Model Selection.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model Selection:\n",
    "\n",
    "## Goal:\n",
    "\n",
    "Find the model that appears to provide best validation results.\n",
    "\n",
    "### Validation Metrics:\n",
    "\n",
    "\n",
    "## Some questions we're trying to answer:\n",
    "1. How complex does the model need to be? ie; how many parameters?\n",
    "2. How big does the latent representation need to be? (ie; deep vs shallow netwoorks)\n",
    "3. How sensitive is the model to hyperparameter tuning?\n",
    "4. What hyperparameters are we finally choosing in the end and why?\n",
    "5. Is the validation loss a good metric compared to cluster comparison?\n",
    "\n",
    "\n",
    "## Chosen Models:\n",
    "\n",
    "All models have between 27,000 and 29,000 parameters, roughly equally complex.\n",
    "\n",
    "- **Model A** is 3 layers deep with latent representation dimensions: [12,27,12]\n",
    "- **Model B** is 2 layers deep with latent representation dimensions: [24,54,14]\n",
    "- **Model C** is 2 layers deep with latent representation dimensions: [4,9,5]\n",
    "\n",
    "## Load Training and Validation data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from numpy import save\n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.keras import backend as K\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers, models\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import MinMaxScaler\n",
    "import pickle\n",
    "from IPython.display import Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "tf.keras.backend.clear_session()\n",
    "\n",
    "# gpu = tf.config.experimental.list_physical_devices('GPU')[0]\n",
    "# tf.config.experimental.set_memory_growth(gpu, True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = np.load('/datasets/home/21/321/ee228sp20ta1/G51/train_set.npy')\n",
    "X_valid = np.load('/datasets/home/21/321/ee228sp20ta1/G51/val_set.npy')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scale and Reshape data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# x_train = X_train[:,:,:216].copy()\n",
    "X_train = X_train.reshape((X_train.shape[0]*96,216))\n",
    "\n",
    "# x_valid = X_valid[:,:,:216].copy()\n",
    "X_valid = X_valid.reshape((X_valid.shape[0]*96,216))\n",
    "\n",
    "\n",
    "# print(x_train.shape)\n",
    "# print(x_valid.shape)\n",
    "\n",
    "scaler = MinMaxScaler().fit(X_train) \n",
    "X_train = scaler.transform(X_train)\n",
    "X_valid = scaler.transform(X_valid)\n",
    "\n",
    "s_train_reshaped = X_train.reshape(-1,96,216,1)\n",
    "s_valid_reshaped = X_valid.reshape(-1,96,216,1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_dataset = tf.data.Dataset.from_tensor_slices((s_train_reshaped[:10800,:,:,:], s_train_reshaped[:10800,:,:,:]))\n",
    "valid_dataset = tf.data.Dataset.from_tensor_slices((s_valid_reshaped, s_valid_reshaped))\n",
    "\n",
    "batch_size = 5\n",
    "shuffle_buffer_size = 100\n",
    "\n",
    "\n",
    "train_dataset = train_dataset.shuffle(shuffle_buffer_size).batch(batch_size)\n",
    "valid_dataset = valid_dataset.batch(batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "del X_train, X_valid"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Models:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ModelA():\n",
    "    \n",
    "    #### Encoder ####\n",
    "    model = models.Sequential()\n",
    "    model.add(keras.Input(shape=(96,216,1)))\n",
    "    model.add(layers.Conv2D(48, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(24, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(12, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    \n",
    "    #### Decoder ####\n",
    "    model.add(layers.Conv2D(12, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(24, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(48, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(1,(3, 3), padding='same'))\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ModelB():\n",
    "    \n",
    "    #### Encoder ####\n",
    "    model = models.Sequential()\n",
    "    model.add(keras.Input(shape=(96,216,1)))\n",
    "    model.add(layers.Conv2D(96, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(14, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    \n",
    "    #### Decoder ####\n",
    "    model.add(layers.Conv2D(14, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(96, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(1,(3, 3), padding='same'))\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ModelC():\n",
    "    \n",
    "    #### Encoder ####\n",
    "    model = models.Sequential()\n",
    "    model.add(keras.Input(shape=(96,216,1)))\n",
    "    model.add(layers.Conv2D(48, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(24, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(12, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((2, 2)))\n",
    "    model.add(layers.Conv2D(5, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.MaxPooling2D((3, 3)))\n",
    "    \n",
    "    #### Decoder ####\n",
    "    model.add(layers.Conv2D(5, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(3,3)))\n",
    "    model.add(layers.Conv2D(12, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(24, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(48, (3, 3), activation='relu', padding='same'))\n",
    "    model.add(layers.UpSampling2D(size=(2,2)))\n",
    "    model.add(layers.Conv2D(1,(3, 3), padding='same'))\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model A training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "AE  = ModelA()\n",
    "# print(AE.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train for 2160 steps, validate for 553 steps\n",
      "Epoch 1/125\n",
      "2160/2160 [==============================] - 21s 10ms/step - loss: 0.0069 - val_loss: 0.0045\n",
      "Epoch 2/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0043 - val_loss: 0.0045\n",
      "Epoch 3/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0041 - val_loss: 0.0039\n",
      "Epoch 4/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0038 - val_loss: 0.0036\n",
      "Epoch 5/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0035 - val_loss: 0.0035\n",
      "Epoch 6/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0034 - val_loss: 0.0034\n",
      "Epoch 7/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0033 - val_loss: 0.0033\n",
      "Epoch 8/125\n",
      "2160/2160 [==============================] - 21s 10ms/step - loss: 0.0032 - val_loss: 0.0032\n",
      "Epoch 9/125\n",
      "2160/2160 [==============================] - 21s 10ms/step - loss: 0.0031 - val_loss: 0.0031\n",
      "Epoch 10/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0030 - val_loss: 0.0031\n",
      "Epoch 11/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0030 - val_loss: 0.0030\n",
      "Epoch 12/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0029 - val_loss: 0.0029\n",
      "Epoch 13/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0029 - val_loss: 0.0029\n",
      "Epoch 14/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0029 - val_loss: 0.0029\n",
      "Epoch 15/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0028 - val_loss: 0.0028\n",
      "Epoch 16/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0028 - val_loss: 0.0028\n",
      "Epoch 17/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0028 - val_loss: 0.0028\n",
      "Epoch 18/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0027 - val_loss: 0.0028\n",
      "Epoch 19/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0027 - val_loss: 0.0027\n",
      "Epoch 20/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0027 - val_loss: 0.0027\n",
      "Epoch 21/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0026 - val_loss: 0.0027\n",
      "Epoch 22/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0026 - val_loss: 0.0026\n",
      "Epoch 23/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0026 - val_loss: 0.0026\n",
      "Epoch 24/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0026 - val_loss: 0.0026\n",
      "Epoch 25/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0026 - val_loss: 0.0026\n",
      "Epoch 26/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0026\n",
      "Epoch 27/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 28/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 29/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 30/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 31/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 32/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0025 - val_loss: 0.0025\n",
      "Epoch 33/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0025\n",
      "Epoch 34/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 35/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 36/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0025\n",
      "Epoch 37/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 38/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 39/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 40/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 41/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 42/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 43/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 44/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 45/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0024 - val_loss: 0.0024\n",
      "Epoch 46/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 47/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 48/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 49/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 50/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 51/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 52/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 53/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 54/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 55/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 56/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 57/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 58/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0026\n",
      "Epoch 59/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 60/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 61/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 62/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 63/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 64/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 65/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 66/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 67/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 68/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 69/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 70/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0024\n",
      "Epoch 71/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 72/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 73/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 74/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 75/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 76/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 77/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 78/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 79/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0023 - val_loss: 0.0023\n",
      "Epoch 80/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 81/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 82/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 83/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 84/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 85/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 86/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 87/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 88/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 89/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 90/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 91/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 92/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 93/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 94/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 95/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 96/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 97/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 98/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 99/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 100/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 101/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 102/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 103/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 104/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 105/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 106/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 107/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 108/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 109/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 110/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 111/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0023\n",
      "Epoch 112/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 113/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 114/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 115/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 116/125\n",
      "2160/2160 [==============================] - 19s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 117/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 118/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 119/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 120/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 121/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 122/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 123/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 124/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n",
      "Epoch 125/125\n",
      "2160/2160 [==============================] - 20s 9ms/step - loss: 0.0022 - val_loss: 0.0022\n"
     ]
    }
   ],
   "source": [
    "opt = keras.optimizers.Adam(learning_rate=0.0005)\n",
    "AE.compile(loss='mse', optimizer=opt)\n",
    "history_modelA = AE.fit(train_dataset, validation_data= valid_dataset, epochs=125)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('ggplot')\n",
    "plt.figure(figsize=(10,4))\n",
    "\n",
    "plt.plot(range(len(history_modelA.history['loss']))[1:], history_modelA.history['loss'][1:], label='Training')\n",
    "plt.plot(range(len(history_modelA.history['loss']))[1:], history_modelA.history['val_loss'][1:], label='Validation')\n",
    "plt.title('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.savefig('ModelA2_125epoch.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "cwd = os.getcwd()\n",
    "AE.save_weights(cwd+'/ExpTestModels228/ModelA2/modelA2')\n",
    "\n",
    "import pickle \n",
    "with open(cwd + '/ExpTestModels228/pickles/modelA2.pickle', 'wb') as handle:\n",
    "    pickle.dump(history_modelA.history, handle)\n",
    "\n",
    "\n",
    "## Model 1 with all relu layers, batchsize 5 and mse loss, 30 epochs, 10,800 samples\n",
    "## Model 2 with inner sigmoid layers, batchsize 5 and mse loss, 40 epochs, 10,800 samples\n",
    "## Model 3 with 2 inner sigmoid layers, batchsize 5 and mse loss, 40 epochs, 10,800 samples\n",
    "## Model 4 with all relu layers, batchsize 15 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "## Model 5 with all relu layers, batchsize 10 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "## Model 6 with all relu layers, batchsize 5 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "\n",
    "#lr = 0.01\n",
    "## Model 7 with all relu layers, batchsize 5 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "## Model 8 with all relu layers, batchsize 10 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "\n",
    "#lr = 0.0005\n",
    "## Model 9 with all relu layers, batchsize 15 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "## Model 10 with all relu layers, batchsize 10 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "## Model 11 with all relu layers, batchsize 5 and mse loss, 50 epochs, 10,800 samples, dataset object\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x7f2bc010e390>"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import os\n",
    "cwd = os.getcwd()\n",
    "AE = AE_02_A()\n",
    "AE.load_weights(cwd+'/ExpTestModels228/ModelA2/modelA2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model B Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train for 2160 steps, validate for 553 steps\n",
      "Epoch 1/125\n",
      "2160/2160 [==============================] - 32s 15ms/step - loss: 0.0035 - val_loss: 0.0017\n",
      "Epoch 2/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0015 - val_loss: 0.0014\n",
      "Epoch 3/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0013 - val_loss: 0.0013\n",
      "Epoch 4/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0012 - val_loss: 0.0012\n",
      "Epoch 5/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0011 - val_loss: 0.0011\n",
      "Epoch 6/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0011 - val_loss: 0.0011\n",
      "Epoch 7/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 0.0010 - val_loss: 0.0010\n",
      "Epoch 8/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 9.8777e-04 - val_loss: 0.0011\n",
      "Epoch 9/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 9.5885e-04 - val_loss: 9.3151e-04\n",
      "Epoch 10/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 9.3407e-04 - val_loss: 9.2908e-04\n",
      "Epoch 11/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 9.1344e-04 - val_loss: 9.0395e-04\n",
      "Epoch 12/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.8241e-04 - val_loss: 8.7076e-04\n",
      "Epoch 13/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.6173e-04 - val_loss: 8.5240e-04\n",
      "Epoch 14/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.4529e-04 - val_loss: 8.4444e-04\n",
      "Epoch 15/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.3021e-04 - val_loss: 8.1180e-04\n",
      "Epoch 16/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.1897e-04 - val_loss: 8.0240e-04\n",
      "Epoch 17/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 8.0694e-04 - val_loss: 8.0476e-04\n",
      "Epoch 18/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.9860e-04 - val_loss: 8.1526e-04\n",
      "Epoch 19/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.8874e-04 - val_loss: 8.0508e-04\n",
      "Epoch 20/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.8126e-04 - val_loss: 7.7485e-04\n",
      "Epoch 21/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.7548e-04 - val_loss: 7.6511e-04\n",
      "Epoch 22/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.6993e-04 - val_loss: 7.5827e-04\n",
      "Epoch 23/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.6313e-04 - val_loss: 7.5210e-04\n",
      "Epoch 24/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.5859e-04 - val_loss: 7.6431e-04\n",
      "Epoch 25/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.5371e-04 - val_loss: 7.4587e-04\n",
      "Epoch 26/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.5009e-04 - val_loss: 7.4594e-04\n",
      "Epoch 27/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.4532e-04 - val_loss: 7.3623e-04\n",
      "Epoch 28/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.4144e-04 - val_loss: 7.2902e-04\n",
      "Epoch 29/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.3804e-04 - val_loss: 7.2407e-04\n",
      "Epoch 30/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.3511e-04 - val_loss: 7.3315e-04\n",
      "Epoch 31/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.3193e-04 - val_loss: 7.4323e-04\n",
      "Epoch 32/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.2900e-04 - val_loss: 7.2451e-04\n",
      "Epoch 33/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.2649e-04 - val_loss: 7.4739e-04\n",
      "Epoch 34/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.2365e-04 - val_loss: 7.4253e-04\n",
      "Epoch 35/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.2102e-04 - val_loss: 7.1055e-04\n",
      "Epoch 36/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.1898e-04 - val_loss: 7.1090e-04\n",
      "Epoch 37/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.1564e-04 - val_loss: 7.0924e-04\n",
      "Epoch 38/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.1432e-04 - val_loss: 7.3939e-04\n",
      "Epoch 39/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.1131e-04 - val_loss: 7.0373e-04\n",
      "Epoch 40/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.0923e-04 - val_loss: 7.0214e-04\n",
      "Epoch 41/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.0808e-04 - val_loss: 7.0117e-04\n",
      "Epoch 42/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.0498e-04 - val_loss: 6.9630e-04\n",
      "Epoch 43/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.0245e-04 - val_loss: 6.9340e-04\n",
      "Epoch 44/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 7.0129e-04 - val_loss: 6.9668e-04\n",
      "Epoch 45/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.9937e-04 - val_loss: 6.9110e-04\n",
      "Epoch 46/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.9674e-04 - val_loss: 6.9170e-04\n",
      "Epoch 47/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.9548e-04 - val_loss: 6.8680e-04\n",
      "Epoch 48/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.9337e-04 - val_loss: 6.8377e-04\n",
      "Epoch 49/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.9143e-04 - val_loss: 6.8218e-04\n",
      "Epoch 50/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8956e-04 - val_loss: 6.8569e-04\n",
      "Epoch 51/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8748e-04 - val_loss: 6.7535e-04\n",
      "Epoch 52/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8665e-04 - val_loss: 7.1622e-04\n",
      "Epoch 53/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8437e-04 - val_loss: 6.9132e-04\n",
      "Epoch 54/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8202e-04 - val_loss: 6.7451e-04\n",
      "Epoch 55/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.8034e-04 - val_loss: 6.7754e-04\n",
      "Epoch 56/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7910e-04 - val_loss: 6.6976e-04\n",
      "Epoch 57/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7760e-04 - val_loss: 6.6623e-04\n",
      "Epoch 58/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7627e-04 - val_loss: 6.7466e-04\n",
      "Epoch 59/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7418e-04 - val_loss: 6.6841e-04\n",
      "Epoch 60/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7341e-04 - val_loss: 6.6903e-04\n",
      "Epoch 61/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7112e-04 - val_loss: 6.7934e-04\n",
      "Epoch 62/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.7078e-04 - val_loss: 6.6353e-04\n",
      "Epoch 63/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6855e-04 - val_loss: 6.6077e-04\n",
      "Epoch 64/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6730e-04 - val_loss: 6.5715e-04\n",
      "Epoch 65/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6573e-04 - val_loss: 6.5680e-04\n",
      "Epoch 66/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6419e-04 - val_loss: 6.6884e-04\n",
      "Epoch 67/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6342e-04 - val_loss: 6.6783e-04\n",
      "Epoch 68/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6250e-04 - val_loss: 6.5150e-04\n",
      "Epoch 69/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.6056e-04 - val_loss: 6.6127e-04\n",
      "Epoch 70/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.6014e-04 - val_loss: 6.5296e-04\n",
      "Epoch 71/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5921e-04 - val_loss: 6.5547e-04\n",
      "Epoch 72/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.5709e-04 - val_loss: 6.5056e-04\n",
      "Epoch 73/125\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5640e-04 - val_loss: 6.4761e-04\n",
      "Epoch 74/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5582e-04 - val_loss: 6.7223e-04\n",
      "Epoch 75/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5425e-04 - val_loss: 6.5518e-04\n",
      "Epoch 76/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5373e-04 - val_loss: 6.4476e-04\n",
      "Epoch 77/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5252e-04 - val_loss: 6.4952e-04\n",
      "Epoch 78/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.5157e-04 - val_loss: 6.4765e-04\n",
      "Epoch 79/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.5127e-04 - val_loss: 6.4773e-04\n",
      "Epoch 80/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4946e-04 - val_loss: 6.4277e-04\n",
      "Epoch 81/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4914e-04 - val_loss: 6.5305e-04\n",
      "Epoch 82/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4933e-04 - val_loss: 6.4106e-04\n",
      "Epoch 83/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4603e-04 - val_loss: 6.4274e-04\n",
      "Epoch 84/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4719e-04 - val_loss: 6.4628e-04\n",
      "Epoch 85/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4572e-04 - val_loss: 6.5480e-04\n",
      "Epoch 86/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4509e-04 - val_loss: 6.6965e-04\n",
      "Epoch 87/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4414e-04 - val_loss: 6.5531e-04\n",
      "Epoch 88/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.4389e-04 - val_loss: 6.3667e-04\n",
      "Epoch 89/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4260e-04 - val_loss: 6.4584e-04\n",
      "Epoch 90/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4268e-04 - val_loss: 6.4050e-04\n",
      "Epoch 91/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4130e-04 - val_loss: 6.3420e-04\n",
      "Epoch 92/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4111e-04 - val_loss: 6.3961e-04\n",
      "Epoch 93/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4014e-04 - val_loss: 6.3564e-04\n",
      "Epoch 94/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.4018e-04 - val_loss: 6.3196e-04\n",
      "Epoch 95/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3939e-04 - val_loss: 6.6139e-04\n",
      "Epoch 96/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3818e-04 - val_loss: 6.4363e-04\n",
      "Epoch 97/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3810e-04 - val_loss: 6.3791e-04\n",
      "Epoch 98/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3812e-04 - val_loss: 6.2945e-04\n",
      "Epoch 99/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3621e-04 - val_loss: 6.7106e-04\n",
      "Epoch 100/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3667e-04 - val_loss: 6.3896e-04\n",
      "Epoch 101/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3576e-04 - val_loss: 6.3086e-04\n",
      "Epoch 102/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3506e-04 - val_loss: 6.2910e-04\n",
      "Epoch 103/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3511e-04 - val_loss: 6.3278e-04\n",
      "Epoch 104/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3416e-04 - val_loss: 6.2880e-04\n",
      "Epoch 105/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3373e-04 - val_loss: 6.2990e-04\n",
      "Epoch 106/125\n",
      "2160/2160 [==============================] - 27s 13ms/step - loss: 6.3320e-04 - val_loss: 6.5801e-04\n",
      "Epoch 107/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3353e-04 - val_loss: 6.2610e-04\n",
      "Epoch 108/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3207e-04 - val_loss: 6.3670e-04\n",
      "Epoch 109/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3224e-04 - val_loss: 6.2775e-04\n",
      "Epoch 110/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3137e-04 - val_loss: 6.2215e-04\n",
      "Epoch 111/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3081e-04 - val_loss: 6.4192e-04\n",
      "Epoch 112/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3140e-04 - val_loss: 6.2405e-04\n",
      "Epoch 113/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.3063e-04 - val_loss: 6.2253e-04\n",
      "Epoch 114/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2965e-04 - val_loss: 6.2618e-04\n",
      "Epoch 115/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2950e-04 - val_loss: 6.2409e-04\n",
      "Epoch 116/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2850e-04 - val_loss: 6.3022e-04\n",
      "Epoch 117/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2823e-04 - val_loss: 6.2478e-04\n",
      "Epoch 118/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2792e-04 - val_loss: 6.1894e-04\n",
      "Epoch 119/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.2798e-04 - val_loss: 6.2964e-04\n",
      "Epoch 120/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 6.2703e-04 - val_loss: 6.2350e-04\n",
      "Epoch 121/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2704e-04 - val_loss: 6.2663e-04\n",
      "Epoch 122/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2705e-04 - val_loss: 6.2383e-04\n",
      "Epoch 123/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2600e-04 - val_loss: 6.1757e-04\n",
      "Epoch 124/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2579e-04 - val_loss: 6.4589e-04\n",
      "Epoch 125/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 6.2600e-04 - val_loss: 6.2486e-04\n"
     ]
    },
    {
     "ename": "FileNotFoundError",
     "evalue": "[Errno 2] No such file or directory: '/home/adantulu/ExpModels228/pickles/modelB1.pickle'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mFileNotFoundError\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-8-98d84ee5f69c>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      8\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      9\u001b[0m \u001b[0;32mimport\u001b[0m \u001b[0mpickle\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 10\u001b[0;31m \u001b[0;32mwith\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcwd\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0;34m'/ExpModels228/pickles/modelB1.pickle'\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'wb'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     11\u001b[0m     \u001b[0mpickle\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdump\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mhistory_modelB\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mhistory\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhandle\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     12\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mFileNotFoundError\u001b[0m: [Errno 2] No such file or directory: '/home/adantulu/ExpModels228/pickles/modelB1.pickle'"
     ]
    }
   ],
   "source": [
    "AE = ModelB()\n",
    "\n",
    "opt = keras.optimizers.Adam(learning_rate=0.0005)\n",
    "AE.compile(loss='mse', optimizer=opt)\n",
    "history_modelB = AE.fit(train_dataset, validation_data= valid_dataset, epochs=125)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "cwd = os.getcwd()\n",
    "AE.save_weights(cwd+'/ExpTestModels228/ModelB1/modelB1')\n",
    "\n",
    "import pickle \n",
    "with open(cwd + '/ExpTestModels228/pickles/modelB1.pickle', 'wb') as handle:\n",
    "    pickle.dump(history_modelB.history, handle)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('ggplot')\n",
    "plt.figure(figsize=(10,4))\n",
    "\n",
    "plt.plot(range(len(history_modelB.history['loss']))[1:], history_modelB.history['loss'][1:], label='Training')\n",
    "plt.plot(range(len(history_modelB.history['loss']))[1:], history_modelB.history['val_loss'][1:], label='Validation')\n",
    "plt.title('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.savefig('ModelB1_125epochs.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#lr = 0.001\n",
    "##Model 1 with all relu layers, batchsize 5 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "##Model 2 with all relu layers, batchsize 10 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "##Model 3 with all relu layers, batchsize 15 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "#lr = 0.0005\n",
    "##Model 4 with all relu layers, batchsize 15 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "##Model 5 with all relu layers, batchsize 5 and mse loss, 50 epochs, 10,800 samples, dataset object\n",
    "##Model 6 with all relu layers, batchsize 10 and mse loss, 50 epochs, 10,800 samples, dataset object\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model B Latent representations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_4\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_26 (Conv2D)           (None, 96, 216, 96)       960       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_11 (MaxPooling (None, 48, 108, 96)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_27 (Conv2D)           (None, 48, 108, 14)       12110     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_12 (MaxPooling (None, 24, 54, 14)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_28 (Conv2D)           (None, 24, 54, 14)        1778      \n",
      "_________________________________________________________________\n",
      "up_sampling2d_11 (UpSampling (None, 48, 108, 14)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_29 (Conv2D)           (None, 48, 108, 96)       12192     \n",
      "_________________________________________________________________\n",
      "up_sampling2d_12 (UpSampling (None, 96, 216, 96)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_30 (Conv2D)           (None, 96, 216, 1)        865       \n",
      "=================================================================\n",
      "Total params: 27,905\n",
      "Trainable params: 27,905\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "AETest = ModelB()\n",
    "AETest.load_weights(cwd+'/ExpTestModels228/ModelB1/modelB1')\n",
    "\n",
    "print(AETest.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "decoded_data = AETest.predict(s_valid_reshaped)\n",
    "mini_batch = s_valid_reshaped[:100]\n",
    "#print(mini_batch.shape)\n",
    "get_hl = K.function([AETest.layers[0].input], [AETest.layers[4].output])\n",
    "AE_h1 = get_hl(mini_batch)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 4  # how many digits we will display\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "    # display original\n",
    "    ax = plt.subplot(3, n, i + 1)\n",
    "    plt.imshow(s_valid_reshaped[i].reshape(96, 216))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display hidden layer representation\n",
    "    ax = plt.subplot(3, n, i + 1 + n)\n",
    "    plt.imshow(AE_h1[i].reshape(168, 108)) # Will have to reshape if you change the latent dimensionality\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display reconstruction\n",
    "    ax = plt.subplot(3, n, i + 1 + n + n)\n",
    "    plt.imshow(decoded_data[i].reshape(96, 216))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model C training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_12\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "conv2d_88 (Conv2D)           (None, 96, 216, 48)       480       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_48 (MaxPooling (None, 48, 108, 48)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_89 (Conv2D)           (None, 48, 108, 24)       10392     \n",
      "_________________________________________________________________\n",
      "max_pooling2d_49 (MaxPooling (None, 24, 54, 24)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_90 (Conv2D)           (None, 24, 54, 12)        2604      \n",
      "_________________________________________________________________\n",
      "max_pooling2d_50 (MaxPooling (None, 12, 27, 12)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_91 (Conv2D)           (None, 12, 27, 5)         545       \n",
      "_________________________________________________________________\n",
      "max_pooling2d_51 (MaxPooling (None, 4, 9, 5)           0         \n",
      "_________________________________________________________________\n",
      "conv2d_92 (Conv2D)           (None, 4, 9, 5)           230       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_28 (UpSampling (None, 12, 27, 5)         0         \n",
      "_________________________________________________________________\n",
      "conv2d_93 (Conv2D)           (None, 12, 27, 12)        552       \n",
      "_________________________________________________________________\n",
      "up_sampling2d_29 (UpSampling (None, 24, 54, 12)        0         \n",
      "_________________________________________________________________\n",
      "conv2d_94 (Conv2D)           (None, 24, 54, 24)        2616      \n",
      "_________________________________________________________________\n",
      "up_sampling2d_30 (UpSampling (None, 48, 108, 24)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_95 (Conv2D)           (None, 48, 108, 48)       10416     \n",
      "_________________________________________________________________\n",
      "up_sampling2d_31 (UpSampling (None, 96, 216, 48)       0         \n",
      "_________________________________________________________________\n",
      "conv2d_96 (Conv2D)           (None, 96, 216, 1)        433       \n",
      "=================================================================\n",
      "Total params: 28,268\n",
      "Trainable params: 28,268\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    }
   ],
   "source": [
    "AE = ModelC()\n",
    "print(AE.summary())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train for 2160 steps, validate for 553 steps\n",
      "Epoch 1/125\n",
      "2160/2160 [==============================] - 38s 18ms/step - loss: 0.0111 - val_loss: 0.0082\n",
      "Epoch 2/125\n",
      "2160/2160 [==============================] - 34s 16ms/step - loss: 0.0081 - val_loss: 0.0079\n",
      "Epoch 3/125\n",
      "2160/2160 [==============================] - 34s 16ms/step - loss: 0.0078 - val_loss: 0.0077\n",
      "Epoch 4/125\n",
      "2160/2160 [==============================] - 34s 16ms/step - loss: 0.0077 - val_loss: 0.0076\n",
      "Epoch 5/125\n",
      "2160/2160 [==============================] - 34s 16ms/step - loss: 0.0076 - val_loss: 0.0076\n",
      "Epoch 6/125\n",
      "2160/2160 [==============================] - 32s 15ms/step - loss: 0.0075 - val_loss: 0.0075\n",
      "Epoch 7/125\n",
      "2160/2160 [==============================] - 32s 15ms/step - loss: 0.0075 - val_loss: 0.0074\n",
      "Epoch 8/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0074 - val_loss: 0.0074\n",
      "Epoch 9/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0074 - val_loss: 0.0073\n",
      "Epoch 10/125\n",
      "2160/2160 [==============================] - 32s 15ms/step - loss: 0.0074 - val_loss: 0.0073\n",
      "Epoch 11/125\n",
      "2160/2160 [==============================] - 31s 15ms/step - loss: 0.0073 - val_loss: 0.0073\n",
      "Epoch 12/125\n",
      "2160/2160 [==============================] - 31s 15ms/step - loss: 0.0073 - val_loss: 0.0073\n",
      "Epoch 13/125\n",
      "2160/2160 [==============================] - 32s 15ms/step - loss: 0.0073 - val_loss: 0.0072\n",
      "Epoch 14/125\n",
      "2160/2160 [==============================] - 31s 15ms/step - loss: 0.0073 - val_loss: 0.0072\n",
      "Epoch 15/125\n",
      "2160/2160 [==============================] - 31s 15ms/step - loss: 0.0073 - val_loss: 0.0072\n",
      "Epoch 16/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0072 - val_loss: 0.0072\n",
      "Epoch 17/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0072 - val_loss: 0.0072\n",
      "Epoch 18/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0072 - val_loss: 0.0072\n",
      "Epoch 19/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0072 - val_loss: 0.0072\n",
      "Epoch 20/125\n",
      "2160/2160 [==============================] - 31s 14ms/step - loss: 0.0072 - val_loss: 0.0072\n",
      "Epoch 21/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0072 - val_loss: 0.0071\n",
      "Epoch 22/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0072 - val_loss: 0.0071\n",
      "Epoch 23/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0072 - val_loss: 0.0071\n",
      "Epoch 24/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0072 - val_loss: 0.0071\n",
      "Epoch 25/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0072 - val_loss: 0.0071\n",
      "Epoch 26/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 27/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 28/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 29/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 30/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 31/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 32/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0071 - val_loss: 0.0070\n",
      "Epoch 33/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0071 - val_loss: 0.0070\n",
      "Epoch 34/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0071 - val_loss: 0.0071\n",
      "Epoch 35/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0071 - val_loss: 0.0070\n",
      "Epoch 36/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 37/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 38/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0070 - val_loss: 0.0071\n",
      "Epoch 39/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 40/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 41/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 42/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 43/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 44/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 45/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 46/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 47/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 48/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 49/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 50/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 51/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 52/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 53/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 54/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 55/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0070\n",
      "Epoch 56/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 57/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 58/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0070 - val_loss: 0.0069\n",
      "Epoch 59/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 60/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 61/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 62/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 63/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 64/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 65/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 66/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 67/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 68/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 69/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 70/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 71/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 72/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0070\n",
      "Epoch 73/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 74/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 75/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 76/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 77/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0068\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 78/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0068\n",
      "Epoch 79/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0068\n",
      "Epoch 80/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 81/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0069 - val_loss: 0.0068\n",
      "Epoch 82/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0069 - val_loss: 0.0069\n",
      "Epoch 83/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 84/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 85/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 86/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 87/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 88/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 89/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 90/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 91/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 92/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 93/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 94/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 95/125\n",
      "2160/2160 [==============================] - 30s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 96/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 97/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 98/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 99/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 100/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 101/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 102/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 103/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 104/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 105/125\n",
      "2160/2160 [==============================] - 28s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 106/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 107/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 108/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 109/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 110/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 111/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 112/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 113/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 114/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 115/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 116/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 117/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 118/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 119/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 120/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 121/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 122/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 123/125\n",
      "2160/2160 [==============================] - 29s 13ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 124/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n",
      "Epoch 125/125\n",
      "2160/2160 [==============================] - 29s 14ms/step - loss: 0.0068 - val_loss: 0.0068\n"
     ]
    }
   ],
   "source": [
    "opt = keras.optimizers.Adam(learning_rate=0.0005)\n",
    "AE.compile(loss='mse', optimizer=opt)\n",
    "history_modelA = AE.fit(train_dataset, validation_data= valid_dataset, epochs=125)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmYAAAEJCAYAAAA+ZuuYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOzdeXxU1fn48c+ZJcskk2UmyxAIEEPYlyhhC8oaBRqpuIB1rUDVVsWC1br3ZWuhKBb0C7ZVRFqV/kQti4i4xIBLIhoQBASUsCYkJCEDWcg2k3t/f0RGYoAEmGSS8Lxfr75k5p5z59xH2u/zPefc5yhd13WEEEIIIYTPGXw9ACGEEEIIUUcSMyGEEEKIVkISMyGEEEKIVkISMyGEEEKIVkISMyGEEEKIVkISMyGEEEKIVkISMyGEEEKIVkISMyFEu3bHHXeQkpLi62EIIUSTSGImhBBCCNFKSGImhLholZWVcffddxMZGYm/vz9JSUl89NFH9drMmTOHSy65BH9/fyIjIxk3bhyVlZUA5Obmcv311xMREUFAQACXXHIJ8+bN88WjCCHaCZOvByCEEL4ybdo0srKyeOONN+jcuTP/+te/uPrqq9m2bRs9e/ZkxYoVzJ07l2XLljFgwACcTicbNmzw9L/nnnuoqKggLS2NsLAw9u/fz5EjR3z3QEKINk8SMyHERSk7O5t33nmHtWvXMm7cOABeeOEFPv/8c5599lleffVVDh48iMPhYPz48ZjNZjp37kxiYqLnHgcPHuTaa6/1fNe1a1dfPIoQoh2RpUwhxEVp586dAIwYMaLe9yNGjOC7774DYMqUKbhcLrp06cIdd9zB66+/TllZmaftzJkzmTNnDkOGDOHhhx/ms88+a7kHEEK0S5KYCSHEGXTs2JHdu3fz6quvEhUVxdNPP02PHj3IyckBYOrUqRw8eJDf/va35OfnM2HCBG699VYfj1oI0ZZJYiaEuCj16dMHoMEs12effUbfvn09n/39/Rk/fjzPPvss27dvp6KiglWrVnmud+jQgalTp/Laa6+xZMkSli1bRmlpacs8hBCi3ZE9ZkKIdq+8vJytW7fW+y4gIIDJkydzzz338NJLL9GlSxf++c9/smPHDv773/8CsGTJEjRNY/DgwYSFhfHJJ59QVlZG7969Abjvvvv4xS9+QY8ePaiqqmLFihXExsZitVpb/BmFEO2DJGZCiHbvq6++4tJLL633XY8ePfj666956KGHuPXWWyktLaVfv36899579OzZE4Dw8HCee+45/vjHP1JdXc0ll1zCyy+/zNixYwHQdZ2ZM2eSk5ODxWJh6NChrFu3DqVUiz+jEKJ9ULqu674ehBBCCCGEkD1mQgghhBCthiRmQgghhBCthCRmQgghhBCthCRmQgghhBCthCRmQgghhBCtRLspl5GXl+fV+0VERHD06FGv3vNiJbH0Domj90gsvUPi6D0SS+9oK3GMiYk54zWZMRNCCCGEaCUkMRNCCCGEaCUkMRNCCCGEaCXazR4zIYQQQpw/XdepqqpC07Q2e6xYQUEB1dXVvh4GUBdPg8FAQEDAOcVTEjMhhBBCUFVVhdlsxmRqu6mByWTCaDT6ehgebrebqqoqAgMDm9xHljKFEEIIgaZpbTopa41MJhOapp1TH0nMhBBCCNFmly9bu3ONqyRmTaDv3Y32v/+ga7W+HooQQggh2jFJzJpAP7QX/YP/QXmpr4cihBBCtDtOp5Mrr7ySK6+8ksTERAYOHOj5XFNT06R7zJo1i+zs7LO2+fe//82KFSu8MeRmI4vJTaBCbegAx50QEu7r4QghhBDtis1m4+OPPwbg73//O0FBQfz2t7+t10bXdc+bjqezYMECTCYTbrf7jL9zxx13eG3MzUVmzJoizFb3z+NO345DCCGEuIjs37+fUaNGcd999zF69GgKCgr44x//yIQJExg9ejQLFizwtJ00aRI7duzA7XbTq1cv5syZQ0pKChMnTvQc0/TMM8+wePFiT/s5c+aQmprKFVdcQVZWFgAVFRXceeedjBo1ijvvvJMJEyawY8eOFntmmTFritC6xEw/7kS2RgohhGjvtDcXo+fs9+o9VWwchl/dec79srOzeeGFFxgwYAAAjz76KOHh4bjdbiZPnkxqairdu3ev16e0tJShQ4fy2GOP8dRTT/Hmm29y3333Nbi3ruusXbuWjz76iOeff55ly5bx6quvEhkZyeLFi/nuu+8YP378+T3weZIZs6YIDav7Z8kx345DCCGEuMh06dLFk5QBrF69mnHjxjF+/Hj27NnDDz/80KBPQEAAY8aMAaB///7k5OSc9t4TJkwAoF+/fp42X3/9Nddccw0Affr0oUePHl59nsY0acZs69atLF26FE3TGDt2LJMmTap33eVysWjRIvbt24fVamXmzJlERUUBsHLlStLT0zEYDEydOpXExEQA3nvvPdLT01FKERsbyz333IOfnx+FhYU8//zzlJWVcckllzBjxgyf11VRJjNYQ2UpUwghxEXhfGa2movFYvH8ed++fbzyyiusXbuW0NBQZsyYcdpK/35+fp4/G41GamtPX1XhZLuztWlpjc6YaZrGkiVLeOyxx1iwYAEZGRnk5ubWa5Oenk5QUBALFy4kNTWVZcuWAZCbm0tmZibz58/n8ccfZ8mSJWiahtPpZN26dcydO5e///3vaJpGZmYmAG+88QapqaksXLiQoKAg0tPTm+Gxz0OoDb1EEjMhhBDCV8rLywkODsZqtVJQUMCGDRu8/huDBg1izZo1AOzateu0M3LNqdHELDs7G4fDQXR0NCaTieTkZM8GuZM2bdrEqFGjABg6dCg7duxA13WysrJITk7GbDYTFRWFw+HwvMqqaRo1NTXU1tZSU1NDeHg4uq7z3XffMXToUABGjRrV4Ld8JixcZsyEEEIIH+rXrx8JCQmMGDGC3//+9wwaNMjrvzFt2jSOHDnCqFGjmD9/Pt27dyckJMTrv3Mmja4ROp1O7Ha757PdbmfPnj1nbGM0GrFYLJSVleF0OklISPC0s9lsOJ1OunfvzsSJE/nd736Hn58fAwYMYMCAAZSWlmKxWDznXJ1s3xqoUBt67gFfD0MIIYRo1/7whz94/hwXF+cpowF1VfQXLlx42n6rVq3ylMvYtWuX5/trrrnGs2fs4Ycfrtf+pKioKDIyMgDw9/dn0aJFBAQEsG/fPm6++WZiYmK883BN4JPNW+Xl5WRlZfHiiy9isViYP38+n332mWf/WVOkpaWRlpYGwNy5c4mIiPDqGE0mU717lsd04sSX6djDw1Gt6IDUtuDnsRTnR+LoPRJL75A4ek9riGVBQYHP93R7w4U+w4kTJ7jhhhtwu93ous5zzz1HQEDAed/P39//nP7dNjp6m81GcXGx53NxcTE2m+20bex2O7W1tVRUVGC1Whv0dTqd2Gw2tm/fTlRUlGdqcMiQIfzwww9cccUVVFRUUFtbi9Fo9LQ/nZSUFFJSUjyfT9Yo8ZaIiIh699T8AkDTOLp/Lyrs9GMSp/fzWIrzI3H0Homld0gcvac1xLK6utqzYtVWNVZgtimCgoJYt25dve8u5J7V1dUN/t2ebQau0T1m8fHx5OfnU1hYiNvtJjMzk6SkpHptBg4c6NmAt3HjRvr06YNSiqSkJDIzM3G5XBQWFpKfn0+3bt2IiIhgz549VFdXo+s627dvp2PHjiil6NOnDxs3bgRgw4YNDX7LV9SPtcyQFwCEEEII0UwanTEzGo1MmzaN2bNno2kao0ePJjY2luXLlxMfH09SUhJjxoxh0aJFzJgxg+DgYGbOnAlAbGwsw4YN44EHHsBgMDB9+nQMBgMJCQkMHTqUhx9+GKPRSNeuXT2zX7fccgvPP/88b775JnFxcZ46JD7nqf5/DLr4dihCCCGEaJ+Uruu6rwfhDXl5eV6938+nlXXnUbSHp6FuuwfDiJatAtzWtYYp+vZA4ug9EkvvkDh6T2uIZUVFRb2aYW2RN5Yyve10cb2gpUzxo5AwUEpKZgghhBCi2Uhi1kTKZILgEEnMhBBCiGZwww03NCgYu3jxYh555JEz9jlZkuvIkSPceefpTyu44YYb+Pbbb8/624sXL6aystLz+bbbbqOkpKSJI/cuSczORZgNXc7LFEIIIbxu0qRJrF69ut53q1evbnAM5Ok4HA4WL1583r/9yiuv1EvMXn/9dUJDQ8/7fhdCErNzEWaXGTMhhBCiGaSmpvLJJ59QU1MDQE5ODgUFBfTt25cpU6Ywbtw4xo4dy4cfftigb05OjudlwcrKSn73u98xcuRIpk+fTlVVlafdI488woQJExg9ejTPPfccAEuWLKGgoIDJkydzww03AHVlvE4WuH/ppZcYM2YMY8aM8SR/OTk5jBw5koceeojRo0dz00031UvsLkTbryTXglRoOPqhvb4ehhBCCNGsXtlUwP5jVY03PAdx4QH8Jin6jNfDw8NJTExk/fr1jBs3jtWrVzNx4kQCAgJYsmQJVqsVp9PJxIkTueqqq1BKnfY+r732GoGBgXz66afs3LmT8eN/emHv4YcfJjw8nNraWm688UZ27tzJ9OnTefnll3n77bcb1E7dtm0bb731Fu+99x66rnP11VczbNgwQkND2b9/Py+++CLz5s3j7rvv5v333+f666+/4DjJjNm5CLNB6XH0VnICvRBCCNGenLqceXIZU9d15s6dS0pKCjfeeCNHjhyhqKjojPf46quvuO666wDo3bs3vXr18lxbs2YN48aNY9y4cXz//fcNjpj8ua+//prx48djsVgICgpiwoQJfPXVV0BdSbC+ffsC0L9/f3Jyci7o2U+SGbNzEWoDXYey43XLmkIIIUQ7dLaZreY0btw4nnrqKbZv305lZSX9+/dn+fLlFBcXs27dOsxmM0OGDKG6uvqc733o0CFeeukl1q5dS1hYGDNnzqy3zHmu/P39PX82Go0XdK9TyYzZOfAcxST7zIQQQgivCwoKIjk5mQceeMCz6b+srIyIiAjMZjMZGRnk5uae9R5DhgzxHFC+e/duz4HmZWVlBAYGEhISQlFREevXr/f0CQ4Opry8/LT3+vDDD6msrKSiooIPPviAIUOGeOtxT0sSs3MRKomZEEII0ZwmTZrEzp07PYnZddddx7fffsvYsWN555136Nat21n733777Zw4cYKRI0fy3HPP0b9/fwD69OlD3759GTFiBPfeey+DBg3y9Lnlllu45ZZbPJv/T+rXrx+TJ08mNTWVq6++mptuusmzfNlcpPL/GZyuCrN+rBjtj1NRt96DYaRU/2+q1lDRuj2QOHqPxNI7JI7e0xpiKZX/m4dU/m9OUv1fCCGEEM1IErNzoIzGuuSsRBIzIYQQQnifJGbnKjQcXWbMhBBCtDPtZGdTq3OucZXE7FyF2mTGTAghRLtjMBha3f6sts7tdmMwnFuqJXXMzpEKs6EfzPb1MIQQQgivCggIoKqqiurq6jNW1W/t/P39z6vGWXPQdR2DwUBAQMA59ZPErAkyDpXy3u5jPJ3SGUOoDcpK0N1ulEnCJ4QQon1QShEYGOjrYVyQ1vB264WSpcwmqHRp7Cyq5OgJV92xTLoOpcd9PSwhhBBCtDOSmDVBdLAZgCPlrp+q/5cc8+GIhBBCCNEeSWLWBNFBfgAUnpwxAygp9uGIhBBCCNEeSWLWBHaLCaOCI2U1EBoOICUzhBBCCOF1kpg1gdGgiAwyU3DCBdYwUAap/i+EEEIIr5PErIkcwWYKyl2nVP+XPWZCCCGE8K4m1XvYunUrS5cuRdM0xo4d6znx/SSXy8WiRYvYt28fVquVmTNnEhUVBcDKlStJT0/HYDAwdepUEhMTycvLY8GCBZ7+hYWFTJkyhdTUVA4cOMDixYupqanBaDTym9/8ptGT5FtCdLAfX+aU1X0Is8lSphBCCCG8rtEZM03TWLJkCY899hgLFiwgIyOD3Nzcem3S09MJCgpi4cKFpKamsmzZMgByc3PJzMxk/vz5PP744yxZsgRN04iJiWHevHnMmzePZ555Bj8/PwYPHgzAG2+8wQ033MC8efOYMmUKb7zxRjM89rmLDjZTWl1Lhau2bp+ZJGZCCCGE8LJGE7Ps7GwcDgfR0dGYTCaSk5PJysqq12bTpk2MGjUKgKFDh7Jjxw50XScrK4vk5GTMZjNRUVE4HA6ys+tXzd++fTsOh4PIyEigrsBdZWUlABUVFYSHh3vjOS/YyZIZhSdLZsixTEIIIYTwskaXMp1OJ3a73fPZbrezZ8+eM7YxGo1YLBbKyspwOp0kJCR42tlsNpzO+glNRkYGw4cP93z+9a9/zezZs3n99dfRNI2//vWvpx1XWloaaWlpAMydO5eIiIjGHuWcmEymevfsUesP5FFhCMQS04kTn5VgDwuT6v9N8PNYivMjcfQeiaV3SBy9R2LpHe0hjj7NKtxuN5s3b+bmm2/2fPfRRx/x61//mqFDh5KZmcm//vUvnnzyyQZ9U1JSSElJ8Xz29hEMPz/Wwd9dC8CevGJ6muvOvTq6bw/KFunV322P2sMRGa2BxNF7JJbeIXH0Homld7SVOMbExJzxWqNLmTabjeLin4qpFhcXY7PZztimtraWiooKrFZrg75Op7Ne3y1bthAXF0dYWJjnu08//ZQhQ4YAMGzYsAZLn75i9TNgMRsoKK9BRdS92EB+7tk7CSGEEEKcg0YTs/j4ePLz8yksLMTtdpOZmUlSUlK9NgMHDmTDhg0AbNy4kT59+qCUIikpiczMTFwuF4WFheTn59d7w/Lny5hQl+Tt3LkTgB07duBwOC70Gb1CKUX0jyUzuKQnGI3o32/z9bCEEEII0Y40upRpNBqZNm0as2fPRtM0Ro8eTWxsLMuXLyc+Pp6kpCTGjBnDokWLmDFjBsHBwcycOROA2NhYhg0bxgMPPIDBYGD69OkYDHW5YFVVFdu2beOuu+6q93t33323pzSH2Wzm7rvvbobHPj/RwWZyS2pQAYEQ1x1993ZfD0kIIYQQ7YjSdV339SC8IS8vz6v3O9069aubC1i35zhv3dgd/d3/oq99G8Pzb6AswV797famraz5t3YSR++RWHqHxNF7JJbe0VbieEF7zMRPooP9qKnVOVZVi+o5AHQNfvjO18MSQgghRDshidk5OFnLrKC8Bi7pAWY/9N2yz0wIIYQQ3iGJ2TlweBIzF8pshm69JDETQgghhNdIYnYOok5JzABUz/5w+CB66XFfDksIIYQQ7YQkZufAz2jAFmjiyKmJGaB/v8OXwxJCCCFEOyGJ2TmKDjZTWF5T96FLNwgIBFnOFEIIIYQXSGJ2jqKDzT/NmBmN0L2v7DMTQgghhFdIYnaOooPNFFe4cdXWlX9TvfpDYR66s8jHIxNCCCFEWyeJ2TlyBPuhA0UnfrbPTE4BEEIIIcQFksTsHEUH/fhm5o+JGTFdIDhE9pkJIYQQ4oJJYnaOoq2nFJkFlMGA6tEP/ftttJPTrYQQQgjhI5KYnSNboAmTQXlqmQHQsz84j0JRvu8GJoQQQog2TxKzc2RQiqggc73ETPVOBEDfluWrYQkhhBCiHZDE7Dw4TimZAaCiOkCnruibM304KiGEEEK0dZKYnYd6RWZ/pAYOh+xd6MeKfTQqIYQQQrR1kpidh+hgM2U1Gidqaj3fqYHDAdC/+dJXwxJCCCFEGyeJ2XnoYPUD4ODxas93qkMn6NgFffMXvhqWEEIIIdo4SczOQ79oCyaD4sucsnrfq8uS65Yzjzt9NDIhhBBCtGWSmJ2HID8jl8UEkXmoDO2U2mUqaTjoOvoWWc4UQgghxLmTxOw8De9s5WiFmx+OVnm+UzGdoUOsvJ0phBBCiPMiidl5GtwpGLNB8cWh0nrfq4HD4Yfv0EuP+WhkQgghhGirJDE7Txbzj8uZB3+2nDkwGXQN/ZuNPhydEEIIIdoiU1Mabd26laVLl6JpGmPHjmXSpEn1rrtcLhYtWsS+ffuwWq3MnDmTqKgoAFauXEl6ejoGg4GpU6eSmJhIXl4eCxYs8PQvLCxkypQppKamArBu3To+/PBDDAYDl112Gbfeequ3nterhne28lVuOd8XVdIrylL3Zccu4OiIvjkDRk3w7QCFEEII0aY0mphpmsaSJUt44oknsNvtPProoyQlJdGpUydPm/T0dIKCgli4cCEZGRksW7aMWbNmkZubS2ZmJvPnz+fYsWM8/fTTvPDCC8TExDBv3jzP/e+++24GDx4MwI4dO9i0aRPz5s3DbDZTUlLSTI9+4QZ5ljPLPImZUgp12XD0de+gl5WgrKE+HqUQQggh2opGlzKzs7NxOBxER0djMplITk4mK6v+mZCbNm1i1KhRAAwdOpQdO3ag6zpZWVkkJydjNpuJiorC4XCQnZ1dr+/27dtxOBxERkYC8NFHH3HNNddgNpsBCA1tvYmNxWxkYMcgMk77dqZWN2smhBBCCNFEjc6YOZ1O7Ha757PdbmfPnj1nbGM0GrFYLJSVleF0OklISPC0s9lsOJ31a3xlZGQwfPhwz+f8/Hx2797Nm2++idls5rbbbqNbt24NxpWWlkZaWhoAc+fOJSIioinP22Qmk6lJ95zQR2fjB9+TV+NHYse6JFK323HGJcAXH2O7/jaUUl4dW1vT1FiKs5M4eo/E0jskjt4jsfSO9hDHJu0xay5ut5vNmzdz8803e77TNI3y8nJmz57N3r17WbBgAYsWLWqQ3KSkpJCSkuL5fPToUa+OLSIiokn37BGi42dUvL8th07+Px1srl0xDv21RRz98jNU9z5eHVtb09RYirOTOHqPxNI7JI7eI7H0jrYSx5iYmDNea3Qp02azUVz808HcxcXF2Gy2M7apra2loqICq9XaoK/T6azXd8uWLcTFxREWFlbvXoMHD0YpRbdu3TAYDJSV1a+w35oEmg0MjAkm81AZtdopy5mDR4IlCH3D+z4cnRBCCCHakkYTs/j4ePLz8yksLMTtdpOZmUlSUlK9NgMHDmTDhg0AbNy4kT59+qCUIikpiczMTFwuF4WFheTn59dblvz5MibAoEGD+O677wDIy8vD7XZjtVov9Dmb1eVdrByrqmXW+wd45vPDvL61iPWHq3APuxL9m0w5okkIIYQQTdLoUqbRaGTatGnMnj0bTdMYPXo0sbGxLF++nPj4eJKSkhgzZgyLFi1ixowZBAcHM3PmTABiY2MZNmwYDzzwAAaDgenTp2Mw1OWCVVVVbNu2jbvuuqve740ZM4Z//OMf/OEPf8BkMnHvvfe2+j1aQzoFc31vGweOV3PgWBVf5ZRRq8Pe2FFMq12F/tmHqF/e5OthCiGEEKKVU7p+yuuEbVheXp5X73ch69S1ms7CjflkHCrjn4UrCT+0G8PcV1Amn27p85m2subf2kkcvUdi6R0SR++RWHpHW4njBe0xE+fOaFDc2C8Ct6azpscvoMQJW+UkACGEEEKcnSRmzaSD1Y8ruoTwwXF/SqO6oK2XlwCEEEIIcXaSmDWjyX3tVLt11gycAj/sQD980NdDEkIIIUQrJolZM4oN9Se5s5X3XVGUB4Sgf7DC10MSQgghRCsmiVkzm9zXTqVb5/3L70DfuB59/57GOwkhhBDioiSJWTOLCw9gcKdg1qhYKsKi0JYvpp28CCuEEEIIL5PErAVM6WvnhEvn41HTYe9u9E1f+HpIQgghhGiFJDFrAQn2QLqG+bPVPwZi49Df+Td6TbWvhyWEEEKIVkYSsxbSOyqQ74ur0KfcCc4i9I9W+XpIQgghhGhlJDFrIb0iLVS5dQ5EdoPLktHXvYN+vLjxjkIIIYS4aEhi1kJ6RwUCsKuoAsMNd4BWi77yDd8OSgghhBCtiiRmLSTCYiYqyMSuokpUpAM1KhX9y/XoR3J9PTQhhBBCtBKSmLWgXpEWdhZVous6asL1YDajr1nu62EJIYQQopWQxKwF9YoM5Film4JyFyokDDXmavSsz9APH/L10IQQQgjRCkhi1oJ6R1kA2FlUCYAady34BaCv+X++HJYQQgghWglJzFpQbKgfQX4GdhVVAKCCQ1ApE9E3Z6Dn7Pfx6IQQQgjha5KYtSCDUvSKCGRnYaXnO3XlJAgMQnv3vz4cmRBCCCFaA0nMWlivKAu5pTWUVrkBUEHBqKuuga1foR/M9vHohBBCCOFLkpi1sN6RP9YzO3rKrNnYX0KQFe1//5EDzoUQQoiLmCRmLaybPQCTQbHr1OXMQAtq0i2w61v0T9b4cHRCCCGE8CVJzFqYn9FAN1uA583Mk9TICTBgMPr//o1+aK+PRieEEEIIX5LEzAd6RwWy11lJtVvzfKeUwnDH/RAcivbyc+hVlWe5gxBCCCHaI1NTGm3dupWlS5eiaRpjx45l0qRJ9a67XC4WLVrEvn37sFqtzJw5k6ioKABWrlxJeno6BoOBqVOnkpiYSF5eHgsWLPD0LywsZMqUKaSmpnq+W7NmDa+//jqvvPIKISEh3njWVqNXZCArdkJ2cRV9oi2e71VwCIbps9DmP4n+5mLUHff7cJRCCCGEaGmNzphpmsaSJUt47LHHWLBgARkZGeTm1j/fMT09naCgIBYuXEhqairLli0DIDc3l8zMTObPn8/jjz/OkiVL0DSNmJgY5s2bx7x583jmmWfw8/Nj8ODBnvsdPXqUbdu2ERER4eXHbR16Rp4sNFvR4Jrq2R81YTJ6Rhpa1uctPTQhhBBC+FCjiVl2djYOh4Po6GhMJhPJyclkZWXVa7Np0yZGjRoFwNChQ9mxYwe6rpOVlUVycjJms5moqCgcDgfZ2fVLQmzfvh2Hw0FkZKTnu//85z/ccsstKKW88IitT4i/ka5h/nyyr4QKV22D62riryC+J/rrL6IXF/pghEIIIYTwhUaXMp1OJ3a73fPZbrezZ8+eM7YxGo1YLBbKyspwOp0kJCR42tlsNpxOZ72+GRkZDB8+3PM5KysLm81G165dzzqutLQ00tLSAJg7d67XZ9dMJlOzztg9ONbM/Su2s/Tb4/xpXPcGSWjtQ3+leNbtGF9bSPhfFqGMxmYbS3Nr7lheLCSO3iOx9A6Jo/dILL2jPcSxSXvMmovb7Wbz5s3cfPPNAFRXV7Ny5aYxVIsAACAASURBVEqeeOKJRvumpKSQkpLi+Xz06FGvji0iIsLr9zxVbAD8ql8E/91WRPcwI1d2C6vfwOgHv7oL19LnKVr2MoZfTG62sTS35o7lxULi6D0SS++QOHqPxNI72kocY2Jiznit0aVMm81GcXGx53NxcTE2m+2MbWpra6moqMBqtTbo63Q66/XdsmULcXFxhIXVJSUFBQUUFhby0EMPce+991JcXMzDDz/M8ePHm/iobcsNfewMcFh4eVMBB45VNbiuho1GJV2O/u5/0Q/sOc0dhBBCCNGeNJqYxcfHk5+fT2FhIW63m8zMTJKSkuq1GThwIBs2bABg48aN9OnTB6UUSUlJZGZm4nK5KCwsJD8/n27dunn6/XwZs3Pnzrzyyiu8+OKLvPjii9jtdp555hlP4tbeGA2KB5JjCDIbmPdFHpUurd51pRTq1nsgJBztlfno1Q2TNyGEEEK0H40mZkajkWnTpjF79mxmzZrFsGHDiI2NZfny5WzatAmAMWPGUF5ezowZM3jvvfe45ZZbAIiNjWXYsGE88MADzJ49m+nTp2Mw1P1kVVUV27ZtY8iQIc34eK1fWKCJB4bHcLi0hpc3HWlwXQUFY5j6eyjMQ39riQ9GKIQQQoiWovR2cjhjXl6eV+/X0uvUy74t4q0dxTwyoiPDYq0NrmvvLEX/cCXqqkmo6+9AGdpObeC2subf2kkcvUdi6R0SR++RWHpHW4njBe0xEy3jxn4RxNv8+edXRzhe5W5wXV13O2p0KvpHq9Bf+Tu6y+WDUQohhBCiOUli1kqYDIqZw2KocGn846sj/HwiUxmMqJvuQl3/a/Ssz9FeeAq9otxHoxVCCCFEc5DErBXpHObPLQMi+Cq3nPX7SxtcV0phGH89avosyN6J9uyj6Medp7mTEEIIIdoiScxamV/2tNE7MpDFmwooOnH65UrD0NEY7v8THC1Am/coenFRC49SCCGEEM1BErNWxmhQ/H5YBzRd52+f5fL5gdIGZTQAVO9LMcz6C5SVoj37CHqhd19+EEIIIUTLk8SsFXJY/ZgxtAPFFW6ey8jj9v/tYc6nuXydW1avnYrvieEPf4WaKrRnH0PPO+SjEQshhBDCGyQxa6Uu7xLCq9d2Y86VnbmqWxjZxVXM/vQwe531i8yqLvEYHvwboKPNewx99zbfDFgIIYQQF0wSs1bMaFD0ibJwZ1I0C6+OI9BkYNXOhpv9VcfOGB76GwRZ0f7+BNryJeiuGh+MWAghhBAXQhKzNiLIz8i4hDC+OFRKQXnDpEtFx2B48nnU6F+gp61G++sD6If2+mCkQgghhDhfkpi1IRN7hqOAd3cfO+115e+P4ebfYvj9U3CiHG3OQ2gfrWpQE00IIYQQrZMkZm1IhMXMyLgQPs4+Tml17Rnbqb6XYfjzQuifhP72q2j/motecaIFRyqEEEKI8yGJWRszqZed6lqdD344/azZSSrIiuF3j6ImT4OtX6HNfgA9d38LjVIIIYQQ50MSszamS5g/A2OCeO/7Y1S7G9Y3O5VSCsNVkzA8OAeqq+uWNte+hX6i7Kz9hBBCCOEbkpi1Qdf2tlFSXcv6/SVNaq8SemP40wLoNQB91Rtof5yKtuyf6EcON/NIhRBCCHEuTL4egDh3faMsJNgDWLXLyWUdgokKNjfaR4WEY5zxJHruAfS0d9G/+Bh9wzpIHIph4o2ozvEtMHIhhBBCnI0kZm2QUoopfe3M/vQwd67eS2yoHwNjghkYE0SPiED8TWeeCFWduqLuuB/9utvQ169DT1+D9vRGGDAYw8SbUF0kQRNCCCF8RentpJZCXp53z4qMiIjg6NGjXr2nt+WWVrP58Ak25ZWzs7ACtwYGVbcPrbs9kO4RAQzqGExowJnzb73iBHr6GvSPV0PFiboE7drbUB27eG2cbSGWbYHE0Xsklt4hcfQeiaV3tJU4xsTEnPGazJi1YZ1C/OkU4s81vWxUuGr5rqCS749W8kNxJV8cLOXD7OOYDDA01sq4bmH0jbZgUKrePZQlCHX1r9DHTERPfw/9o1Vof74fNXQU6pc3oyKiffR0QgghxMVHErN2wmI2MqhTMIM6BQOg6ToHj1fzyd4S1u8v4YuDZXSwmpl6WRRDOlkb9K9L0G5EHzUB/YP/oaevRc/6HHX5lajLr4TO8aifJXVCCCGE8C5JzNopg1LEhQfwm6QAbkuM5MucMlbtcjL3s8PcO8RBSnzYafup4BDUDVPrZtDee/OnlwRiOqOSx6CGjEKF2Vr4aYQQQoiLg5TLuAj4mwyMigtl7lVd6O8IYuHGI6ze1fAw9FMpWwSG2+/D8NxrqFvvgUAL+jv/RnvkN2hvLkYva1qpDiGEEEI0XZNmzLZu3crSpUvRNI2xY8cyadKketddLheLFi1i3759WK1WZs6cSVRUFAArV64kPT0dg8HA1KlTSUxMJC8vjwULFnj6FxYWMmXKFFJTU3n99dfZvHkzJpOJ6Oho7rnnHoKCgrz4yBevAJOBJ0Z2ZEFmPq9+U0hZdS23DIg46xKlCgpGjRwPI8ejHzmM/vEq9PVr0TPSUOOvR6X8EuUf0IJPIYQQQrRfxqeeeuqpszXQNI05c+bw+OOPc+2117J06VJ69+5NSEiIp01aWhoVFRU8+eSTBAQE8MEHHzBs2DByc3N55513ePbZZxk0aBDPP/8848ePJyQkhKuuuoqrrrqKlJQU1q1bx+233+5JwG677TbGjx/P/v372b17N/3792/0QcrKvFvN3mKxUFFR4dV7tgZGg2JorBVnpZs13x9jZ1Elrlodm8VEoPnsE6gqOAQ1YDAq6XL0ogLY8D56Rhq4XNAhFuXvf9p+7TWWLU3i6D0SS++QOHqPxNI72kocrdaGe71PanQpMzs7G4fDQXR0NCaTieTkZLKysuq12bRpE6NGjQJg6NCh7NixA13XycrKIjk5GbPZTFRUFA6Hg+zs7Hp9t2/fjsPhIDIyEoABAwZgNBoB6N69O07n2ZfcxLkzGhT3DnFwe2IkR8pc/OPrI0xdkc2DHxzgv9uK2JJ/ggrXWQ5J79AJ472PYXh4LnTsUneawMPT0P6zED13P+2kAosQQgjR4hpdynQ6ndjtds9nu93Onj17ztjGaDRisVgoKyvD6XSSkJDgaWez2RokWhkZGQwfPvy0v52enk5ycnLTn0Y0mVKK6/vYua63jYPHq/n6cDlf55bz1vZidIpR1NVDG9QxmBv7RWA2NlzuVN16Y5z1F/S8Q+ifvIe+MR39i4/BGgpduqG6dkN16YZ+xdgWfz4hhBCiLfLpW5lut5vNmzdz8803N7i2YsUKjEYjV1xxxWn7pqWlkZaWBsDcuXOJiIjw6thMJpPX79laRUZCUgLcA5yodrOzoJzt+aV8e7iUt78rZs9xF7NTexEWeIajnyIioP9laGUzqcpIw7VnJ+7s3bjffxtd0yh6bSGWcdcSOOF6jLaLI6bN4WL6O9ncJJbeIXH0Homld7SHODaamNlsNoqLiz2fi4uLsdlsp21jt9upra2loqICq9XaoK/T6azXd8uWLcTFxREWVr90w4YNG9i8eTN/+tOfzrgxPSUlhZSUFM9nb1f6bSvVg5tDnAXi4i38Mt7Cp/sDWfTVEaYu+4bHR3aka3gjG/2TRtT9BzBUV8PeXZgz0zjxv9c4sXIZatAVqMuGQWwc2KOkNto5uJj/TnqbxNI7JI7eI7H0jrYSx7NV/m90j1l8fDz5+fkUFhbidrvJzMwkKSmpXpuBAweyYcMGADZu3EifPn1QSpGUlERmZiYul4vCwkLy8/Pp1q2bp9/pljG3bt3K6tWrefjhh/E/w2Zy0XJGxoUy58rOuDSdhz86xJc5ZU3eQ6b8/VG9Ewl7ZC6Gv/4LNXI8+pYv0f4xB+3RO9F+fzO18x5D+3g1uqummZ9ECCGEaP2adFbmN998w3/+8x80TWP06NFcd911LF++nPj4eJKSkqipqWHRokXs37+f4OBgZs6cSXR03VE+K1asYP369RgMBu644w4uvfRSAKqqqrjnnntYtGgRFovF81szZszA7XYTHFxXwT4hIYG77rqr0Qe5GM/KbEnFFS7+9tlh9hRXkWAP4Jc9bSR3tmIy1M141Wo6h0qqOXS8mmGdrfgZf8r5T42lXl0Nhw+g5+yHnH3oB7LhYDbYIlGTbkUNGYEyGH3yjK2d/J30Homld0gcvUdi6R1tJY5nmzGTQ8zPoK38y21JNbUaaXtLWLP7GHllNdgDTSR3sZJXWsPuokpOuDQARseF8PthHTzLlI3FUt/1Ldo7/4ZDe6FTVwypU6DvQFRAYEs8Vpshfye9R2LpHRJH75FYekdbiaMcYi68ws9o4BfdwxmfEMY3eSd4d7eT93Yfo1OoH5d3CaF3VCAHj1ezYqeT7hGB/KJ7eJPuq3oNwPD439E3fVFXeuOlZ8Fkgp4DUAMGofoPQtkim/nphBBCCN+TxEycM4NSJHUMJqljMG5N9yxnQt3h6YeOV7NkcwFx4f70irTU61vp0iipclPl1qh0a1S5dYLMBjqF+mEZPAJ94HDI3om+9Wv0b79CX7YZfdm/wNER1SsR1TsRuvdFWeQ0CCGEEO2PJGbigpyalEFd0jYrOYY/fHCAZz7PY8GErkQAheUuVu128nH2cWpqT796bgs00SnEj8u7xDLuxn7oU6bBkVz07ZvRd21Fz0hDX7+2rnFIGEQ6UFEdoENn1OUpKGtoMz+tEEII0bwkMRNeF+xv5NERHXnow4PM/ewwXSJK+Pj7IgwKRnQNpV+0hUCTgQCzgQCjorS6lpzSGnJLqtnrrOIfXx9BKbiqW1jdUU8dYuGqSeguF+zbjb7veyg6gl6Yj757O3y5Hn3tW6ixV6OumoQKOvNRF0IIIURrJomZaBZdwwO4b4iD+Zn5HCypZmKPcH7Zy0aE5fRFaof8+E+3pvPXDbn88+sj2ANNDOwY7GmjzGbo0Q/Vo1+9vnp+DvqaN9Hffxt9/dq6g9VHjEeF1a+3J4QQQrR28lbmGbSVNztau++PVtKvq4Oa8pIm96lw1fLYx4fIL6thzpVdiLc1UtT2R3ruAY6sWcnrVR24Zf9HOOI7o4aORiUOPeMB622J/J30Homld0gcvUdi6R1tJY4XVGBWiAvRIyKQkIAzHOV0BhazkSdHdcLqZ+Tp9TkUlDet+KyrQ2fmdb2WjKhEXk/+DeTloL/yd7QHb0f7cCW6pp3PIwghhBAtRhIz0SrZLWb+NDqWGk3noQ8P8uJX+WzMKaPCVXvGPq9sKmSvs4pEh4Uva23sfXAhhgdnQ/e+6O8sRZv/JHpxUQs+hRBCCHFuJDETrVbnMH/+PCaW3pEWPj9Qxt8+O8xt7+zhz+k57CmurNc2fV8JH2Yf57reNh4e0ZFQfyOvbzuK6tEPw31PoO64Hw5ko/35frSNG5p8rJQQQgjRkmTzv2jVEuyBPDKiI25NZ3dRJZvzyvlkXwkPfnCQUXEh3JYYSXl1Lf/8+gh9oy3cOiASo0Exua+dVzYXsjX/BIkdglDDU9C790V7dQH6kvnon6xBDRlZd6h6aNMK4QohhBDNTRIz0SaYDIq+0Rb6RluY3NfO/75zsnqXk8xDZQT5GQnyM/Lg8BiMP9ZVG58Qxru7nby2tYgBDgtKKVSkA8NDc9A3rKuribb8FfS3XoWe/VCXX4kaOBxllHM6hRBC+I4sZYo2x2I2cltiJC9OjGNIp2AqXbU8dHkM4YE//f8ZZqOBm/pHstdZRWZOmed7ZTBiGHM1xiefx/CXF1Gpk+FoAfri59Ce+C3a+rV1B60LIYQQPiAzZqLNig7248HLO1Kr6Z6ZslON7BrCip3FLPv2KEM7WRu0UR1iUdfcgj7xJvj2a7QP/of+35fQ3/1/qBHjUEmXQ6eunsPYhRBCiOYmiZlo806XlJ38/rYBkcz57DBvfFvE7YmRp02yvi+uJrTbZTgeGQJ7vkP7YAX6uv+hv/82RMWgBiajBiZD53hJ0oQQQjQrScxEuza4UzBXdQtlxU4nrlqdaQOjMPyYXGm6zlvbi/l/249it5h44RdxWLv3xdi9L3pZCfqWL9E3Z6J/uAJ93Ttgi0AlDmVT3DA+rgzjwSs64m86v90AmYdKibcFEB3s583HFUII0cZJYibaNaUU9wx2EGAy8O7uY1S6Ne4Z7KC6VmNBZj5f55YzpFMwmw6X84+vj/DHy2PqXhSwhqJGjIcR49HLS9G/zULfupHjX37B/1UPpMx8gnWvr+KXfaNQfS9DBVqaPKYt+Sd45vM8OoX4MX9C1/NO7oQQQrQ/kpiJdk8pxbTLogg0G1i+vZiy6loOl9aQV1bDnUlRpHYPZ8XOujc4P9lXQkp8WP3+wSGo4WNh+Fj+/XkOlYfK6aqVslJ15qpX/oa/0qFHX1RCH1S3XhDXHeV/+mOkqt0a//z6COEBRnJLa3j92yJ+MzC6JcIghBCiDZDETFwUlFLc3D+SQJOBf28pItTfyNNjO9M3um6m69reNrbkn2DxpgJ6R1qICWm4xPhNXjmfHjrBjf0iSHR04dGPD/HhLU9zTcFX6Ds2o69ehg5gNNbtR+udiOqXBHEJKENdGY7/t+0oBeUuZqd0JvNQKWt2H2Nwx2D6O4JaMBpCCCFaK0nMxEXl2t52utsDcVjN2C0/neFpUIqZyR34/dr9zM/MY+5VXTCd8lJBpatupqtTiB+T+9gxGw0kOiysLKxmwrW/JmDyVPQT5bBvN/qeneg/7EB//x30tW9BcAiq70D2dRnA6oIYrrzESt9oCwn2ALbkV/B/X+bzQmocQX5SQ00IIS52xqeeeuopXw/CG8rKyhpvdA4sFgsVFRVevefFqrXFMirYjMXcMAmymI10sJp5d/cx8stqCAswYQs0YVCK17YW8k1+BY+O6IjDWjeb1sHqx3s/HCPIbKBXlAXl54eKjkH1GoDh8itRY1KhUxxK13F/t4W/GQagKcUjH/4F85ZMjAe+J8GisaY8FKezhCGdQ1GGM+83a21xbMsklt4hcfQeiaV3tJU4Wq3WM16TGTMhTpHcOYTrelexepeTzw+WYfUz0M8RxMacMsYnhNE76qdN/j0jA7m0QxArdzmZ0D2cQLOB8upa3v6umI+yj9PBaqZPVE/6jruMg0Oq2bvtKA92KMVqvRL9QDb6ji0klHzCdV2v4h1SGPTnvzAk0oTq2R/VawB07HLWRE0IIUT7I4mZED/z60ujuKGPna35J8g6XM7mvBNEWEzcnhjZoO1N/SP444cHeXe3E4vZwPLtRymv0Rgaa6Ws2s26H47z7u5jACTFBHH5qB4oNdjTX6+q4Ff5h/lmcxUv97yefj+8iuXtV+v2qgVZIb4nqlsvVHwvdOvQFoqAaG02HS4HIKljsI9HIoRobk1KzLZu3crSpUvRNI2xY8cyadKketddLheLFi1i3759WK1WZs6cSVRUFAArV64kPT0dg8HA1KlTSUxMJC8vjwULFnj6FxYWMmXKFFJTUykvL2fBggUUFRURGRnJrFmzCA6W/zESLSvIz8jwLiEM7xJCraaj6WA2Niwu2yMikIExQfx321EA+jssTL00iktsdW9l1tRq7CmuIru4ipFdQxoUqFUBFsxxCdwTUslDHxzkzYmP8ZtLjOi7t8Ge79Czd6Jvy0IHCpWCcDtEdkBFx0CkAxXZAaI61P05ILDZ4yJ84+VNBfgblSRmQlwEGk3MNE1jyZIlPPHEE9jtdh599FGSkpLo1KmTp016ejpBQUEsXLiQjIwMli1bxqxZs8jNzSUzM5P58+dz7Ngxnn76aV544QViYmKYN2+e5/533303gwfXzSKsWrWKfv36MWnSJFatWsWqVau49dZbm+nxhWic0aA427b8Oy6LQlHIhO7hDIwJqpd8+RkN9Imy0Cfq7HXOEuyBTOgexvs/HGN0XFe6JY+B5DEA6GWlsHcXgcVHqDiwF70wH/2bL6G8tG5m7aRQW90MW/e+qO59ZCm0nThSVkNBuQuTAdyaXu+lFCFE+9NoYpadnY3D4SA6uq7WUnJyMllZWfUSs02bNjF58mQAhg4dyquvvoqu62RlZZGcnIzZbCYqKgqHw0F2djbdu3f39N2+fTsOh4PIyLploqysLE6+jzBy5EieeuopScxEq9Y51J8nR8de8H1uGRBJ5qEy/vn1EZ4d18Vz1JSyhkDiEIIjIqg6etTTXq8oh6Ij6IVHoCgf8g6hZ+9C/yazLmELDIIu8ajOl9SV7+h8Cdgiz1hjTbRO3x6p28js1iCvrIbOof4+HpEQojk1mpg5nU7sdrvns91uZ8+ePWdsYzQasVgslJWV4XQ6SUhI8LSz2Ww4nc56fTMyMhg+fLjnc0lJCeHh4QCEhYVRUlJy2nGlpaWRlpYGwNy5c4mIiGjsUc6JyWTy+j0vVhLLpokAZo4y8NQH35OR7+a6AR3qXW8Yxwjo3LXBfWoL86nZ+S2uXdtw7fse9/r3wVXjmV1TARYMYeEYwuwYO3TC1OUSTJ0vwdQlHkN4xEVxHmhb+ju5++ujGA2KWk3nuObPZa1o3G0pjq2dxNI72kMcfbr53+12s3nzZm6++ebTXldKnfH/SKSkpJCSkuL5fPSUmQRviIiI8Po9L1YSy6ZLtMEAh4V/Zuynn00RHvjTf0VPxrHarVFeU4vZaCDE/zSLrAYz9E2q+w9gcLvhSC56zn44Xgylx9FKjlFbehzXNxth/fs/9Q201N+71iEW1SUeHB09RXLbg7byd1LTdbIOOhnaKZjMQ2V8l1NE/3Bfj+onbSWObYHE0jvaShxjYmLOeK3RxMxms1FcXOz5XFxcjM1mO20bu91ObW0tFRUVWK3WBn2dTme9vlu2bCEuLo6wsJ+OwAkNDeXYsWOEh4dz7NgxQkJCmvaUQrQDSinuHuTg/rX7mbXuAEFmA7V63csHbn0fZVUuamrr5r6MCib2tHFjP/tp67J57mkyQaeuqE5dT3tdLyuFvIPouQehMK9uD1vOfti6EWpr62ba/PwhNg7VsUvdXrawcFSoDcJsEB4B1tCLYqatpe0/Vk1ZjcbgTsHsO1bFoZIaXw9JCNHMGk3M4uPjyc/Pp7CwEJvNRmZmJvfff3+9NgMHDmTDhg10796djRs30qdPH5RSJCUl8X//939cffXVHDt2jPz8fLp16+bp9/NlTICkpCQ+/fRTJk2axKeffsqgQYO89KhCtA0dQ/x4ILkDnx8sw6DAqBQGBcFBgZg1F8F+RoL8DGQ76+qtfbq/hF9fGsWouIZvfQLouk5OSQ0/FFcSG+pPgj0AwyntlDUEevRD9ehXv19tLRw5jH5oLxzMRj+41/PSAVD/xQOzX12CZo9ERURDRHTdzFtUh7oiu8b2M9vWkr7NPwHAAEcQmYfKOHS82scjEkI0t0YTM6PRyLRp05g9ezaapjF69GhiY2NZvnw58fHxJCUlMWbMGBYtWsSMGTMIDg5m5syZAMTGxjJs2DAeeOABDAYD06dPx/DjW2JVVVVs27aNu+66q97vTZo0iQULFpCenu4plyHExeZkqY5TnW6KfnxCGC9lFfD8l/ms+d5Jl7AAQv2NhPgb8TMpvi+qYlvBCY5X1Xr6hAUYSeoYzKCOwQxwBBFoPv2bm8pohI6dUR07w7DRnu91twtKj0PJMTjuRHceBWcROIvQiwvRt34FZXV7Qz0vIfQegOo7ENX7UrCGgH5KWmf2k9m2M/j2yAm6hPoTHmgiNtSfrMPluGo1zEZ521aI9krpuq433qz1y8vL8+r92so6dVsgsfSOM8VR03XS95Ww7ofjHKtyU1pVi0ur+691eICR/o4g+jssdI8IZL+zylM0t8KlYTIoekcFclmHIC6LCaZzqHeSJL2qEooL0fNyYOcW9B2b4bjz9I39/MEeRbUtmmcjxtApyMS0ATZU5ziUyXz6PheoLfydrKnVuOXtPYxLCOM3A6P5dH8J8zPzeeEXXeka3jrerG0LcWwrJJbe0VbieEF7zIQQrZtBKVLiw0iJr9urqes6VW6dSrdGeICxXqLVOdSfkXGhuGp1dhZV8E3eCbbkneDfW4r495YiAkwKR7AfHaxmHMF+RAebiQ42ExlkJirIjL+paTM1KiCwro5axy4w6HJ0XYfDB9G/3w41J5fjFOgalJVQW1zIC8ZEtqgItlRA6BvvcV1+Rl25j9g4sEWBLQJlj4QwOwRbwT+wXc+07S6qpKZWZ0B0EACdw+rKZBwqqWk1iZkQwvskMROinVFKEWhWZ1yihLpTDAY4ghjgCGLqZXC0wsWWvBMcPF7NkfIackpqyDp8ArdWf0Ld36jQ9LpZOk2H0AAjv+oXwVXdwjx11840prO9gLBsaxEbvyvmjksj2XukhDf4BR16xDNsfwZ61hdwogz42b42oxEswRAShnJ0gphYVExncHQCWyQEWtp04vbtkQoMCvpE153o0DHED4OCnBLZZyZEeyaJmRCCCIuZK7uF1fuuVtM5VuWmqNxFwQkXRSdclNdoGFTdLJ0CdhZV8K+sAtb9cJzpSVEMcARRq/3/9u48OsryXuD49519n2QmmewBsrBXQIKEKAJCXVptraWe2ttW8Hqt4lK1WvFUrV2svfaiuF68rUXFLtpWULBWiywuEWUXwiJgCAnZSGaSTDKZmczMc/8YDMYEgZqaRH+fc+bAzPvOm2d+eWbml2dVVAYivNfQwaGWCGcPdzEpy37cJOm1Ay38taKZ84pSuHiMh65RqTSurubBwGh8V57LyDQrKhyCo2PZVIsfOtqTyVpHO6qlOTlBYUs5PUZmmK3JLaw8acmWu/wCtLzk0h9Dwfb6DkalWbtn3Jr0OrKcJg5JYibE55okZkKIPul1Gmk2I2k2I2OOc45SirergyzdcoS7Xqum2GuhLhilPZoAwGbUsbayjXE+K9+bmM6Y9J5bU+1sCPHYu/VMyLRx1ZQM/epsEQAAIABJREFUNE3DpNf4yYwcbn2linvW1/Cb84bjc9ggOx+y8zleG5iKRpKzSBsOQ6AJAs2oQBM0NaLWvXxskV2jiSZvOnGHC1ypaO4UyMxFG1YEuSN4vTaM22JgYpa9v0J5ytojcQ74w1w63tvj8Xy3iaqW3ktmhLriHG6LUuyV/VKFGOokMRNC/Ms0TaMs30VJjoMXdwd461AbpXlOTsuwcVqmHYdJz6v7W3huZxMLXz3E6Vl2HGZ99/6PrZE4uS4TP56e02MPSLfFwB0zc1n4ShXXrPyAIo+FsT4ro9OtjPJacX9s7ByAZjInW8XyC3qVM7n0Rw3q0AdUHWpgePgIHKmHumrUnu0Q6kABK/Jm8HThVwE439TE/EID5tx8SMv4TBfY3dEYIqGSy2R8VJ7bzDs17UTjCUwfmZn59NYjvLK/haXfKCLFKh/rQgxl8g4WQnxqJr2OueO9zP1YCw/AV0elMrvQzaq9AVbt8WPU68h0Gpma5yDTYWLmCBcOU++kJ99t5ldfzmddZRu7jnTy4h4/z+9KHrObdOS6TOS4TKRYDAQjcVojcVrDMRIKvjIylRnDXcf2G9XraXBn8duIno2xLEZnO/jhN3xku0zJ7s9AMy9tPcTT9TbKItWktdXxYvoZ7NjcyA+f+SVF0ebkRIQRxWjDiyErH2x2sFr/LZMQttd1YDHoGJnWswUs320moeBwW5QRRycAdMUTvFHVRkLBhpog5xcPoq0BhBCnTBIzIcS/ncWgY+44L3PH9U7cPsnwVAvzjiYgkViC/c1hDgTCHG6Lcrgtyra6EK3hGC6zHrfFgMuipzUc58G36/jLzmYuOy2N0jwHL+4J8OyOJnQaXDQ6lXWVQW56uZKrSjI4p8DNP/0GfltvY2qugx9Nn4NBpzG5ys9D78Ltk6/nCvZzQdV61GsrUbFYz0JquuTabPkFaMOLk12iw4ogxfMvJWzVrRHKq4OM91l7tCLCR2ZmtkS6E7NNhztojyYw6TXKD0liJsRQJ4mZEGJIMBt0jMuwMS6j5zg1pVSPBEgpxYaadv60vYlFb9Vi1mtE4orSPAdXTs4g3W7kirIi7ly1k4c21LOmso2KhhCTs+3celZ2dzI0cZiHhzLdLH67jt8dLibr8lmcnmGGmoOoxjoIh6CzM/lvoAl1cD+qYhsH7JlUpBRwbstOzJnZyZmi2fnJGam5w9DsTuqCUbw2Q4/uSIDdR0L8cl0NBp3G9yam94pBttOEXqPH1kxrK1tJtRqYNcLFit1+2sIxXBb5aBdiqJJ3rxBiSOs11kzTmJbnZGqugzergrxdHeScEW6m5Dq6z/E5zfx8dj7P72rmj+818aUMG7dNz+m1or7DrOfWs7L58StV3F9ey/0XDCdj+NHuzI9QSrGzIcTzO5vY2tAJwBp1Drc2vEL2pje6x7B16C38btylrPeMJ4UoFzrbOX+4FUdWFu92Wvmf8nq8NgN3z8oj02nq9VqNeo1s17GZma3hGJsOt3PRaA/Th7l4fpefd2rae82wFUIMHZKYCSE+l3SaxtnDXZw93NXncb1O41vj0zinwI3bYujVbfghs0HHwrNz+NHLB/nvN2r59bn53S1dSim21Hbwpx1N7GsOk2LR872J6eS4TDz6Tj235n6L6+deT5mri4q9h1hcqac5buSilu1UKxvPUMxft0YoWb2Oct9pFITq+cmhdaTUOUj4siArDy07DzJy0czJbsx8t5kD/jAAb1S1EVcwa4SLYSlmMh1Gyg8F+1z6ZPeRTsb6rD32SRVCDD6SmAkhvtC8thNv+5TlNHFjWRb3rD/M/21s4LrSLPY1d/LU1iPsaAiR4TByzRnJ8WofJm1FHgu/efMw971Zy8RMG9vrrWQ4jNxbls3o9DEopag81MiKXc28YZjABH0btxoqsOgTqKr9sKUcEonkEh+allw0NyObPN80yhlO+L3NrD3kZkSquXsngLJ8Jy/s9tMeieMwH5tQsXyXn2Xbj3Dt1EzOldY0IQY1ScyEEOIknJHrZO44L3+taKY2GKWisROXWc9/lfg4rygVo75nS1S63cg9c4bx1NZGVu4NMKfQzX9O9nUvGKtpGgXDMrh5WAZXReLYTDp02tTu56tYFzTWJZf0qK1OrtHWWEvuvo2o4hGUP/si+8dexvxDLxHf1YiWM4zStCKeV9m888ERZo/JBKAuGOXZncm9A1fs9jOn0C2tZkIMYpKYCSHESfrOaWns94fZ1RjiW+O8XDLO051o9cWo17iyJINvn5bW55IgH/po69aHNIPx2KK6k489PqwlDC8d5M8Tv40uqjg7xwo1EdRbqymMrCKt9HbKX93FzN+tQPmyWZJ9IXpdKvN9EZY2wrs1QUrz+u7eFUIMPEnMhBDiJOl1GnfOzCUaT3xiQvZxn5SUnapslxmDDhqiGiXZdjyzrgJAJRIQaKJsUwN/bxpNyF3K5nYD23RpXLlvBeet38BLZ/yY5X+vZkrkbbQRI8GXhZaaBp40SPGiGeQrQYiBJu9CIYQ4BQadhuEz3AWgr5+f4zRT1RphVoG7+3FNpwOvjzNPd/Liq1W8Pu0ynt3ZRJHNyAW33YC+4Zt8bWcDv2vNY0/zBka/+EfgIxvDaxq4UpNJWqo3mbBl5qDlFxLKyOdv7wfx2AxcOMrz2b9oIb5AJDETQoghZkSqmabOLqbkOHodG5lmwWs18MTmBhRw96w8DFYLDC/my7mFPLt8Py+cOZ+xN98C/iPgb0ruKepvSq7HFmiCuhpUxTZUpJO30iewtOgiAmYXOqUo3rGWYp8DLS2DWEERKhwFm5236yOsPtDK3HFexvpsvQsthDgpkpgJIcQQM+90H5eM82I26Hod02ka0/KdrNob4OujUynwWLqPWQw6LhiZyl92NlM7yUfOJ2wMX90S5rcbatjeHKNA18EPA+t4yFbCY82p3PfqQxhUguaj5x62pvNgyQ1EdCY213ZQag7y/bFuskcWSPeoEKdI3jFCCDHEpFoNpH7CZuUXjkolGk9w2Wm9dw/46qhUVuz288JuPwumZvY41hGNU34oyJoPWtl1pBObUcdVJRmcX5yCXjeZq6qD/Pr1w7x01YN8w9mGI9FFc30j9zdkYIzrWBRayxttJlakT2Hj5i4uWPUk32/agMHlAlcKmssNdidY7cm9Rm32o2Pc0iE1TZI4IZDETAghPneynCaunZrV57EUi4FZI9ys+aAVn91IR1ecYCROoDPGew0honFFjsvE9yakM6fQTcpHEsAPd1T40/4Oyr46ivQR2Tz98k4qa1u4c2YuuTlf4jLgvLpG/rj5MKt0Z9GWWcANgdfRtfpR1R9AqAOike5rHhvjpoNUT3Kcm8OF5nCBwwV2R/JmtaPZHcnH3KnJRM9w4jXohBhqJDETQogvmIvHeFjzQSvLth/BoAOnSY/TrGd2gZtzCtwUey3H3YD9qikZXLuykv/d2MClcRMvvd/C10anUvKR8W6eLB/XXegjY2cTz2wH95nX8J+Tfd3XVLEu6AxBRzA5xq25MTnerbkRFWyFYCuqrhra2yAS7r6u+nhh7E5I8UBaBlpaBqT50FK8YDSD0QAGE5gt4E4BZwqafuAmbQhxsiQxE0KIL5hsl4mn5xahoWExaMdNwvqSZjPy/Ynp/N+mBir+sZdCj4XvT/T1ee7ccV5aI3FW7gngsui5dHwacHSNNqc7ecvM7R7jFk8odFrP/U9VrCvZyhZqh452aG9DtQagLQCtAVSgGZoaUHveg0i4d/L2IU0HrpSjrW3uoy1ybnA4k8mbyQwmE5rJDM6UZMKX4kUzSquc+GxJYiaEEF9Ap7IO28edX5zCuspWaoJd3HpWdq9dDz6kaRpXnO4jGI7zh+1NuMx6zi9O7XFOqCvOuzXtlB8KsqW2A7dFz4RMOxMybZyWaSfVakwmVK5jW0n19dOUUtAehFY/xLqgqwu6oslkrTWQfLzFn/x/sBVVfxiCbbQlNJxdoe5r9krsnO5k96nFmryZrWiOoy11bg9aSmoywTNbwGwGs5WI3ojBZMZg+mySOqUUCo67o0NnV4LKQJgx6dZTSsI/bzYdbuexd+qJJo79lt1mPT+ZkUu2yzSAJevppBKzbdu2sXTpUhKJBLNnz+biiy/ucbyrq4tHHnmEDz74AKfTyY033ojPl/wLavny5axZswadTsf8+fOZOHEiAB0dHSxZsoTq6mo0TeOaa65h5MiRHDx4kN/+9rdEo1H0ej1XXnklRUVF/fyyhRBC/Kv0Oo2fz87H5HCjjwQ/8VydpnH9tCyC0Tj/+24Dv9/ciNtiwG3RY9Zr7GkKE0sovFYDcwrdtEbivFMT5LUPWgFwmnTYTXrsR/8tSLVwQXEKmc6eX6SapoHTlbx9zMdTkYRSbDzczgu7/VQ0djLOa+LqsXbyzPFk12lbS3LZkBZ/cgmRUDuEO5O3thZU5fvQ1gIq0SOR69L0PDZqLuszk1s16FQCYyKGPR5mckcV02K1jDe0Y3Q4wGI7msxZwGylMysbhZZsrXO6wO5EM544WWiPxvnv1w9TF4yyYGomp2f3XELlgD/Mb948TF2wi0vHe/nOaWlfyOQs1BXn0XfqsRh0nJF5bDmXtZVtPL3tCAvPzhnA0vV0wsQskUjwxBNPcMcdd+D1ern99tspKSkhNze3+5w1a9Zgt9t5+OGHeeutt/jDH/7ATTfdRE1NDeXl5dx///0EAgF+8Ytf8OCDD6LT6Vi6dCkTJ07kRz/6EbFYjEgkORj0mWeeYe7cuUyaNIktW7bwzDPPcPfdd//bAiCEEOLUWY060pxmmk6QmEFyUdzbpufw6v4WGju6aAvHaY3EaY/GuWBkCmfmOxmVZu1u8YknFJWBCO/Vd3Ak1EVHNEFHNE4wGmflnuSM0im5Di4clcppGbaTTjRaOmO8XR3kxT0BaoNR0m0Gvj46lTUftHLjmwG+MdbLpeMzMefr+myVg2QipNc0LDoFwRZoDUCwjWBnhHsP2dgVNvNVqx93IkI0ruiKK47oDLyRMoZ/al/CEQ8zrWUv393xCs52PyQSALT1GThj9+xV9IbkuUolby43zfnj+IV+IrVdRtLtRn62toY5hW7mn+7DbtSxam+AJ7cewW3WMy3PyXM7m0ko+O6EoZmc1Qej7GwMcTAQoaolefM5jHx3QjoTs+yf+NxndzTj74xx33nDGJVm7X48xWrgT+81sbeps8fjA+mEidn+/fvJzMwkIyMDgLKyMjZu3NgjMdu0aRPf+ta3ACgtLeX3v/89Sik2btxIWVkZRqMRn89HZmYm+/fvJzc3l927d3PttdcmC2EwYDg6TVrTNDo7OwEIhUKkpvZs9hZCCDH0mA06Lhp9crsG6HUaRV4LRV5Lr2PNoS7+sa+FV/a18G5NO+k2A6PSrRR7LRR7reS4TMQTqjsp8nfG2F7fwda6DioDyQaAIo+FH52ZzZn5TvQ6jW+O8/Lk1iP8taKZN6ramD7MRZbTSLbTRJbTRF0wyta65DX2N4cx6TVmjHBxQXEqBcO81Aej/HxdDY3RLm49K4uzho3uVe5ILMHWug7KDwVZe8jK+wVn8NNzcvEYgXCIVKOewKGqZDdrsDU5pi7UAZ0d0NGOUgnQNDRNB5pGVVsXvwiNJKSPc8fOpYxuO8hzw+awQs1ky65D5IUa2Z5azBT/Hq47+CL2RIT/G/ZV/lpxOrG1f+d7R95KjrNzpaC5UpJdtlYbmK20GO081ZJCqQdKM8zHWvaMJjAYkje9EQyGzyzB21Lbzr2vHyYaV5j1GvkpZkpyHLxX38FP11QzMdPG9yf5SEvr/dxDrRFW7vEzp9DdK/n6+mgPf38/wFNbG7lnTv6gSFhPmJj5/X68Xm/3fa/Xy759+457jl6vx2azEQwG8fv9FBcXd5/n8Xjw+/2YTCZcLhePPfYYVVVVFBQUMG/ePCwWC5dffjn33HMPy5YtI5FI8Mtf/rLPcq1evZrVq1cD8Otf/5q0vn4bn4LBYOj3a35RSSz7h8Sx/0gs+8dAxDENGJWfxdUzErz2/hHKK/3sbmjnzarjt9wZdBqnZbs4b0wmU4elUpxu7/EFnAb8IjeTS2paefTNSpbv9hNP9BxtptNgbKaTeWekc6Qjyj/3HuHV/a2Mz3JS0xImoRQPXjKeCTlujicnEy6cBJurW7ht5S7uXHOYBy8ZT1ZWFgaDgfTc4ScVg3erAtz18h5Meh0PTXEy/LS5JJob+YGCmdEqFgWz2GUu5BrrYb42LAR5syAe44fxToyxWlb4pqKlZzKv5V201gCJQwdItLZAIk6VPYNffWk+Ryxm1rck+ME//sa5de/0KkNUZ0CZzNhSPejTM9Cl+dCsNlQ4jAp3osIhQEOfkY0hMwd9Zi46jxfVGaKtJUhFUyf1oQRzRrhwDx+O3peJpu87JXnjQDO/ev0wwz02fnreKPJTreh1yd9fJJZg+Xt1PL2xmptfPsj5Bzu5fvpwUqzJ8X1KKX62fic2k4EbzxlNqq33uL8rS+MsWneAfe16ykYM/JZjAzL4Px6PU1lZyRVXXEFxcTFLly5lxYoVfPvb3+bVV1/l8ssvp7S0lPLycpYsWcKdd97Z6xpz5sxhzpw53febmpr6tYxpaWn9fs0vKoll/5A49h+JZf8Y6Die4dNzhi8dSKclHGN/c5iG9i6Meg2jTsOk17Cb9IxKs2I1frhLQpjm5nCf18uzwK/n5BJPKBo7uqhti1LXHiXVamBChh2H+cMJE3YuG+NiTWUrL7/fgtUAd8zMI8fcdVLxGGaFn52Tx8/WVvODZ7fx83PymFiY84nPVUqxta6Dv1Y0U9HYSY7LxE9n5eJzmAjlHpsVOwy4P6Foj8ZJsYwl+rHr/JdSsKmB5e/DO/nFzB3nYcYIN3oNNlcF+J93jmDRwT2jdTxfpVgy6psEz7yAb9oDaLEo4a4Eq9qdLA95UShmRg9xftNW8ra8kxyfZ7Eea2FLxKFiK4Q7qbH5WJV7Frvdw6mx+VBacizcMzWt/OCJnzKlZR+kZyZb7UzmZOuc0Ui5IYcHbFMYEQtw587XsH1gotmTDp50NK8PbHZmo1E6Bp6v1/HCngbePtDID05P58xCD68fbGNLTStXT8kgHmqlKdQ7tmVZBv7oNPLI6wcotMe7k75/p+zs7OMeO2Fi5vF4aG5u7r7f3NyMx+Pp8xyv10s8HicUCuF0Ons91+/34/F48Hq9eL3e7ta00tJSVqxYAcD69euZP38+ANOmTePxxx8/hZcqhBDiiyjFYuixltqnoddpZB3txjweh1nP10Z7uGhU6ifOiDyeUWlWfjUnn5+uqeb2fx7ivJowHmOCbKeJbKeJuFK0huO0RmL4QzH+eaCVA/4wXquBKyf7+HJRCpY+tuSCZAthiqXvr3dN0/ivkgzGZ9h4bmczD22o5887mpic7eCV/S0MSzFzx8xc0mxGRo9TPPx2HX84CG3eHLKcJp7b2URLOM6UHDt2k55/Vhl5ObuA8ZNsXFCcwtRcB0b9sXJ1xRMs31bHs3uDGDTFGIfirFQjY9KtGExGHt+quPdL8zlb18QVre/g6mylqyuGP6zYkvDyhGsKIyMN3NH4KjYtDoEgascmiEZ7TLywAv8BnGXP5JHRl3LfBsW0VWvY6x5OoQoz5/WXSOxIS3bZqgTEYhCPQ6wLXaid/4i4+B9LCWsf/F9mn38m2pgJp/T77E8nTMwKCwupq6ujsbERj8dDeXk5N9xwQ49zJk+ezLp16xg5ciQbNmxg3LhxaJpGSUkJDz30EBdeeCGBQIC6ujqKiorQ6XR4vV5qa2vJzs5mx44d3WPWPB4Pu3btYty4cezcuZPMzMy+iiWEEEIMOE3TjjtR4ESGp1q498vDeHhDHasqGujsShz33CynkeumZjJzhPu4y5OcLE3TKMt3MS3PyebaDp7d0cTL+1o4I9fBzWXZ3a2LBp3GD8uycFr0rNwTAGC8z8rCs9MZk56c2XjF6TFWH2jlH/sC/ObNWpxmPTNHuPhyYQrxhOKhDXVUBiKcme/kqikZvRLGRfk+/lbRzF8qYLP3Isx6HYHOWHfS9aUMGz+ZMRqrcVb3c7qXRvE3JhcqVurDA5xmMfPfNdWsaKrnz2ljiaHjx43r0DXsQbX4IR7rGQy9HmwOSu0OCguG8SfvVKZresyfKsKfjqaUOu56fB/asmULTz31FIlEglmzZnHJJZfw7LPPUlhYSElJCdFolEceeYTKykocDgc33nhj92SB559/nrVr16LT6Zg3bx6TJk0C4ODBgyxZsoRYLIbP52PBggU4HA727NnTvTSH0WjkyiuvpKCg4IQvpLa29lOGoqeBbqL/PJFY9g+JY/+RWPYPiWP/8Xq9vF9dz+G2KPXtXRh0Gm6zHpdFj8usJ81m/Ld1sSmlOByMku009dnyp5Ri/cE23BYDEzP7ngUbTyi213ew+kAr79QEiSWSy5S4LXqunpLJtHznJ5bhYCDM3yr8GPUa6XYD6fbkTNOx6bZTSkQ/Widr2iI0tnd1LyGiEgkIh0CnT05g0OnRdMda996r7+DO16q5dmom5xal9Hn9/vJJXZknlZgNBZKYDV4Sy/4hcew/Esv+IXHsP5+nWLaGY6yrbKMtEufiMR6c5s9uK6xPG8dNh9uZlGX/t48z+1RjzIQQQgghTpbbYuDrYwZ+duO/or/GKX4afY8cFEIIIYQQnzlJzIQQQgghBglJzIQQQgghBglJzIQQQgghBglJzIQQQgghBglJzIQQQgghBglJzIQQQgghBglJzIQQQgghBonPzcr/QgghhBBDnbSYHcfChQsHugifGxLL/iFx7D8Sy/4hcew/Esv+8XmIoyRmQgghhBCDhCRmQgghhBCDhP7uu+++e6ALMVgVFBQMdBE+NySW/UPi2H8klv1D4th/JJb9Y6jHUQb/CyGEEEIMEtKVKYQQQggxSEhiJoQQQggxSBgGugCD0bZt21i6dCmJRILZs2dz8cUXD3SRhoSmpiYeffRRWlpa0DSNOXPm8JWvfIX29nYeeOABjhw5Qnp6OjfddBMOh2OgizvoJRIJFi5ciMfjYeHChTQ2NrJ48WKCwSAFBQVcf/31GAzyFj6Rjo4OlixZQnV1NZqmcc0115CdnS118hStWrWKNWvWoGkaeXl5LFiwgJaWFqmTJ+Gxxx5jy5YtuN1uFi1aBHDcz0WlFEuXLmXr1q2YzWYWLFgw5MdM9ae+Yrls2TI2b96MwWAgIyODBQsWYLfbAVi+fDlr1qxBp9Mxf/58Jk6cOJDFPzlK9BCPx9V1112n6uvrVVdXl7rllltUdXX1QBdrSPD7/erAgQNKKaVCoZC64YYbVHV1tVq2bJlavny5Ukqp5cuXq2XLlg1kMYeMlStXqsWLF6t7771XKaXUokWL1JtvvqmUUurxxx9Xr7zyykAWb8h4+OGH1erVq5VSSnV1dan29napk6eoublZLViwQEUiEaVUsi6uXbtW6uRJqqioUAcOHFA333xz92PHq4ObN29W99xzj0okEmrv3r3q9ttvH5AyD1Z9xXLbtm0qFosppZJx/TCW1dXV6pZbblHRaFQ1NDSo6667TsXj8QEp96mQrsyP2b9/P5mZmWRkZGAwGCgrK2Pjxo0DXawhITU1tfsvO6vVSk5ODn6/n40bNzJjxgwAZsyYIfE8Cc3NzWzZsoXZs2cDoJSioqKC0tJSAGbOnClxPAmhUIjdu3dzzjnnAGAwGLDb7VIn/wWJRIJoNEo8HicajZKSkiJ18iSNHTu2V4vs8ergpk2bOPvss9E0jZEjR9LR0UEgEPjMyzxY9RXLCRMmoNfrARg5ciR+vx9IxrisrAyj0YjP5yMzM5P9+/d/5mU+VdLm/DF+vx+v19t93+v1sm/fvgEs0dDU2NhIZWUlRUVFtLa2kpqaCkBKSgqtra0DXLrB78knn+S73/0unZ2dAASDQWw2W/eHj8fj6f7wEcfX2NiIy+Xiscceo6qqioKCAubNmyd18hR5PB4uuugirrnmGkwmExMmTKCgoEDq5KdwvDro9/tJS0vrPs/r9eL3+7vPFZ9szZo1lJWVAclYFhcXdx8bKnVUWsxEvwuHwyxatIh58+Zhs9l6HNM0DU3TBqhkQ8PmzZtxu90yrqQfxONxKisrOffcc7nvvvswm82sWLGixzlSJ0+svb2djRs38uijj/L4448TDofZtm3bQBfrc0PqYP94/vnn0ev1TJ8+faCL8qlIi9nHeDwempubu+83Nzfj8XgGsERDSywWY9GiRUyfPp2pU6cC4Ha7CQQCpKamEggEcLlcA1zKwW3v3r1s2rSJrVu3Eo1G6ezs5MknnyQUChGPx9Hr9fj9fqmXJ8Hr9eL1erv/ai4tLWXFihVSJ0/Rjh078Pl83XGaOnUqe/fulTr5KRyvDno8HpqamrrPk++gk7Nu3To2b97MXXfd1Z3kfvz7fKjUUWkx+5jCwkLq6upobGwkFotRXl5OSUnJQBdrSFBKsWTJEnJycrjwwgu7Hy8pKWH9+vUArF+/nilTpgxUEYeE73znOyxZsoRHH32UG2+8kfHjx3PDDTcwbtw4NmzYACQ/hKRenlhKSgper5fa2logmWDk5uZKnTxFaWlp7Nu3j0gkglKqO45SJ/91x6uDJSUlvP766yileP/997HZbNKNeQLbtm3jhRde4LbbbsNsNnc/XlJSQnl5OV1dXTQ2NlJXV0dRUdEAlvTkyMr/fdiyZQtPPfUUiUSCWbNmcckllwx0kYaEPXv2cNddd5Gfn9/9F8tll11GcXExDzzwAE1NTbI0wSmqqKhg5cqVLFy4kIaGBhYvXkx7ezsjRozg+uuvx2g0DnQRB72DBw+yZMkSYrEYPp+PBQsWoJSSOnmKnnvuOcrLy9Hr9QwfPpyrr74av98vdfIkLF68mF27dhEMBnG73Vx66aVMmTKlzzqolOI4G8sqAAAAjElEQVSJJ55g+/btmEwmFixYQGFh4UC/hEGjr1guX76cWCzW/R4uLi7mqquuApLdm2vXrkWn0zFv3jwmTZo0kMU/KZKYCSGEEEIMEtKVKYQQQggxSEhiJoQQQggxSEhiJoQQQggxSEhiJoQQQggxSEhiJoQQQggxSEhiJoQQQggxSEhiJoQQQggxSPw/Zb0BPA7XucMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.style.use('ggplot')\n",
    "plt.figure(figsize=(10,4))\n",
    "\n",
    "plt.plot(range(len(history_modelA.history['loss']))[1:], history_modelA.history['loss'][1:], label='Training')\n",
    "plt.plot(range(len(history_modelA.history['loss']))[1:], history_modelA.history['val_loss'][1:], label='Validation')\n",
    "plt.title('Loss')\n",
    "plt.legend()\n",
    "plt.show()\n",
    "plt.savefig('ModelC_125epoch.png')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "cwd = os.getcwd()\n",
    "AE.save_weights(cwd+'/ExpTestModels228/ModelC/modelC')\n",
    "\n",
    "import pickle \n",
    "with open(cwd + '/ExpTestModels228/pickles/modelC.pickle', 'wb') as handle:\n",
    "    pickle.dump(history_modelA.history, handle)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model C Latent Representations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "decoded_data = AE.predict(s_valid_reshaped)\n",
    "mini_batch = s_valid_reshaped[:100]\n",
    "#print(mini_batch.shape)\n",
    "get_hl = K.function([AE.layers[0].input], [AE.layers[8].output])\n",
    "AE_h1 = get_hl(mini_batch)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x288 with 12 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n = 4  # how many digits we will display\n",
    "plt.figure(figsize=(20, 4))\n",
    "for i in range(n):\n",
    "    # display original\n",
    "    ax = plt.subplot(3, n, i + 1)\n",
    "    plt.imshow(s_valid_reshaped[i].reshape(96, 216))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display hidden layer representation\n",
    "    ax = plt.subplot(3, n, i + 1 + n)\n",
    "    plt.imshow(AE_h1[i].reshape(12, 15)) # Will have to reshape if you change the latent dimensionality\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "\n",
    "    # display reconstruction\n",
    "    ax = plt.subplot(3, n, i + 1 + n + n)\n",
    "    plt.imshow(decoded_data[i].reshape(96, 216))\n",
    "    plt.gray()\n",
    "    ax.get_xaxis().set_visible(False)\n",
    "    ax.get_yaxis().set_visible(False)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "py37-tf2",
   "language": "python",
   "name": "py37-tf2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}