23. TensorBoard with MNIST.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h2 style=\"color:blue\" align=\"center\">Tensorboard demo using handwritten digits classification using neural network</h2>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In this notebook we will classify handwritten digits using a simple neural network which has only input and output layers. We will than add a hidden layer and see how the performance of the model improves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "(X_train, y_train) , (X_test, y_test) = keras.datasets.mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [],
   "source": [
    "One Hot Encoding\n",
    "\n",
    "y_train_categorical = keras.utils.to_categorical(\n",
    "    y_train, num_classes=10, dtype='float32'\n",
    ")\n",
    "y_test_categorical = keras.utils.to_categorical(\n",
    "    y_test, num_classes=10, dtype='float32'\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5\n",
      "[0. 0. 0. 0. 0. 1. 0. 0. 0. 0.]\n"
     ]
    }
   ],
   "source": [
    "print(y_train[0])\n",
    "print(y_train_categorical[0])  # gets 1 at 5th index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "60000"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10000"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28, 28)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7ffc88852f50>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 288x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.matshow(X_train[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "5"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train = X_train / 255\n",
    "X_test = X_test / 255"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_flattened = X_train.reshape(len(X_train), 28*28)\n",
    "X_test_flattened = X_test.reshape(len(X_test), 28*28)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 784)"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_flattened.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h3 style='color:purple'>Using Flatten layer so that we don't have to call .reshape on input dataset</h3>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "   1/1875 [..............................] - ETA: 0s - loss: 2.3442 - accuracy: 0.0625WARNING:tensorflow:Callbacks method `on_train_batch_end` is slow compared to the batch time (batch time: 0.0019s vs `on_train_batch_end` time: 0.0313s). Check your callbacks.\n",
      "1875/1875 [==============================] - 3s 2ms/step - loss: 0.2996 - accuracy: 0.9169\n",
      "Epoch 2/5\n",
      "1875/1875 [==============================] - 3s 2ms/step - loss: 0.1385 - accuracy: 0.9595\n",
      "Epoch 3/5\n",
      "1875/1875 [==============================] - 3s 2ms/step - loss: 0.0988 - accuracy: 0.9701\n",
      "Epoch 4/5\n",
      "1875/1875 [==============================] - 3s 2ms/step - loss: 0.0774 - accuracy: 0.9765\n",
      "Epoch 5/5\n",
      "1875/1875 [==============================] - 3s 2ms/step - loss: 0.0625 - accuracy: 0.9807\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.keras.callbacks.History at 0x7ffc8501a990>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = keras.Sequential([\n",
    "    keras.layers.Flatten(input_shape=(28, 28)),  #flattens X_train to X_train_flattened\n",
    "    \n",
    "    #hidden layer,100 neurons taken randomly,more hidden layers can be added to improve the performance\n",
    "    keras.layers.Dense(100, activation='relu'), #relu is preferred in hidden layers as relu solves vanishing gradient problem\n",
    "    \n",
    "    keras.layers.Dense(10, activation='sigmoid')#output layer, sigmoid is preferred when categorical data(10 categories)\n",
    "])\n",
    "\n",
    "#passing the parameters\n",
    "model.compile(optimizer='adam',\n",
    "              loss='categorical_crossentropy',   # sparse_categorical_crossentropy is used when y_train is not OHE\n",
    "              metrics=['accuracy'])\n",
    "\n",
    "# tb_callback = tf.keras.callbacks.TensorBoard(log_dir=\"logs/adam\", histogram_freq=1)\n",
    "\n",
    "model.fit(X_train, y_train_categorical, epochs=5, callbacks=[tb_callback])  #training the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %load_ext tensorboard\n",
    "# %tensorboard --logdir logs/fit"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Either run above inline magic or go to git bash or your command prompt and type below to run it,\n",
    "`tensorboard --logdir logs/`\n",
    "This will launch tensorboard at this url which you can open in your browser `http://localhost:6006/`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[array([[-0.00872515, -0.01911347, -0.06075284, ...,  0.02473276,\n",
       "         -0.0165142 ,  0.06590307],\n",
       "        [ 0.02018517,  0.01704864,  0.00618984, ..., -0.00770196,\n",
       "         -0.0444339 ,  0.03491807],\n",
       "        [-0.07492869,  0.06215216,  0.02764995, ..., -0.03643103,\n",
       "          0.07159407,  0.01379928],\n",
       "        ...,\n",
       "        [-0.03609845, -0.03093682, -0.02776644, ..., -0.01060833,\n",
       "          0.00518923, -0.0653451 ],\n",
       "        [-0.05540777,  0.04862949, -0.00707511, ...,  0.01807348,\n",
       "         -0.05405346, -0.04024128],\n",
       "        [-0.05392178, -0.06829732, -0.06474487, ..., -0.02665622,\n",
       "          0.03091565,  0.04448262]], dtype=float32),\n",
       " array([-0.02434695,  0.07552375,  0.23955728,  0.0294112 ,  0.10313676,\n",
       "         0.00446706,  0.11074002,  0.07696004,  0.10570896, -0.145429  ,\n",
       "         0.07711296, -0.02282092,  0.07459939,  0.27373767,  0.01669248,\n",
       "         0.01444363,  0.08700529, -0.06301918,  0.1501835 ,  0.1436217 ,\n",
       "        -0.13036308,  0.03680253,  0.04666013,  0.07432528, -0.09189091,\n",
       "         0.44227827, -0.01473343,  0.07111656,  0.11265518,  0.06117533,\n",
       "         0.21018846,  0.03004991,  0.03869489, -0.00429339, -0.070168  ,\n",
       "         0.08030853, -0.05706772, -0.00727534,  0.17769627,  0.08038036,\n",
       "         0.01594802, -0.00548568,  0.04917543, -0.07872794,  0.11835397,\n",
       "        -0.08037189,  0.0967322 ,  0.198713  ,  0.05280869,  0.04824052,\n",
       "         0.16582526,  0.00549677,  0.11592622,  0.11706494, -0.02461421,\n",
       "         0.21529578,  0.11522578,  0.16167139,  0.25778717, -0.01801421,\n",
       "         0.16552101,  0.23732677, -0.02893098,  0.07191815, -0.02940179,\n",
       "         0.0280506 , -0.05452613, -0.06914482,  0.13241787,  0.11693859,\n",
       "         0.00342979, -0.03609999, -0.05111558,  0.08289415,  0.03466756,\n",
       "         0.03315337, -0.11148469, -0.07293869,  0.12203182,  0.14095995,\n",
       "        -0.08869061,  0.27878407,  0.06623887,  0.06685624, -0.00087007,\n",
       "         0.3106244 ,  0.02708532, -0.05024546,  0.17183167,  0.09759334,\n",
       "         0.15050109,  0.28575262, -0.03875093,  0.11956069,  0.21600015,\n",
       "        -0.18693554,  0.17672123, -0.01797042,  0.07861724, -0.00376326],\n",
       "       dtype=float32),\n",
       " array([[-1.45721659e-01,  4.81203437e-01, -7.26448953e-01,\n",
       "         -1.08553983e-01,  6.46308482e-01,  3.24098945e-01,\n",
       "          6.60456810e-03,  6.97240466e-03, -1.89466000e-01,\n",
       "         -7.77451992e-01],\n",
       "        [-4.12337273e-01,  5.83901815e-02, -7.61386156e-01,\n",
       "         -1.66073740e-01,  2.62243301e-01,  2.02351376e-01,\n",
       "         -3.03832982e-02, -1.93213642e-01,  1.83046073e-01,\n",
       "          1.26776204e-01],\n",
       "        [ 2.39079781e-02,  6.53833687e-01, -1.60400406e-01,\n",
       "         -4.13718104e-01,  3.95977676e-01,  3.38028111e-02,\n",
       "          2.52352983e-01,  2.64256448e-01, -5.58848143e-01,\n",
       "         -8.41097414e-01],\n",
       "        [ 1.91844776e-01, -4.59698468e-01, -9.71690357e-01,\n",
       "         -2.13161752e-01,  3.37353766e-01,  1.97477967e-01,\n",
       "          2.49934733e-01,  3.86686862e-01, -5.73768690e-02,\n",
       "         -2.96300799e-01],\n",
       "        [-5.25431633e-01, -5.16507402e-02,  3.75663757e-01,\n",
       "         -1.24579243e-01, -5.20010650e-01, -9.73992795e-02,\n",
       "         -5.63561976e-01,  2.94429123e-01,  9.20057204e-03,\n",
       "         -6.13716170e-02],\n",
       "        [-1.19602248e-01, -3.13205659e-01, -2.91440338e-02,\n",
       "          1.44189849e-01, -4.83338237e-01,  5.46972919e-03,\n",
       "         -4.57768530e-01,  4.24561948e-02, -3.32764052e-02,\n",
       "         -1.14571759e-02],\n",
       "        [-2.10435346e-01, -2.93349057e-01, -2.31847659e-01,\n",
       "          6.51099458e-02, -1.54802144e-01, -4.78590466e-02,\n",
       "         -1.36710107e-01, -5.36566833e-03, -1.62686571e-01,\n",
       "          6.88992888e-02],\n",
       "        [-3.93557340e-01, -5.68417549e-01, -8.40301737e-02,\n",
       "         -6.16332948e-01,  1.21847153e-01,  1.88671082e-01,\n",
       "         -6.21370487e-02,  1.57354474e-01,  1.19742729e-01,\n",
       "         -1.66831866e-01],\n",
       "        [ 1.43786207e-01, -4.22650844e-01, -2.87091374e-01,\n",
       "         -1.77626356e-01, -2.04025671e-01,  4.78448234e-02,\n",
       "         -1.55383289e-01,  4.74186316e-02, -1.32590741e-01,\n",
       "          7.09912702e-02],\n",
       "        [ 1.52177319e-01,  6.11414537e-02,  2.41536051e-01,\n",
       "          1.04491137e-01, -7.27564156e-01, -5.95282018e-02,\n",
       "          1.14016317e-01, -2.86752611e-01,  2.05240026e-01,\n",
       "         -1.97223186e-01],\n",
       "        [-5.01274705e-01, -1.16732009e-01, -9.78402644e-02,\n",
       "          2.04424277e-01,  2.03374252e-01, -4.91920374e-02,\n",
       "         -7.43849456e-01,  1.94564104e-01, -2.53577769e-01,\n",
       "          6.40995661e-03],\n",
       "        [-3.34418528e-02,  5.33272997e-02,  1.43552676e-01,\n",
       "          3.55207235e-01,  2.60055870e-01, -2.20541924e-01,\n",
       "          2.32490405e-01, -2.16118917e-01, -5.92638701e-02,\n",
       "         -8.68470609e-01],\n",
       "        [ 2.28330001e-01,  4.55406606e-01, -3.79949868e-01,\n",
       "         -1.18393295e-01,  1.36190534e-01, -2.20489819e-02,\n",
       "          6.91883452e-03,  2.97688544e-01, -6.13576770e-01,\n",
       "         -2.94373125e-01],\n",
       "        [-4.12321657e-01, -1.11007271e-03,  1.66631430e-01,\n",
       "         -1.21556461e-01, -5.93070507e-01,  3.22169840e-01,\n",
       "         -4.24158216e-01,  6.58671379e-01, -2.04366520e-01,\n",
       "         -4.70197439e-01],\n",
       "        [ 2.23620951e-01,  3.41163993e-01, -3.57920080e-01,\n",
       "         -1.97686493e-01, -2.36802176e-01, -1.22411050e-01,\n",
       "         -5.04433155e-01,  5.22786155e-02, -1.26606375e-02,\n",
       "          1.78072304e-01],\n",
       "        [-2.18039781e-01,  4.32951897e-01,  1.53397337e-01,\n",
       "          1.89843252e-01,  7.15034828e-02, -1.18347276e-02,\n",
       "         -6.55562580e-01, -4.97247338e-01,  4.10087630e-02,\n",
       "         -4.46325988e-02],\n",
       "        [ 1.52243495e-01, -3.96461904e-01, -1.32878155e-01,\n",
       "         -1.81671277e-01, -8.42456222e-02,  8.06663632e-02,\n",
       "         -1.25733361e-01,  3.27105932e-02, -1.02993406e-01,\n",
       "          6.30683228e-02],\n",
       "        [-3.29326957e-01,  4.69766706e-02,  1.27743974e-01,\n",
       "          2.05179960e-01, -3.84219974e-01, -4.09936965e-01,\n",
       "          1.17906757e-01,  3.54921073e-01, -6.63355216e-02,\n",
       "         -9.58555877e-01],\n",
       "        [ 1.55007347e-01, -3.03142279e-01, -2.67698228e-01,\n",
       "         -1.77458465e-01, -2.96830684e-02, -1.10358596e-01,\n",
       "         -3.63105815e-03, -2.78831214e-01, -2.22356737e-01,\n",
       "         -2.50461191e-01],\n",
       "        [-2.01555222e-01,  4.60888185e-02, -4.68575269e-01,\n",
       "         -8.94619703e-01,  4.22185250e-02,  2.24032953e-01,\n",
       "          4.75768566e-01, -2.97590494e-02,  3.00990772e-02,\n",
       "         -1.20311771e-02],\n",
       "        [ 1.42167285e-01, -3.67272884e-01,  6.49618804e-02,\n",
       "          1.34809092e-01,  6.22653440e-02, -2.22200722e-01,\n",
       "         -4.59804922e-01, -1.40595526e-01,  7.03247562e-02,\n",
       "          7.48820677e-02],\n",
       "        [ 1.62979037e-01, -4.93050992e-01, -7.73924530e-01,\n",
       "          1.44846693e-01, -1.87065043e-02,  3.93643498e-01,\n",
       "          1.32795423e-01, -3.93750370e-01,  1.58079788e-01,\n",
       "         -2.23031733e-03],\n",
       "        [-2.82466054e-01,  2.15739459e-02,  3.79945099e-01,\n",
       "         -4.71839845e-01,  3.01358737e-02, -1.98079437e-01,\n",
       "          1.23476587e-01, -3.39397401e-01,  9.66830701e-02,\n",
       "          1.90628335e-01],\n",
       "        [-1.60759270e-01, -1.34745955e-01,  1.87634323e-02,\n",
       "         -2.26533748e-02, -4.28148806e-01, -3.56210163e-04,\n",
       "         -4.76627052e-01,  1.32259741e-01, -1.09068595e-01,\n",
       "         -1.25343725e-01],\n",
       "        [ 3.91184539e-01, -7.36827731e-01,  9.01401788e-02,\n",
       "         -4.56972718e-01,  1.68986216e-01, -2.46491492e-01,\n",
       "          6.99133649e-02,  3.32506858e-02, -4.15302575e-01,\n",
       "          3.73289436e-01],\n",
       "        [-1.77850768e-01, -7.36575499e-02, -2.03159168e-01,\n",
       "         -1.91493839e-01, -4.36496586e-02, -1.09921522e-01,\n",
       "         -2.30765790e-01, -2.36767575e-01, -1.16535284e-01,\n",
       "         -1.73109904e-01],\n",
       "        [-2.14446634e-01, -2.04438567e-01, -1.30970776e-01,\n",
       "         -1.09126098e-01,  8.04581046e-02, -2.00119853e-01,\n",
       "         -2.93146092e-02, -2.06115931e-01, -1.67442873e-01,\n",
       "         -4.08188254e-02],\n",
       "        [-8.46934319e-02, -2.33889014e-01,  2.04560086e-01,\n",
       "         -5.39309561e-01,  1.50693819e-01, -2.91870177e-01,\n",
       "         -1.30271986e-01, -9.00029391e-02,  5.24695158e-01,\n",
       "         -6.89404309e-01],\n",
       "        [-4.87534016e-01,  6.85609728e-02, -8.97774637e-01,\n",
       "          3.85565981e-02,  1.82125121e-01, -2.24467535e-02,\n",
       "         -5.51422298e-01, -1.41089503e-02, -6.87489688e-01,\n",
       "          4.32351708e-01],\n",
       "        [-4.32543606e-01, -4.90153246e-02,  9.71119106e-02,\n",
       "          1.50285214e-01, -1.37018010e-01, -6.64813638e-01,\n",
       "          3.54636669e-01, -1.18632659e-01, -1.17674798e-01,\n",
       "          2.87664086e-02],\n",
       "        [-1.91066265e-01, -5.54003976e-02, -1.27384081e-01,\n",
       "         -2.36670807e-01, -7.95876607e-03, -2.55290627e-01,\n",
       "          1.00607760e-01, -2.06573486e-01, -1.11041978e-01,\n",
       "         -2.22140417e-01],\n",
       "        [ 3.32058847e-01, -6.90744638e-01,  3.38538587e-01,\n",
       "         -3.48395646e-01, -1.75245758e-02, -1.77719504e-01,\n",
       "         -5.72029054e-02, -1.51907355e-01, -1.81313515e-01,\n",
       "          2.45084669e-02],\n",
       "        [-6.88678861e-01, -1.99139312e-01, -2.07873523e-01,\n",
       "          6.65866770e-03,  2.35140324e-01,  9.95428488e-02,\n",
       "         -3.65453988e-01, -8.99138525e-02,  6.76915199e-02,\n",
       "          1.46743402e-01],\n",
       "        [-5.96171260e-01, -7.34646380e-01, -3.19439054e-01,\n",
       "          3.06677222e-01, -1.11758120e-01,  3.08890998e-01,\n",
       "         -5.29184163e-01, -1.53431237e-01, -7.62054250e-02,\n",
       "          1.26858294e-01],\n",
       "        [-1.78773448e-01, -2.29344651e-01,  1.88937649e-01,\n",
       "          1.97489798e-01, -2.02253964e-02, -6.91229478e-02,\n",
       "          1.89138040e-01, -4.91292655e-01,  2.41731182e-01,\n",
       "         -8.63151699e-02],\n",
       "        [ 4.47035208e-02, -6.30875111e-01,  1.56245470e-01,\n",
       "          6.26312122e-02, -2.04015002e-01,  1.11975826e-01,\n",
       "         -1.73769429e-01,  2.39476889e-01, -3.41047704e-01,\n",
       "          7.04466850e-02],\n",
       "        [-7.98671961e-01, -4.89166863e-02, -9.33112055e-02,\n",
       "          1.28059402e-01,  3.49024475e-01,  1.70914963e-01,\n",
       "         -5.04617572e-01, -5.16509891e-01,  2.05355376e-01,\n",
       "          2.16234848e-01],\n",
       "        [ 2.92376995e-01, -3.06940883e-01, -3.40389431e-01,\n",
       "         -3.90546434e-02,  1.78004280e-02, -5.45818090e-01,\n",
       "          9.66533422e-02,  1.39775470e-01, -2.92636693e-01,\n",
       "          2.30001688e-01],\n",
       "        [-1.37609601e-01, -6.58947304e-02,  5.93402097e-03,\n",
       "         -8.80528092e-02, -8.88909623e-02,  1.02540655e-02,\n",
       "         -1.78109929e-02, -3.30429882e-01, -1.91484556e-01,\n",
       "         -1.19467944e-01],\n",
       "        [-1.45465553e-01, -2.94123888e-01,  4.66538817e-02,\n",
       "          3.51336300e-02, -2.07787916e-01, -1.86201990e-01,\n",
       "         -2.92363822e-01, -1.54705361e-01,  1.21500857e-01,\n",
       "         -9.99769270e-02],\n",
       "        [ 9.26411450e-02, -1.91406593e-01, -1.80161417e-01,\n",
       "         -8.44332427e-02, -2.73084700e-01, -1.51551753e-01,\n",
       "         -1.15538687e-01, -4.07473668e-02, -3.73309627e-02,\n",
       "          3.62592302e-02],\n",
       "        [ 2.30629057e-01,  3.10864687e-01, -4.57485825e-01,\n",
       "          2.83030629e-01, -6.31602526e-01,  4.85067368e-01,\n",
       "          1.64668500e-01, -2.38763571e-01, -4.52534854e-01,\n",
       "         -2.22761095e-01],\n",
       "        [ 7.43011758e-02, -4.15961206e-01,  3.75284478e-02,\n",
       "         -3.36236984e-01,  4.04279888e-01, -3.76675501e-02,\n",
       "         -3.22008133e-01,  2.19720796e-01,  1.54605150e-01,\n",
       "         -2.85653502e-01],\n",
       "        [ 1.38510138e-01, -4.82967496e-02,  1.34186208e-01,\n",
       "          4.32283841e-02,  1.29679129e-01, -1.97834700e-01,\n",
       "          1.21244863e-01, -5.73664606e-01, -3.76046777e-01,\n",
       "         -1.75550476e-01],\n",
       "        [-5.33752859e-01,  3.93556803e-01, -2.11194843e-01,\n",
       "         -2.42318623e-02,  1.94822505e-01, -1.07691549e-01,\n",
       "         -3.77812594e-01,  2.78223902e-01, -5.86668074e-01,\n",
       "          1.17759526e-01],\n",
       "        [ 1.32826775e-01, -3.20645273e-01,  2.66913682e-01,\n",
       "         -8.08676630e-02, -8.36794004e-02,  1.80901542e-01,\n",
       "          1.87403128e-01, -5.91205537e-01,  1.32441744e-01,\n",
       "          6.96924105e-02],\n",
       "        [-1.78187937e-01, -1.72518551e-01, -3.42416942e-01,\n",
       "          9.21853706e-02, -2.18329996e-01, -1.15379691e-01,\n",
       "         -2.02103838e-01, -2.92637330e-02, -2.43905470e-01,\n",
       "          7.76209980e-02],\n",
       "        [-3.34782004e-01,  2.37665385e-01, -5.10011256e-01,\n",
       "         -3.15195441e-01, -2.66587315e-03,  2.99933225e-01,\n",
       "         -4.09897536e-01, -3.95061672e-01,  9.21401158e-02,\n",
       "          1.52177766e-01],\n",
       "        [ 1.13695916e-02,  1.27124086e-01, -8.19725469e-02,\n",
       "         -8.14705193e-02, -2.80466080e-01,  2.71087680e-02,\n",
       "          6.21211603e-02, -4.56827611e-01,  5.71620651e-02,\n",
       "         -4.40789796e-02],\n",
       "        [-2.01268103e-02, -3.00515264e-01, -1.42715991e-01,\n",
       "         -2.11509123e-01, -7.07824230e-02, -1.37924686e-01,\n",
       "          2.47372910e-02, -1.20736793e-01,  8.72202665e-02,\n",
       "         -7.52605451e-03],\n",
       "        [-2.87027419e-01, -2.94352293e-01, -1.93176776e-01,\n",
       "         -1.68816522e-02,  1.61262169e-01,  9.43560377e-02,\n",
       "          6.02366701e-02, -3.34896982e-01, -4.62286621e-02,\n",
       "         -2.08579764e-01],\n",
       "        [-4.85087544e-01, -7.55070269e-01, -7.24734133e-03,\n",
       "         -3.97629291e-01, -4.89532873e-02,  4.34941351e-01,\n",
       "         -3.25895250e-01, -5.01338780e-01,  1.16563812e-02,\n",
       "          3.09011012e-01],\n",
       "        [ 3.45659815e-02,  1.85793594e-01,  1.44717827e-01,\n",
       "          7.54906703e-03, -1.06804109e+00,  9.00886133e-02,\n",
       "         -1.55656129e-01, -2.00420283e-02, -9.41834748e-02,\n",
       "         -2.79836446e-01],\n",
       "        [-6.00466967e-01,  2.37365037e-01,  2.21911147e-01,\n",
       "         -6.32786095e-01,  4.23379652e-02, -2.16683466e-02,\n",
       "          1.94478914e-01,  1.38095573e-01,  7.44551271e-02,\n",
       "         -5.66604674e-01],\n",
       "        [ 2.46090442e-01, -8.97646248e-02, -2.98867911e-01,\n",
       "          1.82409838e-01, -4.68216956e-01,  4.08460796e-02,\n",
       "          8.95867273e-02, -3.84669065e-01,  1.73730422e-02,\n",
       "          2.43489374e-03],\n",
       "        [-2.77318388e-01,  1.21363938e-01,  9.98586565e-02,\n",
       "         -2.30842352e-01, -3.23237091e-01, -2.44745597e-01,\n",
       "         -6.89339191e-02, -1.89488307e-01, -5.79817258e-02,\n",
       "         -2.52778649e-01],\n",
       "        [-4.65563275e-02, -4.03692782e-01, -3.70631009e-01,\n",
       "         -1.87521949e-02, -3.66425030e-02,  1.51818573e-01,\n",
       "         -1.21133037e-01, -1.92056876e-02,  1.26665890e-01,\n",
       "         -9.92154181e-02],\n",
       "        [-1.59674287e-01,  2.24075839e-01, -1.02933399e-01,\n",
       "          3.99808884e-02, -3.78126532e-01, -6.16218299e-02,\n",
       "         -1.73224449e-01, -3.20034981e-01, -1.11606002e-01,\n",
       "         -2.57678777e-01],\n",
       "        [-2.05033734e-01,  1.16061062e-01, -4.15369868e-01,\n",
       "         -6.75903916e-01, -4.81253378e-02,  2.23523617e-01,\n",
       "          5.48850060e-01,  1.39578611e-01, -6.56813741e-01,\n",
       "         -3.54355097e-01],\n",
       "        [ 1.81263447e-01, -2.18258187e-01,  7.57295787e-02,\n",
       "          6.45859167e-02, -1.03961611e+00,  2.31785282e-01,\n",
       "          6.86484426e-02, -4.66248065e-01, -4.60919701e-02,\n",
       "          1.98722556e-02],\n",
       "        [-4.36283126e-02, -3.44061971e-01, -2.66301066e-01,\n",
       "         -1.36672854e-01,  1.14868969e-01, -3.49027105e-02,\n",
       "         -2.32123628e-01, -9.20189731e-03, -9.47810411e-02,\n",
       "          6.55034781e-02],\n",
       "        [-2.54081964e-01, -9.92847234e-02, -2.24874437e-01,\n",
       "         -1.74840927e-01, -3.48144844e-02, -1.38230726e-01,\n",
       "         -3.01080108e-01,  9.60784256e-02, -1.74251180e-02,\n",
       "         -5.03492132e-02],\n",
       "        [ 2.70415664e-01,  2.80530334e-01,  1.45623237e-01,\n",
       "         -3.70862693e-01,  1.24756962e-01, -5.95428109e-01,\n",
       "         -3.26305866e-01, -5.74362464e-02, -2.90583789e-01,\n",
       "          1.27142563e-01],\n",
       "        [-2.82609969e-01, -2.25781098e-01,  3.53948735e-02,\n",
       "         -3.90013983e-03, -1.88065007e-01,  5.23980260e-02,\n",
       "         -3.71574253e-01,  3.59230042e-02, -3.87510285e-02,\n",
       "         -1.28914118e-01],\n",
       "        [-3.81658316e-01, -1.38967615e-02, -5.49457014e-01,\n",
       "          2.72406638e-01,  3.46229002e-02,  4.02805537e-01,\n",
       "         -5.03698587e-01, -3.68371457e-01, -2.41412997e-01,\n",
       "          7.38188326e-02],\n",
       "        [-6.67602122e-01,  1.21702127e-01,  4.08558659e-02,\n",
       "          1.66344881e-01, -1.93098500e-01,  1.83596805e-01,\n",
       "         -4.36185092e-01, -3.64117265e-01, -2.45646849e-01,\n",
       "          5.54489568e-02],\n",
       "        [ 1.01128176e-01, -3.95501524e-01, -2.57945210e-02,\n",
       "         -2.83946376e-02,  1.06440879e-01, -1.51987746e-01,\n",
       "          1.06988810e-02, -2.03261107e-01,  1.67989865e-01,\n",
       "         -1.90557674e-01],\n",
       "        [-2.06071615e-01,  1.70857176e-01, -5.03097653e-01,\n",
       "          1.61024824e-01,  3.90254885e-01, -8.23532760e-01,\n",
       "         -2.07022026e-01, -1.04683489e-01,  4.55813222e-02,\n",
       "         -2.31293347e-02],\n",
       "        [-3.11920315e-01,  3.00232083e-01, -1.86881900e-01,\n",
       "         -1.96451455e-01, -2.27444008e-01,  1.33369714e-02,\n",
       "          1.82607308e-01, -2.74908543e-01, -1.32266894e-01,\n",
       "         -3.25464696e-01],\n",
       "        [-2.50273824e-01,  2.27059737e-01,  1.19776346e-01,\n",
       "         -6.01995230e-01,  2.42187202e-01, -4.20478553e-01,\n",
       "         -5.80821574e-01,  3.23594928e-01, -3.96703154e-01,\n",
       "         -1.64215922e-01],\n",
       "        [-2.39021450e-01,  9.53808054e-02,  1.79828927e-02,\n",
       "          1.51263028e-01,  1.72186926e-01, -1.05719173e+00,\n",
       "         -2.96476781e-01,  2.21323684e-01, -8.81061777e-02,\n",
       "          2.24142857e-02],\n",
       "        [-1.44865200e-01, -2.51099467e-01, -5.26798487e-01,\n",
       "         -3.88348967e-01,  1.39834493e-01,  2.80544251e-01,\n",
       "          3.62120837e-01, -3.96000743e-01,  8.62727016e-02,\n",
       "          7.31003731e-02],\n",
       "        [ 7.63144121e-02,  1.06048651e-01, -1.64402664e-01,\n",
       "         -1.31004006e-01, -2.32745707e-01, -2.22158611e-01,\n",
       "          8.62323493e-02, -1.62806600e-01,  1.81317627e-01,\n",
       "          5.72744347e-02],\n",
       "        [-3.26966830e-02, -4.96834576e-01, -1.88697502e-01,\n",
       "          9.71496180e-02, -8.66569858e-03,  1.39403582e-01,\n",
       "         -3.19213606e-02,  9.67338011e-02, -2.43319139e-01,\n",
       "         -2.57124305e-01],\n",
       "        [-1.42017836e-02, -1.36870772e-01,  1.97335914e-01,\n",
       "         -7.01873153e-02, -8.52325141e-01,  2.60103375e-01,\n",
       "          1.86405852e-01, -5.66462874e-01, -3.02175619e-02,\n",
       "          2.73750663e-01],\n",
       "        [-9.40471366e-02,  2.89967172e-02,  1.93851486e-01,\n",
       "         -9.34663832e-01,  3.41721475e-01, -3.97860825e-01,\n",
       "         -1.28677711e-01,  2.44133994e-01, -2.07397625e-01,\n",
       "          2.56583214e-01],\n",
       "        [ 5.18546477e-02,  1.05953120e-01,  6.93604052e-02,\n",
       "          5.36608934e-01,  8.70626420e-02, -1.41307697e-01,\n",
       "          1.80607930e-01,  5.98027050e-01, -8.62323821e-01,\n",
       "         -3.70193422e-01],\n",
       "        [ 1.74224842e-02, -1.20442763e-01, -2.10546002e-01,\n",
       "          9.47663635e-02, -1.99466556e-01, -5.59285246e-02,\n",
       "          6.82233870e-02, -1.50155500e-01,  6.41188622e-02,\n",
       "         -7.74440393e-02],\n",
       "        [-7.15737045e-02, -9.66091454e-02, -1.57081127e-01,\n",
       "         -2.09115565e-01,  6.54462725e-02, -2.16390833e-01,\n",
       "          1.16957314e-01, -1.44821703e-01, -1.02592684e-01,\n",
       "         -2.16073036e-01],\n",
       "        [-2.27551088e-02, -7.11451620e-02, -4.83245105e-01,\n",
       "         -2.60263234e-01, -1.99054822e-01,  4.85058337e-01,\n",
       "          5.21596186e-02, -2.50589013e-01, -4.66281831e-01,\n",
       "          4.70710665e-01],\n",
       "        [ 2.77319700e-01, -2.52727866e-01,  6.66950494e-02,\n",
       "         -5.92886209e-01,  1.27722010e-01, -9.69065353e-02,\n",
       "          2.44026586e-01, -3.30839232e-02, -2.77430654e-01,\n",
       "         -8.98008514e-03],\n",
       "        [-2.47671023e-01, -3.38929623e-01, -1.83313563e-01,\n",
       "         -2.15281576e-01, -2.44617656e-01, -1.81850150e-01,\n",
       "         -1.41873196e-01, -2.75784940e-01,  1.07030526e-01,\n",
       "         -2.11149752e-01],\n",
       "        [ 2.35798702e-01, -3.15361500e-01,  1.18075617e-01,\n",
       "          9.93862152e-02, -8.86504769e-01,  6.60721511e-02,\n",
       "         -1.64393857e-01,  2.03079700e-01, -2.49005884e-01,\n",
       "         -1.40562981e-01],\n",
       "        [ 4.81293648e-01,  1.32903248e-01,  2.80249953e-01,\n",
       "          2.91665465e-01, -6.92318499e-01, -8.19171786e-01,\n",
       "          6.59203455e-02,  1.50623605e-01, -1.39368907e-01,\n",
       "          2.33921155e-01],\n",
       "        [-3.46805230e-02, -7.42556378e-02,  2.13821176e-02,\n",
       "          1.16183236e-01, -3.62693012e-01,  7.41241127e-02,\n",
       "         -6.37330592e-01, -5.49554937e-02, -5.37840370e-03,\n",
       "         -1.23424761e-01],\n",
       "        [-2.78830767e-01, -1.14062130e-01, -6.95910379e-02,\n",
       "         -1.80887654e-01, -3.13673824e-01, -5.38281538e-02,\n",
       "         -1.64445519e-01, -4.39209305e-02, -5.11513576e-02,\n",
       "         -2.19088957e-01],\n",
       "        [-1.83082730e-01, -1.12439923e-01,  7.20223188e-02,\n",
       "         -1.64412335e-02, -3.48522872e-01, -3.72952074e-01,\n",
       "         -1.92721814e-01, -4.89139706e-02, -9.71544906e-02,\n",
       "          9.24774855e-02],\n",
       "        [ 3.47451866e-01, -6.90251663e-02,  1.28038317e-01,\n",
       "         -3.23480546e-01,  1.79866418e-01, -6.50449395e-01,\n",
       "         -5.18214047e-01,  2.16686860e-01,  7.76327308e-03,\n",
       "          2.50702128e-02],\n",
       "        [-1.31315231e-01, -1.70939177e-01, -2.45442554e-01,\n",
       "         -2.10988875e-02, -1.37255371e-01, -2.66254276e-01,\n",
       "         -3.16713691e-01,  5.14764376e-02, -4.78517786e-02,\n",
       "         -1.49704561e-01],\n",
       "        [-5.84859848e-01, -3.50267142e-01,  1.97159663e-01,\n",
       "          1.30020559e-01,  1.15801476e-01,  1.49986088e-01,\n",
       "         -2.61830568e-01, -5.28262734e-01, -4.39257264e-01,\n",
       "          3.87488127e-01],\n",
       "        [-2.07585260e-01,  9.91844907e-02, -2.32877776e-01,\n",
       "         -1.69330370e-02, -1.37815014e-01, -2.50244290e-01,\n",
       "         -1.17849015e-01, -1.60106316e-01, -1.59931764e-01,\n",
       "         -1.18512288e-01],\n",
       "        [-1.02518015e-01, -2.15043426e-01,  3.16012144e-01,\n",
       "         -1.83080614e-01, -1.11022182e-01,  4.18989956e-01,\n",
       "         -5.17493248e-01,  8.55849534e-02, -5.50286889e-01,\n",
       "         -3.77548248e-01],\n",
       "        [ 9.42001715e-02,  2.42516696e-01,  1.30895376e-01,\n",
       "          3.21640112e-02,  1.76098600e-01, -7.46999830e-02,\n",
       "         -2.33719990e-01, -3.30900431e-01,  9.26336944e-02,\n",
       "         -3.39836240e-01],\n",
       "        [-2.34062105e-01, -4.71575141e-01,  4.74519312e-01,\n",
       "         -2.40268096e-01,  3.02356303e-01,  5.52928001e-02,\n",
       "          1.58173636e-01, -8.15808326e-02, -4.58252192e-01,\n",
       "         -5.32823980e-01],\n",
       "        [-2.78816074e-01, -1.85502261e-01, -2.12606177e-01,\n",
       "         -1.63957492e-01, -1.43875718e-01, -2.09919840e-01,\n",
       "         -4.14440483e-02, -1.70275375e-01, -2.01799512e-01,\n",
       "          7.85221905e-02],\n",
       "        [ 2.96663105e-01,  9.54923183e-02,  3.92839432e-01,\n",
       "          4.82474379e-02, -1.46591321e-01, -7.59050310e-01,\n",
       "         -3.99989158e-01,  2.78145447e-02, -5.97593524e-02,\n",
       "          4.80649360e-02],\n",
       "        [-5.43429442e-02,  7.17876911e-01,  2.51223117e-01,\n",
       "          8.74783024e-02,  3.54683679e-03,  3.93863246e-02,\n",
       "         -2.09510595e-01,  1.48458749e-01, -6.68168426e-01,\n",
       "         -6.04870677e-01],\n",
       "        [-6.17851689e-03, -2.20689893e-01,  3.20263267e-01,\n",
       "         -1.14362240e-01, -4.25005816e-02,  1.85051769e-01,\n",
       "         -5.33435633e-03,  2.11527288e-01, -1.17414758e-01,\n",
       "         -7.79547930e-01],\n",
       "        [-2.89978236e-01,  9.68756713e-03,  6.26246929e-02,\n",
       "          3.38514186e-02, -3.46015751e-01, -1.96907789e-01,\n",
       "         -2.09161282e-01,  8.97668228e-02, -2.93042928e-01,\n",
       "         -3.59893292e-01],\n",
       "        [-1.44927979e-01, -3.10385078e-01,  1.39979403e-02,\n",
       "         -3.85810845e-02,  5.93530387e-02, -1.68994248e-01,\n",
       "         -3.96585047e-01, -2.08433166e-01,  5.89027330e-02,\n",
       "          1.76622748e-01]], dtype=float32),\n",
       " array([-2.7621570e-01, -5.6905065e-02, -8.3694488e-02, -1.4356434e-01,\n",
       "        -8.5141376e-02,  1.4569217e-04, -1.2589940e-01, -8.2191937e-02,\n",
       "        -9.4265327e-02, -1.2092391e-01], dtype=float32)]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_weights()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}