Preprocessing.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'utils.ml_utils' from '/home/darylchang/ml-experiments/utils/ml_utils.py'>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import tensorflow_datasets as tfds\n",
    "from tensorflow import keras\n",
    "from tensorflow.keras import layers, initializers\n",
    "\n",
    "import numpy as np\n",
    "import os\n",
    "import random\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import importlib\n",
    "import utils.ml_utils as ml_utils\n",
    "importlib.reload(ml_utils)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Init seeds and env variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "ml_utils.init_env()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Canonical data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Data Summary\n",
      "No. of examples: 18610\n",
      "Class: 0 :::: Count: 9378 :::: Percentage: 50.39226222461043\n",
      "Class: 1 :::: Count: 9232 :::: Percentage: 49.60773777538957\n",
      "\n",
      "Validation Data Summary\n",
      "No. of examples: 4652\n",
      "Class: 0 :::: Count: 2280 :::: Percentage: 49.011177987962164\n",
      "Class: 1 :::: Count: 2372 :::: Percentage: 50.988822012037836\n"
     ]
    }
   ],
   "source": [
    "# Constants\n",
    "IMG_SIZE = 128 # All images in the dataset will be resized to this size\n",
    "BATCH_SIZE = 32 # Batch size\n",
    "\n",
    "# Read in training and validation data\n",
    "train, validation = ml_utils.load_batched_and_resized_dataset(\n",
    "    dataset_name='cats_and_dogs',\n",
    "    batch_size=BATCH_SIZE,\n",
    "    img_size=IMG_SIZE\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "582/582 [==============================] - 40s 69ms/step - loss: 0.6860 - accuracy: 0.5435 - val_loss: 0.6627 - val_accuracy: 0.6279\n",
      "Epoch 2/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6911 - accuracy: 0.5155 - val_loss: 0.6928 - val_accuracy: 0.5099\n",
      "Epoch 3/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6900 - accuracy: 0.5264 - val_loss: 0.6677 - val_accuracy: 0.6131\n",
      "Epoch 4/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6494 - accuracy: 0.6271 - val_loss: 0.6357 - val_accuracy: 0.6464\n",
      "Epoch 5/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6255 - accuracy: 0.6538 - val_loss: 0.6208 - val_accuracy: 0.6582\n",
      "Epoch 6/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6082 - accuracy: 0.6749 - val_loss: 0.5961 - val_accuracy: 0.6930\n",
      "Epoch 7/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5909 - accuracy: 0.6867 - val_loss: 0.5770 - val_accuracy: 0.7096\n",
      "Epoch 8/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5729 - accuracy: 0.7031 - val_loss: 0.5760 - val_accuracy: 0.7016\n",
      "Epoch 9/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5585 - accuracy: 0.7104 - val_loss: 0.5525 - val_accuracy: 0.7197\n",
      "Epoch 10/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5435 - accuracy: 0.7195 - val_loss: 0.5312 - val_accuracy: 0.7341\n",
      "Epoch 11/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5364 - accuracy: 0.7294 - val_loss: 0.5351 - val_accuracy: 0.7489\n",
      "Epoch 12/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5233 - accuracy: 0.7365 - val_loss: 0.5556 - val_accuracy: 0.7197\n",
      "Epoch 13/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5132 - accuracy: 0.7459 - val_loss: 0.4851 - val_accuracy: 0.7700\n",
      "Epoch 14/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.5013 - accuracy: 0.7510 - val_loss: 0.4809 - val_accuracy: 0.7689\n",
      "Epoch 15/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4926 - accuracy: 0.7577 - val_loss: 0.5318 - val_accuracy: 0.7377\n",
      "Epoch 16/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4800 - accuracy: 0.7698 - val_loss: 0.4698 - val_accuracy: 0.7820\n",
      "Epoch 17/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4731 - accuracy: 0.7707 - val_loss: 0.4790 - val_accuracy: 0.7741\n",
      "Epoch 18/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4685 - accuracy: 0.7741 - val_loss: 0.4502 - val_accuracy: 0.7900\n",
      "Epoch 19/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4628 - accuracy: 0.7762 - val_loss: 0.4516 - val_accuracy: 0.7850\n",
      "Epoch 20/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4569 - accuracy: 0.7828 - val_loss: 0.4395 - val_accuracy: 0.7883\n",
      "Epoch 21/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4494 - accuracy: 0.7859 - val_loss: 0.4357 - val_accuracy: 0.7947\n",
      "Epoch 22/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4453 - accuracy: 0.7882 - val_loss: 0.4603 - val_accuracy: 0.7827\n",
      "Epoch 23/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4356 - accuracy: 0.7938 - val_loss: 0.4388 - val_accuracy: 0.7911\n",
      "Epoch 24/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4373 - accuracy: 0.7917 - val_loss: 0.4355 - val_accuracy: 0.8001\n",
      "Epoch 25/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4307 - accuracy: 0.7961 - val_loss: 0.4403 - val_accuracy: 0.7865\n",
      "Epoch 26/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4252 - accuracy: 0.8015 - val_loss: 0.4356 - val_accuracy: 0.7913\n",
      "Epoch 27/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4189 - accuracy: 0.8077 - val_loss: 0.4106 - val_accuracy: 0.8072\n",
      "Epoch 28/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4213 - accuracy: 0.8042 - val_loss: 0.4330 - val_accuracy: 0.7936\n",
      "Epoch 29/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4118 - accuracy: 0.8071 - val_loss: 0.4518 - val_accuracy: 0.7853\n",
      "Epoch 30/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4044 - accuracy: 0.8132 - val_loss: 0.4123 - val_accuracy: 0.8091\n",
      "Epoch 31/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.4017 - accuracy: 0.8119 - val_loss: 0.4116 - val_accuracy: 0.8093\n",
      "Epoch 32/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3991 - accuracy: 0.8150 - val_loss: 0.4009 - val_accuracy: 0.8119\n",
      "Epoch 33/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3952 - accuracy: 0.8180 - val_loss: 0.3920 - val_accuracy: 0.8192\n",
      "Epoch 34/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3907 - accuracy: 0.8235 - val_loss: 0.3878 - val_accuracy: 0.8224\n",
      "Epoch 35/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3867 - accuracy: 0.8242 - val_loss: 0.4230 - val_accuracy: 0.8001\n",
      "Epoch 36/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3831 - accuracy: 0.8219 - val_loss: 0.4248 - val_accuracy: 0.7960\n",
      "Epoch 37/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3848 - accuracy: 0.8237 - val_loss: 0.3934 - val_accuracy: 0.8201\n",
      "Epoch 38/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3766 - accuracy: 0.8275 - val_loss: 0.3940 - val_accuracy: 0.8181\n",
      "Epoch 39/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3721 - accuracy: 0.8298 - val_loss: 0.4245 - val_accuracy: 0.7962\n",
      "Epoch 40/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3667 - accuracy: 0.8347 - val_loss: 0.4021 - val_accuracy: 0.8141\n",
      "Epoch 41/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3636 - accuracy: 0.8341 - val_loss: 0.4226 - val_accuracy: 0.8037\n",
      "Epoch 42/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3650 - accuracy: 0.8326 - val_loss: 0.3992 - val_accuracy: 0.8160\n",
      "Epoch 43/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3646 - accuracy: 0.8346 - val_loss: 0.4001 - val_accuracy: 0.8106\n",
      "Epoch 44/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3574 - accuracy: 0.8368 - val_loss: 0.3845 - val_accuracy: 0.8239\n",
      "Epoch 45/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3610 - accuracy: 0.8340 - val_loss: 0.3737 - val_accuracy: 0.8233\n",
      "Epoch 46/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3513 - accuracy: 0.8395 - val_loss: 0.3925 - val_accuracy: 0.8209\n",
      "Epoch 47/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3523 - accuracy: 0.8424 - val_loss: 0.3995 - val_accuracy: 0.8164\n",
      "Epoch 48/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3422 - accuracy: 0.8476 - val_loss: 0.3691 - val_accuracy: 0.8261\n",
      "Epoch 49/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3462 - accuracy: 0.8440 - val_loss: 0.3781 - val_accuracy: 0.8214\n",
      "Epoch 50/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3397 - accuracy: 0.8486 - val_loss: 0.3985 - val_accuracy: 0.8126\n",
      "Epoch 51/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3431 - accuracy: 0.8475 - val_loss: 0.3889 - val_accuracy: 0.8222\n",
      "Epoch 52/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3328 - accuracy: 0.8495 - val_loss: 0.3731 - val_accuracy: 0.8265\n",
      "Epoch 53/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3283 - accuracy: 0.8558 - val_loss: 0.3992 - val_accuracy: 0.8132\n",
      "Epoch 54/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3260 - accuracy: 0.8539 - val_loss: 0.3845 - val_accuracy: 0.8194\n",
      "Epoch 55/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3278 - accuracy: 0.8552 - val_loss: 0.3840 - val_accuracy: 0.8246\n",
      "Epoch 56/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3281 - accuracy: 0.8557 - val_loss: 0.4000 - val_accuracy: 0.8147\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 57/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3231 - accuracy: 0.8590 - val_loss: 0.4176 - val_accuracy: 0.8102\n",
      "Epoch 58/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3202 - accuracy: 0.8583 - val_loss: 0.4105 - val_accuracy: 0.8136\n",
      "Epoch 59/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3139 - accuracy: 0.8615 - val_loss: 0.3849 - val_accuracy: 0.8199\n",
      "Epoch 60/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3153 - accuracy: 0.8617 - val_loss: 0.3995 - val_accuracy: 0.8145\n",
      "Epoch 61/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3169 - accuracy: 0.8611 - val_loss: 0.3813 - val_accuracy: 0.8184\n",
      "Epoch 62/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3083 - accuracy: 0.8638 - val_loss: 0.3823 - val_accuracy: 0.8261\n",
      "Epoch 63/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3130 - accuracy: 0.8607 - val_loss: 0.3831 - val_accuracy: 0.8280\n",
      "Epoch 64/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3108 - accuracy: 0.8659 - val_loss: 0.3875 - val_accuracy: 0.8259\n",
      "Epoch 65/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3010 - accuracy: 0.8698 - val_loss: 0.4109 - val_accuracy: 0.8100\n",
      "Epoch 66/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3039 - accuracy: 0.8682 - val_loss: 0.3982 - val_accuracy: 0.8218\n",
      "Epoch 67/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3049 - accuracy: 0.8649 - val_loss: 0.3802 - val_accuracy: 0.8261\n",
      "Epoch 68/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2974 - accuracy: 0.8716 - val_loss: 0.3890 - val_accuracy: 0.8220\n",
      "Epoch 69/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.3003 - accuracy: 0.8703 - val_loss: 0.3763 - val_accuracy: 0.8289\n",
      "Epoch 70/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2943 - accuracy: 0.8676 - val_loss: 0.3852 - val_accuracy: 0.8289\n",
      "Epoch 71/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2963 - accuracy: 0.8732 - val_loss: 0.3878 - val_accuracy: 0.8227\n",
      "Epoch 72/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2837 - accuracy: 0.8764 - val_loss: 0.3824 - val_accuracy: 0.8141\n",
      "Epoch 73/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2903 - accuracy: 0.8746 - val_loss: 0.4015 - val_accuracy: 0.8123\n",
      "Epoch 74/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2880 - accuracy: 0.8750 - val_loss: 0.3712 - val_accuracy: 0.8248\n",
      "Epoch 75/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2824 - accuracy: 0.8765 - val_loss: 0.3849 - val_accuracy: 0.8222\n",
      "Epoch 76/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2900 - accuracy: 0.8801 - val_loss: 0.4006 - val_accuracy: 0.8166\n",
      "Epoch 77/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2743 - accuracy: 0.8826 - val_loss: 0.4092 - val_accuracy: 0.8141\n",
      "Epoch 78/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2867 - accuracy: 0.8761 - val_loss: 0.3956 - val_accuracy: 0.8166\n",
      "Epoch 79/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2770 - accuracy: 0.8825 - val_loss: 0.3796 - val_accuracy: 0.8237\n",
      "Epoch 80/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2785 - accuracy: 0.8832 - val_loss: 0.3818 - val_accuracy: 0.8267\n",
      "Epoch 81/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2687 - accuracy: 0.8852 - val_loss: 0.4812 - val_accuracy: 0.7915\n",
      "Epoch 82/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2802 - accuracy: 0.8802 - val_loss: 0.3905 - val_accuracy: 0.8222\n",
      "Epoch 83/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2728 - accuracy: 0.8823 - val_loss: 0.4001 - val_accuracy: 0.8156\n",
      "Epoch 84/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2747 - accuracy: 0.8807 - val_loss: 0.4112 - val_accuracy: 0.8134\n",
      "Epoch 85/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2583 - accuracy: 0.8904 - val_loss: 0.3929 - val_accuracy: 0.8310\n",
      "Epoch 86/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2658 - accuracy: 0.8891 - val_loss: 0.3949 - val_accuracy: 0.8181\n",
      "Epoch 87/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2630 - accuracy: 0.8887 - val_loss: 0.3986 - val_accuracy: 0.8229\n",
      "Epoch 88/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2604 - accuracy: 0.8892 - val_loss: 0.3974 - val_accuracy: 0.8289\n",
      "Epoch 89/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2629 - accuracy: 0.8922 - val_loss: 0.3888 - val_accuracy: 0.8203\n",
      "Epoch 90/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2608 - accuracy: 0.8854 - val_loss: 0.4328 - val_accuracy: 0.8009\n",
      "Epoch 91/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2616 - accuracy: 0.8888 - val_loss: 0.3941 - val_accuracy: 0.8237\n",
      "Epoch 92/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2557 - accuracy: 0.8930 - val_loss: 0.4210 - val_accuracy: 0.8117\n",
      "Epoch 93/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2539 - accuracy: 0.8919 - val_loss: 0.3901 - val_accuracy: 0.8246\n",
      "Epoch 94/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2506 - accuracy: 0.8954 - val_loss: 0.4062 - val_accuracy: 0.8216\n",
      "Epoch 95/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2441 - accuracy: 0.8962 - val_loss: 0.4021 - val_accuracy: 0.8184\n",
      "Epoch 96/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2512 - accuracy: 0.8961 - val_loss: 0.4149 - val_accuracy: 0.8138\n",
      "Epoch 97/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2482 - accuracy: 0.8976 - val_loss: 0.4469 - val_accuracy: 0.8042\n",
      "Epoch 98/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.2530 - accuracy: 0.8927 - val_loss: 0.4063 - val_accuracy: 0.8188\n",
      "Epoch 00098: early stopping\n"
     ]
    }
   ],
   "source": [
    "model = ml_utils.build_model()\n",
    "es = keras.callbacks.EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=50)\n",
    "model_state_canonical_data = ml_utils.train_model(\n",
    "    model,\n",
    "    train,\n",
    "    validation,\n",
    "    epochs=1000,\n",
    "    extra_callbacks=[es],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "ml_utils.save_model_state(model_state_canonical_data, 'model_state_canonical_data')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Try unnormalized data (no division by 255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training Data Summary\n",
      "No. of examples: 18610\n",
      "Class: 0 :::: Count: 9378 :::: Percentage: 50.39226222461043\n",
      "Class: 1 :::: Count: 9232 :::: Percentage: 49.60773777538957\n",
      "\n",
      "Validation Data Summary\n",
      "No. of examples: 4652\n",
      "Class: 0 :::: Count: 2280 :::: Percentage: 49.011177987962164\n",
      "Class: 1 :::: Count: 2372 :::: Percentage: 50.988822012037836\n"
     ]
    }
   ],
   "source": [
    "# Read in unnormalized training and validation data\n",
    "train_unnormalized, validation_unnormalized = ml_utils.load_batched_and_resized_dataset(\n",
    "    dataset_name='cats_and_dogs',\n",
    "    batch_size=BATCH_SIZE,\n",
    "    img_size=IMG_SIZE,\n",
    "    normalize_pixel_values=False,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "582/582 [==============================] - 38s 65ms/step - loss: nan - accuracy: 0.5031 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 2/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 3/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 4/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 5/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 6/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 7/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 8/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 9/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 10/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 11/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 12/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 13/1000\n",
      "582/582 [==============================] - 19s 33ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 14/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 15/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 16/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 17/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 18/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 19/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 20/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 21/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 22/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 23/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 24/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 25/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 26/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 27/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 28/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 29/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 30/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 31/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 32/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 33/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 34/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 35/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 36/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 37/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 38/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 39/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 40/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 41/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 42/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 43/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 44/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 45/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 46/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 47/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 48/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 49/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 50/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: nan - accuracy: 0.5039 - val_loss: nan - val_accuracy: 0.4901\n",
      "Epoch 00050: early stopping\n"
     ]
    }
   ],
   "source": [
    "model_unnormalized_data = ml_utils.build_model()\n",
    "es = keras.callbacks.EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=50)\n",
    "model_state_unnormalized_data = ml_utils.train_model(\n",
    "    model_state_unnormalized_data,\n",
    "    train_unnormalized,\n",
    "    validation_unnormalized,\n",
    "    epochs=1000,\n",
    "    extra_callbacks=[es],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "ml_utils.save_model_state(model_state_unnormalized_data, 'model_state_unnormalized_data')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[[[-1.5690454e+00, -2.0285955e+03,            nan,\n",
       "           -7.7620160e+02],\n",
       "          [-1.3897611e+00, -1.9286010e+03,            nan,\n",
       "           -7.3600848e+02],\n",
       "          [-9.6283960e-01, -1.7204238e+03,            nan,\n",
       "           -6.5323779e+02]],\n",
       " \n",
       "         [[-1.2804551e+00, -2.0264070e+03,            nan,\n",
       "           -7.8193634e+02],\n",
       "          [-1.3969817e+00, -1.9278486e+03,            nan,\n",
       "           -7.4182678e+02],\n",
       "          [-1.4119167e+00, -1.7204136e+03,            nan,\n",
       "           -6.5898407e+02]],\n",
       " \n",
       "         [[-1.2528425e+00, -2.0101455e+03,            nan,\n",
       "           -7.7297058e+02],\n",
       "          [-1.5777417e+00, -1.9109780e+03,            nan,\n",
       "           -7.3252527e+02],\n",
       "          [-1.4428709e+00, -1.7036863e+03,            nan,\n",
       "           -6.5019873e+02]]],\n",
       " \n",
       " \n",
       "        [[[-1.3593225e+00, -2.0297722e+03,            nan,\n",
       "           -7.8896619e+02],\n",
       "          [-1.0609597e+00, -1.9327545e+03,            nan,\n",
       "           -7.4870874e+02],\n",
       "          [-9.3179268e-01, -1.7234253e+03,            nan,\n",
       "           -6.6570020e+02]],\n",
       " \n",
       "         [[-1.3951780e+00, -1.9447755e+03,            nan,\n",
       "           -7.8544202e+02],\n",
       "          [-1.0774081e+00, -1.8489569e+03,            nan,\n",
       "           -7.4619751e+02],\n",
       "          [-1.1659104e+00, -1.6394982e+03,            nan,\n",
       "           -6.6201410e+02]],\n",
       " \n",
       "         [[-1.5475447e+00, -2.0082526e+03,            nan,\n",
       "           -7.7959003e+02],\n",
       "          [-1.1459655e+00, -1.9109827e+03,            nan,\n",
       "           -7.3979486e+02],\n",
       "          [-1.2348763e+00, -1.7013574e+03,            nan,\n",
       "           -6.5650745e+02]]],\n",
       " \n",
       " \n",
       "        [[[-1.3096398e+00, -2.0112871e+03,            nan,\n",
       "           -7.8575372e+02],\n",
       "          [-1.6258224e+00, -1.9151582e+03,            nan,\n",
       "           -7.4583295e+02],\n",
       "          [-9.7873735e-01, -1.7044700e+03,            nan,\n",
       "           -6.6229840e+02]],\n",
       " \n",
       "         [[-1.2539674e+00, -1.9813301e+03,            nan,\n",
       "           -7.8657666e+02],\n",
       "          [-1.2564228e+00, -1.8860168e+03,            nan,\n",
       "           -7.4789392e+02],\n",
       "          [-1.1001216e+00, -1.6749252e+03,            nan,\n",
       "           -6.6329980e+02]],\n",
       " \n",
       "         [[-1.1058891e+00, -1.9736277e+03,            nan,\n",
       "           -7.7933667e+02],\n",
       "          [-1.2168221e+00, -1.8766060e+03,            nan,\n",
       "           -7.4018689e+02],\n",
       "          [-1.0872716e+00, -1.6654314e+03,            nan,\n",
       "           -6.5670197e+02]]]], dtype=float32),\n",
       " array([-7.3328577e-03, -1.3333699e+01,            nan, -4.9924116e+00],\n",
       "       dtype=float32),\n",
       " array([[[[-3.62823121e-02, -2.75069892e-01,  1.75676852e-01,\n",
       "                       nan, -9.55455750e-02,             nan,\n",
       "            3.68035018e-01,  2.18467280e-01],\n",
       "          [-4.07742188e+03, -7.10216980e+02, -5.53445473e+01,\n",
       "                       nan,  5.43372154e-01,             nan,\n",
       "           -1.09799182e+03, -6.68234680e+02],\n",
       "          [-3.40587646e-02, -1.36124939e-01, -1.33688956e-01,\n",
       "                       nan, -1.58370972e-01,             nan,\n",
       "           -4.17731330e-02, -1.42007083e-01],\n",
       "          [-1.47476318e+04, -1.21800972e+02, -1.88129822e+02,\n",
       "                       nan, -7.09621191e-01,             nan,\n",
       "           -2.77558789e+03, -2.24326123e+03]],\n",
       " \n",
       "         [[-1.29921198e-01, -2.53292382e-01, -2.39181519e-01,\n",
       "                       nan,  1.13951107e-02,             nan,\n",
       "           -1.95082903e-01,  1.13738865e-01],\n",
       "          [-4.04592358e+03, -7.13016113e+02, -7.55863266e+01,\n",
       "                       nan,  8.59323919e-01,             nan,\n",
       "           -1.20973511e+03, -6.70440552e+02],\n",
       "          [-6.04415573e-02, -2.01699570e-01,  5.42493686e-02,\n",
       "                       nan,  1.26217812e-01,             nan,\n",
       "            1.40778702e-02, -2.22513363e-01],\n",
       "          [-1.25814062e+04, -1.93297192e+03, -3.49244781e+02,\n",
       "                       nan,  2.89858603e+00,             nan,\n",
       "           -4.80852490e+03, -9.29785034e+02]],\n",
       " \n",
       "         [[ 1.99455887e-01, -1.46543041e-01, -1.93702340e-01,\n",
       "                       nan, -7.27542415e-02,             nan,\n",
       "            9.70910043e-02,  2.86748558e-01],\n",
       "          [-3.69264136e+03, -6.64910339e+02, -9.21754761e+01,\n",
       "                       nan,  2.05115997e-03,             nan,\n",
       "           -1.08713477e+03, -7.27054321e+02],\n",
       "          [ 1.57673851e-01,  1.34597301e-01, -1.65799320e-01,\n",
       "                       nan,  8.97794366e-02,             nan,\n",
       "           -1.61082789e-01,  1.75789908e-01],\n",
       "          [-1.23224277e+04, -1.48718030e+03, -2.37797089e+02,\n",
       "                       nan,  6.79758132e-01,             nan,\n",
       "           -3.64634302e+03, -2.97233960e+03]]],\n",
       " \n",
       " \n",
       "        [[[ 5.65329241e-03, -3.34008157e-01,  1.47189394e-01,\n",
       "                       nan, -1.59211736e-02,             nan,\n",
       "            1.62072498e-02,  3.10071617e-01],\n",
       "          [-3.62589453e+03, -6.51448486e+02, -5.92747040e+01,\n",
       "                       nan,  1.11000431e+00,             nan,\n",
       "           -1.05906616e+03, -7.02550415e+02],\n",
       "          [-1.24320500e-02, -1.62773788e-01, -1.10085253e-02,\n",
       "                       nan,  6.47272915e-02,             nan,\n",
       "           -1.57038495e-01, -2.50469912e-02],\n",
       "          [-1.44294482e+04, -2.65884595e+03, -1.28051010e+02,\n",
       "                       nan,  2.19515443e+00,             nan,\n",
       "           -4.55603564e+03, -2.83563745e+03]],\n",
       " \n",
       "         [[-3.87975335e-01,  5.29076830e-02,  1.96308061e-01,\n",
       "                       nan,  1.03169046e-02,             nan,\n",
       "           -1.34388849e-01, -1.63493350e-01],\n",
       "          [-4.05696240e+03, -7.61154236e+02, -8.67857971e+01,\n",
       "                       nan,  6.94151342e-01,             nan,\n",
       "           -1.18950378e+03, -6.92993225e+02],\n",
       "          [ 1.52245358e-01,  5.54539599e-02,  7.06952214e-02,\n",
       "                       nan,  1.55102639e-02,             nan,\n",
       "            2.27968112e-01,  1.86134428e-01],\n",
       "          [-1.10463584e+04, -1.16563892e+03, -3.38251862e+02,\n",
       "                       nan, -9.07430127e-02,             nan,\n",
       "           -3.97169873e+03, -1.88549634e+03]],\n",
       " \n",
       "         [[-1.25982976e-02, -1.85281381e-01, -2.60324571e-02,\n",
       "                       nan, -2.42998704e-01,             nan,\n",
       "           -5.33968955e-02,  1.29170001e-01],\n",
       "          [-3.49681445e+03, -7.55667297e+02, -9.37520828e+01,\n",
       "                       nan,  8.05511177e-01,             nan,\n",
       "           -1.15411829e+03, -7.52674866e+02],\n",
       "          [ 1.51546309e-02, -1.69918220e-02,  7.77056366e-02,\n",
       "                       nan,  1.59853771e-01,             nan,\n",
       "           -3.45188156e-02, -9.68569666e-02],\n",
       "          [-8.92879492e+03, -3.11877979e+03, -2.11474152e+02,\n",
       "                       nan, -2.99512297e-01,             nan,\n",
       "           -2.53020459e+03, -3.04757422e+03]]],\n",
       " \n",
       " \n",
       "        [[[-1.82168528e-01, -1.93564221e-01,  7.28423148e-02,\n",
       "                       nan,  4.90429774e-02,             nan,\n",
       "            2.33853385e-01,  2.71695048e-01],\n",
       "          [-3.48016162e+03, -7.22337646e+02, -4.60343552e+01,\n",
       "                       nan,  2.05440149e-01,             nan,\n",
       "           -1.21884192e+03, -6.34060547e+02],\n",
       "          [-1.62330016e-01, -1.25964642e-01, -9.60181579e-02,\n",
       "                       nan,  2.24084616e-01,             nan,\n",
       "           -3.79215926e-02,  1.25081837e-01],\n",
       "          [-1.00899033e+04, -2.95839795e+03, -1.90290710e+02,\n",
       "                       nan,  7.62230873e-01,             nan,\n",
       "           -4.43928564e+03, -2.13008447e+03]],\n",
       " \n",
       "         [[ 1.00983642e-01,  1.36549577e-01, -1.55632913e-01,\n",
       "                       nan, -2.37218648e-01,             nan,\n",
       "           -6.52899742e-02,  1.47456318e-01],\n",
       "          [-3.08506006e+03, -6.55469238e+02, -6.46657410e+01,\n",
       "                       nan,  6.30559027e-01,             nan,\n",
       "           -1.12483142e+03, -7.44320190e+02],\n",
       "          [ 1.73795566e-01,  2.22627103e-01, -1.47218764e-01,\n",
       "                       nan,  6.30201399e-02,             nan,\n",
       "            1.84714973e-01,  8.68241563e-02],\n",
       "          [-1.23984160e+04, -2.25069116e+03, -2.43278561e+01,\n",
       "                       nan,  1.45789099e+00,             nan,\n",
       "           -3.53613159e+03, -2.53883228e+03]],\n",
       " \n",
       "         [[ 9.53447595e-02,  2.84593970e-01,  1.06044590e-01,\n",
       "                       nan, -1.21286176e-01,             nan,\n",
       "            2.19758734e-01, -1.01085253e-01],\n",
       "          [-3.23984229e+03, -8.11514709e+02, -6.60881577e+01,\n",
       "                       nan,  5.08288860e-01,             nan,\n",
       "           -1.00453564e+03, -7.27195435e+02],\n",
       "          [-1.06060877e-01, -1.47456646e-01, -1.63890913e-01,\n",
       "                       nan, -7.87033811e-02,             nan,\n",
       "           -1.10154888e-02,  1.61319435e-01],\n",
       "          [-1.48608809e+04, -3.06586548e+03, -2.94374725e+02,\n",
       "                       nan, -1.62907374e+00,             nan,\n",
       "           -3.06298145e+03, -1.92037305e+03]]]], dtype=float32),\n",
       " array([-1.1814677e+01, -2.4813504e+00, -2.4802883e-01,            nan,\n",
       "        -1.1596936e+11,            nan, -3.6238396e+00, -2.2612653e+00],\n",
       "       dtype=float32),\n",
       " array([[[[-3.28852142e+02, -6.17560425e+02,  7.18327910e-02, ...,\n",
       "           -4.19295464e+01, -7.41815552e+02, -1.12791089e+03],\n",
       "          [-1.88195667e+01, -2.66309032e+01, -2.27924228e-01, ...,\n",
       "           -2.03937078e+00, -5.19950142e+01, -7.76581421e+01],\n",
       "          [-1.72282009e+01, -4.91600304e+01, -1.07977800e-01, ...,\n",
       "           -2.84476018e+00, -6.20191841e+01, -7.90372314e+01],\n",
       "          ...,\n",
       "          [-3.39843775e-03, -3.57091665e-01, -1.50484517e-01, ...,\n",
       "           -6.35184199e-02, -3.60413790e-01, -4.32536840e-01],\n",
       "          [-5.99642029e+02, -7.92440857e+02,  5.77768832e-02, ...,\n",
       "           -5.85403900e+01, -1.05590503e+03, -1.69152283e+03],\n",
       "          [-3.59000702e+02, -5.06313202e+02, -3.88574861e-02, ...,\n",
       "           -3.81640091e+01, -9.19542725e+02, -1.22168762e+03]],\n",
       " \n",
       "         [[-3.48401978e+02, -6.62485474e+02, -2.26951718e-01, ...,\n",
       "           -4.94665871e+01, -9.00284973e+02, -1.01122076e+03],\n",
       "          [-3.37568626e+01, -5.02050514e+01, -8.49515125e-02, ...,\n",
       "           -2.42658305e+00, -6.50951080e+01, -7.18801727e+01],\n",
       "          [-3.15871582e+01, -4.76086845e+01, -1.98609918e-01, ...,\n",
       "           -4.18431139e+00, -7.48256149e+01, -7.93405304e+01],\n",
       "          ...,\n",
       "          [-4.76110250e-01, -3.59636754e-01, -8.85893181e-02, ...,\n",
       "            4.43956740e-02, -3.42622846e-01, -4.71115679e-01],\n",
       "          [-6.06692078e+02, -8.46267273e+02, -1.32584214e-01, ...,\n",
       "           -4.71083336e+01, -1.33038696e+03, -1.77329382e+03],\n",
       "          [-4.15741882e+02, -7.15478271e+02, -1.29497156e-01, ...,\n",
       "           -3.30812302e+01, -7.34725769e+02, -9.70144775e+02]],\n",
       " \n",
       "         [[-3.80575409e+02, -6.71951904e+02, -1.17356755e-01, ...,\n",
       "           -5.79717598e+01, -7.32441833e+02, -1.25368323e+03],\n",
       "          [-2.53403378e+01, -3.51279335e+01, -2.66931146e-01, ...,\n",
       "           -1.42158759e+00, -5.83707809e+01, -6.75667801e+01],\n",
       "          [-2.81748867e+01, -3.91036911e+01, -5.27491495e-02, ...,\n",
       "           -2.28316712e+00, -4.14464569e+01, -6.03400154e+01],\n",
       "          ...,\n",
       "          [-2.35438287e-01, -2.01389968e-01, -3.60111706e-02, ...,\n",
       "           -5.92333823e-02, -4.45919961e-01, -7.78820395e-01],\n",
       "          [-4.09155884e+02, -8.19136902e+02,  7.48837441e-02, ...,\n",
       "           -8.58987427e+01, -9.53996826e+02, -1.83538623e+03],\n",
       "          [-3.67463226e+02, -4.88450073e+02, -4.68099490e-02, ...,\n",
       "           -4.83086510e+01, -8.33160950e+02, -9.27278625e+02]]],\n",
       " \n",
       " \n",
       "        [[[-3.93205658e+02, -4.91380066e+02, -1.57064244e-01, ...,\n",
       "           -5.21773338e+01, -8.47046692e+02, -9.86199524e+02],\n",
       "          [-1.26646938e+01, -4.84835358e+01, -1.83056891e-01, ...,\n",
       "           -3.27104950e+00, -3.78271751e+01, -5.14521942e+01],\n",
       "          [-3.39840698e+01, -4.42543144e+01, -1.83076233e-01, ...,\n",
       "           -3.61077189e+00, -4.62076797e+01, -6.00841522e+01],\n",
       "          ...,\n",
       "          [-1.91121444e-01, -6.49560630e-01, -4.18890081e-02, ...,\n",
       "           -5.24261333e-02, -1.44376129e-01, -7.10576117e-01],\n",
       "          [-5.20939941e+02, -7.30516663e+02, -1.63587719e-01, ...,\n",
       "           -9.23995132e+01, -1.25000183e+03, -1.65414746e+03],\n",
       "          [-2.89267120e+02, -5.07554016e+02,  5.62040210e-02, ...,\n",
       "           -4.01084671e+01, -7.03497131e+02, -1.04962427e+03]],\n",
       " \n",
       "         [[-4.22456787e+02, -4.61450348e+02, -2.18037605e-01, ...,\n",
       "           -2.66671371e+01, -8.06905579e+02, -1.19617664e+03],\n",
       "          [-2.96976604e+01, -3.99690437e+01, -1.51546687e-01, ...,\n",
       "           -2.64502859e+00, -5.28815651e+01, -4.63975334e+01],\n",
       "          [-2.46241837e+01, -4.05737915e+01,  7.30767474e-02, ...,\n",
       "           -6.30601072e+00, -6.81261749e+01, -9.50483246e+01],\n",
       "          ...,\n",
       "          [-2.25347117e-01, -6.47708103e-02, -4.32749614e-02, ...,\n",
       "            8.51962641e-02, -8.15733194e-01, -7.68503964e-01],\n",
       "          [-5.77297668e+02, -1.08821741e+03,  5.07599562e-02, ...,\n",
       "           -9.70777206e+01, -1.23117188e+03, -1.45553027e+03],\n",
       "          [-4.13482269e+02, -5.14637695e+02, -5.99510781e-02, ...,\n",
       "           -5.36014824e+01, -9.93305298e+02, -9.22676453e+02]],\n",
       " \n",
       "         [[-3.85315887e+02, -5.95043884e+02, -2.44959116e-01, ...,\n",
       "           -6.93419800e+01, -1.02955273e+03, -9.49807617e+02],\n",
       "          [-2.75939655e+01, -3.37841072e+01,  7.46291503e-03, ...,\n",
       "           -1.79126084e+00, -5.14027977e+01, -5.44236336e+01],\n",
       "          [-3.02147007e+01, -4.89371338e+01, -1.66320652e-01, ...,\n",
       "           -2.69756556e+00, -6.29170761e+01, -6.13714066e+01],\n",
       "          ...,\n",
       "          [-4.18319762e-01, -3.25752050e-01, -1.67842343e-01, ...,\n",
       "           -1.00873604e-01, -7.72398889e-01, -1.13078976e+00],\n",
       "          [-6.32835754e+02, -7.16651672e+02, -1.26144648e-01, ...,\n",
       "           -1.12049454e+02, -1.32097815e+03, -1.24882983e+03],\n",
       "          [-3.84412567e+02, -6.08519775e+02, -8.85470435e-02, ...,\n",
       "           -4.77020683e+01, -7.30655334e+02, -8.98950378e+02]]],\n",
       " \n",
       " \n",
       "        [[[-3.03917786e+02, -5.53939575e+02, -1.58786401e-01, ...,\n",
       "           -5.14833946e+01, -9.24778503e+02, -9.73731567e+02],\n",
       "          [-1.75702934e+01, -2.67822266e+01, -2.67552614e-01, ...,\n",
       "           -2.22497630e+00, -3.97449913e+01, -4.67826462e+01],\n",
       "          [-2.85646381e+01, -4.24576187e+01, -1.32548243e-01, ...,\n",
       "           -5.83427334e+00, -5.91314926e+01, -8.27556305e+01],\n",
       "          ...,\n",
       "          [-3.36306840e-01, -3.13737541e-01, -1.67517930e-01, ...,\n",
       "            4.13783565e-02, -5.40408134e-01, -1.13074720e+00],\n",
       "          [-4.27653381e+02, -8.08354187e+02, -1.66185960e-01, ...,\n",
       "           -6.99731674e+01, -1.24487500e+03, -1.54647021e+03],\n",
       "          [-3.54879211e+02, -6.09904846e+02,  1.93516128e-02, ...,\n",
       "           -5.72432823e+01, -7.32370178e+02, -8.14983704e+02]],\n",
       " \n",
       "         [[-3.25064117e+02, -6.44207031e+02, -1.45424485e-01, ...,\n",
       "           -5.75973816e+01, -7.17967468e+02, -1.15150830e+03],\n",
       "          [-1.73290863e+01, -2.66329365e+01, -1.92645386e-01, ...,\n",
       "           -3.81849098e+00, -5.62637901e+01, -7.43254395e+01],\n",
       "          [-2.73535671e+01, -4.46650124e+01, -4.75865696e-03, ...,\n",
       "           -8.46049213e+00, -7.99718857e+01, -7.39430771e+01],\n",
       "          ...,\n",
       "          [-2.12726086e-01, -5.25602221e-01, -8.03613383e-03, ...,\n",
       "           -1.42945915e-01, -3.68483782e-01, -1.94401398e-01],\n",
       "          [-5.09765503e+02, -7.23191589e+02,  9.65468027e-03, ...,\n",
       "           -7.51648712e+01, -1.24037915e+03, -1.41104871e+03],\n",
       "          [-3.32474091e+02, -6.46094299e+02, -1.07747920e-01, ...,\n",
       "           -6.69576950e+01, -5.91851501e+02, -9.53442261e+02]],\n",
       " \n",
       "         [[-3.25341583e+02, -6.07153870e+02, -3.60321589e-02, ...,\n",
       "           -5.18207855e+01, -9.51418579e+02, -1.04029919e+03],\n",
       "          [-1.51682167e+01, -3.46855812e+01, -1.87440351e-01, ...,\n",
       "           -1.45035052e+00, -6.86003113e+01, -7.49288025e+01],\n",
       "          [-2.32275791e+01, -3.50748825e+01,  1.37906536e-01, ...,\n",
       "           -3.31764603e+00, -6.19974174e+01, -9.30674667e+01],\n",
       "          ...,\n",
       "          [-6.20315671e-02, -2.76184648e-01,  1.14865482e-01, ...,\n",
       "           -1.68577462e-01, -5.32782853e-01, -9.70415056e-01],\n",
       "          [-6.00519226e+02, -1.04804480e+03, -9.28775221e-02, ...,\n",
       "           -9.27041550e+01, -1.38963086e+03, -1.66985156e+03],\n",
       "          [-3.03254852e+02, -6.50247192e+02, -1.71251908e-01, ...,\n",
       "           -5.70224724e+01, -9.35898804e+02, -1.14407996e+03]]]],\n",
       "       dtype=float32),\n",
       " array([-8.35358262e-01, -1.40132213e+00, -1.70656911e-03, -6.99092401e-03,\n",
       "        -2.22204351e+00, -7.19435513e-02, -5.56532154e-03, -2.35522002e-01,\n",
       "        -2.72075248e+00, -8.19050988e+10,             nan, -4.71786968e-03,\n",
       "        -1.16757365e+11, -1.34086251e-01, -1.93487108e+00, -2.40090060e+00],\n",
       "       dtype=float32),\n",
       " array([[[[-6.50430769e-02, -1.04153550e+00, -7.04237976e+02, ...,\n",
       "           -3.04309235e+02, -1.01180941e-01, -2.07605250e-02],\n",
       "          [-1.14887267e-01,  4.69928198e-02, -2.89629730e+02, ...,\n",
       "           -1.37707825e+02,  4.26300131e-02, -1.15002252e-01],\n",
       "          [ 4.04739603e-02,  1.52761880e-02, -7.54779056e-02, ...,\n",
       "            5.68678509e-03, -9.22754779e-02,  1.40090529e-02],\n",
       "          ...,\n",
       "          [ 7.05936700e-02,  3.70620191e-02, -3.13285694e+01, ...,\n",
       "           -1.35378838e+01, -4.98686209e-02,  4.86757867e-02],\n",
       "          [-4.85438593e-02, -5.16144693e-01, -3.73018494e+02, ...,\n",
       "           -1.64914200e+02, -4.86465432e-02, -1.56348675e-01],\n",
       "          [-5.13584837e-02, -7.45128036e-01, -4.87596130e+02, ...,\n",
       "           -2.34351822e+02, -1.81782797e-01, -9.08584371e-02]],\n",
       " \n",
       "         [[ 5.44321202e-02, -1.26662150e-01, -6.73311646e+02, ...,\n",
       "           -2.87870911e+02, -4.85672802e-02, -1.59861967e-02],\n",
       "          [-1.18601710e-01, -6.80977046e-01, -2.98819611e+02, ...,\n",
       "           -1.25068375e+02, -1.21561155e-01, -9.49497372e-02],\n",
       "          [-6.60335720e-02,  2.62166709e-02,  1.33870676e-01, ...,\n",
       "            9.67956409e-02,  1.02286972e-02, -1.61109552e-01],\n",
       "          ...,\n",
       "          [ 5.93645051e-02, -1.76114887e-01, -2.99750824e+01, ...,\n",
       "           -1.23250542e+01,  4.44961376e-02,  5.43886088e-02],\n",
       "          [ 2.15059053e-02, -6.80143476e-01, -3.59496918e+02, ...,\n",
       "           -1.51977020e+02, -2.18537748e-01, -2.63596885e-02],\n",
       "          [-6.00159205e-02, -7.11226225e-01, -5.13063904e+02, ...,\n",
       "           -2.27850647e+02, -1.51112869e-01, -8.89308080e-02]],\n",
       " \n",
       "         [[-1.24175355e-01, -9.36852470e-02, -6.80073364e+02, ...,\n",
       "           -2.81342896e+02,  8.06663111e-02,  7.50062913e-02],\n",
       "          [-7.37408325e-02, -5.72016776e-01, -2.68571289e+02, ...,\n",
       "           -1.15600288e+02, -4.90847118e-02,  7.50046177e-03],\n",
       "          [-8.19031000e-02, -1.16333947e-01, -1.98681727e-02, ...,\n",
       "           -2.01946069e-02, -5.35051432e-03, -2.68379059e-02],\n",
       "          ...,\n",
       "          [ 5.30800633e-02, -6.63700923e-02, -2.36733761e+01, ...,\n",
       "           -1.08950243e+01, -9.08362716e-02,  2.30968948e-02],\n",
       "          [-1.11106327e-02, -6.64804205e-02, -3.46854889e+02, ...,\n",
       "           -1.53647247e+02, -2.26287916e-02, -1.87191904e-01],\n",
       "          [-6.90345094e-02, -7.95854032e-01, -5.13376343e+02, ...,\n",
       "           -2.18060410e+02, -2.15850368e-01, -7.09998012e-02]]],\n",
       " \n",
       " \n",
       "        [[[-4.72103469e-02, -1.10937083e+00, -7.16108337e+02, ...,\n",
       "           -3.65603821e+02,  2.87433085e-03, -2.47688033e-02],\n",
       "          [ 6.01289980e-02, -1.69421092e-01, -3.22819122e+02, ...,\n",
       "           -1.39635040e+02, -1.02869548e-01, -1.00370377e-01],\n",
       "          [-1.23653170e-02,  4.71953787e-02, -7.78933316e-02, ...,\n",
       "            5.77120781e-02,  1.41721228e-02,  1.71440635e-02],\n",
       "          ...,\n",
       "          [ 1.29889604e-02, -1.28070384e-01, -2.83806362e+01, ...,\n",
       "           -1.67371769e+01,  1.00660753e-02, -5.23221083e-02],\n",
       "          [ 6.80633560e-02, -6.76027685e-02, -3.88550995e+02, ...,\n",
       "           -1.91478653e+02, -5.95005676e-02, -1.17458306e-01],\n",
       "          [ 6.32956699e-02, -1.22574341e+00, -5.09014435e+02, ...,\n",
       "           -2.35940781e+02, -7.25239590e-02,  8.31072778e-03]],\n",
       " \n",
       "         [[-1.05213612e-01, -3.89186668e+00, -6.92817444e+02, ...,\n",
       "           -3.58330902e+02, -2.07147449e-02, -6.79278672e-02],\n",
       "          [-9.90279168e-02, -2.72232199e+00, -2.74012054e+02, ...,\n",
       "           -1.40850723e+02,  5.30477241e-02, -2.31152531e-02],\n",
       "          [ 4.62607183e-02, -8.73693228e-02, -7.50431865e-02, ...,\n",
       "           -6.35310933e-02, -1.00948811e-01,  5.40631227e-02],\n",
       "          ...,\n",
       "          [-1.37118567e-02, -5.16117275e-01, -2.26538715e+01, ...,\n",
       "           -1.43578167e+01,  5.51070161e-02, -1.21594742e-01],\n",
       "          [-1.80937618e-01, -3.01346207e+00, -3.59630493e+02, ...,\n",
       "           -1.64144608e+02, -1.94523796e-01, -2.89181590e-01],\n",
       "          [-7.80818537e-02, -2.54692149e+00, -5.55189941e+02, ...,\n",
       "           -2.38133774e+02,  4.23769373e-03,  1.02460328e-02]],\n",
       " \n",
       "         [[-2.66106967e-02, -2.16748619e+00, -7.30616211e+02, ...,\n",
       "           -2.86983398e+02, -2.16686334e-02,  4.45133522e-02],\n",
       "          [ 1.02488464e-02, -1.02467310e+00, -2.68968536e+02, ...,\n",
       "           -1.16016975e+02,  5.00066280e-02,  8.45401082e-03],\n",
       "          [-1.08819619e-01,  4.45083436e-03,  9.03552845e-02, ...,\n",
       "            5.42235859e-02, -9.13554728e-02,  3.32754962e-02],\n",
       "          ...,\n",
       "          [ 6.92063197e-02, -9.00996566e-01, -2.46421356e+01, ...,\n",
       "           -1.16175480e+01,  3.70943770e-02, -1.07602395e-01],\n",
       "          [-1.04817292e-02, -2.81982493e+00, -3.54808685e+02, ...,\n",
       "           -1.56446396e+02, -1.40651107e-01, -1.50032818e-01],\n",
       "          [-5.30497804e-02, -2.86649799e+00, -4.89015656e+02, ...,\n",
       "           -2.20424744e+02, -9.80157554e-02, -2.15513378e-01]]],\n",
       " \n",
       " \n",
       "        [[[-1.02090187e-01, -9.47383225e-01, -6.78770447e+02, ...,\n",
       "           -3.62683990e+02, -1.54123548e-02, -6.25326782e-02],\n",
       "          [ 9.11540911e-02, -4.26108450e-01, -2.96026184e+02, ...,\n",
       "           -1.36713058e+02, -6.10050792e-03,  2.47961376e-02],\n",
       "          [ 3.18548046e-02,  2.39159912e-02,  2.70988923e-02, ...,\n",
       "            1.16431579e-01,  3.60118248e-03,  1.82424970e-02],\n",
       "          ...,\n",
       "          [-2.55287308e-02,  2.70253140e-03, -2.87208023e+01, ...,\n",
       "           -1.28113928e+01,  9.65657737e-03, -8.48554969e-02],\n",
       "          [-3.94901820e-02, -1.16435587e+00, -3.75856018e+02, ...,\n",
       "           -1.83209290e+02, -2.95342710e-02, -1.11266943e-02],\n",
       "          [-1.00100756e-01, -1.55651665e+00, -5.20528748e+02, ...,\n",
       "           -2.50119080e+02, -1.67657107e-01, -6.03213646e-02]],\n",
       " \n",
       "         [[ 1.37222242e-02, -6.25589371e+00, -7.45746521e+02, ...,\n",
       "           -3.29678802e+02, -2.20734552e-02, -8.71039703e-02],\n",
       "          [-9.03457850e-02, -3.07745266e+00, -2.93643982e+02, ...,\n",
       "           -1.43831772e+02, -6.17476925e-02, -9.93375778e-02],\n",
       "          [-4.61532138e-02,  8.69913250e-02,  7.30054677e-02, ...,\n",
       "           -9.25470218e-02,  1.08884603e-01, -1.61040485e-01],\n",
       "          ...,\n",
       "          [ 2.90422197e-02, -3.94496441e-01, -2.38020611e+01, ...,\n",
       "           -1.12735415e+01,  1.00313358e-01, -5.25413193e-02],\n",
       "          [-2.12612405e-01, -3.92987609e+00, -3.49085022e+02, ...,\n",
       "           -1.68760681e+02, -1.03191836e-02, -1.54148027e-01],\n",
       "          [ 1.02956407e-01, -4.24289751e+00, -5.25658936e+02, ...,\n",
       "           -2.76520355e+02, -1.23765737e-01, -1.33562878e-01]],\n",
       " \n",
       "         [[-1.12967499e-01, -6.17565584e+00, -6.29004517e+02, ...,\n",
       "           -3.40535065e+02, -1.43687311e-03,  4.10543084e-02],\n",
       "          [-8.54719654e-02, -2.05357885e+00, -2.72210968e+02, ...,\n",
       "           -1.34428452e+02, -1.06730916e-01, -6.90035895e-02],\n",
       "          [ 3.17688063e-02,  5.29523045e-02, -3.12260222e-02, ...,\n",
       "           -1.44468741e-02, -2.09472794e-02, -3.47315334e-02],\n",
       "          ...,\n",
       "          [-6.05996400e-02, -9.47601870e-02, -2.39154263e+01, ...,\n",
       "           -1.00684052e+01,  1.06392823e-01, -4.70951721e-02],\n",
       "          [-1.31761640e-01, -1.92079306e+00, -3.46098022e+02, ...,\n",
       "           -1.50294983e+02,  6.40106276e-02, -1.15954965e-01],\n",
       "          [-2.86660064e-02, -3.78026628e+00, -5.25451233e+02, ...,\n",
       "           -2.49988037e+02, -6.11802153e-02, -9.77066979e-02]]]],\n",
       "       dtype=float32),\n",
       " array([-7.1289029e+10, -8.8492390e-03, -6.6473788e-01, -6.0671349e+10,\n",
       "        -4.2593059e+10, -4.4115758e-01, -1.3368952e-02, -1.4537679e+11,\n",
       "        -8.3903396e-01, -3.1113291e-01, -1.4128687e+10, -6.8530107e-01,\n",
       "        -6.8484231e+10, -1.8720309e-01, -5.7557458e-01, -2.7774524e-02,\n",
       "        -2.5168663e+10, -1.0238053e+00, -4.7611114e-02, -2.9654811e+10,\n",
       "        -1.4308255e+11, -1.4698394e-01, -2.4718814e-01, -1.9274100e+11,\n",
       "        -5.0421220e-01, -7.6887280e-02, -2.1529664e-01, -5.5798076e+13,\n",
       "                   nan, -3.2154655e-01, -6.8158312e+10, -1.5397143e+10],\n",
       "       dtype=float32),\n",
       " array([[[[-6.06922496e+08,             nan,  1.47072328e-02, ...,\n",
       "            2.04692129e-02, -5.54280192e+08, -1.82591296e+08],\n",
       "          [ 3.80966929e-03,             nan, -1.50944442e-01, ...,\n",
       "           -3.54047529e-02,  1.70240358e-01,  1.23014949e-01],\n",
       "          [-7.10832443e+01,             nan, -3.75745036e-02, ...,\n",
       "           -8.18161368e-02, -4.86520653e+01,  1.94488926e+01],\n",
       "          ...,\n",
       "          [-4.37879944e+01,             nan, -1.39412105e-01, ...,\n",
       "           -7.34908134e-02, -2.53988209e+01,  3.32941937e+00],\n",
       "          [-4.10618912e+08,             nan, -7.06835613e-02, ...,\n",
       "           -1.15477787e-02, -3.64195584e+08, -9.85052960e+07],\n",
       "          [-1.13531880e+07,             nan, -9.79545410e-04, ...,\n",
       "           -1.74594764e-02, -1.20648740e+07, -2.00870625e+06]],\n",
       " \n",
       "         [[-4.89877216e+08,             nan, -1.11423701e-01, ...,\n",
       "           -2.53614839e-02, -4.73677664e+08, -1.80059280e+08],\n",
       "          [ 1.86529621e-01,             nan, -2.17644483e-01, ...,\n",
       "           -1.28368065e-01,  2.64416248e-01,  6.48603812e-02],\n",
       "          [-1.15795593e+02,             nan, -1.24695390e-01, ...,\n",
       "            1.75678525e-02, -6.27226334e+01,  2.67573662e+01],\n",
       "          ...,\n",
       "          [-5.57690010e+01,             nan,  1.38922930e-02, ...,\n",
       "            1.47985062e-02, -2.60653343e+01,  1.47756348e+01],\n",
       "          [-3.24775264e+08,             nan, -1.94096565e-02, ...,\n",
       "            1.80881694e-02, -2.68916960e+08, -1.15645376e+08],\n",
       "          [-1.53073590e+07,             nan, -1.00873694e-01, ...,\n",
       "            1.90425813e-02, -1.97902620e+07, -3.05084750e+06]],\n",
       " \n",
       "         [[-4.53269920e+08,             nan, -1.14414297e-01, ...,\n",
       "            2.64747478e-02, -4.07702368e+08, -1.21605688e+08],\n",
       "          [ 1.21121533e-01,             nan, -1.87761456e-01, ...,\n",
       "            2.84587052e-02,  3.62931728e-01,  8.08992758e-02],\n",
       "          [-8.23135376e+01,             nan, -1.20246850e-01, ...,\n",
       "            3.32048931e-03, -5.75714493e+01,  3.70860901e+01],\n",
       "          ...,\n",
       "          [-3.41186905e+01,             nan, -3.57628465e-02, ...,\n",
       "           -1.00554377e-01, -2.78392563e+01,  1.95546761e+01],\n",
       "          [-3.32675872e+08,             nan,  6.28168210e-02, ...,\n",
       "            1.69679467e-02, -3.41037408e+08, -1.20636688e+08],\n",
       "          [-2.81871780e+07,             nan,  1.29375877e-02, ...,\n",
       "           -1.62243277e-01, -2.45973380e+07, -3.01952950e+06]]],\n",
       " \n",
       " \n",
       "        [[[-5.61966784e+08,             nan, -6.01569265e-02, ...,\n",
       "           -2.58645825e-02, -4.37900896e+08, -1.34720384e+08],\n",
       "          [ 1.48535043e-01,             nan, -2.15450391e-01, ...,\n",
       "            2.00301558e-02,  2.06471205e-01, -9.68778320e-03],\n",
       "          [-7.80872879e+01,             nan, -1.90523952e-01, ...,\n",
       "           -9.36243981e-02, -3.93182411e+01,  1.34561348e+01],\n",
       "          ...,\n",
       "          [-4.73690796e+01,             nan, -7.96808377e-02, ...,\n",
       "           -1.04474418e-01, -2.32161274e+01,  5.73029947e+00],\n",
       "          [-3.28549088e+08,             nan, -8.00817385e-02, ...,\n",
       "            9.10392310e-03, -3.13795232e+08, -1.20344080e+08],\n",
       "          [-1.85295400e+07,             nan, -7.19006360e-02, ...,\n",
       "           -2.58215815e-02, -1.28054030e+07, -4.79412900e+06]],\n",
       " \n",
       "         [[-4.54583712e+08,             nan, -1.38772532e-01, ...,\n",
       "           -2.99684983e-02, -4.15534720e+08, -1.90333952e+08],\n",
       "          [ 1.02163754e-01,             nan, -1.31403118e-01, ...,\n",
       "           -2.12097779e-01,  1.54665485e-02, -4.02633287e-02],\n",
       "          [-1.08024872e+02,             nan, -9.41598937e-02, ...,\n",
       "            7.17607960e-02, -6.91129761e+01,  5.24816895e+01],\n",
       "          ...,\n",
       "          [-4.25039253e+01,             nan, -5.37854321e-02, ...,\n",
       "           -7.38320723e-02, -2.22818050e+01,  1.71610413e+01],\n",
       "          [-3.50887808e+08,             nan, -6.63838536e-02, ...,\n",
       "            3.23743336e-02, -2.64798336e+08, -1.07807928e+08],\n",
       "          [-1.19622380e+07,             nan, -1.84575334e-01, ...,\n",
       "           -1.97497606e-02, -1.08396660e+07, -8.70158700e+06]],\n",
       " \n",
       "         [[-4.89830656e+08,             nan, -1.22534469e-01, ...,\n",
       "           -6.34239018e-02, -4.40765024e+08, -1.94332496e+08],\n",
       "          [ 1.27130717e-01,             nan, -1.59364015e-01, ...,\n",
       "            5.72875030e-02,  1.34272337e-01,  3.63500342e-02],\n",
       "          [-7.69405212e+01,             nan, -1.25509143e-01, ...,\n",
       "           -1.13592461e-01, -4.28064003e+01,  5.27508049e+01],\n",
       "          ...,\n",
       "          [-4.35043907e+01,             nan, -5.46233281e-02, ...,\n",
       "           -8.35600272e-02, -3.96749039e+01,  1.93748779e+01],\n",
       "          [-3.47656096e+08,             nan, -8.54106471e-02, ...,\n",
       "            3.91015522e-02, -2.94253792e+08, -1.18428776e+08],\n",
       "          [-3.20495260e+07,             nan, -1.36922106e-01, ...,\n",
       "            9.77157708e-03, -2.71529100e+07, -1.37286100e+07]]],\n",
       " \n",
       " \n",
       "        [[[-6.38572352e+08,             nan, -1.27158359e-01, ...,\n",
       "           -7.29267150e-02, -5.43160256e+08, -1.75438768e+08],\n",
       "          [ 1.02164090e-01,             nan, -1.82791829e-01, ...,\n",
       "           -1.01756237e-01,  2.55554989e-02,  1.95514038e-02],\n",
       "          [-7.17442780e+01,             nan, -8.44169855e-02, ...,\n",
       "            2.42251065e-02, -5.81829491e+01,  2.18276825e+01],\n",
       "          ...,\n",
       "          [-3.04492416e+01,             nan,  2.00841799e-02, ...,\n",
       "           -5.50139882e-02, -2.04654827e+01,  1.06597481e+01],\n",
       "          [-3.26898624e+08,             nan, -3.13168503e-02, ...,\n",
       "           -9.17524993e-02, -3.13063264e+08, -7.85374560e+07],\n",
       "          [-2.17253540e+07,             nan, -5.20428196e-02, ...,\n",
       "           -9.69418436e-02, -1.41617410e+07, -2.96964475e+06]],\n",
       " \n",
       "         [[-6.56485824e+08,             nan, -7.14172423e-02, ...,\n",
       "           -8.00549090e-02, -5.26057888e+08, -2.18992880e+08],\n",
       "          [ 2.03207925e-01,             nan, -2.09603682e-01, ...,\n",
       "           -7.86805674e-02,  8.53573978e-02,  1.61350101e-01],\n",
       "          [-7.85090485e+01,             nan, -1.62214741e-01, ...,\n",
       "            1.48343453e-02, -4.09546089e+01,  5.44762192e+01],\n",
       "          ...,\n",
       "          [-4.48887405e+01,             nan, -1.14362970e-01, ...,\n",
       "           -8.34402367e-02, -2.34652195e+01,  2.23728981e+01],\n",
       "          [-3.01127488e+08,             nan,  2.01661158e-02, ...,\n",
       "           -3.90119068e-02, -2.85942112e+08, -1.42239408e+08],\n",
       "          [-2.03631440e+07,             nan, -1.52868494e-01, ...,\n",
       "           -1.85385481e-01, -1.61492320e+07, -7.25506900e+06]],\n",
       " \n",
       "         [[-5.58779328e+08,             nan, -8.88953954e-02, ...,\n",
       "           -8.88040811e-02, -5.56713856e+08, -1.89645296e+08],\n",
       "          [ 2.19420537e-01,             nan, -1.31777674e-01, ...,\n",
       "           -9.41251665e-02,  6.04415638e-03,  1.08365178e-01],\n",
       "          [-1.12163147e+02,             nan, -2.61683408e-02, ...,\n",
       "           -5.13654687e-02, -6.07480011e+01,  3.87115707e+01],\n",
       "          ...,\n",
       "          [-5.28846283e+01,             nan, -7.47400820e-02, ...,\n",
       "            6.06409088e-02, -2.76997051e+01,  2.38714428e+01],\n",
       "          [-2.76394304e+08,             nan,  3.10632810e-02, ...,\n",
       "           -6.40572906e-02, -2.94422560e+08, -1.14461008e+08],\n",
       "          [-3.17068840e+07,             nan, -6.95930794e-02, ...,\n",
       "           -6.53197318e-02, -2.53312520e+07, -1.09731700e+07]]]],\n",
       "       dtype=float32),\n",
       " array([-1.6881117e+10,            nan, -4.5610443e-03, -2.8103857e+09,\n",
       "        -9.1296512e+09, -3.1414900e+05, -1.2078454e+09, -6.2044524e+09,\n",
       "        -8.6871040e+09, -6.1381750e+09, -9.4704454e+08, -2.3965998e+08,\n",
       "        -4.6135014e+09, -9.3762591e+09, -5.0520315e+09, -2.1501016e+04,\n",
       "        -1.8622779e+10, -4.1444639e+09, -3.6784061e+08, -1.8336966e+06,\n",
       "        -5.2063027e+09, -8.3731405e+08, -2.7190769e+09, -7.8268549e+09,\n",
       "        -3.0322368e+09, -2.7183878e+08, -8.9952604e+09, -2.3316198e+09,\n",
       "        -5.6585410e+06, -3.9501720e+06, -7.6761944e+05, -6.2706017e+09,\n",
       "        -2.5923356e+09, -1.3200748e+10, -2.2810002e+09, -2.3527250e+07,\n",
       "        -3.9982085e+09, -5.2650450e+09, -2.9171956e+07, -8.6482272e+07,\n",
       "        -1.6723720e+08,            nan, -1.2974345e+09, -6.2965601e+09,\n",
       "        -3.0421665e+06, -9.8654525e+09, -2.9516465e+09, -1.1698741e+10,\n",
       "        -8.8680141e+08, -8.8400036e+09, -2.0483101e+10, -4.9187430e+09,\n",
       "        -2.9253927e+10, -1.0766759e+10, -2.5845238e+07, -5.0992850e+09,\n",
       "        -9.7986816e+08, -7.0114451e+08, -9.6761764e+09, -3.5132073e+09,\n",
       "        -5.6930089e+09, -4.4391967e+09, -1.4600408e+10, -5.7487007e+09],\n",
       "       dtype=float32),\n",
       " array([[-1.2010428e-01,  1.7387300e+00, -1.6173575e+05, ...,\n",
       "          3.8340309e+00,  1.1090167e+00, -3.7311055e-02],\n",
       "        [ 3.1180870e-02,  6.3789561e-02,  4.4629619e-02, ...,\n",
       "         -3.6964803e+33, -6.2295785e+32,  1.8775310e-02],\n",
       "        [-1.5722707e-01, -6.1308935e-02, -5.7922140e-02, ...,\n",
       "         -1.9987298e-02, -2.9642543e-02, -2.2377748e-02],\n",
       "        ...,\n",
       "        [-3.1227463e-01, -1.0890225e+17, -2.7151257e-03, ...,\n",
       "          5.9980171e+14,  4.8289774e+14, -1.7090670e+13],\n",
       "        [-4.5128606e-02, -6.8535662e+00, -1.2225625e+05, ...,\n",
       "          2.7471812e+00,  1.0050371e+00, -3.2143105e-02],\n",
       "        [-3.2787091e-01, -2.8213584e-01, -2.4542275e+04, ...,\n",
       "          8.0733889e-01,  1.0069947e-01,  6.3752733e-02]], dtype=float32),\n",
       " array([-2.11526509e-02, -6.49765400e+07, -3.51713312e+05, -7.98400062e+05,\n",
       "        -5.53944385e+03, -1.78605542e+03,             nan, -2.28906864e+08,\n",
       "        -4.99288928e+08, -6.14746432e+08, -3.58973320e+04, -1.65364380e+03,\n",
       "        -4.43230112e+08, -3.80351200e+08, -2.42307141e+05, -6.77289531e+04,\n",
       "        -1.97615322e+23, -7.94955000e+05, -7.20977295e+02, -1.23573048e+08,\n",
       "        -3.35347656e+03, -5.12238592e+09, -1.71802560e+07, -1.41444992e+08,\n",
       "        -4.73742000e+05, -1.42673156e+05, -3.65534219e+04, -1.54941177e+22,\n",
       "        -1.92339734e+05, -1.62839734e+03, -7.14732715e+03, -4.07350781e+04,\n",
       "        -7.02271719e+04, -9.05726500e+05, -7.56339776e+08, -7.30261517e+22,\n",
       "        -8.91095500e+06, -1.42773608e+21, -5.68344287e+03, -2.18648438e+03,\n",
       "        -4.09976218e+09, -9.39435120e+22, -1.06503008e+04, -1.60245219e+05,\n",
       "        -1.79756660e+07, -2.58542559e+04, -3.39403040e+08, -8.11202752e+08,\n",
       "        -3.16556934e+03,             nan, -7.84866500e+05, -1.59262612e+06,\n",
       "        -1.02533660e+07, -2.37458936e+03, -2.38910398e+22, -9.87958410e-05,\n",
       "        -2.58220618e+22, -2.18256934e+04,             nan, -6.07951611e+03,\n",
       "        -6.43147878e+09, -2.00308740e+22, -9.55610068e+22, -1.31322412e+04,\n",
       "        -7.61362160e+07, -2.59440527e+04, -3.32479900e+23, -3.50497719e+05,\n",
       "        -1.77718292e+10, -1.07944388e+06, -9.42966000e+07, -8.94231500e+06,\n",
       "        -1.49181920e+08, -3.02146500e+07, -2.73142656e+04, -5.26625384e-03,\n",
       "        -2.39755312e+08, -8.41615770e+09, -1.33201050e+07, -9.77066938e+05,\n",
       "        -5.42526688e+05, -1.52568355e+22, -6.89985184e+21, -4.80408242e+04,\n",
       "        -2.15537391e+05, -5.08479961e+04, -6.70920312e+04, -1.89646320e+07,\n",
       "        -5.90274760e+07, -4.20058438e+04, -3.01658867e+23, -2.40008620e+07,\n",
       "        -1.22329790e+22, -4.81856656e+05, -4.67479600e+07, -6.16916800e+07,\n",
       "        -3.62337594e+05,             nan, -9.14931935e+22, -4.47847026e+23,\n",
       "        -7.36420938e+04, -2.48323440e+07, -2.56398112e+08, -5.88268880e+07,\n",
       "        -1.35311456e-02, -4.95956934e+03, -2.83535820e+07, -4.83140670e-04,\n",
       "        -4.40351133e+04, -9.80233312e+05, -7.76619520e+08, -5.87708789e+03,\n",
       "        -8.49273536e+08, -1.29441492e+22, -4.35593568e+08, -4.40579219e+04,\n",
       "        -5.76794379e+22, -3.91156000e+08, -3.12101469e+05, -4.00239320e+07,\n",
       "        -2.36023900e+07,             nan, -3.08469280e+08, -6.90658312e+05,\n",
       "        -1.02307758e+05, -6.98237125e+05,             nan, -1.40926784e+09,\n",
       "         0.00000000e+00, -1.99060352e+04, -7.31720128e+08, -8.59659607e+02,\n",
       "        -7.20404950e+06,             nan, -1.38326802e+23, -4.28245037e+22,\n",
       "        -9.70681840e+07, -1.50594432e+08, -2.00537190e+10, -6.29714074e+09,\n",
       "        -1.12907030e+07, -4.08306040e+07, -1.83891855e+04, -1.09662086e+05,\n",
       "        -1.77432559e+04,             nan, -1.50784096e+08, -4.63832064e+08,\n",
       "        -2.15042304e+08, -6.04359688e+04, -1.97653344e+05, -2.50016720e+08,\n",
       "                    nan, -6.12169000e+05, -1.65581016e+04, -1.69686859e+05,\n",
       "        -7.02381360e+07, -5.64367891e+04, -3.01666528e+03, -3.72978272e+08,\n",
       "        -6.08552198e+19, -5.16584062e+04, -5.92515648e+08, -1.17240909e+09,\n",
       "        -3.90195936e+08, -1.61292109e+05, -1.06158714e+09, -2.52746700e+07,\n",
       "        -7.84609921e+21, -1.84307695e+04, -3.92300312e+05, -5.79403200e+08,\n",
       "        -6.84323264e+08, -6.06918938e+05, -9.79085520e+07, -3.93175633e+23,\n",
       "        -1.38954625e+06, -2.89275315e+09,             nan, -3.54646298e+09,\n",
       "                    nan, -8.15351360e+07, -1.47717627e+04, -1.06859053e+09,\n",
       "        -1.68656858e+09, -1.95126997e+21, -2.94070335e+21, -2.46515050e+06,\n",
       "        -7.98096880e+07, -2.84857158e+22, -2.09648750e+04, -5.14140280e+07,\n",
       "                    nan, -3.17721408e+08, -1.93396378e+01, -4.20190421e+23,\n",
       "        -1.12482764e+04, -5.41160244e+21, -1.96776949e+23, -1.64372781e+05,\n",
       "        -1.04705625e+05, -7.20587891e+04, -1.54902480e+08, -8.23394880e+08,\n",
       "        -3.58819008e+08, -2.78413066e+04, -6.27160339e+02, -1.15778931e+09,\n",
       "        -3.73753491e+03, -1.23365369e+03, -9.54761229e+22, -1.18074945e+05,\n",
       "        -1.52025775e+06, -1.39771718e+23, -7.66649026e+22, -1.42446672e+05,\n",
       "        -5.11248040e+07, -2.01475231e+23, -1.64846266e+23, -9.34520454e+21,\n",
       "        -1.36284922e+05, -1.06131072e+23, -5.68064648e+03, -3.19992070e+04,\n",
       "        -3.02302046e-03,             nan, -3.61056281e+05, -2.06687503e+22,\n",
       "        -7.97023850e+06, -5.99297560e+07, -4.66839129e+22, -1.11599007e+22,\n",
       "        -2.99345352e+04, -7.84894836e+02, -7.32424145e+22,             nan,\n",
       "        -2.68681421e+23, -2.52649422e+05, -4.00405440e+07, -2.36369141e+05,\n",
       "        -1.12695824e+08, -2.24946812e-03, -5.29700947e+21, -5.59878836e+21,\n",
       "        -1.00025446e+09, -6.50770938e+04, -2.52306896e+22, -1.02896019e+09,\n",
       "                    nan, -9.21807680e+07, -3.83453670e+21, -2.15160801e+22,\n",
       "        -7.60149562e+05, -2.84069374e+23, -4.81189489e+22, -6.89224707e+03],\n",
       "       dtype=float32),\n",
       " array([[-3.67062688e-02,  5.99578116e-03,  5.24537414e-02, ...,\n",
       "         -1.43209789e-02,             nan, -5.66571392e-02],\n",
       "        [-1.96799850e+02, -2.40519180e+02, -4.05002682e+11, ...,\n",
       "         -1.34788418e+01,             nan, -6.06656723e+01],\n",
       "        [-2.07827637e+02, -2.63946045e+02, -2.89787984e+00, ...,\n",
       "         -1.21178322e+01,             nan, -6.25166092e+01],\n",
       "        ...,\n",
       "        [-4.19966899e+28, -4.49248404e+28, -5.88178575e+12, ...,\n",
       "         -4.19635359e+27,             nan, -1.16617317e+28],\n",
       "        [-1.28029719e+28, -1.36956384e+28, -4.44322913e+12, ...,\n",
       "         -1.27928642e+27,             nan, -3.55515657e+27],\n",
       "        [ 1.12599581e-01, -5.07779494e-02, -1.33736825e+11, ...,\n",
       "          7.91578218e-02,             nan, -5.76188788e-02]], dtype=float32),\n",
       " array([-5.44788322e+01, -5.82794800e+01, -8.76807421e-02, -5.29690312e+04,\n",
       "        -5.37748003e+00, -1.67366465e+04, -1.00620250e+05, -5.04894066e+00,\n",
       "        -5.10339998e-03,             nan, -1.71208687e+01,             nan,\n",
       "        -9.74311133e+03, -7.34342728e+01, -7.49869406e-01, -3.49221458e+01,\n",
       "        -3.21550255e+01, -2.69992461e+04, -2.65964556e+00, -4.53859024e+01,\n",
       "        -1.31844757e-02, -4.36904053e+03,             nan,             nan,\n",
       "                    nan,             nan, -1.72732620e+01,             nan,\n",
       "        -2.82561550e+01, -2.84211953e+04,             nan, -3.51760197e+00,\n",
       "        -1.55003770e+04, -2.57936406e+00, -2.11355880e-01, -2.81963069e+10,\n",
       "        -9.54508359e+04, -1.22869940e+01, -8.53591156e+00, -1.40910749e+01,\n",
       "        -1.91195612e-03, -3.13098516e+04, -4.04290104e+00, -3.91062227e+04,\n",
       "        -4.33229990e-02, -8.74700625e+04, -6.05979336e+04,             nan,\n",
       "        -1.78467051e+04,             nan,             nan, -7.72720234e+04,\n",
       "        -1.67523635e+10, -6.35464907e-01, -3.34235229e+02, -8.80789105e-03,\n",
       "        -8.28583241e-02,             nan,             nan,             nan,\n",
       "        -4.76438141e+01, -1.11241625e+05,             nan, -2.06353809e+11,\n",
       "        -5.01592891e+04, -7.78655000e+04,             nan, -3.63548901e+03,\n",
       "        -5.12617070e+04, -2.38073444e+01,             nan, -1.22768643e+04,\n",
       "        -5.55446510e+01, -1.98381922e+05, -8.16662064e+01, -7.74038544e+01,\n",
       "        -7.74665222e+01, -4.81617422e+04,             nan, -1.27446602e+04,\n",
       "        -1.23230869e+04, -1.77761844e+05,             nan, -2.51909902e+04,\n",
       "        -7.91967725e+03, -8.88826561e+00,             nan, -9.14852829e+01,\n",
       "                    nan,             nan, -6.02574229e-01,             nan,\n",
       "        -1.01434579e-03, -2.38660736e+01,             nan, -9.22724139e-03,\n",
       "                    nan, -1.94845724e+00, -1.98701266e-02,             nan,\n",
       "        -3.37701797e+01, -1.45842161e+01, -3.98659945e+00, -2.73662186e+01,\n",
       "        -3.85582042e+00, -2.99644432e+01, -5.10087585e+01, -3.05983019e+00,\n",
       "        -1.88410234e+00, -1.15510349e+01, -4.74041271e+00, -4.36739909e-04,\n",
       "        -5.28644421e-04,             nan, -1.20697852e+04,             nan,\n",
       "        -6.22428945e+04, -5.14187559e-04, -4.72947629e-03, -1.70069540e+00,\n",
       "                    nan, -2.58221948e+03, -6.70104294e+01, -9.52769375e+04,\n",
       "        -4.73472942e-03, -8.39509106e+00,             nan,             nan,\n",
       "        -5.04600754e+01,             nan, -1.30765200e+00, -6.33023529e+01,\n",
       "        -3.63432360e+00, -2.04939766e+01, -1.72989845e+01, -8.25119609e+04,\n",
       "                    nan, -8.63877344e+04, -3.31964219e+04,             nan,\n",
       "        -2.88971138e+01,             nan, -9.97033808e-03, -6.34772415e+01,\n",
       "        -1.24690622e-01,             nan, -8.09091719e+04, -3.04159180e+04,\n",
       "        -4.32565391e-01, -1.33737278e+01, -1.38370400e+04, -6.30973434e+01,\n",
       "        -3.05074234e+01,             nan, -2.61516382e+03, -7.22893906e+01,\n",
       "        -5.81679344e+01, -5.26660495e-03, -8.49749279e+00, -3.79927216e+01,\n",
       "        -5.14570742e+04, -7.25889301e+00, -7.61007385e+01, -1.43360031e+02,\n",
       "                    nan,             nan, -4.10400000e+04, -3.01985693e+03,\n",
       "        -6.48684741e-04, -1.24867195e+05,             nan, -4.98813689e-01,\n",
       "        -1.18371410e+01, -8.98457039e-03, -4.41328110e-03,             nan,\n",
       "        -2.62129242e+02,             nan, -2.45664969e+05, -2.24874973e-01,\n",
       "        -8.80088574e+03, -7.07927422e+04,             nan, -7.32799606e+01,\n",
       "        -2.12450809e+01, -2.52560664e+04, -5.44437408e+01, -4.10517070e+04,\n",
       "        -1.09456154e-03,             nan, -8.63533467e-04, -2.85887988e+04,\n",
       "                    nan,             nan, -1.10660961e+05, -6.34414470e-03,\n",
       "        -3.25766487e+01, -5.00550365e+00, -2.55623885e+10, -4.87670097e+01,\n",
       "        -2.29299746e+04, -7.01532568e+03, -7.46947145e+00,             nan,\n",
       "        -5.75833740e+03, -9.85547578e+04, -1.34764075e+00, -1.27581663e+01,\n",
       "                    nan, -5.61383903e-01, -1.92603594e+05,             nan,\n",
       "        -3.17410352e+04, -9.12616216e-03, -1.91893137e+00,             nan,\n",
       "        -2.87856172e+04,             nan, -3.68433189e+01, -1.84749050e+01,\n",
       "        -3.31189270e+01,             nan, -1.59503857e-03,             nan,\n",
       "        -9.70877380e+01, -1.58798254e+00, -1.60458613e+04, -1.43297480e+04,\n",
       "        -1.63922329e+01,             nan, -1.55264629e+04, -2.29327183e-04,\n",
       "        -1.22785702e+01, -3.43075581e-02,             nan, -3.67844653e+00,\n",
       "                    nan,             nan, -1.89411203e+05, -4.51041651e+00,\n",
       "        -9.28878203e+04, -2.14344165e+03, -3.76720963e+01, -2.19923711e+00,\n",
       "        -2.56865254e+04, -4.56495857e+00,             nan, -1.03754902e+01,\n",
       "        -7.68358470e-04,             nan,             nan, -5.22119971e-03,\n",
       "        -1.38389807e+03, -5.44255066e+00,             nan, -1.51279917e+01],\n",
       "       dtype=float32),\n",
       " array([[nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan],\n",
       "        [nan]], dtype=float32),\n",
       " array([nan], dtype=float32)]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model_unnormalized_data.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1000\n",
      "582/582 [==============================] - 22s 37ms/step - loss: 1.8968 - accuracy: 0.5027 - val_loss: 0.7111 - val_accuracy: 0.5277\n",
      "Epoch 2/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 1.0238 - accuracy: 0.5083 - val_loss: 0.6913 - val_accuracy: 0.5430\n",
      "Epoch 3/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.8662 - accuracy: 0.5168 - val_loss: 0.6889 - val_accuracy: 0.5447\n",
      "Epoch 4/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.8057 - accuracy: 0.5166 - val_loss: 0.6870 - val_accuracy: 0.5514\n",
      "Epoch 5/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7701 - accuracy: 0.5189 - val_loss: 0.6856 - val_accuracy: 0.5591\n",
      "Epoch 6/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7483 - accuracy: 0.5223 - val_loss: 0.6855 - val_accuracy: 0.5626\n",
      "Epoch 7/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7393 - accuracy: 0.5213 - val_loss: 0.6855 - val_accuracy: 0.5632\n",
      "Epoch 8/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7305 - accuracy: 0.5202 - val_loss: 0.6858 - val_accuracy: 0.5686\n",
      "Epoch 9/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7180 - accuracy: 0.5296 - val_loss: 0.6861 - val_accuracy: 0.5649\n",
      "Epoch 10/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7125 - accuracy: 0.5336 - val_loss: 0.6861 - val_accuracy: 0.5632\n",
      "Epoch 11/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7163 - accuracy: 0.5208 - val_loss: 0.6862 - val_accuracy: 0.5619\n",
      "Epoch 12/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7096 - accuracy: 0.5298 - val_loss: 0.6863 - val_accuracy: 0.5610\n",
      "Epoch 13/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7037 - accuracy: 0.5339 - val_loss: 0.6862 - val_accuracy: 0.5615\n",
      "Epoch 14/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7021 - accuracy: 0.5323 - val_loss: 0.6863 - val_accuracy: 0.5634\n",
      "Epoch 15/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7013 - accuracy: 0.5322 - val_loss: 0.6862 - val_accuracy: 0.5664\n",
      "Epoch 16/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.7005 - accuracy: 0.5327 - val_loss: 0.6862 - val_accuracy: 0.5677\n",
      "Epoch 17/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6976 - accuracy: 0.5389 - val_loss: 0.6862 - val_accuracy: 0.5716\n",
      "Epoch 18/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6959 - accuracy: 0.5360 - val_loss: 0.6861 - val_accuracy: 0.5724\n",
      "Epoch 19/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6958 - accuracy: 0.5387 - val_loss: 0.6860 - val_accuracy: 0.5727\n",
      "Epoch 20/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6942 - accuracy: 0.5418 - val_loss: 0.6859 - val_accuracy: 0.5739\n",
      "Epoch 21/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6934 - accuracy: 0.5403 - val_loss: 0.6859 - val_accuracy: 0.5755\n",
      "Epoch 22/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6940 - accuracy: 0.5359 - val_loss: 0.6858 - val_accuracy: 0.5755\n",
      "Epoch 23/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6916 - accuracy: 0.5426 - val_loss: 0.6858 - val_accuracy: 0.5770\n",
      "Epoch 24/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6937 - accuracy: 0.5430 - val_loss: 0.6858 - val_accuracy: 0.5765\n",
      "Epoch 25/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6909 - accuracy: 0.5401 - val_loss: 0.6857 - val_accuracy: 0.5782\n",
      "Epoch 26/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6911 - accuracy: 0.5421 - val_loss: 0.6855 - val_accuracy: 0.5782\n",
      "Epoch 27/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6914 - accuracy: 0.5390 - val_loss: 0.6854 - val_accuracy: 0.5808\n",
      "Epoch 28/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6921 - accuracy: 0.5376 - val_loss: 0.6854 - val_accuracy: 0.5817\n",
      "Epoch 29/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6912 - accuracy: 0.5416 - val_loss: 0.6853 - val_accuracy: 0.5815\n",
      "Epoch 30/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6905 - accuracy: 0.5409 - val_loss: 0.6852 - val_accuracy: 0.5845\n",
      "Epoch 31/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6918 - accuracy: 0.5357 - val_loss: 0.6852 - val_accuracy: 0.5856\n",
      "Epoch 32/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6909 - accuracy: 0.5378 - val_loss: 0.6852 - val_accuracy: 0.5856\n",
      "Epoch 33/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6900 - accuracy: 0.5458 - val_loss: 0.6849 - val_accuracy: 0.5881\n",
      "Epoch 34/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6911 - accuracy: 0.5383 - val_loss: 0.6849 - val_accuracy: 0.5881\n",
      "Epoch 35/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6891 - accuracy: 0.5469 - val_loss: 0.6848 - val_accuracy: 0.5864\n",
      "Epoch 36/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6890 - accuracy: 0.5443 - val_loss: 0.6848 - val_accuracy: 0.5892\n",
      "Epoch 37/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6886 - accuracy: 0.5431 - val_loss: 0.6847 - val_accuracy: 0.5873\n",
      "Epoch 38/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6898 - accuracy: 0.5401 - val_loss: 0.6845 - val_accuracy: 0.5911\n",
      "Epoch 39/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6866 - accuracy: 0.5503 - val_loss: 0.6844 - val_accuracy: 0.5916\n",
      "Epoch 40/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6885 - accuracy: 0.5495 - val_loss: 0.6843 - val_accuracy: 0.5929\n",
      "Epoch 41/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6889 - accuracy: 0.5420 - val_loss: 0.6843 - val_accuracy: 0.5931\n",
      "Epoch 42/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6878 - accuracy: 0.5517 - val_loss: 0.6841 - val_accuracy: 0.5959\n",
      "Epoch 43/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6877 - accuracy: 0.5498 - val_loss: 0.6841 - val_accuracy: 0.5980\n",
      "Epoch 44/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6870 - accuracy: 0.5498 - val_loss: 0.6840 - val_accuracy: 0.5978\n",
      "Epoch 45/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6883 - accuracy: 0.5447 - val_loss: 0.6839 - val_accuracy: 0.5982\n",
      "Epoch 46/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6874 - accuracy: 0.5490 - val_loss: 0.6838 - val_accuracy: 0.5989\n",
      "Epoch 47/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6855 - accuracy: 0.5493 - val_loss: 0.6837 - val_accuracy: 0.5980\n",
      "Epoch 48/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6867 - accuracy: 0.5521 - val_loss: 0.6835 - val_accuracy: 0.5991\n",
      "Epoch 49/1000\n",
      "582/582 [==============================] - 19s 33ms/step - loss: 0.6859 - accuracy: 0.5500 - val_loss: 0.6835 - val_accuracy: 0.6002\n",
      "Epoch 50/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6872 - accuracy: 0.5493 - val_loss: 0.6834 - val_accuracy: 0.5989\n",
      "Epoch 51/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6850 - accuracy: 0.5523 - val_loss: 0.6832 - val_accuracy: 0.6019\n",
      "Epoch 52/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6858 - accuracy: 0.5530 - val_loss: 0.6830 - val_accuracy: 0.6017\n",
      "Epoch 53/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6853 - accuracy: 0.5553 - val_loss: 0.6829 - val_accuracy: 0.6032\n",
      "Epoch 54/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6862 - accuracy: 0.5483 - val_loss: 0.6827 - val_accuracy: 0.6066\n",
      "Epoch 55/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6856 - accuracy: 0.5466 - val_loss: 0.6827 - val_accuracy: 0.6038\n",
      "Epoch 56/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6844 - accuracy: 0.5597 - val_loss: 0.6826 - val_accuracy: 0.6017\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 57/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6852 - accuracy: 0.5542 - val_loss: 0.6825 - val_accuracy: 0.6038\n",
      "Epoch 58/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6868 - accuracy: 0.5531 - val_loss: 0.6824 - val_accuracy: 0.6049\n",
      "Epoch 59/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6840 - accuracy: 0.5544 - val_loss: 0.6822 - val_accuracy: 0.6043\n",
      "Epoch 60/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6836 - accuracy: 0.5512 - val_loss: 0.6821 - val_accuracy: 0.6043\n",
      "Epoch 61/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6853 - accuracy: 0.5557 - val_loss: 0.6821 - val_accuracy: 0.6030\n",
      "Epoch 62/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6840 - accuracy: 0.5569 - val_loss: 0.6820 - val_accuracy: 0.6019\n",
      "Epoch 63/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6829 - accuracy: 0.5614 - val_loss: 0.6819 - val_accuracy: 0.6015\n",
      "Epoch 64/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6848 - accuracy: 0.5524 - val_loss: 0.6818 - val_accuracy: 0.6008\n",
      "Epoch 65/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6840 - accuracy: 0.5567 - val_loss: 0.6817 - val_accuracy: 0.6010\n",
      "Epoch 66/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6827 - accuracy: 0.5587 - val_loss: 0.6817 - val_accuracy: 0.6028\n",
      "Epoch 67/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6834 - accuracy: 0.5605 - val_loss: 0.6816 - val_accuracy: 0.6023\n",
      "Epoch 68/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6846 - accuracy: 0.5554 - val_loss: 0.6815 - val_accuracy: 0.6025\n",
      "Epoch 69/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6836 - accuracy: 0.5596 - val_loss: 0.6814 - val_accuracy: 0.6030\n",
      "Epoch 70/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6825 - accuracy: 0.5614 - val_loss: 0.6813 - val_accuracy: 0.6030\n",
      "Epoch 71/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6838 - accuracy: 0.5603 - val_loss: 0.6812 - val_accuracy: 0.6028\n",
      "Epoch 72/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6825 - accuracy: 0.5621 - val_loss: 0.6811 - val_accuracy: 0.6025\n",
      "Epoch 73/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6828 - accuracy: 0.5644 - val_loss: 0.6810 - val_accuracy: 0.6036\n",
      "Epoch 74/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6828 - accuracy: 0.5627 - val_loss: 0.6809 - val_accuracy: 0.6040\n",
      "Epoch 75/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6814 - accuracy: 0.5650 - val_loss: 0.6807 - val_accuracy: 0.6038\n",
      "Epoch 76/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6831 - accuracy: 0.5610 - val_loss: 0.6806 - val_accuracy: 0.6030\n",
      "Epoch 77/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6830 - accuracy: 0.5627 - val_loss: 0.6805 - val_accuracy: 0.6036\n",
      "Epoch 78/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6790 - accuracy: 0.5723 - val_loss: 0.6803 - val_accuracy: 0.6036\n",
      "Epoch 79/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6808 - accuracy: 0.5647 - val_loss: 0.6802 - val_accuracy: 0.6043\n",
      "Epoch 80/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6822 - accuracy: 0.5660 - val_loss: 0.6801 - val_accuracy: 0.6025\n",
      "Epoch 81/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6820 - accuracy: 0.5645 - val_loss: 0.6801 - val_accuracy: 0.6036\n",
      "Epoch 82/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6826 - accuracy: 0.5644 - val_loss: 0.6800 - val_accuracy: 0.6038\n",
      "Epoch 83/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6817 - accuracy: 0.5621 - val_loss: 0.6800 - val_accuracy: 0.6030\n",
      "Epoch 84/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6815 - accuracy: 0.5672 - val_loss: 0.6799 - val_accuracy: 0.6030\n",
      "Epoch 85/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6811 - accuracy: 0.5697 - val_loss: 0.6798 - val_accuracy: 0.6034\n",
      "Epoch 86/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6802 - accuracy: 0.5656 - val_loss: 0.6796 - val_accuracy: 0.6028\n",
      "Epoch 87/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6802 - accuracy: 0.5676 - val_loss: 0.6794 - val_accuracy: 0.6023\n",
      "Epoch 88/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6807 - accuracy: 0.5682 - val_loss: 0.6793 - val_accuracy: 0.6032\n",
      "Epoch 89/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6788 - accuracy: 0.5715 - val_loss: 0.6792 - val_accuracy: 0.6030\n",
      "Epoch 90/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6819 - accuracy: 0.5681 - val_loss: 0.6793 - val_accuracy: 0.6025\n",
      "Epoch 91/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6810 - accuracy: 0.5708 - val_loss: 0.6792 - val_accuracy: 0.6034\n",
      "Epoch 92/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6789 - accuracy: 0.5739 - val_loss: 0.6789 - val_accuracy: 0.6040\n",
      "Epoch 93/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6800 - accuracy: 0.5697 - val_loss: 0.6788 - val_accuracy: 0.6049\n",
      "Epoch 94/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6801 - accuracy: 0.5701 - val_loss: 0.6788 - val_accuracy: 0.6055\n",
      "Epoch 95/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6803 - accuracy: 0.5692 - val_loss: 0.6788 - val_accuracy: 0.6051\n",
      "Epoch 96/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6810 - accuracy: 0.5720 - val_loss: 0.6787 - val_accuracy: 0.6053\n",
      "Epoch 97/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6795 - accuracy: 0.5697 - val_loss: 0.6786 - val_accuracy: 0.6062\n",
      "Epoch 98/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6775 - accuracy: 0.5804 - val_loss: 0.6783 - val_accuracy: 0.6071\n",
      "Epoch 99/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6786 - accuracy: 0.5749 - val_loss: 0.6782 - val_accuracy: 0.6060\n",
      "Epoch 100/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6800 - accuracy: 0.5681 - val_loss: 0.6782 - val_accuracy: 0.6058\n",
      "Epoch 101/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6788 - accuracy: 0.5742 - val_loss: 0.6781 - val_accuracy: 0.6045\n",
      "Epoch 102/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6776 - accuracy: 0.5738 - val_loss: 0.6781 - val_accuracy: 0.6047\n",
      "Epoch 103/1000\n",
      "582/582 [==============================] - 19s 33ms/step - loss: 0.6789 - accuracy: 0.5730 - val_loss: 0.6781 - val_accuracy: 0.6043\n",
      "Epoch 104/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6777 - accuracy: 0.5786 - val_loss: 0.6779 - val_accuracy: 0.6032\n",
      "Epoch 105/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6770 - accuracy: 0.5789 - val_loss: 0.6777 - val_accuracy: 0.6034\n",
      "Epoch 106/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6771 - accuracy: 0.5792 - val_loss: 0.6775 - val_accuracy: 0.6040\n",
      "Epoch 107/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6785 - accuracy: 0.5753 - val_loss: 0.6774 - val_accuracy: 0.6049\n",
      "Epoch 108/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6795 - accuracy: 0.5769 - val_loss: 0.6773 - val_accuracy: 0.6034\n",
      "Epoch 109/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6763 - accuracy: 0.5793 - val_loss: 0.6771 - val_accuracy: 0.6062\n",
      "Epoch 110/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6766 - accuracy: 0.5818 - val_loss: 0.6770 - val_accuracy: 0.6060\n",
      "Epoch 111/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6779 - accuracy: 0.5779 - val_loss: 0.6769 - val_accuracy: 0.6049\n",
      "Epoch 112/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6778 - accuracy: 0.5742 - val_loss: 0.6767 - val_accuracy: 0.6047\n",
      "Epoch 113/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6774 - accuracy: 0.5801 - val_loss: 0.6768 - val_accuracy: 0.6043\n",
      "Epoch 114/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6765 - accuracy: 0.5795 - val_loss: 0.6767 - val_accuracy: 0.6038\n",
      "Epoch 115/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6756 - accuracy: 0.5787 - val_loss: 0.6766 - val_accuracy: 0.6045\n",
      "Epoch 116/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6761 - accuracy: 0.5845 - val_loss: 0.6765 - val_accuracy: 0.6045\n",
      "Epoch 117/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6764 - accuracy: 0.5803 - val_loss: 0.6763 - val_accuracy: 0.6036\n",
      "Epoch 118/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6778 - accuracy: 0.5797 - val_loss: 0.6762 - val_accuracy: 0.6047\n",
      "Epoch 119/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6768 - accuracy: 0.5847 - val_loss: 0.6762 - val_accuracy: 0.6068\n",
      "Epoch 120/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6782 - accuracy: 0.5812 - val_loss: 0.6762 - val_accuracy: 0.6064\n",
      "Epoch 121/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6773 - accuracy: 0.5811 - val_loss: 0.6761 - val_accuracy: 0.6055\n",
      "Epoch 122/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6760 - accuracy: 0.5769 - val_loss: 0.6759 - val_accuracy: 0.6055\n",
      "Epoch 123/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6753 - accuracy: 0.5801 - val_loss: 0.6758 - val_accuracy: 0.6049\n",
      "Epoch 124/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6767 - accuracy: 0.5805 - val_loss: 0.6757 - val_accuracy: 0.6062\n",
      "Epoch 125/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6771 - accuracy: 0.5803 - val_loss: 0.6756 - val_accuracy: 0.6075\n",
      "Epoch 126/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6750 - accuracy: 0.5874 - val_loss: 0.6755 - val_accuracy: 0.6068\n",
      "Epoch 127/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6769 - accuracy: 0.5803 - val_loss: 0.6754 - val_accuracy: 0.6090\n",
      "Epoch 128/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6742 - accuracy: 0.5840 - val_loss: 0.6752 - val_accuracy: 0.6092\n",
      "Epoch 129/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6763 - accuracy: 0.5823 - val_loss: 0.6751 - val_accuracy: 0.6103\n",
      "Epoch 130/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6760 - accuracy: 0.5814 - val_loss: 0.6750 - val_accuracy: 0.6096\n",
      "Epoch 131/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6762 - accuracy: 0.5812 - val_loss: 0.6750 - val_accuracy: 0.6081\n",
      "Epoch 132/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6762 - accuracy: 0.5831 - val_loss: 0.6751 - val_accuracy: 0.6079\n",
      "Epoch 133/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6754 - accuracy: 0.5821 - val_loss: 0.6750 - val_accuracy: 0.6088\n",
      "Epoch 134/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6763 - accuracy: 0.5848 - val_loss: 0.6750 - val_accuracy: 0.6092\n",
      "Epoch 135/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6748 - accuracy: 0.5868 - val_loss: 0.6748 - val_accuracy: 0.6092\n",
      "Epoch 136/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6749 - accuracy: 0.5845 - val_loss: 0.6746 - val_accuracy: 0.6107\n",
      "Epoch 137/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6750 - accuracy: 0.5796 - val_loss: 0.6745 - val_accuracy: 0.6109\n",
      "Epoch 138/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6751 - accuracy: 0.5830 - val_loss: 0.6743 - val_accuracy: 0.6129\n",
      "Epoch 139/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6755 - accuracy: 0.5826 - val_loss: 0.6743 - val_accuracy: 0.6120\n",
      "Epoch 140/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6751 - accuracy: 0.5839 - val_loss: 0.6743 - val_accuracy: 0.6126\n",
      "Epoch 141/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6743 - accuracy: 0.5891 - val_loss: 0.6742 - val_accuracy: 0.6124\n",
      "Epoch 142/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6746 - accuracy: 0.5899 - val_loss: 0.6741 - val_accuracy: 0.6139\n",
      "Epoch 143/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6745 - accuracy: 0.5852 - val_loss: 0.6740 - val_accuracy: 0.6120\n",
      "Epoch 144/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6745 - accuracy: 0.5870 - val_loss: 0.6740 - val_accuracy: 0.6131\n",
      "Epoch 145/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6738 - accuracy: 0.5911 - val_loss: 0.6738 - val_accuracy: 0.6141\n",
      "Epoch 146/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6746 - accuracy: 0.5869 - val_loss: 0.6739 - val_accuracy: 0.6131\n",
      "Epoch 147/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6726 - accuracy: 0.5903 - val_loss: 0.6735 - val_accuracy: 0.6137\n",
      "Epoch 148/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6725 - accuracy: 0.5900 - val_loss: 0.6735 - val_accuracy: 0.6131\n",
      "Epoch 149/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6757 - accuracy: 0.5905 - val_loss: 0.6734 - val_accuracy: 0.6133\n",
      "Epoch 150/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6750 - accuracy: 0.5838 - val_loss: 0.6733 - val_accuracy: 0.6126\n",
      "Epoch 151/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6750 - accuracy: 0.5893 - val_loss: 0.6733 - val_accuracy: 0.6131\n",
      "Epoch 152/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6749 - accuracy: 0.5885 - val_loss: 0.6733 - val_accuracy: 0.6133\n",
      "Epoch 153/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6738 - accuracy: 0.5893 - val_loss: 0.6734 - val_accuracy: 0.6129\n",
      "Epoch 154/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6741 - accuracy: 0.5919 - val_loss: 0.6733 - val_accuracy: 0.6126\n",
      "Epoch 155/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6725 - accuracy: 0.5935 - val_loss: 0.6731 - val_accuracy: 0.6139\n",
      "Epoch 156/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6723 - accuracy: 0.5872 - val_loss: 0.6729 - val_accuracy: 0.6148\n",
      "Epoch 157/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6743 - accuracy: 0.5860 - val_loss: 0.6728 - val_accuracy: 0.6141\n",
      "Epoch 158/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6746 - accuracy: 0.5877 - val_loss: 0.6729 - val_accuracy: 0.6135\n",
      "Epoch 159/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6736 - accuracy: 0.5900 - val_loss: 0.6728 - val_accuracy: 0.6131\n",
      "Epoch 160/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6745 - accuracy: 0.5905 - val_loss: 0.6728 - val_accuracy: 0.6137\n",
      "Epoch 161/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6743 - accuracy: 0.5872 - val_loss: 0.6727 - val_accuracy: 0.6144\n",
      "Epoch 162/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6710 - accuracy: 0.5917 - val_loss: 0.6726 - val_accuracy: 0.6148\n",
      "Epoch 163/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6741 - accuracy: 0.5919 - val_loss: 0.6726 - val_accuracy: 0.6137\n",
      "Epoch 164/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6725 - accuracy: 0.5959 - val_loss: 0.6726 - val_accuracy: 0.6135\n",
      "Epoch 165/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6729 - accuracy: 0.5932 - val_loss: 0.6725 - val_accuracy: 0.6148\n",
      "Epoch 166/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6729 - accuracy: 0.5937 - val_loss: 0.6723 - val_accuracy: 0.6167\n",
      "Epoch 167/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6730 - accuracy: 0.5970 - val_loss: 0.6722 - val_accuracy: 0.6169\n",
      "Epoch 168/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6729 - accuracy: 0.5940 - val_loss: 0.6722 - val_accuracy: 0.6169\n",
      "Epoch 169/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6721 - accuracy: 0.5958 - val_loss: 0.6722 - val_accuracy: 0.6163\n",
      "Epoch 170/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6729 - accuracy: 0.5946 - val_loss: 0.6721 - val_accuracy: 0.6165\n",
      "Epoch 171/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6726 - accuracy: 0.5966 - val_loss: 0.6720 - val_accuracy: 0.6174\n",
      "Epoch 172/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6733 - accuracy: 0.5930 - val_loss: 0.6720 - val_accuracy: 0.6169\n",
      "Epoch 173/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6741 - accuracy: 0.5921 - val_loss: 0.6721 - val_accuracy: 0.6163\n",
      "Epoch 174/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6730 - accuracy: 0.5949 - val_loss: 0.6720 - val_accuracy: 0.6165\n",
      "Epoch 175/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6713 - accuracy: 0.5949 - val_loss: 0.6718 - val_accuracy: 0.6178\n",
      "Epoch 176/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6753 - accuracy: 0.5903 - val_loss: 0.6719 - val_accuracy: 0.6172\n",
      "Epoch 177/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6723 - accuracy: 0.5983 - val_loss: 0.6718 - val_accuracy: 0.6174\n",
      "Epoch 178/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6727 - accuracy: 0.5931 - val_loss: 0.6718 - val_accuracy: 0.6159\n",
      "Epoch 179/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6717 - accuracy: 0.6001 - val_loss: 0.6718 - val_accuracy: 0.6172\n",
      "Epoch 180/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6714 - accuracy: 0.5954 - val_loss: 0.6718 - val_accuracy: 0.6174\n",
      "Epoch 181/1000\n",
      "582/582 [==============================] - 19s 33ms/step - loss: 0.6705 - accuracy: 0.5918 - val_loss: 0.6716 - val_accuracy: 0.6180\n",
      "Epoch 182/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6715 - accuracy: 0.5931 - val_loss: 0.6714 - val_accuracy: 0.6184\n",
      "Epoch 183/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6713 - accuracy: 0.5987 - val_loss: 0.6714 - val_accuracy: 0.6187\n",
      "Epoch 184/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6710 - accuracy: 0.5989 - val_loss: 0.6713 - val_accuracy: 0.6189\n",
      "Epoch 185/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6716 - accuracy: 0.5955 - val_loss: 0.6712 - val_accuracy: 0.6199\n",
      "Epoch 186/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6711 - accuracy: 0.6004 - val_loss: 0.6712 - val_accuracy: 0.6202\n",
      "Epoch 187/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6709 - accuracy: 0.6020 - val_loss: 0.6711 - val_accuracy: 0.6187\n",
      "Epoch 188/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6711 - accuracy: 0.5992 - val_loss: 0.6709 - val_accuracy: 0.6197\n",
      "Epoch 189/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6729 - accuracy: 0.5983 - val_loss: 0.6709 - val_accuracy: 0.6191\n",
      "Epoch 190/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6719 - accuracy: 0.5949 - val_loss: 0.6709 - val_accuracy: 0.6193\n",
      "Epoch 191/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6715 - accuracy: 0.5983 - val_loss: 0.6708 - val_accuracy: 0.6206\n",
      "Epoch 192/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6710 - accuracy: 0.5987 - val_loss: 0.6707 - val_accuracy: 0.6206\n",
      "Epoch 193/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6710 - accuracy: 0.5995 - val_loss: 0.6706 - val_accuracy: 0.6206\n",
      "Epoch 194/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6700 - accuracy: 0.6038 - val_loss: 0.6705 - val_accuracy: 0.6206\n",
      "Epoch 195/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6708 - accuracy: 0.5991 - val_loss: 0.6704 - val_accuracy: 0.6212\n",
      "Epoch 196/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6712 - accuracy: 0.5981 - val_loss: 0.6705 - val_accuracy: 0.6208\n",
      "Epoch 197/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6720 - accuracy: 0.5987 - val_loss: 0.6704 - val_accuracy: 0.6210\n",
      "Epoch 198/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6708 - accuracy: 0.5996 - val_loss: 0.6703 - val_accuracy: 0.6208\n",
      "Epoch 199/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6711 - accuracy: 0.5956 - val_loss: 0.6702 - val_accuracy: 0.6212\n",
      "Epoch 200/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6714 - accuracy: 0.5966 - val_loss: 0.6703 - val_accuracy: 0.6208\n",
      "Epoch 201/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6724 - accuracy: 0.5980 - val_loss: 0.6702 - val_accuracy: 0.6206\n",
      "Epoch 202/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6711 - accuracy: 0.6011 - val_loss: 0.6701 - val_accuracy: 0.6210\n",
      "Epoch 203/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6712 - accuracy: 0.6003 - val_loss: 0.6702 - val_accuracy: 0.6204\n",
      "Epoch 204/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6708 - accuracy: 0.6023 - val_loss: 0.6700 - val_accuracy: 0.6217\n",
      "Epoch 205/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6695 - accuracy: 0.5978 - val_loss: 0.6700 - val_accuracy: 0.6210\n",
      "Epoch 206/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6707 - accuracy: 0.6012 - val_loss: 0.6699 - val_accuracy: 0.6215\n",
      "Epoch 207/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6705 - accuracy: 0.5985 - val_loss: 0.6699 - val_accuracy: 0.6215\n",
      "Epoch 208/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6701 - accuracy: 0.6006 - val_loss: 0.6698 - val_accuracy: 0.6208\n",
      "Epoch 209/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6710 - accuracy: 0.5977 - val_loss: 0.6698 - val_accuracy: 0.6212\n",
      "Epoch 210/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6716 - accuracy: 0.6039 - val_loss: 0.6697 - val_accuracy: 0.6215\n",
      "Epoch 211/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6699 - accuracy: 0.6011 - val_loss: 0.6697 - val_accuracy: 0.6210\n",
      "Epoch 212/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6682 - accuracy: 0.6040 - val_loss: 0.6696 - val_accuracy: 0.6208\n",
      "Epoch 213/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6712 - accuracy: 0.5992 - val_loss: 0.6696 - val_accuracy: 0.6219\n",
      "Epoch 214/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6694 - accuracy: 0.6042 - val_loss: 0.6694 - val_accuracy: 0.6210\n",
      "Epoch 215/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6703 - accuracy: 0.6005 - val_loss: 0.6694 - val_accuracy: 0.6195\n",
      "Epoch 216/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6015 - val_loss: 0.6693 - val_accuracy: 0.6206\n",
      "Epoch 217/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6700 - accuracy: 0.6000 - val_loss: 0.6694 - val_accuracy: 0.6208\n",
      "Epoch 218/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6707 - accuracy: 0.6006 - val_loss: 0.6694 - val_accuracy: 0.6202\n",
      "Epoch 219/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6714 - accuracy: 0.6039 - val_loss: 0.6693 - val_accuracy: 0.6212\n",
      "Epoch 220/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6709 - accuracy: 0.6005 - val_loss: 0.6693 - val_accuracy: 0.6202\n",
      "Epoch 221/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6691 - accuracy: 0.6033 - val_loss: 0.6693 - val_accuracy: 0.6202\n",
      "Epoch 222/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6707 - accuracy: 0.6005 - val_loss: 0.6694 - val_accuracy: 0.6217\n",
      "Epoch 223/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6686 - accuracy: 0.6021 - val_loss: 0.6693 - val_accuracy: 0.6223\n",
      "Epoch 224/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6683 - accuracy: 0.6048 - val_loss: 0.6691 - val_accuracy: 0.6221\n",
      "Epoch 225/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6697 - accuracy: 0.5978 - val_loss: 0.6690 - val_accuracy: 0.6234\n",
      "Epoch 226/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6678 - accuracy: 0.6075 - val_loss: 0.6690 - val_accuracy: 0.6236\n",
      "Epoch 227/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6693 - accuracy: 0.6020 - val_loss: 0.6690 - val_accuracy: 0.6236\n",
      "Epoch 228/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6686 - accuracy: 0.6020 - val_loss: 0.6690 - val_accuracy: 0.6225\n",
      "Epoch 229/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6695 - accuracy: 0.6038 - val_loss: 0.6689 - val_accuracy: 0.6223\n",
      "Epoch 230/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6699 - accuracy: 0.6001 - val_loss: 0.6688 - val_accuracy: 0.6212\n",
      "Epoch 231/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6677 - accuracy: 0.6042 - val_loss: 0.6686 - val_accuracy: 0.6217\n",
      "Epoch 232/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6681 - accuracy: 0.6035 - val_loss: 0.6686 - val_accuracy: 0.6219\n",
      "Epoch 233/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6702 - accuracy: 0.6002 - val_loss: 0.6686 - val_accuracy: 0.6223\n",
      "Epoch 234/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6691 - accuracy: 0.6004 - val_loss: 0.6686 - val_accuracy: 0.6225\n",
      "Epoch 235/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6701 - accuracy: 0.6016 - val_loss: 0.6687 - val_accuracy: 0.6232\n",
      "Epoch 236/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6703 - accuracy: 0.6037 - val_loss: 0.6686 - val_accuracy: 0.6232\n",
      "Epoch 237/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6690 - accuracy: 0.6032 - val_loss: 0.6685 - val_accuracy: 0.6236\n",
      "Epoch 238/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6680 - accuracy: 0.6040 - val_loss: 0.6684 - val_accuracy: 0.6234\n",
      "Epoch 239/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6676 - accuracy: 0.6080 - val_loss: 0.6685 - val_accuracy: 0.6240\n",
      "Epoch 240/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6692 - accuracy: 0.6038 - val_loss: 0.6685 - val_accuracy: 0.6247\n",
      "Epoch 241/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6686 - accuracy: 0.6032 - val_loss: 0.6684 - val_accuracy: 0.6253\n",
      "Epoch 242/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6701 - accuracy: 0.6039 - val_loss: 0.6684 - val_accuracy: 0.6255\n",
      "Epoch 243/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6690 - accuracy: 0.6047 - val_loss: 0.6684 - val_accuracy: 0.6258\n",
      "Epoch 244/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6690 - accuracy: 0.6046 - val_loss: 0.6682 - val_accuracy: 0.6249\n",
      "Epoch 245/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6693 - accuracy: 0.6045 - val_loss: 0.6682 - val_accuracy: 0.6260\n",
      "Epoch 246/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6687 - accuracy: 0.6045 - val_loss: 0.6681 - val_accuracy: 0.6255\n",
      "Epoch 247/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6681 - accuracy: 0.6059 - val_loss: 0.6681 - val_accuracy: 0.6275\n",
      "Epoch 248/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6692 - accuracy: 0.6074 - val_loss: 0.6680 - val_accuracy: 0.6268\n",
      "Epoch 249/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6680 - accuracy: 0.6046 - val_loss: 0.6680 - val_accuracy: 0.6264\n",
      "Epoch 250/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6681 - accuracy: 0.6040 - val_loss: 0.6680 - val_accuracy: 0.6262\n",
      "Epoch 251/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6671 - accuracy: 0.6083 - val_loss: 0.6680 - val_accuracy: 0.6270\n",
      "Epoch 252/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6682 - accuracy: 0.6032 - val_loss: 0.6679 - val_accuracy: 0.6275\n",
      "Epoch 253/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6673 - accuracy: 0.6075 - val_loss: 0.6677 - val_accuracy: 0.6275\n",
      "Epoch 254/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6687 - accuracy: 0.6039 - val_loss: 0.6676 - val_accuracy: 0.6273\n",
      "Epoch 255/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6690 - accuracy: 0.6051 - val_loss: 0.6676 - val_accuracy: 0.6266\n",
      "Epoch 256/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6694 - accuracy: 0.6024 - val_loss: 0.6677 - val_accuracy: 0.6277\n",
      "Epoch 257/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6664 - accuracy: 0.6095 - val_loss: 0.6675 - val_accuracy: 0.6273\n",
      "Epoch 258/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6688 - accuracy: 0.6027 - val_loss: 0.6674 - val_accuracy: 0.6262\n",
      "Epoch 259/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6681 - accuracy: 0.6065 - val_loss: 0.6674 - val_accuracy: 0.6260\n",
      "Epoch 260/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6671 - accuracy: 0.6049 - val_loss: 0.6674 - val_accuracy: 0.6264\n",
      "Epoch 261/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6683 - accuracy: 0.6051 - val_loss: 0.6674 - val_accuracy: 0.6264\n",
      "Epoch 262/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6676 - accuracy: 0.6054 - val_loss: 0.6675 - val_accuracy: 0.6273\n",
      "Epoch 263/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6678 - accuracy: 0.6040 - val_loss: 0.6674 - val_accuracy: 0.6273\n",
      "Epoch 264/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6696 - accuracy: 0.6055 - val_loss: 0.6674 - val_accuracy: 0.6273\n",
      "Epoch 265/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6673 - accuracy: 0.6073 - val_loss: 0.6675 - val_accuracy: 0.6270\n",
      "Epoch 266/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6660 - accuracy: 0.6048 - val_loss: 0.6674 - val_accuracy: 0.6268\n",
      "Epoch 267/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6026 - val_loss: 0.6673 - val_accuracy: 0.6268\n",
      "Epoch 268/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6666 - accuracy: 0.6081 - val_loss: 0.6671 - val_accuracy: 0.6266\n",
      "Epoch 269/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6080 - val_loss: 0.6672 - val_accuracy: 0.6264\n",
      "Epoch 270/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6063 - val_loss: 0.6671 - val_accuracy: 0.6266\n",
      "Epoch 271/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6663 - accuracy: 0.6074 - val_loss: 0.6669 - val_accuracy: 0.6266\n",
      "Epoch 272/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6054 - val_loss: 0.6670 - val_accuracy: 0.6268\n",
      "Epoch 273/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6685 - accuracy: 0.6068 - val_loss: 0.6670 - val_accuracy: 0.6268\n",
      "Epoch 274/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6668 - accuracy: 0.6120 - val_loss: 0.6669 - val_accuracy: 0.6273\n",
      "Epoch 275/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6057 - val_loss: 0.6669 - val_accuracy: 0.6277\n",
      "Epoch 276/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6664 - accuracy: 0.6096 - val_loss: 0.6669 - val_accuracy: 0.6273\n",
      "Epoch 277/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6684 - accuracy: 0.6078 - val_loss: 0.6669 - val_accuracy: 0.6273\n",
      "Epoch 278/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6669 - accuracy: 0.6090 - val_loss: 0.6667 - val_accuracy: 0.6292\n",
      "Epoch 279/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6661 - accuracy: 0.6087 - val_loss: 0.6665 - val_accuracy: 0.6285\n",
      "Epoch 280/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6676 - accuracy: 0.6059 - val_loss: 0.6667 - val_accuracy: 0.6281\n",
      "Epoch 281/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6054 - val_loss: 0.6670 - val_accuracy: 0.6266\n",
      "Epoch 282/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6672 - accuracy: 0.6078 - val_loss: 0.6668 - val_accuracy: 0.6277\n",
      "Epoch 283/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6653 - accuracy: 0.6105 - val_loss: 0.6668 - val_accuracy: 0.6279\n",
      "Epoch 284/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6682 - accuracy: 0.6083 - val_loss: 0.6668 - val_accuracy: 0.6273\n",
      "Epoch 285/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6679 - accuracy: 0.6060 - val_loss: 0.6668 - val_accuracy: 0.6273\n",
      "Epoch 286/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6678 - accuracy: 0.6060 - val_loss: 0.6666 - val_accuracy: 0.6279\n",
      "Epoch 287/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6668 - accuracy: 0.6057 - val_loss: 0.6665 - val_accuracy: 0.6290\n",
      "Epoch 288/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6671 - accuracy: 0.6102 - val_loss: 0.6666 - val_accuracy: 0.6270\n",
      "Epoch 289/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6651 - accuracy: 0.6135 - val_loss: 0.6665 - val_accuracy: 0.6281\n",
      "Epoch 290/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6655 - accuracy: 0.6093 - val_loss: 0.6663 - val_accuracy: 0.6298\n",
      "Epoch 291/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6661 - accuracy: 0.6054 - val_loss: 0.6663 - val_accuracy: 0.6298\n",
      "Epoch 292/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6669 - accuracy: 0.6067 - val_loss: 0.6660 - val_accuracy: 0.6290\n",
      "Epoch 293/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6651 - accuracy: 0.6119 - val_loss: 0.6661 - val_accuracy: 0.6294\n",
      "Epoch 294/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6658 - accuracy: 0.6063 - val_loss: 0.6661 - val_accuracy: 0.6296\n",
      "Epoch 295/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6668 - accuracy: 0.6069 - val_loss: 0.6662 - val_accuracy: 0.6281\n",
      "Epoch 296/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6661 - accuracy: 0.6086 - val_loss: 0.6662 - val_accuracy: 0.6288\n",
      "Epoch 297/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6652 - accuracy: 0.6090 - val_loss: 0.6662 - val_accuracy: 0.6285\n",
      "Epoch 298/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6662 - accuracy: 0.6092 - val_loss: 0.6661 - val_accuracy: 0.6290\n",
      "Epoch 299/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6672 - accuracy: 0.6051 - val_loss: 0.6661 - val_accuracy: 0.6290\n",
      "Epoch 300/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6663 - accuracy: 0.6033 - val_loss: 0.6661 - val_accuracy: 0.6277\n",
      "Epoch 301/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6653 - accuracy: 0.6100 - val_loss: 0.6659 - val_accuracy: 0.6305\n",
      "Epoch 302/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6662 - accuracy: 0.6113 - val_loss: 0.6658 - val_accuracy: 0.6294\n",
      "Epoch 303/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6665 - accuracy: 0.6119 - val_loss: 0.6658 - val_accuracy: 0.6298\n",
      "Epoch 304/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6665 - accuracy: 0.6115 - val_loss: 0.6658 - val_accuracy: 0.6296\n",
      "Epoch 305/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6653 - accuracy: 0.6104 - val_loss: 0.6657 - val_accuracy: 0.6298\n",
      "Epoch 306/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6669 - accuracy: 0.6071 - val_loss: 0.6657 - val_accuracy: 0.6294\n",
      "Epoch 307/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6654 - accuracy: 0.6110 - val_loss: 0.6655 - val_accuracy: 0.6285\n",
      "Epoch 308/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6653 - accuracy: 0.6084 - val_loss: 0.6655 - val_accuracy: 0.6290\n",
      "Epoch 309/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6663 - accuracy: 0.6099 - val_loss: 0.6655 - val_accuracy: 0.6288\n",
      "Epoch 310/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6653 - accuracy: 0.6080 - val_loss: 0.6654 - val_accuracy: 0.6307\n",
      "Epoch 311/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6647 - accuracy: 0.6102 - val_loss: 0.6655 - val_accuracy: 0.6307\n",
      "Epoch 312/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6645 - accuracy: 0.6107 - val_loss: 0.6654 - val_accuracy: 0.6307\n",
      "Epoch 313/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6643 - accuracy: 0.6083 - val_loss: 0.6653 - val_accuracy: 0.6311\n",
      "Epoch 314/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6659 - accuracy: 0.6082 - val_loss: 0.6652 - val_accuracy: 0.6309\n",
      "Epoch 315/1000\n",
      "582/582 [==============================] - 20s 34ms/step - loss: 0.6663 - accuracy: 0.6077 - val_loss: 0.6651 - val_accuracy: 0.6313\n",
      "Epoch 316/1000\n",
      "582/582 [==============================] - 16s 27ms/step - loss: 0.6658 - accuracy: 0.6077 - val_loss: 0.6649 - val_accuracy: 0.6328\n",
      "Epoch 317/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6645 - accuracy: 0.6121 - val_loss: 0.6650 - val_accuracy: 0.6309\n",
      "Epoch 318/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6664 - accuracy: 0.6076 - val_loss: 0.6652 - val_accuracy: 0.6318\n",
      "Epoch 319/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6649 - accuracy: 0.6111 - val_loss: 0.6651 - val_accuracy: 0.6313\n",
      "Epoch 320/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6639 - accuracy: 0.6120 - val_loss: 0.6651 - val_accuracy: 0.6305\n",
      "Epoch 321/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6666 - accuracy: 0.6067 - val_loss: 0.6652 - val_accuracy: 0.6311\n",
      "Epoch 322/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6628 - accuracy: 0.6127 - val_loss: 0.6650 - val_accuracy: 0.6307\n",
      "Epoch 323/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6656 - accuracy: 0.6093 - val_loss: 0.6650 - val_accuracy: 0.6318\n",
      "Epoch 324/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6659 - accuracy: 0.6062 - val_loss: 0.6649 - val_accuracy: 0.6313\n",
      "Epoch 325/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6656 - accuracy: 0.6121 - val_loss: 0.6650 - val_accuracy: 0.6309\n",
      "Epoch 326/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6642 - accuracy: 0.6119 - val_loss: 0.6648 - val_accuracy: 0.6316\n",
      "Epoch 327/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6651 - accuracy: 0.6126 - val_loss: 0.6647 - val_accuracy: 0.6313\n",
      "Epoch 328/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6649 - accuracy: 0.6091 - val_loss: 0.6649 - val_accuracy: 0.6303\n",
      "Epoch 329/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6644 - accuracy: 0.6077 - val_loss: 0.6648 - val_accuracy: 0.6307\n",
      "Epoch 330/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6657 - accuracy: 0.6074 - val_loss: 0.6647 - val_accuracy: 0.6316\n",
      "Epoch 331/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6643 - accuracy: 0.6116 - val_loss: 0.6648 - val_accuracy: 0.6307\n",
      "Epoch 332/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6650 - accuracy: 0.6088 - val_loss: 0.6648 - val_accuracy: 0.6301\n",
      "Epoch 333/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6664 - accuracy: 0.6093 - val_loss: 0.6649 - val_accuracy: 0.6303\n",
      "Epoch 334/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6641 - accuracy: 0.6127 - val_loss: 0.6647 - val_accuracy: 0.6303\n",
      "Epoch 335/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6640 - accuracy: 0.6138 - val_loss: 0.6646 - val_accuracy: 0.6303\n",
      "Epoch 336/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6630 - accuracy: 0.6146 - val_loss: 0.6646 - val_accuracy: 0.6303\n",
      "Epoch 337/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6630 - accuracy: 0.6154 - val_loss: 0.6644 - val_accuracy: 0.6309\n",
      "Epoch 338/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6638 - accuracy: 0.6122 - val_loss: 0.6643 - val_accuracy: 0.6309\n",
      "Epoch 339/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6648 - accuracy: 0.6090 - val_loss: 0.6644 - val_accuracy: 0.6305\n",
      "Epoch 340/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6634 - accuracy: 0.6125 - val_loss: 0.6643 - val_accuracy: 0.6303\n",
      "Epoch 341/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6651 - accuracy: 0.6094 - val_loss: 0.6643 - val_accuracy: 0.6298\n",
      "Epoch 342/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6644 - accuracy: 0.6090 - val_loss: 0.6643 - val_accuracy: 0.6309\n",
      "Epoch 343/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6632 - accuracy: 0.6149 - val_loss: 0.6643 - val_accuracy: 0.6307\n",
      "Epoch 344/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6629 - accuracy: 0.6103 - val_loss: 0.6642 - val_accuracy: 0.6305\n",
      "Epoch 345/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6636 - accuracy: 0.6131 - val_loss: 0.6640 - val_accuracy: 0.6311\n",
      "Epoch 346/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6648 - accuracy: 0.6097 - val_loss: 0.6641 - val_accuracy: 0.6316\n",
      "Epoch 347/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6637 - accuracy: 0.6117 - val_loss: 0.6640 - val_accuracy: 0.6316\n",
      "Epoch 348/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6629 - accuracy: 0.6111 - val_loss: 0.6640 - val_accuracy: 0.6305\n",
      "Epoch 349/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6646 - accuracy: 0.6124 - val_loss: 0.6640 - val_accuracy: 0.6303\n",
      "Epoch 350/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6666 - accuracy: 0.6072 - val_loss: 0.6641 - val_accuracy: 0.6311\n",
      "Epoch 351/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6651 - accuracy: 0.6113 - val_loss: 0.6642 - val_accuracy: 0.6316\n",
      "Epoch 352/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6634 - accuracy: 0.6095 - val_loss: 0.6641 - val_accuracy: 0.6318\n",
      "Epoch 353/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6647 - accuracy: 0.6085 - val_loss: 0.6641 - val_accuracy: 0.6320\n",
      "Epoch 354/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6126 - val_loss: 0.6640 - val_accuracy: 0.6320\n",
      "Epoch 355/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6642 - accuracy: 0.6121 - val_loss: 0.6640 - val_accuracy: 0.6318\n",
      "Epoch 356/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6645 - accuracy: 0.6071 - val_loss: 0.6640 - val_accuracy: 0.6328\n",
      "Epoch 357/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6652 - accuracy: 0.6133 - val_loss: 0.6641 - val_accuracy: 0.6326\n",
      "Epoch 358/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6631 - accuracy: 0.6160 - val_loss: 0.6639 - val_accuracy: 0.6322\n",
      "Epoch 359/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6644 - accuracy: 0.6129 - val_loss: 0.6639 - val_accuracy: 0.6320\n",
      "Epoch 360/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6607 - accuracy: 0.6145 - val_loss: 0.6637 - val_accuracy: 0.6318\n",
      "Epoch 361/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6140 - val_loss: 0.6638 - val_accuracy: 0.6318\n",
      "Epoch 362/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6645 - accuracy: 0.6116 - val_loss: 0.6638 - val_accuracy: 0.6326\n",
      "Epoch 363/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6649 - accuracy: 0.6090 - val_loss: 0.6638 - val_accuracy: 0.6320\n",
      "Epoch 364/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6632 - accuracy: 0.6162 - val_loss: 0.6638 - val_accuracy: 0.6322\n",
      "Epoch 365/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6643 - accuracy: 0.6107 - val_loss: 0.6636 - val_accuracy: 0.6328\n",
      "Epoch 366/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6641 - accuracy: 0.6131 - val_loss: 0.6637 - val_accuracy: 0.6318\n",
      "Epoch 367/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6121 - val_loss: 0.6637 - val_accuracy: 0.6313\n",
      "Epoch 368/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6625 - accuracy: 0.6133 - val_loss: 0.6634 - val_accuracy: 0.6333\n",
      "Epoch 369/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6623 - accuracy: 0.6152 - val_loss: 0.6634 - val_accuracy: 0.6337\n",
      "Epoch 370/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6639 - accuracy: 0.6121 - val_loss: 0.6635 - val_accuracy: 0.6337\n",
      "Epoch 371/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6636 - accuracy: 0.6140 - val_loss: 0.6636 - val_accuracy: 0.6322\n",
      "Epoch 372/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6183 - val_loss: 0.6635 - val_accuracy: 0.6324\n",
      "Epoch 373/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6627 - accuracy: 0.6146 - val_loss: 0.6634 - val_accuracy: 0.6335\n",
      "Epoch 374/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6640 - accuracy: 0.6132 - val_loss: 0.6635 - val_accuracy: 0.6318\n",
      "Epoch 375/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6616 - accuracy: 0.6155 - val_loss: 0.6635 - val_accuracy: 0.6313\n",
      "Epoch 376/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6622 - accuracy: 0.6124 - val_loss: 0.6634 - val_accuracy: 0.6328\n",
      "Epoch 377/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6624 - accuracy: 0.6174 - val_loss: 0.6632 - val_accuracy: 0.6335\n",
      "Epoch 378/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6626 - accuracy: 0.6117 - val_loss: 0.6632 - val_accuracy: 0.6335\n",
      "Epoch 379/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6624 - accuracy: 0.6157 - val_loss: 0.6632 - val_accuracy: 0.6322\n",
      "Epoch 380/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6640 - accuracy: 0.6109 - val_loss: 0.6632 - val_accuracy: 0.6320\n",
      "Epoch 381/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6130 - val_loss: 0.6633 - val_accuracy: 0.6313\n",
      "Epoch 382/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6629 - accuracy: 0.6124 - val_loss: 0.6632 - val_accuracy: 0.6318\n",
      "Epoch 383/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6628 - accuracy: 0.6115 - val_loss: 0.6632 - val_accuracy: 0.6337\n",
      "Epoch 384/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6612 - accuracy: 0.6176 - val_loss: 0.6631 - val_accuracy: 0.6337\n",
      "Epoch 385/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6625 - accuracy: 0.6135 - val_loss: 0.6630 - val_accuracy: 0.6331\n",
      "Epoch 386/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6641 - accuracy: 0.6138 - val_loss: 0.6632 - val_accuracy: 0.6354\n",
      "Epoch 387/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6630 - accuracy: 0.6115 - val_loss: 0.6632 - val_accuracy: 0.6328\n",
      "Epoch 388/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6611 - accuracy: 0.6160 - val_loss: 0.6631 - val_accuracy: 0.6346\n",
      "Epoch 389/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6633 - accuracy: 0.6131 - val_loss: 0.6630 - val_accuracy: 0.6344\n",
      "Epoch 390/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6618 - accuracy: 0.6166 - val_loss: 0.6630 - val_accuracy: 0.6337\n",
      "Epoch 391/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6627 - accuracy: 0.6142 - val_loss: 0.6630 - val_accuracy: 0.6341\n",
      "Epoch 392/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6619 - accuracy: 0.6168 - val_loss: 0.6629 - val_accuracy: 0.6346\n",
      "Epoch 393/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6619 - accuracy: 0.6136 - val_loss: 0.6628 - val_accuracy: 0.6352\n",
      "Epoch 394/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6620 - accuracy: 0.6139 - val_loss: 0.6629 - val_accuracy: 0.6339\n",
      "Epoch 395/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6635 - accuracy: 0.6179 - val_loss: 0.6628 - val_accuracy: 0.6337\n",
      "Epoch 396/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6620 - accuracy: 0.6132 - val_loss: 0.6628 - val_accuracy: 0.6341\n",
      "Epoch 397/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6620 - accuracy: 0.6167 - val_loss: 0.6627 - val_accuracy: 0.6344\n",
      "Epoch 398/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6625 - accuracy: 0.6149 - val_loss: 0.6628 - val_accuracy: 0.6346\n",
      "Epoch 399/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6634 - accuracy: 0.6128 - val_loss: 0.6628 - val_accuracy: 0.6354\n",
      "Epoch 400/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6626 - accuracy: 0.6145 - val_loss: 0.6627 - val_accuracy: 0.6346\n",
      "Epoch 401/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6623 - accuracy: 0.6179 - val_loss: 0.6627 - val_accuracy: 0.6344\n",
      "Epoch 402/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6615 - accuracy: 0.6158 - val_loss: 0.6627 - val_accuracy: 0.6350\n",
      "Epoch 403/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6620 - accuracy: 0.6162 - val_loss: 0.6627 - val_accuracy: 0.6354\n",
      "Epoch 404/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6621 - accuracy: 0.6148 - val_loss: 0.6626 - val_accuracy: 0.6348\n",
      "Epoch 405/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6622 - accuracy: 0.6117 - val_loss: 0.6626 - val_accuracy: 0.6344\n",
      "Epoch 406/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6610 - accuracy: 0.6186 - val_loss: 0.6626 - val_accuracy: 0.6350\n",
      "Epoch 407/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6627 - accuracy: 0.6150 - val_loss: 0.6626 - val_accuracy: 0.6348\n",
      "Epoch 408/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6618 - accuracy: 0.6138 - val_loss: 0.6626 - val_accuracy: 0.6348\n",
      "Epoch 409/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6598 - accuracy: 0.6187 - val_loss: 0.6624 - val_accuracy: 0.6346\n",
      "Epoch 410/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6618 - accuracy: 0.6140 - val_loss: 0.6623 - val_accuracy: 0.6352\n",
      "Epoch 411/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6624 - accuracy: 0.6155 - val_loss: 0.6620 - val_accuracy: 0.6378\n",
      "Epoch 412/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6607 - accuracy: 0.6184 - val_loss: 0.6621 - val_accuracy: 0.6380\n",
      "Epoch 413/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6625 - accuracy: 0.6183 - val_loss: 0.6622 - val_accuracy: 0.6363\n",
      "Epoch 414/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6616 - accuracy: 0.6157 - val_loss: 0.6622 - val_accuracy: 0.6356\n",
      "Epoch 415/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6617 - accuracy: 0.6124 - val_loss: 0.6623 - val_accuracy: 0.6356\n",
      "Epoch 416/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6638 - accuracy: 0.6107 - val_loss: 0.6625 - val_accuracy: 0.6359\n",
      "Epoch 417/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6601 - accuracy: 0.6138 - val_loss: 0.6622 - val_accuracy: 0.6363\n",
      "Epoch 418/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6616 - accuracy: 0.6143 - val_loss: 0.6622 - val_accuracy: 0.6367\n",
      "Epoch 419/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6618 - accuracy: 0.6150 - val_loss: 0.6622 - val_accuracy: 0.6363\n",
      "Epoch 420/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6623 - accuracy: 0.6142 - val_loss: 0.6622 - val_accuracy: 0.6367\n",
      "Epoch 421/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6618 - accuracy: 0.6180 - val_loss: 0.6621 - val_accuracy: 0.6371\n",
      "Epoch 422/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6610 - accuracy: 0.6187 - val_loss: 0.6620 - val_accuracy: 0.6376\n",
      "Epoch 423/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6620 - accuracy: 0.6170 - val_loss: 0.6621 - val_accuracy: 0.6367\n",
      "Epoch 424/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6622 - accuracy: 0.6112 - val_loss: 0.6623 - val_accuracy: 0.6361\n",
      "Epoch 425/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6607 - accuracy: 0.6143 - val_loss: 0.6622 - val_accuracy: 0.6359\n",
      "Epoch 426/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6594 - accuracy: 0.6184 - val_loss: 0.6620 - val_accuracy: 0.6376\n",
      "Epoch 427/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6609 - accuracy: 0.6167 - val_loss: 0.6621 - val_accuracy: 0.6384\n",
      "Epoch 428/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6600 - accuracy: 0.6179 - val_loss: 0.6621 - val_accuracy: 0.6382\n",
      "Epoch 429/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6610 - accuracy: 0.6157 - val_loss: 0.6619 - val_accuracy: 0.6382\n",
      "Epoch 430/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6613 - accuracy: 0.6135 - val_loss: 0.6619 - val_accuracy: 0.6387\n",
      "Epoch 431/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6592 - accuracy: 0.6169 - val_loss: 0.6619 - val_accuracy: 0.6380\n",
      "Epoch 432/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6608 - accuracy: 0.6123 - val_loss: 0.6618 - val_accuracy: 0.6384\n",
      "Epoch 433/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6605 - accuracy: 0.6169 - val_loss: 0.6619 - val_accuracy: 0.6389\n",
      "Epoch 434/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6602 - accuracy: 0.6172 - val_loss: 0.6620 - val_accuracy: 0.6387\n",
      "Epoch 435/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6604 - accuracy: 0.6164 - val_loss: 0.6621 - val_accuracy: 0.6371\n",
      "Epoch 436/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6608 - accuracy: 0.6207 - val_loss: 0.6621 - val_accuracy: 0.6367\n",
      "Epoch 437/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6618 - accuracy: 0.6127 - val_loss: 0.6621 - val_accuracy: 0.6376\n",
      "Epoch 438/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6617 - accuracy: 0.6153 - val_loss: 0.6619 - val_accuracy: 0.6387\n",
      "Epoch 439/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6600 - accuracy: 0.6188 - val_loss: 0.6618 - val_accuracy: 0.6387\n",
      "Epoch 440/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6604 - accuracy: 0.6177 - val_loss: 0.6620 - val_accuracy: 0.6376\n",
      "Epoch 441/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6605 - accuracy: 0.6191 - val_loss: 0.6618 - val_accuracy: 0.6382\n",
      "Epoch 442/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6603 - accuracy: 0.6196 - val_loss: 0.6617 - val_accuracy: 0.6387\n",
      "Epoch 443/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6611 - accuracy: 0.6139 - val_loss: 0.6617 - val_accuracy: 0.6380\n",
      "Epoch 444/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6604 - accuracy: 0.6141 - val_loss: 0.6617 - val_accuracy: 0.6391\n",
      "Epoch 445/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6607 - accuracy: 0.6179 - val_loss: 0.6616 - val_accuracy: 0.6387\n",
      "Epoch 446/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6603 - accuracy: 0.6175 - val_loss: 0.6615 - val_accuracy: 0.6387\n",
      "Epoch 447/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6605 - accuracy: 0.6185 - val_loss: 0.6616 - val_accuracy: 0.6389\n",
      "Epoch 448/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6606 - accuracy: 0.6145 - val_loss: 0.6614 - val_accuracy: 0.6387\n",
      "Epoch 449/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6606 - accuracy: 0.6163 - val_loss: 0.6615 - val_accuracy: 0.6395\n",
      "Epoch 450/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6603 - accuracy: 0.6201 - val_loss: 0.6614 - val_accuracy: 0.6397\n",
      "Epoch 451/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6610 - accuracy: 0.6176 - val_loss: 0.6615 - val_accuracy: 0.6404\n",
      "Epoch 452/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6592 - accuracy: 0.6216 - val_loss: 0.6614 - val_accuracy: 0.6395\n",
      "Epoch 453/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6595 - accuracy: 0.6147 - val_loss: 0.6614 - val_accuracy: 0.6387\n",
      "Epoch 454/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6612 - accuracy: 0.6194 - val_loss: 0.6615 - val_accuracy: 0.6393\n",
      "Epoch 455/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6586 - accuracy: 0.6164 - val_loss: 0.6614 - val_accuracy: 0.6391\n",
      "Epoch 456/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6601 - accuracy: 0.6192 - val_loss: 0.6613 - val_accuracy: 0.6395\n",
      "Epoch 457/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6602 - accuracy: 0.6165 - val_loss: 0.6614 - val_accuracy: 0.6393\n",
      "Epoch 458/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6589 - accuracy: 0.6172 - val_loss: 0.6611 - val_accuracy: 0.6397\n",
      "Epoch 459/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6600 - accuracy: 0.6181 - val_loss: 0.6612 - val_accuracy: 0.6393\n",
      "Epoch 460/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6599 - accuracy: 0.6161 - val_loss: 0.6612 - val_accuracy: 0.6389\n",
      "Epoch 461/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6583 - accuracy: 0.6205 - val_loss: 0.6611 - val_accuracy: 0.6395\n",
      "Epoch 462/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6601 - accuracy: 0.6186 - val_loss: 0.6610 - val_accuracy: 0.6395\n",
      "Epoch 463/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6591 - accuracy: 0.6137 - val_loss: 0.6609 - val_accuracy: 0.6406\n",
      "Epoch 464/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6591 - accuracy: 0.6181 - val_loss: 0.6611 - val_accuracy: 0.6397\n",
      "Epoch 465/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6582 - accuracy: 0.6203 - val_loss: 0.6610 - val_accuracy: 0.6391\n",
      "Epoch 466/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6582 - accuracy: 0.6179 - val_loss: 0.6611 - val_accuracy: 0.6389\n",
      "Epoch 467/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6593 - accuracy: 0.6193 - val_loss: 0.6609 - val_accuracy: 0.6389\n",
      "Epoch 468/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6603 - accuracy: 0.6210 - val_loss: 0.6610 - val_accuracy: 0.6384\n",
      "Epoch 469/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6581 - accuracy: 0.6210 - val_loss: 0.6608 - val_accuracy: 0.6380\n",
      "Epoch 470/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6604 - accuracy: 0.6181 - val_loss: 0.6608 - val_accuracy: 0.6384\n",
      "Epoch 471/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6582 - accuracy: 0.6205 - val_loss: 0.6609 - val_accuracy: 0.6387\n",
      "Epoch 472/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6592 - accuracy: 0.6181 - val_loss: 0.6607 - val_accuracy: 0.6408\n",
      "Epoch 473/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6590 - accuracy: 0.6179 - val_loss: 0.6608 - val_accuracy: 0.6387\n",
      "Epoch 474/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6589 - accuracy: 0.6162 - val_loss: 0.6609 - val_accuracy: 0.6397\n",
      "Epoch 475/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6603 - accuracy: 0.6176 - val_loss: 0.6608 - val_accuracy: 0.6397\n",
      "Epoch 476/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6594 - accuracy: 0.6180 - val_loss: 0.6609 - val_accuracy: 0.6395\n",
      "Epoch 477/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6588 - accuracy: 0.6204 - val_loss: 0.6608 - val_accuracy: 0.6402\n",
      "Epoch 478/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6595 - accuracy: 0.6173 - val_loss: 0.6608 - val_accuracy: 0.6395\n",
      "Epoch 479/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6597 - accuracy: 0.6196 - val_loss: 0.6608 - val_accuracy: 0.6389\n",
      "Epoch 480/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6566 - accuracy: 0.6231 - val_loss: 0.6606 - val_accuracy: 0.6399\n",
      "Epoch 481/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6596 - accuracy: 0.6167 - val_loss: 0.6606 - val_accuracy: 0.6395\n",
      "Epoch 482/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6593 - accuracy: 0.6193 - val_loss: 0.6608 - val_accuracy: 0.6395\n",
      "Epoch 483/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6590 - accuracy: 0.6192 - val_loss: 0.6608 - val_accuracy: 0.6387\n",
      "Epoch 484/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6582 - accuracy: 0.6163 - val_loss: 0.6607 - val_accuracy: 0.6395\n",
      "Epoch 485/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6578 - accuracy: 0.6180 - val_loss: 0.6603 - val_accuracy: 0.6395\n",
      "Epoch 486/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6584 - accuracy: 0.6205 - val_loss: 0.6604 - val_accuracy: 0.6391\n",
      "Epoch 487/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6578 - accuracy: 0.6191 - val_loss: 0.6606 - val_accuracy: 0.6391\n",
      "Epoch 488/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6578 - accuracy: 0.6201 - val_loss: 0.6606 - val_accuracy: 0.6395\n",
      "Epoch 489/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6587 - accuracy: 0.6191 - val_loss: 0.6603 - val_accuracy: 0.6380\n",
      "Epoch 490/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6590 - accuracy: 0.6172 - val_loss: 0.6603 - val_accuracy: 0.6382\n",
      "Epoch 491/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6579 - accuracy: 0.6172 - val_loss: 0.6605 - val_accuracy: 0.6395\n",
      "Epoch 492/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6562 - accuracy: 0.6218 - val_loss: 0.6603 - val_accuracy: 0.6391\n",
      "Epoch 493/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6571 - accuracy: 0.6199 - val_loss: 0.6602 - val_accuracy: 0.6389\n",
      "Epoch 494/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6573 - accuracy: 0.6181 - val_loss: 0.6603 - val_accuracy: 0.6389\n",
      "Epoch 495/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6583 - accuracy: 0.6191 - val_loss: 0.6602 - val_accuracy: 0.6384\n",
      "Epoch 496/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6580 - accuracy: 0.6198 - val_loss: 0.6601 - val_accuracy: 0.6376\n",
      "Epoch 497/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6574 - accuracy: 0.6209 - val_loss: 0.6600 - val_accuracy: 0.6382\n",
      "Epoch 498/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6593 - accuracy: 0.6180 - val_loss: 0.6602 - val_accuracy: 0.6397\n",
      "Epoch 499/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6568 - accuracy: 0.6211 - val_loss: 0.6601 - val_accuracy: 0.6387\n",
      "Epoch 500/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6580 - accuracy: 0.6221 - val_loss: 0.6600 - val_accuracy: 0.6393\n",
      "Epoch 501/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6589 - accuracy: 0.6200 - val_loss: 0.6603 - val_accuracy: 0.6391\n",
      "Epoch 502/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6592 - accuracy: 0.6183 - val_loss: 0.6602 - val_accuracy: 0.6402\n",
      "Epoch 503/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6578 - accuracy: 0.6205 - val_loss: 0.6602 - val_accuracy: 0.6404\n",
      "Epoch 504/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6571 - accuracy: 0.6224 - val_loss: 0.6600 - val_accuracy: 0.6399\n",
      "Epoch 505/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6586 - accuracy: 0.6182 - val_loss: 0.6600 - val_accuracy: 0.6399\n",
      "Epoch 506/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6556 - accuracy: 0.6214 - val_loss: 0.6599 - val_accuracy: 0.6393\n",
      "Epoch 507/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6590 - accuracy: 0.6244 - val_loss: 0.6600 - val_accuracy: 0.6395\n",
      "Epoch 508/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6589 - accuracy: 0.6197 - val_loss: 0.6599 - val_accuracy: 0.6384\n",
      "Epoch 509/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6588 - accuracy: 0.6175 - val_loss: 0.6602 - val_accuracy: 0.6369\n",
      "Epoch 510/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6579 - accuracy: 0.6180 - val_loss: 0.6600 - val_accuracy: 0.6374\n",
      "Epoch 511/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6572 - accuracy: 0.6218 - val_loss: 0.6600 - val_accuracy: 0.6378\n",
      "Epoch 512/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6567 - accuracy: 0.6223 - val_loss: 0.6597 - val_accuracy: 0.6384\n",
      "Epoch 513/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6588 - accuracy: 0.6184 - val_loss: 0.6599 - val_accuracy: 0.6382\n",
      "Epoch 514/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6582 - accuracy: 0.6177 - val_loss: 0.6598 - val_accuracy: 0.6371\n",
      "Epoch 515/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6568 - accuracy: 0.6214 - val_loss: 0.6598 - val_accuracy: 0.6371\n",
      "Epoch 516/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6583 - accuracy: 0.6203 - val_loss: 0.6598 - val_accuracy: 0.6369\n",
      "Epoch 517/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6564 - accuracy: 0.6282 - val_loss: 0.6597 - val_accuracy: 0.6387\n",
      "Epoch 518/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6577 - accuracy: 0.6219 - val_loss: 0.6597 - val_accuracy: 0.6369\n",
      "Epoch 519/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6573 - accuracy: 0.6241 - val_loss: 0.6596 - val_accuracy: 0.6378\n",
      "Epoch 520/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6586 - accuracy: 0.6201 - val_loss: 0.6594 - val_accuracy: 0.6380\n",
      "Epoch 521/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6587 - accuracy: 0.6204 - val_loss: 0.6595 - val_accuracy: 0.6378\n",
      "Epoch 522/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6284 - val_loss: 0.6593 - val_accuracy: 0.6378\n",
      "Epoch 523/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6570 - accuracy: 0.6275 - val_loss: 0.6593 - val_accuracy: 0.6387\n",
      "Epoch 524/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6587 - accuracy: 0.6215 - val_loss: 0.6596 - val_accuracy: 0.6389\n",
      "Epoch 525/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6566 - accuracy: 0.6197 - val_loss: 0.6596 - val_accuracy: 0.6380\n",
      "Epoch 526/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6562 - accuracy: 0.6208 - val_loss: 0.6597 - val_accuracy: 0.6384\n",
      "Epoch 527/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6569 - accuracy: 0.6234 - val_loss: 0.6595 - val_accuracy: 0.6382\n",
      "Epoch 528/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6569 - accuracy: 0.6210 - val_loss: 0.6593 - val_accuracy: 0.6376\n",
      "Epoch 529/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6569 - accuracy: 0.6229 - val_loss: 0.6594 - val_accuracy: 0.6380\n",
      "Epoch 530/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6565 - accuracy: 0.6211 - val_loss: 0.6592 - val_accuracy: 0.6367\n",
      "Epoch 531/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6568 - accuracy: 0.6212 - val_loss: 0.6593 - val_accuracy: 0.6391\n",
      "Epoch 532/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6564 - accuracy: 0.6232 - val_loss: 0.6593 - val_accuracy: 0.6382\n",
      "Epoch 533/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6575 - accuracy: 0.6276 - val_loss: 0.6592 - val_accuracy: 0.6378\n",
      "Epoch 534/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6572 - accuracy: 0.6201 - val_loss: 0.6592 - val_accuracy: 0.6369\n",
      "Epoch 535/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6567 - accuracy: 0.6264 - val_loss: 0.6592 - val_accuracy: 0.6374\n",
      "Epoch 536/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6558 - accuracy: 0.6239 - val_loss: 0.6591 - val_accuracy: 0.6365\n",
      "Epoch 537/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6570 - accuracy: 0.6250 - val_loss: 0.6593 - val_accuracy: 0.6376\n",
      "Epoch 538/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6564 - accuracy: 0.6222 - val_loss: 0.6592 - val_accuracy: 0.6378\n",
      "Epoch 539/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6583 - accuracy: 0.6165 - val_loss: 0.6593 - val_accuracy: 0.6376\n",
      "Epoch 540/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6294 - val_loss: 0.6591 - val_accuracy: 0.6369\n",
      "Epoch 541/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6566 - accuracy: 0.6200 - val_loss: 0.6590 - val_accuracy: 0.6367\n",
      "Epoch 542/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6575 - accuracy: 0.6225 - val_loss: 0.6590 - val_accuracy: 0.6369\n",
      "Epoch 543/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6565 - accuracy: 0.6240 - val_loss: 0.6591 - val_accuracy: 0.6380\n",
      "Epoch 544/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6575 - accuracy: 0.6202 - val_loss: 0.6590 - val_accuracy: 0.6378\n",
      "Epoch 545/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6240 - val_loss: 0.6589 - val_accuracy: 0.6374\n",
      "Epoch 546/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6226 - val_loss: 0.6589 - val_accuracy: 0.6369\n",
      "Epoch 547/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6256 - val_loss: 0.6587 - val_accuracy: 0.6374\n",
      "Epoch 548/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6564 - accuracy: 0.6221 - val_loss: 0.6587 - val_accuracy: 0.6369\n",
      "Epoch 549/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6558 - accuracy: 0.6273 - val_loss: 0.6585 - val_accuracy: 0.6389\n",
      "Epoch 550/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6568 - accuracy: 0.6221 - val_loss: 0.6587 - val_accuracy: 0.6384\n",
      "Epoch 551/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6562 - accuracy: 0.6212 - val_loss: 0.6587 - val_accuracy: 0.6382\n",
      "Epoch 552/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6559 - accuracy: 0.6243 - val_loss: 0.6588 - val_accuracy: 0.6378\n",
      "Epoch 553/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6228 - val_loss: 0.6588 - val_accuracy: 0.6369\n",
      "Epoch 554/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6231 - val_loss: 0.6587 - val_accuracy: 0.6382\n",
      "Epoch 555/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6568 - accuracy: 0.6260 - val_loss: 0.6588 - val_accuracy: 0.6378\n",
      "Epoch 556/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6557 - accuracy: 0.6246 - val_loss: 0.6587 - val_accuracy: 0.6382\n",
      "Epoch 557/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6559 - accuracy: 0.6246 - val_loss: 0.6587 - val_accuracy: 0.6387\n",
      "Epoch 558/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6567 - accuracy: 0.6244 - val_loss: 0.6588 - val_accuracy: 0.6352\n",
      "Epoch 559/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6225 - val_loss: 0.6587 - val_accuracy: 0.6376\n",
      "Epoch 560/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6573 - accuracy: 0.6228 - val_loss: 0.6587 - val_accuracy: 0.6378\n",
      "Epoch 561/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6267 - val_loss: 0.6585 - val_accuracy: 0.6389\n",
      "Epoch 562/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6566 - accuracy: 0.6204 - val_loss: 0.6585 - val_accuracy: 0.6387\n",
      "Epoch 563/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6261 - val_loss: 0.6583 - val_accuracy: 0.6397\n",
      "Epoch 564/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6577 - accuracy: 0.6221 - val_loss: 0.6585 - val_accuracy: 0.6380\n",
      "Epoch 565/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6559 - accuracy: 0.6243 - val_loss: 0.6584 - val_accuracy: 0.6384\n",
      "Epoch 566/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6276 - val_loss: 0.6583 - val_accuracy: 0.6389\n",
      "Epoch 567/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6551 - accuracy: 0.6250 - val_loss: 0.6584 - val_accuracy: 0.6376\n",
      "Epoch 568/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6200 - val_loss: 0.6586 - val_accuracy: 0.6363\n",
      "Epoch 569/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6550 - accuracy: 0.6253 - val_loss: 0.6585 - val_accuracy: 0.6363\n",
      "Epoch 570/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6269 - val_loss: 0.6582 - val_accuracy: 0.6380\n",
      "Epoch 571/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6551 - accuracy: 0.6271 - val_loss: 0.6581 - val_accuracy: 0.6387\n",
      "Epoch 572/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6249 - val_loss: 0.6582 - val_accuracy: 0.6397\n",
      "Epoch 573/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6540 - accuracy: 0.6285 - val_loss: 0.6581 - val_accuracy: 0.6389\n",
      "Epoch 574/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6545 - accuracy: 0.6236 - val_loss: 0.6579 - val_accuracy: 0.6391\n",
      "Epoch 575/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6219 - val_loss: 0.6580 - val_accuracy: 0.6397\n",
      "Epoch 576/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6551 - accuracy: 0.6251 - val_loss: 0.6580 - val_accuracy: 0.6402\n",
      "Epoch 577/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6561 - accuracy: 0.6270 - val_loss: 0.6580 - val_accuracy: 0.6399\n",
      "Epoch 578/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6536 - accuracy: 0.6247 - val_loss: 0.6579 - val_accuracy: 0.6393\n",
      "Epoch 579/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6558 - accuracy: 0.6237 - val_loss: 0.6580 - val_accuracy: 0.6399\n",
      "Epoch 580/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6565 - accuracy: 0.6220 - val_loss: 0.6582 - val_accuracy: 0.6397\n",
      "Epoch 581/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6542 - accuracy: 0.6246 - val_loss: 0.6581 - val_accuracy: 0.6395\n",
      "Epoch 582/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6543 - accuracy: 0.6245 - val_loss: 0.6580 - val_accuracy: 0.6371\n",
      "Epoch 583/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6548 - accuracy: 0.6255 - val_loss: 0.6580 - val_accuracy: 0.6384\n",
      "Epoch 584/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6558 - accuracy: 0.6202 - val_loss: 0.6579 - val_accuracy: 0.6410\n",
      "Epoch 585/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6547 - accuracy: 0.6262 - val_loss: 0.6578 - val_accuracy: 0.6406\n",
      "Epoch 586/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6542 - accuracy: 0.6259 - val_loss: 0.6578 - val_accuracy: 0.6410\n",
      "Epoch 587/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6542 - accuracy: 0.6218 - val_loss: 0.6578 - val_accuracy: 0.6414\n",
      "Epoch 588/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6280 - val_loss: 0.6578 - val_accuracy: 0.6412\n",
      "Epoch 589/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6544 - accuracy: 0.6262 - val_loss: 0.6575 - val_accuracy: 0.6410\n",
      "Epoch 590/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6556 - accuracy: 0.6243 - val_loss: 0.6578 - val_accuracy: 0.6410\n",
      "Epoch 591/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6543 - accuracy: 0.6253 - val_loss: 0.6577 - val_accuracy: 0.6402\n",
      "Epoch 592/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6540 - accuracy: 0.6265 - val_loss: 0.6575 - val_accuracy: 0.6421\n",
      "Epoch 593/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6535 - accuracy: 0.6274 - val_loss: 0.6575 - val_accuracy: 0.6421\n",
      "Epoch 594/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6567 - accuracy: 0.6247 - val_loss: 0.6579 - val_accuracy: 0.6393\n",
      "Epoch 595/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6542 - accuracy: 0.6269 - val_loss: 0.6579 - val_accuracy: 0.6393\n",
      "Epoch 596/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6237 - val_loss: 0.6576 - val_accuracy: 0.6414\n",
      "Epoch 597/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6239 - val_loss: 0.6574 - val_accuracy: 0.6421\n",
      "Epoch 598/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6551 - accuracy: 0.6238 - val_loss: 0.6575 - val_accuracy: 0.6419\n",
      "Epoch 599/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6518 - accuracy: 0.6291 - val_loss: 0.6574 - val_accuracy: 0.6423\n",
      "Epoch 600/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6537 - accuracy: 0.6258 - val_loss: 0.6575 - val_accuracy: 0.6417\n",
      "Epoch 601/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6530 - accuracy: 0.6267 - val_loss: 0.6574 - val_accuracy: 0.6397\n",
      "Epoch 602/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6542 - accuracy: 0.6282 - val_loss: 0.6575 - val_accuracy: 0.6412\n",
      "Epoch 603/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6536 - accuracy: 0.6277 - val_loss: 0.6573 - val_accuracy: 0.6423\n",
      "Epoch 604/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6541 - accuracy: 0.6274 - val_loss: 0.6573 - val_accuracy: 0.6423\n",
      "Epoch 605/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6525 - accuracy: 0.6276 - val_loss: 0.6571 - val_accuracy: 0.6436\n",
      "Epoch 606/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6545 - accuracy: 0.6263 - val_loss: 0.6572 - val_accuracy: 0.6436\n",
      "Epoch 607/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6533 - accuracy: 0.6233 - val_loss: 0.6572 - val_accuracy: 0.6436\n",
      "Epoch 608/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6264 - val_loss: 0.6571 - val_accuracy: 0.6442\n",
      "Epoch 609/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6518 - accuracy: 0.6247 - val_loss: 0.6570 - val_accuracy: 0.6429\n",
      "Epoch 610/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6528 - accuracy: 0.6276 - val_loss: 0.6570 - val_accuracy: 0.6432\n",
      "Epoch 611/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6526 - accuracy: 0.6244 - val_loss: 0.6570 - val_accuracy: 0.6414\n",
      "Epoch 612/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6546 - accuracy: 0.6259 - val_loss: 0.6572 - val_accuracy: 0.6421\n",
      "Epoch 613/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6521 - accuracy: 0.6258 - val_loss: 0.6572 - val_accuracy: 0.6421\n",
      "Epoch 614/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6534 - accuracy: 0.6267 - val_loss: 0.6569 - val_accuracy: 0.6432\n",
      "Epoch 615/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6531 - accuracy: 0.6297 - val_loss: 0.6569 - val_accuracy: 0.6434\n",
      "Epoch 616/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6527 - accuracy: 0.6282 - val_loss: 0.6568 - val_accuracy: 0.6436\n",
      "Epoch 617/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6539 - accuracy: 0.6261 - val_loss: 0.6570 - val_accuracy: 0.6429\n",
      "Epoch 618/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6527 - accuracy: 0.6323 - val_loss: 0.6567 - val_accuracy: 0.6451\n",
      "Epoch 619/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6523 - accuracy: 0.6305 - val_loss: 0.6565 - val_accuracy: 0.6447\n",
      "Epoch 620/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6524 - accuracy: 0.6285 - val_loss: 0.6567 - val_accuracy: 0.6445\n",
      "Epoch 621/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6532 - accuracy: 0.6245 - val_loss: 0.6568 - val_accuracy: 0.6427\n",
      "Epoch 622/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6516 - accuracy: 0.6298 - val_loss: 0.6565 - val_accuracy: 0.6438\n",
      "Epoch 623/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6527 - accuracy: 0.6253 - val_loss: 0.6564 - val_accuracy: 0.6442\n",
      "Epoch 624/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6305 - val_loss: 0.6568 - val_accuracy: 0.6432\n",
      "Epoch 625/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6539 - accuracy: 0.6236 - val_loss: 0.6568 - val_accuracy: 0.6427\n",
      "Epoch 626/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6530 - accuracy: 0.6251 - val_loss: 0.6570 - val_accuracy: 0.6404\n",
      "Epoch 627/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6511 - accuracy: 0.6275 - val_loss: 0.6568 - val_accuracy: 0.6423\n",
      "Epoch 628/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6255 - val_loss: 0.6565 - val_accuracy: 0.6429\n",
      "Epoch 629/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6528 - accuracy: 0.6286 - val_loss: 0.6564 - val_accuracy: 0.6434\n",
      "Epoch 630/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6515 - accuracy: 0.6312 - val_loss: 0.6564 - val_accuracy: 0.6425\n",
      "Epoch 631/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6544 - accuracy: 0.6288 - val_loss: 0.6563 - val_accuracy: 0.6438\n",
      "Epoch 632/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6525 - accuracy: 0.6269 - val_loss: 0.6563 - val_accuracy: 0.6442\n",
      "Epoch 633/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6520 - accuracy: 0.6283 - val_loss: 0.6564 - val_accuracy: 0.6432\n",
      "Epoch 634/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6519 - accuracy: 0.6289 - val_loss: 0.6564 - val_accuracy: 0.6440\n",
      "Epoch 635/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6538 - accuracy: 0.6274 - val_loss: 0.6560 - val_accuracy: 0.6460\n",
      "Epoch 636/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6526 - accuracy: 0.6253 - val_loss: 0.6562 - val_accuracy: 0.6447\n",
      "Epoch 637/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6522 - accuracy: 0.6299 - val_loss: 0.6560 - val_accuracy: 0.6453\n",
      "Epoch 638/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6517 - accuracy: 0.6290 - val_loss: 0.6559 - val_accuracy: 0.6462\n",
      "Epoch 639/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6511 - accuracy: 0.6292 - val_loss: 0.6560 - val_accuracy: 0.6432\n",
      "Epoch 640/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6512 - accuracy: 0.6313 - val_loss: 0.6559 - val_accuracy: 0.6462\n",
      "Epoch 641/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6536 - accuracy: 0.6286 - val_loss: 0.6562 - val_accuracy: 0.6432\n",
      "Epoch 642/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6511 - accuracy: 0.6289 - val_loss: 0.6560 - val_accuracy: 0.6449\n",
      "Epoch 643/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6517 - accuracy: 0.6277 - val_loss: 0.6562 - val_accuracy: 0.6429\n",
      "Epoch 644/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6516 - accuracy: 0.6274 - val_loss: 0.6559 - val_accuracy: 0.6445\n",
      "Epoch 645/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6537 - accuracy: 0.6306 - val_loss: 0.6559 - val_accuracy: 0.6445\n",
      "Epoch 646/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6509 - accuracy: 0.6292 - val_loss: 0.6558 - val_accuracy: 0.6442\n",
      "Epoch 647/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6514 - accuracy: 0.6301 - val_loss: 0.6558 - val_accuracy: 0.6442\n",
      "Epoch 648/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6521 - accuracy: 0.6285 - val_loss: 0.6557 - val_accuracy: 0.6451\n",
      "Epoch 649/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6516 - accuracy: 0.6311 - val_loss: 0.6557 - val_accuracy: 0.6451\n",
      "Epoch 650/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6512 - accuracy: 0.6319 - val_loss: 0.6556 - val_accuracy: 0.6457\n",
      "Epoch 651/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6513 - accuracy: 0.6313 - val_loss: 0.6557 - val_accuracy: 0.6447\n",
      "Epoch 652/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6512 - accuracy: 0.6289 - val_loss: 0.6557 - val_accuracy: 0.6468\n",
      "Epoch 653/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6515 - accuracy: 0.6279 - val_loss: 0.6556 - val_accuracy: 0.6457\n",
      "Epoch 654/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6499 - accuracy: 0.6346 - val_loss: 0.6555 - val_accuracy: 0.6460\n",
      "Epoch 655/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6509 - accuracy: 0.6303 - val_loss: 0.6554 - val_accuracy: 0.6457\n",
      "Epoch 656/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6506 - accuracy: 0.6278 - val_loss: 0.6554 - val_accuracy: 0.6475\n",
      "Epoch 657/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6511 - accuracy: 0.6282 - val_loss: 0.6555 - val_accuracy: 0.6462\n",
      "Epoch 658/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6518 - accuracy: 0.6306 - val_loss: 0.6553 - val_accuracy: 0.6472\n",
      "Epoch 659/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6509 - accuracy: 0.6311 - val_loss: 0.6554 - val_accuracy: 0.6464\n",
      "Epoch 660/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6520 - accuracy: 0.6293 - val_loss: 0.6553 - val_accuracy: 0.6481\n",
      "Epoch 661/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6498 - accuracy: 0.6345 - val_loss: 0.6551 - val_accuracy: 0.6488\n",
      "Epoch 662/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6516 - accuracy: 0.6265 - val_loss: 0.6551 - val_accuracy: 0.6490\n",
      "Epoch 663/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6500 - accuracy: 0.6355 - val_loss: 0.6550 - val_accuracy: 0.6494\n",
      "Epoch 664/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6506 - accuracy: 0.6324 - val_loss: 0.6551 - val_accuracy: 0.6472\n",
      "Epoch 665/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6513 - accuracy: 0.6300 - val_loss: 0.6552 - val_accuracy: 0.6468\n",
      "Epoch 666/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6500 - accuracy: 0.6306 - val_loss: 0.6551 - val_accuracy: 0.6479\n",
      "Epoch 667/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6496 - accuracy: 0.6315 - val_loss: 0.6548 - val_accuracy: 0.6492\n",
      "Epoch 668/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6507 - accuracy: 0.6333 - val_loss: 0.6549 - val_accuracy: 0.6481\n",
      "Epoch 669/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6496 - accuracy: 0.6340 - val_loss: 0.6548 - val_accuracy: 0.6494\n",
      "Epoch 670/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6506 - accuracy: 0.6304 - val_loss: 0.6546 - val_accuracy: 0.6526\n",
      "Epoch 671/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6510 - accuracy: 0.6311 - val_loss: 0.6549 - val_accuracy: 0.6485\n",
      "Epoch 672/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6505 - accuracy: 0.6301 - val_loss: 0.6548 - val_accuracy: 0.6507\n",
      "Epoch 673/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6508 - accuracy: 0.6350 - val_loss: 0.6550 - val_accuracy: 0.6462\n",
      "Epoch 674/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6502 - accuracy: 0.6312 - val_loss: 0.6549 - val_accuracy: 0.6468\n",
      "Epoch 675/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6509 - accuracy: 0.6327 - val_loss: 0.6548 - val_accuracy: 0.6483\n",
      "Epoch 676/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6504 - accuracy: 0.6280 - val_loss: 0.6548 - val_accuracy: 0.6479\n",
      "Epoch 677/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6484 - accuracy: 0.6344 - val_loss: 0.6545 - val_accuracy: 0.6520\n",
      "Epoch 678/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6486 - accuracy: 0.6354 - val_loss: 0.6545 - val_accuracy: 0.6500\n",
      "Epoch 679/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6509 - accuracy: 0.6322 - val_loss: 0.6545 - val_accuracy: 0.6507\n",
      "Epoch 680/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6488 - accuracy: 0.6328 - val_loss: 0.6542 - val_accuracy: 0.6528\n",
      "Epoch 681/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6489 - accuracy: 0.6309 - val_loss: 0.6543 - val_accuracy: 0.6515\n",
      "Epoch 682/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6506 - accuracy: 0.6337 - val_loss: 0.6543 - val_accuracy: 0.6511\n",
      "Epoch 683/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6495 - accuracy: 0.6341 - val_loss: 0.6544 - val_accuracy: 0.6503\n",
      "Epoch 684/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6508 - accuracy: 0.6308 - val_loss: 0.6543 - val_accuracy: 0.6498\n",
      "Epoch 685/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6497 - accuracy: 0.6327 - val_loss: 0.6544 - val_accuracy: 0.6481\n",
      "Epoch 686/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6491 - accuracy: 0.6310 - val_loss: 0.6540 - val_accuracy: 0.6505\n",
      "Epoch 687/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6491 - accuracy: 0.6289 - val_loss: 0.6542 - val_accuracy: 0.6490\n",
      "Epoch 688/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6496 - accuracy: 0.6347 - val_loss: 0.6540 - val_accuracy: 0.6511\n",
      "Epoch 689/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6491 - accuracy: 0.6325 - val_loss: 0.6539 - val_accuracy: 0.6511\n",
      "Epoch 690/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6506 - accuracy: 0.6302 - val_loss: 0.6540 - val_accuracy: 0.6496\n",
      "Epoch 691/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6492 - accuracy: 0.6365 - val_loss: 0.6541 - val_accuracy: 0.6481\n",
      "Epoch 692/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6497 - accuracy: 0.6278 - val_loss: 0.6538 - val_accuracy: 0.6507\n",
      "Epoch 693/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6494 - accuracy: 0.6362 - val_loss: 0.6538 - val_accuracy: 0.6498\n",
      "Epoch 694/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6469 - accuracy: 0.6373 - val_loss: 0.6537 - val_accuracy: 0.6503\n",
      "Epoch 695/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6490 - accuracy: 0.6313 - val_loss: 0.6538 - val_accuracy: 0.6500\n",
      "Epoch 696/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6477 - accuracy: 0.6385 - val_loss: 0.6536 - val_accuracy: 0.6513\n",
      "Epoch 697/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6500 - accuracy: 0.6334 - val_loss: 0.6537 - val_accuracy: 0.6500\n",
      "Epoch 698/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6492 - accuracy: 0.6348 - val_loss: 0.6534 - val_accuracy: 0.6515\n",
      "Epoch 699/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6507 - accuracy: 0.6326 - val_loss: 0.6538 - val_accuracy: 0.6500\n",
      "Epoch 700/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6484 - accuracy: 0.6353 - val_loss: 0.6535 - val_accuracy: 0.6500\n",
      "Epoch 701/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6482 - accuracy: 0.6340 - val_loss: 0.6533 - val_accuracy: 0.6511\n",
      "Epoch 702/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6482 - accuracy: 0.6335 - val_loss: 0.6533 - val_accuracy: 0.6507\n",
      "Epoch 703/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6477 - accuracy: 0.6370 - val_loss: 0.6532 - val_accuracy: 0.6528\n",
      "Epoch 704/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6469 - accuracy: 0.6358 - val_loss: 0.6530 - val_accuracy: 0.6539\n",
      "Epoch 705/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6482 - accuracy: 0.6352 - val_loss: 0.6532 - val_accuracy: 0.6518\n",
      "Epoch 706/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6477 - accuracy: 0.6331 - val_loss: 0.6532 - val_accuracy: 0.6513\n",
      "Epoch 707/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6482 - accuracy: 0.6353 - val_loss: 0.6534 - val_accuracy: 0.6522\n",
      "Epoch 708/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6480 - accuracy: 0.6335 - val_loss: 0.6531 - val_accuracy: 0.6535\n",
      "Epoch 709/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6479 - accuracy: 0.6343 - val_loss: 0.6533 - val_accuracy: 0.6533\n",
      "Epoch 710/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6472 - accuracy: 0.6351 - val_loss: 0.6532 - val_accuracy: 0.6535\n",
      "Epoch 711/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6468 - accuracy: 0.6388 - val_loss: 0.6533 - val_accuracy: 0.6520\n",
      "Epoch 712/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6471 - accuracy: 0.6392 - val_loss: 0.6528 - val_accuracy: 0.6546\n",
      "Epoch 713/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6490 - accuracy: 0.6387 - val_loss: 0.6532 - val_accuracy: 0.6520\n",
      "Epoch 714/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6476 - accuracy: 0.6327 - val_loss: 0.6531 - val_accuracy: 0.6533\n",
      "Epoch 715/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6459 - accuracy: 0.6390 - val_loss: 0.6529 - val_accuracy: 0.6543\n",
      "Epoch 716/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6488 - accuracy: 0.6352 - val_loss: 0.6528 - val_accuracy: 0.6552\n",
      "Epoch 717/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6467 - accuracy: 0.6418 - val_loss: 0.6527 - val_accuracy: 0.6558\n",
      "Epoch 718/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6480 - accuracy: 0.6371 - val_loss: 0.6527 - val_accuracy: 0.6556\n",
      "Epoch 719/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6485 - accuracy: 0.6361 - val_loss: 0.6528 - val_accuracy: 0.6546\n",
      "Epoch 720/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6461 - accuracy: 0.6402 - val_loss: 0.6525 - val_accuracy: 0.6563\n",
      "Epoch 721/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6473 - accuracy: 0.6384 - val_loss: 0.6527 - val_accuracy: 0.6541\n",
      "Epoch 722/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6470 - accuracy: 0.6374 - val_loss: 0.6525 - val_accuracy: 0.6548\n",
      "Epoch 723/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6473 - accuracy: 0.6390 - val_loss: 0.6527 - val_accuracy: 0.6546\n",
      "Epoch 724/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6464 - accuracy: 0.6377 - val_loss: 0.6525 - val_accuracy: 0.6552\n",
      "Epoch 725/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6437 - accuracy: 0.6437 - val_loss: 0.6524 - val_accuracy: 0.6546\n",
      "Epoch 726/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6467 - accuracy: 0.6354 - val_loss: 0.6526 - val_accuracy: 0.6539\n",
      "Epoch 727/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6458 - accuracy: 0.6431 - val_loss: 0.6522 - val_accuracy: 0.6565\n",
      "Epoch 728/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6462 - accuracy: 0.6383 - val_loss: 0.6520 - val_accuracy: 0.6558\n",
      "Epoch 729/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6465 - accuracy: 0.6406 - val_loss: 0.6522 - val_accuracy: 0.6561\n",
      "Epoch 730/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6471 - accuracy: 0.6337 - val_loss: 0.6521 - val_accuracy: 0.6567\n",
      "Epoch 731/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6484 - accuracy: 0.6362 - val_loss: 0.6521 - val_accuracy: 0.6563\n",
      "Epoch 732/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6448 - accuracy: 0.6391 - val_loss: 0.6519 - val_accuracy: 0.6558\n",
      "Epoch 733/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6459 - accuracy: 0.6400 - val_loss: 0.6518 - val_accuracy: 0.6569\n",
      "Epoch 734/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6457 - accuracy: 0.6380 - val_loss: 0.6517 - val_accuracy: 0.6565\n",
      "Epoch 735/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6463 - accuracy: 0.6352 - val_loss: 0.6518 - val_accuracy: 0.6571\n",
      "Epoch 736/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6458 - accuracy: 0.6360 - val_loss: 0.6519 - val_accuracy: 0.6567\n",
      "Epoch 737/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6472 - accuracy: 0.6353 - val_loss: 0.6519 - val_accuracy: 0.6567\n",
      "Epoch 738/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6465 - accuracy: 0.6394 - val_loss: 0.6516 - val_accuracy: 0.6576\n",
      "Epoch 739/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6461 - accuracy: 0.6443 - val_loss: 0.6516 - val_accuracy: 0.6578\n",
      "Epoch 740/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6433 - accuracy: 0.6425 - val_loss: 0.6514 - val_accuracy: 0.6582\n",
      "Epoch 741/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6466 - accuracy: 0.6363 - val_loss: 0.6516 - val_accuracy: 0.6580\n",
      "Epoch 742/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6454 - accuracy: 0.6407 - val_loss: 0.6514 - val_accuracy: 0.6580\n",
      "Epoch 743/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6450 - accuracy: 0.6396 - val_loss: 0.6514 - val_accuracy: 0.6578\n",
      "Epoch 744/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6447 - accuracy: 0.6398 - val_loss: 0.6514 - val_accuracy: 0.6582\n",
      "Epoch 745/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6465 - accuracy: 0.6360 - val_loss: 0.6513 - val_accuracy: 0.6586\n",
      "Epoch 746/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6454 - accuracy: 0.6422 - val_loss: 0.6513 - val_accuracy: 0.6589\n",
      "Epoch 747/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6440 - accuracy: 0.6436 - val_loss: 0.6513 - val_accuracy: 0.6591\n",
      "Epoch 748/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6456 - accuracy: 0.6386 - val_loss: 0.6513 - val_accuracy: 0.6571\n",
      "Epoch 749/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6443 - accuracy: 0.6425 - val_loss: 0.6510 - val_accuracy: 0.6591\n",
      "Epoch 750/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6422 - accuracy: 0.6403 - val_loss: 0.6508 - val_accuracy: 0.6586\n",
      "Epoch 751/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6446 - accuracy: 0.6418 - val_loss: 0.6507 - val_accuracy: 0.6593\n",
      "Epoch 752/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6448 - accuracy: 0.6381 - val_loss: 0.6509 - val_accuracy: 0.6591\n",
      "Epoch 753/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6466 - accuracy: 0.6385 - val_loss: 0.6509 - val_accuracy: 0.6586\n",
      "Epoch 754/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6444 - accuracy: 0.6406 - val_loss: 0.6508 - val_accuracy: 0.6593\n",
      "Epoch 755/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6451 - accuracy: 0.6420 - val_loss: 0.6508 - val_accuracy: 0.6586\n",
      "Epoch 756/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6460 - accuracy: 0.6379 - val_loss: 0.6508 - val_accuracy: 0.6582\n",
      "Epoch 757/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6447 - accuracy: 0.6394 - val_loss: 0.6508 - val_accuracy: 0.6589\n",
      "Epoch 758/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6437 - accuracy: 0.6431 - val_loss: 0.6506 - val_accuracy: 0.6589\n",
      "Epoch 759/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6477 - accuracy: 0.6358 - val_loss: 0.6507 - val_accuracy: 0.6589\n",
      "Epoch 760/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6431 - accuracy: 0.6436 - val_loss: 0.6505 - val_accuracy: 0.6584\n",
      "Epoch 761/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6444 - accuracy: 0.6441 - val_loss: 0.6504 - val_accuracy: 0.6593\n",
      "Epoch 762/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6441 - accuracy: 0.6414 - val_loss: 0.6504 - val_accuracy: 0.6591\n",
      "Epoch 763/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6462 - accuracy: 0.6403 - val_loss: 0.6504 - val_accuracy: 0.6586\n",
      "Epoch 764/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6446 - accuracy: 0.6409 - val_loss: 0.6505 - val_accuracy: 0.6586\n",
      "Epoch 765/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6431 - accuracy: 0.6436 - val_loss: 0.6500 - val_accuracy: 0.6595\n",
      "Epoch 766/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6447 - accuracy: 0.6398 - val_loss: 0.6502 - val_accuracy: 0.6589\n",
      "Epoch 767/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6430 - accuracy: 0.6430 - val_loss: 0.6503 - val_accuracy: 0.6586\n",
      "Epoch 768/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6430 - accuracy: 0.6412 - val_loss: 0.6501 - val_accuracy: 0.6591\n",
      "Epoch 769/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6447 - accuracy: 0.6412 - val_loss: 0.6502 - val_accuracy: 0.6586\n",
      "Epoch 770/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6437 - accuracy: 0.6438 - val_loss: 0.6499 - val_accuracy: 0.6578\n",
      "Epoch 771/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6439 - accuracy: 0.6407 - val_loss: 0.6499 - val_accuracy: 0.6582\n",
      "Epoch 772/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6437 - accuracy: 0.6412 - val_loss: 0.6500 - val_accuracy: 0.6582\n",
      "Epoch 773/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6428 - accuracy: 0.6444 - val_loss: 0.6497 - val_accuracy: 0.6586\n",
      "Epoch 774/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6436 - accuracy: 0.6425 - val_loss: 0.6497 - val_accuracy: 0.6582\n",
      "Epoch 775/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6439 - accuracy: 0.6436 - val_loss: 0.6496 - val_accuracy: 0.6591\n",
      "Epoch 776/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6427 - accuracy: 0.6429 - val_loss: 0.6494 - val_accuracy: 0.6582\n",
      "Epoch 777/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6448 - accuracy: 0.6407 - val_loss: 0.6497 - val_accuracy: 0.6578\n",
      "Epoch 778/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6424 - accuracy: 0.6405 - val_loss: 0.6492 - val_accuracy: 0.6589\n",
      "Epoch 779/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6433 - accuracy: 0.6438 - val_loss: 0.6496 - val_accuracy: 0.6582\n",
      "Epoch 780/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6426 - accuracy: 0.6438 - val_loss: 0.6494 - val_accuracy: 0.6582\n",
      "Epoch 781/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6427 - accuracy: 0.6442 - val_loss: 0.6495 - val_accuracy: 0.6580\n",
      "Epoch 782/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6420 - accuracy: 0.6417 - val_loss: 0.6493 - val_accuracy: 0.6578\n",
      "Epoch 783/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6419 - accuracy: 0.6461 - val_loss: 0.6490 - val_accuracy: 0.6593\n",
      "Epoch 784/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6438 - accuracy: 0.6402 - val_loss: 0.6493 - val_accuracy: 0.6574\n",
      "Epoch 785/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6431 - accuracy: 0.6408 - val_loss: 0.6493 - val_accuracy: 0.6563\n",
      "Epoch 786/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6424 - accuracy: 0.6440 - val_loss: 0.6491 - val_accuracy: 0.6569\n",
      "Epoch 787/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6420 - accuracy: 0.6410 - val_loss: 0.6488 - val_accuracy: 0.6593\n",
      "Epoch 788/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6430 - accuracy: 0.6422 - val_loss: 0.6487 - val_accuracy: 0.6599\n",
      "Epoch 789/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6419 - accuracy: 0.6458 - val_loss: 0.6489 - val_accuracy: 0.6593\n",
      "Epoch 790/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6438 - accuracy: 0.6457 - val_loss: 0.6487 - val_accuracy: 0.6606\n",
      "Epoch 791/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6415 - accuracy: 0.6461 - val_loss: 0.6485 - val_accuracy: 0.6621\n",
      "Epoch 792/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6428 - accuracy: 0.6437 - val_loss: 0.6486 - val_accuracy: 0.6593\n",
      "Epoch 793/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6421 - accuracy: 0.6434 - val_loss: 0.6487 - val_accuracy: 0.6580\n",
      "Epoch 794/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6422 - accuracy: 0.6425 - val_loss: 0.6483 - val_accuracy: 0.6621\n",
      "Epoch 795/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6432 - accuracy: 0.6442 - val_loss: 0.6485 - val_accuracy: 0.6595\n",
      "Epoch 796/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6418 - accuracy: 0.6462 - val_loss: 0.6483 - val_accuracy: 0.6619\n",
      "Epoch 797/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6414 - accuracy: 0.6472 - val_loss: 0.6483 - val_accuracy: 0.6623\n",
      "Epoch 798/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6415 - accuracy: 0.6418 - val_loss: 0.6484 - val_accuracy: 0.6612\n",
      "Epoch 799/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6411 - accuracy: 0.6471 - val_loss: 0.6484 - val_accuracy: 0.6606\n",
      "Epoch 800/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6405 - accuracy: 0.6462 - val_loss: 0.6480 - val_accuracy: 0.6634\n",
      "Epoch 801/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6421 - accuracy: 0.6473 - val_loss: 0.6481 - val_accuracy: 0.6621\n",
      "Epoch 802/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6417 - accuracy: 0.6450 - val_loss: 0.6480 - val_accuracy: 0.6617\n",
      "Epoch 803/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6409 - accuracy: 0.6429 - val_loss: 0.6479 - val_accuracy: 0.6625\n",
      "Epoch 804/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6398 - accuracy: 0.6458 - val_loss: 0.6478 - val_accuracy: 0.6629\n",
      "Epoch 805/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6404 - accuracy: 0.6476 - val_loss: 0.6479 - val_accuracy: 0.6629\n",
      "Epoch 806/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6414 - accuracy: 0.6465 - val_loss: 0.6478 - val_accuracy: 0.6627\n",
      "Epoch 807/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6405 - accuracy: 0.6497 - val_loss: 0.6477 - val_accuracy: 0.6632\n",
      "Epoch 808/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6410 - accuracy: 0.6473 - val_loss: 0.6479 - val_accuracy: 0.6619\n",
      "Epoch 809/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6415 - accuracy: 0.6442 - val_loss: 0.6476 - val_accuracy: 0.6629\n",
      "Epoch 810/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6416 - accuracy: 0.6452 - val_loss: 0.6475 - val_accuracy: 0.6627\n",
      "Epoch 811/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6400 - accuracy: 0.6471 - val_loss: 0.6475 - val_accuracy: 0.6644\n",
      "Epoch 812/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6413 - accuracy: 0.6462 - val_loss: 0.6474 - val_accuracy: 0.6632\n",
      "Epoch 813/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6419 - accuracy: 0.6436 - val_loss: 0.6476 - val_accuracy: 0.6634\n",
      "Epoch 814/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6400 - accuracy: 0.6499 - val_loss: 0.6471 - val_accuracy: 0.6640\n",
      "Epoch 815/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6396 - accuracy: 0.6482 - val_loss: 0.6471 - val_accuracy: 0.6636\n",
      "Epoch 816/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6403 - accuracy: 0.6488 - val_loss: 0.6470 - val_accuracy: 0.6644\n",
      "Epoch 817/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6404 - accuracy: 0.6460 - val_loss: 0.6469 - val_accuracy: 0.6634\n",
      "Epoch 818/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6387 - accuracy: 0.6474 - val_loss: 0.6467 - val_accuracy: 0.6629\n",
      "Epoch 819/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6400 - accuracy: 0.6491 - val_loss: 0.6467 - val_accuracy: 0.6634\n",
      "Epoch 820/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6402 - accuracy: 0.6473 - val_loss: 0.6469 - val_accuracy: 0.6638\n",
      "Epoch 821/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6399 - accuracy: 0.6487 - val_loss: 0.6467 - val_accuracy: 0.6636\n",
      "Epoch 822/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6394 - accuracy: 0.6456 - val_loss: 0.6468 - val_accuracy: 0.6629\n",
      "Epoch 823/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6399 - accuracy: 0.6495 - val_loss: 0.6466 - val_accuracy: 0.6644\n",
      "Epoch 824/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6411 - accuracy: 0.6427 - val_loss: 0.6464 - val_accuracy: 0.6647\n",
      "Epoch 825/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6396 - accuracy: 0.6481 - val_loss: 0.6465 - val_accuracy: 0.6647\n",
      "Epoch 826/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6399 - accuracy: 0.6471 - val_loss: 0.6464 - val_accuracy: 0.6647\n",
      "Epoch 827/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6401 - accuracy: 0.6458 - val_loss: 0.6463 - val_accuracy: 0.6644\n",
      "Epoch 828/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6391 - accuracy: 0.6474 - val_loss: 0.6461 - val_accuracy: 0.6647\n",
      "Epoch 829/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6381 - accuracy: 0.6502 - val_loss: 0.6459 - val_accuracy: 0.6655\n",
      "Epoch 830/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6388 - accuracy: 0.6516 - val_loss: 0.6459 - val_accuracy: 0.6655\n",
      "Epoch 831/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6390 - accuracy: 0.6489 - val_loss: 0.6457 - val_accuracy: 0.6664\n",
      "Epoch 832/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6403 - accuracy: 0.6477 - val_loss: 0.6462 - val_accuracy: 0.6627\n",
      "Epoch 833/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6378 - accuracy: 0.6471 - val_loss: 0.6460 - val_accuracy: 0.6629\n",
      "Epoch 834/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6376 - accuracy: 0.6499 - val_loss: 0.6459 - val_accuracy: 0.6640\n",
      "Epoch 835/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6378 - accuracy: 0.6485 - val_loss: 0.6457 - val_accuracy: 0.6651\n",
      "Epoch 836/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6382 - accuracy: 0.6492 - val_loss: 0.6456 - val_accuracy: 0.6651\n",
      "Epoch 837/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6391 - accuracy: 0.6482 - val_loss: 0.6454 - val_accuracy: 0.6672\n",
      "Epoch 838/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6397 - accuracy: 0.6457 - val_loss: 0.6456 - val_accuracy: 0.6647\n",
      "Epoch 839/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6399 - accuracy: 0.6488 - val_loss: 0.6455 - val_accuracy: 0.6651\n",
      "Epoch 840/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6380 - accuracy: 0.6510 - val_loss: 0.6454 - val_accuracy: 0.6653\n",
      "Epoch 841/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6389 - accuracy: 0.6470 - val_loss: 0.6453 - val_accuracy: 0.6668\n",
      "Epoch 842/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6385 - accuracy: 0.6523 - val_loss: 0.6450 - val_accuracy: 0.6675\n",
      "Epoch 843/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6375 - accuracy: 0.6500 - val_loss: 0.6449 - val_accuracy: 0.6664\n",
      "Epoch 844/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6382 - accuracy: 0.6484 - val_loss: 0.6450 - val_accuracy: 0.6666\n",
      "Epoch 845/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6379 - accuracy: 0.6521 - val_loss: 0.6450 - val_accuracy: 0.6657\n",
      "Epoch 846/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6353 - accuracy: 0.6521 - val_loss: 0.6448 - val_accuracy: 0.6664\n",
      "Epoch 847/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6365 - accuracy: 0.6515 - val_loss: 0.6448 - val_accuracy: 0.6649\n",
      "Epoch 848/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6377 - accuracy: 0.6511 - val_loss: 0.6446 - val_accuracy: 0.6660\n",
      "Epoch 849/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6390 - accuracy: 0.6502 - val_loss: 0.6447 - val_accuracy: 0.6649\n",
      "Epoch 850/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6385 - accuracy: 0.6533 - val_loss: 0.6446 - val_accuracy: 0.6638\n",
      "Epoch 851/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6383 - accuracy: 0.6499 - val_loss: 0.6444 - val_accuracy: 0.6653\n",
      "Epoch 852/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6395 - accuracy: 0.6475 - val_loss: 0.6446 - val_accuracy: 0.6629\n",
      "Epoch 853/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6369 - accuracy: 0.6502 - val_loss: 0.6445 - val_accuracy: 0.6632\n",
      "Epoch 854/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6360 - accuracy: 0.6504 - val_loss: 0.6440 - val_accuracy: 0.6668\n",
      "Epoch 855/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6372 - accuracy: 0.6506 - val_loss: 0.6440 - val_accuracy: 0.6653\n",
      "Epoch 856/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6368 - accuracy: 0.6503 - val_loss: 0.6441 - val_accuracy: 0.6636\n",
      "Epoch 857/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6376 - accuracy: 0.6497 - val_loss: 0.6438 - val_accuracy: 0.6655\n",
      "Epoch 858/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6363 - accuracy: 0.6503 - val_loss: 0.6437 - val_accuracy: 0.6651\n",
      "Epoch 859/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6378 - accuracy: 0.6505 - val_loss: 0.6438 - val_accuracy: 0.6644\n",
      "Epoch 860/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6363 - accuracy: 0.6515 - val_loss: 0.6437 - val_accuracy: 0.6649\n",
      "Epoch 861/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6346 - accuracy: 0.6536 - val_loss: 0.6433 - val_accuracy: 0.6664\n",
      "Epoch 862/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6350 - accuracy: 0.6567 - val_loss: 0.6434 - val_accuracy: 0.6653\n",
      "Epoch 863/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6377 - accuracy: 0.6520 - val_loss: 0.6437 - val_accuracy: 0.6623\n",
      "Epoch 864/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6355 - accuracy: 0.6531 - val_loss: 0.6432 - val_accuracy: 0.6655\n",
      "Epoch 865/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6354 - accuracy: 0.6563 - val_loss: 0.6428 - val_accuracy: 0.6685\n",
      "Epoch 866/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6352 - accuracy: 0.6511 - val_loss: 0.6433 - val_accuracy: 0.6649\n",
      "Epoch 867/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6369 - accuracy: 0.6521 - val_loss: 0.6432 - val_accuracy: 0.6647\n",
      "Epoch 868/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6354 - accuracy: 0.6532 - val_loss: 0.6431 - val_accuracy: 0.6653\n",
      "Epoch 869/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6349 - accuracy: 0.6523 - val_loss: 0.6430 - val_accuracy: 0.6644\n",
      "Epoch 870/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6350 - accuracy: 0.6524 - val_loss: 0.6428 - val_accuracy: 0.6662\n",
      "Epoch 871/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6359 - accuracy: 0.6565 - val_loss: 0.6426 - val_accuracy: 0.6668\n",
      "Epoch 872/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6349 - accuracy: 0.6553 - val_loss: 0.6426 - val_accuracy: 0.6664\n",
      "Epoch 873/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6349 - accuracy: 0.6560 - val_loss: 0.6423 - val_accuracy: 0.6679\n",
      "Epoch 874/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6368 - accuracy: 0.6541 - val_loss: 0.6426 - val_accuracy: 0.6644\n",
      "Epoch 875/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6355 - accuracy: 0.6519 - val_loss: 0.6421 - val_accuracy: 0.6679\n",
      "Epoch 876/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6363 - accuracy: 0.6497 - val_loss: 0.6422 - val_accuracy: 0.6670\n",
      "Epoch 877/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6346 - accuracy: 0.6557 - val_loss: 0.6423 - val_accuracy: 0.6662\n",
      "Epoch 878/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6336 - accuracy: 0.6570 - val_loss: 0.6422 - val_accuracy: 0.6664\n",
      "Epoch 879/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6359 - accuracy: 0.6510 - val_loss: 0.6421 - val_accuracy: 0.6672\n",
      "Epoch 880/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6343 - accuracy: 0.6536 - val_loss: 0.6420 - val_accuracy: 0.6677\n",
      "Epoch 881/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6331 - accuracy: 0.6568 - val_loss: 0.6416 - val_accuracy: 0.6694\n",
      "Epoch 882/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6368 - accuracy: 0.6493 - val_loss: 0.6419 - val_accuracy: 0.6666\n",
      "Epoch 883/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6335 - accuracy: 0.6534 - val_loss: 0.6418 - val_accuracy: 0.6683\n",
      "Epoch 884/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6344 - accuracy: 0.6523 - val_loss: 0.6416 - val_accuracy: 0.6687\n",
      "Epoch 885/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6357 - accuracy: 0.6567 - val_loss: 0.6416 - val_accuracy: 0.6675\n",
      "Epoch 886/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6352 - accuracy: 0.6521 - val_loss: 0.6413 - val_accuracy: 0.6687\n",
      "Epoch 887/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6328 - accuracy: 0.6558 - val_loss: 0.6414 - val_accuracy: 0.6679\n",
      "Epoch 888/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6328 - accuracy: 0.6565 - val_loss: 0.6413 - val_accuracy: 0.6681\n",
      "Epoch 889/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6335 - accuracy: 0.6581 - val_loss: 0.6410 - val_accuracy: 0.6694\n",
      "Epoch 890/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6339 - accuracy: 0.6510 - val_loss: 0.6412 - val_accuracy: 0.6675\n",
      "Epoch 891/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6325 - accuracy: 0.6523 - val_loss: 0.6410 - val_accuracy: 0.6685\n",
      "Epoch 892/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6334 - accuracy: 0.6565 - val_loss: 0.6410 - val_accuracy: 0.6668\n",
      "Epoch 893/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6345 - accuracy: 0.6528 - val_loss: 0.6410 - val_accuracy: 0.6668\n",
      "Epoch 894/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6328 - accuracy: 0.6547 - val_loss: 0.6407 - val_accuracy: 0.6670\n",
      "Epoch 895/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6336 - accuracy: 0.6503 - val_loss: 0.6407 - val_accuracy: 0.6683\n",
      "Epoch 896/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6336 - accuracy: 0.6554 - val_loss: 0.6405 - val_accuracy: 0.6681\n",
      "Epoch 897/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6330 - accuracy: 0.6520 - val_loss: 0.6407 - val_accuracy: 0.6668\n",
      "Epoch 898/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6341 - accuracy: 0.6527 - val_loss: 0.6406 - val_accuracy: 0.6679\n",
      "Epoch 899/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6336 - accuracy: 0.6525 - val_loss: 0.6406 - val_accuracy: 0.6664\n",
      "Epoch 900/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6332 - accuracy: 0.6555 - val_loss: 0.6405 - val_accuracy: 0.6692\n",
      "Epoch 901/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6321 - accuracy: 0.6563 - val_loss: 0.6404 - val_accuracy: 0.6679\n",
      "Epoch 902/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6318 - accuracy: 0.6567 - val_loss: 0.6402 - val_accuracy: 0.6694\n",
      "Epoch 903/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6321 - accuracy: 0.6538 - val_loss: 0.6401 - val_accuracy: 0.6690\n",
      "Epoch 904/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6321 - accuracy: 0.6595 - val_loss: 0.6400 - val_accuracy: 0.6698\n",
      "Epoch 905/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6316 - accuracy: 0.6535 - val_loss: 0.6400 - val_accuracy: 0.6696\n",
      "Epoch 906/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6318 - accuracy: 0.6591 - val_loss: 0.6400 - val_accuracy: 0.6687\n",
      "Epoch 907/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6300 - accuracy: 0.6592 - val_loss: 0.6395 - val_accuracy: 0.6709\n",
      "Epoch 908/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6309 - accuracy: 0.6576 - val_loss: 0.6396 - val_accuracy: 0.6702\n",
      "Epoch 909/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6319 - accuracy: 0.6558 - val_loss: 0.6394 - val_accuracy: 0.6707\n",
      "Epoch 910/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6321 - accuracy: 0.6553 - val_loss: 0.6396 - val_accuracy: 0.6690\n",
      "Epoch 911/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6318 - accuracy: 0.6562 - val_loss: 0.6398 - val_accuracy: 0.6687\n",
      "Epoch 912/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6300 - accuracy: 0.6577 - val_loss: 0.6392 - val_accuracy: 0.6685\n",
      "Epoch 913/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6306 - accuracy: 0.6565 - val_loss: 0.6394 - val_accuracy: 0.6675\n",
      "Epoch 914/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6317 - accuracy: 0.6537 - val_loss: 0.6392 - val_accuracy: 0.6670\n",
      "Epoch 915/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6319 - accuracy: 0.6580 - val_loss: 0.6391 - val_accuracy: 0.6677\n",
      "Epoch 916/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6323 - accuracy: 0.6577 - val_loss: 0.6392 - val_accuracy: 0.6670\n",
      "Epoch 917/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6328 - accuracy: 0.6535 - val_loss: 0.6393 - val_accuracy: 0.6694\n",
      "Epoch 918/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6315 - accuracy: 0.6571 - val_loss: 0.6387 - val_accuracy: 0.6692\n",
      "Epoch 919/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6307 - accuracy: 0.6564 - val_loss: 0.6389 - val_accuracy: 0.6679\n",
      "Epoch 920/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6313 - accuracy: 0.6566 - val_loss: 0.6387 - val_accuracy: 0.6679\n",
      "Epoch 921/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6314 - accuracy: 0.6577 - val_loss: 0.6389 - val_accuracy: 0.6662\n",
      "Epoch 922/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6309 - accuracy: 0.6560 - val_loss: 0.6387 - val_accuracy: 0.6677\n",
      "Epoch 923/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6308 - accuracy: 0.6581 - val_loss: 0.6383 - val_accuracy: 0.6685\n",
      "Epoch 924/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6304 - accuracy: 0.6563 - val_loss: 0.6384 - val_accuracy: 0.6677\n",
      "Epoch 925/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6315 - accuracy: 0.6555 - val_loss: 0.6384 - val_accuracy: 0.6675\n",
      "Epoch 926/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6321 - accuracy: 0.6548 - val_loss: 0.6385 - val_accuracy: 0.6670\n",
      "Epoch 927/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6310 - accuracy: 0.6573 - val_loss: 0.6382 - val_accuracy: 0.6668\n",
      "Epoch 928/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6318 - accuracy: 0.6544 - val_loss: 0.6382 - val_accuracy: 0.6660\n",
      "Epoch 929/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6312 - accuracy: 0.6587 - val_loss: 0.6380 - val_accuracy: 0.6679\n",
      "Epoch 930/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6311 - accuracy: 0.6569 - val_loss: 0.6380 - val_accuracy: 0.6700\n",
      "Epoch 931/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6302 - accuracy: 0.6568 - val_loss: 0.6380 - val_accuracy: 0.6683\n",
      "Epoch 932/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6309 - accuracy: 0.6594 - val_loss: 0.6378 - val_accuracy: 0.6683\n",
      "Epoch 933/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6280 - accuracy: 0.6612 - val_loss: 0.6374 - val_accuracy: 0.6707\n",
      "Epoch 934/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6329 - accuracy: 0.6542 - val_loss: 0.6381 - val_accuracy: 0.6672\n",
      "Epoch 935/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6297 - accuracy: 0.6595 - val_loss: 0.6376 - val_accuracy: 0.6690\n",
      "Epoch 936/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6299 - accuracy: 0.6588 - val_loss: 0.6371 - val_accuracy: 0.6713\n",
      "Epoch 937/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6305 - accuracy: 0.6566 - val_loss: 0.6373 - val_accuracy: 0.6700\n",
      "Epoch 938/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6298 - accuracy: 0.6589 - val_loss: 0.6372 - val_accuracy: 0.6696\n",
      "Epoch 939/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6287 - accuracy: 0.6618 - val_loss: 0.6375 - val_accuracy: 0.6711\n",
      "Epoch 940/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6297 - accuracy: 0.6603 - val_loss: 0.6374 - val_accuracy: 0.6690\n",
      "Epoch 941/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6286 - accuracy: 0.6595 - val_loss: 0.6374 - val_accuracy: 0.6685\n",
      "Epoch 942/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6272 - accuracy: 0.6601 - val_loss: 0.6371 - val_accuracy: 0.6687\n",
      "Epoch 943/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6311 - accuracy: 0.6565 - val_loss: 0.6367 - val_accuracy: 0.6700\n",
      "Epoch 944/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6282 - accuracy: 0.6625 - val_loss: 0.6372 - val_accuracy: 0.6705\n",
      "Epoch 945/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6295 - accuracy: 0.6564 - val_loss: 0.6371 - val_accuracy: 0.6698\n",
      "Epoch 946/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6296 - accuracy: 0.6579 - val_loss: 0.6366 - val_accuracy: 0.6698\n",
      "Epoch 947/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6291 - accuracy: 0.6591 - val_loss: 0.6369 - val_accuracy: 0.6698\n",
      "Epoch 948/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6289 - accuracy: 0.6565 - val_loss: 0.6366 - val_accuracy: 0.6700\n",
      "Epoch 949/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6304 - accuracy: 0.6605 - val_loss: 0.6368 - val_accuracy: 0.6705\n",
      "Epoch 950/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6278 - accuracy: 0.6615 - val_loss: 0.6365 - val_accuracy: 0.6707\n",
      "Epoch 951/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6307 - accuracy: 0.6569 - val_loss: 0.6369 - val_accuracy: 0.6692\n",
      "Epoch 952/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6274 - accuracy: 0.6567 - val_loss: 0.6364 - val_accuracy: 0.6713\n",
      "Epoch 953/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6300 - accuracy: 0.6587 - val_loss: 0.6361 - val_accuracy: 0.6724\n",
      "Epoch 954/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6282 - accuracy: 0.6610 - val_loss: 0.6362 - val_accuracy: 0.6720\n",
      "Epoch 955/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6289 - accuracy: 0.6566 - val_loss: 0.6362 - val_accuracy: 0.6718\n",
      "Epoch 956/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6289 - accuracy: 0.6608 - val_loss: 0.6362 - val_accuracy: 0.6692\n",
      "Epoch 957/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6281 - accuracy: 0.6576 - val_loss: 0.6362 - val_accuracy: 0.6709\n",
      "Epoch 958/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6290 - accuracy: 0.6610 - val_loss: 0.6361 - val_accuracy: 0.6715\n",
      "Epoch 959/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6275 - accuracy: 0.6625 - val_loss: 0.6355 - val_accuracy: 0.6713\n",
      "Epoch 960/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6305 - accuracy: 0.6557 - val_loss: 0.6359 - val_accuracy: 0.6711\n",
      "Epoch 961/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6297 - accuracy: 0.6577 - val_loss: 0.6358 - val_accuracy: 0.6705\n",
      "Epoch 962/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6287 - accuracy: 0.6591 - val_loss: 0.6357 - val_accuracy: 0.6705\n",
      "Epoch 963/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6274 - accuracy: 0.6592 - val_loss: 0.6360 - val_accuracy: 0.6696\n",
      "Epoch 964/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6271 - accuracy: 0.6592 - val_loss: 0.6358 - val_accuracy: 0.6709\n",
      "Epoch 965/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6285 - accuracy: 0.6569 - val_loss: 0.6357 - val_accuracy: 0.6705\n",
      "Epoch 966/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6259 - accuracy: 0.6577 - val_loss: 0.6350 - val_accuracy: 0.6730\n",
      "Epoch 967/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6282 - accuracy: 0.6594 - val_loss: 0.6356 - val_accuracy: 0.6711\n",
      "Epoch 968/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6283 - accuracy: 0.6587 - val_loss: 0.6356 - val_accuracy: 0.6709\n",
      "Epoch 969/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6281 - accuracy: 0.6605 - val_loss: 0.6352 - val_accuracy: 0.6700\n",
      "Epoch 970/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6269 - accuracy: 0.6659 - val_loss: 0.6354 - val_accuracy: 0.6713\n",
      "Epoch 971/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6282 - accuracy: 0.6603 - val_loss: 0.6357 - val_accuracy: 0.6685\n",
      "Epoch 972/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6271 - accuracy: 0.6591 - val_loss: 0.6353 - val_accuracy: 0.6690\n",
      "Epoch 973/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6264 - accuracy: 0.6613 - val_loss: 0.6352 - val_accuracy: 0.6700\n",
      "Epoch 974/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6268 - accuracy: 0.6615 - val_loss: 0.6350 - val_accuracy: 0.6705\n",
      "Epoch 975/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6267 - accuracy: 0.6588 - val_loss: 0.6348 - val_accuracy: 0.6702\n",
      "Epoch 976/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6287 - accuracy: 0.6608 - val_loss: 0.6349 - val_accuracy: 0.6700\n",
      "Epoch 977/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6278 - accuracy: 0.6565 - val_loss: 0.6347 - val_accuracy: 0.6707\n",
      "Epoch 978/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6268 - accuracy: 0.6615 - val_loss: 0.6347 - val_accuracy: 0.6709\n",
      "Epoch 979/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6265 - accuracy: 0.6604 - val_loss: 0.6347 - val_accuracy: 0.6705\n",
      "Epoch 980/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6270 - accuracy: 0.6613 - val_loss: 0.6344 - val_accuracy: 0.6722\n",
      "Epoch 981/1000\n",
      "582/582 [==============================] - 13s 22ms/step - loss: 0.6281 - accuracy: 0.6595 - val_loss: 0.6345 - val_accuracy: 0.6702\n",
      "Epoch 982/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6275 - accuracy: 0.6622 - val_loss: 0.6348 - val_accuracy: 0.6687\n",
      "Epoch 983/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6274 - accuracy: 0.6594 - val_loss: 0.6346 - val_accuracy: 0.6696\n",
      "Epoch 984/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6269 - accuracy: 0.6615 - val_loss: 0.6344 - val_accuracy: 0.6702\n",
      "Epoch 985/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6272 - accuracy: 0.6607 - val_loss: 0.6346 - val_accuracy: 0.6690\n",
      "Epoch 986/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6263 - accuracy: 0.6583 - val_loss: 0.6342 - val_accuracy: 0.6709\n",
      "Epoch 987/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6248 - accuracy: 0.6643 - val_loss: 0.6341 - val_accuracy: 0.6713\n",
      "Epoch 988/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6271 - accuracy: 0.6602 - val_loss: 0.6339 - val_accuracy: 0.6713\n",
      "Epoch 989/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6256 - accuracy: 0.6608 - val_loss: 0.6340 - val_accuracy: 0.6715\n",
      "Epoch 990/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6265 - accuracy: 0.6598 - val_loss: 0.6343 - val_accuracy: 0.6696\n",
      "Epoch 991/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6279 - accuracy: 0.6580 - val_loss: 0.6340 - val_accuracy: 0.6711\n",
      "Epoch 992/1000\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6265 - accuracy: 0.6600 - val_loss: 0.6342 - val_accuracy: 0.6696\n",
      "Epoch 993/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6260 - accuracy: 0.6613 - val_loss: 0.6339 - val_accuracy: 0.6709\n",
      "Epoch 994/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6261 - accuracy: 0.6606 - val_loss: 0.6338 - val_accuracy: 0.6713\n",
      "Epoch 995/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6270 - accuracy: 0.6595 - val_loss: 0.6338 - val_accuracy: 0.6707\n",
      "Epoch 996/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6268 - accuracy: 0.6586 - val_loss: 0.6341 - val_accuracy: 0.6687\n",
      "Epoch 997/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6257 - accuracy: 0.6607 - val_loss: 0.6339 - val_accuracy: 0.6700\n",
      "Epoch 998/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6239 - accuracy: 0.6611 - val_loss: 0.6340 - val_accuracy: 0.6694\n",
      "Epoch 999/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6261 - accuracy: 0.6616 - val_loss: 0.6335 - val_accuracy: 0.6715\n",
      "Epoch 1000/1000\n",
      "582/582 [==============================] - 12s 21ms/step - loss: 0.6269 - accuracy: 0.6613 - val_loss: 0.6340 - val_accuracy: 0.6685\n"
     ]
    }
   ],
   "source": [
    "model_unnormalized_data_2 = ml_utils.build_model(optimizer=keras.optimizers.SGD(0.3e-4))\n",
    "es = keras.callbacks.EarlyStopping(monitor='val_loss', mode='min', verbose=1, patience=50)\n",
    "model_state_unnormalized_data_2 = ml_utils.train_model(\n",
    "    model_unnormalized_data_2,\n",
    "    train_unnormalized,\n",
    "    validation_unnormalized,\n",
    "    epochs=1000,\n",
    "    extra_callbacks=[es],\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "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": 4
}