10b3 Segment then small LSTM 32Node-Copy1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Look at accelerometer data \n",
    "\n",
    "Finding Zero velocity times by rail axis acceleration noise levels, making summary statistics for the noise levels across the whole day files.  Spot check graphs to see what works"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Standard Header used on the projects\n",
    "\n",
    "#first the major packages used for math and graphing\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from cycler import cycler\n",
    "import scipy.special as sp\n",
    "\n",
    "#Custome graph format style sheet\n",
    "#plt.style.use('Prospectus.mplstyle')\n",
    "\n",
    "#If being run by a seperate file, use the seperate file's graph format and saving paramaeters\n",
    "#otherwise set what is needed\n",
    "if not 'Saving' in locals():\n",
    "    Saving = False\n",
    "if not 'Titles' in locals():\n",
    "    Titles = True\n",
    "if not 'Ledgends' in locals():\n",
    "    Ledgends = True\n",
    "if not 'FFormat' in locals():\n",
    "    FFormat = '.png'\n",
    "\n",
    "#Standard cycle to make black and white images and dashed and line styles\n",
    "default_cycler = (cycler('color', ['0.00', '0.40', '0.60', '0.70']) + cycler(linestyle=['-', '-', '-', '-']))\n",
    "plt.rc('axes', prop_cycle=default_cycler)\n",
    "my_cmap = plt.get_cmap('gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extra Headers:\n",
    "import os as os\n",
    "import pywt as py\n",
    "import statistics as st\n",
    "import os as os\n",
    "import random\n",
    "import multiprocessing\n",
    "from joblib import Parallel, delayed\n",
    "import platform\n",
    "\n",
    "from time import time as ti"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "import CoreFunctions as cf\n",
    "from skimage.restoration import denoise_wavelet"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choosing Platform\n",
    "Working is beinging conducted on several computers, and author needs to be able to run code on all without rewriting..  This segment of determines which computer is being used, and sets the directories accordingly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "HostName = platform.node()\n",
    "\n",
    "if HostName == \"Server\":\n",
    "    Computer = \"Desktop\"   \n",
    "elif HostName[-6:] == 'wm.edu':\n",
    "    Computer = \"SciClone\"\n",
    "elif HostName == \"SchoolLaptop\":\n",
    "    Computer = \"LinLap\"\n",
    "elif HostName == \"WTC-TAB-512\":\n",
    "    Computer = \"PortLap\"\n",
    "else:\n",
    "    Computer = \"WinLap\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "if Computer == \"SciClone\":\n",
    "    location = '/sciclone/home20/dchendrickson01/image/'\n",
    "elif Computer == \"WinLap\":\n",
    "    location = 'C:\\\\Data\\\\'\n",
    "elif Computer == \"Desktop\":\n",
    "    location = \"E:\\\\Backups\\\\Dan\\\\CraneData\\\\\"\n",
    "elif Computer == \"LinLap\":\n",
    "    location = '/home/dan/Output/'\n",
    "elif Computer == 'PortLap':\n",
    "    location = 'C:\\\\users\\\\dhendrickson\\\\Desktop\\\\AccelData\\\\'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "if Computer ==  \"SciClone\":\n",
    "    rootfolder = '/sciclone/home20/dchendrickson01/'\n",
    "    folder = '/sciclone/scr10/dchendrickson01/Recordings2/'\n",
    "    imageFolder = '/sciclone/scr10/dchendrickson01/Move3Dprint/'\n",
    "elif Computer == \"Desktop\":\n",
    "    rootfolder = location\n",
    "    folder = rootfolder + \"Recordings2\\\\\"\n",
    "elif Computer ==\"WinLap\":\n",
    "    rootfolder = location\n",
    "    folder = rootfolder + \"Recordings2\\\\\"   \n",
    "elif Computer == \"LinLap\":\n",
    "    rootfolder = '/home/dan/Data/'\n",
    "    folder = rootfolder + 'Recordings2/'\n",
    "elif Computer =='PortLap':\n",
    "    rootfolder = location \n",
    "    folder = rootfolder + 'Recordings2\\\\'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Global Variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "Saving = False\n",
    "location = folder\n",
    "Titles = True\n",
    "Ledgends = True\n",
    "\n",
    "f = 0\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "files = ['230418 recording1.csv','230419 recording1.csv','230420 recording1.csv','230421 recording1.csv',\n",
    "         '230418 recording2.csv','230419 recording2.csv','230420 recording2.csv','230421 recording2.csv']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "BeforeTamping = ['221206 recording1.csv','221207 recording1.csv','221208 recording1.csv','221209 recording1.csv',\n",
    "         '221206 recording2.csv','221207 recording2.csv','221208 recording2.csv','221209 recording2.csv']\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Project Specific Functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Smooth = cf.Smoothing(ODataSet[:,3],2) #,50)\n",
    "def SmoothMoves(file):\n",
    "    #    if file[-3:] =='csv':\n",
    "    ODataSet = np.genfromtxt(open(folder+file,'r'), delimiter=',',skip_header=0,missing_values=0,invalid_raise=False)\n",
    "    SmoothX = denoise_wavelet(ODataSet[:,3], method='VisuShrink', mode='soft', wavelet_levels=3, wavelet='sym2', rescale_sigma='True')\n",
    "    SmoothY = denoise_wavelet(ODataSet[:,4], method='VisuShrink', mode='soft', wavelet_levels=3, wavelet='sym2', rescale_sigma='True')\n",
    "    SmoothZ = denoise_wavelet(ODataSet[:,5], method='VisuShrink', mode='soft', wavelet_levels=3, wavelet='sym2', rescale_sigma='True')\n",
    "    SmoothX -= np.average(SmoothX)\n",
    "    SmoothY -= np.average(SmoothY)\n",
    "    SmoothZ -= np.average(SmoothZ)\n",
    "    MoveMatrix = np.matrix([SmoothX, SmoothY, SmoothZ])\n",
    "    return MoveMatrix\n",
    "    #else:\n",
    "    #    pass\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Process Files"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "LoopFiles = 8\n",
    "loops = int(len(files) / LoopFiles) \n",
    "if len(files)%LoopFiles != 0:\n",
    "    loops += 1\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['230418 recording1.csv',\n",
       " '230419 recording1.csv',\n",
       " '230420 recording1.csv',\n",
       " '230421 recording1.csv',\n",
       " '230418 recording2.csv',\n",
       " '230419 recording2.csv',\n",
       " '230420 recording2.csv',\n",
       " '230421 recording2.csv']"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "files\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 1.792790945370992\n"
     ]
    }
   ],
   "source": [
    "\n",
    "st = ti()\n",
    "\n",
    "Moves = []\n",
    "\n",
    "for k in range(loops):\n",
    "    if k == loops -1:\n",
    "        tfiles = files[k*LoopFiles:]\n",
    "    else:\n",
    "        tfiles = files[k*LoopFiles:(k+1)*LoopFiles]\n",
    "    #Results = Parallel(n_jobs=LoopFiles)(delayed(DeviationVelocity)(file) for file in tfiles)\n",
    "    Results = Parallel(n_jobs=LoopFiles)(delayed(SmoothMoves)(file) for file in tfiles)\n",
    "    #Results =[]\n",
    "    #for file in tfiles:\n",
    "    #    Results.append(DeviationVelocity(file))\n",
    "    #    print(file, (ti()-st)/60.0)\n",
    "    for result in Results:\n",
    "        Moves.append(result)\n",
    "    print(k, (ti()-st)/60.0)\n",
    "        \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Try LSTM Stuff"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "https://machinelearningmastery.com/how-to-develop-lstm-models-for-time-series-forecasting/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "TimeSteps = 250\n",
    "StepSize = 15\n",
    "Features = np.shape(Moves[0])[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(Moves)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# split a multivariate sequence into samples\n",
    "def split_sequences(sequences, n_steps, s_step = 1):\n",
    "    X, y = list(), list()\n",
    "    Steps_to_take = int(len(sequences) / s_step)\n",
    "    for j in range(Steps_to_take):\n",
    "        i = j * s_step\n",
    "        # find the end of this pattern\n",
    "        end_ix = i + n_steps\n",
    "        # check if we are beyond the dataset\n",
    "        if end_ix > len(sequences)-1:\n",
    "            break\n",
    "        # gather input and output parts of the pattern\n",
    "        seq_x, seq_y = sequences[i:end_ix, :], sequences[end_ix, :]\n",
    "        X.append(seq_x)\n",
    "        y.append(seq_y)\n",
    "    return np.array(X), np.array(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "Sequences = []\n",
    "Outputs = []\n",
    "for move in Moves:\n",
    "    Seq, Out = split_sequences(move.T,TimeSteps,StepSize)\n",
    "    Sequences.append(Seq)\n",
    "    Outputs.append(Out)\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "MoveSegments = []\n",
    "for seq in Sequences:\n",
    "    for mv in seq:\n",
    "        MoveSegments.append(mv)\n",
    "NextDataPoint = []\n",
    "for out in Outputs:\n",
    "    for pt in out:\n",
    "        NextDataPoint.append(np.reshape(pt,(1,3)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8569639"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(MoveSegments)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.layers import LSTM, Dense, RepeatVector, TimeDistributed, Masking\n",
    "from keras.models import Sequential\n",
    "import tensorflow as tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "#tf.config.threading.set_intra_op_parallelism_threads(32)\n",
    "#tf.config.threading.set_inter_op_parallelism_threads(32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "class LSTM_Autoencoder:\n",
    "  def __init__(self, optimizer='adam', loss='mse'):\n",
    "    self.optimizer = optimizer\n",
    "    self.loss = loss\n",
    "    self.n_features = Features\n",
    "    self.timesteps = TimeSteps\n",
    "    \n",
    "  def build_model(self):\n",
    "    timesteps = self.timesteps\n",
    "    n_features = self.n_features\n",
    "    model = Sequential()\n",
    "    \n",
    "    # Padding\n",
    "    #model.add(Masking(mask_value=0.0, input_shape=(timesteps, n_features)))\n",
    "\n",
    "    # Encoder\n",
    "    model.add(LSTM(timesteps, activation='relu', input_shape=(TimeSteps, Features), return_sequences=True))\n",
    "    model.add(LSTM(35, activation='relu', return_sequences=True))\n",
    "    model.add(LSTM(6, activation='relu'))\n",
    "    model.add(RepeatVector(timesteps))\n",
    "    \n",
    "    # Decoder\n",
    "    model.add(LSTM(timesteps, activation='relu', return_sequences=True))\n",
    "    model.add(LSTM(35, activation='relu', return_sequences=True))\n",
    "    model.add(TimeDistributed(Dense(n_features)))\n",
    "    \n",
    "    model.compile(optimizer=self.optimizer, loss=self.loss, metrics=['accuracy'])\n",
    "    model.summary()\n",
    "    self.model = model\n",
    "    \n",
    "  def simple_model(self):\n",
    "    \n",
    "    # define model\n",
    "    model = Sequential()\n",
    "    model.add(LSTM(150, input_shape=(self.timesteps, self.n_features)))\n",
    "    model.add(RepeatVector(self.timesteps))\n",
    "    model.add(LSTM(150, return_sequences=True))\n",
    "    model.add(TimeDistributed(Dense(self.n_features, activation='softmax')))\n",
    "    model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])\n",
    "    model.summary()\n",
    "    \n",
    "    self.model = model\n",
    "    \n",
    "  def fit(self, X, epochs=3, batch_size=32):\n",
    "    #self.timesteps = np.shape(X)[0]\n",
    "    self.build_model()\n",
    "    \n",
    "    #input_X = np.expand_dims(X, axis=1)\n",
    "    self.model.fit(X, X, epochs=epochs, batch_size=batch_size)\n",
    "    \n",
    "  def predict(self, X):\n",
    "    #input_X = np.expand_dims(X, axis=1)\n",
    "    output_X = self.model.predict(X)\n",
    "    reconstruction = np.squeeze(output_X)\n",
    "    return np.linalg.norm(X - reconstruction, axis=-1)\n",
    "  \n",
    "  def plot(self, scores, timeseries, threshold=0.95):\n",
    "    sorted_scores = sorted(scores)\n",
    "    threshold_score = sorted_scores[round(len(scores) * threshold)]\n",
    "    \n",
    "    plt.title(\"Reconstruction Error\")\n",
    "    plt.plot(scores)\n",
    "    plt.plot([threshold_score]*len(scores), c='r')\n",
    "    plt.show()\n",
    "    \n",
    "    anomalous = np.where(scores > threshold_score)\n",
    "    normal = np.where(scores <= threshold_score)\n",
    "    \n",
    "    plt.title(\"Anomalies\")\n",
    "    plt.scatter(normal, timeseries[normal][:,-1], s=3)\n",
    "    plt.scatter(anomalous, timeseries[anomalous][:,-1], s=5, c='r')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "#split = int(len(Moves)*.9)\n",
    "#Train_data = Moves[:split]\n",
    "#Test_data = Moves[split:]\n",
    "#Train_data = tf.ragged.constant(Train_data)\n",
    "#Test_data = tf.ragged.constant(Test_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "lstm_autoencoder2 = LSTM_Autoencoder(optimizer='adam', loss='mse')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " lstm_2 (LSTM)               (None, 150)               92400     \n",
      "                                                                 \n",
      " repeat_vector_1 (RepeatVect  (None, 250, 150)         0         \n",
      " or)                                                             \n",
      "                                                                 \n",
      " lstm_3 (LSTM)               (None, 250, 150)          180600    \n",
      "                                                                 \n",
      " time_distributed_1 (TimeDis  (None, 250, 3)           453       \n",
      " tributed)                                                       \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 273,453\n",
      "Trainable params: 273,453\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "lstm_autoencoder2.simple_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "Batches = 32\n",
    "NumbBatches = 1000\n",
    "\n",
    "SamplesPerSet = Batches * NumbBatches\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32000"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "SamplesPerSet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8569639 268\n"
     ]
    }
   ],
   "source": [
    "SetsNeeded = int(len(MoveSegments) / SamplesPerSet)\n",
    "if  int(len(MoveSegments) / SamplesPerSet) != 0:\n",
    "    SetsNeeded += 1\n",
    "print(len(MoveSegments), SetsNeeded)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "PercentPerSet = 1.0 / float(SetsNeeded)\n",
    "\n",
    "PercentHoldOutForNext=1.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 334ms/step - loss: -7.9985e-04 - accuracy: 0.3265\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0052 - accuracy: 0.3352\n",
      "1 of 268 12.223058021068573 54.188891669379345\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 334ms/step - loss: -0.0056 - accuracy: 0.3614\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0056 - accuracy: 0.3614\n",
      "2 of 268 23.5812428911527 52.07524496313599\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0071 - accuracy: 0.3630\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0071 - accuracy: 0.3630\n",
      "3 of 268 34.825545084476474 51.07746623423364\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0062 - accuracy: 0.3621\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0062 - accuracy: 0.3621\n",
      "4 of 268 45.980844442049666 50.387342129680846\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0060 - accuracy: 0.3613\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0061 - accuracy: 0.3614\n",
      "5 of 268 57.1813072125117 49.93834173931016\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0056 - accuracy: 0.3611\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0056 - accuracy: 0.3613\n",
      "6 of 268 68.51659281253815 49.6745299210151\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0038 - accuracy: 0.3600\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0038 - accuracy: 0.3600\n",
      "7 of 268 80.0064582546552 49.527807552010295\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0075 - accuracy: 0.3620\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0074 - accuracy: 0.3620\n",
      "8 of 268 91.49916993379593 49.37142713979714\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0082 - accuracy: 0.3625\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0083 - accuracy: 0.3626\n",
      "9 of 268 102.70151855945588 49.06850337202902\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0061 - accuracy: 0.3625\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0067 - accuracy: 0.3633\n",
      "10 of 268 113.97177883386613 48.817911965648335\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 328s 328ms/step - loss: -0.0073 - accuracy: 0.3641\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0072 - accuracy: 0.3640\n",
      "11 of 268 125.12833726008733 48.534627819109446\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0079 - accuracy: 0.3627\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0082 - accuracy: 0.3630\n",
      "12 of 268 136.39876768191655 48.30789691087273\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0121 - accuracy: 0.3693\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 327s 328ms/step - loss: -0.0147 - accuracy: 0.3734\n",
      "13 of 268 147.51073682308197 48.03554766919368\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0131 - accuracy: 0.3745\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 326s 326ms/step - loss: -0.0140 - accuracy: 0.3759\n",
      "14 of 268 158.56619288921357 47.75862716163434\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0165 - accuracy: 0.3774\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 322s 322ms/step - loss: -0.0164 - accuracy: 0.3773\n",
      "15 of 268 169.5331784963608 47.46929001405504\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0151 - accuracy: 0.3769\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0150 - accuracy: 0.3768\n",
      "16 of 268 180.69919537703197 47.245310481385225\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0168 - accuracy: 0.3771\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0166 - accuracy: 0.3767\n",
      "17 of 268 192.0254445552826 47.065059969682316\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0142 - accuracy: 0.3752\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0143 - accuracy: 0.3754\n",
      "18 of 268 203.41748507817587 46.89903129623996\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0163 - accuracy: 0.3763\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0158 - accuracy: 0.3735\n",
      "19 of 268 214.76786630153657 46.72143057820742\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0347 - accuracy: 0.3799\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0347 - accuracy: 0.3799\n",
      "20 of 268 226.09795502026876 46.53849575272533\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0357 - accuracy: 0.3806\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0356 - accuracy: 0.3806\n",
      "21 of 268 237.37745407422383 46.34512200633685\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: -0.0349 - accuracy: 0.3792\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0348 - accuracy: 0.3792\n",
      "22 of 268 248.7721508105596 46.17361892338353\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 335ms/step - loss: -0.0360 - accuracy: 0.3823\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0361 - accuracy: 0.3824\n",
      "23 of 268 260.14457561969755 45.996577150870635\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0352 - accuracy: 0.3810\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0352 - accuracy: 0.3810\n",
      "24 of 268 271.37950824896495 45.79529203082952\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0341 - accuracy: 0.3786\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0340 - accuracy: 0.3786\n",
      "25 of 268 282.6476461927096 45.600486932664445\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0356 - accuracy: 0.3822\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0355 - accuracy: 0.3820\n",
      "26 of 268 293.80606877009075 45.389270894909515\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0366 - accuracy: 0.3838\n",
      "Epoch 2/2\n",
      " 485/1000 [=============>................] - ETA: 2:49 - loss: -0.0344 - accuracy: 0.3827"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0344 - accuracy: 0.3793\n",
      "28 of 268 316.318430741628 45.000062479509246\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0334 - accuracy: 0.3820\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0335 - accuracy: 0.3820\n",
      "29 of 268 327.5477195382118 44.80250417447639\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0340 - accuracy: 0.3804\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0338 - accuracy: 0.3802\n",
      "30 of 268 338.6815383036931 44.59306921952964\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0344 - accuracy: 0.3810\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0344 - accuracy: 0.3810\n",
      "31 of 268 350.10967006285983 44.42251729844718\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0339 - accuracy: 0.3795\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: -0.0339 - accuracy: 0.3794\n",
      "32 of 268 361.62882203261057 44.26186104095437\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0353 - accuracy: 0.3788\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0351 - accuracy: 0.3786\n",
      "33 of 268 372.9963386297226 44.08138548120103\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0343 - accuracy: 0.3797\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0343 - accuracy: 0.3798\n",
      "34 of 268 384.4195011695226 43.90673715446315\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0332 - accuracy: 0.3794\n",
      "Epoch 2/2\n",
      " 641/1000 [==================>...........] - ETA: 1:58 - loss: -0.0328 - accuracy: 0.3797"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0335 - accuracy: 0.3795\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 345s 345ms/step - loss: -0.0335 - accuracy: 0.3795\n",
      "37 of 268 418.2961110472679 43.33698449833614\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0333 - accuracy: 0.3782\n",
      "Epoch 2/2\n",
      " 256/1000 [======>.......................] - ETA: 4:08 - loss: -0.0329 - accuracy: 0.3763"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0324 - accuracy: 0.3785\n",
      "39 of 268 440.85323236385983 42.95493033797822\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 333ms/step - loss: -0.0321 - accuracy: 0.3761\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0319 - accuracy: 0.3759\n",
      "40 of 268 452.13778704007467 42.76469903406833\n",
      "Epoch 1/2\n",
      " 846/1000 [========================>.....] - ETA: 51s - loss: -0.0331 - accuracy: 0.3791"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0304 - accuracy: 0.3748\n",
      "42 of 268 474.63103036880494 42.37777057707309\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0298 - accuracy: 0.3753\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0299 - accuracy: 0.3754\n",
      "43 of 268 485.9128108859062 42.187778936410446\n",
      "Epoch 1/2\n",
      "  33/1000 [..............................] - ETA: 5:12 - loss: -0.0303 - accuracy: 0.3893"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 326s 325ms/step - loss: -0.0312 - accuracy: 0.3788\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0311 - accuracy: 0.3788\n",
      "45 of 268 508.24661492506664 41.78916612187138\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 325s 325ms/step - loss: -0.0310 - accuracy: 0.3769\n",
      "Epoch 2/2\n",
      " 583/1000 [================>.............] - ETA: 2:17 - loss: -0.0301 - accuracy: 0.3769"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0304 - accuracy: 0.3759\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0297 - accuracy: 0.3751\n",
      "48 of 268 541.6056706309319 41.18459787781454\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0307 - accuracy: 0.3772\n",
      "Epoch 2/2\n",
      " 378/1000 [==========>...................] - ETA: 3:38 - loss: -0.0313 - accuracy: 0.3787"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0305 - accuracy: 0.3742\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0303 - accuracy: 0.3740\n",
      "51 of 268 575.3609174132347 40.6137118208681\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 327s 327ms/step - loss: -0.0313 - accuracy: 0.3754\n",
      "Epoch 2/2\n",
      " 180/1000 [====>.........................] - ETA: 4:34 - loss: -0.0295 - accuracy: 0.3703"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 328s 328ms/step - loss: -0.0283 - accuracy: 0.3728\n",
      "53 of 268 597.5308717926343 40.211197036533484\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 327s 327ms/step - loss: -0.0319 - accuracy: 0.3773\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0318 - accuracy: 0.3773\n",
      "54 of 268 608.6181954542795 40.01101100235074\n",
      "Epoch 1/2\n",
      " 455/1000 [============>.................] - ETA: 3:01 - loss: -0.0301 - accuracy: 0.3768"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 324s 324ms/step - loss: -0.0293 - accuracy: 0.3764\n",
      "56 of 268 630.8272254427274 39.614447793377536\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0294 - accuracy: 0.3747\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0294 - accuracy: 0.3746\n",
      "57 of 268 642.1818404992422 39.43221827942029\n",
      "Epoch 1/2\n",
      " 205/1000 [=====>........................] - ETA: 4:31 - loss: -0.0279 - accuracy: 0.3750"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0314 - accuracy: 0.3768\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0310 - accuracy: 0.3765\n",
      "59 of 268 664.8850362022717 39.066691396181426\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 330ms/step - loss: -0.0274 - accuracy: 0.3740\n",
      "Epoch 2/2\n",
      " 890/1000 [=========================>....] - ETA: 36s - loss: -0.0266 - accuracy: 0.3731"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 346s 346ms/step - loss: -0.0279 - accuracy: 0.3719\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0294 - accuracy: 0.3735\n",
      "62 of 268 698.7144554336866 38.50442564127907\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0296 - accuracy: 0.3754\n",
      "Epoch 2/2\n",
      " 681/1000 [===================>..........] - ETA: 1:49 - loss: -0.0299 - accuracy: 0.3771"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0302 - accuracy: 0.3759\n",
      "64 of 268 721.6901900450389 38.151851200004\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0296 - accuracy: 0.3748\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0302 - accuracy: 0.3755\n",
      "65 of 268 732.9512048681577 37.96311369718649\n",
      "Epoch 1/2\n",
      " 804/1000 [=======================>......] - ETA: 1:06 - loss: -0.0281 - accuracy: 0.3746"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0304 - accuracy: 0.3760\n",
      "67 of 268 755.5115949869156 37.587641544469555\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0293 - accuracy: 0.3742\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: -0.0305 - accuracy: 0.3755\n",
      "68 of 268 766.8409030914306 37.40228915013519\n",
      "Epoch 1/2\n",
      " 437/1000 [============>.................] - ETA: 3:09 - loss: -0.0319 - accuracy: 0.3778"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0289 - accuracy: 0.3745\n",
      "70 of 268 789.4462977647781 37.02879063664921\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0285 - accuracy: 0.3731\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0282 - accuracy: 0.3727\n",
      "71 of 268 800.9116735816002 36.84945728711306\n",
      "Epoch 1/2\n",
      "  92/1000 [=>............................] - ETA: 5:09 - loss: -0.0333 - accuracy: 0.3780"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0323 - accuracy: 0.3773\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 339ms/step - loss: -0.0313 - accuracy: 0.3762\n",
      "73 of 268 823.6300270954767 36.48041672996099\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: -0.0321 - accuracy: 0.3774\n",
      "Epoch 2/2\n",
      " 868/1000 [=========================>....] - ETA: 44s - loss: -0.0333 - accuracy: 0.3781"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 342s 341ms/step - loss: -0.0336 - accuracy: 0.3813\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 336ms/step - loss: -0.0338 - accuracy: 0.3816\n",
      "76 of 268 857.5827196637789 35.92067970847962\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0358 - accuracy: 0.3821\n",
      "Epoch 2/2\n",
      " 594/1000 [================>.............] - ETA: 2:17 - loss: -0.0354 - accuracy: 0.3827"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0333 - accuracy: 0.3803\n",
      "78 of 268 880.4607028961182 35.55706685150281\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0340 - accuracy: 0.3797\n",
      "Epoch 2/2\n",
      " 941/1000 [===========================>..] - ETA: 20s - loss: -0.0342 - accuracy: 0.3806"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 344s 344ms/step - loss: -0.0374 - accuracy: 0.4118\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0369 - accuracy: 0.4114\n",
      "81 of 268 914.4611688137054 34.997896586700726\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0371 - accuracy: 0.4095\n",
      "Epoch 2/2\n",
      " 563/1000 [===============>..............] - ETA: 2:25 - loss: -0.0389 - accuracy: 0.4097"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0350 - accuracy: 0.4093\n",
      "83 of 268 936.9386945883433 34.617815223822156\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0340 - accuracy: 0.4078\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0339 - accuracy: 0.4075\n",
      "84 of 268 948.310743077596 34.432711507458066\n",
      "Epoch 1/2\n",
      " 911/1000 [==========================>...] - ETA: 30s - loss: -0.0359 - accuracy: 0.4090"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0338 - accuracy: 0.4057\n",
      "86 of 268 970.817479177316 34.05386894664715\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0344 - accuracy: 0.4077\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0346 - accuracy: 0.4080\n",
      "87 of 268 981.9471739888191 33.86024738106006\n",
      "Epoch 1/2\n",
      " 521/1000 [==============>...............] - ETA: 2:38 - loss: -0.0340 - accuracy: 0.4085"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0351 - accuracy: 0.4061\n",
      "89 of 268 1004.4685119072597 33.482283731963896\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0334 - accuracy: 0.4059\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0332 - accuracy: 0.4055\n",
      "90 of 268 1015.734030298392 33.29350432826634\n",
      "Epoch 1/2\n",
      " 119/1000 [==>...........................] - ETA: 4:42 - loss: -0.0297 - accuracy: 0.3997"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0331 - accuracy: 0.4065\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -0.0316 - accuracy: 0.4041\n",
      "92 of 268 1038.2525842229525 32.915616355313055\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 327s 327ms/step - loss: -0.0299 - accuracy: 0.3987\n",
      "Epoch 2/2\n",
      " 766/1000 [=====================>........] - ETA: 1:16 - loss: -0.0255 - accuracy: 0.3918"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0033 - accuracy: 0.3562\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0019 - accuracy: 0.3544\n",
      "95 of 268 1072.0887285391489 32.35074760127695\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: 4.1712e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      " 335/1000 [=========>....................] - ETA: 3:46 - loss: -9.8712e-04 - accuracy: 0.3541"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 336s 336ms/step - loss: -1.4260e-04 - accuracy: 0.3534\n",
      "97 of 268 1094.7250074068704 31.976503654138053\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: 4.9084e-04 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: 5.5100e-04 - accuracy: 0.3529\n",
      "98 of 268 1105.8974527994792 31.785147880147512\n",
      "Epoch 1/2\n",
      " 300/1000 [========>.....................] - ETA: 3:43 - loss: 4.6629e-04 - accuracy: 0.3525"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 323s 323ms/step - loss: -0.0013 - accuracy: 0.3535\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0013 - accuracy: 0.3535\n",
      "100 of 268 1128.0148546179137 31.396413454527988\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -6.8131e-04 - accuracy: 0.3538\n",
      "Epoch 2/2\n",
      " 962/1000 [===========================>..] - ETA: 12s - loss: 8.4258e-05 - accuracy: 0.3534"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 331s 331ms/step - loss: -5.2592e-04 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 327s 327ms/step - loss: -5.2592e-04 - accuracy: 0.3530\n",
      "103 of 268 1161.6734944105149 30.827581406831744\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -2.6744e-04 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      " 608/1000 [=================>............] - ETA: 2:07 - loss: -8.1817e-05 - accuracy: 0.3532"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 340s 340ms/step - loss: 0.0019 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: 0.0019 - accuracy: 0.3523\n",
      "106 of 268 1195.365926082929 30.260049388810398\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: 0.0022 - accuracy: 0.3509\n",
      "Epoch 2/2\n",
      " 379/1000 [==========>...................] - ETA: 3:26 - loss: 0.0020 - accuracy: 0.3499"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 337s 337ms/step - loss: 4.0899e-04 - accuracy: 0.3512\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 4.0899e-04 - accuracy: 0.3512\n",
      "109 of 268 1229.4410314043364 29.702092197307753\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: 5.4350e-04 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "  24/1000 [..............................] - ETA: 5:18 - loss: -0.0142 - accuracy: 0.3549"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 331s 331ms/step - loss: -2.1157e-04 - accuracy: 0.3537\n",
      "111 of 268 1251.8412502328554 29.32240766408387\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: 4.2028e-04 - accuracy: 0.3533\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 358s 358ms/step - loss: 4.2028e-04 - accuracy: 0.3533\n",
      "112 of 268 1263.4433195988338 29.14192180754174\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: 4.9460e-04 - accuracy: 0.3527\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: 4.9460e-04 - accuracy: 0.3527\n",
      "113 of 268 1274.9293842116992 28.95857303537157\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 328s 328ms/step - loss: 9.4151e-04 - accuracy: 0.3532\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: 9.4151e-04 - accuracy: 0.3532\n",
      "114 of 268 1286.1736437916757 28.769673614578643\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 348s 348ms/step - loss: 0.0017 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: 0.0017 - accuracy: 0.3523\n",
      "115 of 268 1297.7729113141695 28.58862065588621\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0027 - accuracy: 0.3552\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0027 - accuracy: 0.3552\n",
      "116 of 268 1309.1072582681975 28.40160862227167\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -1.6493e-04 - accuracy: 0.3526\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -1.6493e-04 - accuracy: 0.3526\n",
      "117 of 268 1320.3676007231077 28.212982923008102\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 1.1979e-04 - accuracy: 0.3538\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 1.1979e-04 - accuracy: 0.3538\n",
      "118 of 268 1331.6470882217088 28.02477629292932\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: 0.0025 - accuracy: 0.3514\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: 0.0025 - accuracy: 0.3514\n",
      "119 of 268 1343.067617090543 27.839496825519397\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: 3.2431e-04 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 3.2431e-04 - accuracy: 0.3528\n",
      "120 of 268 1354.392256573836 27.652175239268278\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0015 - accuracy: 0.3535\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -0.0015 - accuracy: 0.3535\n",
      "121 of 268 1365.4403736154238 27.459269222268503\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 0.0025 - accuracy: 0.3511\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: 0.0025 - accuracy: 0.3511\n",
      "122 of 268 1376.6193023880323 27.26909820403612\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: -9.6911e-04 - accuracy: 0.3549\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -9.6911e-04 - accuracy: 0.3549\n",
      "123 of 268 1387.7601316213609 27.078246472146773\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -1.4040e-05 - accuracy: 0.3527\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 343ms/step - loss: -1.4039e-05 - accuracy: 0.3527\n",
      "124 of 268 1399.1860876758894 26.89295840839866\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: 7.8611e-04 - accuracy: 0.3522\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 346s 346ms/step - loss: 7.8611e-04 - accuracy: 0.3522\n",
      "125 of 268 1410.7570235093435 26.710332979325184\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: 0.0013 - accuracy: 0.3524\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: 0.0013 - accuracy: 0.3524\n",
      "126 of 268 1422.2510396083196 26.52611066044638\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 0.0019 - accuracy: 0.3524\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 0.0019 - accuracy: 0.3524\n",
      "127 of 268 1433.5950890580814 26.339017385331974\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: -0.0011 - accuracy: 0.3537\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 347s 347ms/step - loss: -0.0011 - accuracy: 0.3537\n",
      "128 of 268 1445.077559554577 26.154398540070073\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0019 - accuracy: 0.3539\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -0.0019 - accuracy: 0.3539\n",
      "129 of 268 1456.3572269042334 25.966059085216944\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 6.8495e-04 - accuracy: 0.3516\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0013 - accuracy: 0.3540\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: -0.0013 - accuracy: 0.3540\n",
      "131 of 268 1479.099834438165 25.592567110994388\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -9.2450e-04 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -9.2450e-04 - accuracy: 0.3525\n",
      "132 of 268 1490.485957356294 25.40601063757382\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: 8.1870e-04 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: 8.1870e-04 - accuracy: 0.3528\n",
      "133 of 268 1501.9928508202236 25.221433836967623\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: -7.4258e-04 - accuracy: 0.3542\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: -7.4258e-04 - accuracy: 0.3542\n",
      "134 of 268 1513.4868004600207 25.036535382120565\n",
      "Epoch 1/2\n",
      " 920/1000 [==========================>...] - ETA: 26s - loss: -3.8166e-04 - accuracy: 0.3534"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 344s 344ms/step - loss: 1.8068e-04 - accuracy: 0.3524\n",
      "136 of 268 1536.3345361193021 24.664194147171436\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 347s 347ms/step - loss: -0.0022 - accuracy: 0.3541\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0022 - accuracy: 0.3541\n",
      "137 of 268 1547.8551500121753 24.47946101153177\n",
      "Epoch 1/2\n",
      " 573/1000 [================>.............] - ETA: 2:25 - loss: -0.0015 - accuracy: 0.3538"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 342s 342ms/step - loss: -9.3422e-04 - accuracy: 0.3539\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: -9.3422e-04 - accuracy: 0.3539\n",
      "139 of 268 1570.762783018748 24.10763024404514\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: -7.3276e-04 - accuracy: 0.3529\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -7.3276e-04 - accuracy: 0.3529\n",
      "140 of 268 1582.231929520766 23.921839887726396\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 325s 325ms/step - loss: 0.0022 - accuracy: 0.3522\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 0.0022 - accuracy: 0.3522\n",
      "141 of 268 1593.2656536420186 23.729488459047698\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -0.0011 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0011 - accuracy: 0.3530\n",
      "142 of 268 1604.5646605014801 23.54114818916075\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 346s 346ms/step - loss: -0.0037 - accuracy: 0.3542\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: -0.0037 - accuracy: 0.3542\n",
      "143 of 268 1616.0775570869446 23.355899427748042\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 347s 347ms/step - loss: -5.8394e-04 - accuracy: 0.3541\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -5.8394e-04 - accuracy: 0.3541\n",
      "144 of 268 1627.5819973270097 23.170438157068915\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 5.4754e-04 - accuracy: 0.3522\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 5.4754e-04 - accuracy: 0.3522\n",
      "145 of 268 1638.8618882258734 22.981741422083644\n",
      "Epoch 1/2\n",
      " 345/1000 [=========>....................] - ETA: 3:39 - loss: -0.0030 - accuracy: 0.3552"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 339s 339ms/step - loss: 0.0017 - accuracy: 0.3526\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 0.0017 - accuracy: 0.3526\n",
      "147 of 268 1661.562670036157 22.60629483130513\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0026 - accuracy: 0.3540\n",
      "Epoch 2/2\n",
      " 484/1000 [=============>................] - ETA: 2:53 - loss: -0.0034 - accuracy: 0.3546"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 334s 334ms/step - loss: -7.3512e-04 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -7.3512e-04 - accuracy: 0.3523\n",
      "150 of 268 1695.5412578384082 22.042036352660922\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 0.0025 - accuracy: 0.3507\n",
      "Epoch 2/2\n",
      " 122/1000 [==>...........................] - ETA: 5:11 - loss: 0.0046 - accuracy: 0.3489"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 343s 343ms/step - loss: -5.1352e-04 - accuracy: 0.3538\n",
      "152 of 268 1718.4667603453001 21.669262878036026\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: 4.7646e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 4.7646e-04 - accuracy: 0.3531\n",
      "153 of 268 1729.7567190289496 21.48063899546018\n",
      "Epoch 1/2\n",
      " 777/1000 [======================>.......] - ETA: 1:13 - loss: 0.0023 - accuracy: 0.3514"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 333s 333ms/step - loss: 5.8623e-04 - accuracy: 0.3530\n",
      "155 of 268 1751.8550494591395 21.097609198770645\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 1.0740e-04 - accuracy: 0.3541\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: 1.0740e-04 - accuracy: 0.3541\n",
      "156 of 268 1763.031604429086 20.907746591276332\n",
      "Epoch 1/2\n",
      " 394/1000 [==========>...................] - ETA: 3:23 - loss: 8.4468e-04 - accuracy: 0.3529"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 329s 329ms/step - loss: 3.5647e-04 - accuracy: 0.3533\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 3.5647e-04 - accuracy: 0.3533\n",
      "158 of 268 1785.1504719138145 20.525464287355394\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 319s 319ms/step - loss: 2.7714e-04 - accuracy: 0.3526\n",
      "Epoch 2/2\n",
      " 478/1000 [=============>................] - ETA: 2:52 - loss: 0.0037 - accuracy: 0.3505"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 326s 326ms/step - loss: 8.9765e-04 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: 8.9765e-04 - accuracy: 0.3523\n",
      "161 of 268 1818.1008519927661 19.950174981085407\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: 8.8416e-04 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      " 114/1000 [==>...........................] - ETA: 4:59 - loss: 0.0047 - accuracy: 0.3490"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 342s 342ms/step - loss: 0.0029 - accuracy: 0.3515\n",
      "163 of 268 1840.5945292870204 19.572784361012065\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0018 - accuracy: 0.3538\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: -0.0018 - accuracy: 0.3538\n",
      "164 of 268 1851.8694649338722 19.3844059850744\n",
      "Epoch 1/2\n",
      " 746/1000 [=====================>........] - ETA: 1:24 - loss: 0.0015 - accuracy: 0.3528"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 339s 339ms/step - loss: 0.0010 - accuracy: 0.3530\n",
      "166 of 268 1874.3310764153798 19.006770956112515\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -0.0019 - accuracy: 0.3539\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0019 - accuracy: 0.3539\n",
      "167 of 268 1885.6069487015407 18.818432622312386\n",
      "Epoch 1/2\n",
      " 349/1000 [=========>....................] - ETA: 3:40 - loss: -0.0037 - accuracy: 0.3555"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 334s 334ms/step - loss: 2.9179e-05 - accuracy: 0.3523\n",
      "169 of 268 1908.0849575718244 18.441057775772617\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 328s 328ms/step - loss: 0.0015 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: 0.0015 - accuracy: 0.3528\n",
      "170 of 268 1919.1041216810545 18.250303902703948\n",
      "Epoch 1/2\n",
      "   3/1000 [..............................] - ETA: 5:00 - loss: -0.0274 - accuracy: 0.3730"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 330s 330ms/step - loss: -0.0027 - accuracy: 0.3534\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0027 - accuracy: 0.3534\n",
      "172 of 268 1941.4120213707288 17.87152539104031\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -7.1395e-04 - accuracy: 0.3534\n",
      "Epoch 2/2\n",
      " 663/1000 [==================>...........] - ETA: 1:50 - loss: 4.4160e-05 - accuracy: 0.3532"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 328s 328ms/step - loss: -0.0021 - accuracy: 0.3548\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 322s 322ms/step - loss: -0.0021 - accuracy: 0.3548\n",
      "175 of 268 1974.6362264593442 17.301574556465003\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 329s 329ms/step - loss: 1.7048e-04 - accuracy: 0.3535\n",
      "Epoch 2/2\n",
      " 300/1000 [========>.....................] - ETA: 3:55 - loss: 0.0028 - accuracy: 0.3516"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 328s 328ms/step - loss: -0.0020 - accuracy: 0.3531\n",
      "177 of 268 1996.695516359806 16.921148445113456\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: 0.0010 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 0.0010 - accuracy: 0.3525\n",
      "178 of 268 2007.7969865441323 16.731641554865572\n",
      "Epoch 1/2\n",
      " 651/1000 [==================>...........] - ETA: 1:56 - loss: -9.6464e-04 - accuracy: 0.3535"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 338s 338ms/step - loss: -7.0626e-04 - accuracy: 0.3527\n",
      "180 of 268 2030.0839351614316 16.35345392327287\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 332s 332ms/step - loss: 0.0011 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: 0.0011 - accuracy: 0.3530\n",
      "181 of 268 2041.234365594387 16.164471035274424\n",
      "Epoch 1/2\n",
      " 253/1000 [======>.......................] - ETA: 4:01 - loss: -6.3423e-04 - accuracy: 0.3531"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 331s 331ms/step - loss: 0.0014 - accuracy: 0.3520\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 0.0014 - accuracy: 0.3520\n",
      "183 of 268 2063.3617957433066 15.78528149772863\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 2.6955e-04 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      " 969/1000 [============================>.] - ETA: 10s - loss: -2.2285e-05 - accuracy: 0.3525"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 338s 338ms/step - loss: -7.8618e-04 - accuracy: 0.3531\n",
      "185 of 268 2085.829997376601 15.40883421503567\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 5.1102e-04 - accuracy: 0.3519\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: 5.1101e-04 - accuracy: 0.3519\n",
      "186 of 268 2097.2304173231123 15.22183367420696\n",
      "Epoch 1/2\n",
      " 186/1000 [====>.........................] - ETA: 4:46 - loss: -0.0013 - accuracy: 0.3535"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 339s 339ms/step - loss: 6.6508e-04 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: 6.6508e-04 - accuracy: 0.3525\n",
      "188 of 268 2119.7480486710865 14.845753178538379\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: 7.4969e-04 - accuracy: 0.3529\n",
      "Epoch 2/2\n",
      " 799/1000 [======================>.......] - ETA: 1:06 - loss: 6.1517e-04 - accuracy: 0.3535"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 356s 356ms/step - loss: 0.0016 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 369s 369ms/step - loss: 0.0016 - accuracy: 0.3531\n",
      "191 of 268 2154.602109917005 14.288809804103966\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 372s 372ms/step - loss: -5.2614e-04 - accuracy: 0.3533\n",
      "Epoch 2/2\n",
      " 406/1000 [===========>..................] - ETA: 3:31 - loss: 3.7684e-04 - accuracy: 0.3531"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 373s 373ms/step - loss: -0.0026 - accuracy: 0.3549\n",
      "193 of 268 2179.1438779592513 13.925444470615515\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -4.7202e-06 - accuracy: 0.3532\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 356s 356ms/step - loss: -4.7205e-06 - accuracy: 0.3532\n",
      "194 of 268 2190.7979083657265 13.739540147134457\n",
      "Epoch 1/2\n",
      " 987/1000 [============================>.] - ETA: 4s - loss: -8.7285e-04 - accuracy: 0.3534"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 358s 357ms/step - loss: -3.9364e-04 - accuracy: 0.3535\n",
      "196 of 268 2214.9938061237335 13.372836754753548\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 371s 371ms/step - loss: 0.0018 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 352s 352ms/step - loss: 0.0018 - accuracy: 0.3530\n",
      "197 of 268 2227.0746252934136 13.18910522565691\n",
      "Epoch 1/2\n",
      " 669/1000 [===================>..........] - ETA: 1:51 - loss: 0.0017 - accuracy: 0.3523"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 349s 349ms/step - loss: 7.3935e-04 - accuracy: 0.3535\n",
      "199 of 268 2250.8512924075126 12.818918583238677\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 359s 359ms/step - loss: -7.1462e-05 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 348s 348ms/step - loss: -7.1461e-05 - accuracy: 0.3525\n",
      "200 of 268 2262.6770835876464 12.633280383322965\n",
      "Epoch 1/2\n",
      " 349/1000 [=========>....................] - ETA: 3:57 - loss: 3.1722e-04 - accuracy: 0.3530"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 376s 376ms/step - loss: -9.6940e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 360s 361ms/step - loss: -9.6940e-04 - accuracy: 0.3531\n",
      "202 of 268 2286.9155303955076 12.264811012665886\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 364s 364ms/step - loss: -0.0014 - accuracy: 0.3545\n",
      "Epoch 2/2\n",
      " 830/1000 [=======================>......] - ETA: 1:00 - loss: -0.0012 - accuracy: 0.3541"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 371s 371ms/step - loss: 2.8868e-04 - accuracy: 0.3540\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 367s 367ms/step - loss: 2.8868e-04 - accuracy: 0.3540\n",
      "205 of 268 2323.103461976846 11.709952409929375\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 377s 377ms/step - loss: 3.3074e-04 - accuracy: 0.3534\n",
      "Epoch 2/2\n",
      " 293/1000 [=======>......................] - ETA: 4:38 - loss: 9.3093e-04 - accuracy: 0.3531"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 376s 376ms/step - loss: 4.2660e-04 - accuracy: 0.3531\n",
      "207 of 268 2348.5646534641583 11.345722963649678\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 378s 378ms/step - loss: 0.0012 - accuracy: 0.3519\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 373s 373ms/step - loss: 0.0012 - accuracy: 0.3519\n",
      "208 of 268 2361.1210422794024 11.162351081173263\n",
      "Epoch 1/2\n",
      " 888/1000 [=========================>....] - ETA: 42s - loss: -0.0013 - accuracy: 0.3547"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 368s 368ms/step - loss: -7.1090e-04 - accuracy: 0.3531\n",
      "210 of 268 2385.9875695705414 10.793753290641876\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 367s 367ms/step - loss: 2.3913e-04 - accuracy: 0.3518\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 373s 373ms/step - loss: 2.3913e-04 - accuracy: 0.3518\n",
      "211 of 268 2398.3553045988083 10.608838629945135\n",
      "Epoch 1/2\n",
      " 241/1000 [======>.......................] - ETA: 4:58 - loss: 0.0035 - accuracy: 0.3538"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 395s 395ms/step - loss: 0.0013 - accuracy: 0.3516\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 372s 371ms/step - loss: 0.0013 - accuracy: 0.3516\n",
      "213 of 268 2424.1957341273624 10.243080567092589\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 381s 381ms/step - loss: -8.2003e-04 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      " 621/1000 [=================>............] - ETA: 2:22 - loss: -0.0021 - accuracy: 0.3531"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 371s 371ms/step - loss: 0.0010 - accuracy: 0.3526\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 387s 387ms/step - loss: 0.0010 - accuracy: 0.3526\n",
      "216 of 268 2462.6124812165895 9.690836152683614\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 365s 365ms/step - loss: -9.2972e-04 - accuracy: 0.3543\n",
      "Epoch 2/2\n",
      " 121/1000 [==>...........................] - ETA: 4:44 - loss: -9.3869e-04 - accuracy: 0.3535"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 343s 343ms/step - loss: 0.0024 - accuracy: 0.3510\n",
      "218 of 268 2485.744539443652 9.312039941012795\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 349s 349ms/step - loss: 0.0011 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 0.0011 - accuracy: 0.3531\n",
      "219 of 268 2497.242683843772 9.12234770337832\n",
      "Epoch 1/2\n",
      " 500/1000 [==============>...............] - ETA: 2:53 - loss: -0.0028 - accuracy: 0.3548"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 343s 343ms/step - loss: 7.2658e-05 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 7.2658e-05 - accuracy: 0.3530\n",
      "221 of 268 2520.11059991916 8.742465127725703\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 337s 337ms/step - loss: 8.4416e-05 - accuracy: 0.3526\n",
      "Epoch 2/2\n",
      " 895/1000 [=========================>....] - ETA: 35s - loss: -5.0421e-04 - accuracy: 0.3531"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 343s 343ms/step - loss: 6.6377e-04 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 6.6377e-04 - accuracy: 0.3528\n",
      "224 of 268 2554.5218817035357 8.172949472276104\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: 0.0015 - accuracy: 0.3524\n",
      "Epoch 2/2\n",
      " 281/1000 [=======>......................] - ETA: 4:10 - loss: 0.0026 - accuracy: 0.3526"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 344s 344ms/step - loss: 0.0014 - accuracy: 0.3520\n",
      "226 of 268 2577.380089863141 7.7929633982457585\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -2.4063e-04 - accuracy: 0.3540\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: -2.4063e-04 - accuracy: 0.3540\n",
      "227 of 268 2588.7660742084186 7.602837222028869\n",
      "Epoch 1/2\n",
      " 680/1000 [===================>..........] - ETA: 1:48 - loss: 0.0010 - accuracy: 0.3526"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 341s 341ms/step - loss: -7.6542e-04 - accuracy: 0.3534\n",
      "229 of 268 2611.630030186971 7.222848700396587\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 9.5965e-04 - accuracy: 0.3527\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 9.5965e-04 - accuracy: 0.3527\n",
      "230 of 268 2622.9586127916973 7.032570193387178\n",
      "Epoch 1/2\n",
      " 195/1000 [====>.........................] - ETA: 4:36 - loss: 0.0013 - accuracy: 0.3511  "
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 340s 340ms/step - loss: -6.0246e-04 - accuracy: 0.3532\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 343s 343ms/step - loss: -6.0246e-04 - accuracy: 0.3532\n",
      "232 of 268 2645.7968527674675 6.652506454322074\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 345s 345ms/step - loss: 3.3050e-04 - accuracy: 0.3528\n",
      "Epoch 2/2\n",
      " 781/1000 [======================>.......] - ETA: 1:16 - loss: -7.8074e-04 - accuracy: 0.3535"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 342s 342ms/step - loss: 6.6135e-04 - accuracy: 0.3529\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 6.6135e-04 - accuracy: 0.3529\n",
      "235 of 268 2680.1894395709037 6.0826994370286895\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 0.0012 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      " 287/1000 [=======>......................] - ETA: 4:00 - loss: 8.8227e-04 - accuracy: 0.3526"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 333s 333ms/step - loss: -0.0025 - accuracy: 0.3545\n",
      "237 of 268 2702.8742645462353 5.7022663805789895\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: 1.1363e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 1.1363e-04 - accuracy: 0.3531\n",
      "238 of 268 2714.128835527102 5.5118862903162675\n",
      "Epoch 1/2\n",
      " 767/1000 [======================>.......] - ETA: 1:18 - loss: -0.0012 - accuracy: 0.3529"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 327s 327ms/step - loss: 7.1388e-04 - accuracy: 0.3528\n",
      "240 of 268 2736.3427851835886 5.13064272189803\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -7.5721e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -7.5721e-04 - accuracy: 0.3531\n",
      "241 of 268 2747.495328092575 4.940171405418563\n",
      "Epoch 1/2\n",
      " 179/1000 [====>.........................] - ETA: 4:29 - loss: 8.6031e-04 - accuracy: 0.3511"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 332s 332ms/step - loss: 9.4778e-04 - accuracy: 0.3527\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 330s 330ms/step - loss: 9.4778e-04 - accuracy: 0.3527\n",
      "243 of 268 2769.7240369757014 4.55921652141095\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: -8.8910e-04 - accuracy: 0.3541\n",
      "Epoch 2/2\n",
      " 578/1000 [================>.............] - ETA: 2:22 - loss: 3.1341e-04 - accuracy: 0.3545"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 336s 336ms/step - loss: -7.9610e-04 - accuracy: 0.3533\n",
      "245 of 268 2792.269664112727 4.178906979851438\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 331s 331ms/step - loss: -2.5058e-04 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: -2.5058e-04 - accuracy: 0.3525\n",
      "246 of 268 2803.340994445483 3.9884932843084893\n",
      "Epoch 1/2\n",
      " 988/1000 [============================>.] - ETA: 3s - loss: -0.0021 - accuracy: 0.3546"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 334s 334ms/step - loss: 0.0026 - accuracy: 0.3512\n",
      "248 of 268 2825.680064670245 3.6080592218968053\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 1.4684e-04 - accuracy: 0.3529\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -1.0557e-04 - accuracy: 0.3529\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -1.0557e-04 - accuracy: 0.3529\n",
      "250 of 268 2848.593870762984 3.228406386509208\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 0.0013 - accuracy: 0.3525\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 0.0013 - accuracy: 0.3525\n",
      "251 of 268 2859.9037686785064 3.0384103780393827\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 342s 342ms/step - loss: -0.0010 - accuracy: 0.3538\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: -0.0010 - accuracy: 0.3538\n",
      "252 of 268 2871.2535873413085 2.848465859937294\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 6.6277e-04 - accuracy: 0.3524\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: 6.6277e-04 - accuracy: 0.3524\n",
      "253 of 268 2882.497596995036 2.6584299307928125\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -0.0014 - accuracy: 0.3540\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: -0.0014 - accuracy: 0.3540\n",
      "254 of 268 2893.7577516118686 2.4684285276017426\n",
      "Epoch 1/2\n",
      " 783/1000 [======================>.......] - ETA: 1:14 - loss: 5.3737e-04 - accuracy: 0.3537"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1000/1000 [==============================] - 340s 340ms/step - loss: 0.0021 - accuracy: 0.3522\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 335s 335ms/step - loss: 0.0021 - accuracy: 0.3522\n",
      "256 of 268 2916.380101931095 2.0885534583161482\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 346s 346ms/step - loss: -1.0625e-04 - accuracy: 0.3530\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 334s 334ms/step - loss: -1.0625e-04 - accuracy: 0.3530\n",
      "257 of 268 2927.723980863889 1.8986536835616588\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 339s 339ms/step - loss: -9.3068e-04 - accuracy: 0.3542\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 345s 345ms/step - loss: -9.3068e-04 - accuracy: 0.3542\n",
      "258 of 268 2939.1344880104066 1.7087991205373854\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 344s 344ms/step - loss: 2.4009e-04 - accuracy: 0.3523\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 338s 338ms/step - loss: 2.4009e-04 - accuracy: 0.3523\n",
      "259 of 268 2950.5052461147307 1.5189216190990797\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: 2.9460e-04 - accuracy: 0.3536\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 333s 333ms/step - loss: 2.9460e-04 - accuracy: 0.3536\n",
      "260 of 268 2961.751649308205 1.3289911241434635\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 336s 336ms/step - loss: -5.2244e-04 - accuracy: 0.3533\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: -5.2244e-04 - accuracy: 0.3533\n",
      "261 of 268 2973.030351952712 1.1390920883173548\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -2.0982e-04 - accuracy: 0.3531\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0011 - accuracy: 0.3527\n",
      "Epoch 2/2\n",
      "1000/1000 [==============================] - 341s 341ms/step - loss: -0.0011 - accuracy: 0.3527\n",
      "263 of 268 2995.8336660901705 0.7594001683435393\n",
      "Epoch 1/2\n",
      "1000/1000 [==============================] - 340s 340ms/step - loss: 8.9922e-04 - accuracy: 0.3524\n",
      "Epoch 2/2\n",
      "  24/1000 [..............................] - ETA: 5:28 - loss: -0.0137 - accuracy: 0.3603"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "st = ti()\n",
    "\n",
    "for i in range(SetsNeeded-1):\n",
    "    PercentHoldOutForNext = 1.0 - (SamplesPerSet / len(MoveSegments))\n",
    "    seq_train, seq_test, out_train, out_test = train_test_split(MoveSegments, NextDataPoint, test_size=PercentHoldOutForNext, shuffle=True, random_state=0)\n",
    "    seq_train = np.asarray(seq_train)\n",
    "    lstm_autoencoder2.model.fit(seq_train, seq_train, epochs=2, batch_size=32, verbose=1)\n",
    "    MoveSegments = seq_test\n",
    "    NextDataPoint = out_test\n",
    "    print(str(i+1)+' of ' + str(SetsNeeded), (ti()-st)/60, (((ti()-st)/(i+1) * ( SetsNeeded -1) - (ti()-st) )/60/60))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING:absl:Found untraced functions such as lstm_cell_2_layer_call_fn, lstm_cell_2_layer_call_and_return_conditional_losses, lstm_cell_3_layer_call_fn, lstm_cell_3_layer_call_and_return_conditional_losses while saving (showing 4 of 4). These functions will not be directly callable after loading.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: LSTM_sm2_FullDays/assets\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO:tensorflow:Assets written to: LSTM_sm2_FullDays/assets\n"
     ]
    }
   ],
   "source": [
    "\n",
    "lstm_autoencoder2.model.save(\"LSTM_sm2_FullDays\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "MoveSegments = np.asarray(MoveSegments)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "802/802 [==============================] - 101s 124ms/step\n"
     ]
    }
   ],
   "source": [
    "output_X = lstm_autoencoder2.model.predict(MoveSegments)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "reconstruction = np.squeeze(output_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = np.linalg.norm(MoveSegments - output_X, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "output_X[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "maxes = np.max(scores, axis = 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(250,)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.shape(maxes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold_score=0.95\n",
    "anomalous = np.where(maxes > threshold_score)\n",
    "normal = np.where(maxes <= threshold_score)\n",
    "\n",
    "plt.title(\"Anomalies\")\n",
    "plt.scatter(normal, seq_train[normal][:,-1], s=3)\n",
    "#plt.scatter(anomalous, seq_train[anomalous][:,-1], s=5, c='r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([  0,   1,   2,   3,   4,   5,   6,   7,   8,   9,  10,  11,  12,\n",
       "         13,  14,  15,  16,  17,  18,  19,  20,  21,  22,  23,  24,  25,\n",
       "         26,  27,  28,  29,  30,  31,  32,  33,  34,  35,  36,  37,  38,\n",
       "         39,  40,  41,  42,  43,  44,  45,  46,  47,  48,  49,  50,  51,\n",
       "         52,  53,  54,  55,  56,  57,  58,  59,  60,  61,  62,  63,  64,\n",
       "         65,  66,  67,  68,  69,  70,  71,  72,  73,  74,  75,  76,  77,\n",
       "         78,  79,  80,  81,  82,  83,  84,  85,  86,  87,  88,  89,  90,\n",
       "         91,  92,  93,  94,  95,  96,  97,  98,  99, 100, 101, 102, 103,\n",
       "        104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116,\n",
       "        117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129,\n",
       "        130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142,\n",
       "        143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155,\n",
       "        156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168,\n",
       "        169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181,\n",
       "        182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194,\n",
       "        195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207,\n",
       "        208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220,\n",
       "        221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233,\n",
       "        234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246,\n",
       "        247, 248, 249]),)"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "anomalous"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAiMAAAGxCAYAAACwbLZkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAAB8GUlEQVR4nO2deXgUVfr9T3X2hYQkhCXssstmBBFwAxnRoDhuXxdEQUZxw1GRUYIjoI6jP1RkZhzEZZRhXJkRcF9wWBUUUBAVFISwhyVA9pClu35/ZG6lqrp6r053VZ/P8/QD6a7uvlVdde+p8773vZIsyzIIIYQQQiKEI9INIIQQQkhsQzFCCCGEkIhCMUIIIYSQiEIxQgghhJCIQjFCCCGEkIhCMUIIIYSQiEIxQgghhJCIQjFCCCGEkIhCMUIIIYSQiEIxQkgQLFy4EJIkKY/4+Hi0a9cO119/PXbu3Bnp5pnO/PnzsXDhwoi24c0338S8efMMX5MkCbNnz27W9gDu54H+sWrVqmZvEyFWJD7SDSDEyrz22mvo3bs3Tp06ha+++gpPPPEEVq5ciZ9//hlZWVmRbp5pzJ8/H61atcLEiRMj1oY333wTP/74I+677z6319avX48OHTo0f6P+hzgP9Jx++ukRaA0h1oNihJAQ6NevHwYPHgwAGDFiBJxOJ2bNmoVly5bhlltuiXDrIkN9fb3iFjUXQ4cObbbvMkJ9HviLLMs4deoUUlJS3F6rqalBcnIyJEkKuk3V1dVITU0N+v2ENCcM0xBiImJAOnLkiOb5TZs24fLLL0d2djaSk5ORn5+PxYsXu73/4MGDmDx5Mjp27IjExETk5eXhmmuu0Xzevn37MH78eLRu3RpJSUno06cPnn32WbhcLmWbPXv2QJIkPPPMM5g7dy66du2K9PR0DBs2DF9//bXmO3fv3o3rr78eeXl5SEpKQps2bTBq1Chs2bIFANClSxf89NNPWL16tRJ+6NKlCwBg1apVkCQJ//rXv/DAAw+gffv2SEpKwq+//orZs2cbDqYitLFnzx7N82+++SaGDRuG9PR0pKen44wzzsA//vEPAI1C76OPPsLevXs1YRCBUZjmxx9/xG9/+1tkZWUhOTkZZ5xxBv75z39qthHtf+utt/Dwww8jLy8PGRkZ+M1vfoNffvnFre2hIEkSpkyZggULFqBPnz5ISkrCP//5T+V4fP7555g0aRJyc3ORmpqK2tpauFwuzJkzB71790ZSUhJat26Nm2++GQcOHNB89ogRI9CvXz+sWbMGw4cPR2pqKiZNmmRq+wkJJ3RGCDGRoqIiAEDPnj2V51auXIlLLrkEZ599NhYsWIDMzEy8/fbbuO6661BdXa2EPg4ePIizzjoL9fX1mDFjBgYMGIDjx4/js88+w8mTJ9GmTRscO3YMw4cPR11dHR5//HF06dIFH374IaZNm4Zdu3Zh/vz5mvb8/e9/R+/evZVci0ceeQRjxoxBUVERMjMzAQBjxoyB0+nEnDlz0KlTJ5SUlGDdunUoLS0FACxduhTXXHMNMjMzlc9PSkrSfE9hYSGGDRuGBQsWwOFwoHXr1gEdt5kzZ+Lxxx/HVVddhQceeACZmZn48ccfsXfvXgCNYaLJkydj165dWLp0qc/P++WXXzB8+HC0bt0af/3rX5GTk4PXX38dEydOxJEjR/Dggw9qtp8xYwbOOeccvPLKKygvL8dDDz2EsWPHYvv27YiLi/P5fU6nEw0NDZrnJElye++yZcuwdu1azJw5E23btkXr1q2xceNGAMCkSZNw6aWX4l//+heqqqqQkJCAO++8Ey+99BKmTJmCyy67DHv27MEjjzyCVatW4bvvvkOrVq2Uzy4uLsb48ePx4IMP4s9//jMcDt5rEgshE0IC5rXXXpMByF9//bVcX18vV1RUyJ9++qnctm1b+fzzz5fr6+uVbXv37i3n5+drnpNlWb7sssvkdu3ayU6nU5ZlWZ40aZKckJAgb9u2zeP3Tp8+XQYgf/PNN5rn77zzTlmSJPmXX36RZVmWi4qKZABy//795YaGBmW7DRs2yADkt956S5ZlWS4pKZEByPPmzfO6v3379pUvuOACt+dXrlwpA5DPP/98t9dmzZolG3Ux4tgVFRXJsizLu3fvluPi4uQbb7zRaxsuvfRSuXPnzoavAZBnzZql/H399dfLSUlJ8r59+zTbFRQUyKmpqXJpaamm/WPGjNFst3jxYhmAvH79eq9tEvti9IiLi3NrY2ZmpnzixAnDz7j55ps1z2/fvl0GIN91112a57/55hsZgDxjxgzluQsuuEAGIP/3v//12l5CohVKZ0JCYOjQoUhISECLFi1wySWXICsrC++9956SL/Hrr7/i559/xo033ggAaGhoUB5jxoxBcXGxEg745JNPMHLkSPTp08fj961YsQKnn346hgwZonl+4sSJkGUZK1as0Dx/6aWXau7OBwwYAACK45CdnY1u3brh6aefxty5c7F582ZNuMdfrr766oDfI1i+fDmcTifuvvvuoD9Dz4oVKzBq1Ch07NhR8/zEiRNRXV2N9evXa56//PLLNX/rj5MvFi1ahI0bN2oe33zzjdt2F154ocfEZv0xXLlypdJmNUOGDEGfPn3w3//+V/N8VlYWLrzwQr/aS0i0QTFCSAiIQWjFihW4/fbbsX37dtxwww3K6yLXY9q0aUhISNA87rrrLgBASUkJAODYsWM+Z4QcP34c7dq1c3s+Ly9PeV1NTk6O5m8RXqmpqQHQGEr473//i4svvhhz5szBmWeeidzcXPz+979HRUWF38fBqE3+cuzYMQAwdTaM2cfJF3369MHgwYM1j0GDBrlt5+046V8TbfS0H/p9COU3ICTSMGeEkBAQgxAAjBw5Ek6nE6+88gr+85//4JprrlFi+oWFhbjqqqsMP6NXr14AgNzcXLfERD05OTkoLi52e/7QoUMAoMkh8JfOnTsriaI7duzA4sWLMXv2bNTV1WHBggV+fYZRompycjIAoLa2VpNjIsSXIDc3FwBw4MABNycjWMJxnMzA2+wY/WtCIBUXF7sJtUOHDrntQygzbwiJNHRGCDGROXPmICsrCzNnzoTL5UKvXr3Qo0cPfP/99253zuLRokULAEBBQQFWrlzpdRbHqFGjsG3bNnz33Xea5xctWgRJkjBy5MiQ2t+zZ0/88Y9/RP/+/TXfkZSU5LdLIBAzbrZu3ap5/oMPPtD8PXr0aMTFxeGFF17w+nmBtGHUqFFYsWKFIj4EixYtQmpqasSnAvuDCLm8/vrrmuc3btyI7du3Y9SoUZFoFiFhgc4IISaSlZWFwsJCPPjgg3jzzTcxfvx4vPjiiygoKMDFF1+MiRMnon379jhx4gS2b9+O7777Dv/+978BAI899hg++eQTnH/++ZgxYwb69++P0tJSfPrpp5g6dSp69+6N+++/H4sWLcKll16Kxx57DJ07d8ZHH32E+fPn484779TM4vGHrVu3YsqUKfi///s/9OjRA4mJiVixYgW2bt2K6dOnK9v1798fb7/9Nt555x2cdtppSE5ORv/+/b1+9pgxY5CdnY3f/e53eOyxxxAfH4+FCxdi//79mu26dOmCGTNm4PHHH0dNTQ1uuOEGZGZmYtu2bSgpKcGjjz6qtGHJkiV44YUXMGjQIDgcDo+1PWbNmoUPP/wQI0eOxMyZM5GdnY033ngDH330EebMmaPMJDKLH3/80W02DQB069ZNcX4CpVevXpg8eTL+9re/weFwoKCgQJlN07FjR9x///2hNpuQ6CHSGbSEWBExA2Ljxo1ur9XU1MidOnWSe/Toocxk+f777+Vrr71Wbt26tZyQkCC3bdtWvvDCC+UFCxZo3rt//3550qRJctu2beWEhAQ5Ly9Pvvbaa+UjR44o2+zdu1ceN26cnJOTIyckJMi9evWSn376aWVWjiw3zaZ5+umn3doH1cyTI0eOyBMnTpR79+4tp6Wlyenp6fKAAQPk5557TjMLZ8+ePfLo0aPlFi1ayACUWS1iNsq///1vw+O0YcMGefjw4XJaWprcvn17edasWfIrr7yimU0jWLRokXzWWWfJycnJcnp6upyfny+/9tpryusnTpyQr7nmGrlly5ayJEmamTrQzaaRZVn+4Ycf5LFjx8qZmZlyYmKiPHDgQM3neWu/OH767fV4m00DQH755Zc1bbz77rs9fobRueR0OuX/9//+n9yzZ085ISFBbtWqlTx+/Hh5//79mu0uuOACuW/fvl7bSkg0I8myLDez/iGEEEIIUWDOCCGEEEIiCsUIIYQQQiIKxQghhBBCIgrFCCGEEEIiCsUIIYQQQiIKxQghhBBCIoolip65XC4cOnQILVq0YMljQgghxCLIsoyKigrk5eXB4fDsf1hCjBw6dMi0NSsIIYQQ0rzs37/f62KYlhAjYu2O/fv3IyMjI8KtIYQQQog/lJeXo2PHjso47glLiBERmsnIyKAYIYQQQiyGrxQLJrASQgghJKJQjBBCCCEkolCMEEIIISSiUIwQQgghJKJQjBBCCCEkolCMEEIIISSiUIwQQgghJKJQjBBCCCEkolCMEEIIISSiUIwQQgghJKJQjBBCCCEkolCMEEIIISSiUIwQQgixBfv378f/+3//DydPnox0U0iAUIwQQgixBU8//TSmT5+ORYsWRbopJEAoRgghhNiC48ePAwBKS0sj2xASMBQjhBBCbEFNTQ0AoK6uLsItIYFCMUIIIcQWnDp1CgBQX18f4ZaQQKEYIYQQYgvojFgXihFCCCG2gM6IdaEYIYQQYgvojFgXihFCCCG2QIgROiPWg2KEEEKILRBhGjoj1oNihBBCiC2gM2JdKEYIIYTYAuaMWBeKEUIIIbaAs2msC8UIIYQQy9PQ0ICGhgYAdEasCMUIIYQQyyNcEYDOiBWhGCGEEGJ5RL4IQGfEilCMEEIIsTxqMUJnxHpQjBBCCLE8DNNYG4oRQgghlodhGmtDMUIIIcTy0BmxNhQjhBBCLA+dEWtDMUIIIcTyMIHV2gQsRtasWYOxY8ciLy8PkiRh2bJlPt9TW1uLhx9+GJ07d0ZSUhK6deuGV199NZj2EkIIIW6owzR0RqxHfKBvqKqqwsCBA3HLLbfg6quv9us91157LY4cOYJ//OMf6N69O44ePapUyiOEEEJChc6ItQlYjBQUFKCgoMDv7T/99FOsXr0au3fvRnZ2NgCgS5cugX4tIYQQ4hHmjFibsOeMvP/++xg8eDDmzJmD9u3bo2fPnpg2bZrmxNFTW1uL8vJyzYMQQgjxBGfTWJuAnZFA2b17N7788kskJydj6dKlKCkpwV133YUTJ054zBt58skn8eijj4a7aYQQQmyC+gbX5XLB6XQiLi4ugi0igRB2Z8TlckGSJLzxxhsYMmQIxowZg7lz52LhwoUe3ZHCwkKUlZUpj/3794e7mYQQQiyM2hkB6I5YjbA7I+3atUP79u2RmZmpPNenTx/IsowDBw6gR48ebu9JSkpCUlJSuJtGCCHEJuhvbuvq6pCcnByh1pBACbszcs455+DQoUOorKxUntuxYwccDgc6dOgQ7q8nhBASA+jFCJ0RaxGwGKmsrMSWLVuwZcsWAEBRURG2bNmCffv2AWgMsdx8883K9uPGjUNOTg5uueUWbNu2DWvWrMEf/vAHTJo0CSkpKebsBSGEkJhGH6bhjBprEbAY2bRpE/Lz85Gfnw8AmDp1KvLz8zFz5kwAQHFxsSJMACA9PR3Lly9HaWkpBg8ejBtvvBFjx47FX//6V5N2gRBCSKxDZ8TaBJwzMmLECMiy7PH1hQsXuj3Xu3dvLF++PNCvIoQQQvyCzoi14do0hBBCLA+dEWtDMUIIIcTyUIxYG4oRQgghlodhGmtDMUIIIcTy0BmxNhQjhBBCLI9R0TNiHShGCCGEWB6Wg7c2FCOEEEIsj3BGJEkCQGfEalCMEEIIsTzCGcnIyABAZ8RqUIwQQojN2bRpEy655BJ8//33kW5K2BDOiBAjdEasBcUIIYTYnJdeegmfffYZ3n777Ug3JSzIsuwmRuiMWAuKEUIIsTm7d+8G4J7kaRfq6+uVZUrojFgTihFCCLE5RUVFAIDa2toItyQ8qKf10hmxJhQjhBBiY5xOp7KSut3FiCRJSE9PB0BnxGpQjBBCiI05cOAAGhoaANh3gBbhp+TkZCQmJgKgM2I1KEYIIcTGiHwRwP7OiFqM2FV42RWKEUIIsTEiXwSwrxgRzkhKSgoSEhIA0BmxGhQjhBBiY9RixK5ugXBGUlJSGKaxKBQjhBBiY2LBGVGHaYQzYlfhZVcoRgghxMbEghhRh2nojFgTihFCCLEx6gRWu7oF6jANnRFrQjFCCCE2paamBocPH1b+trszwqm91oVihBBCbMqePXs0f9tVjNAZsT4UI4QQYlPWrVsHoLEyKWDfAdoogZXOiLWgGCGExDQnT56MdBPCwuzZs3HbbbcBAIYPHw7Avs6IUQKrXYWXXaEYIYTELIsXL0Z2djYWLFgQ6aaYyu7du/Hoo49ClmXcfvvtmDdvHgD7ihGjMA2dEWtBMUIIiVm+++47AMDmzZsj3BJzOXr0KACgS5cuWLBgAbKysgDY1y1gOXjrQzFCCIlZqqurATTZ/HZB7FdaWhoAICkpCYB9nRGWg7c+FCOEkJilqqoKgP3EiDpsAUAz3VWW5Yi1K1zQGbE+FCOEkJjF7s5IamoqgCZnBLDnIE1nxPpQjBBCYpZYFCN2DNUYLZRnR9FlZyhGCCExixi07TZA68WIGKABew7SrDNifShGCCExS6w4Iw6HA/Hx8QDsJ7wALpRnByhGCCExS6wksAL2nlHDcvDWh2KEEBKzxIozAsDWuRRGs2nojFgLihFCSMwSS2LEzs6I0WwaO4ouO0MxQgiJWShG7AGdEetDMUIIiVliSYzYOUxDZ8T6xEe6AYQQEgmcTqfiEthNjMRyAquYNURnxFpQjBBCYhLhHgCNA7Qsy5AkKYItMg9vzoidxUhycjIcjkbDn86ItWCYhhASk6jFCGCvQdpbzojdBmlZlpXfTl8O3o7r8NgVihFCSEyiFyN2CtXEUgKr+ndTFz2TZRlOpzNSzSIBwjANISQmsbMYMcoZsWuYRv27JScna16rr69XckhIdENnhBASk9hZjMRSmEYIr7i4OCQkJNh+HR67QjFCCIlJRCl4QayIEbs5I+rkVQBKzgjAGTVWgmKEEBKTxJozYtc6I+oaI0DjooBxcXEA7LevdoZihBASk8SaGLG7M6LOj1HPqCHWgGKEEBKT2FWM1NfXK7NIYqHomT5MA4Al4S0IxQghJCaxqxhR71cshmkAsCS8BaEYIYTEJHZNYBVixOFwaGaW2N0ZMZrGTGfEOlCMEEJiErs7I6mpqZry9nYVI+J3U4dp6IxYD4oRQkhMEgtiRI1dwzR0RuwBxQghJCaxqxgxGpwB+zojRgmsdEasB8UIISQmsetCeZ6cEbuKEW8JrHRGrEPAYmTNmjUYO3Ys8vLyIEkSli1b5nX7VatWQZIkt8fPP/8cbJsJISRk7OqMMEzTJLzs8pvGAgGvIFRVVYWBAwfilltuwdVXX+33+3755RdkZGQof+fm5gb61YQQYhpiNk1cXBycTqdtBq5Yc0aMwjRi38VrJPoJWIwUFBSgoKAg4C9q3bo1WrZs6de2tbW1mgumvLw84O8jhBBviEE7KysLJSUlthEjnnJG7L5qr3p/hRjRu18kemm2nJH8/Hy0a9cOo0aNwsqVK71u++STTyIzM1N5dOzYsZlaSQiJFcRAlZOTA8A+lr4vZ8SuYRojZ4RixDqEXYy0a9cOL730Et59910sWbIEvXr1wqhRo7BmzRqP7yksLERZWZny2L9/f7ibSQiJMcRAlZ2dDSB2xAidERKNBBymCZRevXqhV69eyt/Dhg3D/v378cwzz+D88883fE9SUpJy4RBCSDiINTFi1zCNUViKYsR6RGRq79ChQ7Fz585IfDUhhABoSmCNFTHCMA2JZiIiRjZv3ox27dpF4qsJIQSAfZ2RWCt65i1Mw9k01iHgME1lZSV+/fVX5e+ioiJs2bIF2dnZ6NSpEwoLC3Hw4EEsWrQIADBv3jx06dIFffv2RV1dHV5//XW8++67ePfdd83bC0IICZBYS2CNpTojdEasR8BiZNOmTRg5cqTy99SpUwEAEyZMwMKFC1FcXIx9+/Ypr9fV1WHatGk4ePAgUlJS0LdvX3z00UcYM2aMCc0nhJDAcblcivjIysoCYH8xYndnRB2mEcKEYsQ6BCxGRowYAVmWPb6+cOFCzd8PPvggHnzwwYAbRggh4UJt39stTBNrYoTOiD3g2jSEkJhDPUjFijNi9zANE1itDcUIISTmEDNpkpOTlYHLLmKECawUI1aEYoQQEnOIQSotLU25o7aLGPEVpnG5XGhoaGj2doULhmnsAcUIISTmUA/YsSJGRJgGsFeohmEae0AxQgiJOWJRjKirWtspVMMwjT2gGCGExByxKEbi4+MhSRIA+4gRp9OJ+vp6ACx6ZnUoRgghMYdIYNWLEW9lC6yCpwRWSZJsN6NGLSAZprE2FCOEkJjDyBkB7DFIi33TixHAfjNq1M6Hp6JndhCYsQDFCCEk5jCaTQNYP1RTX1+vhC3S0tLcXrfbyr1CjCQkJCAuLk55XjgjsizbZl/tDsUIISTmUDsj6lkmVhcj6rCEkRix28q9Rsmr+r8ZqrEGFCOEkJhDLUYkSVLcEavfRYtcGIfDoRFZAruGafRiJCEhAQkJCQAoRqwCxQghJOZQJ7ACsM2MGnX4ScycUWPXMI061CZgEqu1oBghhMQc+umvdhEjQmQZhWiA2AnTABQjVoNihBASc6gdBMB+YkRfY0QQK2EagGLEalCMEEJiDrs6I3qRpceuYRpvYoSFz6wBxQghJObQixHhGFhdjPhyRsSgbfX9FNAZsQ8UI4SQmCNWnRG7DdBMYLUP8ZFuQMT5352EIXFxgPok97atwwGo1Xkg21ZXA56qBEoSoL7LCWTbmhrA5fLcDnWHFci2p04BTqc526amNrYbAGprAW9LmweybUpK43EGgLo64H+FoELeNjm58bwIdNv6+sbtPZGUBMTHB75tQ0PjsfBEYiLwvymOAW3rdDb+dp5ISGjcPtBtXa7Gc82MbePjG48F0HhNeBt0dNs6y8uRCqCFwwFUVaFlQgJSAdSXljbui7/XfZT1EVVVVUgBkJ2UZPj5Lf93bJUB2uJ9RH1pKVIBtExIaNxf1bYtEhORCqDu5EnjY80+ohH1dR9JZAtQVlYmA5DLysrM//DGS9z4MWaMdtvUVM/bXnCBdttWrTxvO3iwdtvOnT1ve/rp2m1PP93ztp07a7cdPNjztq1aabe94ALP26amarcdM8b7cVNzzTXet62sbNp2wgTv2x492rTtXXd537aoqGnbadO8b/vjj03bzprlfdsNG5q2nTPH+7YrVzZt+/zz3rf98MOmbV97zfu2ixc3bbt4sfdtX3utadsPP/S+7fPPN227cqX3befMadp2wwbv286a1bTtjz9633batKZti4q8b3vXXU3bHj3qfdsJE5q2raz0vu0118gavG0bZX3EvHnz5B+9tPdYeroMQH7iiSca32DjPuKTbt28b8s+ovGh7iPCgL/jN8M0hBBiE3yFJETtkVgIXcTH0/i3EpIsy3KkG+GL8vJyZGZmoqysDBkZGeZ+uIUs2IC3tbgFG9K2tGAbYZjGcNvenTph/4EDWLN6NQYNGoTJkyfjjTffxGOPPooHHnzQsmGaP/7xj5j7xBO48/bb8eyzz7pt+sjMmfjT3Lm4//77MXfuXMv3ETNnzsSzc+fi7rvuwpw5czTb3jN5Ml59+WU8PGMGZsyY4f657CMaCXOYxt/xm9LRQ6JXs27rIfM95G0NMsxN2dYgWcyUbZOSmgYMM7dNTGwa4CK1bUKC/xd8INvGxzd1OmZuGxfn/zkcyLYOR3i2laSAtj1+6hSqASTn5ABpaUjNzUU1gJKaGvdz1kJ9RHV1NWoAJLRsafj5CZmZynYALN9HVLhcqAYQl5Hhtr+JLVqgGkBZQ4PvY80+IuIwTEMIiTn0U2Az/zdIl5WVRaxNZuCrAqvdZphwaq99oBghhMQULpdLGcTEoG03MeKpzojdBmgWPbMPFCOEkJhCXUvEbs6Iv3VG7DJA0xmxDxQjhJCYQj04iUHMLmIkVp0RFj2zPhQjhJCYQgxOSUlJiPvfTAa7iBFfzogQX3YZoL05I3bbV7tDMUIIiSn0peAB+4gRJrA2Ybd9tTsUI4SQmMIolCHESGlpaSSaZBqxGqahGLE+FCOEkJjCKJQhxEhFRQVc3gp7RTmxulAexYj1oRghhMQU3sI0siyjoqIiIu0yg1hzRsTMKIoR60MxQgiJKYzESHJyMhL+V9HSynkjdEaasNu+2h2KEUJITGEkRiRJQsuWLQFYV4w4nU7U/m8NEl9ipKGhAfXe1kyxCCx6Zh8oRgghMYWnUIbVZ9RUqRbe8xWmAezhGPjrjFhgPdiYh2KEEBJTeAplWF2MiP2SJMmwCBgAJCYmwvG/1WetLkacTqfi7ngregZoq+6S6IRihBASUxiFaQDrixG14yNJkuE2kiTZJpdCHX7xVvQM0LpGJDqhGCGExBR2FSO+klcFsSJG4uLikJ2dDQDYv39/s7WLBAfFCCEkprCrGPFVfVVgNzGiDj3pGTBgAABg69atzdYuEhwUI4SQmMLuYsRT8qrAbmLEyBURUIxYB4oRQkhM4clBsHpJ+FgL03greCagGLEOFCOEkJiCzog9xAidEXtBMUIIiSnsKkZizRnxR4z07dsXDocDR48exeHDh5uraSQIKEYIITGFXcWIvwmsYvC2emVSf8RIamoqevToAYDuSLRDMUIIiSk8iRGrl4OP1TCNpwJvAoZqrAHFCCEkpvCVwGpVMcIwjTEUI9aAYoQQElPYPUwTa86Iv2Lk+++/D3ubSPBQjBBCYgpfYqSiogIul6vZ2xUqdEaM6dWrFwCgqKgo7G0iwUMxQgiJKXyJEVmWUVFR0eztCpVYrcDqS4y0a9cOQKPI5Bo10QvFCCEkZpBl2aMYSU5ORmJiIgBrhmpiLUzjT9EzAGjRooWyzZEjR8LeLhIcFCOEkJjh1KlTkGUZgPGgbeUqrAzTGCNJEtq0aQOAYiSaoRghhMQM6gHYSIy0atUKAFBSUtJsbTKLWHNG/BUjANC2bVsAYOGzKIZihBASM4gBOCEhAfHx8W6vi0GruLi4WdtlBp7CT3ooRkg0ErAYWbNmDcaOHYu8vDxIkoRly5b5/d6vvvoK8fHxOOOMMwL9WkIICRnhHqSnpxu+LpIdrThoicE51sSIr6JnABimsQABi5GqqioMHDgQzz//fEDvKysrw80334xRo0YF+pWEEGIKlZWVADyLESs7IyKh09fgbDcxQmfEHrj7lD4oKChAQUFBwF90++23Y9y4cYiLiwvITSGEELPwJUas7IxQjHiGzkj00yw5I6+99hp27dqFWbNm+bV9bW0tysvLNQ9CCAkVIUY8zTixsjPi7+Aci2KEzkj0E3YxsnPnTkyfPh1vvPGGYcKYEU8++SQyMzOVR8eOHcPcSkJILGDnnBE6I56hGIl+wipGnE4nxo0bh0cffRQ9e/b0+32FhYUoKytTHvv37w9jKwkhsYJdc0ZkWfZbjIjBu7q6Wqm5YkX8LXoGaMM0Vt5nOxNwzkggVFRUYNOmTdi8eTOmTJkCAHC5XJBlGfHx8fj8889x4YUXur0vKSkJSUlJ4WwaISQG8Tdn5OTJk6itrbVMP1RfX68Msv46I7Iso7a21q/ZKNFIMDkjNTU1qKioQEZGRljbRgInrGIkIyMDP/zwg+a5+fPnY8WKFfjPf/6Drl27hvPrCSFEgy8xkpWVhYSEBNTX1+PIkSPo1KlTczYvaMTADPifMwI0uiOxIEbS0tLQokULVFRU4PDhwxQjUUjAYqSyshK//vqr8ndRURG2bNmC7OxsdOrUCYWFhTh48CAWLVoEh8OBfv36ad7funVrJCcnuz1PCCHhxpcYkSQJbdu2xf79+1FcXGwZMSJCFgCU9XU8IQq+NTQ0oLq6GtnZ2eFuXlgIRIwAje5IRUUFjhw5ElDaAGkeAs4Z2bRpE/Lz85Gfnw8AmDp1KvLz8zFz5kwAjbHWffv2mdtKQggxAV9iBLBmEqs6X0SSJJ/bi9lEVl7FNpCiZwCTWKOdgJ2RESNGeE0AWrhwodf3z549G7Nnzw70awkhJGTE4OttMTkrJrEGkswJNIajysrKcOLEiXA2K6wE6oxQjEQ3XJuGEBIz2NUZCdQlyMnJAQAcP348bG0KJ/X19XA6nQACC9MALHwWrVCMEEJiBn/EiJWdEX/FiMgTsaozEkjCroDOSHRDMUIIiRns6owEKkas7oyoxYi/+0xnJLqhGCGExAx2dUYCzZ+wuhgJNGEXaJrSrBYyJHqgGCGExAy+ysEDdEasQKDiC2ic0gwADQ0NYWkTCQ2KEUJIzOBroTxAm1vgcrmapV2hQjHiGyFG6uvrw9ImEhoUI4SQmMHfMI0kSaivr0dJSUlzNS0kKEZ8IxZqpTMSnVCMEEJiAlmW/RIjiYmJSrLjgQMHmqVtoRJrOSOBTmUG6IxEOxQjhJCY4NSpU0rYxZsYAYAOHToAgGVWDKcz4hs6I9ENxQghJCYQrgigXSzOiI4dOwKwjjNCMeIbOiPRDcUIISQmEDNpUlNTERcX53Vb4YxYTYwEGqapqamx5FTXUJwRipHohGKEEBIT+DOTRmC1ME2gORQZGRnK4GzFKqyc2ms/KEYIITGBP8mrAqs6I/6KEUmSlJLwVgzVBOoEAQzTRDsUI4SQmCAQMWL3nBEAlhYjTGC1HxQjhJCYIFhnRJblsLbLDIIZnK2cxBrrCaz19fX46aefLHFu+gvFCCEkJvCnFLwgLy8PAFBbW2uJwmfBOCOxJkbs5IzMmDED/fr1wwcffBDpppgGxQghJCYIxBlJSkqyVOGzWBUjsVr07OeffwYA7N69O8ItMQ+KEUJITBDIbBrAWjNqYlWMxGqYpry8HABQV1cX4ZaYB8UIISQmCMQZAaw1o4Y5I74RYRqn02n5XAuKEUIIsSiBihErzaihM+Ib4YwA1s8boRghhBCLEqwzYvcwTawUPRPOCGAfMVJbWxvhlpgHxQghJCYIZDYNwDBNNBNK0TPA+nkjFRUVAOiMEEKI5QjUGWnfvj0A4NChQ2Frk1kwTOMbuzgjtbW1iiNCMUIIIRYjUDHSqlUrANYYrEOpwHry5EnLJXQGI0bUiyNa2RkRrghAMUIIIZYj0Km9QoycOHECTqczbO0yg2DESGZmJoDG2SUihGUVghEjkiTZYnqvyBcBKEYIIcRyBOqMCOdAlmWUlpaGq1khI8tyUINzamqqErqI5v0zIpj9BexRhZVihBBCLEygCayJiYnIyMgAgKguCa8ekAJxRiRJQsuWLQEAZWVlZjcrrARTgRWwR+EzhmkIIcTCBOqMANZI8hQhGiDwwVmEauiMWAc6I4QQYmHEHWUgYkTkjUSzMyLEiCRJSExMDOi9VnRGgg1LAfZwRihGCCHEotTX1ysDmHAD/MEKYkQdspAkKaD3WtEZqaurU2b/BOuM2EWMsOgZIYRYCHUH3qJFC7/fZ6UwTaAhGqDJGbGSGFGHpYJ1RhimiT4oRgghtkeEIdLS0jTFr3xhBWfEDDFipTCNcIKCCUvZIUzDBFZCCLEo4m5SzI7xFyuIkWDzJwBrhmnU+xtoWIoJrNELxQghxPaIO/9A8kUAhmmikVDElx2cEYoRQgixKHZ2RkIRI0KcWTFME4wYoTMSvVCMEEJsjxhsgxUjdEaih2ALngF0RqIZihFCiO0RHXiwYZpodkZCcQqsnMAaijNiZTHCBFZCCLEooYZpTpw4AZfLZXq7zMCMMI0VnZFQckYYpok+KEYIIbYn1ARWl8sVtQN2rIZpmMDKomeEEGIpgnVGEhMTlSJp0RqqibU6I2J/mcBKZ4QQQixFsM4IEP1JrGbUGampqbHMwBbLzogsy5qcEZfLBafTGcEWmQfFCCHE9gTrjADRn8QaijOiPh5WcUdieWpvVVWVsi6PwCoi0hcUI4QQ2xPs1F4g+muNhCJG4uLilDCUVfJGYtkZUYdoBBQjhBBiEYKd2gvYO0wDWC+JlWKk6TcDKEYIIcQyhOKM2DlMA1gviTWUomdWD9Oow41CWPkSIy6XCydOnAh720KFYoQQYnvMcEbsKkasVmuEzkijGBErFvsSI5MmTUKbNm2wc+fOsLcvFChGCCG2RpblkBJY09PTAQDV1dWmtssszHJGYkGM2MkZ8VeMbNmyBQ0NDfj555/D3r5QoBghhNiaU6dOKXfCwYgRfzv9SBFqzojVFsuLZWdETOsNRIyI16P1/BVQjBBCbI24m5QkSXE5AiEpKQlA9Fa7jDVnJJaLnhk5I77OS4oRQgiJAsQdf4sWLeBwBN7lRbszEqsJrNHmjBw6dAirV682/XPVCDHSokULOiOEEGIlQkleBaLfGQlldgnABFazuOmmmzBixAj88MMPpn+2QO2MiPOSYoQQQixAKNN6Aes4I6wz4ptwhmkOHjwIACguLjb9swXBJLAKER3teTIUI4QQWxOqM+JvbD5SmDW1l2Ga0BCiIJz5KExgVbFmzRqMHTsWeXl5kCQJy5Yt87r9l19+iXPOOQc5OTlISUlB79698dxzzwXbXkIICYhQnRF/7fBIEaoYSU1NBdA0yEc7oexvOJ0RIXDCKUaCcUasIkbiA31DVVUVBg4ciFtuuQVXX321z+3T0tIwZcoUDBgwAGlpafjyyy9x++23Iy0tDZMnTw6q0YQQ4i+h1BgBoj9ME+rU3mjfPz2hiJHmcEbCGQ4JNIHV5XIp4ijaf9+AxUhBQQEKCgr83j4/Px/5+fnK3126dMGSJUuwdu1aj2KktrZWY4kaLQ5ECCH+YPcE1lCdkWjfPz10Rvx3RtTCKNrFSLPnjGzevBnr1q3DBRdc4HGbJ598EpmZmcqjY8eOzdhCQoidsHMCqyzLIYuRaN4/I6LdGYkmMaIWmExg/R8dOnRAUlISBg8ejLvvvhu33nqrx20LCwtRVlamPPbv399czSSE2Aw7OyPqgYjOiG/CKUaawxkxSmD19rupz49oF5sBh2mCZe3ataisrMTXX3+N6dOno3v37rjhhhsMt01KSlIuEEIICQU7OyPqpNNYyBlxOp3KYB9NYRpZlqMyTEMxYkDXrl0BAP3798eRI0cwe/Zsj2KEEELMws5Te4VLIEmSctcfKFZyRtRtDOaGNVzOiNPphCzLAMInRurq6pTf298EViuJkYjUGZFl2RInPiHE+lRWVgJAUOvSAE2DXn19vTLgRAvqkIUkSUF9hpWcEbG/QHQ5I2pxE67cDBGiARrFiD9Tzq0kRgJ2RiorK/Hrr78qfxcVFWHLli3Izs5Gp06dUFhYiIMHD2LRokUAgL///e/o1KkTevfuDaCx7sgzzzyDe+65x6RdIIQQz4RaoVQM1kDjQKP+O9KEOq0X0DojsiwHLWqaA/FbxsXFKcIiEMLljKgH+nA5I8LhS0lJQUJCgu0SWAP+NTdt2oSRI0cqf0+dOhUAMGHCBCxcuBDFxcXYt2+f8rrL5UJhYSGKiooQHx+Pbt264amnnsLtt99uQvMJIcQ7Zk19BRo792gSI6HuG9AktmRZhtPpDGqQby7E4Brs/op9M3tgVn9euMSIOnkV8M/RsrUzMmLECK9W5cKFCzV/33PPPXRBCCERw6ypr0D0dehmiBG92IpmMSL2N9gJDsIZMVswNKczYlcxwrVpCCG2JtQBOy4uDnFxcQCiL8nTTGcEiP4BK9T9DVeYpjmcEXX1VYBihBBCLIWZA3a0dehm5IzEx8creSLRJrb0hPpbNkcCK52R4KAYIYTYGjFg21GMmCG0JEmK2v3TE2rOSHMksIYrUdSTGPEmIK2UwEoxQgixNWbmVUSbc2DGvgHRu396Qs0ZsbIzYvcEVooRQohtaWhoCKlipyBanQOzxEi07p+eaM0Zac4EVuaMEEKIxVDf6dvRGTEjZwSI3v3TE61ihDkjoUMxQgixLeqKnaGsdxWtzkGsOSNm1Rmxw9Reu1VgpRghhNgWMVjHx8eHVD8jWp0D5owEBp2R6IVihBBiW+zuHJgVponW/dNjham94V6bJhAxwtk0hBASBdhdjMSqMxJtOSNMYA0dihFCiG2x+2Btd7Glx6ycEafTaeoKzAzThA7FCCHEtth9sLa72NJjVs4IYK5o4No0oUMxQgixLXYfrJkzEhhqMWJmqCaSzoi3c5JihBBCogA6I/4RrWJLj1kJrIC1nBFZlkNOYHU6nXC5XKa3zSwoRgghtsVsMRJtg7XdxZYeKzgj4Zi1Ul1drQiJYBNYw9U2s6AYIYTYFrOdg2gbrM1YBBCwjjMi2hdszojD0TTkmelghDtMI0I0kiQhLS0NQHBiJNrOXzUUI4QQ22J350DsH3NG/EOSpLBM7w13mEadLyJJEoDAK7D62jbSUIwQQmyL3XMq7L5/eszY33AUPmsuZ0TkiwB0RgghxDLEijNi1/3TY8b+WtEZ0SevAk2/mcvlgtPpNHyfXlwyZ4QQQiKA3Z2DWFu1N9ScESA8YiTczkhZWRkAYzECeBaRdEYIISQKsLtzYPf90xOtYRr1cQuH+1BaWgoAyMzMVJ6jGCGEEItg9wRPuzs/eqI1TNNczkjLli2V59TTlD39bhQjhBASBdh9sKYzEjhWTGAVYkTtjKhnBtEZISQMVFZWRnWlQGId7DxYy7JsmvMTrWJLT6gL5QHWTGA1CtMAvs9LJrASEiTFxcVo27Ytxo0bF+mmEBtgZ2dE3RY7ii0jQl0oD7C2M6IO0wC+fzc6I4QEybZt21BVVYUNGzZEuinEBtjZGRH7BthTbBkRrTkj0eqMUIwQEiSisxFTFu3ODz/8gOHDh2PFihWRbootsXO5dLFv6tyBYIlGsWVEtIqRcK9N48kZ8VWFVTwvqrZG8+9LMUKiilgTI0uXLsX69evxxhtvRLoptiQWnJGUlBRlsAmWaBRbepxOp+I6RFsCa7idEaMEVsD3Ao6iXWI9m2g6f/VQjJCoItbESHV1teZfYi6xIEZC3TcgOvdPj3rAjbWiZyJMo3dGROKyOmSnRhwzsdIvE1gJ8RNxUdXV1XkscWwnxP5SjIQHOyewmilGonH/9JiVI2PlBFa9M5KamgrAc/9BZ4SQIFF3OLHgjoh9jIV9jQR2dkbMKgUPROf+6RG/ZVxcnCIogsFqCayyLPsUI1VVVV7blZ6e7tbOaINihEQV6juzWBig6YyEFzoj/hGN+6fHjBojQJMzYpUwTU1NjfL5+jCNN2dElmWKEUKChc4IMRM7OyOxljNiRo0RoMkZCVcCq8vlMrVoo3BFHA6HIioE3sSIev8oRggJkFgTI3RGwovZzkg0deax5oyYtb/hTmAFzBU66hoj+llT3sSI+lxlAishARJrYkTsI8VIeDDbGYmmwTpWc0bMCtOEyxkx+7M95YsA3sWI+lxlAishAaIWI7EwQDNME14YpvEPOiOhEU5nJFgxoi54JgRrNJ2/eihGSFQRa84IwzTho6GhQZkeblaYpr6+PmoWcTSruiwQnWJLjxBKoeaMhHtqr9mf7anGCOCfGElMTIzKMKMeihESVcSaGFGHaWRZjnBr7IWZa7eIwRqInrh7eXk5ACAjIyPkzxKDVUNDQ9SILT3R7IxEY5hGLUasIDYpRkhUEWtiROyvehoeMQf1uRTq3bT6/dESyjh58iQAICsrK+TPUoutaD0Po1mM6D/LzM8O1RlJSkoKyz6bDcUIiSpiTYyo95GhGnMR51JCQgLi4uJC+iz1QnSRHKwff/xxPPHEEwC8D1KBEo1iS4/ZCax2cEZEYqq3BFarOCPBl7EjJAzEmhjR768Zd7mkETMTPOPi4hAXFwen0xmxDr2qqgozZ84EANxzzz2mOiPRIra8YVbOSE5ODgDg6NGjIbdJEM0JrFYRI3RGSFQRq7NpgNjY3+bETDECRH7Gifr8KCkpMdUZcTgciiCxuzNy2mmnAQCKiopCbpMgnM6IGQmsFCOEBEgsOSOyLMec+GpOzBYjke7Q1SLhxIkTpooRIPL75wuzfs+uXbsCAHbv3h1ymwDA6XQqyecOR+OQ2tzOiNHaNBQjhIRALIiRDRs24KeffnK7A7Xr/kYKuzkj6mvj+PHjpoZpgMjvny/Mdkb27dtnSt6I+jNEPY9wiBEmsBLSjNhdjOzbtw/nnXceLrroIrf9ozNiLnZ2Ro4fPx5zzohZOSNt27ZFcnIyXC4X9u/fH3K71MdLiANPg/4zzzyDnj174uDBg35/vrocvB5/KrDSGSEkCOwuRpYtW4a6ujoUFxcrdzyCcIiRn3/+GWvXrjX9c71RUlKCffv2Net3GiHOJTPKpQORdw7U36vOGTHbGdm+fTv++te/RuT6++STTzB37lzDmjtmiUtJkkwN1QTijLzzzjvYuXMnVq9e7ffnx0oCK2fTkKjC7mLkvffeU/5/7NgxzWvh2N9LL70Ue/fuxcGDB9GmTRvTP9+I4cOH49ChQyguLlYW6IoEdnNG1NfG3r17leJkZjsj9913H44cOYJWrVph3Lhxpny2v0yePBkHDhzARRddhP79+2teM/P37Nq1K7Zv325KEqs4HxwOh6Z4nBEVFRUA3K99bzCBlZAIYOeEzhMnTmjuiPQdUjj298CBA3A6nQHZwqHgcrmwc+dOVFVVobi4uFm+0xN2EyNqZ2TXrl1Km8zOiTly5AiAwAZMsygpKQFg7FiY+XuKvBEznZGEhAQlN8MsMeJyuZT3eHNG6urq3L6TYoSQELCzM/Lxxx8ra6UA4XdG1DUx9CGhcKHeB6MM/0D49ttv8dprrwVdJt9uCaxGYiQrK8ttWflgUVdhBZr/ZqChoUH5zYzCfGbljABNM2rMdEYSExN9rnsTqBipqKhQzn9vYgRw7z+MElgpRgjxA6fTqYm/1tTUYNu2bbj33ntx+PDhCLbMHNQhGsC96JLZnb9a2Il1TMKNWoCo/79nzx5MmTJFGUT94Xe/+x0mTZqErVu3BtUWuzkj6t9THEezQjSA+yAfqpgMFDFQAzBMLBXtUQ/AwRIuZ8SbGJFlGZWVlQCaHCBfiBlTSUlJhuex+jl9/2GUwMrZNIT4gdFU12effRZ//etf8frrr0eoVeYhEklFafJwh2nUd0rN5YyoBzD1/ixYsAB///vfsWDBAr8/S4i1YIWomavaAtHljIh9M7Nib6SdEbUY2bdvH2RZxr///W9FMIjXzchDMtMZEQO82hkxGvSrqqoUl8NfZ0SIEVE1Vo8kSR7zRhimISRI1Hd+QGOHKwaiEydOmP59LpcLl1xyCa677rpmWTFXCIJ27doBCH+YRt05hUOM7NmzB+effz6WLVumPOfJGQnmdxTtF3eTgWIlZ0SWZdx444144IEHPG5jJILs5Iyof+d9+/Zh9erVuPbaa3HrrbcCCI8YKSkp0YigYBDngy9nRO1O+itGxPXiTXR6Wp/G9mJkzZo1GDt2LPLy8iBJkqYjMmLJkiW46KKLkJubi4yMDAwbNgyfffZZsO2NGoqLi4PuJIkxRmJEXIyiw7jttttw+eWXm2I37tmzB5999hkWL14c9t9SHQ8XYiTcYZpwOyNLlizB2rVr8eqrryrPeRIj+t/RH0T7o0WMiME6HB36oUOH8Oabb+K5555TZsno0V8fgLliJNqckQ0bNgCAknwtXs/IyAj5uzIyMhS3IVR3xN8wjXr//A3TiOsmOzvb4zb+OCO2zBmpqqrCwIED8fzzz/u1/Zo1a3DRRRfh448/xrfffouRI0di7Nix2Lx5c8CNjRZOnDiB0047DSNHjox0U2yFvrOtrq7WDGJ1dXV45ZVX8MEHH+C///1vyN934MAB5f9iBkG4UA/Mbdu2BWD9MM2ePXsAeBYg6v05fvw4AP/FSENDg9JxBnvnKs4nMxIegabBOhxhGvFbqfMK9Bh9r5lhGv1xiqQYKS4uVsYIce4KZ8Gs6eJmrVFjNOj7EiPHjx/XJLN7wgwxkpSUZAlnJOA6IwUFBSgoKPB7+3nz5mn+/vOf/4z33nsPH3zwAfLz8wP9+qhgz549OHXqFLZv3x7pptgKI2dEXMAVFRUam/Odd97BJZdcEtL3qcXI4cOH0b1795A+zxtigImLi0OrVq0AhD9ME24xsnfvXgCexYj6/4GKEXXbg3VGxPvS09ODer8e4bCEY5aXPtnY6O4/3GEavTMSyQRWl8uFL774AkCTCDEzTANAuQ5DDQEH44zIsowTJ04gNzfX62eLnJFgxAgTWH0g5k17O7i1tbUoLy/XPKIJ0clVV1c3S65BrGDkjIiLUS9Gli5dGrLKV2fsh3u2jnpgFJ1pczoj4biGhBhRt9uXGPG3HerPDNYZ8Va5MhjE7xaOkJ5aaHg6RuEO00TaGdEfVxHKqKmpQV1dnfK6WWJEfE6w59e+fftw9dVXY9WqVQB8T+3Vf48/eSP+5Ix4WizP9jkjofLss8+iqqoK1157rcdtnnzySWRmZiqPjh07NkvbHnnkETz00EM+txM/un7VVRIaelu9oqJCiZ/rxUhZWRk+//zzkL6vOcM0RmJEnEeiToTVwjS+nBGxPy6XK+CcEfWxCHbwN1uMCIclHGJE3Y+oj9HBgwfxyiuv4NSpU2EP00STM6Ln8OHDyo2f2WIk2N/z5ZdfxpIlS/DMM88A0DojRg6Efv/8yRsxK2dE/LbV1dUYP3485s6d6/O7m5tmFSNvvfUWZs+ejXfeeQetW7f2uF1hYSHKysqUhxmLGfmiqqoKf/rTnzBnzhzlbtzbtgK7VQmNJKJDNupg9WIEaAzVhEKknBF92EDsr5XCNOXl5UqZal/OSHl5uUZU+oMZzog4X8xIeATCK0Y8OSMzZ87EbbfdhsWLFzf7bJpI5ozoEUmsDofDlDojQOjOyLfffgtAGw4JlzMSjBhR59iIXBYAeOONN7zO2ooUzSZG3nnnHfzud7/D4sWL8Zvf/MbrtklJScjIyNA8wo06big6WU+oO6PmvnuwM/6KERG7//DDD/1KAvOEPmcknIjzRO2MCERHYyVnRLgigO+cERGiAZo3Z8SqzohajIiS+keOHDF0YcPhjIg6ONHkjIhrtUWLFqZVnA1FjMiyrIgRQSA5I4B/YkTcGPsTptH3H2oho3e9opFmESNvvfUWJk6ciDfffBOXXnppc3xlwKgFiK+Om85IeBCdrdHdnlqMnH322WjRogVKS0uDrs4JaJ2RSIRpBEKMWMkZ0YsRYaEbXRtqMVJXV+fXbBQznREriBH1MVHvrzgO1dXVzeaMiETuSDkjRi6AWoyYRShi5NChQ25T8wOZTQOYH6Y5fvw4Lr30UmXSiFrIGImRaMsfCViMVFZWYsuWLdiyZQuAxmlRW7ZsUdYSKCwsxM0336xs/9Zbb+Hmm2/Gs88+i6FDh+Lw4cM4fPhws1WE9Bd1aCYQMUJnxDyEGElNTXWzjNViJDs7G+eddx4ABLQUt5ra2lpNZ9JcYZq0tDRbOCNiWi/QmBMiOjZfzgjgX+dvZs6IFcI0npwRcQybQ4z06dMHADB69GjNdzcX4rj27dvX7TURpokWMaJ3RYDgnJEdO3Z4XVAyEDHyySef4OOPP8Zzzz2nea8nMRJtY1fAYmTTpk3Iz89XpuVOnToV+fn5mDlzJoBGW1G9yNGLL76IhoYG3H333WjXrp3yuPfee03aheCpqqpCUVERTp48GZAzou6M6IyYh7pIVUpKiua1hoYGxdbMyMjABRdcAABKJnug6FexjWTOSHOIkcrKypBCWnrUzgjQ1LEZiRH91MlAxUgwg4Wv1U6DIRLOiFqMiOtDPTCZGaYZP3489u7diz/+8Y8AGq9HTwXYBGvWrMETTzzhczt/EPutFiOnn346gOgTI999953bc/6KEVFs7bvvvsOAAQPQo0cPvPrqq4YzMwMRI6LUhHBc1NOCRehNTbQV7QxYjIwYMQKyLLs9Fi5cCABYuHChZoBYtWqV1+0jybhx43Daaadh8eLFdEYCYOHChRg9erTP3Bqg0V71t8CdNzECNHVIGRkZGDFiBIDGzjCYjlDYvuJ7jhw54tc07V27dmHw4MFYvHhxQN/nLUwjOqdwhmmAwDpdWZaxdetWj3UJ9GJEiAdfYRp/2xGqM1JZWel1tdNgCESM7N+/H++9957fU/8DcUbat2+vvG7Wvgk6deqklBcX3+uN3//+9/jjH/+INWvWhPzd4rzIz89Hz549MXz4cPTo0QNAeMSI+D3NckZ8rU0jvkcUW/vyyy9RW1uLqqoq/O53v8PTTz+t2b62tlY5/v7kjIhzqLq6GlVVVUr/LFZ2nj17Nm677bawTlEPhZhem0YUnDl27JjpOSO1tbUYPnw4fvvb39quFsnzzz+P5cuX+1XWf/To0TjrrLM8Og9qIRGIGDnzzDORnp6OkydP4ocffgh4H4QYOeOMMwA0xk/9EVeikvCiRYsC+j5/ckbC6YwAgYVqFi9ejIEDB2L27NmGrwfijOjFiD+1RkIVI2JfExISTKvA6mvw+vnnn5VzeNKkSbjiiiv8HqQ9iRGjnBGxnECLFi2Uwc9M1Neer3NSXNfqsF2wqHNGfvrpJ3z55ZeK2BLXq5mTGcwI0/Ts2VN5zl9nRIgRQb9+/QA0Lq+gRtwgS5LkVXSqxaNg165dyrgjhMysWbPw0ksvKaG9aLuRjmkxIirwlZSUaJwRX52lP7NpNm3ahPXr1+P999/Hxo0bTWht9CA6evVsFCOqqqqwfft2OJ1Ot8ELAJ577jmkpaVh3bp1ALRixGj6nlqMxMfH49xzzwUQXKhGJK92795duTj9CdWIfQ/0QlbPpvEUpqmtrVVCKR999BE+/vjjgL5Dj34gCUSMbNu2TfOvHk/OiNE031CdEV/bG4l99Uwas2ZfeHNGPv30U/Tp0wcPPfSQZqaFv2UJAgnT5OXlATA3X0SNw+FQBIm381yWZaXfPHToUMjfqy5qFh8fD0mSFPERTWGa4uJiFBcXw+Fw4PLLL1ee93dqr16MTJ8+HYD7LE51zofD4XmoNuord+7cqbymF+PhDDeGQkyLkXA6I2oB8vrrr4fQyuhDXFS+xMiuXbuU/xsd008++QSnTp3Ciy++CMC3MyI6PNFBiVDNO++8E3CoRrS9Y8eOaNOmDQD/ZtQIoRroheyPMwI0HoPy8nJceeWVuOqqq0JaByUUZ0QMMkalsk+dOqUcKxFiMnJGxLURTM6Iuu3V1dUe810eeughdOrUyW0wNHsmDaAtkqUXQEI4Ll++HEePHlWOn78VZ42cEZfLpRwHtTNyzjnnIC8vD5dddlkIe+MdTyvBqqmurlYSl80QI0bl3sXvJ74nGsSIEOg9evRA7969lecTEhL8mk2jFiNDhgxB//79AcCtvpU/1VcB72LE6L0UI1GIWowEkjPijzOiFiNvv/12VK8JECj+ipFff/1V+b9RCEQkpH788cdwOp0+xYiY/SLEyLhx45Camor169drVo71B3HH2qFDB2XhukCckVDEiKeiZ0BjB79jxw7U19ejtrY2pHUz9GIkkJLw4nowKgAoEtTT0tLQoUMHAIGFaQJ1RvSfK3A6nViwYAEOHDjg5j6aPZMGaOrEnU6nm0gU379jxw5lpiHg/zE3KnqmF2Rim7y8POzfvx/z588PfCf8xFOJcTXqc0OfEO4P+hsIIzGi//3CJUYCCaWLfigvLw+dOnVSnvfXGenatavy3JVXXqk4XCdPntS0w591aQCKEVugDtOE0xk5duwYXnzxRSxbtizqToBAcTqdyj776oDUYsTomIqLuqSkBF9//bVHMSIucHGhig6qY8eOePzxxwEAf/jDHwKaEWPkjPjz/lCdkbS0NE15ZvGcsFJramqwY8cO5TV/8lg8EYozIr7XSAyp7wz1d9B6MSLLsiJGROdfXl6OiooKr+eP/royOt5btmxxW0RNYHbBM0Abm1e3p76+XhEgTqcTS5cuVV4LxhkR+6LvZ9TXhzfb3gz8cUbU50agzsiOHTvQqlUrPPbYY8pzYr/VYr05xIj6RsgfRL+Vm5urESP+5oy0bt0a7du3R1xcHK6++mpFMNTX12uuWX9m0gDexYjRe43ESFlZmWGbm5OYFiPBOiO+ZtOcPHlSORnGjx8PALjnnntw5ZVX4tFHHw253ZFEfQKHEqaRZVlTgfCDDz7wKEb0axOpO6jf//73GDRoEEpLS/G3v/3N7/0Qc/vz8vIUZ8SfMI0Zzgig7VTVOTLCGRFESox4C9OIaY1nnnmmMmgZOSNOpxN1dXWKGBF3hBUVFRg1ahS6devmVjhKoB8EjdwUda6Q/vVwhGni4uKU81L9+2/btk0zmKkTEUNxRvT9jNjGrIRcb4TbGVm9ejVOnjyJjz76CEDj4C2OoTdnJBxOFxBYqEb0W7m5uZq+yV9npEWLFvjkk0+wevVq9OjRA+np6crUW/UxNUOM+OuMTJgwAUlJSfjnP//p9bvCCcUIQgvTGN05bNq0CUBjbHDGjBlITU1VTtKVK1eG3O5Ior5oDx065LV2hbcwTVlZmSZ05U2MdOnSRfNedYcUHx+Pu+++G0BgBdDUg1WwYZpArF1vYiQlJSWsYkR8ZzBipKamxu2u0UiMGDkj4nnRqYrfsbS0FN999x1qa2vx448/Gn6/P86I+vf25IyYvZSEUUeuDxGpRXYoOSN6Z6Q5xUigzsjhw4cDqmMjbgaEUFUfz+YK0zgcDmU/fYmRtWvXYv78+ZqbqNzcXKSmpip5U94WytOLrf79++Occ84B0DhbRogGIzESTM6I6Mv8FSP79u2Dy+VS9iUSxLQYEWGa2tpazV1+qM6I6JzOOuss9OnTBydPnsQvv/wCANi6daulV/pVX7ROp9Orm+AtTCMu6OTkZMTHx2Pbtm2K/a8XI2orFHDvoMSsmo0bN/p1bF0ul3IhZmRkBBWmaWhoCKicsno2jfpfoFGMiP3Vh2l8LdroDSFGhNgKJkxj1Aa1GFHfQcuy7HY9lJWVKcdMiJEdO3YoA5cnd82XM+J0OjXTZvWDfjjCNIBxRy5uPsTNjZpQwjTqY6AP04SbQJ0Rp9Pp11orAr0YEfusn4odTjGi/jxfYmTSpEm4++678f3332vECNDUP3kL06g/32gf1HkjgmByRvRiwt8wjcih07vQzUlMi5G0tDTlwlZ3vr46EF85I2oxAjTad127dkVubi7q6+vx/fffB9XeLVu2BJUoZib6i9bTYFJbW6uZ1qgfCIU93759e6XWh4i7q8VIRkaGm7rXd1Ddu3dHmzZtUFdX59c0av1dmBisvZVlNtqPQEI1/oZpqqqqFIsVMMcZEWJLtH3t2rXo3r27YpEL5s2bh5deegmAcacIaKc1DhgwQHMHXVdXp4gMkdMgzgFJkpROWz1d2F8xoj/WW7du1fwWzRGmAbyLERGSVeOvAFSHaWpqatDQ0GApZwTwHKrZtWsXXn/9dY2TKK610tJSNDQ0GCavAu6/X7jEiLdrWZZl5TzdvXu3mxgRA7i32TRi//T5YgLRx6mv92DCNEOGDDH8XDX6c7impkap2qq/8WtOYlqMSJJkeDfjrQPR3/0Z3TmIO8fBgwdrvkucKBs2bFCef+yxx/Dkk0/6bOtXX32FQYMGaea1RwJ9p++pAyoqKtJ0PvpBVX1B9+rVCwCUgUw9OGdnZ7t1QPq/JUlS3JEvv/zS732Ij49HUlKSMtVOXSjIE2qhapYYUTsje/bs0XyHGWJEiC3xue+++y527dqFt99+W9n24MGDuP/++3HXXXehvLzcMJEOgFJNt3fv3khLS9OIKPW1IDpQMfOmZcuWyt2fukCWJzGiz3fRH2t9bZnmDtOI76utrVUWa7zlllvctlf/lrW1tXj88ccNKxLrHb2KigrN8VRX44yWnBG9GPGUxHrHHXfgpptuwgcffKA8pxb+J0+eNExeBaLDGVG7UgcPHlRupFq3bg0AuPHGG9G9e3dcdNFFPp0RT+33FqYJRIycffbZhp+rRi9GxA1DWlpa2GrX+ENMixGgKVSjpqKiwmPdipqaGs2Apb9zcDqdygAtShkL9GLk5MmTmDVrFmbMmOH1YpBlGYWFhXC5XNi6dasp60AEi7/OiDpEA3h2Rlq3bq2pYghoB+ecnBzNBZyUlGTYGYuF89auXetzH8QAkZGRAUmScNppp8HhcKCystJrqMblchmW6vaFLMt+OyN618xMMSJ+A9H5FBUVKdsKt8LpdCohRYF60FGHaABoEljF8UhISFA6NfFd6t9Rff14KgomrivhsKjPO1mW8a9//QuANilWTXOFabZv3476+npkZ2ejX79+SihKOK7q8+U///kPZs6cicLCQrfP1U8VLi8vdzu/xADXHGEaf5wRffjO042J6Av++9//Ks+phcvx48c1Bc/UhFuM+FMSXh1+OnDggJszcu2112Lnzp3Iz88PWowYhWn8zRlRz/ISbrzAnzCNOkRjVoHAYIh5MaJ2RoR9Jsuyx5NT30Ho/z5+/Lhyh693XfRiRH2Se8u9+Pzzz5VBtqGhwa+lp4148sknMXbs2JDKjgcqRsRg4ClnRO2MCNRhGr0z4ulOVzgj69at85lIp+8YRBgNgCZfQ48+adVfZ0RdWdVTzogQI+oaFUB4ckZE57N7925lW7HIlv7/+jaIyqJCjKgTb8W1kJaWpnSQRmJEja8wjbhZUB/rzz77DJs3b0ZaWhruueceAO6h1eYK0whR3aFDB0iSpBSwGjRokFu7hCNilFuhd0bKy8s9XqfR5owIwagWGDt37sSpU6cgy7LyvMjxcblcGtF//Phxj4O1L3ESKv44I+r+ds+ePcr1YOSqexIj4jzw5YwY5Wr5ckbS09Mxbdo0TJs2TVlYUP+5+u0BdzESyRANQDGiOaHatGmjCBJPoRr9AKTvMISoaNWqlRI/FAjVumPHDpSWlmqKQXkSI/X19W53UsFWO5w3bx4+/PBDvP/++0G9H/BfjIhpvWLQCsQZSU5OVizQjh07+iVGBg4ciPT0dJSVlXmcoaHfB/XnijZ4EyP6Aa+yshLl5eU+FwJUnzNikBbfHR8fj/j4eEUwrF+/HkCTMA6nM1JcXKxsoxYg+hLwgToj6vCNEKWtW7c2/O18iRGR76I+75544gkAjfa/cCIiNZtGb6Vfe+21SE9Px0033QRAe86INZSMBj69M6IP06iJlpwRMVh2794dQFO/tHTpUvTs2ROFhYU4ceKEkuj9/fffo6ysDMePH9cM1t7ESGJiosYJikSYRi1GREhOkiRDkeBpNk2gYRqXy6WEskRf6I2nn34aTz/9tJvT748YEaHUSCavAhQjmh8vKytLuZMqLS3F119/7bMSpP5vofhFJ6omJydHyU/YtGmTRox4qrcwffp0bN68GRkZGejcuTOA4MWI6KBF4mJpaWnAIR9xUQnnwmgwWb58Od58800ATQLMW86IkRi5/vrr8cILL+DRRx/1S4zEx8crOTq+EoTVYRqBkRjZsGGDMp0PcBdUlZWVmDBhAs4880x8/fXXHr9PnCPJyclKPQGxT6KjnT59Olq0aKF8l0jq9SZGXn31VTz77LOGeS4ul0u521aLkbq6Oo3wFWvM/Pzzz8pz6v8DTQNuZWWl0nENGDAAgG9nRIiczp07G3bEJSUlhjOgxHUnOmLRcX711Vf48ssvkZiYiKlTp3ocTMIVptEnPIprWMxiGD9+PMrKynDttdcCaHQ8xGAsBjKjgc/IGfEkRowSIM0mEGdELPQmwjR//etfATT+Vuq+SpZlfPXVV26J4t7ECKD9DfU5JaESqBgRIcycnBzlWlZjVpjm4MGDqKqqQnx8vKZiqy/UBRSBwMM0kSTmxYjaGWnZsqUyQC1cuBDDhg3DH/7wB832RnUU1IiOXgwAevLz8wEAP/30k09nZMmSJZg7d67Snr59+wIITozU1tYqd1+ffPIJNmzYgPbt2yt3cP4iLioRWtGLkU8//RQXX3wxTpw4gTPPPBO33XYbgMbOVT1oqp0RdVlxoCmH4o477kCHDh38EiNAU/6AGDB97YM3Z2TXrl0YNWoU7r77bnz11VfKPqiprKxUBm6jGidFRUXKdEBA25GK/wtR1717d2WNHqAppOdJjNTU1GDy5MmYNm2a4qaoUQ9u4lwsLS3FwYMHNb+DCNX4E6ZRL1Qo7rg8OSP6MI0nMaL+XDV6MSJ+M5G4etVVVyEvL8/jYNJcYRqjJEOHw6HZ14qKCpSUlCg3Kkaz9cTvJQY4T2IkMTEx7NVXgcCcESFGDh06hN27dyu/0d69e936qjVr1riJkZKSEo8JrEDTNZ+SkmL6KsX+iBF1WE1cO0YhGgA+Z9P464yIfqVbt25uDrs3JElyu8HW48kZYZgmwqhPKrUzIha+Uk+zBJp+QHGCBOKMAE3Lfx85csSrGJFlGdOmTQMATJs2DVdeeaWyWmcwYkTdAR4/fhz/93//h+rqar9mn6gRF1WfPn0AwG1we+uttyDLMq644gp89dVXaN++PQBtbQ9AW1IZ0C7FrU/Q81eMiIspVDHS0NCAm266SWmv6BiMnBExGOlzPQDgL3/5C+bPn6+UrFd3tHpnBABuuOEG/OlPf8LYsWNx5ZVXAvCcM7Jr1y4lD+Xll192e109G0XY6GVlZW6uUVFREU6ePKk5//TJx2IfhfBUC0f1oGUkRgSdOnVy++3E8TBKYhXt1zsjIulWnH/iM8Vv+vLLL2PlypURC9MI4uPjFXehvLxcCdEAxgv/iRsFMZBUVFQYCoHmCNEAwTkjhw4dwsKFC5XXjx49qpxLIjFy7dq1hs6IpwRWoOk3NDtEo/5Mf50RgafQSaizacTNh3Bg1Avx+YtajBjNjqEzEqXofzghRsTJoL8zFRenOBkDdUaESDl69KjXMM2GDRtQVFSEtLQ0pYS8WWIEaBqwjxw5ElAlUb0zUltbq7lYf/rpJwDATTfdhOTkZCQnJyvCTX0sxd2GOI7qJNZgxYgIY/kSI97CNLt27cKcOXM0boMYAPVipKKiQumQjfJGhMMghIqRGNEvCPjwww/j/fff17gZRqhF8jvvvOPWNjGYx8fHo2XLloprJES2YPfu3W5hGRG6E+EAvRgRAhOA4dRedc6IQO+MOBwOJe9E766pC8rpc0bEbyH2R73ezfbt2zF58mRcfvnlyvsj4YwIxPlVVlamhGgE+twz4YwIce7JGWkuMeLJGdm+fTuGDx+ODz74QDnnhBgpKSnBvHnzNNuL6+iCCy4A0FiDSb1MBOA7TCOOo9nCUv19gYoRT86IWWEacU0GI0ZEyDAjI8PQSVKfw7IsU4xEC56cEYEvMVJfX69JVvLljKiXq/fmjLzzzjsAgN/+9rdKxy4GgWDEiKeE3Nra2oBWcxUXVU5OjtIeYfW7XC5lABYhJUmS3GbUuFwut+lxkXZGOnTogOTkZNTX1yuLd4nBUuyf/jgVFxcrnc6OHTvcBg8R8hHbeAvT6BGdU2lpqaFYVLsXNTU1So6O+jn154tZHkKMiDvVoqIi5TfT2/9iwBcdpDdnxFOYRtC5c2dNwad27dopyad6MaJ2dTw5I3ox0tDQoAg0T6XFzcCTGDEqoy3OVb0zIp5TI5wRcT14SmBtjmm9gDYXSM2SJUuwfv16PProo8p52a1bNwwbNgxAY7uzsrKU61mIkXPOOQcdO3ZEfX290reJfsFfMRJNzkigYkTc3Hqq4+EpTBOKM+JpSrA4h6uqqnDy5EnlXKYYiTD6nBFfYkT8cOr3qS9YX86I6Fy9OSMulwuLFy8GAFx33XXK82Y4I23atEFaWhoyMjKUji2Q1W7VnYaw/8XAuHfvXlRXVyMxMRHdunVT3iMuQCFGSktL3aY/m+GMqMWIN7fHyBlxOBxKXZja2lr06NEDDz30EADPzohI/gQaw2rqu9+amhrN64BWjAwcOBDx8fFKoqoeccycTqfhoCQGXnGevfbaa5rX9WJEJJyK/Azxd1FRkdLxiemoAvEbegvTqActcR3oxUhiYqJy3ovfskOHDkrnt2vXLjz77LNK9Vz19STOj8rKSjQ0NChCUwgZ9TFVT1UW32V2foW+LoW4ho2cEdGXlJeXuzkj+sEvWp0RfRtE/yZmVYlVqNeuXYv169fjz3/+M9577z3leha/Sfv27XHJJZcAaDp3haMS7WJE3Dipa3D4EiPqG9RNmzZh6dKlkCQJV1xxheH7PIVp9GUP/MFfMaL+npycHMM1bpqTmBcj+mQf/WB38uRJzcAmLs7s7Gylo1NfsGY4I1999RUOHjyIzMxMXHzxxcrzZoiRrl27YuPGjfjuu++UQcWf1WoF3sSICNH07t1bYw+qZygBTcIrMzNTuVNWOyP6Dlc9sHkTI2J/hOL3Zx/UqNswffp0Zf88iRG9A6POGzGq5qrejx49euDIkSN49dVXDduYkpKiHBujfRHHfMqUKQAaBwf1WjmenBHB+eefD6BxsBDOyMiRIzXbiJlf/uSMeAvTdOrUSblWxO/XsWNH5XNee+01TJs2Dffddx+AJjGSmpqqGSwOHjwIp9OJxMRE5VqIi4tTvktv/5sdogGCC9OUlpYq14ZIUNUXcRPOiBBtnuqMNHfOiKcwtDi3xX7HxcVh6NChKCwsxHnnnee2uGVeXh4KCgo0z6nFiOgbvM2mibQzIvoDwH9nRJ37N378eGUCgx4hHKqqqnDixAnlWgtFjHiqT5KSkqIIK3HtR9oVAShGNKLCyBlpaGjQXJDqSppGcdVgc0bE++rr65WEx6uuukrT+YgO+MiRI4ZLVHtDndDXp08fdOvWTSOM/EU9kAsnQS9GRIhGoA/T6PNFgEaRNHLkSPzmN79xs/gdDocyCHgTIykpKcpnegvV+BIjHTp0wPjx45XB+OjRo6iqqlIEnciB8SZG9FVMAfeZAupzT48kSZpQjR5xdzly5Ei0aNECTqcTO3fuxJ49e3DHHXcov4knMSIq1paXlysF9URcXyD2X0wB9+WMeArTqLP01c6I+Bzhkonz0EiMVFZWKqKwc+fOmuMmzoloFSPff/89qqurkZSUpAwu6sFPLSKNwjTqqbzNFabx5Izoc9s83X2L/C1BXl4eLrzwQs1Nijgnjx8/rvQd+mn+QOSdESFG1C6mv7Npli9fjtWrVyMpKQl/+tOfPH6H+lz95ptvADT2j74KnhkhwueexiBJkpTzWNQUivRMGoBiBA6HQ4n3GuWMANrBwOjuTzzX0NCgDLSenBExWOpt/LKyMtTU1GDixIlYvnw5UlJScO+992rem5ubi7i4OLhcLo91ScQiW3qMQhPiZPUWptm1a5cmvm3kjIiB0ZMY0Ydp9DNpgMbfYcWKFVi+fLlhSWLRafhKYvMnb8ToWADAhAkTcNZZZ+GFF15AYmKi23oqov1CFOqPmzqJ1R8x4gtPYqSmpkYRBj179lSqLm7fvh2PPfYYXnzxRTz11FMAmsRIjx49NMK2Z8+emvojPXr0wIUXXqiJaQsxIsrge3NG6uvrlXbqxYh6YDISIwLxfrUYUYdF9Pki+s/Uh2nCLUZkWfYrZ0RUre3Ro4dyfPVr1giMwjTq6yRanBGBp8HSyBnJzMzE8OHDleeEM3LkyBEcOnRIU8FWjRAoRkIlVHyVg3e5XMpN48CBA5Xn/XVGRLh90qRJXgf8uLg45XwRNYuCyRcBgOuvvx6PPvooZs6c6XEbvRihMxIliDv8rl27BiRG9M5ISUkJZFl2m+utRv0+ve3/l7/8BW+++Sbi4+Px7rvvak5+oPGEFQOIUaimtLQU3bt3x4UXXuj2mlERKF/OyJ49e9CrVy9cdtllynP+hGk8OSPiOBo5I74wU4x4ckZ69eqFDRs2aPZXDHy7d+9WBhD1bBKg6Q7vhx9+0CS0AtqBOFgxog/TCAcgMzMTOTk5ihjZtm2bstSAyAMRYiQ+Pl5TKrpjx47KvmVkZOD9999HcnKy5m6qbdu2yvsPHTqk3CEaiRGg6XfVh2nUx0AMKGeccQZ69OiB7OxsTbxclmVNiEk9+Asxoh/oxO8oXhdWuBCNZqJuT3l5ueLqGDkE4lwV+RU9evRwm4oMaGvCGIkRdV8SiZwRo/pAAn/EiCRJSl+jDtWIfkJ8fvfu3Q3dj5tuugkbN27Eww8/HMSeeMeXM3Ly5Elldpm6P/Znaq/L5VIKTIqp+t4Q51CoYiQjIwMzZ870GuIR57EQyuoQVKSgGEFjbYwvvvgC/fr18ylG1GEavTMi7pRzc3O9FufRn8iigxKLf91///1u8VWBt7yRlStX4tChQ1i7dq2b0DFyA3yJkY0bN8LpdGpmAqgHcpHgePz4cZw4cUKJP+rXR9CHaQIpcywQF6avGGogzog/tq9wB4qKitycEcFZZ52F5ORknDp1Svle4Yz89re/VbYLVIwYrVcBNDlRPXr0gCRJSs2Nb775RrnTEaJIPVtHiKbU1FRkZWXh//7v/9CqVSu88847yvEVdXDE94vBRpTYT0lJ0Qy8CQkJSh6EWox4CtM8//zz+O6773DhhRciLS0Nv/zyi5LcKZJ11c6IOF/r6+uVjtOTMyLCHTNmzMALL7yguENmog4bCVdEvbCjGvXUXkA72BqJkcTERMVhOXbsmFvhN6D5Z9O4XC7luNbX12tCy4B/YZo2bdoo/eGll14KoFHQ5+TkaMSVp2Ruh8OBwYMHh6XyrPr3mDVrllsiuBDgmZmZmvPOH2fk22+/xeHDh5Genq7kaHlD3HwIMRJMvoi/iL5IiEt9nx0JKEbQ2FmOGjUKgNY5ECeDP86Iy+VSBnVPIRqB+vX4+HhlUBcDiT6RUI03MaKuAqoPExhVpPQVphGDXmlpKerq6pQH0HgRp6enK5/xxRdfoKamBklJSZqZNOrvFJ2ycFLEQO8Pb731Fnbu3KkMvJ4IxBnxp26B6IDUYkTvjOTm5iqdk+i8hDOizp43K0wjjp+4mxEdyeeff+6WNKt2KMQMGrE65/3334+jR48qsxwAbZw5KytLGWyEIBULwgkkSVKuA9GxeQvTpKamIj8/X/mMVq1aoX379konXlpa6iZG9NOS9WJE/zu2bdsWd9xxh9t5aAbiN6yvr1euG0/ugL5dajFiFKZJTk5WjtXevXsNZ+41d5gGaOrzhNiMi4tTrmlP+56Tk6OcA2rx3r9/f7z//vtYsmQJJEnShLc8iZFwor4heeyxxzB58mRNnoy4nlu1aoXOnTsjJSVFcSSNUM+m+fDDDwEAF198sV+/m7jWxLkxYsSIwHfIT/R9EcVIFCI64+7duytxNF85I//617+QmpqK5557TvMZnlCLkZycHDfxIubsGxGsGDGqSOnLGVEX1jp27Jjmbk5cxCLEJWoH9O7d223NBv2gKgZT8V5/SE1N9ctKDCVMY4RajHgK02RnZ2vESElJiXLXfPbZZyt3tmaJEbUzAjR1JEa5Quo79lGjRkGSJM35pc/PUZ+7LVu2VAYbtRjRI64DT2LEV3KcPllXLUYAYPLkyZrtPTkjgkAct0BR75c4x4IRI0bOSFJSEjp06ACHw4FTp04Zrg7bXGIkISFBScYUv4c610usOeXJGZEkSQnV6J3EsWPHKssdqENQkRAj+mT5hoYGZYo5oBUj6enpWLVqFVauXOnR+VY7Ix988AEAaMK+3lAfy1GjRil1jsKBvgCjvk+LBBQjOgYMGIBXXnkFb7zxhuFgYDSb5t1330VtbS0+++wzAL6dEXVnqRcjffv29VgYB2i6sPXreZw8eVJT6tuTMxKsGDl69KhmkTxx0QmBsGTJEgDAb37zG7fP0Tsj+sHUTDyJkfXr16NXr174/PPPQ3ZG9J1rdna20qkeO3ZMOfYdO3ZEamqqEmv2lEfkCU85I+L4iWMv7tiMUD9/xhln4PDhw3jllVc8fqcQI2lpaUhISFAGWpH3YCRG9DlQeXl5mjtrf5LjvImR8ePHa8ITnnJGBJ4sdDNISEhQBIE4xzzdJevPL085I2pnJDEx0dB5EzRXmAZo+l03b96M2tpajfN73333YdiwYR7rZgBNjpi33J1IOyNGyfLr1q1T/q8vzjhkyBCP03OBptk0JSUl2Lx5MyRJwpgxY/xqi7rfFzWOwoVajJx++umGx6G5oRjRIUkSfve732HIkCGGYsTIGdFb46E4I+pscyNER6UfbL/88ktNO/Qlvr2FaTyVhPfkjKg7f7VbIVZS1aMWI8ePH1cG13DY6EKMHDp0SFN4aP78+dixY4dmFd5AckZ2797tMUyTk5OjCI2SkhJlVocQW3PnzsVTTz2lxMv9xShnpK6uTrlzE2EXh8OhCV+pBYhepLRu3dpwtVGBOCfEd4u7M1Ey2ugOSi08UlNTcdZZZykCKT8/36+7eW9ipGXLlsoquKmpqW5iQ/07ivL34UR05GI2nCdnRH2tJSUloX379oZhGrUzAriLrUgksAJNx/Xyyy9Hfn6+Jtfr0ksvxbp167yGTYVj5+2mQ4iR3NxcTb5Sc/Lqq69i2rRpmDNnDgCtGFE7I/6gd0zOPPNMv506cR7l5+cb3tSZiVqM+Ap9NxcUI14wujP1VvJaEEjOSE5OjuZk9SVGRB6LXmyIlTLF3UggYZra2lrDhFd15rzaGfEkRm655RbDuyD11F4Romnfvn1YKv7l5uYiJSUFsixjwYIFyvNiQUAx/dbhcHh0E9R07twZCQkJqKysVO5gjZwRdZhGhNDEwN2vXz889NBDfn2fGiMxvG7dOlRVVaF169aKGAG0MV+16An0O8VAKAaGqVOnasIs3pwRoLGYWmJiIrKysvDrr78armZshDcxAgB333034uLiMGTIELe7OPX5mJubG/a7PH/FiPpa69atm2Y1X0/OCOA+E0X9+c0pRmbOnIkzzjgDDocD27dvVxIrffVvgsLCQvzzn//EHXfc4XEbIUYGDhwYsbvzW265BU8//bSSo7F+/XplBk2oYuTcc8/1ux0TJkzA6NGjsWDBgmY7h4HoyBcBKEa84itMo+4su3Xrppki7A1vYRpfYkScOAcOHNDcXYlO/9ZbbwXQ6GqoVwY1CtOkpKQonaM+VKNfudWTGBHiKC4uzqO1qJ7aG84QDdAoMkTFw9///veYN28eDh48iD179gBocpQyMjL8uuCTk5Nxyy23aJ7T38GpwzTqpeJ9OWS+MDr/RChw9OjRmsJf4u6mV69eGhs5UDFy7rnnYt68efj73/8OoFFo/O1vf1Ne9yVGRCI40Cja/C1Upd5XMbVXfX0NGTIE33//Pf7973+7vVd9ToczRCMQ4kCERf0RI0K0e5vaayRG9Dc9zRmmufXWW7F582blfPI3DC3Izs7GzTff7DVXShyXQAbtcHHGGWcgJSUFJ06cwI4dOyDLstJv+Hte6cXIOeec4/f39+/fH5999pmSTxNOKEYshn4wKC0tVToRfScxdOhQfPzxx5g3b55mOqcRemdEDFo5OTk+B+mWLVsqd+ZiKm19fb3SMY4bNw5JSUmora11K6oGuBeCUodqAGDhwoW46aab3FahPXbsmOGU2AEDBuCxxx7Da6+95lGENacYAYBHH31UqUnwhz/8Ae+++67bNoFUc3zkkUc0f6enp2uEgD5MI+zsUG1ncf6pp1N+/vnnABrFiJorrrgCWVlZuOWWWzThr0DFiMPhwL333qskKAKNNv0dd9yB0047zXDQUIsGtRgJBCNnRN/2vn37Gt6hqn/LcCavCsQxEKErf3JGxPkeaJhGX7OlOZ0RgcjlEAOzmcd4ypQp+OSTT/Dggw+a9pnBkpCQoJz3r7/+Oq688kql7/A3lBGKGGlOKEYshrqDnDlzJrKyshRhkpGRoekkBg8ejO7du+Pee+/1WmMEcHdGzjvvPIwbNw7PPPOMX3fr4uQRRcZ2796NhoYGpKWloUuXLkrHJ0I1six7rDoqhNHhw4fhcrkwdepUvP76627V+zw5I5Ik4ZFHHsFNN93ksb0iJFNVVYW3334bQHiL7EiShD/96U8YOnQoGhoa8Mc//tFtm0DESIcOHTR1X9TllIHG/Apx53Ts2DHTxIhwnX788Ufs3LkTR48eVRJJL7roIs22p59+Ok6cOIGHHnpIM2U6UDHiiRdeeAG7du0yHHiFKM/JyXEr1Ocv6lCeUZjGG/owTbjRC8FAnBF/wjRqUZ+amhpxMaL/Tf11RvwhOTkZl1xyiWnnaagIZ/qJJ57Ae++9h4SEBDz22GN+z4hR9/0dO3YMS+E9M1CvGq4v3R8pKEa8oBYjr7/+OoDGBMmpU6eiffv2Gmdk8ODBfn+u3hlJSkrCG2+8gYkTJ/r1fnXFTaApf6Rnz55wOBzKICbEyKlTp5Rpn3pnRD2j5vvvv1fyY0Teg7hL8yRG/CElJUXJuhf1N8LpjAjuvPNOAE0dv1ro+TOTRs2iRYtw3nnnYdasWQC0F3NKSkpYwjRdunTBpZdeClmW8be//Q1ffPEFgMbBwdtnh0OMeEMMliNHjgx6lVxfOSPeaG4xMmLECGXWBOBZjBjlVnkL03hyRiIVphHoZ7mYKUaiDVFzR5IkXH755di0aRMeeeQRv3M41OfF2WefHZY2moHov/r06WP6ytbBEh2tiFJEB1lcXKxYlJs2bcKzzz4LSZKUztLhcAQ0LS0rK0tR0J4sXm/oxYgQHaKKpvhXiBQRolEXqBKowzQrV650+y5hM4YiRgDgxhtv1PzdHGLk2muv1QwU6gz1QPehVatWWLNmDWbPng2g6WIWv184wjQAlPWJXn31VcyYMQMANCs5G5Gdna2IzuYQIyNHjkRycjImTJgQ9GeoxYg6FOoPamHZHGGa9PR0Ta0WT2IkMTERrVq10lTJVYdptm7discee0y5ARBCQ9QaAaIjTKNOlAaa5xhHigsuuADr1q3Dr7/+ivfee89t332hdkaiWYxccMEF6Natm1s+XCShGPGC6CAPHToEWZY1iYpAU2fZp0+fgApaqddqCLT2BODZGdGLkS+++ALV1dWaEI2nIle//PKLIkbUSYrCtvQ0tddfLrroIs2+hmNarx518mnXrl01JZlDXQFU/N5iIBJ35MePH1eOkxli5De/+Q1OP/10VFVVYe/evejRowfuu+8+r++RJElxR8IxY0nP+PHjUVlZ6beVbYRajHibRmxEczsjgDZU4+2GYsmSJVi6dKmyL+py8lOnTsWsWbPw1ltvAWgSGgkJCco1GA1iRF8K3c7OCNBYdDKQ6tBq1GIkWvNFgMb+99dff8WUKVMi3RQFihEv6OsV9OnTRzOYjxw5EmeccQZ+//vfB/zZM2fOxPXXXx+UehZiRJSM1ouRyy67DHl5edi1axemTZvmMXkVaOpUly5dihUrVgAAXnrpJaSlpSEjI0NJ1jt69KiSLxPMQJ6QkIDrrrsOQKPYaa4Y8QMPPIBRo0bhkUce0eSpBBqm0aMXI/q7Y/VMpVCQJAnTp08H0JgkvW7dOr9Ezh/+8AdcfPHFQSeUBoq32iX+oBYjIvHa31h2cyewAtqcHW/LvJ933nmahHZx3smyrCwVLxLRjQq76cVIJMI0gDZU01yCz4okJCTgzjvvxE033RTVzkg04j3TMsbRixH9KoqdOnVym3XiL5MnT3Yrc+0vYjrwkSNHsH37dkWMiFyRli1bYuHChRg9ejReeOEFpTMzGoCHDBmCgoICfPLJJ6iurkbLli0xevRofPPNN5BlWRkQqqqq8NVXX2m+J1Buv/12LFy40OMigOGgXbt2Sq7Fpk2blOfNckbEXXF8fDyysrIUy71du3am1QoYP348Bg0ahJ49e/pMjhbccMMNuOGGG0z5/uZA7UKKqpfBiJHmGigHDRqEAQMGoLa2NiCnICUlBQ6HAy6XSykTIHJk1K5Hly5dsGbNGqSmpmqEeyScEaAxT2np0qXIysoKy4J1dmL+/PmRboIloTPihdTUVE3nH+ySzuFAuCNr1qzByZMnIUmSJg/joosuwt133w2g6eIwckYAKHkQQGPRqri4OPTt2xf9+vVDenq6cjcmZu8MHTo0qDb3798fR44cwYsvvhjU+0NFHRoK1RkRITr1XbF6IDSzmqQkSTj99NP9FiJWRC1GgEaB4W8l1UiIkbi4OGzevBk//PBDQL+LJEkehbDa9RBhgszMTMTHxysCIFJiRNQaidbZIcT6UIx4Qb2AFxA9ZXOBxqqeQOOUS6DxLlKfHyCSRkURKU8D8JAhQ3D55ZcDgFvJckmSNB18bm6uz6Ju3khLS4tYpcWsrCxFPJidMwJo839CnUkTa+iFR+fOnf0+T5KSknDGGWegffv2zTpN0eFwaGZP+Iun61AtRm699VbcfvvtuOeeewA05f5EKkxTUFCAe++9F0899VREvp/YH4oRH6hXUowmZ+TWW2+FJEnYtWsXAOPQyeDBgzUzEry5AW+88Qbee+89pYKrGnUcfujQoVGxqFKw6Os9BMuYMWPQunVrZSogoBUjkVpnw6oYiZFA+Oabb7Bz586IOQeB4OncU7e9ffv2WLBggXLTIcRIpPYvISEB8+bNCylJmRBvUIz4QHSSSUlJbgtYRZIBAwZocgKMhFJCQoKmYqanMA3QeKd/+eWXG84514sRK3PFFVcgLS0t5P245pprcPjwYWU9CyB8YZpYQB8SDVSMJCYmRk3hLF/4E6bRI5w2Jo8Su0Ix4gMhRnr27BnyjAGzefTRR5U2eXJtRo4cqfw/2DwJtRixeoZ4YWEhSktLldVoQ0HvEDFMEzz6kGi0VIUMB56uQ2+ux8KFC/HGG2+gf//+4WoWIRGFYsQHooOMphCNoHv37nj88cfRo0cPj/apWox4c0a8Ie7GJEnSrFliVcKVCMowTWjEihgRzogkSZrp5t6ckf79+2PcuHFhbxshkYJixAfiDlfEbqONwsJC7Nixw3A1VQA488wzlc4vVGekb9++Ic9CsTMUI6ERa2LktNNOQ8+ePZXnrZDvQki4sO9cQZOYNm0acnNzcdddd0W6KUERHx+P0aNH491330WnTp2C+oxzzz0XiYmJStEyYow6ns8wTeDEmhjp27evRrRGaqYMIdEAxYgPOnXq5LaEvNV44YUXcPPNN2PMmDFBvX/YsGEoKytjZ+kD4Yw4HA4mGgaBECOJiYm2Ljku3JBzzz0XdXV1yvN0RkgsQzESA+Tm5ip1RIKFQsQ3PXv2RFpaGk4//fSoS3a2AkKMdOrUKWpWEg0Hd911F8455xwMGDAAb775pvI8rzESy1CMEGISWVlZKCoqCmjRRNKEECN2DtEAjdVbxWyujh07Ks9TjJBYhmKEEBNheCZ4RP6EOqnT7qjFCMM0JJahGCGERAWTJk2CJEkxlSjdvn175f90RkgsY9/ALCHEUrRs2RL3339/TC3GlpKSorgjYgVoQmIROiOEEBJBFi9ejN27d4e0ACUhVodihBBCIsjQoUMtv+YTIaEScJhmzZo1GDt2LPLy8iBJEpYtW+Z1++LiYowbNw69evWCw+HAfffdF2RTCSGEEGJHAhYjVVVVGDhwIJ5//nm/tq+trUVubi4efvhhDBw4MOAGEkIIIcTeBBymKSgoQEFBgd/bd+nSBX/5y18AAK+++mqgX0cIIYQQmxOVOSO1tbWora1V/i4vL49gawghhBASTqJyau+TTz6JzMxM5aEuDEQIIYQQexGVYqSwsBBlZWXKY//+/ZFuEiGEEELCRFSGaZKSklgamRBCCIkRotIZIYQQQkjsELAzUllZiV9//VX5u6ioCFu2bEF2djY6deqEwsJCHDx4EIsWLVK22bJli/LeY8eOYcuWLUhMTMTpp58e+h4QQgghxNJIsizLgbxh1apVGDlypNvzEyZMwMKFCzFx4kTs2bMHq1atavoSSXLbvnPnztizZ49f31leXo7MzEyUlZUhIyMjkOYSQgghJEL4O34HLEYiAcUIIYQQYj38Hb+ZM0IIIYSQiEIxQgghhJCIEpVTe/WISBIrsRJCCCHWQYzbvjJCLCFGKioqAICVWAkhhBALUlFRgczMTI+vWyKB1eVy4dChQ2jRooXhzJxgKS8vR8eOHbF//34mxoYZHuvmgce5+eCxbj54rJuHcBxnWZZRUVGBvLw8OByeM0Ms4Yw4HA506NAhbJ+fkZHBE7yZ4LFuHnicmw8e6+aDx7p5MPs4e3NEBExgJYQQQkhEoRghhBBCSESJaTGSlJSEWbNmcVG+ZoDHunngcW4+eKybDx7r5iGSx9kSCayEEEIIsS8x7YwQQgghJPJQjBBCCCEkolCMEEIIISSiUIwQQgghJKJQjBBCCCEkosS0GJk/fz66du2K5ORkDBo0CGvXro10kyzN7NmzIUmS5tG2bVvldVmWMXv2bOTl5SElJQUjRozATz/9FMEWW4c1a9Zg7NixyMvLgyRJWLZsmeZ1f45tbW0t7rnnHrRq1QppaWm4/PLLceDAgWbci+jH13GeOHGi2zk+dOhQzTY8zr558skncdZZZ6FFixZo3bo1rrjiCvzyyy+abXhOm4M/xzoazuuYFSPvvPMO7rvvPjz88MPYvHkzzjvvPBQUFGDfvn2Rbpql6du3L4qLi5XHDz/8oLw2Z84czJ07F88//zw2btyItm3b4qKLLlIWQiSeqaqqwsCBA/H8888bvu7Psb3vvvuwdOlSvP322/jyyy9RWVmJyy67DE6ns7l2I+rxdZwB4JJLLtGc4x9//LHmdR5n36xevRp33303vv76ayxfvhwNDQ0YPXo0qqqqlG14TpuDP8caiILzWo5RhgwZIt9xxx2a53r37i1Pnz49Qi2yPrNmzZIHDhxo+JrL5ZLbtm0rP/XUU8pzp06dkjMzM+UFCxY0UwvtAQB56dKlyt/+HNvS0lI5ISFBfvvtt5VtDh48KDscDvnTTz9ttrZbCf1xlmVZnjBhgvzb3/7W43t4nIPj6NGjMgB59erVsizznA4n+mMty9FxXsekM1JXV4dvv/0Wo0eP1jw/evRorFu3LkKtsgc7d+5EXl4eunbtiuuvvx67d+8GABQVFeHw4cOaY56UlIQLLriAxzxE/Dm23377Lerr6zXb5OXloV+/fjz+AbJq1Sq0bt0aPXv2xG233YajR48qr/E4B0dZWRkAIDs7GwDP6XCiP9aCSJ/XMSlGSkpK4HQ60aZNG83zbdq0weHDhyPUKutz9tlnY9GiRfjss8/w8ssv4/Dhwxg+fDiOHz+uHFcec/Px59gePnwYiYmJyMrK8rgN8U1BQQHeeOMNrFixAs8++yw2btyICy+8ELW1tQB4nINBlmVMnToV5557Lvr16weA53S4MDrWQHSc1/GmfIpFkSRJ87csy27PEf8pKChQ/t+/f38MGzYM3bp1wz//+U8lGYrHPHwEc2x5/APjuuuuU/7fr18/DB48GJ07d8ZHH32Eq666yuP7eJw9M2XKFGzduhVffvml22s8p83F07GOhvM6Jp2RVq1aIS4uzk3RHT161E2Jk+BJS0tD//79sXPnTmVWDY+5+fhzbNu2bYu6ujqcPHnS4zYkcNq1a4fOnTtj586dAHicA+Wee+7B+++/j5UrV6JDhw7K8zynzcfTsTYiEud1TIqRxMREDBo0CMuXL9c8v3z5cgwfPjxCrbIftbW12L59O9q1a4euXbuibdu2mmNeV1eH1atX85iHiD/HdtCgQUhISNBsU1xcjB9//JHHPwSOHz+O/fv3o127dgB4nP1FlmVMmTIFS5YswYoVK9C1a1fN6zynzcPXsTYiIue1KWmwFuTtt9+WExIS5H/84x/ytm3b5Pvuu09OS0uT9+zZE+mmWZYHHnhAXrVqlbx7927566+/li+77DK5RYsWyjF96qmn5MzMTHnJkiXyDz/8IN9www1yu3bt5PLy8gi3PPqpqKiQN2/eLG/evFkGIM+dO1fevHmzvHfvXlmW/Tu2d9xxh9yhQwf5iy++kL/77jv5wgsvlAcOHCg3NDREareiDm/HuaKiQn7ggQfkdevWyUVFRfLKlSvlYcOGye3bt+dxDpA777xTzszMlFetWiUXFxcrj+rqamUbntPm4OtYR8t5HbNiRJZl+e9//7vcuXNnOTExUT7zzDM1U51I4Fx33XVyu3bt5ISEBDkvL0++6qqr5J9++kl53eVyybNmzZLbtm0rJyUlyeeff778ww8/RLDF1mHlypUyALfHhAkTZFn279jW1NTIU6ZMkbOzs+WUlBT5sssuk/ft2xeBvYlevB3n6upqefTo0XJubq6ckJAgd+rUSZ4wYYLbMeRx9o3RMQYgv/baa8o2PKfNwdexjpbzWvpfYwkhhBBCIkJM5owQQgghJHqgGCGEEEJIRKEYIYQQQkhEoRghhBBCSEShGCGEEEJIRKEYIYQQQkhEoRghhBBCSEShGCGEEEJIRKEYIYQQQkhEoRghhBBCSEShGCGEEEJIRPn/TpaYWMFXH88AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "ename": "ValueError",
     "evalue": "x and y must be the same size",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[62], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m \u001b[43mlstm_autoencoder2\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mplot\u001b[49m\u001b[43m(\u001b[49m\u001b[43mmaxes\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mseq_train\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mthreshold\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m0.95\u001b[39;49m\u001b[43m)\u001b[49m\n",
      "Cell \u001b[0;32mIn[40], line 70\u001b[0m, in \u001b[0;36mLSTM_Autoencoder.plot\u001b[0;34m(self, scores, timeseries, threshold)\u001b[0m\n\u001b[1;32m     67\u001b[0m normal \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39mwhere(scores \u001b[38;5;241m<\u001b[39m\u001b[38;5;241m=\u001b[39m threshold_score)\n\u001b[1;32m     69\u001b[0m plt\u001b[38;5;241m.\u001b[39mtitle(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mAnomalies\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m---> 70\u001b[0m \u001b[43mplt\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscatter\u001b[49m\u001b[43m(\u001b[49m\u001b[43mnormal\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtimeseries\u001b[49m\u001b[43m[\u001b[49m\u001b[43mnormal\u001b[49m\u001b[43m]\u001b[49m\u001b[43m[\u001b[49m\u001b[43m:\u001b[49m\u001b[43m,\u001b[49m\u001b[38;5;241;43m-\u001b[39;49m\u001b[38;5;241;43m1\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43ms\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m3\u001b[39;49m\u001b[43m)\u001b[49m\n\u001b[1;32m     71\u001b[0m plt\u001b[38;5;241m.\u001b[39mscatter(anomalous, timeseries[anomalous][:,\u001b[38;5;241m-\u001b[39m\u001b[38;5;241m1\u001b[39m], s\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m5\u001b[39m, c\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mr\u001b[39m\u001b[38;5;124m'\u001b[39m)\n\u001b[1;32m     72\u001b[0m plt\u001b[38;5;241m.\u001b[39mshow()\n",
      "File \u001b[0;32m/sciclone/home20/dchendrickson01/.conda/envs/simple/lib/python3.10/site-packages/matplotlib/pyplot.py:2835\u001b[0m, in \u001b[0;36mscatter\u001b[0;34m(x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, edgecolors, plotnonfinite, data, **kwargs)\u001b[0m\n\u001b[1;32m   2830\u001b[0m \u001b[38;5;129m@_copy_docstring_and_deprecators\u001b[39m(Axes\u001b[38;5;241m.\u001b[39mscatter)\n\u001b[1;32m   2831\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21mscatter\u001b[39m(\n\u001b[1;32m   2832\u001b[0m         x, y, s\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, c\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, marker\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, cmap\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, norm\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m,\n\u001b[1;32m   2833\u001b[0m         vmin\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, vmax\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, alpha\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, linewidths\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;241m*\u001b[39m,\n\u001b[1;32m   2834\u001b[0m         edgecolors\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, plotnonfinite\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m, data\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[0;32m-> 2835\u001b[0m     __ret \u001b[38;5;241m=\u001b[39m \u001b[43mgca\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscatter\u001b[49m\u001b[43m(\u001b[49m\n\u001b[1;32m   2836\u001b[0m \u001b[43m        \u001b[49m\u001b[43mx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43my\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43ms\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43ms\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mc\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mc\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mmarker\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mmarker\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcmap\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcmap\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnorm\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mnorm\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2837\u001b[0m \u001b[43m        \u001b[49m\u001b[43mvmin\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mvmin\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mvmax\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mvmax\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43malpha\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43malpha\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mlinewidths\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mlinewidths\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2838\u001b[0m \u001b[43m        \u001b[49m\u001b[43medgecolors\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43medgecolors\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mplotnonfinite\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mplotnonfinite\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m   2839\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43m{\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mdata\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m:\u001b[49m\u001b[43m \u001b[49m\u001b[43mdata\u001b[49m\u001b[43m}\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mif\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43mdata\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;129;43;01mis\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;129;43;01mnot\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;28;43;01mNone\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[38;5;28;43;01melse\u001b[39;49;00m\u001b[43m \u001b[49m\u001b[43m{\u001b[49m\u001b[43m}\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   2840\u001b[0m     sci(__ret)\n\u001b[1;32m   2841\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m __ret\n",
      "File \u001b[0;32m/sciclone/home20/dchendrickson01/.conda/envs/simple/lib/python3.10/site-packages/matplotlib/__init__.py:1442\u001b[0m, in \u001b[0;36m_preprocess_data.<locals>.inner\u001b[0;34m(ax, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1439\u001b[0m \u001b[38;5;129m@functools\u001b[39m\u001b[38;5;241m.\u001b[39mwraps(func)\n\u001b[1;32m   1440\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21minner\u001b[39m(ax, \u001b[38;5;241m*\u001b[39margs, data\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mNone\u001b[39;00m, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs):\n\u001b[1;32m   1441\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m data \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m-> 1442\u001b[0m         \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mfunc\u001b[49m\u001b[43m(\u001b[49m\u001b[43max\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;28;43mmap\u001b[39;49m\u001b[43m(\u001b[49m\u001b[43msanitize_sequence\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[38;5;241;43m*\u001b[39;49m\u001b[43mkwargs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m   1444\u001b[0m     bound \u001b[38;5;241m=\u001b[39m new_sig\u001b[38;5;241m.\u001b[39mbind(ax, \u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[1;32m   1445\u001b[0m     auto_label \u001b[38;5;241m=\u001b[39m (bound\u001b[38;5;241m.\u001b[39marguments\u001b[38;5;241m.\u001b[39mget(label_namer)\n\u001b[1;32m   1446\u001b[0m                   \u001b[38;5;129;01mor\u001b[39;00m bound\u001b[38;5;241m.\u001b[39mkwargs\u001b[38;5;241m.\u001b[39mget(label_namer))\n",
      "File \u001b[0;32m/sciclone/home20/dchendrickson01/.conda/envs/simple/lib/python3.10/site-packages/matplotlib/axes/_axes.py:4584\u001b[0m, in \u001b[0;36mAxes.scatter\u001b[0;34m(self, x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, edgecolors, plotnonfinite, **kwargs)\u001b[0m\n\u001b[1;32m   4582\u001b[0m y \u001b[38;5;241m=\u001b[39m np\u001b[38;5;241m.\u001b[39mma\u001b[38;5;241m.\u001b[39mravel(y)\n\u001b[1;32m   4583\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m x\u001b[38;5;241m.\u001b[39msize \u001b[38;5;241m!=\u001b[39m y\u001b[38;5;241m.\u001b[39msize:\n\u001b[0;32m-> 4584\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mx and y must be the same size\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[1;32m   4586\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m s \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m   4587\u001b[0m     s \u001b[38;5;241m=\u001b[39m (\u001b[38;5;241m20\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m mpl\u001b[38;5;241m.\u001b[39mrcParams[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m_internal.classic_mode\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;28;01melse\u001b[39;00m\n\u001b[1;32m   4588\u001b[0m          mpl\u001b[38;5;241m.\u001b[39mrcParams[\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mlines.markersize\u001b[39m\u001b[38;5;124m'\u001b[39m] \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39m \u001b[38;5;241m2.0\u001b[39m)\n",
      "\u001b[0;31mValueError\u001b[0m: x and y must be the same size"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lstm_autoencoder2.plot(maxes, seq_train, threshold=0.95)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 2.5975878477096557\n"
     ]
    }
   ],
   "source": [
    "LoopFiles = 9\n",
    "loops = int(len(BeforeTamping) / LoopFiles) \n",
    "if len(files)%LoopFiles != 0:\n",
    "    loops += 1\n",
    "\n",
    "st = ti()\n",
    "\n",
    "Moves = []\n",
    "\n",
    "for k in range(loops):\n",
    "    if k == loops -1:\n",
    "        tfiles = BeforeTamping[k*LoopFiles:]\n",
    "    else:\n",
    "        tfiles = BeforeTamping[k*LoopFiles:(k+1)*LoopFiles]\n",
    "    #Results = Parallel(n_jobs=LoopFiles)(delayed(DeviationVelocity)(file) for file in tfiles)\n",
    "    Results = Parallel(n_jobs=LoopFiles)(delayed(SmoothMoves)(file) for file in tfiles)\n",
    "    #Results =[]\n",
    "    #for file in tfiles:\n",
    "    #    Results.append(DeviationVelocity(file))\n",
    "    #    print(file, (ti()-st)/60.0)\n",
    "    for result in Results:\n",
    "        Moves.append(result)\n",
    "    print(k, (ti()-st)/60.0)\n",
    "        \n",
    "Sequences = []\n",
    "Outputs = []\n",
    "for move in Moves:\n",
    "    Seq, Out = split_sequences(move.T,TimeSteps,StepSize)\n",
    "    Sequences.append(Seq)\n",
    "    Outputs.append(Out)\n",
    "print('Sequence Made ', (ti()-st)/60.0)\n",
    "MoveSegments = []\n",
    "for seq in Sequences:\n",
    "    for mv in seq:\n",
    "        MoveSegments.append(mv)\n",
    "NextDataPoint = []\n",
    "print('Moves Segmented ', (ti()-st)/60.0)\n",
    "for out in Outputs:\n",
    "    for pt in out:\n",
    "        NextDataPoint.append(np.reshape(pt,(1,3)))\n",
    "print('Done Prep ',(ti()-st)/60.0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32/32 [==============================] - 4s 111ms/step\n"
     ]
    }
   ],
   "source": [
    "Check1 = np.asarray(MoveSegments[:1000])\n",
    "\n",
    "output_X = lstm_autoencoder2.model.predict(Check1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "scores = np.linalg.norm(Check1 - output_X, axis=-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "maxes = np.max(scores, axis = 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1000,)"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.shape(maxes)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.title('random Comparison')\n",
    "#plt.plot(MoveSegments[1][:,0])\n",
    "plt.plot(output_X[99])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "threshold = 0.95\n",
    "\n",
    "sorted_scores = sorted(maxes)\n",
    "threshold_score = sorted_scores[round(len(maxes) * threshold)]\n",
    "\n",
    "plt.title(\"Reconstruction Error\")\n",
    "plt.plot(maxes)\n",
    "plt.plot([threshold_score]*len(scores), c='r')\n",
    "plt.show()\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.4"
  },
  "vscode": {
   "interpreter": {
    "hash": "178f6c3502586c94dc93af50f98dbd15c5205250cbf2345a6eb57380f8c77d96"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}