Deep_Learning_Intro_ao_Keras.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Deep_Learning_Intro_ao_Keras",
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MEsC4zB7bBeG"
      },
      "source": [
        "#Aula 01"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GWZi-kaGGLrF"
      },
      "source": [
        "Repositório com informações sobre o dataset:\n",
        "\n",
        "https://github.com/zalandoresearch/fashion-mnist"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SbwAwfVXa3sC"
      },
      "source": [
        "import tensorflow\n",
        "from tensorflow import keras\n",
        "\n",
        "dataset = keras.datasets.fashion_mnist"
      ],
      "execution_count": 1,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GnBIeoNqcaGz"
      },
      "source": [
        "((imagens_treino,identificacoes_treino),(imagens_teste,identificacoes_teste)) = dataset.load_data()"
      ],
      "execution_count": 2,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PyKXGb1xc7vQ",
        "outputId": "d08b7be4-48fc-4fb5-8c27-43f8a9684afa"
      },
      "source": [
        "imagens_treino"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]],\n",
              "\n",
              "       [[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]],\n",
              "\n",
              "       [[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]],\n",
              "\n",
              "       ...,\n",
              "\n",
              "       [[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]],\n",
              "\n",
              "       [[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]],\n",
              "\n",
              "       [[0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        ...,\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0],\n",
              "        [0, 0, 0, ..., 0, 0, 0]]], dtype=uint8)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2rCLQ2JxdhGY",
        "outputId": "1b8e231b-2179-41cc-ced6-bb54b0e88797"
      },
      "source": [
        "len(imagens_treino)\n",
        "imagens_treino.shape"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(60000, 28, 28)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 4
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s09xIvgsd7d7",
        "outputId": "a48bf8a5-05f2-4aed-92f5-556283ba180f"
      },
      "source": [
        "identificacoes_teste.shape"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(10000,)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "hY8lzS1peCjC",
        "outputId": "4d6bc752-4624-4d8f-f917-c1cc34f11b47"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(imagens_treino[0])"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f32eb161a50>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uXgjk-U9EdF1"
      },
      "source": [
        "#Aula 02"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_aRXJo7-ea78",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 299
        },
        "outputId": "67a0eb4c-1340-4906-a7b0-e800535ce539"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "plt.imshow(imagens_treino[0])\n",
        "plt.title(identificacoes_treino[0])"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, '9')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 7
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4Zn4_A_lEwKh",
        "outputId": "243af685-f6d2-46a4-fdc3-e156aea13bd1"
      },
      "source": [
        "print(\"O menor argumento é %.d e o maior é %.d\" % (identificacoes_treino.min(),identificacoes_treino.max()))"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "O menor argumento é 0 e o maior é 9\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HIfCdVcIE5Uk"
      },
      "source": [
        "total_de_classificacoes = identificacoes_treino.max()-identificacoes_treino.min()+1"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 240
        },
        "id": "uOm4dE0ZFBuu",
        "outputId": "f788d6ad-e474-41d0-9fc7-1cebacaee1e9"
      },
      "source": [
        "for imagem in range(10):\n",
        "  plt.subplot(2,5,imagem+1)\n",
        "  plt.imshow(imagens_treino[imagem])\n",
        "  plt.title(identificacoes_treino[imagem])\n",
        "  "
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "T8LIRiTqFvZG"
      },
      "source": [
        "nomes_de_classificacoes = ['Camiseta','Calca','Pullover',\n",
        "                           'Vestido','Casaco','Sandália',\n",
        "                           'Camisa','Tenis','Bolsa','Bota']"
      ],
      "execution_count": 11,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 240
        },
        "id": "oqImm8oNGKV9",
        "outputId": "432d3fb6-54d4-4b83-8de9-f637ca12aa6e"
      },
      "source": [
        "for imagem in range(10):\n",
        "  plt.subplot(2,5,imagem+1)\n",
        "  plt.imshow(imagens_treino[imagem])\n",
        "  plt.title(nomes_de_classificacoes[identificacoes_treino[imagem]])\n",
        "  "
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 10 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "GvO0MACSH2ii"
      },
      "source": [
        "## Criando um modelo"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cVGjxNjZGqGT"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28))     \n",
        "     #processamento\n",
        "     #saida \n",
        "     ])"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1RPjfHmaKHFe"
      },
      "source": [
        "Uma biblioteca para pré-processamento da imagem é a ndimage.\n",
        "\n",
        "`from scipy import ndimage`"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kYvOBTYiKaQk"
      },
      "source": [
        "#Aula 03"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CjQ1Qro6KOeJ"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "#modelo.fit(imagens_treino,identificacoes_treino)"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HyBWd44ZTD2-"
      },
      "source": [
        "Treinar este modelo não é possível sem antes compilar. Seguimos para a próxima aula."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "naoJkClkPlXA"
      },
      "source": [
        "#Aula 04"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BZddAw90PmoZ",
        "outputId": "41da411d-2105-4cf9-88f1-ad9518fb4eb2"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy')\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1875/1875 [==============================] - 6s 3ms/step - loss: 12.0310\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32e6e88e90>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "W24YSTaZQSOH"
      },
      "source": [
        "Verificando se há cor na imagem"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 283
        },
        "id": "nXMDxd4gPvXp",
        "outputId": "2967e45d-59f7-4792-a5b8-0a30176eb5d6"
      },
      "source": [
        "plt.imshow(imagens_treino[0])\n",
        "plt.colorbar()  "
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.colorbar.Colorbar at 0x7f32e46f6250>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SmJFtZKFQ9ac"
      },
      "source": [
        "Vamos tratar a imagem pois há muita informação de cor"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NUsy7M41Q7rP",
        "outputId": "ffb4dd41-4d2f-4254-c4c3-56e3bb290722"
      },
      "source": [
        "imagens_treino = imagens_treino/float(255)\n",
        "\n",
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy')\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.6130\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32e469b2d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ea1FmdP7RemJ"
      },
      "source": [
        "Houve uma perda bem menor do que o exemplo anterior. Isto se deve ao fato de que as imagens de treino estão normalizados."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "aNkK-fJBR7Of"
      },
      "source": [
        "Vamos testar uma nova camada"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MK57szZqRc7M",
        "outputId": "81859d82-2f81-45a7-b267-91a93b8765f6"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(128,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(64,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy')\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino)"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1875/1875 [==============================] - 7s 4ms/step - loss: 0.6289\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32e459db10>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "u9MLOursSqvy"
      },
      "source": [
        "Incrementar as camadas gerou uma perda maior. Deve-se continuar apenas com 1 camada oculta."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "29gXQMZ3Sv83",
        "outputId": "a0566007-3c5b-4c6d-a949-eb6d1531b4f4"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy')\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino)"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.6149\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32e3474210>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ne64suM7WM1G"
      },
      "source": [
        "#Aula 05"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qFXT6pJpTSGG",
        "outputId": "72200078-077c-451a-c5e1-3ca7df8a6286"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino,epochs=5)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.6037 - accuracy: 0.7905\n",
            "Epoch 2/5\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.3682 - accuracy: 0.8678\n",
            "Epoch 3/5\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.3203 - accuracy: 0.8836\n",
            "Epoch 4/5\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.3030 - accuracy: 0.8885\n",
            "Epoch 5/5\n",
            "1875/1875 [==============================] - 6s 3ms/step - loss: 0.2860 - accuracy: 0.8952\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32eaa4c450>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "myT6MdaaXjxl"
      },
      "source": [
        "Utilizando o modelo para testar"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "3OcIletkXZD4",
        "outputId": "b4f44aae-80b8-4390-ba20-f1b891c1ff43"
      },
      "source": [
        "import numpy as np\n",
        "\n",
        "testes = modelo.predict(imagens_teste/float(255))\n",
        "\n",
        "print('Resultado do teste:',np.argmax(testes[0]))\n",
        "print('Número da Imagem de Teste:',identificacoes_teste[0])"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Resultado do teste: 9\n",
            "Número da Imagem de Teste: 9\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "99DG2rUvYKUn",
        "outputId": "46b13197-f2dd-4f77-db98-6ff51668eaec"
      },
      "source": [
        "print('Resultado do teste:',np.argmax(testes[1]))\n",
        "print('Número da Imagem de Teste:',identificacoes_teste[1])"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Resultado do teste: 2\n",
            "Número da Imagem de Teste: 2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qBdStzVdYcxe"
      },
      "source": [
        "imagens_teste = imagens_teste/float(255)"
      ],
      "execution_count": 23,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pcpgaQyNYOhb",
        "outputId": "be18c03e-82db-4d9c-f26c-74d879c2bde4"
      },
      "source": [
        "perda_teste, acuracia_teste = modelo.evaluate(imagens_teste,identificacoes_teste)\n",
        "print('Perda do teste:',perda_teste)\n",
        "print('Acurácia do teste:',acuracia_teste)"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 1s 1ms/step - loss: 0.3464 - accuracy: 0.8733\n",
            "Perda do teste: 0.3464018702507019\n",
            "Acurácia do teste: 0.8733000159263611\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BGPVhOUNY55g",
        "outputId": "e37fd57e-ef42-4773-d1d1-77bfca4b05da"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "modelo.fit(imagens_treino,identificacoes_treino,epochs=5, validation_split=0.2)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.6367 - accuracy: 0.7770 - val_loss: 0.3903 - val_accuracy: 0.8621\n",
            "Epoch 2/5\n",
            "1500/1500 [==============================] - 4s 3ms/step - loss: 0.3813 - accuracy: 0.8612 - val_loss: 0.4164 - val_accuracy: 0.8475\n",
            "Epoch 3/5\n",
            "1500/1500 [==============================] - 4s 3ms/step - loss: 0.3427 - accuracy: 0.8717 - val_loss: 0.3466 - val_accuracy: 0.8767\n",
            "Epoch 4/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3115 - accuracy: 0.8855 - val_loss: 0.3641 - val_accuracy: 0.8748\n",
            "Epoch 5/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.2903 - accuracy: 0.8933 - val_loss: 0.3203 - val_accuracy: 0.8853\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7f32e65a7610>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Fwxtp_A4a75P"
      },
      "source": [
        "#Aula 06"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8lhmBq1ba1-Y",
        "outputId": "c6513290-620f-458a-9366-7aed15150447"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "historico = modelo.fit(imagens_treino,identificacoes_treino,epochs=5, validation_split=0.2)"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.6318 - accuracy: 0.7782 - val_loss: 0.4233 - val_accuracy: 0.8508\n",
            "Epoch 2/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3841 - accuracy: 0.8589 - val_loss: 0.3924 - val_accuracy: 0.8508\n",
            "Epoch 3/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3340 - accuracy: 0.8784 - val_loss: 0.3430 - val_accuracy: 0.8788\n",
            "Epoch 4/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3140 - accuracy: 0.8827 - val_loss: 0.3548 - val_accuracy: 0.8744\n",
            "Epoch 5/5\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.2878 - accuracy: 0.8911 - val_loss: 0.3392 - val_accuracy: 0.8778\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "qmLM_nhObBfp",
        "outputId": "e1778451-425a-4d53-f1f7-6e422bab8c0f"
      },
      "source": [
        "plt.plot(historico.history['accuracy'])\n",
        "plt.plot(historico.history['val_accuracy'])\n",
        "plt.title(\"Acurácia por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Acurácia')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32eab76cd0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 27
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "GAhqKmpab9Ug",
        "outputId": "35baf1a8-0d7d-4301-e826-5f90bb4c167b"
      },
      "source": [
        "plt.plot(historico.history['loss'])\n",
        "plt.plot(historico.history['val_loss'])\n",
        "plt.title(\"Perdas por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Perda')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32ea9d8750>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ktbvGPIxcvF7",
        "outputId": "d38c3f11-a054-49d0-8202-2c98029f2f76"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "historico = modelo.fit(imagens_treino,identificacoes_treino,epochs=10, validation_split=0.2)"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/10\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.6256 - accuracy: 0.7833 - val_loss: 0.4080 - val_accuracy: 0.8552\n",
            "Epoch 2/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3872 - accuracy: 0.8607 - val_loss: 0.3566 - val_accuracy: 0.8698\n",
            "Epoch 3/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3360 - accuracy: 0.8758 - val_loss: 0.3416 - val_accuracy: 0.8769\n",
            "Epoch 4/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.3123 - accuracy: 0.8844 - val_loss: 0.3310 - val_accuracy: 0.8822\n",
            "Epoch 5/10\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.2924 - accuracy: 0.8893 - val_loss: 0.3371 - val_accuracy: 0.8827\n",
            "Epoch 6/10\n",
            "1500/1500 [==============================] - 7s 4ms/step - loss: 0.2745 - accuracy: 0.8988 - val_loss: 0.3383 - val_accuracy: 0.8810\n",
            "Epoch 7/10\n",
            "1500/1500 [==============================] - 5s 4ms/step - loss: 0.2620 - accuracy: 0.9033 - val_loss: 0.3253 - val_accuracy: 0.8857\n",
            "Epoch 8/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.2433 - accuracy: 0.9077 - val_loss: 0.3264 - val_accuracy: 0.8818\n",
            "Epoch 9/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.2407 - accuracy: 0.9086 - val_loss: 0.3172 - val_accuracy: 0.8879\n",
            "Epoch 10/10\n",
            "1500/1500 [==============================] - 5s 3ms/step - loss: 0.2280 - accuracy: 0.9152 - val_loss: 0.3215 - val_accuracy: 0.8878\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "Vrz9GNnocxc1",
        "outputId": "d114f3d8-0a36-4463-d4d2-133458d5ff95"
      },
      "source": [
        "plt.plot(historico.history['accuracy'])\n",
        "plt.plot(historico.history['val_accuracy'])\n",
        "plt.title(\"Acurácia por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Acurácia')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32dfa14d50>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEaCAYAAAAL7cBuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXhU5fn/8fedBQIhCwlbSAgBWcISwhIERRTBBZFFVFqpG7ZqXerW5fe11n5Fu9l+rdW2VusGalEqKCqI4oILiiIJS9gRIZANSICQhJBlkvv3xxkgxAQGyORkuV/XNVdmzjb3HGU+8zznnOeIqmKMMcbUFOB2AcYYYxonCwhjjDG1soAwxhhTKwsIY4wxtbKAMMYYUysLCGOMMbWygDCmCRORFBGZ6nYdpnmygDCmiRKRzsATwFdu12KaJwsI02KISLCIrBGRy31c/j0RudHfdZ0uVd2jquep6m63azHNk9iV1MZNIvIpkAx0UdUyP7/X/wKtVPVBf76Pv4nIGGApUFJj1sWqaq0JU2+sBWFcIyIJwGhAgcl+2L6ISID3eSBQAPxvfb+PP4lIUB2zclS1XY2HhYOpVxYQxk03AF8Ds4HjunJEpJuIvCkieSKyT0T+6Z0+U0T+U225BBHRI1+kIvKpiPxBRL7E+YXdU0RuAtYDfwC2ichPa7zXFG/XU6GIfCci46tt62bv87NEZKm3lnwRmSMikXV9MG9Nd4vIdu/y/1ctrAJE5EER2Skie0XkZRGJqPF5fiIiu3BaCqfEW/efROQb72d6W0Siqs2fLCIbRKTAu2w/H/b7CT+/iPyPiGSLSJGIbBGRcadat2l8LCCMm24A5ngfl3oPuh75tb8I2AkkALHA3FPY7vXArUCYdxv5wEQgHLgJ+JuIDPW+19nAy8CvgEjgfCCjlm0K8CegK9AP6AbMPEkdU4EUYCgwBfixd/oM7+NCoCfQDvhnjXUv8L7PpSd5j7rc4H2/GMAD/B1ARPoArwH3Ah2BxcBCEWl1kv1e5+cXkb7Az4DhqhrmrTnjNOs2jYmq2sMeDf4AzgMqgA7e15uB+7zPzwHygKBa1psJ/Kfa6wScLqog7+tPgUdO8t5vAfd4n/8b+Fsdy30K3FzHvCuA1Sd4DwXGV3t9B/Cx9/nHwB3V5vX17ougap+n5wm2PQaowukyq/4IrVb3o9WW7w+UA4HAb4HXq80LALK926xzv5/o8wO9gL3ARUCw2/9v2aP+HtaCMG65EfhAVfO9r1/lWDdTN2CnqnpOc9uZ1V+IyDhvV8ouEcnA+SLrUO29vjvZBkWks4jM9XajFAL/qbYNX+rYifPrG+/fnTXmBQGd6/oMtchR1cgaj0MneO9gb73HvbeqVnmXjeUE+/1En19Vt+G0SGYCe73Lda25DdP0WECYBicibYAfABeIyG4R2Q3cBySLSDLOF1Z8HQdoDwFtq73uUssyR0/NE5FWwNvAX4HuqpqA8wtevItkAmf5UPYfvdtNUtVw4Lpq26hLt2rP44Ec7/McoHuNeR5gT22f4TTVfO8KnK62495bRMS7bDYn3u8n/Pyq+qqqnufdtgJ/PsP6TSNgAWHccAVQidP1Mdj76Acsw+k7/wbIBR4VkVARCRGRUd511wDni0i898Dur0/yXq2BNjjBgohcBlxcbf4LwE3eVkaAiMSKSGIt2wkDioGDIhKLc8ziZH4lIu1FpBtwD/Bf7/TXgPtEpIeItMP58v3vGbSYanOdiPQXkbbAI8B8Va0EXgcu937eYOAXQBmwnBPv9zo/v4j0FZGxItIaKAUO43SBmSbOAsK44UZglqruUtXdRx44B2qvxfllOgmnb3sXkAX8EEBVP8T5ok0H0nAOqtZJVYuAu3G+lA8APwLeqTb/G7wHroGDwGcc/+v+iIdxDjYfBN4F3vThc77trXGNd50XvNNfBF4BPgd24Hyp3uXD9qrrKiLFNR5XVZv/Cs7ZYbuBEJx9gKpuwfn1/w+cFsUkYJKqlnsDpNb9fpLP3xp41Lu93UAnTh7cpgmwC+WM8QMRUaC3t3++od/7U5wD+c839Hub5sVaEMYYY2plAWGMMaZW1sVkjDGmVtaCMMYYU6u6BgJrcjp06KAJCQlul2GMMU1KWlpavqp2rG1eswmIhIQEUlNT3S7DGGOaFBHZWdc862IyxhhTKwsIY4wxtbKAMMYYU6tmcwyiNhUVFWRlZVFaWup2KU1eSEgIcXFxBAcHu12KMaaBNOuAyMrKIiwsjISEBJxBK83pUFX27dtHVlYWPXr0cLscY0wDadZdTKWlpURHR1s4nCERITo62lpixrQwzTogAAuHemL70ZiWp1l3MRljTHOVX1zG5twiNuUWEhMZwsRB9X8TPwsIP9q3bx/jxo0DYPfu3QQGBtKxo3PB4jfffEOrVq3qXPeZZ56hbdu23HDDDQ1SqzGmcSr3VLE9v5hNuYVszi1iY24hm3cXkVdUdnSZycldLSCamujoaNasWQPAzJkzadeuHb/85S+Pzvd4PAQF1f6f4LbbbmuQGo0xjUd+cdnRINiUW8im3UVs21tERaUzqGqrwAB6d27H+b070i8mjH4x4SR2CSO6XWu/1GMB0cBmzJhBSEgIq1evZtSoUdx5553ceeed5OXl0bZtW5577jkSExOPC5QxY8YwYsQIPvnkEwoKCnjhhRcYPXo0paWl3H777aSmphIUFMTjjz/OhRde6PZHNMacRLmniu/yitm8u5BNR8Igt4j84mOtgs7hrUnsEs4FfY6FQY8OoQQHNtyh4xYTEA8v3MDGnMJ63Wb/ruE8NGnAKa+XlZXF8uXLCQwMZNy4cTzzzDP07t2bFStWcMcdd7B06dLvrePxePjmm29YvHgxDz/8MB999BFPPfUUIsK6devYvHkzl1xyCVu3biUkJKQ+Pp4xph7kFZV5g+BYF9F3ecXHWgVBAfTp3I4xfTvSLyacfl3CSIwJJyq07i7ohtJiAqIxmTZtGoGBgRQXF7N8+XKmTZt2dF5ZWVmt61x55ZUADBs2jIyMDAC++OIL7rrLuZVxYmIi3bt3Z+vWrQwaNMi/H8AY8z1HWgWbvMcIamsVdAkPITEmjAsTO5HYJYz+3lZBUAO2Ck5FiwmI0/ml7y+hoaEAVFVVERkZefQ4xYm0bu30MQYGBuLxePxanzHmxFSV9KyDrNix72gXUW2tggv7diQxJpx+MWEkdmkcrYJT0WICojEKDw+nR48ezJs3j2nTpjn/06Wnk5yc7NP6o0ePZs6cOYwdO5atW7eya9cu+vbt6+eqjWmZVJVNuUUsTM9hUXoOmfsPA06roJ+3VXCki6gxtwpOhQWEy+bMmcPtt9/O73//eyoqKrjmmmt8Dog77riD22+/naSkJIKCgpg9e/bRloYxpn5s21vEwrW5LEzPYXveIQIDhFG9OnDX2N6MS+zktzOIGoNmc0/qlJQUrXnDoE2bNtGvXz+XKmp+bH+almLnvkMsSs9l4docNu8uQgRG9ohmYnIM4wd0aVahICJpqppS2zxrQRhjDJBdcJh303NYlJ5LetZBAIZ1b8/MSf2ZkBRDp/CWd3agBYQxpsXaW1jK4nW5LEzPJW3nAQAGxUXwwIRELh/UldjINi5X6C4LCGNMi7L/UDnvrXe6j1bs2I8qJHYJ41eX9mXioBi6R4e6XWKjYQFhjGn2Dh6uYMmG3SxKz+XLbflUVik9O4Zy99jeTEqOoVenMLdLbJQsIIwxzVJxmYePNu5hUXoOn2/Np7yyim5Rbfjp+T2ZOKgr/WLCbBj7k7CAMMY0G4fLK/lky14Wrs1h6ea9lHmqiIkI4cZzuzNxUFcGxUVYKJyCpn8lRyN24YUXsmTJkuOmPfHEE9x+++21Lj9mzBiOnKo7YcIECgoKvrfMzJkzeeyxx06rnieeeIKRI0cybdo01q1bd1rbMKaxKfNU8uHGPdwzdzXDfv8hd8xZxcqMA1wzvBvzbzuHL/9nLL+5vD/J3SItHE6RtSD8aPr06cydO5dLL7306LS5c+fyl7/85aTrLl68uN7ruffee7n33nvrfbvGNLSKyiq+3JbPovRclmzYTVGph/Ztg5kyOJZJg2IY0TOawAALgzNlLQg/uvrqq3n33XcpLy8HICMjg5ycHF577TVSUlIYMGAADz30UK3rJiQkkJ+fD8Af/vAH+vTpw3nnnceWLVuOLvPcc88xfPhwkpOTueqqqygpKQFgz549TJ06leTkZAYPHkxqairFxcWMGzeOoUOHkpSUxNtvv310O48//jgDBw5k4MCBPPHEE/7aHcactsoqZUPOQV5ansHPXl3F2X/4iBmzVrJk/W4uHdCF2TcN55vfXMSfrkzi3F4dLBzqSctpQbx3P+yu526VLklw2aN1zo6KiuLss8/mvffeY8qUKcydO5cf/OAHPPDAA0RFRVFZWcm4ceNIT0+vcwTWtLQ05s6dy5o1a/B4PAwdOpRhw4YBzgivt9xyCwAPPvggL7zwAnfddRd33303Y8eOZcGCBXg8HkpKSggJCWHBggWEh4eTn5/PyJEjmTx5MqtWrWLWrFmsWLECVWXEiBFccMEFDBkypH73lTGn4HB5JWsyC0jN2M/KnQdYtfMAxWXOIJUxESGM6duJCUkxnN+nA62DAl2utvlqOQHhkiPdTEcC4oUXXuD111/n2WefxePxkJuby8aNG+sMiGXLljF16lTatm0LwOTJk4/OW79+PQ8++CAFBQUUFxcf7cpaunQpr7zyCgBBQUGEh4dTUVHBAw88wOeff05AQADZ2dns2bOHL774gqlTpx4dYfbKK69k2bJlFhCmQe0rLiN15wEnEDIOsD77IJ4qRQT6dg7jiiFdGZ4QRUpCVIu/eK0htZyAOMEvfX+aMmUK9913H6tWraKkpISoqCgee+wxVq5cSfv27ZkxYwalpaWnte0ZM2bw1ltvkZyczOzZs/n000/rXHbOnDnk5eWRlpZGcHAwCQkJp/2+xpwJVWXnvhJWZuwnNeMAK3fuZ3veIcAZJntwXCS3nt+T4QlRDI1vT0TbYJcrbrlaTkC4pF27dlx44YX8+Mc/Zvr06RQWFhIaGkpERAR79uzhvffeY8yYMXWuf/755zNjxgx+/etf4/F4WLhwIT/96U8BKCoqIiYmhoqKCubMmUNsbCwA48aN49///jd33XXX0S6mgwcP0qlTJ4KDg/nkk0/YuXMn4AwZPmPGDO6//35UlQULFhxtfRhTHzyVVWzMLWRlxrEWwpGb6ES0CWZ4Qnt+kNKN4QntGRgbYV1GjYgFRAOYPn06U6dOZe7cuSQmJjJkyBASExPp1q0bo0aNOuG6Q4cO5Yc//CHJycl06tSJ4cOHH533u9/9jhEjRtCxY0dGjBhBUVERAE8++SS33HILjz76KNHR0cyaNYtrr72WSZMmkZSUREpKComJiUe3P2PGDM4++2wAbr75ZuteMmfkUJmH1bsKnBbCzv2s3lVASXklAN2i2nB+7w6kJEQxPKE9Z3VsR4AdUG60bLjvZmz58uVs2bKFm266qV6219L3p6nd3qJS0jIO8I23y2hjbiGVVUqAQL+YcO+xg/akdI+iS0TLGxG1sbPhvlug1157jd/+9rc8+OCDbpdimhFVZXv+oaNdRakZ+8nY55xeHRIcwJBu7blzzFmkJEQxJD6SsBA7ftCUWUA0U9OnT2f69Olul2GauIMlFazNKmBN5rHH/kPOdT3Roa1ISWjPdSO7k5IQxYCu4QQ3g9tsmmP8GhAiMh54EggEnlfVR2vM7w68CHQE9gPXqWqWd96NwJGfv79X1ZdOpwZVtcvr60Fz6Yo0dSv3VLF5d6ETBLucMNie75xdJAK9OrZjXGInUhLaMzwhih4dQu3fVjPnt4AQkUDgKeBiIAtYKSLvqOrGaos9Brysqi+JyFjgT8D1IhIFPASkAAqkedc9cCo1hISEsG/fPqKjo+1/5DOgquzbt4+QEOs/bi5Ulcz9h1mdeYA1mQWszSxgfU4h5Z4qADqGtWZwt0iuGhbH4G6RJMVFEG7dRS2OP1sQZwPbVHU7gIjMBaYA1QOiP/Bz7/NPgLe8zy8FPlTV/d51PwTGA6+dSgFxcXFkZWWRl5d32h/COEJCQoiLi3O7DHOaDpZUsCbLaRkc6TI60lUUEhzAoNhIbjynO4O7tWdwfCRdI0LsR5Xxa0DEApnVXmcBI2ossxa4EqcbaioQJiLRdawbW/MNRORW4FaA+Pj47xUQHBxMjx49Tv8TGNMElXuq2JRbePSYwdoaXUW9OzldRYPjIxncLZK+ncMIsmMHphZuH6T+JfBPEZkBfA5kA5W+rqyqzwLPgnOaqz8KNKYxU1V27S9hTWYBq72tgw11dBUN8XYV2ZlFxlf+DIhsoFu113HeaUepag5OCwIRaQdcpaoFIpINjKmx7qd+rNWYJqGgpPy4lsHarINHu4raBAeSFBvBjHMTGNzNaR3EWFeROQP+DIiVQG8R6YETDNcAP6q+gIh0AParahXwa5wzmgCWAH8Ukfbe15d45xvT4mzdU8Tzy7azMuMAO2p0FV3Ur5Nz3KBbJH06t7OuIlOv/BYQquoRkZ/hfNkHAi+q6gYReQRIVdV3cFoJfxIRxeliutO77n4R+R1OyAA8cuSAtTEtxfa8Yp78+FveWZtDaKsgzj0rmmkpcQyOs64i0zCa9VAbxjRFmftL+PvH3/Lm6mxaBQYwY1QCt47uSfvQVm6XZpohG2rDmCYg9+Bh/rl0G6+nZiIizDg3gdsuOIuOYa3dLs20UBYQxrhsb1EpT3/6HXNW7EJVuWZ4PHde2MsGtjOus4AwxiX7D5Xz78+/46XlGVRUKlcPjeOucb2Ia9/W7dKMASwgjGlwBw9X8MKy7bzwxQ5KKiq5YnAs94zrTUKHULdLM+Y4FhDGNJDiMg+zv9zBs59vp7DUw+VJMdx7UW96dw5zuzRjamUBYYyfHS6v5JWvM3jms+3sP1TORf06c9/FvRnQNcLt0ow5IQsIY/ykzFPJayt28dSn35FXVMb5fTry84v7MLhbpNulGeMTCwhj6llFZRXzUrP4x9JvyT1YyogeUfzr2qEMT4hyuzRjTokFhDH1xFNZxVtrcnjy461k7j/M0PhIHpuWzLln2f1ITNNkAWHMGaqqUhaty+WJD7eyPf8QA2PDeeSmgYzp09GCwTRpFhDGnCZVZcmGPfztw61s2VNE385h/Pv6YVzSv7MFg2kWLCCMOUWqyqdb8vjrh1tYn11Iz46h/GP6EC5PiiEgwILBNB8WEMb4SFVZ/t0+HvtgC6t3FdAtqg1/nZbMlMFdbZht0yxZQBjjg5UZ+/nrB1v4evt+ukaE8Kcrk7h6WBzBFgymGbOAMKYOqkrazgP8fek2Pt+aR8ew1jw8eQDXnN2N1kGBbpdnjN9ZQBhTjaqyMbeQRem5LErPIXP/YaJCW/GbCf24bmR32rSyYDAthwWEMcC3e4pY6A2F7XmHCAwQzuvVgbvH9uaypBjatbZ/Kqblsf/rTYuVkX+IRek5LErPZfPuIgIERvaM5ubzejJ+YBei7A5upoWzgDAtSnbBYd5Nz2Hh2lzWZR8EIKV7ex6ePIDLkrrQKcxu0mPMERYQptnbW1jKu+tyWZSeS9rOAwAkx0Xw4OX9mJAUQ9fINi5XaEzjZAFhmqV9xWW8t343i9JzWLFjP6rQLyacX13al0mDuhIfbXdtM6dJFTKWwaqXYe8mEAEJhIBAkIAazwO8zwOP/RWpMe3IOjXXPdl2Ao6t0zER+o6v949qAWGajYMlFSzZuJuFa3NY/t0+KquUszqGcs+43kwc1JVendq5XaJpyor3wppXYdVLsH87hERA/LnOF35VJWjlsb+q3ufl1aZVQVVVteVqPP/etErv8jWmadX3axt4lQWEMTUVl3n4aOMeFq7N4fNv86ioVOKj2nLbBT2ZOKgriV3CbFwkc/qqqmD7J5A2G7YshiqPEwoX3A/9J0OwC92Tqt5HtdAQ/1ywaQFhmpzD5ZUs3byXRek5LN28lzJPFV0jQphxbgKTkruSFBthoWDOTGEOrJ4Dq1+Ggl3QJgpG3AZDb4SOfdytTcR5EACBwX59KwsI0ySUeSr5bEsei9Jz+WjTHkrKK+nQrjXTz45n4qAYhsa3t4HyGkKlBwKb6ddGVSV8+6HThbT1facrp8cFcNFMSJwIQa3drrDBNdP/0qY5qKis4ott+Sxam8sHG3ZTVOahfdtgpgyOZVJyDCN6RBNooeBfqrBng9O9svldyF0LnQdCj9GQMBq6nwttmvgtVAsyYfUrsPo/UJgNoZ1g1D0w5HqIPsvt6lxlAWEanb2FpTzz2XYWrM7iQEkFYSFBXDqwC5OSu3LuWdE2QJ6/VXpg13LYvBi2vOt0sSAQlwLn3Am70yH1Rfj6X07fd0yyExY9zof4kdA6zO1PcHKVFU4rIe0l2PaRM63XOBj/KPS9zO9dN02FBYRpNPKLy3jm0+945eudeKqUCUkxTE7uyvl9OtjgeP5WVgzffey0ErYugdICCGwNPcfA6F9An8sgrPOx5StKITsVdnwOO5bB10/D8r87p1zGDjvWwug2Alo1olOK9293Tk9d8yoU74GwrnD+r2Do9RAZ73Z1jY6oqts11IuUlBRNTU11uwxzGg4cKuffn2/npeUZlHkquWJILPeM60336FC3S2veinbDlvec7qPtn0FlGbRpD33GQ98JcNZYaO3jqcHlJZC5wgmMjGWQvco5uyawFcQN97YwRjvPG7ov31PmBF/abNjxmdPq6X0pDJsBvS5qvsdUfCQiaaqaUus8CwjjloMlFTz/xXZe/GIHJRWVTBrUlXsu6s1ZHe16Bb9QhbwtTrfR5sVOCwAgsjskXu6EQvw59fOFWVYEu772tjA+d7qltAqCQqDb2U53VML5EDvUf905+d86obD2NSjZBxHxMPQGGHIthHf1z3s2QRYQplEpKq3gxS8yeP6L7RSVepiQ1IV7L+pDn85NoO+6qamqdH7Zb37XaSns3+5M7zoE+l4OiROgU3/vaZN+dLgAdi53Whc7Poc9653pwaHOcYse5zstjC7JZxZQFYdh4zvOmUg7v4SAICf4ht0IPcc6Vx6b41hAmEbhUJmH2cszeG7ZdgpKKrikf2fuvagP/buGu1dUVZXTteIpdfrVPaVOl4TnsPdvKbQOh/BYCO3YNL5gykuci7s2L4at7zm/ngOCnS/hxAnO8YSIWHdrPLQPdn7hHL/IWAZ5m53prcOdM6N6nO90S3Ue6Ns+37PRCYW1c53jJ1E9ndbC4GuhXSf/fpYm7kQB0bI730yDOFxeyStfZ/DMZ9vZf6icsYmduO+iPiTFRZxgpQPO2TOeMudX4ZEva0+1L/Fap59oXlmNECh1wsFXAUEQFuN0T4R3dULj6OtY529YF3fOgCnOc87K2bIYvvvECbjWEdDnEucXdK+LIMTFIK4pNBr6T3EeAEV7nKDIWOaExtb3nelt2kP3Uc71CD1GO2MOHWntlB+CDQucbqSslc7xjn6TndZC9/OaRpg3ctaCMH5TWlHJqyt28a9PvyO/uIzRvTtw38V9GBrfvo4VDjq/eje8Cd8tdYY18IUEQFAb5+BnsPdvUEi1h/d1cPXXbXybHtjaqasw27m6tjAHinKOPa8oqVmM84u1rgAJ7+pMq48ze/K3HTuekLkCUAiPc1oJfSdAwnlN93TNg9nHwiLjc++ptjituITznJbGhgVQVggd+jqhMOgaJ3jMKbEuJtOgyjyVvL4yk39+so09hWWc0zOa+y7uw9k9or6/cPkh50yaDQucq1gryyCiGwyY6hzMDG5T48s+5PtB4NaXoKrTnVGY6w2M7O8HSGG2EzA1tWlfR4DEHHveOvz4YwNVVZCd5g2FdyF/qzO9S9Kx4wldBvn/eIIbDuw8dvxixzI4vN/5f2Tojc4xjOb4mRuIawEhIuOBJ4FA4HlVfbTG/HjgJSDSu8z9qrpYRIKB54GhON1gL6vqn070XhYQ7quorGJ+Whb/XLqN7ILDpHRvz88v6cO5Z3WosWApbPsQ1r/hnHNfUQLtujj/4Ade6ZwK2Zz+wZcf8oZIHQFSmAuH9n5/veDQY62ONpHOWUHFe5xrDRJGOcM/9L2s5Z2/f2Sk1BZ+emp9ceUYhIgEAk8BFwNZwEoReUdVN1Zb7EHgdVV9WkT6A4uBBGAa0FpVk0SkLbBRRF5T1Qx/1WtOn6eyigWrs/n70m/J3H+Ywd0iefSqJM7r1eHYoHmecufA6fo3nV+/5UXQNhqSpzuhEH+OM859c9QqFDr0ch518ZRDUW7dAVKw09lHiZdD74udFkhLJWLh0ED8uZfPBrap6nYAEZkLTAGqB4QCR46cRQA51aaHikgQ0AYoBwr9WKs5DZVVysK1OTz58bfsyD9EUmwEj8wYyJi+HZ1gqPQ43QLr34BNC53umJAIGDDFGb8+4Xz7h35EUCto3915GNNI+PNfZyyQWe11FjCixjIzgQ9E5C4gFLjIO30+TpjkAm2B+1R1f803EJFbgVsB4uNbWDPbRVVVyuL1uTzx0bds21tMYpcwnr1+GBf374yoOue7r38DNr4NJfnQKszpHx9wpXN1blArtz+CMcYHbv98mw7MVtW/isg5wCsiMhCn9VEJdAXaA8tE5KMjrZEjVPVZ4FlwjkE0bOktj6qyZMMenvhoK5t3F9G7Uzv+de1QxvfvTEBOGix50jnYXJTrnA3Ud7wTCr0vdufGKsaYM+LPgMgGulV7HeedVt1PgPEAqvqViIQAHYAfAe+ragWwV0S+BFKA7ZgGp6os3byXxz/cyoacQnp2COXJHyYzsVMegRufho8XOKchBraCXhc7xxT6jPd9HB9jTKPkz4BYCfQWkR44wXANzhd/dbuAccBsEekHhAB53uljcVoUocBI4Ak/1mpqoap8tjWPv324lbVZB4mPasuz40O5yLOMgGU/h/3fOReP9bwQxjzgdCOFnODiN2NMk+K3gFBVj4j8DFiCcwrri6q6QUQeAVJV9R3gF8BzInIfzoHpGaqqIvIUMEtENgACzFLVdH/Vao6nqiz/bh+Pf7iVtJ0HGBF+gIWDNjHwwEfIp3F9I0oAABkjSURBVJudC9MSRsOou50rV9vWcn2DMabJswvlzHE25hTy8MINZGds4Zq2qUwPTSW6cJMzM/5cp/uo/xQb38aYZsLGYjI+Wbo+k6Wv/4PfBHzCoNZbndMEwobBOX+E/le4P8CbMaZB+RQQInI5MADnGAEAqvqIv4oyDaz0IKve+CsDtr7E2IACPB36weCZzpXN7RPcrs4Y45KTBoSIPINzLcKFOMNfXA184+e6TEMozEG/+hflK15gaFUJ69sOI+KK3xDSZ2zzGurCGHNafGlBnKuqg0QkXVUfFpG/Au/5uzDjR3s3w/K/o+mvU1VVyXuVI8npfws//eFUAgMsGIwxDl8C4rD3b4mIdAX2ATH+K8n4hSrs+gq+fBK2vo8GteXd1pfxaME4Zkw4n9vP63Fs3CRjjMG3gFgkIpHA/wGrcE5Hfd6vVZn6U1Xp3ETmyyedm6q0jebAiF9xw7oktha24okfDeayJMt7Y8z3nTQgVPV33qdviMgiIERVaxng3jQqFaWQPheW/wP2bXMONl/+V1a1v4yfvLoBEeG1W1PqvnmPMabFqzMgRGSsqi4VkStrmYeqvunf0sxpOXwAUl+Er59x7jEQMximzYZ+k1m4bg+/mL2WuMg2zLppON2jQ92u1hjTiJ2oBXEBsBSYVMs8BSwgGpODWfD10879ecuLnXsQj7oHEkajwDOfbefP729meEJ7nr0+hfahNqKqMebE6gwIVX3I+/emhivHnLI9G+DLv8P6+c6B6KSr4dy7octAwLmZz0PvbGDOil1MSu7K/109iJDgZnpjHmNMvfLlOog/An9R1QLv6/bAL1T1QX8XZ+qgChlfOAeet33o3Jry7Fth5O3H3X6yuMzDz15dxadb8rhjzFn88pK+BNhprMYYH/lyFtNlqvrAkReqekBEJuDcLtQ0pKpK585sXz4JOasgtCOMfRBSfvK9AfP2FJZy06yVbNlTxB+nJvGjEXZDJWPMqfElIAJFpLWqlgGISBugtX/LMsepOAxrXnXOSDqwA6J6wsS/OfdzruVGPJt3F3LTrJUUHq7ghRtTGNPXBtYzxpw6XwJiDvCxiMzyvr4JeMl/JZmjSvbDyhdgxTPOrTu7DoWLH4bEiRBQ+3GEZd/mcft/VtGudRDzbjuX/l3Da13OGGNOxpfrIP4sIuk4N/YB+J2qLvFvWS1cwS746l+w6mWoOAS9L3HOSOo+6oRjJL2+MpMHFqyjV6d2zLppODERdptPY8zp82k0V1V9Dxt/yf92r/OekfSGEwRJ05wzkjr3P+FqqsrjH27lH0u3Mbp3B/517VDCQoIbqGhjTHNVa0CISDtVLfY+Hwn8E+iLc+whEDikqtZ3UV9U4dM/wWd/hlbtnLORRt4OEXEnXbXMU8n/zE/nrTU5XDO8G7+7YiDBgQENULQxprmrqwVxnXdgvodwwuFa4BngIuAGoE/DlNcCqMKS38DXT8Hga+HSP0Ab34a/OFhSwa2vpLJix35+dWlf7hhzlg24Z4ypN7X+1FTVZ4C1OMGAqm4BglW1UlVnAeMbrsRmrKoKFt3rhMOI22DyP30Oh8z9JVz59Jes3lXAk9cM5s4Le1k4GGPq1YmupH4DQERuFZFWwGbvRXN5ON1M5kxUeuDtOyD9v3Dez2Hc//p8k541mQXc/NJKKiqVV35yNiN6Rvu5WGNMS+RLZ/X13uXuA0qBeJy7ypnT5SmDeTc64TD2t3DRQz6HwwcbdnPNs1/RplUgb9x+roWDMcZvTngWk4gEAn9U1WtxwsHuQ32mKg7Df6+DbR/B+Eedg9E+mvXlDh5ZtJFBcZG8cGMKHdrZ9YrGGP85YUCoaqWIdBeRVqpa3lBFNVtlRfDadGccpUl/h2E3+rRaZZXy+3c3MuvLDC4d0JknfjiENq2sl88Y41++XAexHfhSRN4BDh2ZqKqP+62q5ujwAfjP1ZCzGq58DgZN82218krumbuaDzbu4cejevCby/vZfaONMQ3Cl4D4zvsIAML8W04zVZwHr0yF/C3wg5eh30SfVssrKuPml1NJzyrgoUn9uWlUDz8Xaowxx/gy1MbDDVFIs1WYAy9PgYJMmP6acyMfH2zbW8xNs78hr6iMf183jEsGdPFzocYYczxf7gfxCc4d5I6jqmP9UlFzcmAnvDwZDuXDdW9AwiifVvt6+z5ufTmVVkEB/PfWc0juFunnQo0x5vt86WL6ZbXnIcBVgMc/5TQj+d86LYfyQ3DDOxA3zKfV3l6Tza/mpRMf3ZZZM4bTLaqtnws1xpja+dLFlFZj0pci8o2f6mkedq+HV65wns949+jtP09EVXnqk2089sFWRvaM4t/XpRDR1gbcM8a4x5cupuq3KgsAhgERfquoqctOg1euhOC2cMPb0NG3YaveXJXNYx9sZeqQWB69KonWQXYaqzHGXb50MaXhHIMQnK6lHcBP/FlUk7VzOcz5gXP7zxvfgfYJPq/66je76NWpHY//INnGVDLGNAq+dDHZuZW++G4pvPYjZ4juG96GiFjfV80rJm3nAX59WaKFgzGm0TjpWEwicqeIRFZ73V5E7vBvWU3M5sXw6g8h+iy46b1TCgeA+WlZBAYIU4ee2nrGGONPvgzWd4uqFhx5oaoHgFv8V1ITs26+M7ZSlyS4cSG063hKq1dWKW+uymJMn450CgvxU5HGGHPqfAmIQKnW7+EdwK+V/0pqQla9Am/cDPEjnW6ltlEnX6eGz7/NY09hGdNSTn73OGOMaUi+BMT7wH9FZJyIjANew8f7U4vIeBHZIiLbROT+WubHi8gnIrJaRNJFZEK1eYNE5CsR2SAi60Skcf28XvFveOdncNZYuHY+tD69UUjmp2YRFdqKsYmd67lAY4w5M76cxfQ/wK3Abd7X6cBJx33wtjSeAi4GsoCVIvKOqm6sttiDwOuq+rSI9AcWAwkiEgT8B7heVdeKSDRQ4euH8rtlf4WPH4HEiXD1ixB0esNuF5SU8+HGPVw7Mp5WQXYfaWNM43LSbyVVrQJWABnA2cBYYJMP2z4b2Kaq271Dhc8FptTcPBDufR4B5HifXwKkq+pabw37VLXSh/f0L1UnGD5+BJKmwbTZpx0OAG+vyaG8soppw7rVX43GGFNP6mxBiEgfYLr3kQ/8F0BVL/Rx27FAZrXXWcCIGsvMBD4QkbuAUODISHZ9ABWRJUBHYK6q/qWWGm/Fad0QHx/vY1mnSRXe/zWseBqG3gATn4CAM7uYbV5aJgO6htO/a/jJFzbGmAZ2ohbEZpzWwkRVPU9V/wHU96/46cBsVY0DJgCviEgATnCdB1zr/TvVe/zjOKr6rKqmqGpKx46ndvbQKamqhIV3O+Ew4nbnZj9nGA6bcgtZn13ItGF2cNoY0zidKCCuBHKBT0TkOe8X9KlcxZUNVO87ifNOq+4nwOsAqvoVzmCAHXBaG5+rar6qluAcmxh6Cu9dfyorYMFPYdXLMPqXMP5PPt8/+kTmpWbRKjCAKYPt2gdjTONUZ0Co6luqeg2QCHwC3At0EpGnReQSH7a9EugtIj1EpBVwDfBOjWV2AeMARKQfTkDkAUuAJBFp6z1gfQGwkYbmKYN5M2DdPBj3EIz7bb2EQ7mnirfWZHNR/060D7Uzho0xjZMvB6kPqeqrqjoJpxWwGufMppOt5wF+hvNlvwnnbKUNIvKIiEz2LvYL4BYRWYtz+uwMdRwAHscJmTXAKlV99zQ+3+krL3HuH715EVz2Fxj983rb9NLNe9l/qNwOThtjGjVR/d69gJqklJQUTU1NrZ+NlRbCa9c4g+9N/gcMvb5+tut180srSc86yPL7xxIUaKe3GmPcIyJpqppS2zz7dqqpZL9zL4fMFXDV8/UeDnuLSvlkSx5Th8ZaOBhjGjVfLpRrOYrznHDI3wo/eAUSJ5x8nVP01upsKqvUupeMMY2eBcQRhTnw0mQ4mAU/+q8zhEY9U1XmpWYxJD6SXp3a1fv2jTGmPlkfB8CBDHhxPBTthuvf9Es4AKzNOsi3e4ut9WCMaRKsBZH/rdNy8ByGG9+G2GF+e6t5qZmEBAcwMTnGb+9hjDH1xVoQQSEQHgMz3vVrOJRWVPLO2hzGD+hCeEiw397HGGPqi7UgIrvBzR/XywVwJ7Jkw26KSj1MS7HuJWNM02AtCPB7OIBzW9HYyDac0zPa7+9ljDH1wQKiAWQXHOaLbflcNSyOgAD/h5ExxtQHC4gG8GZaFqrYyK3GmCbFAsLPVJX5q7IY2TOKblFt3S7HGGN8ZgHhZ9/s2M/OfSV27YMxpsmxgPCzeWlZtGsdxGVJJ72NtzHGNCoWEH50qMzD4nW5XJ4UQ9tWdkaxMaZpsYDwo3fX5VJSXsm0FDs4bYxpeiwg/Gh+ahY9O4QyrHt7t0sxxphTZgHhJxn5h/gmYz9XDYtDGuBCPGOMqW8WEH4yPy2LAIGrhlr3kjGmabKA8IPKKuWNVVmM7t2RLhEhbpdjjDGnxQLCD77clk/uwVI7OG2MadIsIPxgXloWEW2CuahfZ7dLMcaY02YBUc8OllSwZMNupgzuSkhwoNvlGGPMabOAqGfvpOdQ7qmyoTWMMU2eBUQ9m5+aSWKXMAbGhrtdijHGnBELiHq0dU8Ra7MOcrVd+2CMaQYsIOrRvNRMggKEqUNi3S7FGGPOmAVEPamorGLB6mzGJnYiul1rt8sxxpgzZgFRTz7dkkd+cTnTUuzgtDGmebCAqCfz0zLp0K4VY/p2dLsUY4ypFxYQ9WBfcRkfb9rLFYNjCQ60XWqMaR7s26wevLUmB0+VWveSMaZZsYA4Q6rKvNRMBsVF0LdLmNvlGGNMvbGAOEMbcgrZvLuIacNsYD5jTPNiAXGG5qVm0ioogMnJdu2DMaZ5sYA4A2WeSt5em8Ml/TsT0TbY7XKMMaZe+TUgRGS8iGwRkW0icn8t8+NF5BMRWS0i6SIyoZb5xSLyS3/Webo+2riXgpIKOzhtjGmW/BYQIhIIPAVcBvQHpotI/xqLPQi8rqpDgGuAf9WY/zjwnr9qPFPz0jKJiQjhvF4d3C7FGGPqnT9bEGcD21R1u6qWA3OBKTWWUeDIsKcRQM6RGSJyBbAD2ODHGk/b7oOlfL41jyuHxhIYYAPzGWOaH38GRCyQWe11lndadTOB60QkC1gM3AUgIu2A/wEePtEbiMitIpIqIql5eXn1VbdP3lydRZXC1XbfB2NMM+X2QerpwGxVjQMmAK+ISABOcPxNVYtPtLKqPquqKaqa0rFjww1xoarMT81ieEJ7enQIbbD3NcaYhhTkx21nA9V/Xsd5p1X3E2A8gKp+JSIhQAdgBHC1iPwFiASqRKRUVf/px3p9tmrXAbbnH+K2C85yuxRjjPEbfwbESqC3iPTACYZrgB/VWGYXMA6YLSL9gBAgT1VHH1lARGYCxY0lHADmpWbRJjiQCYNi3C7FGGP8xm9dTKrqAX4GLAE24ZyttEFEHhGRyd7FfgHcIiJrgdeAGaqq/qqpPpSUe1iUnsuEpBjatfZnvhpjjLv8+g2nqotxDj5Xn/a/1Z5vBEadZBsz/VLcaXp//W6KyzxMS7GhNYwxzZvbB6mbnHmpWcRHtWVEjyi3SzHGGL+ygDgFmftL+Gr7Pq4eFoeIXftgjGneLCBOwfy0LETgKhu51RjTAlhA+KiqSpmflsWoszoQG9nG7XKMMcbvLCB89PX2fWQXHLaD08aYFsMCwkfz0rIICwni0gFd3C7FGGMahAWEDwpLK3hvfS6TkrsSEhzodjnGGNMgLCB88G56LqUVVXZbUWNMi2IB4YN5qZn06tSOwd0i3S7FGGMajAXESWzbW8yqXQVMs2sfjDEtjAXEScxPyyIwQJg6tOatLIwxpnmzgDgBT2UVb67KYkyfjnQKC3G7HGOMaVAWECew7Nt89haV2bUPxpgWyQLiBOalZdK+bTBjEzu7XYoxxjQ4C4g6HDhUzkcb9zJlcCytgmw3GWNaHvvmq8Pba7Ipr6yy7iVjTItlAVGHeWlZ9I8JZ0DXCLdLMcYYV1hA1GJjTiEbcgqt9WCMadEsIGoxLy2T4EBhymC79sEY03JZQNRQ7qni7TU5XNSvM1GhrdwuxxhjXGMBUcPSzXvYf6jcupeMMS2eBUQN81Kz6BTWmvN7d3S7FGOMcZUFRDV7i0r5dGseU4fGEhRou8YY07LZt2A1C1ZlU1mlTBvWze1SjDHGdRYQXqrKvLQshsRH0qtTO7fLMcYY11lAeK3JLGDb3mJrPRhjjJcFhNe8tCxCggOYmBzjdinGGNMoWEAApRWVLFybw/gBXQgPCXa7HGOMaRQsIIAlG3ZTVOphWop1LxljzBEWEDi3FY2NbMM5PaPdLsUYYxqNFh8Q2QWH+WJbPlcNiyMgQNwuxxhjGo0WHxAlZR4u7NuJq4fa0BrGGFNdkNsFuK135zBenDHc7TKMMabRafEtCGOMMbWzgDDGGFMrvwaEiIwXkS0isk1E7q9lfryIfCIiq0UkXUQmeKdfLCJpIrLO+3esP+s0xhjzfX47BiEigcBTwMVAFrBSRN5R1Y3VFnsQeF1VnxaR/sBiIAHIByapao6IDASWAHZ7N2OMaUD+bEGcDWxT1e2qWg7MBabUWEaBcO/zCCAHQFVXq2qOd/oGoI2ItPZjrcYYY2rwZ0DEApnVXmfx/VbATOA6EcnCaT3cVct2rgJWqWpZzRkicquIpIpIal5eXv1UbYwxBnD/IPV0YLaqxgETgFdE5GhNIjIA+DPw09pWVtVnVTVFVVM6drQ7wBljTH3yZ0BkA9UHN4rzTqvuJ8DrAKr6FRACdAAQkThgAXCDqn7nxzqNMcbUwp8Xyq0EeotID5xguAb4UY1ldgHjgNki0g8nIPJEJBJ4F7hfVb/05c3S0tLyRWTnGdTbAefguLF9UZPtj2NsXxyvOeyP7nXNEFX127t6T1t9AggEXlTVP4jII0Cqqr7jPXPpOaAdzgHr/6eqH4jIg8CvgW+rbe4SVd3rx1pTVTXFX9tvSmxfHM/2xzG2L47X3PeHXwOiKWnu/6FPhe2L49n+OMb2xfGa+/5w+yC1McaYRsoC4phn3S6gEbF9cTzbH8fYvjhes94f1sVkjDGmVtaCMMYYU6sWHxAnG1CwJRGRbt7BEzeKyAYRucftmtwmIoHewSQXuV2L20QkUkTmi8hmEdkkIue4XZObROQ+77+T9SLymoiEuF1TfWvRAVFtQMHLgP7AdO+pty2VB/iFqvYHRgJ3tvD9AXAPsMntIhqJJ4H3VTURSKYF7xcRiQXuBlJUdSDOqfzXuFtV/WvRAYFvAwq2GKqaq6qrvM+LcL4AWuwout6r+S8Hnne7FreJSARwPvACgKqWq2qBu1W5LghnINEgoC3ewUabk5YeEL4MKNgiiUgCMARY4W4lrnoC+H9AlduFNAI9gDxglrfL7XkRCXW7KLeoajbwGM5oELnAQVX9wN2q6l9LDwhTCxFpB7wB3KuqhW7X4wYRmQjsVdU0t2tpJIKAocDTqjoEOAS02GN2ItIep7ehB9AVCBWR69ytqv619IDwZUDBFkVEgnHCYY6qvul2PS4aBUwWkQycrsexIvIfd0tyVRaQpapHWpTzcQKjpboI2KGqeapaAbwJnOtyTfWupQfE0QEFRaQVzkGmd1yuyTUiIjh9zJtU9XG363GTqv5aVeNUNQHn/4ulqtrsfiH6SlV3A5ki0tc7aRyw8QSrNHe7gJEi0tb772YczfCgvT9Hc230VNUjIj/DuaXpkQEFN7hclptGAdcD60RkjXfaA6q62MWaTONxFzDH+2NqO3CTy/W4RlVXiMh8YBXO2X+raYZXVduV1MYYY2rV0ruYjDHG1MECwhhjTK0sIIwxxtTKAsIYY0ytLCCMMcbUygLCGB+ISICIvC8i8W7XYkxDsdNcjfGBiJwFxKnqZ27XYkxDsYAw5iREpBJYV23SXFV91K16jGkoFhDGnISIFKtqO7frMKah2TEIY06TiGSIyF9EZJ2IfCMivbzTE0RkqYiki8jHR45biEhnEVkgImu9j3O9098SkTTv3clu9U4LFJHZ3ruVrROR+9z7pKalatFjMRnjozbVxqYC+JOq/tf7/KCqJonIDTj3j5gI/AN4SVVfEpEfA38HrvD+/UxVp3rvZnikVfJjVd0vIm2AlSLyBpAAxHrvVoaIRPr7QxpTk3UxGXMSdXUxeYcCH6uq273DpO9W1WgRyQdiVLXCOz1XVTuISB7Oge6yGtuZCUz1vkwALgW2AKnAYuBd4ANVtRsXmQZlXUzGnBmt47lPRGQMzr0FzlHVZJxRQUNU9QDOfZ8/BW7DbntqXGABYcyZ+WG1v195ny/n2A3srwWWeZ9/DNwOR48xRAARwAFVLRGRRGCkd34HIEBV3wAepGXfnMe4xLqYjDmJWk5zfV9V7/d2Mf0XuAwoA6ar6jYR6Q7MAjrg3Mf5JlXdJSKdce4Z0BOoxAmLVcBbOF1LW4BIYCZwwLuNIz/ifq2q7/nxYxrzPRYQxpwmb0CkqGq+27UY4w/WxWSMMaZW1oIwxhhTK2tBGGOMqZUFhDHGmFpZQBhjjKmVBYQxxphaWUAYY4yplQWEMcaYWv1/cKjuobs2x7AAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "RQRBUFk_czYU",
        "outputId": "ac6ea133-b80e-46d4-8c5d-297b2fbe6298"
      },
      "source": [
        "plt.plot(historico.history['loss'])\n",
        "plt.plot(historico.history['val_loss'])\n",
        "plt.title(\"Perdas por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Perda')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32df99b690>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "atosM-DJdA34"
      },
      "source": [
        "Se o gráfico da validação vem caindo e o teste subindo, estamos tendo uma situação de **overfitting**. Caso o contrário, estamos com uma situação de **underfitting**."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "j-oISribdomw"
      },
      "source": [
        "\"Adormecendo\" camadas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7NWCA87_dKgE",
        "outputId": "fa86dff1-ffd9-460b-9a76-cb92591b9e71"
      },
      "source": [
        "modelo = keras.Sequential([                    \n",
        "     keras.layers.Flatten(input_shape=(28,28)), \n",
        "     keras.layers.Dense(256,activation=tensorflow.nn.relu),\n",
        "     keras.layers.Dropout(0.2),\n",
        "     keras.layers.Dense(10,activation=tensorflow.nn.softmax)\n",
        "])\n",
        "\n",
        "\n",
        "modelo.compile(optimizer='adam',\n",
        "               loss='sparse_categorical_crossentropy',\n",
        "               metrics=['accuracy'])\n",
        "\n",
        "historico = modelo.fit(imagens_treino,identificacoes_treino,epochs=5, validation_split=0.2)"
      ],
      "execution_count": 32,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.6640 - accuracy: 0.7648 - val_loss: 0.4176 - val_accuracy: 0.8471\n",
            "Epoch 2/5\n",
            "1500/1500 [==============================] - 7s 4ms/step - loss: 0.4052 - accuracy: 0.8522 - val_loss: 0.3679 - val_accuracy: 0.8680\n",
            "Epoch 3/5\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3681 - accuracy: 0.8653 - val_loss: 0.3538 - val_accuracy: 0.8757\n",
            "Epoch 4/5\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3407 - accuracy: 0.8735 - val_loss: 0.3611 - val_accuracy: 0.8662\n",
            "Epoch 5/5\n",
            "1500/1500 [==============================] - 6s 4ms/step - loss: 0.3251 - accuracy: 0.8802 - val_loss: 0.3328 - val_accuracy: 0.8826\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "ZHoLs5E-d01I",
        "outputId": "aed18721-1ce1-49b9-9a02-f26975039c76"
      },
      "source": [
        "plt.plot(historico.history['accuracy'])\n",
        "plt.plot(historico.history['val_accuracy'])\n",
        "plt.title(\"Acurácia por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Acurácia')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32df7744d0>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 33
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 317
        },
        "id": "Ea0xuGv0d2ex",
        "outputId": "52d7b20d-5480-4032-d85d-953c045cf91c"
      },
      "source": [
        "plt.plot(historico.history['loss'])\n",
        "plt.plot(historico.history['val_loss'])\n",
        "plt.title(\"Perdas por Épocas\")\n",
        "plt.xlabel('Épocas')\n",
        "plt.ylabel('Perda')\n",
        "plt.legend(['Treino','Validação'])"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.legend.Legend at 0x7f32df714590>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 34
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2w_BS4iYd8xM"
      },
      "source": [
        "modelo.save('modelo_fashion_mnist.h5')"
      ],
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Ncw6_hreGXu"
      },
      "source": [
        "from tensorflow.keras.models import load_model\n",
        "modelo_salvo = load_model('/content/modelo_fashion_mnist.h5')"
      ],
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NgxU6N-ieYeI",
        "outputId": "d6d45141-a7c1-44ca-84a1-83f900f9f1c6"
      },
      "source": [
        "testes_modelos_salvo = modelo_salvo.predict(imagens_teste)\n",
        "print('Resultado do Teste:',np.argmax(testes[1]))\n",
        "print('Resultado do Modelo Salvo:',np.argmax(testes_modelos_salvo[1]))\n",
        "print('Número da Imagem de Teste:',identificacoes_teste[1])"
      ],
      "execution_count": 37,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Resultado do Teste: 2\n",
            "Resultado do Modelo Salvo: 2\n",
            "Número da Imagem de Teste: 2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eZ24K1kjfCgi"
      },
      "source": [
        "Validado que tem como importar o modelo gerado para os próximos códigos."
      ]
    }
  ]
}