[player_goes_first]_deep_learning_tic_tac_toe_model_training.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "[player goes first] deep learning tic-tac-toe model training.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyMTo5CeFJxBVz8V3WYBH48a",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/ZackAkil/deep-tic-tac-toe/blob/master/%5Bplayer_goes_first%5D_deep_learning_tic_tac_toe_model_training.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nKBz663XdQhj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kU3ye6JadXQU",
        "colab_type": "code",
        "cellView": "both",
        "colab": {}
      },
      "source": [
        "#@title\n",
        "# rewards\n",
        "WIN = 1\n",
        "LOSE = -2\n",
        "TIE = 0.2\n",
        "INVALID = -3\n",
        "\n",
        "\n",
        "class XandOs():\n",
        "  def __init__(self, log = False):\n",
        "    self.state = np.zeros([3,3], dtype=np.int16)\n",
        "    self.end = False\n",
        "    self.history = []\n",
        "    self.log = log\n",
        "\n",
        "  def make_move(self, pos, player=1):\n",
        "  \n",
        "    if self.state.flatten()[pos]:\n",
        "      # invalid move\n",
        "      if self.log:\n",
        "        print('player {0} INVALID move {1}'.format(player, pos))\n",
        "      return False\n",
        "    else:\n",
        "      # make move\n",
        "      if self.log:\n",
        "        print('player {0} making move {1}'.format(player, pos))\n",
        "      grid = self.state.flatten()\n",
        "      grid[pos] = player\n",
        "      self.state = np.reshape(grid, [3,3])\n",
        "      return True\n",
        "\n",
        "  def check_win(self, player=1):\n",
        "    bool_grid = self.state == player\n",
        "    return win(bool_grid)\n",
        "\n",
        "\n",
        "  def step(self, action):\n",
        "\n",
        "    self.history.append([self.state, action])\n",
        "    reward = 0\n",
        "    pos = action\n",
        "    valid_move = self.make_move(pos, player=1)\n",
        "\n",
        "    if not valid_move:\n",
        "      reward = INVALID\n",
        "      self.end = True\n",
        "\n",
        "    elif self.check_win(player=1):\n",
        "      if self.log:\n",
        "        print('player 1 win')\n",
        "      reward = WIN\n",
        "      self.end = True\n",
        "\n",
        "    elif self.game_full():\n",
        "      if self.log:\n",
        "        print('player 1 ended game')\n",
        "      self.end = True\n",
        "      reward = TIE\n",
        "\n",
        "    if self.end:  \n",
        "      return reward, self.state, self.end\n",
        "\n",
        "    # player 2 move\n",
        "    player_2_move = self.rand_agent_move(player=2)\n",
        "    valid_move = self.make_move(player_2_move, player=2)\n",
        "\n",
        "    if not valid_move:\n",
        "      reward = WIN\n",
        "      self.end = True\n",
        "\n",
        "    elif self.check_win(player=2):\n",
        "      if self.log:\n",
        "        print('player 2 win')\n",
        "      reward = LOSE\n",
        "      self.end = True\n",
        "\n",
        "    elif self.game_full():\n",
        "      if self.log:\n",
        "        print('player 2 ended game')\n",
        "      self.end = True\n",
        "      reward = TIE\n",
        "\n",
        "    return reward, self.state, self.end\n",
        "\n",
        "\n",
        "\n",
        "  def render(self):\n",
        "    render_game(self.state)\n",
        "\n",
        "  def game_full(self):\n",
        "    return (self.state==0).sum() == 0\n",
        "\n",
        "\n",
        "  def rand_agent_move(self, player = 2):\n",
        "\n",
        "    flat_state = self.state.flatten() == 0\n",
        "    empty_indexs, = np.where(flat_state)\n",
        "\n",
        "    for indx in empty_indexs:\n",
        "      grid = self.state.flatten() == player\n",
        "      grid[indx] = True\n",
        "      if win(np.reshape(grid,[3,3])):\n",
        "        if self.log:\n",
        "          print('found winning move ')\n",
        "        return indx\n",
        "    if len(empty_indexs):\n",
        "      return np.random.choice(empty_indexs)\n",
        "    else:\n",
        "      return 0\n",
        "\n",
        "def win(bool_grid):\n",
        "  # horizontal\n",
        "  for row in bool_grid:\n",
        "    if row.sum() == 3:\n",
        "      return True\n",
        "\n",
        "  # vertical\n",
        "  for row in bool_grid.T:\n",
        "    if row.sum() == 3:\n",
        "      return True\n",
        "\n",
        "  # diangonal 1\n",
        "  if np.reshape(bool_grid,[-1])[::4].sum() == 3:\n",
        "    return True\n",
        "\n",
        "  # diangonal 2\n",
        "  if np.reshape(bool_grid,[-1])[2:7:2].sum() == 3:\n",
        "    return True\n",
        "\n",
        "  return False\n",
        "\n",
        "def render_game(state):\n",
        "\n",
        "    # data = np.reshape(pred, [3,3])\n",
        "    fig, ax = plt.subplots()\n",
        "    im = ax.imshow(state/2, cmap=plt.get_cmap('cool'), interpolation='nearest',\n",
        "              vmin=0, vmax=1)\n",
        "    text_num = 0\n",
        "    for i in range(3):\n",
        "      for j in range(3):\n",
        "        ax.text(j, i, str(text_num))\n",
        "        text_num += 1\n",
        "\n",
        "      # Minor ticks\n",
        "    fig.colorbar(im)\n",
        "    plt.show()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P1KUiRucnHqD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "game = XandOs()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H9CbboUHnAz4",
        "colab_type": "code",
        "outputId": "438e72c6-0375-43d5-97ee-3380a34b30e1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        }
      },
      "source": [
        "print(game.step(1))\n",
        "game.render()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(0, array([[0, 1, 0],\n",
            "       [0, 0, 0],\n",
            "       [2, 0, 0]], dtype=int16), False)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nPz1M1ydIVeyEAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XcfHE9Ekj8i0",
        "colab_type": "code",
        "outputId": "259a5713-7260-4723-c43b-aa5ef3b5b0de",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 926
        }
      },
      "source": [
        "game = XandOs()\n",
        "\n",
        "while(not game.end):\n",
        "  move = np.random.choice(range(9))\n",
        "  print(game.step(move))\n",
        "  game.render()\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(0, array([[0, 0, 0],\n",
            "       [1, 0, 0],\n",
            "       [0, 2, 0]], dtype=int16), False)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\naU5yZlZpTnJmVmlOcmZWaf8foRfjVNvEvysAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "(0, array([[0, 0, 1],\n",
            "       [1, 2, 0],\n",
            "       [0, 2, 0]], dtype=int16), False)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "(-3, array([[0, 0, 1],\n",
            "       [1, 2, 0],\n",
            "       [0, 2, 0]], dtype=int16), True)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZItV8JI04SqL",
        "colab_type": "code",
        "outputId": "464387a6-43d6-4798-d7f6-dde8b421fda0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(game.history)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ry3PE3K3o2iY",
        "colab_type": "code",
        "outputId": "15226134-5613-4429-e2c5-30beb197b05e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 80
        }
      },
      "source": [
        "import keras\n",
        "from keras.datasets import mnist\n",
        "from keras.models import Sequential\n",
        "from keras.layers import Dense, Dropout, Flatten\n",
        "from keras.layers import Conv2D, MaxPooling2D, Conv2DTranspose\n",
        "from keras import backend as K"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<p style=\"color: red;\">\n",
              "The default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.<br>\n",
              "We recommend you <a href=\"https://www.tensorflow.org/guide/migrate\" target=\"_blank\">upgrade</a> now \n",
              "or ensure your notebook will continue to use TensorFlow 1.x via the <code>%tensorflow_version 1.x</code> magic:\n",
              "<a href=\"https://colab.research.google.com/notebooks/tensorflow_version.ipynb\" target=\"_blank\">more info</a>.</p>\n"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oB9NMGULt20B",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from collections import deque"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hOF8pY1VcDw0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Agent():\n",
        "\n",
        "  def __init__(self):\n",
        "    self.model = self.create_model()\n",
        "\n",
        "  def create_model(self):\n",
        "    model = Sequential()\n",
        "    # model.add(Flatten(input_shape=[3,3,2]))\n",
        "    model.add(Conv2D(8, kernel_size=(1, 1),\n",
        "                 activation='relu',\n",
        "                 input_shape=[3,3,2],\n",
        "                 strides=(1,1), padding='same'))\n",
        "    model.add(Conv2D(8, (3, 3), activation='relu',padding='same'))\n",
        "    model.add(Flatten())\n",
        "    model.add(Dense(100, activation='relu'))\n",
        "    model.add(Dropout(0.1))\n",
        "    model.add(Dense(100, activation='relu'))\n",
        "    model.add(Dropout(0.1))\n",
        "    model.add(Dense(50, activation='relu'))\n",
        "    model.add(Dense(9, activation='sigmoid'))\n",
        "    model.compile(loss='binary_crossentropy', optimizer='adam')\n",
        "    return model\n",
        "\n",
        "  def show_heat_map(self, pred):\n",
        "    data = np.reshape(pred, [3,3])\n",
        "    fig, ax = plt.subplots()\n",
        "    im = ax.imshow(data, cmap=plt.get_cmap('hot'), interpolation='nearest',\n",
        "              vmin=0, vmax=1)\n",
        "    fig.colorbar(im)\n",
        "    plt.show()\n",
        "\n",
        "  def decide_move(self, env, show=True, only_valid=False):\n",
        "    model_input = np.zeros([1,3,3,2])\n",
        "    model_input[0,:,:,0] = env==1\n",
        "    model_input[0,:,:,1] = env==2\n",
        "    \n",
        "    pred = np.squeeze(self.model.predict(model_input))\n",
        "    # print(pred.shape, (env != 0).flatten().shape)\n",
        "    if only_valid:\n",
        "      pred[(env != 0).flatten()] = 0\n",
        "    \n",
        "    if show:\n",
        "      self.show_heat_map(pred)\n",
        "    return pred.argmax(), pred\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "m97oo1yaq8Fc",
        "colab_type": "text"
      },
      "source": [
        "# The game"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "s8QvSz-4i0OZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.preprocessing import normalize"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zRoDqfUBjet5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def render_all(things,vmin=0, vmax=2, cmap='cool', show_max=True):\n",
        "  display_all = np.zeros([3,3*(len(things))])\n",
        "  for i, thing in enumerate(things):\n",
        "    if show_max:\n",
        "      print('make move: ', thing.argmax())\n",
        "    step = 3*i\n",
        "    display_all[:,step:step+3] = np.reshape(thing, [3,3])\n",
        "  plt.imshow(display_all, vmin=vmin, vmax=vmax, cmap=cmap)\n",
        "  plt.show()\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nfKcoz4Hn2Xx",
        "colab_type": "code",
        "outputId": "a9c1fa62-3aed-4fc0-f6a5-8a22fb034acd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        }
      },
      "source": [
        "game = XandOs()\n",
        "agent = Agent()\n",
        "game.render()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:66: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:541: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:4432: The name tf.random_uniform is deprecated. Please use tf.random.uniform instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:148: The name tf.placeholder_with_default is deprecated. Please use tf.compat.v1.placeholder_with_default instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3733: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/optimizers.py:793: The name tf.train.Optimizer is deprecated. Please use tf.compat.v1.train.Optimizer instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3657: The name tf.log is deprecated. Please use tf.math.log instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/ops/nn_impl.py:183: where (from tensorflow.python.ops.array_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.where in 2.0, which has the same broadcast rule as np.where\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "q-FzFabMxZXw",
        "colab_type": "code",
        "outputId": "1f53d4ad-67f9-4840-ab94-1879dff9f4b2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        }
      },
      "source": [
        "!wget https://github.com/ZackAkil/deep-tic-tac-toe/raw/master/python%20model%20weights/winer_weights.keras \n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-02-03 23:48:48--  https://github.com/ZackAkil/deep-tic-tac-toe/raw/master/python%20model%20weights/winer_weights.keras\n",
            "Resolving github.com (github.com)... 192.30.253.113\n",
            "Connecting to github.com (github.com)|192.30.253.113|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/ZackAkil/deep-tic-tac-toe/master/python%20model%20weights/winer_weights.keras [following]\n",
            "--2020-02-03 23:48:48--  https://raw.githubusercontent.com/ZackAkil/deep-tic-tac-toe/master/python%20model%20weights/winer_weights.keras\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 39168 (38K) [application/octet-stream]\n",
            "Saving to: ‘winer_weights.keras.1’\n",
            "\n",
            "\rwiner_weights.keras   0%[                    ]       0  --.-KB/s               \rwiner_weights.keras 100%[===================>]  38.25K  --.-KB/s    in 0.01s   \n",
            "\n",
            "2020-02-03 23:48:48 (3.16 MB/s) - ‘winer_weights.keras.1’ saved [39168/39168]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kDnWC4ryte6-",
        "colab_type": "code",
        "outputId": "6d9a0831-51d2-496b-9195-cafde5bdd41c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 307
        }
      },
      "source": [
        "agent.model.load_weights('winer_weights.keras')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ValueError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-15-5d46dedabfe5>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0magent\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mload_weights\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'winer_weights.keras'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/engine/saving.py\u001b[0m in \u001b[0;36mload_wrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    456\u001b[0m                 \u001b[0mos\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mremove\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtmp_filepath\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    457\u001b[0m             \u001b[0;32mreturn\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 458\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mload_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    459\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    460\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mload_wrapper\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/engine/network.py\u001b[0m in \u001b[0;36mload_weights\u001b[0;34m(self, filepath, by_name, skip_mismatch, reshape)\u001b[0m\n\u001b[1;32m   1215\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1216\u001b[0m                 saving.load_weights_from_hdf5_group(\n\u001b[0;32m-> 1217\u001b[0;31m                     f, self.layers, reshape=reshape)\n\u001b[0m\u001b[1;32m   1218\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mf\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'close'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1219\u001b[0m                 \u001b[0mf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclose\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/engine/saving.py\u001b[0m in \u001b[0;36mload_weights_from_hdf5_group\u001b[0;34m(f, layers, reshape)\u001b[0m\n\u001b[1;32m   1169\u001b[0m                          \u001b[0;34m'containing '\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mstr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlayer_names\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1170\u001b[0m                          \u001b[0;34m' layers into a model with '\u001b[0m \u001b[0;34m+\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1171\u001b[0;31m                          str(len(filtered_layers)) + ' layers.')\n\u001b[0m\u001b[1;32m   1172\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1173\u001b[0m     \u001b[0;31m# We batch weight value assignments in a single backend call\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mValueError\u001b[0m: You are trying to load a weight file containing 4 layers into a model with 5 layers."
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eE7Un1NM25KK",
        "colab_type": "code",
        "outputId": "2e3b6ace-de9f-4cb3-ff80-ef3c257d988d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 459
        }
      },
      "source": [
        "agent.model.summary()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d_1 (Conv2D)            (None, 3, 3, 8)           24        \n",
            "_________________________________________________________________\n",
            "conv2d_2 (Conv2D)            (None, 3, 3, 8)           584       \n",
            "_________________________________________________________________\n",
            "flatten_1 (Flatten)          (None, 72)                0         \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 100)               7300      \n",
            "_________________________________________________________________\n",
            "dropout_1 (Dropout)          (None, 100)               0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 100)               10100     \n",
            "_________________________________________________________________\n",
            "dropout_2 (Dropout)          (None, 100)               0         \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 50)                5050      \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 9)                 459       \n",
            "=================================================================\n",
            "Total params: 23,517\n",
            "Trainable params: 23,517\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UnU5iMlkcgty",
        "colab_type": "code",
        "outputId": "1b329b92-5256-46c1-bc28-15ad464bd69c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        }
      },
      "source": [
        "move, pred = agent.decide_move(game.state , only_valid=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3005: The name tf.Session is deprecated. Please use tf.compat.v1.Session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:190: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:197: The name tf.ConfigProto is deprecated. Please use tf.compat.v1.ConfigProto instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:207: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:216: The name tf.is_variable_initialized is deprecated. Please use tf.compat.v1.is_variable_initialized instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:223: The name tf.variables_initializer is deprecated. Please use tf.compat.v1.variables_initializer instead.\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nmoOcmWXNQc7MsuYgZ2ZZ+z/pYqk9/t/ctwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wc18BX5C_sAd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from copy import deepcopy\n",
        "\n",
        "def apply_rewards(reward, predictions, decay = 0.5):\n",
        "  # copy and reverse (more reward applied to latest moves)\n",
        "  output = deepcopy(predictions)[::-1]\n",
        "  applied_reward = reward\n",
        "  for pred, move in output:\n",
        "    move_index =  np.unravel_index(move, pred.shape)\n",
        "    pred[ move_index ] = np.clip(pred[ move_index ] + applied_reward, 0, 1)\n",
        "    applied_reward *= decay\n",
        "  # un-reverse \n",
        "  return [o[0] for o in output[::-1]]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L5SuTyfwDbbA",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def get_states(memory):\n",
        "  return [m[0] for m in memory]\n",
        "\n",
        "def get_new_qs(memory):\n",
        "  return [m[1] for m in memory]\n",
        "\n",
        "def render_memory(memory):\n",
        "  states = get_states(memory)\n",
        "  updated = get_new_qs(memory)\n",
        "  current = [m[2][0] for m in memory]\n",
        "  print('states')\n",
        "  render_all(states)\n",
        "  print('current qs')\n",
        "  render_all(current)\n",
        "  print('upated qs')\n",
        "  render_all(updated)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6S8fhzd2Ei1O",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def agent_perf(agent, n= 100,lose_score = LOSE, show_lose=False):\n",
        "  game_scores = []\n",
        "  for i in range(n):\n",
        "    # play game\n",
        "    game = XandOs()\n",
        "    while not game.end:\n",
        "      move, pred = agent.decide_move(game.state, False)\n",
        "      predictions.append(pred)\n",
        "      _reward, _state, _end = game.step(move)\n",
        "    if show_lose:\n",
        "      if(_reward == lose_score):\n",
        "        render_all(np.array([h[0] for h in game.history] ) )\n",
        "    game_scores.append(_reward)\n",
        "  return np.mean(game_scores), game_scores"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g6sxi8Qkz1AI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "memory = []\n",
        "EPSILON = 0.01\n",
        "\n",
        "for i in range(1_000):\n",
        "\n",
        "  # print('NEW GAME')\n",
        "  game = XandOs()\n",
        "  predictions = []\n",
        "  # randomly who starts\n",
        "  if np.random.random() > 0.5:\n",
        "    game.make_move(game.rand_agent_move(), player=2) \n",
        "    # print('player 2 first')  \n",
        "  \n",
        "  while not game.end:\n",
        "    # game.render()\n",
        "    move, pred = agent.decide_move(game.state, False, only_valid=True)\n",
        "    if np.random.random() < EPSILON :\n",
        "      move = game.rand_agent_move(player = 1)\n",
        "    predictions.append([pred, move])\n",
        "    _reward, _state, _end = game.step(move)\n",
        "  # print(_reward)\n",
        "\n",
        "  if _reward == INVALID:\n",
        "    # invalid move reward\n",
        "    # strongly punish last move\n",
        "    updated_q = apply_rewards(_reward, predictions, decay=0)\n",
        "  else:\n",
        "    # valid move reward\n",
        "    # normally reward previous moves\n",
        "    updated_q = apply_rewards(_reward, predictions, decay=0.9)\n",
        "\n",
        "  states = [h[0] for h in game.history]\n",
        "  for j in range(len(updated_q)):\n",
        "    memory.append( [states[j], updated_q[j], predictions[j]])\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kk-WI3b-ELhU",
        "colab_type": "code",
        "outputId": "2e0a1cca-7cec-4920-ab6e-62e3c8f4d6d0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 770
        }
      },
      "source": [
        "render_memory(memory[-10:])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "states\n",
            "make move:  1\n",
            "make move:  1\n",
            "make move:  1\n",
            "make move:  0\n",
            "make move:  0\n",
            "make move:  0\n",
            "make move:  5\n",
            "make move:  5\n",
            "make move:  5\n",
            "make move:  1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAABACAYAAAAK/4xcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQY+IKEQKekREIVLQIyIK8X9j1Y6Q6TAjvwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "current qs\n",
            "make move:  6\n",
            "make move:  3\n",
            "make move:  4\n",
            "make move:  2\n",
            "make move:  3\n",
            "make move:  7\n",
            "make move:  2\n",
            "make move:  4\n",
            "make move:  8\n",
            "make move:  0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAABACAYAAAAK/4xcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ns0y4oZuZZeJ/vUeFzlLzDXcAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "upated qs\n",
            "make move:  6\n",
            "make move:  3\n",
            "make move:  4\n",
            "make move:  2\n",
            "make move:  3\n",
            "make move:  7\n",
            "make move:  2\n",
            "make move:  4\n",
            "make move:  8\n",
            "make move:  0\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAABACAYAAAAK/4xcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nH/NQhc0nVDI6AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6cgNw84GEdPf",
        "colab_type": "code",
        "outputId": "939c687d-ee7c-44a3-ce46-1d7324382113",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "moves = [p[2][0] for p in memory]\n",
        "moves_np = np.squeeze(np.array(moves).sum(axis=1))\n",
        "moves_np.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(3266,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 399
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cFYEAUPBGREe",
        "colab_type": "code",
        "outputId": "123e6745-1fbf-4b40-f477-6bf762d570f5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "_X = np.array(get_states(memory))\n",
        "X = np.zeros([*_X.shape,2])\n",
        "X[:,:,:,0] = (_X == 1)\n",
        "X[:,:,:,1] = (_X == 2)\n",
        "y = np.squeeze(np.array(get_new_qs(memory)))\n",
        "\n",
        "y_sum = y.sum(axis=1)\n",
        "\n",
        "pos = (y_sum > moves_np)\n",
        "\n",
        "positive_rewards = pos.sum()\n",
        "negative_rewards = (~pos).sum()\n",
        "print('{} positive change, {} negative change'.format(positive_rewards ,negative_rewards))\n",
        "updated_moves = y.argmax(axis=1)\n",
        "updated_moves\n",
        "\n",
        "moves = [p[2][1] for p in memory]\n",
        "mask = (moves_np != updated_moves)\n",
        "\n",
        "X = X[mask]\n",
        "y = y[mask]\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2834 positive change, 432 negative change\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "e4Qr4yBla44P",
        "colab_type": "code",
        "outputId": "71276871-277f-42a3-b47e-a70b0cad5cbb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "moves_np"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([4.716607 , 1.8522296, 2.5783396, ..., 3.4604878, 2.5777278,\n",
              "       1.6078216], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 401
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lFMmMswGa09R",
        "colab_type": "code",
        "outputId": "1ec08d93-3b7c-4716-eacb-9f594aac16b5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "moves_np != updated_moves"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ True,  True,  True, ...,  True,  True,  True])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 402
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q822Gf4hTJM1",
        "colab_type": "code",
        "outputId": "d9f48ae5-17bf-4a52-86fb-8c3913c14837",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "pos_weight =  negative_rewards/len(y)\n",
        "neg_weight =  positive_rewards/len(y)\n",
        "\n",
        "weights = np.zeros(len(y))\n",
        "weights[pos] = pos_weight\n",
        "weights[~pos] = neg_weight\n",
        "pos_weight,neg_weight\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(0.13227189222290264, 0.8677281077770974)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 403
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3iVK6U0aanJO",
        "colab_type": "code",
        "outputId": "bb666632-b20d-49ee-c89f-66a52b5ad831",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "mask.sum()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "3266"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 404
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TggbFTPFGtKp",
        "colab_type": "code",
        "outputId": "15d6a924-0d7e-4aaa-8609-67f6db295fe3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "X.shape, y.shape"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "((3266, 3, 3, 2), (3266, 9))"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 405
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jy8o3j_GKi0p",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# mean, scores = agent_perf(agent,n=1000)\n",
        "# mean"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "O4dfe2vsG3J2",
        "colab_type": "code",
        "outputId": "a1c56496-4259-491f-fcd3-dfa21346ea94",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "\n",
        "hist = agent.model.fit(X, y, epochs=150, verbose=1, batch_size=124) # , sample_weight=weights)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/150\n",
            "3266/3266 [==============================] - 0s 71us/step - loss: 0.3687\n",
            "Epoch 2/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3676\n",
            "Epoch 3/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3668\n",
            "Epoch 4/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3663\n",
            "Epoch 5/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3661\n",
            "Epoch 6/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3662\n",
            "Epoch 7/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3656\n",
            "Epoch 8/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3659\n",
            "Epoch 9/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3659\n",
            "Epoch 10/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3656\n",
            "Epoch 11/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3656\n",
            "Epoch 12/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3656\n",
            "Epoch 13/150\n",
            "3266/3266 [==============================] - 0s 47us/step - loss: 0.3655\n",
            "Epoch 14/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3653\n",
            "Epoch 15/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3652\n",
            "Epoch 16/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3653\n",
            "Epoch 17/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3653\n",
            "Epoch 18/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3651\n",
            "Epoch 19/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3652\n",
            "Epoch 20/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3652\n",
            "Epoch 21/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3652\n",
            "Epoch 22/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3652\n",
            "Epoch 23/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3652\n",
            "Epoch 24/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3653\n",
            "Epoch 25/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3653\n",
            "Epoch 26/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3653\n",
            "Epoch 27/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3651\n",
            "Epoch 28/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3652\n",
            "Epoch 29/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3650\n",
            "Epoch 30/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3647\n",
            "Epoch 31/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3650\n",
            "Epoch 32/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3653\n",
            "Epoch 33/150\n",
            "3266/3266 [==============================] - 0s 38us/step - loss: 0.3653\n",
            "Epoch 34/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3647\n",
            "Epoch 35/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3652\n",
            "Epoch 36/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3650\n",
            "Epoch 37/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3652\n",
            "Epoch 38/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3652\n",
            "Epoch 39/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3651\n",
            "Epoch 40/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3650\n",
            "Epoch 41/150\n",
            "3266/3266 [==============================] - 0s 38us/step - loss: 0.3648\n",
            "Epoch 42/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3654\n",
            "Epoch 43/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3656\n",
            "Epoch 44/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3651\n",
            "Epoch 45/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3651\n",
            "Epoch 46/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3649\n",
            "Epoch 47/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3649\n",
            "Epoch 48/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3653\n",
            "Epoch 49/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3648\n",
            "Epoch 50/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 51/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3648\n",
            "Epoch 52/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3650\n",
            "Epoch 53/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3651\n",
            "Epoch 54/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 55/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3651\n",
            "Epoch 56/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3652\n",
            "Epoch 57/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3651\n",
            "Epoch 58/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3649\n",
            "Epoch 59/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3649\n",
            "Epoch 60/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 61/150\n",
            "3266/3266 [==============================] - 0s 47us/step - loss: 0.3650\n",
            "Epoch 62/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 63/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3649\n",
            "Epoch 64/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3650\n",
            "Epoch 65/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3650\n",
            "Epoch 66/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3651\n",
            "Epoch 67/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3648\n",
            "Epoch 68/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3647\n",
            "Epoch 69/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 70/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 71/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3651\n",
            "Epoch 72/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3648\n",
            "Epoch 73/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3650\n",
            "Epoch 74/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3649\n",
            "Epoch 75/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3648\n",
            "Epoch 76/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3647\n",
            "Epoch 77/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3646\n",
            "Epoch 78/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3648\n",
            "Epoch 79/150\n",
            "3266/3266 [==============================] - 0s 39us/step - loss: 0.3648\n",
            "Epoch 80/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3646\n",
            "Epoch 81/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 82/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3645\n",
            "Epoch 83/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3649\n",
            "Epoch 84/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3647\n",
            "Epoch 85/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3647\n",
            "Epoch 86/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3647\n",
            "Epoch 87/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3649\n",
            "Epoch 88/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3648\n",
            "Epoch 89/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3645\n",
            "Epoch 90/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3648\n",
            "Epoch 91/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3649\n",
            "Epoch 92/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3653\n",
            "Epoch 93/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3648\n",
            "Epoch 94/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3651\n",
            "Epoch 95/150\n",
            "3266/3266 [==============================] - 0s 40us/step - loss: 0.3650\n",
            "Epoch 96/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3647\n",
            "Epoch 97/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 98/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3646\n",
            "Epoch 99/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3647\n",
            "Epoch 100/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3650\n",
            "Epoch 101/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3648\n",
            "Epoch 102/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3648\n",
            "Epoch 103/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3647\n",
            "Epoch 104/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3648\n",
            "Epoch 105/150\n",
            "3266/3266 [==============================] - 0s 47us/step - loss: 0.3648\n",
            "Epoch 106/150\n",
            "3266/3266 [==============================] - 0s 47us/step - loss: 0.3648\n",
            "Epoch 107/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3650\n",
            "Epoch 108/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3645\n",
            "Epoch 109/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3648\n",
            "Epoch 110/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3647\n",
            "Epoch 111/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3645\n",
            "Epoch 112/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3646\n",
            "Epoch 113/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 114/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3651\n",
            "Epoch 115/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3645\n",
            "Epoch 116/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3647\n",
            "Epoch 117/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3646\n",
            "Epoch 118/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3646\n",
            "Epoch 119/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3649\n",
            "Epoch 120/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3647\n",
            "Epoch 121/150\n",
            "3266/3266 [==============================] - 0s 51us/step - loss: 0.3647\n",
            "Epoch 122/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3648\n",
            "Epoch 123/150\n",
            "3266/3266 [==============================] - 0s 47us/step - loss: 0.3649\n",
            "Epoch 124/150\n",
            "3266/3266 [==============================] - 0s 48us/step - loss: 0.3646\n",
            "Epoch 125/150\n",
            "3266/3266 [==============================] - 0s 49us/step - loss: 0.3648\n",
            "Epoch 126/150\n",
            "3266/3266 [==============================] - 0s 51us/step - loss: 0.3651\n",
            "Epoch 127/150\n",
            "3266/3266 [==============================] - 0s 52us/step - loss: 0.3649\n",
            "Epoch 128/150\n",
            "3266/3266 [==============================] - 0s 51us/step - loss: 0.3648\n",
            "Epoch 129/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3647\n",
            "Epoch 130/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3648\n",
            "Epoch 131/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 132/150\n",
            "3266/3266 [==============================] - 0s 53us/step - loss: 0.3647\n",
            "Epoch 133/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3646\n",
            "Epoch 134/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3647\n",
            "Epoch 135/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3648\n",
            "Epoch 136/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3648\n",
            "Epoch 137/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3647\n",
            "Epoch 138/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3649\n",
            "Epoch 139/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3647\n",
            "Epoch 140/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3648\n",
            "Epoch 141/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3646\n",
            "Epoch 142/150\n",
            "3266/3266 [==============================] - 0s 49us/step - loss: 0.3647\n",
            "Epoch 143/150\n",
            "3266/3266 [==============================] - 0s 46us/step - loss: 0.3648\n",
            "Epoch 144/150\n",
            "3266/3266 [==============================] - 0s 43us/step - loss: 0.3647\n",
            "Epoch 145/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3646\n",
            "Epoch 146/150\n",
            "3266/3266 [==============================] - 0s 45us/step - loss: 0.3646\n",
            "Epoch 147/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3646\n",
            "Epoch 148/150\n",
            "3266/3266 [==============================] - 0s 41us/step - loss: 0.3652\n",
            "Epoch 149/150\n",
            "3266/3266 [==============================] - 0s 42us/step - loss: 0.3648\n",
            "Epoch 150/150\n",
            "3266/3266 [==============================] - 0s 44us/step - loss: 0.3652\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p7nhXsEATbyo",
        "colab_type": "code",
        "outputId": "bf20e80d-6ec7-4c6d-a061-aa65efdd4e55",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        }
      },
      "source": [
        "plt.plot(hist.history['loss'])"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7f0fd6eb9390>]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 408
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD4CAYAAADlwTGnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OmsckKHSDDO1",
        "colab_type": "code",
        "outputId": "f39a094d-8734-47f8-a6b1-11da9615a850",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 286
        }
      },
      "source": [
        "means = []\n",
        "for i in range(10):\n",
        "  mean, scores = agent_perf(agent,n=100,show_lose=True)\n",
        "  means.append(mean)\n",
        "  plt.plot(sorted(scores))\n",
        "plt.ylim(INVALID,WIN)\n",
        "plt.show()\n",
        "max(means)\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjIsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy8li6FKAAASxklEQVR4nO3dfYxd9X3n8feH8QwEkwRICA82BLdY\nTWkeSDpiEzW7SoqbAk2h7TYSaKWQNJFVqajd1VYRCG2iTXelVKm2u1VRslaSLa2ikJaWxZs4AUNS\noWpFwrAlhMdgKCk2EAYID8tDsPF3/7jH3cl0xp7xuTN3cn/vl3Q15+E35/s7OtbHZ3733N9NVSFJ\nGn9HjLoDkqTVYeBLUiMMfElqhIEvSY0w8CWpEQa+JDViKIGf5AtJHk9y5yL7k+RPkuxKckeSdwyj\nriRp6YZ1h/9nwLkH2X8esLl7bQU+M6S6kqQlGkrgV9XNwFMHaXIh8Oc1cAtwbJKTh1FbkrQ061ap\nzgbg4Tnru7ttj85vmGQrg78CWL9+/c+/6U1vWnax5554ACZeAeAfJ05mXyYOo8uSNBoTtZ83v+aY\nw/rd22677YmqOmGhfasV+EtWVduAbQDT09M1MzPT63gbbryNt+x+hN//6VOH0T1JWnHr1k3wnne9\n5bB+N8n3Fz3uYfdoefYAcxN3Y7dtZYs+9gSvTExw3Et72fIvz1rpcpK0pq3WY5nbgQ92T+u8E3im\nqv7ZcM6wfW/XoMQx+50gTpKGcoef5EvAe4DXJ9kNfAKYBKiqzwI7gPOBXcALwIeHUfdQHnr8STju\nWF57RFajnCStaUMJ/Kq6+BD7C/idYdRajseeeR6OO5bjp6ZWu7QkrTlj/Unbp370IwBOfM3RI+6J\nJI3eWAf+s93Y/WlvOH7EPZGk0RvrwH+uG7vfvOmkEfdEkkZvrAP/+YkJjtj/ChtOet2ouyJJIzfW\ngf/i1ARH7d3L5OTkqLsiSSM31oH/wuQ6jtq7d9TdkKQ1YawD/6WpSY562cCXJBj3wJ+c5FUGviQB\nYx74L05NctTefaPuhiStCWMd+D+anORoA1+SgDEO/Ed/8BT7Jtaxft8ro+6KJK0JYxv49//DIwCs\nf8WZMiUJxjjwv//YkwC8Js6UKUkwxoH/6LPPA3D8lB+6kiQY48B/8sXBTJlvOOZVI+6JJK0NYxv4\nz76yH4CNJxw74p5I0towtoH/XDd0/zObThltRyRpjRjbwH9+3QRH7N/PGzeeMOquSNKaMJTAT3Ju\nkvuS7Epy2QL7P5RkNsnt3eujw6h7MC9MruPIfc6UKUkH9P5O2yQTwJXALwG7gVuTbK+qu+c1/XJV\nXdq33lK9OLXOidMkaY5h3OGfDeyqqger6mXgauDCIRy3lxedGlmSfswwAn8D8PCc9d3dtvn+dZI7\nklyT5NQh1D2ol6YmDXxJmmO13rT9X8DpVfVWYCdw1WINk2xNMpNkZnZ29rALDqZGduI0STpgGIG/\nB5h7x76x2/ZPqurJqvpRt/o54OcXO1hVbauq6aqaPuGEw3/CxsCXpB83jMC/FdicZFOSKeAiYPvc\nBklOnrN6AXDPEOou6ulnnmPvuknW7zPwJemA3k/pVNW+JJcC1wMTwBeq6q4knwRmqmo78LtJLgD2\nAU8BH+pb92Du3dXNlLnfmTIl6YDegQ9QVTuAHfO2fXzO8uXA5cOotRQPPfIEHLOe16xWQUn6CTCW\nn7Td88NnATjeD11J0j8Zy8A/MFPmCeudKVOSDhjLwH96/+BrDTe+7rUj7okkrR1jGfjPdT9/+rQT\nR9oPSVpLxjLwn584glRxxuknH7qxJDViLAP/hcl1HLl3L0ceNTXqrkjSmjGWgf/i1CDwJUn/33gG\n/qQTp0nSfGMZ+C9NreNVzoUvST9mPAN/csrAl6R5xjTwJzl6rxOnSdJcYxf4e/fu5ef2PMbPOjWy\nJP2YoUyetpZMTk7ytQ//6qi7IUlrztjd4UuSFmbgS1IjDHxJaoSBL0mNMPAlqREGviQ1YiiBn+Tc\nJPcl2ZXksgX2H5nky93+byU5fRh1JUlL1zvwk0wAVwLnAWcCFyc5c16zjwA/rKozgD8G/rBvXUnS\n8gzjDv9sYFdVPVhVLwNXAxfOa3MhcFW3fA1wTpIMobYkaYmGEfgbgIfnrO/uti3Ypqr2Ac8Ar1vo\nYEm2JplJMjM7OzuE7kmSYA2+aVtV26pquqqmTzjhhFF3R5LGxjACfw9w6pz1jd22BdskWQe8Fnhy\nCLUlSUs0jMC/FdicZFOSKeAiYPu8NtuBS7rl3wS+UVU1hNqSpCXqPVtmVe1LcilwPTABfKGq7kry\nSWCmqrYDnwf+Isku4CkG/ylIklbRUKZHrqodwI552z4+Z/kl4APDqCVJOjxr7k1bSdLKMPAlqREG\nviQ1wsCXpEYY+JLUCANfkhph4EtSIwx8SWqEgS9JjTDwJakRBr4kNcLAl6RGGPiS1AgDX5IaYeBL\nUiMMfElqhIEvSY0w8CWpEb0CP8nxSXYmub/7edwi7V5Jcnv3mv8F55KkVdD3Dv8y4Kaq2gzc1K0v\n5MWqOqt7XdCzpiTpMPQN/AuBq7rlq4Bf63k8SdIK6Rv4J1bVo93yY8CJi7Q7KslMkluSHPQ/hSRb\nu7Yzs7OzPbsnSTpg3aEaJLkROGmBXVfMXamqSlKLHOaNVbUnyU8B30jy3ap6YKGGVbUN2AYwPT29\n2PEkSct0yMCvqi2L7UvygyQnV9WjSU4GHl/kGHu6nw8m+Vvg7cCCgS9JWhl9h3S2A5d0y5cA181v\nkOS4JEd2y68HfgG4u2ddSdIy9Q38TwG/lOR+YEu3TpLpJJ/r2vwsMJPkO8A3gU9VlYEvSavskEM6\nB1NVTwLnLLB9Bvhot/y/gbf0qSNJ6s9P2kpSIwx8SWqEgS9JjTDwJakRBr4kNcLAl6RGGPiS1AgD\nX5IaYeBLUiMMfElqhIEvSY0w8CWpEQa+JDXCwJekRhj4ktQIA1+SGmHgS1IjDHxJakSvwE/ygSR3\nJdmfZPog7c5Ncl+SXUku61NTknR4+t7h3wn8BnDzYg2STABXAucBZwIXJzmzZ11J0jL1/RLzewCS\nHKzZ2cCuqnqwa3s1cCFwd5/akqTlWY0x/A3Aw3PWd3fbFpRka5KZJDOzs7Mr3jlJasUh7/CT3Aic\ntMCuK6rqumF3qKq2AdsApqena9jHl6RWHTLwq2pLzxp7gFPnrG/stkmSVtFqDOncCmxOsinJFHAR\nsH0V6kqS5uj7WOavJ9kNvAv4apLru+2nJNkBUFX7gEuB64F7gL+sqrv6dVuStFx9n9K5Frh2ge2P\nAOfPWd8B7OhTS5LUj5+0laRGGPiS1AgDX5IaYeBLUiMMfElqhIEvSY0w8CWpEQa+JDXCwJekRhj4\nktQIA1+SGmHgS1IjDHxJaoSBL0mNMPAlqREGviQ1wsCXpEYY+JLUiL7fafuBJHcl2Z9k+iDtHkry\n3SS3J5npU1OSdHh6factcCfwG8B/X0Lb91bVEz3rSZIOU98vMb8HIMlweiNJWjGrNYZfwA1Jbkuy\n9WANk2xNMpNkZnZ2dpW6J0nj75B3+EluBE5aYNcVVXXdEuu8u6r2JHkDsDPJvVV180INq2obsA1g\nenq6lnh8SdIhHDLwq2pL3yJVtaf7+XiSa4GzgQUDX5K0MlZ8SCfJ+iSvPrAMvI/Bm72SpFXU97HM\nX0+yG3gX8NUk13fbT0myo2t2IvB3Sb4DfBv4alV9vU9dSdLy9X1K51rg2gW2PwKc3y0/CLytTx1J\nUn9+0laSGmHgS1IjDHxJaoSBL0mNMPAlqREGviQ1wsCXpEYY+JLUCANfkhph4EtSIwx8SWqEgS9J\njTDwJakRBr4kNcLAl6RGGPiS1AgDX5IaYeBLUiP6fqftp5Pcm+SOJNcmOXaRducmuS/JriSX9akp\nSTo8fe/wdwJvrqq3At8DLp/fIMkEcCVwHnAmcHGSM3vWlSQtU6/Ar6obqmpft3oLsHGBZmcDu6rq\nwap6GbgauLBPXUnS8g1zDP+3gK8tsH0D8PCc9d3dtgUl2ZpkJsnM7OzsELsnSW1bd6gGSW4ETlpg\n1xVVdV3X5gpgH/DFvh2qqm3ANoDp6enqezxJ0sAhA7+qthxsf5IPAe8HzqmqhQJ6D3DqnPWN3TZJ\n0irq+5TOucDHgAuq6oVFmt0KbE6yKckUcBGwvU9dSdLy9R3D/1Pg1cDOJLcn+SxAklOS7ADo3tS9\nFLgeuAf4y6q6q2ddSdIyHXJI52Cq6oxFtj8CnD9nfQewo08tSVI/ftJWkhph4EtSIwx8SWqEgS9J\njTDwJakRBr4kNcLAl6RGGPiS1AgDX5IaYeBLUiMMfElqhIEvSY0w8CWpEQa+JDXCwJekRhj4ktQI\nA1+SGmHgS1Ijen3FYZJPA78KvAw8AHy4qp5eoN1DwHPAK8C+qpruU1eStHx97/B3Am+uqrcC3wMu\nP0jb91bVWYa9JI1Gr8Cvqhuqal+3eguwsX+XJEkrYZhj+L8FfG2RfQXckOS2JFuHWFOStESHHMNP\nciNw0gK7rqiq67o2VwD7gC8ucph3V9WeJG8Adia5t6puXqTeVmArwGmnnbaEU5AkLcUhA7+qthxs\nf5IPAe8HzqmqWuQYe7qfjye5FjgbWDDwq2obsA1genp6weNJkpav15BOknOBjwEXVNULi7RZn+TV\nB5aB9wF39qkrSVq+vmP4fwq8msEwze1JPguQ5JQkO7o2JwJ/l+Q7wLeBr1bV13vWlSQtU6/n8Kvq\njEW2PwKc3y0/CLytTx1JUn9+0laSGmHgS1IjDHxJaoSBL0mNMPAlqREGviQ1wsCXpEYY+JLUCANf\nkhph4EtSIwx8SWqEgS9JjTDwJakRBr4kNcLAl6RGGPiS1AgDX5IaYeBLUiMMfElqRO/AT/IHSe7o\nvsT8hiSnLNLukiT3d69L+taVJC3PMO7wP11Vb62qs4CvAB+f3yDJ8cAngH8BnA18IslxQ6gtSVqi\n3oFfVc/OWV0P1ALNfhnYWVVPVdUPgZ3AuX1rS5KWbt0wDpLkPwMfBJ4B3rtAkw3Aw3PWd3fbFjrW\nVmBrt/p/k9x3mN16PfDEYf7uT6oWzxnaPO8WzxnaPO/lnvMbF9uRqoVuyOc1Sm4ETlpg1xVVdd2c\ndpcDR1XVJ+b9/u932/9Tt/4fgBer6o+W1v/lSzJTVdMrdfy1qMVzhjbPu8VzhjbPe5jnvKQ7/Kra\nssTjfRHYwWC8fq49wHvmrG8E/naJx5QkDcEwntLZPGf1QuDeBZpdD7wvyXHdm7Xv67ZJklbJMMbw\nP5XkZ4D9wPeB3wZIMg38dlV9tKqeSvIHwK3d73yyqp4aQu2D2bbCx1+LWjxnaPO8WzxnaPO8h3bO\nSxrDlyT95POTtpLUCANfkhoxdoGf5Nwk9yXZleSyUfdnpSQ5Nck3k9yd5K4kv9dtPz7Jzm4Ki53j\n+InmJBNJ/j7JV7r1TUm+1V3zLyeZGnUfhy3JsUmuSXJvknuSvGvcr3WSf9f9274zyZeSHDWO1zrJ\nF5I8nuTOOdsWvLYZ+JPu/O9I8o7l1BqrwE8yAVwJnAecCVyc5MzR9mrF7AP+fVWdCbwT+J3uXC8D\nbqqqzcBN3fq4+T3gnjnrfwj8cVWdAfwQ+MhIerWy/hvw9ap6E/A2Buc/ttc6yQbgd4HpqnozMAFc\nxHhe6z/jn888sNi1PQ/Y3L22Ap9ZTqGxCnwG8/TsqqoHq+pl4GoGj4qOnap6tKr+T7f8HIMA2MDg\nfK/qml0F/NpoergykmwEfgX4XLce4BeBa7om43jOrwX+FfB5gKp6uaqeZsyvNYOnCF+VZB1wNPAo\nY3itq+pmYP5Ti4td2wuBP6+BW4Bjk5y81FrjFvhLnsJhnCQ5HXg78C3gxKp6tNv1GHDiiLq1Uv4r\n8DEGjwEDvA54uqr2devjeM03AbPA/+iGsj6XZD1jfK2rag/wR8A/Mgj6Z4DbGP9rfcBi17ZXxo1b\n4DcnyTHAXwP/dt5EdtTgmduxee42yfuBx6vqtlH3ZZWtA94BfKaq3g48z7zhmzG81scxuJvdBJzC\nYGLGJidcHOa1HbfA3wOcOmd9Y7dtLCWZZBD2X6yqv+k2/+DAn3jdz8dH1b8V8AvABUkeYjBc94sM\nxraP7f7sh/G85ruB3VX1rW79Ggb/AYzztd4C/ENVzVbVXuBvGFz/cb/WByx2bXtl3LgF/q3A5u6d\n/CkGb/JsH3GfVkQ3dv154J6q+i9zdm0HDnzBzCXAdfN/9ydVVV1eVRur6nQG1/YbVfVvgG8Cv9k1\nG6tzBqiqx4CHu0+0A5wD3M0YX2sGQznvTHJ092/9wDmP9bWeY7Frux34YPe0zjuBZ+YM/RxaVY3V\nCzgf+B7wAIPZPEfepxU6z3cz+DPvDuD27nU+gzHtm4D7gRuB40fd1xU6//cAX+mWfwr4NrAL+Cvg\nyFH3bwXO9yxgprve/xM4btyvNfAfGczNdSfwF8CR43itgS8xeJ9iL4O/5j6y2LUFwuBJxAeA7zJ4\nimnJtZxaQZIaMW5DOpKkRRj4ktQIA1+SGmHgS1IjDHxJaoSBL0mNMPAlqRH/D0Pr7QWT5FVgAAAA\nAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 409
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "N1q8FW6Cbdwg",
        "colab_type": "code",
        "outputId": "d928e3e1-fa75-488f-e102-8cdbddc8a01a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "mean, scores = agent_perf(agent,n=100, show_lose=True)\n",
        "mean"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "1.0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 412
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cwubgEAk70o6",
        "colab_type": "text"
      },
      "source": [
        "# Lets play"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cmN8dSHCtK7A",
        "colab_type": "code",
        "outputId": "a6723553-d42f-43ca-dd10-5b542cef9be2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "game = XandOs()\n",
        "\n",
        "game.render()\n",
        "player_move = int(input())\n",
        "player_reward = game.make_move(player_move, player=2)\n",
        "\n",
        "while not game.end:\n",
        "  move, pred = agent.decide_move(game.state)\n",
        "  reward = game.make_move(move)\n",
        "  game.render()\n",
        "  print('valid move', reward)\n",
        "  print(pred)\n",
        "\n",
        "  if not game.end:\n",
        "    player_move = int(input())\n",
        "    player_reward = game.make_move(player_move, player=2)\n",
        "    print(player_move, player_reward)\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "valid move True\n",
            "[9.5911449e-01 4.0012732e-01 4.3898532e-01 3.2782555e-07 5.9564728e-01\n",
            " 3.5727382e-01 6.1504680e-01 3.2291222e-01 6.2474960e-01]\n",
            "4\n",
            "4 True\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZe3/AFWzg4Z1z0FsAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkjOzSnOSM7NKc5Izs0r7/4Wx8T/ONeuWAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "valid move True\n",
            "[0.         0.02969357 0.16465813 0.         0.         0.9955013\n",
            " 0.19414705 0.02345785 0.04613736]\n",
            "6\n",
            "6 True\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQc7MsvZ/ESOFJToQmZEAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAATkAAAD8CAYAAAAMs9NCAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "valid move True\n",
            "[0.0000000e+00 2.2417915e-01 9.9990451e-01 0.0000000e+00 0.0000000e+00\n",
            " 2.9802322e-08 0.0000000e+00 2.6210201e-01 7.2412694e-01]\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/ipykernel/kernelbase.py\u001b[0m in \u001b[0;36m_input_request\u001b[0;34m(self, prompt, ident, parent, password)\u001b[0m\n\u001b[1;32m    729\u001b[0m             \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 730\u001b[0;31m                 \u001b[0mident\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mreply\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msession\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstdin_socket\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    731\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/jupyter_client/session.py\u001b[0m in \u001b[0;36mrecv\u001b[0;34m(self, socket, mode, content, copy)\u001b[0m\n\u001b[1;32m    802\u001b[0m         \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 803\u001b[0;31m             \u001b[0mmsg_list\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msocket\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecv_multipart\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmode\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcopy\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    804\u001b[0m         \u001b[0;32mexcept\u001b[0m \u001b[0mzmq\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mZMQError\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/zmq/sugar/socket.py\u001b[0m in \u001b[0;36mrecv_multipart\u001b[0;34m(self, flags, copy, track)\u001b[0m\n\u001b[1;32m    465\u001b[0m         \"\"\"\n\u001b[0;32m--> 466\u001b[0;31m         \u001b[0mparts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrecv\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mflags\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcopy\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrack\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mtrack\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    467\u001b[0m         \u001b[0;31m# have first part already, only loop while more to receive\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32mzmq/backend/cython/socket.pyx\u001b[0m in \u001b[0;36mzmq.backend.cython.socket.Socket.recv\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;32mzmq/backend/cython/socket.pyx\u001b[0m in \u001b[0;36mzmq.backend.cython.socket.Socket.recv\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;32mzmq/backend/cython/socket.pyx\u001b[0m in \u001b[0;36mzmq.backend.cython.socket._recv_copy\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/zmq/backend/cython/checkrc.pxd\u001b[0m in \u001b[0;36mzmq.backend.cython.checkrc._check_rc\u001b[0;34m()\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: ",
            "\nDuring handling of the above exception, another exception occurred:\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-392-449913397003>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     13\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     14\u001b[0m   \u001b[0;32mif\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mend\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 15\u001b[0;31m     \u001b[0mplayer_move\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minput\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     16\u001b[0m     \u001b[0mplayer_reward\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgame\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmake_move\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mplayer_move\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mplayer\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     17\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mplayer_move\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mplayer_reward\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/ipykernel/kernelbase.py\u001b[0m in \u001b[0;36mraw_input\u001b[0;34m(self, prompt)\u001b[0m\n\u001b[1;32m    703\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_parent_ident\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    704\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_parent_header\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 705\u001b[0;31m             \u001b[0mpassword\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    706\u001b[0m         )\n\u001b[1;32m    707\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/ipykernel/kernelbase.py\u001b[0m in \u001b[0;36m_input_request\u001b[0;34m(self, prompt, ident, parent, password)\u001b[0m\n\u001b[1;32m    733\u001b[0m             \u001b[0;32mexcept\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    734\u001b[0m                 \u001b[0;31m# re-raise KeyboardInterrupt, to truncate traceback\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 735\u001b[0;31m                 \u001b[0;32mraise\u001b[0m \u001b[0mKeyboardInterrupt\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    736\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    737\u001b[0m                 \u001b[0;32mbreak\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dQs8vaNeeuSO",
        "colab_type": "code",
        "outputId": "2bd2860c-d9db-4d52-a0b5-2b212f99e395",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "|game.end"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "False"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 64
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "p_DKqBLVDyW9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "game.end"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rH7Uum0Gevf4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "agent.model.save('v4.keras')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mpJwa-ZQfG5x",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "agent.model.save_weights('winer_weights_v3.keras')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "D_UWux3XgSuV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}