Trajectory Prediction - LSTM - levelB1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "    Author: Eswar Lakshminarayanan"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Importing dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from datetime import datetime\n",
    "from random import sample\n",
    "import matplotlib.pyplot as plt \n",
    "from folium.plugins import HeatMap\n",
    "import folium\n",
    "import folium.plugins as plugins\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Random sampling"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "n = 1764258 #number of records in file\n",
    "s = 200000 #desired sample size\n",
    "skip = sorted(sample(range(1,n),n-s))\n",
    "df = pd.read_csv(\"new_levelB1_trajectory.csv\",header=0,skiprows=skip)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Filter really short and long trajectories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.loc[(df.seq_len>10) & (df.seq_len<1500)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(200000, 5)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Trajectory Length Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(8,5))\n",
    "df.seq_len.hist(bins =100)\n",
    "plt.xlabel(\"Length of Trajectory\",fontsize=26)\n",
    "plt.ylabel(\"Frequency\",fontsize=26)\n",
    "plt.title(\"Level B1\",fontsize=26)\n",
    "plt.xticks(np.arange(min(df.seq_len), max(df.seq_len)+1, 100),fontsize=20,rotation=70)\n",
    "plt.yticks(fontsize=16)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Extract Lat Lon from Trajectories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "lat_lng =[]\n",
    "#lat_lng_time = []\n",
    "counter = 0\n",
    "counter2 = 0\n",
    "def fn(row):\n",
    "    l1 = []\n",
    "    l2 = []\n",
    "    for each_recording in eval(row[\"loc_info\"]):\n",
    "        _,lat,lng,time = each_recording\n",
    "        l1.append([lat,lng])\n",
    "        #l2.append([lat,lng,time])\n",
    "    lat_lng.append(l1)\n",
    "    #lat_lng_time.append(l2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[[\"loc_info\"]].apply(lambda x: fn(x),axis=1)\n",
    "df=None"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "200000"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(lat_lng)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##  train split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = []\n",
    "Y_train = []\n",
    "\n",
    "for each_traj in lat_lng[0:5000]:\n",
    "    X_train.append(each_traj[:len(each_traj)-1])\n",
    "    Y_train.append(each_traj[-1])\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Model Architectire"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/opc/anaconda3/envs/tf_keras/lib/python3.7/site-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Colocations handled automatically by placer.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "lstm_1 (LSTM)                (None, None, 256)         265216    \n",
      "_________________________________________________________________\n",
      "lstm_2 (LSTM)                (None, None, 128)         197120    \n",
      "_________________________________________________________________\n",
      "lstm_3 (LSTM)                (None, None, 64)          49408     \n",
      "_________________________________________________________________\n",
      "lstm_4 (LSTM)                (None, 32)                12416     \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 2)                 66        \n",
      "=================================================================\n",
      "Total params: 524,226\n",
      "Trainable params: 524,226\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "from keras.models import Sequential\n",
    "from keras.layers import Dense, Dropout,TimeDistributed\n",
    "from keras.layers import Embedding\n",
    "from keras.layers import LSTM\n",
    "import numpy as np\n",
    "import keras.backend as K\n",
    "\n",
    "\n",
    "def euclidean_distance_loss(y_true, y_pred):\n",
    "    \"\"\"\n",
    "    Euclidean distance loss\n",
    "    :param y_true: TensorFlow/Theano tensor\n",
    "    :param y_pred: TensorFlow/Theano tensor of the same shape as y_true\n",
    "    :return: float\n",
    "    \"\"\"\n",
    "    euclidean_distance_loss.__name__ = \"Custom Loss\"\n",
    "    return K.sqrt(K.sum(K.square(y_pred - y_true), axis=-1))\n",
    "\n",
    "model = Sequential()\n",
    "model.add(LSTM(256, return_sequences=True,input_shape=(None, 2))) \n",
    "model.add(LSTM(128, return_sequences=True)) \n",
    "model.add(LSTM(64, return_sequences=True))\n",
    "model.add(LSTM(32))\n",
    "model.add(Dense(2))\n",
    "model.compile(loss=euclidean_distance_loss, optimizer='adam')\n",
    "model.summary()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Online Learning (Batch Size = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From /home/opc/anaconda3/envs/tf_keras/lib/python3.7/site-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "Use tf.cast instead.\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 103.8003\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 103.0727\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 102.6120\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 218ms/step - loss: 102.2236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 490ms/step - loss: 101.8763\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 103ms/step - loss: 101.5809\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 101.2730\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 978ms/step - loss: 101.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 182ms/step - loss: 100.7536\n",
      "# 10 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 154ms/step - loss: 100.5171\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 100.3013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 100.1282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 99.9331\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 99.7194\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 99.5364\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 207ms/step - loss: 99.3540\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 99.2215\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 519ms/step - loss: 98.9687\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 98.7666\n",
      "# 20 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 352ms/step - loss: 98.5584\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 98.3612\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 98.1667\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 97.9866\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 97.8341\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 805ms/step - loss: 97.6834\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 468ms/step - loss: 97.5382\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 97.4005\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 512ms/step - loss: 97.2628\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 97.1219\n",
      "# 30 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 400ms/step - loss: 96.9824\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 483ms/step - loss: 96.8517\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 96.7387\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 96.6428\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 934ms/step - loss: 96.5588\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 787ms/step - loss: 96.4765\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 283ms/step - loss: 96.3996\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 96.3315\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 474ms/step - loss: 96.2679\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 96.2082\n",
      "# 40 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 96.1486\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 96.0963\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 334ms/step - loss: 96.0454\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 494ms/step - loss: 95.9965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 288ms/step - loss: 95.9474\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 95.9027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 209ms/step - loss: 95.8586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 95.8147\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 223ms/step - loss: 95.7706\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 118ms/step - loss: 95.7265\n",
      "# 50 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 95.6825\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 440ms/step - loss: 95.6384\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 95.5942\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 95.5502\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 95.5061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 95.4619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 95.4176\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 301ms/step - loss: 95.3731\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 179ms/step - loss: 95.3287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 162ms/step - loss: 95.2840\n",
      "# 60 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 95.2393\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 95.1946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 311ms/step - loss: 95.1494\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 95.1068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 95.0674\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 575ms/step - loss: 95.0280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 157ms/step - loss: 94.9887\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 170ms/step - loss: 94.9495\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 94.9103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 94.8717\n",
      "# 70 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 94.8324\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 205ms/step - loss: 94.7936\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 445ms/step - loss: 94.7548\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 94.7167\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 94.6775\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 94.6390\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 94.6007\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 94.5625\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 695ms/step - loss: 94.5243\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 94.4866\n",
      "# 80 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 904ms/step - loss: 94.4484\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 94.4099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 158ms/step - loss: 94.3721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 94.3344\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 94.2966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 94.2589\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 94.2214\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 94.1839\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 94.1464\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 500ms/step - loss: 94.1090\n",
      "# 90 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 7s 7s/step - loss: 94.0718\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 810ms/step - loss: 94.0345\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 158ms/step - loss: 93.9973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 538ms/step - loss: 93.9601\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 427ms/step - loss: 93.9230\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 93.8860\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 93.8491\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 93.8122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 221ms/step - loss: 93.7753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 435ms/step - loss: 93.7385\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 100 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 487ms/step - loss: 93.7018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 93.6651\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 93.6283\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 93.5923\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 93.5556\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 93.5191\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 93.4824\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 406ms/step - loss: 93.4459\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 568ms/step - loss: 93.4095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 93.3731\n",
      "# 110 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 93.3369\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 93.3006\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 224ms/step - loss: 93.2643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 93.2282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 93.1921\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 122ms/step - loss: 93.1559\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 245ms/step - loss: 93.1198\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 93.0838\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 93.0477\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 93.0120\n",
      "# 120 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 92.9757\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 170ms/step - loss: 92.9398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 92.9040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 92.8683\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 92.8327\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 133ms/step - loss: 92.7965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 92.7608\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 409ms/step - loss: 92.7250\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 352ms/step - loss: 92.6893\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 136ms/step - loss: 92.6536\n",
      "# 130 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 205ms/step - loss: 92.6179\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 289ms/step - loss: 92.5823\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 781ms/step - loss: 92.5467\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 170ms/step - loss: 92.5111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 92.4755\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 596ms/step - loss: 92.4400\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 99ms/step - loss: 92.4045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 183ms/step - loss: 92.3695\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 92.3339\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 92.2986\n",
      "# 140 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 92.2628\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 92.2271\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 92.1926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 152ms/step - loss: 92.1565\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 92.1212\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 116ms/step - loss: 92.0859\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 285ms/step - loss: 92.0506\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 302ms/step - loss: 92.0153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 91.9800\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 195ms/step - loss: 91.9448\n",
      "# 150 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 196ms/step - loss: 91.9095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 91.8744\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 91.8394\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 91.8041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 91.7692\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 91.7336\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 91.6988\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 91.6633\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 420ms/step - loss: 91.6286\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 91.5933\n",
      "# 160 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 91.5582\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 91.5232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 91.4883\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 91.4531\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 91.4186\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 551ms/step - loss: 91.3831\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 336ms/step - loss: 91.3480\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 278ms/step - loss: 91.3131\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 91.2782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 91.2436\n",
      "# 170 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 91.2086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 91.1735\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 91.1391\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 540ms/step - loss: 91.1038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 91.0691\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 91.0343\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 310ms/step - loss: 90.9991\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 90.9645\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 533ms/step - loss: 90.9301\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 428ms/step - loss: 90.8952\n",
      "# 180 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 473ms/step - loss: 90.8604\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 90.8256\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 90.7911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 90.7559\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 90.7211\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 90.6866\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 90.6515\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 90.6174\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 549ms/step - loss: 90.5821\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 90.5481\n",
      "# 190 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 90.5129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 90.4782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 90.4436\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 90.4088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 668ms/step - loss: 90.3742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 90.3396\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 90.3050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 90.2704\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 90.2358\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 90.2012\n",
      "# 200 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 661ms/step - loss: 90.1670\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 90.1322\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 90.0978\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 90.0632\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 90.0286\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 404ms/step - loss: 89.9946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 89.9599\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 319ms/step - loss: 89.9252\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 89.8907\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 89.8562\n",
      "# 210 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 89.8218\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 89.7873\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 160ms/step - loss: 89.7525\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 451ms/step - loss: 89.7181\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 168ms/step - loss: 89.6836\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 364ms/step - loss: 89.6497\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 89.6149\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 89.5805\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 89.5461\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 493ms/step - loss: 89.5118\n",
      "# 220 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 89.4772\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 89.4429\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 89.4087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 718ms/step - loss: 89.3742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 89.3398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 89.3056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 7s 7s/step - loss: 89.2712\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 89.2368\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 567ms/step - loss: 89.2024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 314ms/step - loss: 89.1682\n",
      "# 230 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 89.1337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 89.0993\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 149ms/step - loss: 89.0652\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 322ms/step - loss: 89.0308\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 291ms/step - loss: 88.9965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 536ms/step - loss: 88.9623\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 354ms/step - loss: 88.9280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 331ms/step - loss: 88.8937\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 88.8595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 770ms/step - loss: 88.8252\n",
      "# 240 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 88.7908\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 709ms/step - loss: 88.7567\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 621ms/step - loss: 88.7229\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 885ms/step - loss: 88.6884\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 88.6548\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 88.6199\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 88.5855\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 88.5514\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 88.5170\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 88.4833\n",
      "# 250 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 88.4485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 161ms/step - loss: 88.4146\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 88.3804\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 88.3462\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 607ms/step - loss: 88.3119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 88.2781\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 262ms/step - loss: 88.2435\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 326ms/step - loss: 88.2099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 598ms/step - loss: 88.1753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 88.1413\n",
      "# 260 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 88.1071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 88.0730\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 88.0388\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 735ms/step - loss: 88.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 87.9707\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 87.9363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 899ms/step - loss: 87.9027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 87.8684\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 315ms/step - loss: 87.8344\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 87.8005\n",
      "# 270 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 87.7660\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 87.7323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.6978\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.6635\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 419ms/step - loss: 87.6298\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 87.5954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 154ms/step - loss: 87.5614\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.5274\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 87.4932\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 815ms/step - loss: 87.4592\n",
      "# 280 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 87.4252\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.3911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.3574\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 821ms/step - loss: 87.3232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 87.2894\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.2554\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 282ms/step - loss: 87.2212\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 87.1872\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 87.1530\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 351ms/step - loss: 87.1190\n",
      "# 290 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 87.0852\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 87.0516\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 87.0172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 86.9834\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 229ms/step - loss: 86.9493\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 86.9154\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 566ms/step - loss: 86.8813\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 86.8473\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 86.8137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 586ms/step - loss: 86.7792\n",
      "# 300 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 682ms/step - loss: 86.7452\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 86.7116\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 86.6775\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 337ms/step - loss: 86.6435\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 1s 518ms/step - loss: 86.6098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 319ms/step - loss: 86.5758\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 86.5418\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 86.5081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 86.4739\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 370ms/step - loss: 86.4403\n",
      "# 310 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 472ms/step - loss: 86.4062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 86.3722\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 86.3384\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 86.3044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 86.2710\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 86.2366\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 86.2026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 318ms/step - loss: 86.1686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 86.1349\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 86.1014\n",
      "# 320 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 86.0673\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 710ms/step - loss: 86.0336\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 267ms/step - loss: 85.9993\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 183ms/step - loss: 85.9654\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 85.9317\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 85.8979\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 85.8639\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 85.8300\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 85.7969\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 85.7623\n",
      "# 330 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 235ms/step - loss: 85.7285\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 260ms/step - loss: 85.6946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 85.6608\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 187ms/step - loss: 85.6269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 106ms/step - loss: 85.5931\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 303ms/step - loss: 85.5592\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 224ms/step - loss: 85.5254\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 85.4916\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 199ms/step - loss: 85.4577\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 85.4239\n",
      "# 340 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 85.3901\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 85.3563\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 85.3225\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 157ms/step - loss: 85.2886\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 356ms/step - loss: 85.2548\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 246ms/step - loss: 85.2210\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 85.1874\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 85.1535\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 85.1198\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 473ms/step - loss: 85.0860\n",
      "# 350 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 85.0522\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 232ms/step - loss: 85.0184\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 84.9847\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 84.9511\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 84.9170\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 84.8829\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 84.8494\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 84.8161\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 84.7819\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 84.7480\n",
      "# 360 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 84.7143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 84.6806\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 175ms/step - loss: 84.6468\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 165ms/step - loss: 84.6135\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 84.5794\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 84.5460\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 84.5119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 84.4782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 84.4442\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 975ms/step - loss: 84.4103\n",
      "# 370 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 84.3775\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 84.3429\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 84.3097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 84.2758\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 84.2419\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 84.2083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 84.1743\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 84.1407\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 118ms/step - loss: 84.1070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 226ms/step - loss: 84.0733\n",
      "# 380 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 84.0396\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 84.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 83.9721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 83.9384\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 174ms/step - loss: 83.9046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 83.8710\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 83.8374\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 83.8037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 739ms/step - loss: 83.7702\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 83.7366\n",
      "# 390 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 83.7031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 83.6688\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 83.6353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 172ms/step - loss: 83.6018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 83.5677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 83.5344\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 972ms/step - loss: 83.5005\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 83.4669\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 83.4331\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 971ms/step - loss: 83.3993\n",
      "# 400 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 83.3657\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 83.3322\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 148ms/step - loss: 83.2983\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 83.2648\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 83.2312\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 83.1976\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 59ms/step - loss: 83.1641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 83.1307\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 83.0964\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 83.0629\n",
      "# 410 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 83.0297\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 82.9958\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 82.9622\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 716ms/step - loss: 82.9284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 82.8951\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 82.8611\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 82.8275\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 82.7936\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 82.7599\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 82.7264\n",
      "# 420 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 915ms/step - loss: 82.6926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 563ms/step - loss: 82.6594\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 341ms/step - loss: 82.6255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 82.5924\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 82.5587\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 204ms/step - loss: 82.5247\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 82.4914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 82.4572\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 82.4244\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 106ms/step - loss: 82.3901\n",
      "# 430 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 82.3566\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 82.3232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 260ms/step - loss: 82.2895\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 82.2557\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 82.2221\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 82.1885\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 82.1552\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 447ms/step - loss: 82.1213\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 213ms/step - loss: 82.0877\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 82.0542\n",
      "# 440 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 82.0208\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 81.9868\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 81.9539\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 81.9200\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 81.8863\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 81.8526\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 81.8192\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 81.7858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 81.7523\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 81.7182\n",
      "# 450 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 81.6850\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 81.6512\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 81.6178\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 81.5840\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 81.5508\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 476ms/step - loss: 81.5170\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 81.4838\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 81.4499\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 81.4161\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 81.3828\n",
      "# 460 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 81.3490\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 81.3156\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 81.2823\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 81.2489\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 718ms/step - loss: 81.2153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 81.1818\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 81.1482\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 741ms/step - loss: 81.1143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 188ms/step - loss: 81.0807\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 81.0471\n",
      "# 470 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 81.0136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 80.9801\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 80.9465\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 121ms/step - loss: 80.9130\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 534ms/step - loss: 80.8795\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 388ms/step - loss: 80.8464\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 80.8127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 80.7793\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 964ms/step - loss: 80.7453\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 736ms/step - loss: 80.7115\n",
      "# 480 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 80.6783\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 80.6449\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 80.6113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 80.5781\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 80.5441\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 80.5108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 283ms/step - loss: 80.4771\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 80.4440\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 80.4102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 80.3765\n",
      "# 490 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 80.3431\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 80.3095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 80.2759\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 80.2426\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 80.2089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 80.1753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 80.1420\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 455ms/step - loss: 80.1088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 171ms/step - loss: 80.0749\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 80.0414\n",
      "# 500 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 80.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 79.9748\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 495ms/step - loss: 79.9414\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 79.9075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 138ms/step - loss: 79.8737\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 79.8405\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 79.8070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 79.7736\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 1s 1s/step - loss: 79.7398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 79.7068\n",
      "# 510 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 79.6731\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 79.6396\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 79.6065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 79.5726\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 450ms/step - loss: 79.5391\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 79.5056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 79.4721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 79.4386\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 79.4053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 79.3715\n",
      "# 520 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 79.3384\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 79.3045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 79.2713\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 79.2377\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 79.2043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 79.1708\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 79.1373\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 79.1039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 297ms/step - loss: 79.0705\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 785ms/step - loss: 79.0370\n",
      "# 530 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 79.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 935ms/step - loss: 78.9701\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 78.9364\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 616ms/step - loss: 78.9032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 992ms/step - loss: 78.8700\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 211ms/step - loss: 78.8363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 78.8027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 78.7695\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 78.7358\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 78.7026\n",
      "# 540 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 78.6687\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 468ms/step - loss: 78.6354\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 581ms/step - loss: 78.6019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 239ms/step - loss: 78.5683\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 78.5349\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 78.5019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 78.4681\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 78.4347\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 78.4015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 78.3684\n",
      "# 550 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 78.3348\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 522ms/step - loss: 78.3010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 78.2677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 78.2342\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 78.2006\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 78.1672\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 212ms/step - loss: 78.1337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 150ms/step - loss: 78.1004\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 320ms/step - loss: 78.0670\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 78.0336\n",
      "# 560 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 78.0002\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 77.9667\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 77.9333\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 426ms/step - loss: 77.8997\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 77.8664\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 77.8332\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 77.7992\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 77.7660\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.7325\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 77.6991\n",
      "# 570 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 795ms/step - loss: 77.6660\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 77.6323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 77.5989\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.5654\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.5320\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.4986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 280ms/step - loss: 77.4652\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 77.4318\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.3984\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 77.3650\n",
      "# 580 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 77.3317\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 77.2987\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 77.2647\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 77.2313\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 77.1984\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 77.1648\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 77.1313\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 77.0979\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 77.0647\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 77.0311\n",
      "# 590 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 76.9976\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 307ms/step - loss: 76.9643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 76.9311\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 76.8975\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 76.8643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 76.8307\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 76.7971\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 76.7636\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 76.7303\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 76.6974\n",
      "# 600 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 275ms/step - loss: 76.6636\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 76.6304\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 76.5973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 76.5633\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 76.5301\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 76.4966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 76.4635\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 76.4296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 76.3968\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 773ms/step - loss: 76.3633\n",
      "# 610 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 76.3297\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 76.2963\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 76.2629\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 76.2296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 76.1961\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 76.1627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 76.1295\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 76.0961\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 76.0625\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 76.0295\n",
      "# 620 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 75.9960\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 830ms/step - loss: 75.9626\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 75.9290\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 75.8956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 75.8624\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 75.8292\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 486ms/step - loss: 75.7956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 183ms/step - loss: 75.7621\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 75.7287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 75.6953\n",
      "# 630 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 75.6619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 867ms/step - loss: 75.6286\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 75.5959\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 75.5620\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 718ms/step - loss: 75.5287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 75.4954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 75.4621\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 75.4289\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 75.3954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 75.3620\n",
      "# 640 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 75.3284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 75.2956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 75.2618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 278ms/step - loss: 75.2282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 204ms/step - loss: 75.1950\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 75.1618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 75.1283\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 75.0949\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 75.0616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 75.0283\n",
      "# 650 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 74.9952\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 74.9615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 95ms/step - loss: 74.9282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 74.8949\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 74.8619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 74.8285\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 74.7948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 74.7613\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 74.7282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 74.6947\n",
      "# 660 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 74.6617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 74.6279\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 74.5948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 608ms/step - loss: 74.5617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 74.5284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 74.4948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 74.4617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 74.4280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 74.3950\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 74.3616\n",
      "# 670 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 74.3284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 103ms/step - loss: 74.2948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 74.2612\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 74.2282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 74.1948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 74.1614\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 74.1280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 74.0946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 74.0616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 74.0279\n",
      "# 680 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 73.9949\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 73.9614\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 73.9280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 140ms/step - loss: 73.8947\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 73.8615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 306ms/step - loss: 73.8281\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 73.7947\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 73.7617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 73.7285\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 73.6947\n",
      "# 690 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 73.6616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 73.6281\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 990ms/step - loss: 73.5946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 73.5614\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 73.5281\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 73.4947\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 73.4617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 73.4282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 73.3948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 73.3614\n",
      "# 700 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 73.3281\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 73.2952\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 73.2617\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 73.2287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 73.1954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 73.1616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 261ms/step - loss: 73.1283\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 73.0948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 73.0620\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 73.0283\n",
      "# 710 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 72.9949\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 72.9618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 72.9288\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 72.8954\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 344ms/step - loss: 72.8622\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 72.8285\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 72.7955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 952ms/step - loss: 72.7619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 72.7286\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 116ms/step - loss: 72.6952\n",
      "# 720 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 72.6623\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 72.6289\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 72.5955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 72.5621\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 72.5290\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 72.4954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 72.4624\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 167ms/step - loss: 72.4288\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 72.3956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 72.3623\n",
      "# 730 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 72.3290\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 72.2956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 72.2627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 72.2291\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 72.1957\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 72.1628\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 347ms/step - loss: 72.1292\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 72.0957\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 72.0630\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 72.0293\n",
      "# 740 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 71.9959\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 71.9627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 414ms/step - loss: 71.9295\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 608ms/step - loss: 71.8962\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 837ms/step - loss: 71.8627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 71.8294\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 71.7962\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 71.7631\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 71.7299\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 71.6965\n",
      "# 750 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 71.6633\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 158ms/step - loss: 71.6299\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 71.5966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 71.5631\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 71.5297\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 71.4966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 71.4637\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 71.4301\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 71.3968\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 71.3635\n",
      "# 760 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 71.3302\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 71.2970\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 71.2637\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 332ms/step - loss: 71.2306\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 71.1974\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 71.1642\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 71.1307\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 71.0975\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 71.0640\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 71.0307\n",
      "# 770 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 70.9975\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 70.9643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 491ms/step - loss: 70.9308\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 495ms/step - loss: 70.8978\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 70.8643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 70.8310\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 175ms/step - loss: 70.7978\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 70.7645\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 70.7312\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 70.6980\n",
      "# 780 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 177ms/step - loss: 70.6647\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 190ms/step - loss: 70.6314\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 70.5981\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 70.5649\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 70.5316\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 345ms/step - loss: 70.4983\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 70.4651\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 70.4323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 548ms/step - loss: 70.3986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 70.3652\n",
      "# 790 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 70.3320\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 70.2986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 70.2657\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 70.2324\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 70.1989\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 70.1657\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 70.1325\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 70.0994\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 70.0660\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 70.0328\n",
      "# 800 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 69.9995\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 69.9661\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 69.9330\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 69.8999\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 69.8666\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 954ms/step - loss: 69.8335\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 69.8002\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 808ms/step - loss: 69.7666\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 69.7334\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 69.7000\n",
      "# 810 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 69.6671\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 69.6335\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 69.6007\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 69.5671\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 154ms/step - loss: 69.5338\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 189ms/step - loss: 69.5005\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 69.4672\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 145ms/step - loss: 69.4340\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 69.4007\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 218ms/step - loss: 69.3675\n",
      "# 820 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 69.3342\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 69.3010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 148ms/step - loss: 69.2677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 69.2345\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 116ms/step - loss: 69.2012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 69.1679\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 69.1348\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 69.1015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 69.0680\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 69.0350\n",
      "# 830 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 348ms/step - loss: 69.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 659ms/step - loss: 68.9686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 68.9353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 254ms/step - loss: 68.9021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 68.8689\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 68.8354\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 68.8023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 68.7690\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 68.7357\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 68.7025\n",
      "# 840 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 68.6693\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 495ms/step - loss: 68.6360\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 68.6028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 68.5695\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 155ms/step - loss: 68.5363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 68.5030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 68.4698\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 68.4369\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 68.4035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 453ms/step - loss: 68.3706\n",
      "# 850 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 712ms/step - loss: 68.3367\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 916ms/step - loss: 68.3037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 68.2704\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 68.2374\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 354ms/step - loss: 68.2041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 150ms/step - loss: 68.1707\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 68.1372\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 68.1042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 68.0710\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 237ms/step - loss: 68.0377\n",
      "# 860 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 508ms/step - loss: 68.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 746ms/step - loss: 67.9712\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 67.9380\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 793ms/step - loss: 67.9048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 900ms/step - loss: 67.8715\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 882ms/step - loss: 67.8383\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 67.8051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 424ms/step - loss: 67.7718\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 67.7386\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 67.7055\n",
      "# 870 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 67.6721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 67.6389\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 67.6055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 67.5723\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 67.5395\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 67.5060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 67.4727\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 67.4395\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 67.4064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 67.3732\n",
      "# 880 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 67.3403\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 67.3064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 67.2733\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 67.2401\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 67.2067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 67.1738\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 67.1405\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 67.1073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 67.0742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 67.0411\n",
      "# 890 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 208ms/step - loss: 67.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 66.9742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 66.9412\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 66.9081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 281ms/step - loss: 66.8746\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 66.8415\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 389ms/step - loss: 66.8082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 66.7753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 66.7421\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 66.7084\n",
      "# 900 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 66.6754\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 66.6422\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 66.6090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 66.5759\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 66.5424\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 66.5097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 66.4760\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 66.4430\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 66.4097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 334ms/step - loss: 66.3763\n",
      "# 910 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 66.3432\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 314ms/step - loss: 66.3105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 66.2768\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 66.2439\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 66.2102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 66.1772\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 66.1438\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 66.1105\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 66.0778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 156ms/step - loss: 66.0445\n",
      "# 920 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 66.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 348ms/step - loss: 65.9778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 65.9446\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 65.9114\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 65.8782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 65.8450\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 65.8118\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 65.7788\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 65.7457\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 456ms/step - loss: 65.7122\n",
      "# 930 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 65.6790\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 65.6458\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 367ms/step - loss: 65.6125\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 65.5793\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 153ms/step - loss: 65.5461\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 65.5129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 65.4796\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 65.4465\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 65.4135\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 65.3805\n",
      "# 940 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 65.3470\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 65.3136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 65.2806\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 65.2473\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 65.2141\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 65.1814\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 65.1476\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 65.1143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 65.0816\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 65.0480\n",
      "# 950 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 65.0149\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 64.9822\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 64.9485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 64.9155\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 64.8821\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 64.8490\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 64.8157\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 64.7825\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 64.7491\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 64.7162\n",
      "# 960 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 64.6831\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 64.6499\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 64.6168\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 64.5832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 64.5504\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 64.5169\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 64.4837\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 64.4507\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 64.4177\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 64.3842\n",
      "# 970 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 64.3511\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 64.3178\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 64.2843\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 64.2512\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 64.2182\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 64.1848\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 64.1517\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 64.1190\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 64.0856\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 142ms/step - loss: 64.0525\n",
      "# 980 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 122ms/step - loss: 64.0188\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 63.9858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 63.9526\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 63.9197\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 103ms/step - loss: 63.8863\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 63.8530\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 63.8198\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 63.7865\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 63.7534\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 63.7203\n",
      "# 990 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 63.6871\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 63.6538\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 63.6207\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 63.5874\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 63.5544\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 63.5213\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 814ms/step - loss: 63.4881\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 63.4548\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 63.4217\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 63.3885\n",
      "# 1000 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 63.3551\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 63.3220\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 63.2886\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 63.2558\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 63.2223\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 63.1892\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 63.1561\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 63.1227\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 63.0902\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 63.0569\n",
      "# 1010 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 63.0234\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 62.9902\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 62.9572\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 62.9241\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 637ms/step - loss: 62.8910\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 292ms/step - loss: 62.8573\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 62.8245\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 62.7911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 62.7582\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 62.7246\n",
      "# 1020 examples trained\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 59ms/step - loss: 62.6914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 62.6583\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 632ms/step - loss: 62.6251\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 62.5919\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 62.5589\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 62.5255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 407ms/step - loss: 62.4928\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 766ms/step - loss: 62.4595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 328ms/step - loss: 62.4264\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 62.3934\n",
      "# 1030 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 62.3595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 62.3266\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 62.2935\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 62.2604\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 62.2271\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 62.1944\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 62.1607\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 62.1273\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 62.0942\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 62.0613\n",
      "# 1040 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 62.0278\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 287ms/step - loss: 61.9947\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 61.9619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 61.9283\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 61.8956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 61.8626\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 61.8289\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 61.7962\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 61.7626\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 61.7296\n",
      "# 1050 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 61.6962\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 61.6632\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 61.6303\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 61.5968\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 61.5638\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 61.5304\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 61.4971\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 61.4640\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 546ms/step - loss: 61.4308\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 61.3981\n",
      "# 1060 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 61.3645\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 61.3314\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 61.2983\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 61.2651\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 61.2323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 61.1987\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 61.1660\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 61.1325\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 61.0990\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 61.0662\n",
      "# 1070 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 61.0329\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 60.9997\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 60.9665\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 60.9334\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 60.9006\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 60.8672\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 60.8338\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 60.8012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 60.7677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 60.7347\n",
      "# 1080 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 186ms/step - loss: 60.7012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 213ms/step - loss: 60.6680\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 60.6349\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 60.6017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 271ms/step - loss: 60.5685\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 60.5354\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 60.5024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 60.4691\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 60.4359\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 60.4026\n",
      "# 1090 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 60.3697\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 60.3366\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 60.3033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 60.2705\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 60.2369\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 60.2037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 60.1708\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 162ms/step - loss: 60.1376\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 60.1042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 60.0712\n",
      "# 1100 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 60.0380\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 60.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 59.9715\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 59.9384\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 59.9057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 59.8721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 59.8391\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 59.8060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 59.7729\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 59.7396\n",
      "# 1110 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 59.7063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 59.6734\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 59.6402\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 59.6069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 59.5740\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 579ms/step - loss: 59.5408\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 59.5077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 59.4743\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 59.4412\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 59.4080\n",
      "# 1120 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 59.3749\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 59.3419\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 516ms/step - loss: 59.3089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 59.2754\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 59.2424\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 59.2091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 59.1759\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 59.1431\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 59.1096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 59.0767\n",
      "# 1130 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 59.0433\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 59.0102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 58.9770\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 58.9441\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 58.9110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 58.8776\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 58.8447\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 576ms/step - loss: 58.8116\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 58.7780\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 58.7450\n",
      "# 1140 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 58.7121\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 58.6787\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 58.6453\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 58.6124\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 58.5792\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 58.5464\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 58.5132\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 58.4798\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 58.4467\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 58.4133\n",
      "# 1150 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 58.3805\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 58.3473\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 58.3143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 58.2807\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 58.2478\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 58.2150\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 58.1814\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 58.1484\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 58.1152\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 58.0820\n",
      "# 1160 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 58.0489\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 58.0156\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 57.9825\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 57.9496\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 57.9168\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 57.8831\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 57.8500\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 57.8171\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 57.7837\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 57.7507\n",
      "# 1170 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 57.7176\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 57.6844\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 57.6512\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 57.6180\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 57.5851\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 57.5517\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 57.5186\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 57.4858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 57.4522\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 57.4190\n",
      "# 1180 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 57.3858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 57.3529\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 57.3197\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 336ms/step - loss: 57.2865\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 57.2534\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 57.2204\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 57.1871\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 57.1540\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 57.1208\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 57.0875\n",
      "# 1190 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 57.0546\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 57.0215\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 56.9887\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 56.9552\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 56.9220\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 56.8890\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 56.8555\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 56.8228\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 255ms/step - loss: 56.7895\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 56.7563\n",
      "# 1200 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 56.7232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 56.6902\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 56.6569\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 56.6236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 56.5907\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 56.5576\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 56.5241\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 56.4913\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 56.4583\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 56.4250\n",
      "# 1210 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 56.3921\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 56.3587\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 56.3255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 784ms/step - loss: 56.2926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 56.2592\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 56.2261\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 56.1931\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 56.1599\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 56.1266\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 56.0938\n",
      "# 1220 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 56.0603\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 56.0277\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 55.9942\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 55.9613\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 147ms/step - loss: 55.9282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 55.8949\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 55.8615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 55.8287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 55.7955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 55.7624\n",
      "# 1230 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 55.7292\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 55.6961\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 55.6633\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 55.6298\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 55.5965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 55.5636\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 55.5307\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 745ms/step - loss: 55.4973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 55.4642\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 55.4308\n",
      "# 1240 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 55.3982\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 55.3647\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 55.3316\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 55.2985\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 55.2654\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 55.2326\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 55.1991\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 55.1658\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 55.1329\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 55.0995\n",
      "# 1250 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 55.0671\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 55.0337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 55.0006\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 54.9671\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 54.9345\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 54.9011\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 54.8679\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 54.8348\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 54.8016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 54.7684\n",
      "# 1260 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 54.7356\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 338ms/step - loss: 54.7026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 54.6689\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 54.6362\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 54.6028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 54.5696\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 373ms/step - loss: 54.5365\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 54.5037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 54.4705\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 54.4372\n",
      "# 1270 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 54.4041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 458ms/step - loss: 54.3713\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 282ms/step - loss: 54.3383\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 531ms/step - loss: 54.3051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 275ms/step - loss: 54.2720\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 54.2388\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 54.2051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 54.1726\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 54.1392\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 54.1062\n",
      "# 1280 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 54.0730\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 54.0398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 54.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 991ms/step - loss: 53.9733\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 53.9405\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 53.9070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 53.8741\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 53.8408\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 53.8078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 53.7746\n",
      "# 1290 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 188ms/step - loss: 53.7417\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 53.7086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 53.6755\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 53.6421\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 53.6090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 180ms/step - loss: 53.5758\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 53.5432\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 53.5097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 53.4765\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 53.4436\n",
      "# 1300 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 53.4106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 53.3774\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 53.3446\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 53.3110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 53.2778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 53.2448\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 53.2117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 53.1784\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 53.1457\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 53.1125\n",
      "# 1310 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 53.0792\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 53.0462\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 53.0130\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 52.9798\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 52.9466\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 52.9141\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 52.8804\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 52.8476\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 52.8143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 145ms/step - loss: 52.7815\n",
      "# 1320 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 52.7485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 52.7151\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 52.6818\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 52.6487\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 52.6156\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 52.5823\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 52.5496\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 52.5162\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 52.4832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 52.4500\n",
      "# 1330 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 52.4169\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 52.3839\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 52.3509\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 52.3174\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 52.2844\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 52.2515\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 52.2182\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 52.1857\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 166ms/step - loss: 52.1520\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 52.1193\n",
      "# 1340 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 52.0859\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 52.0528\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 52.0196\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 51.9866\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 51.9534\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 51.9202\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 51.8869\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 51.8539\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 51.8209\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 51.7878\n",
      "# 1350 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 51.7547\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 51.7217\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 51.6882\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 51.6556\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 51.6226\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 51.5892\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 51.5565\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 154ms/step - loss: 51.5229\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 51.4897\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 51.4566\n",
      "# 1360 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 51.4235\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 51.3904\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 51.3573\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 263ms/step - loss: 51.3242\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 493ms/step - loss: 51.2911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 51.2580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 51.2249\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 51.1917\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 51.1591\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 51.1257\n",
      "# 1370 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 51.0929\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 51.0595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 51.0263\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 50.9932\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 50.9601\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 50.9268\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 50.8936\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 50.8608\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 50.8274\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 50.7950\n",
      "# 1380 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 50.7614\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 50.7282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 50.6951\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 50.6622\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 50.6290\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 50.5958\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 50.5632\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 50.5296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 50.4965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 50.4634\n",
      "# 1390 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 50.4306\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 107ms/step - loss: 50.3976\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 50.3641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 50.3309\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 50.2979\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 50.2646\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 50.2319\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 50.1987\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 734ms/step - loss: 50.1656\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 50.1325\n",
      "# 1400 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 50.0993\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 50.0662\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 50.0332\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 50.0000\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 49.9670\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 714ms/step - loss: 49.9335\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 49.9005\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 49.8676\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 49.8345\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 49.8013\n",
      "# 1410 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 204ms/step - loss: 49.7681\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 49.7352\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 49.7019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 49.6690\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 49.6358\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 49.6028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 49.5696\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 49.5364\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 49.5034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 49.4703\n",
      "# 1420 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 49.4372\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 49.4043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 49.3711\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 49.3380\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 49.3048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 49.2717\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 49.2389\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 49.2053\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 49.1725\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 49.1398\n",
      "# 1430 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 49.1062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 49.0729\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 49.0400\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 49.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 48.9738\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 48.9408\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 48.9079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 48.8744\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 48.8415\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 48.8084\n",
      "# 1440 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 48.7753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 197ms/step - loss: 48.7420\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 48.7091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 48.6759\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 48.6428\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 48.6097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 48.5766\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 48.5434\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 48.5104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 48.4773\n",
      "# 1450 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 48.4443\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 48.4113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 48.3778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 48.3447\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 48.3119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 48.2792\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 48.2458\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 48.2129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 48.1793\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 48.1467\n",
      "# 1460 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 48.1134\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 48.0803\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 48.0470\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 48.0137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 47.9812\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 47.9478\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 47.9148\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 47.8815\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 47.8485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 47.8156\n",
      "# 1470 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 47.7826\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 289ms/step - loss: 47.7495\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 47.7160\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 47.6832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 47.6502\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 47.6169\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 780ms/step - loss: 47.5840\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 47.5509\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 47.5176\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 47.4845\n",
      "# 1480 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 47.4515\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 47.4187\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 47.3852\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 47.3519\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 47.3190\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 47.2857\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 47.2529\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 47.2197\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 47.1869\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 47.1535\n",
      "# 1490 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 146ms/step - loss: 47.1204\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 121ms/step - loss: 47.0873\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 47.0542\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 47.0211\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 205ms/step - loss: 46.9880\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 195ms/step - loss: 46.9549\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 300ms/step - loss: 46.9218\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 46.8887\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 46.8557\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 46.8225\n",
      "# 1500 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 46.7898\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 46.7566\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 46.7238\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 46.6903\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 46.6570\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 46.6238\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 46.5907\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 46.5580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 46.5249\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 46.4917\n",
      "# 1510 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 46.4586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 46.4255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 46.3924\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 46.3595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 46.3261\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 46.2932\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 46.2600\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 46.2270\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 46.1939\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 46.1607\n",
      "# 1520 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 46.1277\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 46.0948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 46.0618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 46.0282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 45.9954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 45.9623\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 45.9294\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 45.8961\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 45.8633\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 45.8300\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 1530 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 45.7967\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 45.7637\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 45.7308\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 45.6977\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 45.6646\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 45.6314\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 45.5986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 45.5653\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 45.5321\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 45.4992\n",
      "# 1540 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 45.4662\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 45.4332\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 45.4002\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 45.3666\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 45.3335\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 45.3004\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 45.2673\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 45.2347\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 45.2013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 45.1686\n",
      "# 1550 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 45.1351\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 45.1021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 45.0693\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 45.0363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 45.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 44.9698\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 44.9366\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 44.9037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 44.8704\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 44.8374\n",
      "# 1560 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 44.8041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 44.7714\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 44.7379\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 44.7050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 44.6721\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 44.6387\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 44.6056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 44.5729\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 44.5395\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 44.5065\n",
      "# 1570 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 44.4735\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 44.4404\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 44.4077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 44.3742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 44.3416\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 44.3084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 44.2750\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 44.2419\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 44.2088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 44.1760\n",
      "# 1580 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 44.1425\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 44.1095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 44.0765\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 44.0436\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 44.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 43.9771\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 43.9439\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 43.9110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 43.8779\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 43.8447\n",
      "# 1590 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 43.8118\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 43.7788\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 43.7458\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 43.7127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 43.6799\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 43.6463\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 43.6130\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 43.5802\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 43.5470\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 43.5145\n",
      "# 1600 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 43.4809\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 43.4478\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 369ms/step - loss: 43.4149\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 43.3816\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 43.3485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 43.3158\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 43.2827\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 43.2494\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 43.2164\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 43.1833\n",
      "# 1610 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 43.1501\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 43.1170\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 43.0839\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 43.0507\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 43.0180\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 42.9850\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 42.9518\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 42.9186\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 42.8854\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 42.8525\n",
      "# 1620 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 42.8195\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 42.7866\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 42.7533\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 42.7201\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 42.6872\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 169ms/step - loss: 42.6537\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 42.6209\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 42.5883\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 975ms/step - loss: 42.5547\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 42.5220\n",
      "# 1630 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 42.4885\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 42.4558\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 42.4227\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 42.3894\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 42.3562\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 42.3231\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 42.2901\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 42.2570\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 116ms/step - loss: 42.2238\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 42.1911\n",
      "# 1640 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 42.1578\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 42.1247\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 42.0914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 42.0586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 42.0254\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 41.9928\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 41.9595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 41.9261\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 41.8930\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 41.8602\n",
      "# 1650 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 41.8269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 41.7939\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 41.7609\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 41.7280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 41.6946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 41.6616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 41.6287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 41.5955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 41.5625\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 41.5291\n",
      "# 1660 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 41.4967\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 41.4636\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 41.4303\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 41.3975\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 41.3639\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 41.3308\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 41.2977\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 41.2649\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 41.2319\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 41.1990\n",
      "# 1670 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 41.1658\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 41.1326\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 41.0994\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 41.0662\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 41.0334\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 41.0003\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 40.9672\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 40.9341\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 40.9013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 40.8680\n",
      "# 1680 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 40.8350\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 40.8017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 40.7687\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 40.7359\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 40.7025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 40.6697\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 40.6365\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 40.6033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 40.5704\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 40.5370\n",
      "# 1690 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 40.5042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 40.4711\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 40.4379\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 270ms/step - loss: 40.4048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 40.3718\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 40.3387\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 40.3058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 40.2728\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 40.2396\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 40.2067\n",
      "# 1700 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 40.1735\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 40.1403\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 40.1071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 40.0742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 40.0414\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 40.0079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 39.9751\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 39.9422\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 39.9088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 39.8758\n",
      "# 1710 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 39.8428\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 39.8096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 39.7763\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 39.7437\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 39.7108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 39.6772\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 39.6442\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 39.6114\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 39.5782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 39.5453\n",
      "# 1720 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 39.5119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 39.4790\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 39.4457\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 39.4127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 39.3798\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 39.3466\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 39.3135\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 39.2803\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 39.2475\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 39.2143\n",
      "# 1730 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 39.1813\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 39.1483\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 39.1153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 39.0819\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 39.0491\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 39.0160\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 38.9830\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 38.9497\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 38.9172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 38.8837\n",
      "# 1740 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 38.8507\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 38.8175\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 38.7844\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 38.7514\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 122ms/step - loss: 38.7183\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 38.6853\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 38.6523\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 38.6190\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 38.5862\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 38.5530\n",
      "# 1750 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 38.5199\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 38.4872\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 106ms/step - loss: 38.4537\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 38.4207\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 408ms/step - loss: 38.3876\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 38.3544\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 38.3217\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 38.2885\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 38.2556\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 38.2225\n",
      "# 1760 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 38.1894\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 38.1567\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 38.1233\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 38.0901\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 38.0571\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 38.0240\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 37.9908\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 37.9580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 37.9248\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 37.8917\n",
      "# 1770 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 37.8587\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 37.8255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 37.7924\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 37.7594\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 37.7263\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 37.6938\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 37.6603\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 37.6274\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 37.5940\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 37.5609\n",
      "# 1780 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 37.5278\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 37.4948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 37.4619\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 37.4289\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 37.3958\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 37.3627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 37.3297\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 37.2966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 37.2636\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 37.2308\n",
      "# 1790 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 37.1973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 37.1646\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 37.1315\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 37.0984\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 95ms/step - loss: 37.0648\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 37.0322\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 36.9994\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 36.9659\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 36.9331\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 36.8998\n",
      "# 1800 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 36.8668\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 36.8339\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 36.8005\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 36.7676\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 36.7343\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 36.7015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 36.6686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 36.6358\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 36.6021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 36.5691\n",
      "# 1810 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 36.5361\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 36.5031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 36.4702\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 36.4369\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 36.4039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 36.3711\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 36.3377\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 36.3047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 36.2716\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 36.2387\n",
      "# 1820 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 36.2058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 36.1724\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 36.1393\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 188ms/step - loss: 36.1062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 36.0732\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 36.0405\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 36.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 35.9741\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 35.9409\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 35.9077\n",
      "# 1830 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 35.8750\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 240ms/step - loss: 35.8422\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 35.8088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 35.7757\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 35.7425\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 35.7099\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 35.6766\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 35.6436\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 35.6106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 35.5773\n",
      "# 1840 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 35.5443\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 35.5113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 35.4781\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 35.4451\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 35.4122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 35.3789\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 35.3456\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 35.3128\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 35.2796\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 35.2469\n",
      "# 1850 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 35.2134\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 35.1809\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 35.1475\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 35.1144\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 35.0815\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 35.0482\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 35.0153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 34.9822\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 34.9496\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 34.9162\n",
      "# 1860 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 34.8832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 34.8502\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 34.8172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 34.7837\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 34.7508\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 34.7176\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 34.6847\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 34.6517\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 193ms/step - loss: 34.6189\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 34.5856\n",
      "# 1870 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 34.5526\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 34.5195\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 95ms/step - loss: 34.4862\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 34.4534\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 34.4204\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 34.3871\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 34.3542\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 34.3211\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 34.2885\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 34.2549\n",
      "# 1880 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 34.2223\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 34.1891\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 34.1559\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 34.1232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 34.0896\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 536ms/step - loss: 34.0564\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 34.0240\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 33.9910\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 33.9573\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 33.9244\n",
      "# 1890 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 33.8914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 33.8582\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 33.8252\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 33.7926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 33.7594\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 33.7260\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 33.6931\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 33.6600\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 33.6269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 33.5938\n",
      "# 1900 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 33.5612\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 33.5275\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 33.4946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 33.4615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 33.4285\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 33.3956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 33.3625\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 33.3296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 33.2964\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 33.2633\n",
      "# 1910 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 33.2304\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 33.1973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 33.1641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 33.1312\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 33.0980\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 33.0651\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 33.0317\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 32.9989\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 32.9658\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 32.9327\n",
      "# 1920 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 32.8997\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.8666\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 32.8336\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 32.8005\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 32.7677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 589ms/step - loss: 32.7347\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.7016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 32.6683\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 32.6350\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.6022\n",
      "# 1930 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.5694\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 32.5359\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 32.5031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 32.4700\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.4372\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 32.4039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 32.3710\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 32.3378\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 32.3048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 32.2720\n",
      "# 1940 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 32.2387\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 32.2057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 32.1725\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 32.1395\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 32.1064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 32.0734\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 32.0404\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 32.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 31.9742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 31.9412\n",
      "# 1950 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 31.9078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 31.8755\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 31.8418\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 31.8091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 31.7758\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 31.7432\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 31.7098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 31.6767\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 31.6437\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 31.6105\n",
      "# 1960 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 31.5778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 31.5446\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 31.5113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 31.4783\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 31.4454\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 31.4125\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 451ms/step - loss: 31.3793\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 31.3462\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 222ms/step - loss: 31.3131\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 31.2800\n",
      "# 1970 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 31.2470\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 31.2140\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 144ms/step - loss: 31.1809\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 31.1478\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 449ms/step - loss: 31.1148\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 31.0818\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 31.0488\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 31.0155\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 30.9826\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 30.9494\n",
      "# 1980 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 30.9164\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 30.8834\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 30.8505\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 30.8172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 30.7841\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 30.7514\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 30.7183\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 30.6850\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 30.6524\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 30.6191\n",
      "# 1990 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 30.5862\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 30.5531\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 30.5199\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 30.4870\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 30.4541\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 30.4205\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 30.3878\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 30.3547\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 30.3219\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 30.2885\n",
      "# 2000 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 30.2556\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 30.2225\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 30.1896\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 30.1563\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 30.1237\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 30.0905\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 30.0573\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 30.0241\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.9914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 29.9583\n",
      "# 2010 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.9251\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 29.8918\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 29.8594\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 29.8258\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 29.7931\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 29.7602\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.7269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.6941\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 29.6607\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 29.6277\n",
      "# 2020 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 29.5946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 29.5616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.5283\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 29.4955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 29.4624\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 29.4296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 29.3964\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 29.3635\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 29.3301\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 29.2972\n",
      "# 2030 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 29.2643\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 29.2310\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 29.1982\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 29.1651\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 29.1324\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 29.0988\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 29.0659\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 29.0327\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 28.9999\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 28.9666\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 2040 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 28.9337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 28.9009\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 28.8676\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 28.8344\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 28.8016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 28.7686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 28.7351\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 28.7023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 28.6692\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 28.6361\n",
      "# 2050 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 28.6029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 28.5701\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 28.5373\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 28.5039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 28.4712\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 28.4378\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 28.4051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 28.3717\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 28.3388\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 28.3056\n",
      "# 2060 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 28.2728\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 28.2398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 28.2066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 28.1735\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 28.1403\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 28.1077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 28.0745\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 28.0415\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 28.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 27.9752\n",
      "# 2070 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 27.9425\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 27.9091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 27.8761\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 27.8430\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 27.8100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 27.7771\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 27.7444\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 27.7109\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 27.6780\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 27.6448\n",
      "# 2080 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 27.6119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 27.5788\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 27.5457\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 27.5126\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 27.4800\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 27.4466\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 27.4137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 27.3808\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 27.3476\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 27.3143\n",
      "# 2090 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 27.2817\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 27.2484\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 27.2150\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 27.1822\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 27.1491\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 27.1163\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 27.0830\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 27.0500\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 27.0169\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 26.9841\n",
      "# 2100 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.9509\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 26.9177\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 26.8848\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 26.8519\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 26.8187\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 26.7861\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 26.7527\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 26.7195\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 26.6866\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 26.6535\n",
      "# 2110 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 26.6208\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 26.5874\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.5545\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 26.5212\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 26.4882\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.4552\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 26.4221\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 26.3891\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 26.3561\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 26.3229\n",
      "# 2120 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.2900\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.2570\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.2238\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 26.1911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 26.1580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 26.1248\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 26.0918\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 26.0589\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 26.0255\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 25.9926\n",
      "# 2130 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 25.9598\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 25.9265\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 25.8936\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 25.8604\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 25.8273\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 25.7946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 25.7613\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 172ms/step - loss: 25.7282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 25.6955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 25.6620\n",
      "# 2140 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 25.6292\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 25.5960\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 25.5629\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 25.5300\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 25.4970\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 25.4639\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 25.4309\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 25.3977\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 25.3648\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 25.3322\n",
      "# 2150 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 25.2990\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 25.2659\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 25.2329\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 25.1996\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 25.1668\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 25.1335\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 25.1004\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 25.0675\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 25.0345\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 25.0013\n",
      "# 2160 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 24.9684\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 24.9353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 24.9023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 24.8690\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 24.8359\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.8035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 122ms/step - loss: 24.7699\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 24.7368\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 24.7041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 24.6714\n",
      "# 2170 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 131ms/step - loss: 24.6377\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.6046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 24.5718\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 24.5388\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 24.5056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.4725\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 24.4398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 24.4067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.3736\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 24.3404\n",
      "# 2180 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 24.3076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 24.2747\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 24.2412\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 24.2086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.1753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 24.1424\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 47ms/step - loss: 24.1093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 24.0762\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 24.0431\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 24.0101\n",
      "# 2190 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 23.9771\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 23.9444\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 23.9110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 23.8782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 23.8447\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 23.8118\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 23.7786\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 23.7457\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 23.7127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 23.6797\n",
      "# 2200 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 23.6470\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 23.6136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 23.5805\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 23.5475\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 23.5146\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 23.4814\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 23.4483\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 23.4153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 23.3824\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 23.3493\n",
      "# 2210 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 23.3163\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 23.2834\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 23.2503\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 23.2170\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 23.1846\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 23.1512\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 23.1183\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 23.0849\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 23.0522\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 23.0190\n",
      "# 2220 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 22.9858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 22.9529\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 22.9199\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 22.8867\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 22.8538\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 22.8206\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 136ms/step - loss: 22.7876\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 22.7547\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 22.7217\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 22.6883\n",
      "# 2230 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 22.6555\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 22.6226\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 22.5895\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 22.5563\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 22.5237\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 22.4904\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 22.4575\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 22.4243\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 22.3914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 22.3582\n",
      "# 2240 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 22.3251\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 22.2925\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 22.2595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 22.2260\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 22.1928\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 22.1599\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 22.1269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 22.0938\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 22.0609\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 22.0277\n",
      "# 2250 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 21.9948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 21.9618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 21.9288\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 21.8956\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 21.8626\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 21.8293\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 21.7964\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 21.7638\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 21.7306\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 21.6974\n",
      "# 2260 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 21.6645\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 21.6311\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 21.5983\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 21.5654\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 21.5322\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 21.4993\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 21.4662\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 21.4332\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 21.4002\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 21.3669\n",
      "# 2270 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 21.3339\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 21.3010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 21.2678\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 21.2353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 21.2021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 21.1686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 21.1358\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 21.1029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 21.0695\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 21.0364\n",
      "# 2280 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 21.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 20.9706\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 20.9379\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 20.9044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 20.8716\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 20.8385\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 20.8055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 20.7723\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 20.7394\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 20.7066\n",
      "# 2290 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 20.6738\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 20.6404\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 20.6073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 20.5744\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 20.5412\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 20.5079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 20.4752\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 20.4420\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 20.4089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 20.3761\n",
      "# 2300 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 20.3431\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 20.3097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 20.2770\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 20.2438\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 20.2108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 20.1778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 20.1452\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 20.1118\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 20.0787\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 20.0456\n",
      "# 2310 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 20.0127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 19.9800\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 19.9465\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 19.9137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 19.8804\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 19.8472\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 19.8142\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 19.7814\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 19.7483\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 19.7151\n",
      "# 2320 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 19.6823\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 19.6493\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 19.6162\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 19.5832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 19.5500\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 19.5171\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 19.4841\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 19.4508\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 19.4184\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 19.3849\n",
      "# 2330 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 19.3520\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 19.3188\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 19.2859\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 19.2528\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 19.2196\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 19.1867\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 19.1538\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 19.1206\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 19.0877\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 19.0546\n",
      "# 2340 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 19.0220\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 18.9886\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 18.9555\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 18.9226\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 18.8894\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 18.8564\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 18.8236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 18.7901\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 18.7573\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 18.7242\n",
      "# 2350 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 18.6911\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 18.6582\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 18.6252\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 18.5922\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 18.5592\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 18.5262\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 18.4932\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 18.4600\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 18.4269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 18.3940\n",
      "# 2360 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 18.3610\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 18.3278\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 18.2948\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 18.2618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 18.2291\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 18.1955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 18.1627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 18.1298\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 18.0966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 18.0635\n",
      "# 2370 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 18.0309\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 17.9975\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 17.9646\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 17.9313\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 17.8986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 17.8655\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 17.8322\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 17.7992\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 17.7663\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 17.7333\n",
      "# 2380 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 17.7002\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 17.6673\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 17.6342\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 17.6012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 17.5682\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 17.5353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 17.5021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 17.4688\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 17.4360\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 17.4028\n",
      "# 2390 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 17.3697\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 17.3370\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 17.3038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 17.2709\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 17.2378\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 17.2050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 17.1720\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 17.1391\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 17.1059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 17.0727\n",
      "# 2400 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 17.0396\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 17.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 16.9737\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 16.9409\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 16.9075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 470ms/step - loss: 16.8746\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 16.8416\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 16.8084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 16.7753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.7426\n",
      "# 2410 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 16.7096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.6765\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 16.6434\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.6105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.5774\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.5442\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 16.5111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 16.4782\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 16.4452\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 16.4120\n",
      "# 2420 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.3792\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 204ms/step - loss: 16.3459\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 16.3129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 16.2798\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 16.2472\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 16.2141\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 16.1810\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 16.1477\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 16.1147\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 16.0817\n",
      "# 2430 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 16.0485\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 16.0159\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 15.9829\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 15.9498\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 15.9164\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 15.8837\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 15.8506\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 15.8174\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 15.7846\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 15.7515\n",
      "# 2440 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 15.7183\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 15.6858\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 15.6525\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 15.6193\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 15.5862\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 15.5532\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 15.5201\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 15.4873\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 15.4542\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 15.4213\n",
      "# 2450 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 15.3881\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 15.3554\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 15.3222\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 95ms/step - loss: 15.2889\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 15.2558\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 15.2229\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 15.1903\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 15.1569\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 15.1239\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 15.0908\n",
      "# 2460 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 15.0579\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 174ms/step - loss: 15.0247\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 14.9917\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 14.9586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 14.9256\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 14.8926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 859ms/step - loss: 14.8595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 220ms/step - loss: 14.8265\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 14.7935\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 472ms/step - loss: 14.7605\n",
      "# 2470 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 14.7273\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 14.6946\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 14.6612\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 14.6284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 14.5955\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 14.5623\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 14.5293\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 14.4964\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 14.4635\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 14.4302\n",
      "# 2480 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 14.3974\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 14.3641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 14.3309\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 14.2980\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 14.2649\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 14.2319\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 14.1992\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 14.1659\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 14.1332\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 14.0999\n",
      "# 2490 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 14.0670\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 14.0337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 149ms/step - loss: 14.0008\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 13.9677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 13.9350\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 13.9020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 13.8688\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 13.8357\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 13.8027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 13.7695\n",
      "# 2500 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 13.7365\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 13.7034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 13.6706\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 13.6376\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 13.6048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 13.5715\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 13.5383\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 432ms/step - loss: 13.5053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 13.4726\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 13.4394\n",
      "# 2510 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 13.4065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 13.3737\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 13.3401\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 13.3074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 13.2741\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 13.2412\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 13.2086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 13.1754\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 13.1423\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 13.1089\n",
      "# 2520 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 13.0759\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 13.0429\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 13.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 12.9768\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 12.9439\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 12.9112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 12.8778\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 12.8452\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 12.8121\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 12.7789\n",
      "# 2530 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 12.7458\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 12.7127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 12.6798\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 12.6467\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 12.6137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 12.5811\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 12.5476\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 12.5146\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 12.4817\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 436ms/step - loss: 12.4489\n",
      "# 2540 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 12.4156\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 12.3822\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 12.3496\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 12.3164\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 12.2834\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 12.2505\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 12.2173\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 12.1843\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 12.1514\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 12.1184\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 2550 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 12.0857\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 12.0520\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 12.0192\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 11.9862\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 11.9530\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 11.9201\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 11.8870\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 11.8543\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 11.8211\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 11.7880\n",
      "# 2560 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 11.7551\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 11.7219\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 11.6889\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 11.6559\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 11.6229\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 11.5899\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 11.5568\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 11.5236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 11.4907\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 11.4581\n",
      "# 2570 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 11.4249\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 11.3914\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 11.3590\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 11.3256\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 11.2929\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 11.2596\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 11.2266\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 11.1937\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 11.1606\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 11.1274\n",
      "# 2580 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 11.0945\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 107ms/step - loss: 11.0615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 11.0282\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 10.9953\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 10.9624\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 10.9294\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 249ms/step - loss: 10.8966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 10.8630\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 10.8305\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 10.7971\n",
      "# 2590 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 144ms/step - loss: 10.7641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 10.7311\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 10.6982\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 10.6652\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 10.6320\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 10.5994\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 10.5662\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 10.5331\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 10.5000\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 10.4670\n",
      "# 2600 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 10.4341\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 10.4009\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 10.3679\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 10.3348\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 144ms/step - loss: 10.3021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 10.2686\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 10.2357\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 10.2029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 10.1697\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 10.1366\n",
      "# 2610 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 10.1038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 10.0709\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 10.0376\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 10.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 9.9715\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 9.9386\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 9.9057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 9.8723\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 9.8398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 9.8066\n",
      "# 2620 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 9.7737\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 9.7402\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 9.7076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 9.6744\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 9.6416\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 9.6084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 9.5753\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 9.5422\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 9.5094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 806ms/step - loss: 9.4765\n",
      "# 2630 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 9.4434\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 9.4102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 9.3771\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 9.3440\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 9.3112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 9.2780\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 9.2451\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 9.2122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 9.1792\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 9.1460\n",
      "# 2640 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 9.1133\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 9.0802\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 9.0471\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 9.0138\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 8.9808\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 8.9482\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 8.9148\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 8.8816\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 8.8486\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 8.8157\n",
      "# 2650 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 8.7832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 8.7497\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 56ms/step - loss: 8.7168\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 8.6839\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 8.6506\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 8.6176\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 8.5847\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 8.5515\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 8.5187\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 8.4856\n",
      "# 2660 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 8.4527\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 8.4195\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 8.3865\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 8.3535\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 149ms/step - loss: 8.3203\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 8.2873\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 8.2542\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 8.2214\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 463ms/step - loss: 8.1883\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 8.1554\n",
      "# 2670 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 8.1223\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 251ms/step - loss: 8.0892\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 8.0564\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 8.0232\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 7.9905\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 7.9572\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 7.9244\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 7.8912\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 7.8580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 7.8252\n",
      "# 2680 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 7.7921\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 7.7590\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 7.7259\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 7.6932\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 7.6599\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 7.6269\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 7.5942\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 141ms/step - loss: 7.5609\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 325ms/step - loss: 7.5278\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 205ms/step - loss: 7.4948\n",
      "# 2690 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 7.4618\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 7.4288\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 7.3958\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 7.3627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 7.3297\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 7.2965\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 7.2637\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 7.2306\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 7.1977\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 7.1647\n",
      "# 2700 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 7.1317\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 7.0986\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 7.0656\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 7.0323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 6.9995\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 6.9665\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 6.9337\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 6.9004\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 6.8677\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 6.8348\n",
      "# 2710 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 6.8013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 6.7685\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 6.7351\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 6.7023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 6.6692\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 6.6363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 6.6035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 6.5706\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 6.5371\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 6.5039\n",
      "# 2720 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 6.4710\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 6.4383\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 6.4053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 6.3724\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 6.3391\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 6.3060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 6.2728\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 6.2400\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 6.2071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 6.1740\n",
      "# 2730 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 6.1410\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 6.1079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 6.0750\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 6.0419\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 6.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 5.9761\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 5.9427\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 5.9098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 5.8767\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 5.8439\n",
      "# 2740 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 5.8108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 5.7779\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 5.7451\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 5.7117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 5.6787\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 5.6456\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 5.6129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 5.5797\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 5.5465\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 5.5136\n",
      "# 2750 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 5.4805\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 5.4477\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 5.4145\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 5.3817\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 5.3489\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 5.3155\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 5.2824\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 5.2494\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 5.2162\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 5.1833\n",
      "# 2760 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 5.1503\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 5.1172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 5.0843\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 5.0514\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 5.0183\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 4.9852\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 4.9520\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 4.9191\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 4.8861\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 4.8535\n",
      "# 2770 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 4.8204\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 4.7870\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 4.7541\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 4.7212\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 4.6879\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 4.6553\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 4.6220\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 4.5890\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 4.5561\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 4.5229\n",
      "# 2780 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 4.4902\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 4.4567\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 4.4241\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 4.3907\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 4.3580\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 4.3249\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 4.2919\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 4.2586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 4.2259\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 4.1929\n",
      "# 2790 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 4.1595\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 4.1267\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 4.0940\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 4.0605\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 4.0275\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 3.9945\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 3.9615\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 3.9284\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 3.8954\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 3.8624\n",
      "# 2800 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 3.8295\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 3.7962\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 3.7634\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 3.7305\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 354ms/step - loss: 3.6973\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 3.6646\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 3.6314\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 3.5983\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 3.5653\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 3.5326\n",
      "# 2810 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 3.4995\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 3.4664\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 3.4333\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 3.4003\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 3.3670\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 3.3340\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 3.3013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 3.2680\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 3.2351\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 3.2020\n",
      "# 2820 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 3.1690\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 3.1361\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 3.1031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 3.0699\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 3.0371\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 3.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 2.9711\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 2.9378\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 2.9048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 2.8722\n",
      "# 2830 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 2.8387\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 2.8057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 2.7727\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 2.7398\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 2.7066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 2.6742\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 2.6405\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 2.6079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 2.5746\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 2.5415\n",
      "# 2840 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 2.5090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 145ms/step - loss: 2.4755\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 2.4426\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 2.4095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 2.3768\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 2.3437\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 2.3106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 2.2774\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 2.2445\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 2.2114\n",
      "# 2850 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 2.1785\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 173ms/step - loss: 2.1456\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 2.1127\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 2.0794\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 2.0465\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 2.0134\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 1.9804\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 1.9473\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 1.9143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 1.8816\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "# 2860 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 1.8483\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 1.8152\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 1.7823\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 1.7493\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 1.7161\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 1.6832\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 1.6502\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 1.6175\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 1.5841\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 1.5513\n",
      "# 2870 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 1.5180\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 1.4852\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 1.4520\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 1.4193\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 1.3860\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 1.3531\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 1.3202\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 1.2870\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 1.2536\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 1.2207\n",
      "# 2880 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 1.1879\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 1.1548\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 1.1219\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 1.0889\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 1.0558\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 1.0230\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.9896\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.9567\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.9236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.8908\n",
      "# 2890 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.8576\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.8246\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.7918\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.7589\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.7259\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.6926\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.6594\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.6272\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 169ms/step - loss: 0.5940\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 146ms/step - loss: 0.5604\n",
      "# 2900 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.5276\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.4950\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.4616\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.4287\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.3961\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 130ms/step - loss: 0.3627\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.3296\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.2966\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.2641\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.2314\n",
      "# 2910 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.1982\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.1650\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.1315\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0987\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0655\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0363\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 0.0699\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0319\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0882\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.1037\n",
      "# 2920 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0922\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.1009\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 0.1207\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.1231\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.1106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.1017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.1043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0998\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0817\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0629\n",
      "# 2930 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0586\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0430\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0167\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0189\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0236\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0353\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0379\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0440\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0446\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0386\n",
      "# 2940 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0348\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0261\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0143\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0323\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0359\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0271\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0381\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0404\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0274\n",
      "# 2950 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0277\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0253\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0154\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0145\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0224\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0211\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.0190\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0158\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0078\n",
      "# 2960 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0086\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0228\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0200\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0141\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0210\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0103\n",
      "# 2970 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0142\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0128\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0162\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0142\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0219\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0234\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0123\n",
      "# 2980 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0219\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0222\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.0210\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0240\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0230\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0280\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0167\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0132\n",
      "# 2990 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0192\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0199\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0275\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0168\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 0.0166\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0190\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 0.0244\n",
      "# 3000 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0157\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0174\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0145\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0206\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0126\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0189\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0091\n",
      "# 3010 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0198\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 0.0168\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0141\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0197\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0097\n",
      "# 3020 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 98ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0189\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0092\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0148\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0176\n",
      "# 3030 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0133\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0187\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0130\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0052\n",
      "# 3040 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0132\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0173\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0124\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0172\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0083\n",
      "# 3050 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0153\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0142\n",
      "# 3060 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0130\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0074\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 0.0120\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0063\n",
      "# 3070 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0134\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 140ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0137\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0079\n",
      "# 3080 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0115\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 0.0126\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0118\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0127\n",
      "# 3090 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0124\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0120\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 555ms/step - loss: 0.0062\n",
      "# 3100 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0122\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 0.0115\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0082\n",
      "# 3110 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0109\n",
      "# 3120 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0114\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0056\n",
      "# 3130 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 295ms/step - loss: 0.0106\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 657ms/step - loss: 0.0105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0071\n",
      "# 3140 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0109\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.0106\n",
      "# 3150 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0051\n",
      "# 3160 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0066\n",
      "# 3170 examples trained\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 144ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0096\n",
      "# 3180 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0092\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0058\n",
      "# 3190 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 160ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0069\n",
      "# 3200 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 125ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0091\n",
      "# 3210 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0043\n",
      "# 3220 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0063\n",
      "# 3230 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0092\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0081\n",
      "# 3240 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0054\n",
      "# 3250 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0063\n",
      "# 3260 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0078\n",
      "# 3270 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0045\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 213ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0043\n",
      "# 3280 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0044\n",
      "# 3290 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0085\n",
      "# 3300 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0040\n",
      "# 3310 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 136ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0059\n",
      "# 3320 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 231ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0086\n",
      "# 3330 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0050\n",
      "# 3340 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0058\n",
      "# 3350 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0071\n",
      "# 3360 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0031\n",
      "# 3370 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 235ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 129ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 98ms/step - loss: 0.0083\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0049\n",
      "# 3380 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 473ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 361ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 6s 6s/step - loss: 0.0077\n",
      "# 3390 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 7s 7s/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 178ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0038\n",
      "# 3400 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 7s 7s/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0054\n",
      "# 3410 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0051\n",
      "# 3420 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 431ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 218ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 174ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0022\n",
      "# 3430 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 105ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0022\n",
      "# 3440 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 3.5544e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0061\n",
      "# 3450 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 218ms/step - loss: 5.5728e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0066\n",
      "# 3460 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 8.9159e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 273ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 107ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 723ms/step - loss: 0.0033\n",
      "# 3470 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 206ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 162ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0091\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 5s 5s/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 477ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0125\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 777ms/step - loss: 0.0117\n",
      "# 3480 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0032\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 128ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 786ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 252ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 971ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0048\n",
      "# 3490 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 314ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 836ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0024\n",
      "# 3500 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 306ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 151ms/step - loss: 7.6576e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 5.1411e-04\n",
      "# 3510 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 150ms/step - loss: 0.0031\n",
      "# 3520 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 208ms/step - loss: 0.0039\n",
      "# 3530 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 107ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 293ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 209ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 208ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0023\n",
      "# 3540 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 43ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 182ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0031\n",
      "# 3550 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0033\n",
      "# 3560 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 134ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0034\n",
      "# 3570 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 102ms/step - loss: 0.0022\n",
      "# 3580 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0018\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0032\n",
      "# 3590 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0027\n",
      "# 3600 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0032\n",
      "# 3610 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0023\n",
      "# 3620 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 335ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0023\n",
      "# 3630 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 108ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 147ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0034\n",
      "# 3640 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0020\n",
      "# 3650 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 147ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 54ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0027\n",
      "# 3660 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 691ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.0026\n",
      "# 3670 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.0016\n",
      "# 3680 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0027\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0024\n",
      "# 3690 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 475ms/step - loss: 8.7321e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 173ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0015\n",
      "# 3700 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 173ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0034\n",
      "# 3710 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 144ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 148ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0015\n",
      "# 3720 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 80ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 425ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 2.5633e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0086\n",
      "# 3730 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0037\n",
      "# 3740 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 259ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0033\n",
      "# 3750 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0060\n",
      "# 3760 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 156ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 464ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0035\n",
      "# 3770 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0041\n",
      "# 3780 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0067\n",
      "# 3790 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 364ms/step - loss: 0.0030\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 152ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 243ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0027\n",
      "# 3800 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 918ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 266ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0041\n",
      "# 3810 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0077\n",
      "# 3820 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 195ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0012\n",
      "# 3830 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 8.7358e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 1.0033e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0042\n",
      "# 3840 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0029\n",
      "# 3850 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0018\n",
      "# 3860 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0015\n",
      "# 3870 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0015\n",
      "# 3880 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 9.5867e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0050\n",
      "# 3890 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 5.6768e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0047\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0031\n",
      "# 3900 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0027\n",
      "# 3910 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0030\n",
      "# 3920 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0034\n",
      "# 3930 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 156ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 112ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 110ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0019\n",
      "# 3940 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0014\n",
      "# 3950 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 272ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 341ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0014\n",
      "# 3960 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 260ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0015\n",
      "# 3970 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0027\n",
      "# 3980 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0034\n",
      "# 3990 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 153ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 225ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0017\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0029\n",
      "# 4000 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 3.7207e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0026\n",
      "# 4010 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 9.7006e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 7.8968e-04\n",
      "# 4020 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 98ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 7.1616e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 5.5199e-04\n",
      "# 4030 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 198ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 193ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0053\n",
      "# 4040 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 114ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 138ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 458ms/step - loss: 0.0045\n",
      "# 4050 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0028\n",
      "# 4060 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0039\n",
      "# 4070 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 734ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0059\n",
      "# 4080 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 127ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 203ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0032\n",
      "# 4090 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 264ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0034\n",
      "# 4100 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0050\n",
      "# 4110 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 349ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 413ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 450ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 132ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 325ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 798ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 203ms/step - loss: 0.0027\n",
      "# 4120 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 205ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0028\n",
      "# 4130 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 47ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 4s 4s/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 47ms/step - loss: 0.0043\n",
      "# 4140 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0071\n",
      "# 4150 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 117ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0022\n",
      "# 4160 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0040\n",
      "# 4170 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 8.4261e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0069\n",
      "# 4180 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0018\n",
      "# 4190 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 47ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 1.6033e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "# 4200 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 186ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 3.2934e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0109\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0087\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0119\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0113\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 111ms/step - loss: 0.0060\n",
      "# 4210 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0061\n",
      "# 4220 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0017\n",
      "# 4230 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 2.4270e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0074\n",
      "# 4240 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 284ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0011\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0021\n",
      "# 4250 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 164ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0043\n",
      "# 4260 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 415ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 227ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 0.0047\n",
      "# 4270 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 9.2208e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "# 4280 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 8.0194e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0050\n",
      "# 4290 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0021\n",
      "# 4300 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 265ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0023\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "# 4310 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0048\n",
      "# 4320 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 137ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 4.9611e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0077\n",
      "# 4330 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 230ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 104ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 145ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0042\n",
      "# 4340 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0029\n",
      "# 4350 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0028\n",
      "# 4360 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 120ms/step - loss: 9.7715e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 4.1132e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 6.7307e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0011\n",
      "# 4370 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 6.1612e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 9.4954e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 216ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0086\n",
      "# 4380 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 9.1582e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0052\n",
      "# 4390 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0066\n",
      "# 4400 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 201ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 606ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 518ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 122ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 217ms/step - loss: 0.0077\n",
      "# 4410 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 139ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 0.0071\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 209ms/step - loss: 0.0107\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 152ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 310ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 270ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 311ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0065\n",
      "# 4420 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 174ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 161ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 622ms/step - loss: 0.0069\n",
      "# 4430 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0095\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 109ms/step - loss: 0.0046\n",
      "# 4440 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0071\n",
      "# 4450 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 355ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0050\n",
      "# 4460 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0105\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 135ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 119ms/step - loss: 0.0067\n",
      "# 4470 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 88ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0038\n",
      "# 4480 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0089\n",
      "# 4490 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 47ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0081\n",
      "# 4500 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0123\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 123ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0081\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0060\n",
      "# 4510 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0024\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0017\n",
      "# 4520 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 3s 3s/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 8.2258e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0064\n",
      "# 4530 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 282ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 8.6404e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0031\n",
      "# 4540 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0053\n",
      "# 4550 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 358ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 7.5819e-04\n",
      "# 4560 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 124ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0136\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0123\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0096\n",
      "# 4570 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0025\n",
      "# 4580 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 83ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 206ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 569ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 524ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 307ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0034\n",
      "# 4590 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0025\n",
      "# 4600 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 226ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0028\n",
      "# 4610 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 539ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0028\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0026\n",
      "# 4620 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 233ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 94ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 0.0025\n",
      "# 4630 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0019\n",
      "# 4640 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 395ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 87ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0016\n",
      "# 4650 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 100ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 9.1768e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 97ms/step - loss: 0.0046\n",
      "# 4660 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 4.0847e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 850ms/step - loss: 0.0129\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0110\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0117\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0063\n",
      "# 4670 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0047\n",
      "# 4680 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 175ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0012\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0053\n",
      "# 4690 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0057\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 3.1191e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 8.4871e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0011\n",
      "# 4700 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 8.9736e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 3.7174e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0052\n",
      "# 4710 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 812ms/step - loss: 3.7187e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 166ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 157ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 294ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 187ms/step - loss: 0.0041\n",
      "# 4720 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 145ms/step - loss: 0.0022\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0037\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 106ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 165ms/step - loss: 9.8282e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0082\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 73ms/step - loss: 0.0027\n",
      "# 4730 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 44ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0021\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 204ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0034\n",
      "# 4740 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 58ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0029\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 316ms/step - loss: 0.0053\n",
      "# 4750 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0026\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 0.0035\n",
      "# 4760 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 90ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0050\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0023\n",
      "# 4770 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0044\n",
      "# 4780 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 672ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0036\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0028\n",
      "# 4790 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 229ms/step - loss: 0.0028\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0045\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 1s 928ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0042\n",
      "# 4800 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0052\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0042\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 427ms/step - loss: 0.0085\n",
      "# 4810 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0010\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0015\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0074\n",
      "# 4820 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0101\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0049\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0077\n",
      "Epoch 1/1\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0093\n",
      "# 4830 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0083\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0062\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0052\n",
      "# 4840 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 61ms/step - loss: 0.0080\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 45ms/step - loss: 0.0038\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 57ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 84ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0075\n",
      "# 4850 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0024\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 0.0102\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0048\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0040\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0073\n",
      "# 4860 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0111\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 91ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 65ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 53ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0095\n",
      "# 4870 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 126ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0047\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0084\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0030\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 106ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 79ms/step - loss: 0.0050\n",
      "# 4880 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 253ms/step - loss: 0.0077\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0031\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 71ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0044\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0076\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 266ms/step - loss: 0.0067\n",
      "# 4890 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0017\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0070\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0037\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 96ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 113ms/step - loss: 0.0062\n",
      "# 4900 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 48ms/step - loss: 0.0112\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 46ms/step - loss: 0.0103\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0056\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 299ms/step - loss: 0.0034\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0066\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0063\n",
      "# 4910 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 165ms/step - loss: 0.0059\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 92ms/step - loss: 0.0014\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0090\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0072\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0074\n",
      "# 4920 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0114\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 279ms/step - loss: 0.0060\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0061\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0013\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0093\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 115ms/step - loss: 0.0073\n",
      "# 4930 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0016\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0073\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 51ms/step - loss: 0.0108\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0035\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 85ms/step - loss: 0.0071\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0018\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0066\n",
      "# 4940 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 324ms/step - loss: 0.0099\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0085\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0041\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0065\n",
      "# 4950 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 93ms/step - loss: 0.0019\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0067\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0100\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 101ms/step - loss: 0.0086\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0058\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 72ms/step - loss: 0.0094\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 64ms/step - loss: 0.0087\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0043\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0040\n",
      "# 4960 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 250ms/step - loss: 0.0075\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 77ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 81ms/step - loss: 0.0051\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 86ms/step - loss: 0.0079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 56ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 50ms/step - loss: 6.0117e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0069\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0081\n",
      "# 4970 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0027\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0104\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0098\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 82ms/step - loss: 0.0053\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 69ms/step - loss: 0.0032\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 59ms/step - loss: 0.0065\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 76ms/step - loss: 0.0055\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 9.8637e-04\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 55ms/step - loss: 0.0071\n",
      "# 4980 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 62ms/step - loss: 0.0096\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0074\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0020\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 75ms/step - loss: 0.0064\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 60ms/step - loss: 0.0093\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 68ms/step - loss: 0.0079\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0033\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0054\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0088\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 143ms/step - loss: 0.0080\n",
      "# 4990 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0023\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 66ms/step - loss: 0.0063\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0097\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 78ms/step - loss: 0.0089\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0039\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0046\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 70ms/step - loss: 0.0078\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 67ms/step - loss: 0.0068\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 89ms/step - loss: 0.0025\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 74ms/step - loss: 0.0057\n",
      "# 5000 examples trained\n",
      "Epoch 1/1\n",
      "1/1 [==============================] - 0s 63ms/step - loss: 0.0087\n"
     ]
    }
   ],
   "source": [
    "counter = 0\n",
    "loss = []\n",
    "for x_seq,label in zip(X_train,Y_train):\n",
    "    counter +=1\n",
    "    if counter %10 ==0:\n",
    "        print(\"# {} examples trained\".format(counter))\n",
    "    X_seq = np.array(x_seq).reshape(-1,len(x_seq), 2)\n",
    "    y = np.array(label).reshape(1, 2) # predict next element\n",
    "    history = model.fit(X_seq, y, batch_size=1, verbose=1, shuffle=False)\n",
    "    loss.extend(history.history['loss'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Loss VS # of training examples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "y = loss\n",
    "x = list(range(1,len(loss)+1))\n",
    "plt.figure(figsize=(5,5))\n",
    "plt.plot(x, y)\n",
    "plt.xlabel(\"# of training examples\",fontsize=18)\n",
    "plt.ylabel(\"euclidean distance loss\",fontsize=18)\n",
    "plt.title(\"Level B1 Loss\",fontsize=18)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.spatial.distance import euclidean\n",
    "\n",
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "\n",
    "test_set =  lat_lng[10000:15000]\n",
    "predicted_val =[]\n",
    "for i,each_traj in enumerate(test_set):\n",
    "    x_test = each_traj[:len(each_traj) -1]\n",
    "    x_test = np.array(x_test).reshape(1, len(x_test), 2)\n",
    "    y_test = each_traj[-1]\n",
    "    y_test = np.array(y_test).reshape(1,2)\n",
    "    y_pred = model.predict(x_test)\n",
    "    predicted_val.append(y_pred.tolist()[0])\n",
    "        \n",
    "new_predicted_val = []\n",
    "for sublst in predicted_val:\n",
    "    for i, val in enumerate(sublst):\n",
    "        sublst[i] = round(val, 7)\n",
    "    new_predicted_val.append(sublst)    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Test Error"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "test error 0.007865015328699107\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "error = []\n",
    "for each_traj,y_pred in zip(test_set,new_predicted_val):\n",
    "    y_test = each_traj[-1]\n",
    "    error.extend(euclidean_distances([y_test],[y_pred]))\n",
    "        \n",
    "print(\"test error\",np.mean(error))   "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Example plot Predicted vs True Trajectory"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "  <tr>\n",
    "    <td> <img src=\"http://drive.google.com/uc?export=view&id=1gjgEflz3Q8GmGZ21dhjPOlJMb-6ZQ80G\" alt=\"Drawing\" style=\"width: 500px;\"/> </td>\n",
    "    <td> <img src=\"http://drive.google.com/uc?export=view&id=1T1K2H2_DTuE_xWXz1pfeAtB57QksoTEe\" alt=\"Drawing\" style=\"width: 500px;\"/> </td>\n",
    "    </tr>\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Centroids from K- Means models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "centroids = [[1.2901505322971618, 103.85151526415854],\n",
    " [1.290761314882777, 103.85183701616937],\n",
    " [1.2898866212491749, 103.85139324211389],\n",
    " [1.2903683538500839, 103.85163185742208],\n",
    " [1.2908584102736864, 103.8519792337393],\n",
    " [1.290061721528241, 103.85165098408669],\n",
    " [1.290055102350767, 103.85144672301637],\n",
    " [1.2906218405153758, 103.85173958428693],\n",
    " [1.2897596030983194, 103.85130943752183],\n",
    " [1.2902503772023322, 103.85156459994279]]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate Visitors near each Cluster - Using True Label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "from collections import Counter\n",
    "calc_cluster = []\n",
    "for each_traj in lat_lng[10000:15000]:\n",
    "    label = each_traj[-1]\n",
    "    label = np.array(label).reshape(1,2)   \n",
    "    local_cluster = []\n",
    "    for center in centroids:\n",
    "        #print(float(euclidean_distances(label.tolist(), [center])[0][0]))\n",
    "        local_cluster.append(euclidean_distances(label.tolist(), [center])[0][0]) \n",
    "    calc_cluster.append(np.argmin(local_cluster))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({7: 359,\n",
       "         1: 272,\n",
       "         8: 399,\n",
       "         6: 568,\n",
       "         0: 940,\n",
       "         4: 728,\n",
       "         2: 571,\n",
       "         9: 548,\n",
       "         3: 308,\n",
       "         5: 307})"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(calc_cluster)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Calculate Visitors near each Cluster - Using Predicted Label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics.pairwise import euclidean_distances\n",
    "from collections import Counter\n",
    "calc_cluster = []\n",
    "for each_traj in new_predicted_val:\n",
    "    label = each_traj\n",
    "    label = np.array(label).reshape(1,2)   \n",
    "    local_cluster = []\n",
    "    for center in centroids:\n",
    "        #print(float(euclidean_distances(label.tolist(), [center])[0][0]))\n",
    "        local_cluster.append(euclidean_distances(label.tolist(), [center])[0][0]) \n",
    "    calc_cluster.append(np.argmin(local_cluster))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Counter({7: 396,\n",
       "         5: 244,\n",
       "         1: 168,\n",
       "         3: 132,\n",
       "         8: 25,\n",
       "         9: 1801,\n",
       "         4: 718,\n",
       "         0: 690,\n",
       "         6: 474,\n",
       "         2: 352})"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "Counter(calc_cluster)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def np_haversine(latlon1, latlon2):\n",
    "#     \"\"\"\n",
    "#     Numpy version of the Haversine function to calculate distances between two sets of points.\n",
    "#     Returns the distance in miles.\n",
    "#     \"\"\"\n",
    "#     lat1 = latlon1[:, 0]\n",
    "#     lon1 = latlon1[:, 1]\n",
    "#     lat2 = latlon2[:, 0]\n",
    "#     lon2 = latlon2[:, 1]\n",
    "    \n",
    "#     REarth = 3961\n",
    "#     lat = np.abs(lat1 - lat2) * np.pi / 180\n",
    "#     lon = np.abs(lon1 - lon2) * np.pi / 180\n",
    "#     lat1 = lat1 * np.pi / 180\n",
    "#     lat2 = lat2 * np.pi / 180\n",
    "#     a = np.sin(lat / 2) * np.sin(lat / 2) + np.cos(lat1) * np.cos(lat2) * np.sin(lon / 2) * np.sin(lon / 2)\n",
    "#     d = 2 * np.arctan2(np.sqrt(a), np.sqrt(1 - a))\n",
    "#     return REarth * d\n",
    "# for each_traj in lat_lng[15000:20000]:\n",
    "#     label = each_traj[-1]\n",
    "#     new_predicted_val.append(label)\n",
    "    "
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}