1dCNN&RNN.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "1dCNN&RNN.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyPv+Pj/jlD8NgRj6V3qcTN/",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "TPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/HanzhouLiu/Deep-Learning-with-Python-Exercises/blob/main/1dCNN%26RNN.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "x6w2Om6F4N5G",
        "outputId": "2271bf40-7b2b-4839-8792-809a48643e5d"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xN3N4ZxG9b5b"
      },
      "source": [
        "import os\n",
        "\n",
        "data_dir = '/content/drive/MyDrive/636/homework'\n",
        "fname = os.path.join(data_dir, 'jena_climate_2009_2016.csv')\n",
        "\n",
        "f = open(fname)\n",
        "data = f.read()\n",
        "f.close()\n",
        "\n",
        "lines = data.split('\\n')\n",
        "header = lines[0].split(',')\n",
        "lines = lines[1:]\n",
        "\n",
        "float_data = np.zeros((len(lines), len(header) - 1))\n",
        "for i, line in enumerate(lines):\n",
        "    values = [float(x) for x in line.split(',')[1:]]\n",
        "    float_data[i, :] = values\n",
        "    \n",
        "mean = float_data[:200000].mean(axis=0)\n",
        "float_data -= mean\n",
        "std = float_data[:200000].std(axis=0)\n",
        "float_data /= std\n",
        "\n",
        "def generator(data, lookback, delay, min_index, max_index,\n",
        "              shuffle=False, batch_size=128, step=6):\n",
        "    if max_index is None:\n",
        "        max_index = len(data) - delay - 1\n",
        "    i = min_index + lookback\n",
        "    while 1:\n",
        "        if shuffle:\n",
        "            rows = np.random.randint(\n",
        "                min_index + lookback, max_index, size=batch_size)\n",
        "        else:\n",
        "            if i + batch_size >= max_index:\n",
        "                i = min_index + lookback\n",
        "            rows = np.arange(i, min(i + batch_size, max_index))\n",
        "            i += len(rows)\n",
        "\n",
        "        samples = np.zeros((len(rows),\n",
        "                           lookback // step,\n",
        "                           data.shape[-1]))\n",
        "        targets = np.zeros((len(rows),))\n",
        "        for j, row in enumerate(rows):\n",
        "            indices = range(rows[j] - lookback, rows[j], step)\n",
        "            samples[j] = data[indices]\n",
        "            targets[j] = data[rows[j] + delay][1]\n",
        "        yield samples, targets\n",
        "        \n",
        "lookback = 1440\n",
        "step = 6\n",
        "delay = 144\n",
        "batch_size = 128\n",
        "\n",
        "train_gen = generator(float_data,\n",
        "                      lookback=lookback,\n",
        "                      delay=delay,\n",
        "                      min_index=0,\n",
        "                      max_index=200000,\n",
        "                      shuffle=True,\n",
        "                      step=step, \n",
        "                      batch_size=batch_size)\n",
        "val_gen = generator(float_data,\n",
        "                    lookback=lookback,\n",
        "                    delay=delay,\n",
        "                    min_index=200001,\n",
        "                    max_index=300000,\n",
        "                    step=step,\n",
        "                    batch_size=batch_size)\n",
        "test_gen = generator(float_data,\n",
        "                     lookback=lookback,\n",
        "                     delay=delay,\n",
        "                     min_index=300001,\n",
        "                     max_index=None,\n",
        "                     step=step,\n",
        "                     batch_size=batch_size)\n",
        "\n",
        "# This is how many steps to draw from `val_gen`\n",
        "# in order to see the whole validation set:\n",
        "val_steps = (300000 - 200001 - lookback) // batch_size\n",
        "\n",
        "# This is how many steps to draw from `test_gen`\n",
        "# in order to see the whole test set:\n",
        "test_steps = (len(float_data) - 300001 - lookback) // batch_size"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "C6Y3cIytHUos",
        "outputId": "39caae0d-2216-4240-fd26-d31278f2f7fc"
      },
      "source": [
        "from keras.models import Sequential\n",
        "from keras import layers\n",
        "from keras.optimizers import RMSprop\n",
        "\n",
        "model = Sequential()\n",
        "model.add(layers.Conv1D(32, 5, activation='relu',\n",
        "                        input_shape=(None, float_data.shape[-1])))\n",
        "model.add(layers.MaxPooling1D(3))\n",
        "model.add(layers.Conv1D(32, 5, activation='relu'))\n",
        "model.add(layers.MaxPooling1D(3))\n",
        "model.add(layers.Conv1D(32, 5, activation='relu'))\n",
        "model.add(layers.GlobalMaxPooling1D())\n",
        "model.add(layers.Dense(1))\n",
        "\n",
        "model.compile(optimizer=RMSprop(), loss='mae')\n",
        "history = model.fit_generator(train_gen,\n",
        "                              steps_per_epoch=500,\n",
        "                              epochs=20,\n",
        "                              validation_data=val_gen,\n",
        "                              validation_steps=val_steps)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/20\n",
            "500/500 [==============================] - 45s 88ms/step - loss: 0.4562 - val_loss: 0.4482\n",
            "Epoch 2/20\n",
            "500/500 [==============================] - 43s 87ms/step - loss: 0.3710 - val_loss: 0.4713\n",
            "Epoch 3/20\n",
            "500/500 [==============================] - 43s 87ms/step - loss: 0.3442 - val_loss: 0.4443\n",
            "Epoch 4/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.3251 - val_loss: 0.4581\n",
            "Epoch 5/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.3088 - val_loss: 0.4523\n",
            "Epoch 6/20\n",
            "500/500 [==============================] - 43s 86ms/step - loss: 0.3013 - val_loss: 0.4321\n",
            "Epoch 7/20\n",
            "500/500 [==============================] - 43s 86ms/step - loss: 0.2927 - val_loss: 0.4610\n",
            "Epoch 8/20\n",
            "500/500 [==============================] - 43s 86ms/step - loss: 0.2860 - val_loss: 0.4533\n",
            "Epoch 9/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2772 - val_loss: 0.4547\n",
            "Epoch 10/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2733 - val_loss: 0.4528\n",
            "Epoch 11/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2722 - val_loss: 0.4619\n",
            "Epoch 12/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2657 - val_loss: 0.4511\n",
            "Epoch 13/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2616 - val_loss: 0.4462\n",
            "Epoch 14/20\n",
            "500/500 [==============================] - 43s 86ms/step - loss: 0.2572 - val_loss: 0.4491\n",
            "Epoch 15/20\n",
            "500/500 [==============================] - 43s 86ms/step - loss: 0.2554 - val_loss: 0.4567\n",
            "Epoch 16/20\n",
            "500/500 [==============================] - 43s 87ms/step - loss: 0.2502 - val_loss: 0.4552\n",
            "Epoch 17/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2512 - val_loss: 0.4524\n",
            "Epoch 18/20\n",
            "500/500 [==============================] - 44s 87ms/step - loss: 0.2486 - val_loss: 0.4506\n",
            "Epoch 19/20\n",
            "500/500 [==============================] - 43s 87ms/step - loss: 0.2455 - val_loss: 0.4851\n",
            "Epoch 20/20\n",
            "500/500 [==============================] - 43s 87ms/step - loss: 0.2460 - val_loss: 0.4573\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "VjW8VstLK5Kg",
        "outputId": "96405759-f80f-402b-b9e7-0388cfc16fe3"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(loss))\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": 34,
      "outputs": [
        {
          "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": "ZAzOE_MaK_HK"
      },
      "source": [
        "# This was previously set to 6 (one point per hour).\n",
        "# Now 3 (one point per 30 min).\n",
        "step = 3\n",
        "lookback = 720  # Unchanged\n",
        "delay = 144 # Unchanged\n",
        "\n",
        "train_gen = generator(float_data,\n",
        "                      lookback=lookback,\n",
        "                      delay=delay,\n",
        "                      min_index=0,\n",
        "                      max_index=200000,\n",
        "                      shuffle=True,\n",
        "                      step=step)\n",
        "val_gen = generator(float_data,\n",
        "                    lookback=lookback,\n",
        "                    delay=delay,\n",
        "                    min_index=200001,\n",
        "                    max_index=300000,\n",
        "                    step=step)\n",
        "test_gen = generator(float_data,\n",
        "                     lookback=lookback,\n",
        "                     delay=delay,\n",
        "                     min_index=300001,\n",
        "                     max_index=None,\n",
        "                     step=step)\n",
        "val_steps = (300000 - 200001 - lookback) // 128\n",
        "test_steps = (len(float_data) - 300001 - lookback) // 128"
      ],
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h-i90x7mLFcP",
        "outputId": "47c1b821-d9da-481b-b168-db1de6a69464"
      },
      "source": [
        "model = Sequential()\n",
        "model.add(layers.Conv1D(32, 5, activation='relu',\n",
        "                        input_shape=(None, float_data.shape[-1])))\n",
        "model.add(layers.MaxPooling1D(3))\n",
        "model.add(layers.Conv1D(32, 5, activation='relu'))\n",
        "model.add(layers.GRU(32, dropout=0.1, recurrent_dropout=0.5))\n",
        "model.add(layers.Dense(1))\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer=RMSprop(), loss='mae')\n",
        "history = model.fit_generator(train_gen,\n",
        "                              steps_per_epoch=500,\n",
        "                              epochs=20,\n",
        "                              validation_data=val_gen,\n",
        "                              validation_steps=val_steps)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_3\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv1d_3 (Conv1D)            (None, None, 32)          2272      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_2 (MaxPooling1 (None, None, 32)          0         \n",
            "_________________________________________________________________\n",
            "conv1d_4 (Conv1D)            (None, None, 32)          5152      \n",
            "_________________________________________________________________\n",
            "gru_1 (GRU)                  (None, 32)                6336      \n",
            "_________________________________________________________________\n",
            "dense_4 (Dense)              (None, 1)                 33        \n",
            "=================================================================\n",
            "Total params: 13,793\n",
            "Trainable params: 13,793\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/20\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "500/500 [==============================] - 91s 177ms/step - loss: 0.3636 - val_loss: 0.2791\n",
            "Epoch 2/20\n",
            "500/500 [==============================] - 88s 175ms/step - loss: 0.2998 - val_loss: 0.2858\n",
            "Epoch 3/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2856 - val_loss: 0.2776\n",
            "Epoch 4/20\n",
            "500/500 [==============================] - 87s 175ms/step - loss: 0.2736 - val_loss: 0.2799\n",
            "Epoch 5/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2706 - val_loss: 0.2829\n",
            "Epoch 6/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2612 - val_loss: 0.2789\n",
            "Epoch 7/20\n",
            "500/500 [==============================] - 88s 177ms/step - loss: 0.2537 - val_loss: 0.2878\n",
            "Epoch 8/20\n",
            "500/500 [==============================] - 89s 177ms/step - loss: 0.2490 - val_loss: 0.2869\n",
            "Epoch 9/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2421 - val_loss: 0.2891\n",
            "Epoch 10/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2389 - val_loss: 0.2882\n",
            "Epoch 11/20\n",
            "500/500 [==============================] - 87s 175ms/step - loss: 0.2334 - val_loss: 0.3090\n",
            "Epoch 12/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2283 - val_loss: 0.2968\n",
            "Epoch 13/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2238 - val_loss: 0.2917\n",
            "Epoch 14/20\n",
            "500/500 [==============================] - 88s 177ms/step - loss: 0.2200 - val_loss: 0.3005\n",
            "Epoch 15/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2175 - val_loss: 0.2959\n",
            "Epoch 16/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2125 - val_loss: 0.3165\n",
            "Epoch 17/20\n",
            "500/500 [==============================] - 89s 177ms/step - loss: 0.2078 - val_loss: 0.3076\n",
            "Epoch 18/20\n",
            "500/500 [==============================] - 88s 177ms/step - loss: 0.2060 - val_loss: 0.3008\n",
            "Epoch 19/20\n",
            "500/500 [==============================] - 88s 176ms/step - loss: 0.2038 - val_loss: 0.3057\n",
            "Epoch 20/20\n",
            "500/500 [==============================] - 88s 177ms/step - loss: 0.1993 - val_loss: 0.3073\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "_-w6fbo9LJRa",
        "outputId": "40ea2248-10a8-49d6-f571-00efc8da434a"
      },
      "source": [
        "loss = history.history['loss']\n",
        "val_loss = history.history['val_loss']\n",
        "\n",
        "epochs = range(len(loss))\n",
        "\n",
        "plt.figure()\n",
        "\n",
        "plt.plot(epochs, loss, 'bo', label='Training loss')\n",
        "plt.plot(epochs, val_loss, 'b', label='Validation loss')\n",
        "plt.title('Training and validation loss')\n",
        "plt.legend()\n",
        "\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAEICAYAAABRSj9aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3de5yN5fr48c9lHAcdZKIcBtUmSmQcYhMdiah+OkghlSjfXd+y24oih/b+xu7r265tp9JBbNLBptKBnDoyNMYhCpEpInJWjLl+f9zPYs2YNbNm1nHWXO/Xa71mred4rWetuZ573ff93I+oKsYYYxJXmVgHYIwxJrIs0RtjTIKzRG+MMQnOEr0xxiQ4S/TGGJPgLNEbY0yCs0RvikRE5opI33AvG0sisllErojAdlVEzvWe/0tEHgtm2WLsp7eIfFTcOAvYbkcRyQr3dk30lY11ACbyROSA38tk4HfgmPf6HlWdGuy2VLVLJJZNdKo6MBzbEZF6wPdAOVXN9rY9FQj6MzSljyX6UkBVq/iei8hm4C5VnZd3OREp60sexpjEYVU3pZjvp7mI/EVEtgMvi8jpIvKuiOwUkV+957X91lkoInd5z/uJyKciMt5b9nsR6VLMZeuLyGIR2S8i80TkORF5PUDcwcQ4WkQ+87b3kYhU95t/u4hsEZFdIjKsgOPTWkS2i0iS37TrRSTTe95KRL4QkT0isk1EnhWR8gG29YqIjPF7/WdvnZ9EpH+eZbuKyNcisk9EtorISL/Zi72/e0TkgIhc4ju2fuu3FZFlIrLX+9s22GNTEBE531t/j4isEZHufvOuEZG13jZ/FJEh3vTq3uezR0R2i8gSEbG8E2V2wE1NoBqQCgzAfSde9l7XBQ4DzxawfmtgPVAdeAp4SUSkGMtOA5YCZwAjgdsL2GcwMd4K3AGcCZQHfImnMTDR2/7Z3v5qkw9V/Qo4CFyWZ7vTvOfHgP/23s8lwOXAvQXEjRdDZy+eK4HzgLztAweBPsBpQFdgkIhc583r4P09TVWrqOoXebZdDXgPeMZ7b08D74nIGXnew0nHppCYywFzgI+89f4LmCoiDb1FXsJVA1YFLgA+8aY/BGQBKUAN4FHAxl2JMkv0JgcYoaq/q+phVd2lqm+p6iFV3Q+MBS4tYP0tqvqCqh4DXgXOwv1DB72siNQFWgKPq+oRVf0UmB1oh0HG+LKqfquqh4E3gGbe9J7Au6q6WFV/Bx7zjkEg/wZ6AYhIVeAabxqqulxVv1TVbFXdDDyfTxz5ucmLb7WqHsSd2Pzf30JVXaWqOaqa6e0vmO2COzF8p6pTvLj+DawDrvVbJtCxKUgboArwN+8z+gR4F+/YAEeBxiJyiqr+qqor/KafBaSq6lFVXaI2wFbUWaI3O1X1N98LEUkWkee9qo19uKqC0/yrL/LY7nuiqoe8p1WKuOzZwG6/aQBbAwUcZIzb/Z4f8ovpbP9te4l2V6B94UrvN4hIBeAGYIWqbvHi+INXLbHdi+NJXOm+MLliALbkeX+tRWSBVzW1FxgY5HZ9296SZ9oWoJbf60DHptCYVdX/pOi/3f+HOwluEZFFInKJN30csAH4SEQ2icjQ4N6GCSdL9CZv6eohoCHQWlVP4URVQaDqmHDYBlQTkWS/aXUKWD6UGLf5b9vb5xmBFlbVtbiE1oXc1TbgqoDWAed5cTxanBhw1U/+puF+0dRR1VOBf/ltt7DS8E+4Ki1/dYEfg4irsO3WyVO/fny7qrpMVXvgqnVm4X4poKr7VfUhVW0AdAceFJHLQ4zFFJElepNXVVyd9x6vvndEpHfolZDTgZEiUt4rDV5bwCqhxPgm0E1E/ug1nI6i8P+DacD9uBPKzDxx7AMOiEgjYFCQMbwB9BORxt6JJm/8VXG/cH4TkVa4E4zPTlxVU4MA234f+IOI3CoiZUXkZqAxrpolFF/hSv8Pi0g5EemI+4yme59ZbxE5VVWP4o5JDoCIdBORc722mL24do2CqspMBFiiN3lNACoBvwBfAh9Eab+9cQ2au4AxwAxcf//8FDtGVV0D3IdL3tuAX3GNhQXx1ZF/oqq/+E0fgkvC+4EXvJiDiWGu9x4+wVVrfJJnkXuBUSKyH3gcr3TsrXsI1ybxmdeTpU2ebe8CuuF+9ewCHga65Ym7yFT1CC6xd8Ed938CfVR1nbfI7cBmrwprIO7zBNfYPA84AHwB/FNVF4QSiyk6sXYRE49EZAawTlUj/ovCmERnJXoTF0SkpYicIyJlvO6HPXB1vcaYENmVsSZe1ATexjWMZgGDVPXr2IZkTGKwqhtjjElwVnVjjDEJLu6qbqpXr6716tWLdRjGGFOiLF++/BdVTclvXtwl+nr16pGenh7rMIwxpkQRkbxXRB9nVTfGGJPgLNEbY0yCs0RvjDEJzhK9McYkOEv0xhiT4BIm0U+dCvXqQZky7u9Uu1WyMcYAcdi9sjimToUBA+CQd9uKLVvca4DevQOvZ4wxpUFClOiHDTuR5H0OHXLTjTGmtEuIRP/DD0WbbowxpUlCJPq6eW/EVsh0Y4wpTRIi0Y8dC8nJuaclJ7vpxhhT2iVEou/dGyZNgtRUEHF/J02yhlhjjIEE6XUDLqlbYjfGmJMlRIneGGNMYJbojTEmwVmiN8aYBGeJ3hhjEpwlemOMSXCW6I0xJsFZojfGmAQXVKIXkc4isl5ENojI0HzmDxSRVSKSISKfikhjb/qVIrLcm7dcRC4L9xswxhhTsEITvYgkAc8BXYDGQC9fIvczTVUvVNVmwFPA0970X4BrVfVCoC8wJWyRG2OMCUowJfpWwAZV3aSqR4DpQA//BVR1n9/LyoB6079W1Z+86WuASiJSIfSwjTHGBCuYIRBqAVv9XmcBrfMuJCL3AQ8C5YH8qmj+H7BCVX8vRpzGGGOKKWyNsar6nKqeA/wFGO4/T0SaAP8D3JPfuiIyQETSRSR9586d4QrJGGMMwSX6H4E6fq9re9MCmQ5c53shIrWBd4A+qroxvxVUdZKqpqlqWkpKShAhGWOMCVYwiX4ZcJ6I1BeR8sAtwGz/BUTkPL+XXYHvvOmnAe8BQ1X1s/CEbIwxpigKTfSqmg0MBj4EvgHeUNU1IjJKRLp7iw0WkTUikoGrp+/rmw6cCzzudb3MEJEzw/82jDHGBCKqGusYcklLS9P09PRYh2GMibIhQ+D332H8eKhgffOKTESWq2pafvMS5sYjxpiSKz0d/v5393zlSnj7bahePbYxJRIbAsEYE3OPPQZnnAEvvghLl0KbNrBuXayjShyW6I0xMfX55/DBB/Dww3DnnbBwIezf75L9vHmxji4xWKI3xsTUY4/BmWfCffe5123auFJ93brQuTM8/3xs40sEluiNMTGzcCF88gk88ghUrnxiemoqfPopXH01DBwI//3fcOxYzMIs8SzRG2NiQtWV5s8+2yXzvE45Bf7zH7j/fpgwAXr0cFU6pugs0RtjYuKjj1ypffhwqFgx/2XKlnVJ/p//dPX47drBDz9EN85EYIneGBN1vtJ8aqprgC3MoEHw/vuwZQu0agVffRX5GBOJJXpjTNS9+y4sW+aSffnywa1z1VXwxReQnAwdO8Ibb0Q0xIRiid4YE1U5OfD443DOOdCnT9HWbdzYlebT0uDmm2H0aPfrwBTMEr0xYfLBB7BmTayjiH/vvAMZGTBiBJQrV/T1U1Jc//rbb3cnjNtvh99+C3+cicQSvTFhsHEjdOsGHTrAt9/GOpr4deyYS86NGsGttxZ/OxUqwKuvwtixMHUqXH457NgRvjgTjSV6Y8JgzBhXOi1TBrp2hV9+iXVE8WnGDFi7FkaOhKSk0LYlAo8+6urqV6yA1q1h9eqwhBkzkbpWwEavNCZE333nSqj33w833gidOkGLFjB/fuBug6VRdrarY69Y0VXdlAljMXPpUujeHX791TXwPvxw8I280aYKP/7oxvJZv949fM/PPdd9b4rDRq80JoJGjXJVCX/5C9SoAVOmwE03Qb9+MG1aeBNaSfb66+6k+M474T8mrVq5k8f997tEP2MGvPCCG04hVg4dctV4/ol83To37eDBE8tVqQING0L79u5XSSRYid6YEKxbB02awEMPwVNPnZg+bpwrVQ4dCn/9a+ziixdHjrhfPdWquW6VIpHb15w5cO+9rtQ8eLCrx69aNXL7A1fl8p//uCEdfEnd/8IuEXfNQMOG7jg0bHji+Vlnhed4WInemAgZNQoqVYI//zn39CFDXAPt3/4GDRrA3XfHJr548fLL8P338OyzkU3yANdeC5deCsOGuf3NmgUTJ7q2k3A7fNg1Co8f7z7vqlVPlM79E/p557nvScyoalw9WrRoocaUBKtXq4qoPvJI/vOPHlXt3Fk1KUn1ww+jG1s8OXxYtXZt1TZtVHNyorvvzz5TbdxYFVRvuUV1+/bwbHfXLtXRo1VTUty2W7ZUffNN1ezs8Gy/OIB0DZBXY57Y8z4s0ZuS4sYbVatWVf3ll8DL7N2r2rSpWy4zM3qxxZNnnnGZ5uOPY7P/335TfeIJ1fLlVU8/XXXy5OKfcLZsUX3gAdXKld176tJFdeHC6J/A8mOJ3pgwW7nS/fcMH174slu3qp59tmqdOqo//hj52OLJwYOqNWuqdugQ+2S4dq1qu3buc7v8ctUNG4JfNzNT9fbbVcuWdY/bb3ffgXhSUKK3/gDGFMPIkW4Y3QcfLHzZ2rXd2C67d7v64wMHIh5e3Jg4EbZvd0MVRLpuvjDnnw+LF7uYli6FCy5wDejZ2fkvr+oaV6+5Bpo2dfexHTzY1cW/9pqbVmIEOgPE6mElehPvli93pcKRI4u23nvvqZYpo9qtW2zrcqNl3z7V6tVVr7wy1pGcLCtLtUcP9zk2b+4+U5/sbFff3rKlm5+SojpmjKuXj2eEWqIXkc4isl5ENojI0HzmDxSRVSKSISKfikhjv3mPeOutF5Grw3iOMiYmRo6E006DBx4o2nrXXON6gbz7rltX46tnc9j94x/uCuHRo2Mdyclq1XL9+d98E7Ztg5YtXU+p5593vWV69nS/wCZOdEMjDxvmuoaWWIHOAL4HkARsBBoA5YGVQOM8y5zi97w78IH3vLG3fAWgvredpIL2ZyV6E8+WLnWlvDFjir+Nhx5y2/jf/w1fXKE6dkz1qafcY/fu0Le3Z49r+OzaNfRtRdqvv6refbf7TEA1LU115syS96uLUBpjgUuAD/1ePwI8UsDyvYC5+S0LfAhcUtD+LNGbeNali2q1aq5aoriOHVO94QbXNfOdd8IXW3Ht33+iGgNUk5NV771Xdf364m9zxAi3Lf8qkXiXnq766aexbzQuroISfTBVN7WArX6vs7xpuYjIfSKyEXgK+FMR1x0gIukikr5z584gQjIm+r74AubOdVe8hnKlZZkybpiEli3dCI7LloUvxqLautVd3DNnDjzzDKxc6cZ5f/FFd6FPt25u7JWiVDPt2gX/+79w/fVw8cWRiz3cWrRwtyqMdaNxJISt142qPqeq5wB/AYYXcd1JqpqmqmkpKSnhCsmYsBoxwo2Fft99oW8rORlmz4aaNV1PnM2bQ99mUS1d6saI2bgR3nsP/uu/XE+SyZPd5fsjR7qT0BVXwEUXuenBjPs+fry7ifcTT0T8LZggBZPofwTq+L2u7U0LZDpwXTHXNSYuLVkCH3/sBi6rUiU826xRwyXY3393l+fv2ROe7QbjjTfcMAGVKrlfKp07nxzbiBGuIXLyZDftzjvdeC0jR8LPP+e/3R073C+Dm2+GCy+M6FswRRGoTkdP1KuXBTbhGlN9jbFN8ixznt/za/HqioAm5G6M3UScNsa+/rpqaqqrN01Nda+N8enUSbVGDXcBULgtWKBarpzqZZep/v57+LfvLyfHXSUKqn/8o+qOHcGvN3++6xoK7irTfv1OvmjowQddF9J168IfuykYoV4ZC1wDfIvrNTPMmzYK6O49/z9gDZABLPA/EQDDvPXWA10K21csEv3rr7sGKF9jlK9BypK9UVX95BP3nZgwIXL7ePVVt4927VTnzYtMg+ChQ6q9ern99OnjhgYojvXrXWOt73/msstU58xxfdMrVnTbNtEXcqKP5iMWiT41NXeS9z1SU6MeiokzOTmq7du7IQwOHYrsviZPdvsB1UsucRdYhSvhb9um2rq12/Zf/xqe7e7apfq3v6nWquW2W7myG8CtKEMLmPApKNHbEAjkHjc6mOmm9Jg/39XPP/po5IeZveMO1zA6cSL89JOrt2/Rwl16n5NT/O1mZrpG11Wr3LaGDg1Pz5Jq1Vybxfffw7//7XrYPPIInHNO6Ns24WU3HgHq1XONTnmlpsamN4SJD6quu93WrbBhg7uLVLQcPeruyPTkk27fjRu7qzNvugnKFuEuEnPmQK9e7kreOXOgefPIxWxiq6Abj1iJHncHmuTk3NOSk910U3p99JHrkTJ8eHSTPLgbjd9xh7tb0bRprgTeu7cbmOvll92JoCCq8Pe/Q48ebp2lSy3Jl2aW6HH/QJMmuRK875Zfkya56aZ0UoXHH3ffhTvuiF0cSUmuRJ6Z6apdqlaF/v3dHYsmTsy/X/uRI+6OVkOGuDFbFi2Cs8+Ofuwmflii9/Tu7appcnLcX0vypdv777tS8GOPQfnysY7GXU17/fWwfLnre3/WWe6+qA0auKtQfTeb3rULrroKXnrJxT59+sm/Vk3pY3X0xuShCmlp7gKmdetcNUq8UYUFC9zIkAsXuit2773X1etnZblEb4WV0sXq6I0pgtmzYcUKVyKOxyQProrxsstcsl+yxPXOeeIJN/TAggWW5E1uVqI3xk9OjusmePAgfPNN0Xq4xNqqVa5kX7NmrCMxsVBQib4EfY2Nibx33nEjOE6ZUrKSPNjYMiawEvZVNiZ/Bw+6nik5OXDmma5ke+qpRbswKCfHDeTVsKHr6WJMorBEb0qco0ddNcWyZa5nzLJlsGbNyVePlivnEn5Kyonk7/ub37S5c912/v1v163RmERhid6EzdatsHcvVK8OZ5wRnobMnBxYv94lc98jI8MN7QvuMvyWLd2FQWlpULGiGyp3586T/27c6J4fOBB4f02awI03hh63MfHEEr0Ji8mTYeDA3FdsnnaaS/rVq7sSs+953te+56ee6sYX8k/qy5fDvn1ue5Uru94lgwe75N6yJdSvX/RxWw4fdonf/ySwc6e7kXXPnlaaN4nHEr0JSU6OGyRr3Di48kq46y530Y4vcf7yi3u+dSt8/bV77iuN51WmzInql3Ll3F2Nevc+kdTPPz88SbhSJahb1z2MKQ0s0fvZsMFd8h6vfafjzYEDcNtt8J//wKBB8H//V/ixU3UNp74TgO9k4HucfbZL6hddFP3xZYxJVJbocSXQIUPglVegdWuYORPq1Cl0tVItK8vd6zQz0906bvDg4KpQRNyt+KpUcaOGGmMir1RfGavqxgI5/3x36Xj//rB2rbtgZt68WEcXv9LTT9xU+t133U2lwzG+uTEmMkptov/hB+jWzfWXrlfPNfq99JJrAKxRww0MNXZs8Dd8mDrVbadMGfd36tQIBh9Db74JHTq4gb4+/xy6dIl1RMaYwpS6RH/smKtLbtzYDd86YYIbc7xpUze/YUP46it3Ahg+HK67Dn79teBtTp0KAwa4m5eour8DBiRWsld1N8G48UZo1sz1X7/gglhHZYwJSqB7DMbqEcl7xmZmqrZq5e5v2aWL6ubNgZfNyVH9xz9Uy5ZVbdBA9euvAy8by3vOHjum+tVXqk88odqpk7tp87Jl4b259G+/qd5+u3tPt96qevhw+LZtjAkPSvvNwQ8fVn30UZe0U1JUp00LPhF+/rm7+XHFiqovv5z/MiL5J3qRsL2FXHbsUJ0yRbV3b9Xq1U/s66KLXJygeuGFqk8/7ZYNdV/t2rltjhoV3hOIMSZ8Qk70QGdgPbABGJrP/AeBtUAmMB9I9Zv3FLAG+AZ4Bm/EzECPcCf6hQtVzzvPvdO+fVV/+aXo2/j5Z1daBtW77z65RBvpEn12tjvhPPaYasuWJ04sKSmupD1tmurOnW7ZX39V/de/TvxyKVtW9frrVWfPVj16tGj7XbNGtX59d/KYMSM878UYExkhJXogCdgINADKAyuBxnmW6QQke88HATO8522Bz7xtJAFfAB0L2l+4Ev3u3ap33eXeYYMGqh9/HNr2jh5VHTrUba9FC9Xvvz8x7/XXVZOTcyf55GQ3vbi2b1d95RXVW25RrVbNbbNMGdW2bVVHj3bVM8eOFbyN1atVH3pI9cwz3fo1a6r++c+qa9cWvv8PPlA95RTVGjVc1ZAxJr6FmugvAT70e/0I8EgByzcHPvNbdzlQCUgG0oHzC9pfqIk+J0d15kyXoJKSXGI7eDCkTeYya5ZLgNWqqc6de2L666+7EryI+1vUJH/woOqiRarDhqlefPGJE0bNmqr9+rkS9a5dxYv5yBEXd48eroQPqm3aqD7/vOqePScv/49/uJNK06aqW7YUb5/GmOgKNdH3BF70e3078GwByz8LDPd7PR7YA+wFxgZYZ4B3EkivW7dusd/o1q2q3bu7d9W8uery5cXeVIG++84lQRHVkSMLL1nndeyY6jffuDr/gQNdrElJLu6kJNX27VWffNI1ABd124XZvl11/HjVJk3c/ipVUr3tNtVPPnEnhPvuc9OvvVZ1377w7tsYEzlRS/TAbcCXQAXv9bnAe0AV7/EF0L6g/RW3RL9+vWrVqi5xjRtX9Proojp48ERPlM6dCy5t79yp+u67ro79yitVTz31RIn9lFNUr7jCleRnz3Z17NGQk6O6dKk70fjiqVrV/X3oIdcuYIwpOQpK9MEMgfAj4D8gQG1vWi4icgUwDLhUVX3DVl0PfKmqB7xl5nrVOUuC2G+RnHeeu0LzzjuhQYNwb/1kycnw6qvQti3cf7+7mvatt1zf8owM1xf/q6/gyy9h0ya3Tpkyrr/+Lbe4oRbatHH99svE4GoGkRODhT39tLuz0qxZ0LUr9O0b/XiMMZFT6D1jRaQs8C1wOS7BLwNuVdU1fss0B94EOqvqd37TbwbuxvXaEeADYIKqzgm0v5J4z9ilS93wttu3uwR65IibXquWS+atW7tHixZuqF1jjAm3kO4Zq6rZIjIY+BDXc2ayqq4RkVG4nwqzgXG4qpmZ4gY9+UFVu+OS/2XAKkCBDwpK8iVVq1awYgU88YQbAteX3GvVinVkxhgTRIk+2kpiid4YY2KtoBJ9qRvrxhhjShtL9MYYk+As0RtjTIKzRG+MMQnOEr0xxiQ4S/RxorTcocoYE312c/A44LtD1aFD7rXvDlUAvXvHLi5jTGKwEn0cGDbsRJL3OXTITTfGmFBZoo8DP/xQtOnGGFMUlujjQN26RZtujDFFYYk+Dowd60bD9Jec7KYHyxpzjTGBWKKPA717w6RJkJrqRr9MTXWvg22I9TXmbtniRrn3NeZasjfGgA1qlhDq1XPJPa/UVNi8OdrRGGNiwQY1S3DWmGuMKYgl+gRgjbnGmIJYok8A4WjMNcYkLkv0CSDUxlxjTGKzIRASRO/eltiNMfmzEr0xxiQ4S/TGGJPgLNEbY0yCs0RvjDEJLqhELyKdRWS9iGwQkaH5zH9QRNaKSKaIzBeRVL95dUXkIxH5xlumXvjCN+FiY+UYk7gKTfQikgQ8B3QBGgO9RKRxnsW+BtJUtSnwJvCU37zXgHGqej7QCtgRjsBN+NhYOcYktmBK9K2ADaq6SVWPANOBHv4LqOoCVfXdOuNLoDaAd0Ioq6ofe8sd8FvOxAm78YkxiS2YRF8L2Or3OsubFsidwFzv+R+APSLytoh8LSLjvF8IuYjIABFJF5H0nTt3Bhu7CRMbK8eYxBbWxlgRuQ1IA8Z5k8oC7YEhQEugAdAv73qqOklV01Q1LSUlJZwhmSDYWDnGJLZgEv2PQB2/17W9abmIyBXAMKC7qv7uTc4CMrxqn2xgFnBxaCGbcLOxcoxJbMEk+mXAeSJSX0TKA7cAs/0XEJHmwPO4JL8jz7qniYivmH4ZsDb0sE042Vg5xiS2Qse6UdVsERkMfAgkAZNVdY2IjALSVXU2rqqmCjBTRAB+UNXuqnpMRIYA88XNWA68EKk3Y4rPxsoxJnHZHaaMMSYB2B2mjDGmFLNEb4wxCc4SvTHGJDhL9CYsbKwcY+KX3WHKhMw3Vo5vGAXfWDlgPXmMiQdWojchs7FyjIlvluhNyGysHGPimyV6EzIbK8eY+GaJ3oTMxsoxJr5Zojchs7FyjIlv1uvGhIWNlWNM/LISvTHGJDhL9CYu2AVXxkSOVd2YmLMLroyJLCvRm5izC66MiSxL9Cbm7IIrYyLLEr2JObvgypjIskRvYs4uuDImsizRm5izC66MiSzrdWPigl1wZUzkWIneJATrh29MYEElehHpLCLrRWSDiAzNZ/6DIrJWRDJFZL6IpOaZf4qIZInIs+EK3BgfXz/8LVtA9UQ/fEv2xjiFJnoRSQKeA7oAjYFeItI4z2JfA2mq2hR4E3gqz/zRwOLQwzXmZNYP35iCBVOibwVsUNVNqnoEmA708F9AVReoqu9f7Uugtm+eiLQAagAfhSdkY3KzfvjGFCyYRF8L2Or3OsubFsidwFwAESkD/B0YUtAORGSAiKSLSPrOnTuDCMmYE6wfvjEFC2tjrIjcBqQB47xJ9wLvq2pWQeup6iRVTVPVtJSUlHCGZEoB64dvTMGCSfQ/AnX8Xtf2puUiIlcAw4Duqvq7N/kSYLCIbAbGA31E5G8hRWxMHuHoh2+9dkwiE1UteAGRssC3wOW4BL8MuFVV1/gt0xzXCNtZVb8LsJ1+uAbbwQXtLy0tTdPT04vyHowJSd7RM8H9IrCLtkxJIiLLVTUtv3mFluhVNRsYDHwIfAO8oaprRGSUiHT3FhsHVAFmikiGiMwOU+zGRJz12jGJrtASfbRZid5EW5kyrv99XiKQkxP9eIwpjpBK9MYkOuu1YxKdJXpT6oWj14415pp4ZonelHqh9tqxIRhMvLM6emNCVK+eS+55pabC5s3RjsaUVlZHb0wE2RAMJt5ZojcmRNaYa+KdJXpjQmRDMJh4Z1ckzvUAABIoSURBVInemBDZrRBNvLNbCRoTBnYrRBPPrERvjDEJzhK9MXHALrgykWRVN8bEWN7RM30XXIFVB5nwsBK9MTFmo2eaSLNEb0yM2QVXJtIs0RsTY3bBlYk0S/TGxJhdcGUizRK9MTFm97w1kWa9boyJA6FccGW9dkxhrERvTAlnvXZMYSzRG1PCWa8dUxhL9MaUcNZrxxQmqEQvIp1FZL2IbBCRofnMf1BE1opIpojMF5FUb3ozEflCRNZ4824O9xswprSzXjumMIUmehFJAp4DugCNgV4i0jjPYl8DaaraFHgTeMqbfgjoo6pNgM7ABBE5LVzBG2NsmGRTuGB63bQCNqjqJgARmQ70ANb6FlDVBX7Lfwnc5k3/1m+Zn0RkB5AC7Ak9dGOMjw2TbAoSTNVNLWCr3+ssb1ogdwJz804UkVZAeWBjUQI0xkSe9cNPbGHtRy8itwFpwKV5pp8FTAH6qmpOPusNAAYA1LUWJGOiyvrhJ75gSvQ/AnX8Xtf2puUiIlcAw4Duqvq73/RTgPeAYar6ZX47UNVJqpqmqmkpKSlFid8YEyLrh5/4ginRLwPOE5H6uAR/C3Cr/wIi0hx4Huisqjv8ppcH3gFeU9U3ixvk0aNHycrK4rfffivuJkwUVaxYkdq1a1OuXLlYh2KCEI5++FOnuhPDDz+4bp1jx9qvgXhSaKJX1WwRGQx8CCQBk1V1jYiMAtJVdTYwDqgCzBQRgB9UtTtwE9ABOENE+nmb7KeqGUUJMisri6pVq1KvXj287Zs4pars2rWLrKws6tevH+twTBDq1nXVNflND4ZV/cQ/UdVYx5BLWlqapqen55r2zTff0KhRI0vyJYSqsm7dOs4///xYh2KCkDdRg+uHH2wXzXr18j9RpKbC5s3hitIURkSWq2pafvNKzJWxluRLDvusSpZQ++HbEAzxz0avNMaE1A8/1KofE3klpkRfFOHuE7xr1y6aNWtGs2bNqFmzJrVq1Tr++siRIwWum56ezp/+9KdC99G2bdvQgvQsXLiQbt26hWVbxgTDhmCIfwlXoo9Ew9AZZ5xBRoZrPx45ciRVqlRhyJAhx+dnZ2dTtmz+hzItLY20tHyrzXL5/PPPixecMTHm+7+yXjfxK+FK9NHqE9yvXz8GDhxI69atefjhh1m6dCmXXHIJzZs3p23btqxfvx7IXcIeOXIk/fv3p2PHjjRo0IBnnnnm+PaqVKlyfPmOHTvSs2dPGjVqRO/evfE1mL///vs0atSIFi1a8Kc//anQkvvu3bu57rrraNq0KW3atCEzMxOARYsWHf9F0rx5c/bv38+2bdvo0KEDzZo144ILLmDJkiXhPWAmofXu7Rpec3LcX0vy8SXhSvTRbBjKysri888/JykpiX379rFkyRLKli3LvHnzePTRR3nrrbdOWmfdunUsWLCA/fv307BhQwYNGnRSf/Ovv/6aNWvWcPbZZ9OuXTs+++wz0tLSuOeee1i8eDH169enV69ehcY3YsQImjdvzqxZs/jkk0/o06cPGRkZjB8/nueee4527dpx4MABKlasyKRJk7j66qsZNmwYx44d41Des6UxEWT98CMr4RJ9NBuGbrzxRpKSkgDYu3cvffv25bvvvkNEOHr0aL7rdO3alQoVKlChQgXOPPNMfv75Z2rXrp1rmVatWh2f1qxZMzZv3kyVKlVo0KDB8b7pvXr1YtKkSQXG9+mnnx4/2Vx22WXs2rWLffv20a5dOx588EF69+7NDTfcQO3atWnZsiX9+/fn6NGjXHfddTRr1iykY2NMsKwffuQlXNVNNBuGKleufPz5Y489RqdOnVi9ejVz5swJeBVvhQoVjj9PSkoiOzu7WMuEYujQobz44oscPnyYdu3asW7dOjp06MDixYupVasW/fr147XXXgvrPo0JxIZgiLyES/SxGpt779691KrlBvV85ZVXwr79hg0bsmnTJjZ7V6DMmDGj0HXat2/PVK/L0cKFC6levTqnnHIKGzdu5MILL+Qvf/kLLVu2ZN26dWzZsoUaNWpw9913c9ddd7FixYqwvwdj8mP98CMv4apuIDZjcz/88MP07duXMWPG0LVr17Bvv1KlSvzzn/+kc+fOVK5cmZYtWxa6jq/xt2nTpiQnJ/Pqq68CMGHCBBYsWECZMmVo0qQJXbp0Yfr06YwbN45y5cpRpUoVK9GbqLF++FGgqnH1aNGihea1du3ak6aVRvv371dV1ZycHB00aJA+/fTTMY4oMPvMTLBef101OVkVTjySk930omwjNVVVxP0tyrqJAjf2WL55NeGqbhLZCy+8QLNmzWjSpAl79+7lnnvuiXVIxoQs1OpWX2Puli3uNOFrzLWbp5xQYgY1swGyShb7zEy02KBqTkIMamaMMfkJ13j6iXwrRUv0xpgSLVCjbVHH00/kqh9L9MaYEi3Ua2dKQz9+S/TGmBLNxtMvnCX6IHTq1IkPP/ww17QJEyYwaNCggOt07NgRX6PyNddcw549e05aZuTIkYwfP77Afc+aNYu1a9cef/34448zb968ooSfLxvO2CSSUAZVC7XqpySwRB+EXr16MX369FzTpk+fHtTAYuBGnTzttNOKte+8iX7UqFFcccUVxdqWMeZk4Rg2Jd4bc0vclbEPPAAZRbq1eOGaNYMJEwLP79mzJ8OHD+fIkSOUL1+ezZs389NPP9G+fXsGDRrEsmXLOHz4MD179uSJJ544af169eqRnp5O9erVGTt2LK+++ipnnnkmderUoUWLFoDrIz9p0iSOHDnCueeey5QpU8jIyGD27NksWrSIMWPG8NZbbzF69Gi6detGz549mT9/PkOGDCE7O5uWLVsyceJEKlSoQL169ejbty9z5szh6NGjzJw5k0aNGgV8f7t376Z///5s2rSJ5ORkJk2aRNOmTVm0aBH3338/4G4PuHjxYg4cOMDNN9/Mvn37yM7OZuLEibRv3z60D8CYGAp1PP2SMCibleiDUK1aNVq1asXcuXMBV5q/6aabEBHGjh1Leno6mZmZLFq06PiY7/lZvnw506dPJyMjg/fff59ly5Ydn3fDDTewbNkyVq5cyfnnn89LL71E27Zt6d69O+PGjSMjI4Nzzjnn+PK//fYb/fr1Y8aMGaxatep40vWpXr06K1asYNCgQYVWD/mGM87MzOTJJ5+kT58+AMeHM87IyGDJkiVUqlSJadOmcfXVV5ORkcHKlSttlEuTEEKp+ikJjbklrkRfUMk7knzVNz169GD69Om89NJLALzxxhtMmjSJ7Oxstm3bxtq1a2natGm+21iyZAnXX389yd7vxO7dux+ft3r1aoYPH86ePXs4cOAAV199dYHxrF+/nvr16/OHP/wBgL59+/Lcc8/xwAMPAO7EAdCiRQvefvvtArdlwxkbU3wloTE3qBK9iHQWkfUiskFEhuYz/0ERWSsimSIyX0RS/eb1FZHvvEffcAYfTT169GD+/PmsWLGCQ4cO0aJFC77//nvGjx/P/PnzyczMpGvXrgGHJy5Mv379ePbZZ1m1ahUjRowo9nZ8fEMdhzLMsQ1nbEzhwtGYG+k6/kITvYgkAc8BXYDGQC8RaZxnsa+BNFVtCrwJPOWtWw0YAbQGWgEjROT08IUfPVWqVKFTp07079//eCPsvn37qFy5Mqeeeio///zz8aqdQDp06MCsWbM4fPgw+/fvZ86cOcfn7d+/n7POOoujR48eH1oYoGrVquzfv/+kbTVs2JDNmzezYcMGAKZMmcKll15arPdmwxkbU3yhNuZG44KtYEr0rYANqrpJVY8A04Ee/guo6gJV9dVSfQn4bpl0NfCxqu5W1V+Bj4HO4Qk9+nr16sXKlSuPJ/qLLrqI5s2b06hRI2699VbatWtX4PoXX3wxN998MxdddBFdunTJNdTw6NGjad26Ne3atcvVcHrLLbcwbtw4mjdvzsaNG49Pr1ixIi+//DI33ngjF154IWXKlGHgwIHFel8jR45k+fLlNG3alKFDh+YazviCCy6gadOmlCtXji5durBw4cLj73vGjBnHG2uNKa1C7ccfjTr+Qgc1E5GeQGdVvct7fTvQWlUHB1j+WWC7qo4RkSFARVUd4817DDisquPzrDMAGABQt27dFlvyjFBkA2SVPPaZGROcMmVcST4vEdc4HKyoDWomIrcBacC4oqynqpNUNU1V01JSUsIZkjHGxLVoXLAVTKL/Eajj97q2Ny0XEbkCGAZ0V9Xfi7KuMcaUVtG4z3UwiX4ZcJ6I1BeR8sAtwGz/BUSkOfA8Lsnv8Jv1IXCViJzuNcJe5U0rsngbN98EZp+VMcGLxn2uC+1Hr6rZIjIYl6CTgMmqukZERuFuXTUbV1VTBZgpIgA/qGp3Vd0tIqNxJwuAUaq6u6hBVqxYkV27dnHGGWfgbd/EKVVl165dVKxYMdahGFNiRPo+1yXiDlNHjx4lKysr5L7lJjoqVqxI7dq1KVeuXKxDMabUKKgxtkRcGVuuXDnq168f6zCMMaZEsrFujDEmwVmiN8aYBGeJ3hhjElzcNcaKyE5gS6ELBlYd+CVM4USCxRcaiy80Fl9o4jm+VFXN94rTuEv0oRKR9EAtz/HA4guNxRcaiy808R5fIFZ1Y4wxCc4SvTHGJLhETPSTYh1AISy+0Fh8obH4QhPv8eUr4erojTHG5JaIJXpjjDF+LNEbY0yCK5GJPoiblVcQkRne/K9EpF4UY6sjIgu8m6WvEZGT7rUnIh1FZK+IZHiPx6MVn18Mm0Vklbf/9Hzmi4g84x3DTBG5OIqxNfQ7Nhkisk9EHsizTFSPoYhMFpEdIrLab1o1EfnYu/H9x4Huhywifb1lvhORvlGMb5yIrPM+v3dE5LQA6xb4XYhgfCNF5Ee/z/CaAOsW+P8ewfhm+MW2WUQyAqwb8eMXMlUtUQ/cUMkbgQZAeWAl0DjPMvcC//Ke3wLMiGJ8ZwEXe8+rAt/mE19H4N0YH8fNQPUC5l8DzAUEaAN8FcPPezvuYpCYHUOgA3AxsNpv2lPAUO/5UOB/8lmvGrDJ+3u69/z0KMV3FVDWe/4/+cUXzHchgvGNBIYE8fkX+P8eqfjyzP878Hisjl+oj5JYoi/0ZuXe61e9528Cl0uUBrJX1W2qusJ7vh/4BqgVjX2HWQ/gNXW+BE4TkbNiEMflwEZVDeVq6ZCp6mIg770U/L9nrwLX5bPq1cDHqrpbVX8FPgY6RyM+Vf1IVbO9l1/i7vAWEwGOXzCC+X8PWUHxebnjJuDf4d5vtJTERF8L2Or3OouTE+nxZbwv+l7gjKhE58erMmoOfJXP7EtEZKWIzBWRJlENzFHgIxFZ7t2cPa9gjnM03ELgf7BYH8MaqrrNe74dqJHPMvFyHPvjfqHlp7DvQiQN9qqWJgeo+oqH49ce+FlVvwswP5bHLyglMdGXCCJSBXgLeEBV9+WZvQJXFXER8A9gVrTjA/6oqhcDXYD7RKRDDGIokLhbV3YHZuYzOx6O4XHqfsPHZV9lERkGZANTAywSq+/CROAcoBmwDVc9Eo96UXBpPu7/l0piog/mhuPHlxGRssCpwK6oROf2WQ6X5Keq6tt556vqPlU94D1/HygnItWjFZ+33x+9vzuAd3A/kf3Fw43duwArVPXnvDPi4RgCP/uqs7y/O/JZJqbHUUT6Ad2A3t7J6CRBfBciQlV/VtVjqpoDvBBgv7E+fmWBG4AZgZaJ1fEripKY6Au9Wbn32te7oSfwSaAvebh59XkvAd+o6tMBlqnpazMQkVa4zyGaJ6LKIlLV9xzXaLc6z2KzgT5e75s2wF6/aopoCViSivUx9Ph/z/oC/8lnmQ+Bq0TkdK9q4ipvWsSJSGfgYaC7qh4KsEww34VIxeff5nN9gP0G8/8eSVcA61Q1K7+ZsTx+RRLr1uDiPHA9Qr7FtcYP86aNwn2hASrifu5vAJYCDaIY2x9xP+EzgQzvcQ0wEBjoLTMYWIPrQfAl0DbKx6+Bt++VXhy+Y+gfowDPecd4FZAW5Rgr4xL3qX7TYnYMcSecbcBRXD3xnbh2n/nAd8A8oJq3bBrwot+6/b3v4gbgjijGtwFXv+37Hvp6op0NvF/QdyFK8U3xvluZuOR9Vt74vNcn/b9HIz5v+iu+75zfslE/fqE+bAgEY4xJcCWx6sYYY0wRWKI3xpgEZ4neGGMSnCV6Y4xJcJbojTEmwVmiN8aYBGeJ3hhjEtz/BwS9oZofvFSVAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F306hyUYSdlS"
      },
      "source": [
        "model.save(\"/content/drive/MyDrive/636/homework/1dCNN&RNN.h5\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RcamKrVKiyhg"
      },
      "source": [
        "from tensorflow import keras\n",
        "model = keras.models.load_model('/content/drive/MyDrive/636/homework/1dCNN.h5')"
      ],
      "execution_count": 32,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mDzzkbcGokG1",
        "outputId": "523eb05c-f6e8-4ccb-ddd9-e3cdade68c4c"
      },
      "source": [
        "#! mkdir /tmp/my_log_dir\n",
        "callbacks = [\n",
        "    keras.callbacks.TensorBoard(\n",
        "        log_dir='/tmp/my_log_dir',\n",
        "        histogram_freq=1,\n",
        "        embeddings_freq=1,\n",
        "    )\n",
        "]\n",
        "model = Sequential()\n",
        "model.add(layers.Conv1D(32, 5, activation='relu',\n",
        "                        input_shape=(None, float_data.shape[-1])))\n",
        "model.add(layers.MaxPooling1D(3))\n",
        "model.add(layers.Conv1D(32, 5, activation='relu'))\n",
        "model.add(layers.GRU(32, dropout=0.1, recurrent_dropout=0.5))\n",
        "model.add(layers.Dense(1))\n",
        "\n",
        "model.summary()\n",
        "\n",
        "model.compile(optimizer=RMSprop(), loss='mae')\n",
        "history = model.fit_generator(train_gen,\n",
        "                              steps_per_epoch=500,\n",
        "                              epochs=20,\n",
        "                              validation_data=val_gen,\n",
        "                              validation_steps=val_steps,\n",
        "                              callbacks=callbacks\n",
        "                              )"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_4\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv1d_5 (Conv1D)            (None, None, 32)          2272      \n",
            "_________________________________________________________________\n",
            "max_pooling1d_3 (MaxPooling1 (None, None, 32)          0         \n",
            "_________________________________________________________________\n",
            "conv1d_6 (Conv1D)            (None, None, 32)          5152      \n",
            "_________________________________________________________________\n",
            "gru_2 (GRU)                  (None, 32)                6336      \n",
            "_________________________________________________________________\n",
            "dense_5 (Dense)              (None, 1)                 33        \n",
            "=================================================================\n",
            "Total params: 13,793\n",
            "Trainable params: 13,793\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "Epoch 1/20\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "500/500 [==============================] - 93s 180ms/step - loss: 0.3584 - val_loss: 0.2820\n",
            "Epoch 2/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.3000 - val_loss: 0.2877\n",
            "Epoch 3/20\n",
            "500/500 [==============================] - 90s 180ms/step - loss: 0.2871 - val_loss: 0.2886\n",
            "Epoch 4/20\n",
            "500/500 [==============================] - 90s 180ms/step - loss: 0.2752 - val_loss: 0.2859\n",
            "Epoch 5/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.2702 - val_loss: 0.2830\n",
            "Epoch 6/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2609 - val_loss: 0.2886\n",
            "Epoch 7/20\n",
            "500/500 [==============================] - 90s 180ms/step - loss: 0.2574 - val_loss: 0.2848\n",
            "Epoch 8/20\n",
            "500/500 [==============================] - 89s 179ms/step - loss: 0.2501 - val_loss: 0.2892\n",
            "Epoch 9/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.2466 - val_loss: 0.3030\n",
            "Epoch 10/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2404 - val_loss: 0.2900\n",
            "Epoch 11/20\n",
            "500/500 [==============================] - 89s 179ms/step - loss: 0.2361 - val_loss: 0.3162\n",
            "Epoch 12/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2301 - val_loss: 0.3007\n",
            "Epoch 13/20\n",
            "500/500 [==============================] - 90s 180ms/step - loss: 0.2255 - val_loss: 0.2975\n",
            "Epoch 14/20\n",
            "500/500 [==============================] - 90s 180ms/step - loss: 0.2210 - val_loss: 0.3086\n",
            "Epoch 15/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.2188 - val_loss: 0.3078\n",
            "Epoch 16/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.2156 - val_loss: 0.3102\n",
            "Epoch 17/20\n",
            "500/500 [==============================] - 89s 179ms/step - loss: 0.2106 - val_loss: 0.3104\n",
            "Epoch 18/20\n",
            "500/500 [==============================] - 89s 179ms/step - loss: 0.2095 - val_loss: 0.3126\n",
            "Epoch 19/20\n",
            "500/500 [==============================] - 89s 178ms/step - loss: 0.2056 - val_loss: 0.3142\n",
            "Epoch 20/20\n",
            "500/500 [==============================] - 90s 179ms/step - loss: 0.2018 - val_loss: 0.3138\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sC5FQGVGtAoa",
        "outputId": "b74a1ce3-a353-424c-8f6d-8bcd4a5cb0b6"
      },
      "source": [
        "! tensorboard --logdir=/tmp/my_log_dir"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2021-03-16 20:58:30.072532: I tensorflow/stream_executor/platform/default/dso_loader.cc:49] Successfully opened dynamic library libcudart.so.11.0\n",
            "Serving TensorBoard on localhost; to expose to the network, use a proxy or pass --bind_all\n",
            "TensorBoard 2.4.1 at http://localhost:6006/ (Press CTRL+C to quit)\n",
            "^C\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 644
        },
        "id": "iIx_hxNQ1NjD",
        "outputId": "009b9250-a18a-4057-c97f-11c6a13bba86"
      },
      "source": [
        "from keras.utils import plot_model\n",
        "plot_model(model, show_shapes=True, to_file='model.png')"
      ],
      "execution_count": 42,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<IPython.core.display.Image object>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 42
        }
      ]
    }
  ]
}