \n", " \n", " X \n", "on move: O\n", " O \n", " X \n", " X \n", "on move: X\n", " O \n", "O X \n", " X \n", "on move: O\n", " O \n", "O X \n", " X X \n", "on move: X\n", "O O \n", "O X \n", " X X \n", "on move: O\n", "O O \n", "O X \n", "X X X \n", "Episode 99, Total Reward: 1\n", "Average Reward: 0.25252525252525254\n", "on move: O\n", " X \n", " \n", " \n", "on move: X\n", " X \n", " \n", " O \n", "on move: O\n", " X \n", " X \n", " O \n", "on move: X\n", " X \n", " X \n", "O O \n", "on move: O\n", " X X \n", " X \n", "O O \n", "on move: X\n", " X X \n", "O X \n", "O O \n", "on move: O\n", " X X \n", "O X \n", "O O X \n", "Episode 100, Total Reward: 1\n", "Average Reward: 0.26\n" ] } ], "source": [ "# Основной цикл обучения (работа с отдельным классом агента)\n", "\n", "# Создание среды для игры в крестики-нолики\n", "environment = TicTacToeEnv()\n", "\n", "# Создание агента (играющего крестиками)\n", "agent = Agent(symbol=1)\n", "\n", "num_episodes = 100 # Количество эпизодов (игр) для обучения\n", "collected_rewards = [] # Список для хранения наград/побед в каждом эпизоде \n", "\n", "# Переменная для отслеживания символа и текущего игрока\n", "oom = 1\n", "\n", "for i in range(num_episodes):\n", " # Сброс среды и начало нового эпизода\n", " state, _ = environment.reset() \n", "\n", " # Общая награда за эпизод\n", " total_reward = 0\n", "\n", " # Флаг завершения игры\n", " done = False\n", " om = oom \n", "\n", " # Максимум 9 ходов, поскольку поле 3x3 \n", " for j in range(9): \n", " moves = environment.move_generator() \n", "\n", " # Ходов нет, заканчиваем игру\n", " if not moves:\n", " break\n", "\n", " \n", " if len(moves) == 1:\n", " move = moves[0] # Если остался один ход на основе стратегии\n", " else:\n", " move = agent.get_action(moves) # Агент выбирает ход на основе стратегии\n", "\n", " # Выполнение хода и обновление состояния игры\n", " next_state, reward, done, info = environment.step(move)\n", " total_reward += reward\n", " state = next_state\n", "\n", " # Отображаем текущее состояние игры\n", " environment.render()\n", "\n", " if done:\n", " break\n", "\n", " om = -om # Смена игрока\n", "\n", " collected_rewards.append(total_reward)\n", "\n", " print(f\"Episode {i+1}, Total Reward: {total_reward}\")\n", " average_reward = sum(collected_rewards) / len(collected_rewards)\n", " print(f\"Average Reward: {average_reward}\")" ] } ], "metadata": { "kernelspec": { "display_name": "aimenv", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.12.6" } }, "nbformat": 4, "nbformat_minor": 2 }