simple_image_recognition.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyMHf5c3OAXM/HsCEP1GDnoH",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/gratusrichard/text-recognition/blob/main/simple_image_recognition.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "rgMsjA5v-DZx",
        "outputId": "e678292b-a4a7-4711-d96a-aeb9552ded12"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'2.14.0'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "2jNMsRkQAvO3"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "tf.__version__\n",
        "\n",
        "from tensorflow.keras import layers,models\n",
        "from tensorflow.keras.datasets import mnist\n",
        "from tensorflow.keras.utils import to_categorical"
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "xQP1-YSF-DBe"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "(train_images,train_labels),(test_images,test_labels) = mnist.load_data()\n",
        "print(train_images.shape)\n",
        "print(test_images.shape)\n",
        "train_images = train_images.reshape((60000, 28, 28, 1)).astype('float32') / 255\n",
        "test_images = test_images.reshape((10000, 28, 28, 1)).astype('float32') / 255\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dalpaAMFA7Pp",
        "outputId": "e90a6f7d-bbe2-48e8-ea02-f1bdd10c3c9c"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28)\n",
            "(10000, 28, 28)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_labels = to_categorical(train_labels)\n",
        "test_labels = to_categorical(test_labels)"
      ],
      "metadata": {
        "id": "VfUprg3DBcbd"
      },
      "execution_count": 8,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model = models.Sequential()\n",
        "model.add(layers.Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)))\n",
        "model.add(layers.MaxPooling2D((2, 2)))\n",
        "model.add(layers.Conv2D(64, (3, 3), activation='relu'))\n",
        "model.add(layers.MaxPooling2D((2, 2)))\n",
        "model.add(layers.Conv2D(64, (3, 3), activation='relu'))\n",
        "model.add(layers.Flatten())\n",
        "model.add(layers.Dense(64, activation='relu'))\n",
        "model.add(layers.Dense(10, activation='softmax'))"
      ],
      "metadata": {
        "id": "o8m8FCW2BgK0"
      },
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer='adam',\n",
        "              loss='categorical_crossentropy',\n",
        "              metrics=['accuracy'])\n"
      ],
      "metadata": {
        "id": "iXutz47JCKkp"
      },
      "execution_count": 10,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.fit(train_images, train_labels, epochs=5, batch_size=64, validation_split=0.2)\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kYoe5crSCOFC",
        "outputId": "0bd4da99-6f29-4160-b9f9-b53284748443"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "750/750 [==============================] - 14s 5ms/step - loss: 0.2038 - accuracy: 0.9360 - val_loss: 0.0706 - val_accuracy: 0.9788\n",
            "Epoch 2/5\n",
            "750/750 [==============================] - 4s 6ms/step - loss: 0.0564 - accuracy: 0.9821 - val_loss: 0.0559 - val_accuracy: 0.9825\n",
            "Epoch 3/5\n",
            "750/750 [==============================] - 4s 6ms/step - loss: 0.0378 - accuracy: 0.9880 - val_loss: 0.0435 - val_accuracy: 0.9877\n",
            "Epoch 4/5\n",
            "750/750 [==============================] - 4s 5ms/step - loss: 0.0295 - accuracy: 0.9906 - val_loss: 0.0448 - val_accuracy: 0.9861\n",
            "Epoch 5/5\n",
            "750/750 [==============================] - 4s 5ms/step - loss: 0.0232 - accuracy: 0.9926 - val_loss: 0.0412 - val_accuracy: 0.9883\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.src.callbacks.History at 0x79fe93f93490>"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_loss, test_acc = model.evaluate(test_images, test_labels)\n",
        "print(f'Test accuracy: {test_acc}')"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DilkpR8aDZa_",
        "outputId": "4decb0f9-1921-448a-b279-61498e9ccd39"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 1s 4ms/step - loss: 0.0319 - accuracy: 0.9902\n",
            "Test accuracy: 0.9901999831199646\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "(_, _), (test_images, test_labels) = mnist.load_data()\n",
        "sample_image = test_images[4]\n",
        "sample_label = test_labels[4]"
      ],
      "metadata": {
        "id": "vn8lMEy-Dcfd"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "sample_image = sample_image.reshape((28, 28, 1)).astype('float32') / 255\n",
        "sample_image = np.expand_dims(sample_image, axis=0)"
      ],
      "metadata": {
        "id": "j89atRjwGFO5"
      },
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "predictions = model.predict(sample_image)\n",
        "\n",
        "predicted_digit = np.argmax(predictions)\n",
        "\n",
        "plt.imshow(sample_image.squeeze(), cmap='gray')\n",
        "plt.title(f'True Digit: {sample_label}, Predicted Digit: {predicted_digit}')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 469
        },
        "id": "bULy7wIEGH0G",
        "outputId": "f23fcee8-7109-4c32-a7cb-d41052c9c88a"
      },
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "1/1 [==============================] - 0s 235ms/step\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAaAAAAGzCAYAAABpdMNsAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAqK0lEQVR4nO3de1xVdb7/8ff2tr0BiiiIF0QsNa+PTBk0FdNE7FiWXbzU0WbSbFBTp2zoV2qXcyhnTjmeQ2pzHElNmywvM9XRHBSdGiw1zfFxjBEOiiZYamwQFQ2+vz98uKctIGzc8AV8PR+P7+PBXmt99vqw9oI3a6/F2g5jjBEAANWsnu0GAAA3JwIIAGAFAQQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIPtepUydNmTKlUrXR0dGKjo72aT91mcPh0MKFC92Pk5KS5HA4dPToUWs9XevaHn0lJSVFDodDKSkpXtcePXpUDodDSUlJPu8LFUcAWeJwOCo0KvPD5UvR0dHuXurVqyd/f3917dpVjz32mLZt21bl6z958qQWLlyoAwcO+PR57777bjkcDs2YMaPSz7Fw4UKP16pp06a67bbb9MILLygvL8+H3Va9tWvXavHixdbWfzUQro6GDRsqKChIAwcO1PPPP6+srKwq7+GTTz7xeVBmZGSocePGcjgc2rt3r0+fuy5oYLuBm9Xq1as9Hq9atUrbtm0rMb179+7V2Vap2rdvr4SEBElSQUGB0tPTtWHDBq1Zs0YPP/yw1qxZo4YNG7qXT0tLU716lfvb5tNPP/V4fPLkSb300kvq1KmT+vbtW+nv4ac2bNig1NRUnzyXJC1dulTNmzfXuXPn9Omnn+rf/u3ftH37dn3++edyOBw+W09FPPbYYxo/frycTqdXdWvXrtWhQ4c0e/bsqmmsgiZMmKDRo0eruLhYP/zwg/bs2aPFixfrd7/7nVasWKHx48e7lx0yZIguXLigRo0aeb2esLAwXbhwwWO//eSTT5SYmOjTEJozZ44aNGigwsJCnz1nXUIAWfLoo496PN69e7e2bdtWYvq1zp8/r6ZNm1ZlayUEBASU6Ou1117TrFmz9NZbb6lTp056/fXX3fO8/eX3U5X5ZeKNixcv6le/+pWee+45zZ8/3yfP+eCDDyooKEiSNH36dI0bN04bNmzQ7t27FRUVVWpNVb2O9evXV/369X3+vNXl9ttvL7GvHTt2TCNHjtTkyZPVvXt39enTR5JUr149NW7cuFLrcTgcla6tqK1bt2rr1q2aN2+eXn311SpdV23FW3A1WHR0tHr27Kl9+/ZpyJAhatq0qZ5//nlJZb+vXtr5l9zcXM2ePVsdOnSQ0+lUly5d9Prrr6u4uLjSvdWvX19LlizRbbfdpv/6r/+Sy+W6bg8HDx7U0KFD1aRJE7Vv316vvvqqVq5cWeJ8xU/PAaWkpKh///6SpMcff9z99szV9+3Pnz+vb775RqdPn65w34sWLVJxcbGeeeaZSn3fFXHXXXdJkjIzMyVd/3UsLCzUggUL1KVLFzmdTnXo0EHz5s0r8RdzYWGh5syZo9atW8vPz0/33nuvTpw4UWLdZZ0D+p//+R8NHTpUfn5+8vf3V//+/bV27Vp3fx9//LGOHTvm3sadOnXyWLcve/RWWFiYkpKSdOnSJS1atMg9vaxzQImJiercubOaNGmiAQMG6K9//WuJc4vXngOaMmWKEhMTJXm+PX5Vdna2vvnmG12+fLlCPV++fFlPP/20nn76aUVERFTuG78JcARUw505c0axsbEaP368Hn30UQUHB3tVf/78eQ0dOlTffvutnnzySXXs2FF/+9vfFB8fr+zs7Bt6379+/fqaMGGCXnzxRX322We65557Sl3u22+/1bBhw+RwOBQfH69mzZrpv//7v8s9UurevbtefvllzZ8/X9OmTdPgwYMlSQMHDpQkffnllxo2bJgWLFhQobdNsrKy9Nprr+kPf/iDmjRp4t0364WMjAxJUqtWrdzTSnsdi4uLde+99+qzzz7TtGnT1L17d/3973/Xm2++qX/84x/atGmTu/6JJ57QmjVrNHHiRA0cOFDbt28vc3tfKykpST//+c/Vo0cPxcfHq0WLFtq/f7+2bNmiiRMn6v/9v/8nl8ulEydO6M0335QkNW/eXJKqrcfyREVFKSIiotzzjkuXLtWMGTM0ePBgzZkzR0ePHtXYsWPVsmVLtW/fvsy6J598UidPniz1bXBJio+P1zvvvKPMzEyPcC7L4sWL9cMPP+iFF17Qhg0byl3+pmVQI8TFxZlrX46hQ4caSWbZsmUllpdkFixYUGJ6WFiYmTx5svvxK6+8Ypo1a2b+8Y9/eCz361//2tSvX99kZWVdt6+hQ4eaHj16lDl/48aNRpL53e9+V2YPM2fONA6Hw+zfv9897cyZMyYwMNBIMpmZmR7rGzp0qPvxnj17jCSzcuXKEuvesWNHmduhNA8++KAZOHCg+7EkExcXV6Ha0ixYsMBIMmlpaeb77783mZmZZvny5cbpdJrg4GBTUFDg/p5Kex1Xr15t6tWrZ/761796TF+2bJmRZD7//HNjjDEHDhwwkswvf/lLj+UmTpxY4vtfuXKlxzbNzc01fn5+JjIy0ly4cMGjvri42P31PffcY8LCwkp8j1XRY2kyMzONJPOb3/ymzGXuu+8+I8m4XC5jzD9f/x07dhhjjCksLDStWrUy/fv3N5cvX3bXJSUlGUke+9XV9f10vyrtZ/CqyZMnl9hXy5KdnW38/PzM8uXLjTH/fE327NlTbu3Nhrfgajin06nHH3+80vXr16/X4MGD1bJlS50+fdo9RowYoaKiIu3ateuG+rv6l3J+fn6Zy2zZskVRUVEeFxEEBgZq0qRJN7Tu6OhoGWMqdPSzY8cOffjhh1VypVfXrl3VunVrhYeH68knn1SXLl308ccfe5zjKe11XL9+vbp3765u3bp5vDZX38LbsWOHpCsnxyVp1qxZHvUVuWBg27Ztys/P169//esS5zwqcoFEdfRYUeXta3v37tWZM2c0depUNWjwzzd3Jk2apJYtW97QupOSkmSMqdDRz3PPPafOnTvriSeeuKF13gx4C66Ga9eu3Q2dmD9y5IgOHjyo1q1blzr/u+++q/RzS9K5c+ckSX5+fmUuc+zYsVJPxnfp0uWG1l1RP/74o2bNmqXHHnvMfU7Jlz788EP5+/urYcOGat++fanv+Zf2Oh45ckSHDx8u97U5duyY6tWrV+J5u3btWm5vV98O7NmzZ4W+l2tVR48VVd6+duzYMUkl96sGDRpUKDh8Yffu3Vq9erWSk5MrfSXozYQAquG8PVdRVFTk8bi4uFh333235s2bV+ryt956a6V7k6RDhw5Jqr4wqYxVq1YpLS1Ny5cvL3FyPj8/X0ePHlWbNm0qfVXakCFD3FfBlaW017G4uFi9evXSG2+8UWpNhw4dKtWPL9WkHg8dOqQ2bdrI39+/2tbprXnz5mnw4MEKDw9372tXL5LJzs5WVlaWOnbsaLHDmoUAqqVatmyp3Nxcj2mXLl1Sdna2x7SIiAidO3dOI0aM8HkPRUVFWrt2rZo2bao777yzzOXCwsKUnp5eYnpp067li/+jycrK0uXLlzVo0KAS81atWqVVq1Zp48aNGjt27A2vyxsRERH6+uuvNXz48Ot+n2FhYSouLlZGRobHEUVaWlqF1iFd+eV9vT8Sylp/dfRYEampqcrIyLjuvymEhYVJurJfDRs2zD39xx9/1NGjR9W7d+/rrsNX+9qxY8cUHh5eYt69996rgICAEj+3NzOOEWupiIiIEudv3n777RJHQA8//LBSU1O1devWEs+Rm5urH3/8sVLrLyoq0qxZs3T48GHNmjXrun+VxsTEKDU11eNuBmfPntW7775b7nqaNWvm7vVaFb0Me/z48dq4cWOJIUmjR4/Wxo0bFRkZWW4vvvbwww/r22+/1e9///sS8y5cuKCCggJJUmxsrCRpyZIlHstU5HzWyJEj5efnp4SEBF28eNFjnjHG/XWzZs08LqWvzh7Lc+zYMU2ZMkWNGjXSs88+W+Zyd9xxh1q1aqXf//73Hvv1u+++qx9++KHc9VxvX6voZdhvv/12if1s5syZkqTf/va3FdrnbyYcAdVSTzzxhPufHu+++259/fXX2rp1a4m3gp599ln96U9/0r/8y79oypQp6tevnwoKCvT3v/9dH3zwgY4ePVru20cul0tr1qyRdOWX/tU7IWRkZGj8+PF65ZVXrls/b948rVmzRnfffbdmzpzpvgy7Y8eOOnv27HX/8oyIiFCLFi20bNky+fn5qVmzZoqMjFR4eHiFL8Pu1q2bunXrVuq88PDwEkc+0dHR2rlzp8cv6Krw2GOP6f3339f06dO1Y8cODRo0SEVFRfrmm2/0/vvva+vWrbrjjjvUt29fTZgwQW+99ZZcLpcGDhyo5OTkCh1B+vv7680339QTTzyh/v37a+LEiWrZsqW+/vprnT9/Xu+8844kqV+/fvrjH/+ouXPnqn///mrevLnGjBlTLT3+1FdffaU1a9aouLhYubm52rNnjz788EM5HA6tXr36ukcxjRo10sKFCzVz5kzdddddevjhh3X06FElJSUpIiKi3COcfv36SbpyIUVMTIzq16/vvvNCRS/DHjlyZIlpVwNt6NChuuOOO8rZAjcZuxfh4aqyLsMu6xLooqIi89xzz5mgoCDTtGlTExMTY9LT00tcAm2MMfn5+SY+Pt506dLFNGrUyAQFBZmBAwea3/72t+bSpUvX7evqJcRXR/Pmzc0tt9xiHn30UfPpp5+WWlNaD/v37zeDBw82TqfTtG/f3iQkJJglS5YYSSYnJ8djfT+9XNYYYzZv3mxuu+0206BBA49LZ729DPtaKuMy7H79+pmQkJBy669ehv39999fd7nrvY6XLl0yr7/+uunRo4dxOp2mZcuWpl+/fuall15yX25sjDEXLlwws2bNMq1atTLNmjUzY8aMMcePHy/3Muyr/vSnP5mBAweaJk2aGH9/fzNgwACzbt069/xz586ZiRMnmhYtWhhJHpdk+7rH0ly9LPrqaNCggQkMDDSRkZEmPj7eHDt2rETNtZdhX7VkyRITFhZmnE6nGTBggPn8889Nv379zKhRo0qs76eXYf/4449m5syZpnXr1sbhcHj8PHpzGfa1uAy7bA5jqvjPPKAMs2fP1vLly3Xu3Lkac/uY/Px8BQYGavHixYqLi7PdDnyguLhYrVu31gMPPFDqW4mwh3NAqBYXLlzweHzmzBmtXr1ad955Z40JH0natWuX2rVrp6lTp9puBZVw8eLFEm+drlq1SmfPnuVjPmogjoBQLfr27avo6Gh1795dp06d0ooVK3Ty5EklJydryJAhtttDHZGSkqI5c+booYceUqtWrfTVV19pxYoV6t69u/bt21flN7uFd7gIAdVi9OjR+uCDD/T222/L4XDo9ttv14oVKwgf+FSnTp3UoUMHLVmyRGfPnlVgYKD+9V//Va+99hrhUwNxBAQAsIJzQAAAKwggAIAVNe4cUHFxsU6ePCk/P79q/zhjAMCNM8YoPz9foaGh170pa40LoJMnT9aImzACAG7M8ePHr/tBgDXuLbjr3dYfAFB7lPf7vMoCKDExUZ06dVLjxo0VGRmpL7/8skJ1vO0GAHVDeb/PqySArt7UcMGCBfrqq6/Up08fxcTE3PCHnwEA6pCquMHcgAEDPG7yWFRUZEJDQ01CQkK5tS6Xy+OmhAwGg8GoneOnN6stjc+PgC5duqR9+/Z5fABavXr1NGLECKWmppZYvrCwUHl5eR4DAFD3+TyATp8+raKiIgUHB3tMDw4OVk5OTonlExISFBAQ4B5cAQcANwfrV8HFx8fL5XK5x/Hjx223BACoBj7/P6CgoCDVr19fp06d8ph+6tQphYSElFje6XTK6XT6ug0AQA3n8yOgRo0aqV+/fkpOTnZPKy4uVnJysqKiony9OgBALVUld0KYO3euJk+erDvuuEMDBgzQ4sWLVVBQoMcff7wqVgcAqIWqJIAeeeQRff/995o/f75ycnLUt29fbdmypcSFCQCAm1eN+zygvLw8BQQE2G4DAHCDXC6X/P39y5xv/So4AMDNiQACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWNLDdAFCe22+/3euaDRs2VGpdnTp1qlQdKmfkyJFe1xw+fNjrmuPHj3tdg6rHERAAwAoCCABghc8DaOHChXI4HB6jW7duvl4NAKCWq5JzQD169NBf/vKXf66kAaeaAACeqiQZGjRooJCQkKp4agBAHVEl54COHDmi0NBQde7cWZMmTVJWVlaZyxYWFiovL89jAADqPp8HUGRkpJKSkrRlyxYtXbpUmZmZGjx4sPLz80tdPiEhQQEBAe7RoUMHX7cEAKiBfB5AsbGxeuihh9S7d2/FxMTok08+UW5urt5///1Sl4+Pj5fL5XIPrtcHgJtDlV8d0KJFC916661KT08vdb7T6ZTT6azqNgAANUyV/x/QuXPnlJGRobZt21b1qgAAtYjPA+iZZ57Rzp07dfToUf3tb3/T/fffr/r162vChAm+XhUAoBbz+VtwJ06c0IQJE3TmzBm1bt1ad955p3bv3q3WrVv7elUAgFrM5wH03nvv+fopcZOLiYnxuobzirXDmDFjvK75+c9/7nXN+PHjva5B1eNecAAAKwggAIAVBBAAwAoCCABgBQEEALCCAAIAWEEAAQCsIIAAAFYQQAAAKwggAIAVBBAAwAoCCABgRZV/IB3wUw0aeL/LjR49ugo6QU2wb98+r2vmzp3rdU2zZs28rpGkgoKCStWhYjgCAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIAGAFAQQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBXcDRvVatiwYV7XREVFeV2zaNEir2tQ/Vq2bOl1zW233eZ1TdOmTb2ukbgbdlXjCAgAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArOBmpKi0nj17el2zbt06r2syMjK8rvn3f/93r2tQ/e677z7bLcAijoAAAFYQQAAAKwggAIAVBBAAwAoCCABgBQEEALCCAAIAWEEAAQCsIIAAAFYQQAAAKwggAIAVBBAAwApuRopKe+GFF7yuadasmdc1o0aN8rrm3LlzXtfgxgQGBnpdM3ToUK9riouLva5BzcQREADACgIIAGCF1wG0a9cujRkzRqGhoXI4HNq0aZPHfGOM5s+fr7Zt26pJkyYaMWKEjhw54qt+AQB1hNcBVFBQoD59+igxMbHU+YsWLdKSJUu0bNkyffHFF2rWrJliYmJ08eLFG24WAFB3eH0RQmxsrGJjY0udZ4zR4sWL9cILL7g/6XDVqlUKDg7Wpk2bNH78+BvrFgBQZ/j0HFBmZqZycnI0YsQI97SAgABFRkYqNTW11JrCwkLl5eV5DABA3efTAMrJyZEkBQcHe0wPDg52z7tWQkKCAgIC3KNDhw6+bAkAUENZvwouPj5eLpfLPY4fP267JQBANfBpAIWEhEiSTp065TH91KlT7nnXcjqd8vf39xgAgLrPpwEUHh6ukJAQJScnu6fl5eXpiy++UFRUlC9XBQCo5by+Cu7cuXNKT093P87MzNSBAwcUGBiojh07avbs2Xr11Vd1yy23KDw8XC+++KJCQ0M1duxYX/YNAKjlvA6gvXv3atiwYe7Hc+fOlSRNnjxZSUlJmjdvngoKCjRt2jTl5ubqzjvv1JYtW9S4cWPfdQ0AqPUcxhhju4mfysvLU0BAgO02bioPPvhgper+8Ic/eF1z7Ngxr2t69erldQ2q33/8x394XTN79myva1JSUryuqcwNbSXp8uXLlarDFS6X67rn9a1fBQcAuDkRQAAAKwggAIAVBBAAwAoCCABgBQEEALCCAAIAWEEAAQCsIIAAAFYQQAAAKwggAIAVBBAAwAoCCABghdcfx4C656GHHqpUXdOmTb2ueeuttyq1LlSvTp06eV0zadIkr2uKioq8rnn11Ve9ruGu1jUTR0AAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAU3I61jAgICvK752c9+VgWdlG7p0qXVti5U3rRp07yuCQoK8rrm8OHDXtfs2LHD6xrUTBwBAQCsIIAAAFYQQAAAKwggAIAVBBAAwAoCCABgBQEEALCCAAIAWEEAAQCsIIAAAFYQQAAAKwggAIAV3Iy0jnE6nV7XtGvXrlLrWrduXaXqUPNFRERUy3oOHTpULetBzcQREADACgIIAGAFAQQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIAGAFAQQAsIIAAgBYwc1I65j8/Hyvaw4cOFCpdfXu3dvrmsDAQK9rzp4963UNrmjTpk2l6h588EEfd1K6zz77rFrWg5qJIyAAgBUEEADACq8DaNeuXRozZoxCQ0PlcDi0adMmj/lTpkyRw+HwGKNGjfJVvwCAOsLrACooKFCfPn2UmJhY5jKjRo1Sdna2e/DBZQCAa3l9EUJsbKxiY2Ovu4zT6VRISEilmwIA1H1Vcg4oJSVFbdq0UdeuXfXUU0/pzJkzZS5bWFiovLw8jwEAqPt8HkCjRo3SqlWrlJycrNdff107d+5UbGysioqKSl0+ISFBAQEB7tGhQwdftwQAqIF8/n9A48ePd3/dq1cv9e7dWxEREUpJSdHw4cNLLB8fH6+5c+e6H+fl5RFCAHATqPLLsDt37qygoCClp6eXOt/pdMrf399jAADqvioPoBMnTujMmTNq27ZtVa8KAFCLeP0W3Llz5zyOZjIzM3XgwAEFBgYqMDBQL730ksaNG6eQkBBlZGRo3rx56tKli2JiYnzaOACgdvM6gPbu3athw4a5H189fzN58mQtXbpUBw8e1DvvvKPc3FyFhoZq5MiReuWVV+R0On3XNQCg1vM6gKKjo2WMKXP+1q1bb6gh3JgLFy54XZORkVGpdY0bN87rmo8//tjrmjfeeMPrmpquZ8+eXtd07tzZ65pOnTp5XSPpuj/jvlRcXFwt60HNxL3gAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYIXDVNdtbysoLy9PAQEBttu4qXTr1q1SdS+//LLXNffcc4/XNXXxozxOnz7tdU1lflSDgoK8rpEkh8NRqTpv+fn5eV1TmTu+ww6Xy3XdT7nmCAgAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArOBmpKhWffv29bqmS5cuvm/Esg8++KBa1vPOO+9Uqm7SpEk+7qR0DRo0qJb1wA5uRgoAqJEIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAV3AkS1OnDgQLXU4Ir/+7//s93CdfXs2dPrmkOHDlVBJ7CBIyAAgBUEEADACgIIAGAFAQQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIAGAFAQQAsIKbkQJ1mMPhqNY6b3Fj0ZsbR0AAACsIIACAFV4FUEJCgvr37y8/Pz+1adNGY8eOVVpamscyFy9eVFxcnFq1aqXmzZtr3LhxOnXqlE+bBgDUfl4F0M6dOxUXF6fdu3dr27Ztunz5skaOHKmCggL3MnPmzNGf//xnrV+/Xjt37tTJkyf1wAMP+LxxAEDt5tVFCFu2bPF4nJSUpDZt2mjfvn0aMmSIXC6XVqxYobVr1+quu+6SJK1cuVLdu3fX7t279bOf/cx3nQMAarUbOgfkcrkkSYGBgZKkffv26fLlyxoxYoR7mW7duqljx45KTU0t9TkKCwuVl5fnMQAAdV+lA6i4uFizZ8/WoEGD3J/rnpOTo0aNGqlFixYeywYHBysnJ6fU50lISFBAQIB7dOjQobItAQBqkUoHUFxcnA4dOqT33nvvhhqIj4+Xy+Vyj+PHj9/Q8wEAaodK/SPqjBkz9NFHH2nXrl1q3769e3pISIguXbqk3Nxcj6OgU6dOKSQkpNTncjqdcjqdlWkDAFCLeXUEZIzRjBkztHHjRm3fvl3h4eEe8/v166eGDRsqOTnZPS0tLU1ZWVmKioryTccAgDrBqyOguLg4rV27Vps3b5afn5/7vE5AQICaNGmigIAA/eIXv9DcuXMVGBgof39/zZw5U1FRUVwBBwDw4FUALV26VJIUHR3tMX3lypWaMmWKJOnNN99UvXr1NG7cOBUWFiomJkZvvfWWT5oFANQdXgWQMabcZRo3bqzExEQlJiZWuikAvlGRn1lf1gHe4F5wAAArCCAAgBUEEADACgIIAGAFAQQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIAGAFAQQAsKJSn4gKoHZo3Lhxta3rwoUL1bYu1A0cAQEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFdyMFKjDHn/88UrV5ebmel3zyiuvVGpduHlxBAQAsIIAAgBYQQABAKwggAAAVhBAAAArCCAAgBUEEADACgIIAGAFAQQAsIIAAgBYQQABAKwggAAAVnAzUqAO27NnT6Xq3njjDa9rduzYUal14ebFERAAwAoCCABgBQEEALCCAAIAWEEAAQCsIIAAAFYQQAAAKwggAIAVBBAAwAoCCABgBQEEALCCAAIAWOEwxhjbTfxUXl6eAgICbLcBALhBLpdL/v7+Zc7nCAgAYAUBBACwwqsASkhIUP/+/eXn56c2bdpo7NixSktL81gmOjpaDofDY0yfPt2nTQMAaj+vAmjnzp2Ki4vT7t27tW3bNl2+fFkjR45UQUGBx3JTp05Vdna2eyxatMinTQMAaj+vPhF1y5YtHo+TkpLUpk0b7du3T0OGDHFPb9q0qUJCQnzTIQCgTrqhc0Aul0uSFBgY6DH93XffVVBQkHr27Kn4+HidP3++zOcoLCxUXl6exwAA3ARMJRUVFZl77rnHDBo0yGP68uXLzZYtW8zBgwfNmjVrTLt27cz9999f5vMsWLDASGIwGAxGHRsul+u6OVLpAJo+fboJCwszx48fv+5yycnJRpJJT08vdf7FixeNy+Vyj+PHj1vfaAwGg8G48VFeAHl1DuiqGTNm6KOPPtKuXbvUvn376y4bGRkpSUpPT1dERESJ+U6nU06nszJtAABqMa8CyBijmTNnauPGjUpJSVF4eHi5NQcOHJAktW3btlINAgDqJq8CKC4uTmvXrtXmzZvl5+ennJwcSVJAQICaNGmijIwMrV27VqNHj1arVq108OBBzZkzR0OGDFHv3r2r5BsAANRS3pz3URnv861cudIYY0xWVpYZMmSICQwMNE6n03Tp0sU8++yz5b4P+FMul8v6+5YMBoPBuPFR3u9+bkYKAKgS3IwUAFAjEUAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABW1LgAMsbYbgEA4APl/T6vcQGUn59vuwUAgA+U9/vcYWrYIUdxcbFOnjwpPz8/ORwOj3l5eXnq0KGDjh8/Ln9/f0sd2sd2uILtcAXb4Qq2wxU1YTsYY5Sfn6/Q0FDVq1f2cU6DauypQurVq6f27dtfdxl/f/+bege7iu1wBdvhCrbDFWyHK2xvh4CAgHKXqXFvwQEAbg4EEADAiloVQE6nUwsWLJDT6bTdilVshyvYDlewHa5gO1xRm7ZDjbsIAQBwc6hVR0AAgLqDAAIAWEEAAQCsIIAAAFYQQAAAK2pNACUmJqpTp05q3LixIiMj9eWXX9puqdotXLhQDofDY3Tr1s12W1Vu165dGjNmjEJDQ+VwOLRp0yaP+cYYzZ8/X23btlWTJk00YsQIHTlyxE6zVai87TBlypQS+8eoUaPsNFtFEhIS1L9/f/n5+alNmzYaO3as0tLSPJa5ePGi4uLi1KpVKzVv3lzjxo3TqVOnLHVcNSqyHaKjo0vsD9OnT7fUcelqRQD98Y9/1Ny5c7VgwQJ99dVX6tOnj2JiYvTdd9/Zbq3a9ejRQ9nZ2e7x2Wef2W6pyhUUFKhPnz5KTEwsdf6iRYu0ZMkSLVu2TF988YWaNWummJgYXbx4sZo7rVrlbQdJGjVqlMf+sW7dumrssOrt3LlTcXFx2r17t7Zt26bLly9r5MiRKigocC8zZ84c/fnPf9b69eu1c+dOnTx5Ug888IDFrn2vIttBkqZOneqxPyxatMhSx2UwtcCAAQNMXFyc+3FRUZEJDQ01CQkJFruqfgsWLDB9+vSx3YZVkszGjRvdj4uLi01ISIj5zW9+456Wm5trnE6nWbdunYUOq8e128EYYyZPnmzuu+8+K/3Y8t133xlJZufOncaYK699w4YNzfr1693LHD582EgyqampttqsctduB2OMGTp0qHn66aftNVUBNf4I6NKlS9q3b59GjBjhnlavXj2NGDFCqampFjuz48iRIwoNDVXnzp01adIkZWVl2W7JqszMTOXk5HjsHwEBAYqMjLwp94+UlBS1adNGXbt21VNPPaUzZ87YbqlKuVwuSVJgYKAkad++fbp8+bLH/tCtWzd17NixTu8P126Hq959910FBQWpZ8+eio+P1/nz5220V6Yadzfsa50+fVpFRUUKDg72mB4cHKxvvvnGUld2REZGKikpSV27dlV2drZeeuklDR48WIcOHZKfn5/t9qzIycmRpFL3j6vzbhajRo3SAw88oPDwcGVkZOj5559XbGysUlNTVb9+fdvt+VxxcbFmz56tQYMGqWfPnpKu7A+NGjVSixYtPJaty/tDadtBkiZOnKiwsDCFhobq4MGDeu6555SWlqYNGzZY7NZTjQ8g/FNsbKz76969eysyMlJhYWF6//339Ytf/MJiZ6gJxo8f7/66V69e6t27tyIiIpSSkqLhw4db7KxqxMXF6dChQzfFedDrKWs7TJs2zf11r1691LZtWw0fPlwZGRmKiIio7jZLVePfggsKClL9+vVLXMVy6tQphYSEWOqqZmjRooVuvfVWpaen227Fmqv7APtHSZ07d1ZQUFCd3D9mzJihjz76SDt27PD4/LCQkBBdunRJubm5HsvX1f2hrO1QmsjISEmqUftDjQ+gRo0aqV+/fkpOTnZPKy4uVnJysqKioix2Zt+5c+eUkZGhtm3b2m7FmvDwcIWEhHjsH3l5efriiy9u+v3jxIkTOnPmTJ3aP4wxmjFjhjZu3Kjt27crPDzcY36/fv3UsGFDj/0hLS1NWVlZdWp/KG87lObAgQOSVLP2B9tXQVTEe++9Z5xOp0lKSjL/+7//a6ZNm2ZatGhhcnJybLdWrX71q1+ZlJQUk5mZaT7//HMzYsQIExQUZL777jvbrVWp/Px8s3//frN//34jybzxxhtm//795tixY8YYY1577TXTokULs3nzZnPw4EFz3333mfDwcHPhwgXLnfvW9bZDfn6+eeaZZ0xqaqrJzMw0f/nLX8ztt99ubrnlFnPx4kXbrfvMU089ZQICAkxKSorJzs52j/Pnz7uXmT59uunYsaPZvn272bt3r4mKijJRUVEWu/a98rZDenq6efnll83evXtNZmam2bx5s+ncubMZMmSI5c491YoAMsaY//zP/zQdO3Y0jRo1MgMGDDC7d++23VK1e+SRR0zbtm1No0aNTLt27cwjjzxi0tPTbbdV5Xbs2GEklRiTJ082xly5FPvFF180wcHBxul0muHDh5u0tDS7TVeB622H8+fPm5EjR5rWrVubhg0bmrCwMDN16tQ690daad+/JLNy5Ur3MhcuXDC//OUvTcuWLU3Tpk3N/fffb7Kzs+01XQXK2w5ZWVlmyJAhJjAw0DidTtOlSxfz7LPPGpfLZbfxa/B5QAAAK2r8OSAAQN1EAAEArCCAAABWEEAAACsIIACAFQQQAMAKAggAYAUBBACwggACAFhBAAEArCCAAABW/H/x+5M7SsBKNgAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [],
      "metadata": {
        "id": "a53rvfCCGNKZ"
      },
      "execution_count": 15,
      "outputs": []
    }
  ]
}