Artificial_Neural_Networks_with_Keras.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Na8EnUFiwf-A"
   },
   "source": [
    "<a href=\"https://colab.research.google.com/github/Shrey-Viradiya/HandsOnMachineLearning/blob/master/Artificial_Neural_Networks_with_Keras.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WkCq751Bwf-D"
   },
   "source": [
    "# Artificial Neural Networks with Keras"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "uq4VtSDPwf-E"
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import tensorflow as tf\n",
    "from tensorflow import keras\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "C-BCZbgIwf-M"
   },
   "source": [
    "Let's start by loading the fashion MNIST dataset. Keras has a number of functions to load popular datasets in keras.datasets. The dataset is already split for you between a training set and a test set, but it can be useful to split the training set further to have a validation set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Fu-B_muPwf-N"
   },
   "outputs": [],
   "source": [
    "fashion_mnist = keras.datasets.fashion_mnist"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 157
    },
    "colab_type": "code",
    "id": "JVzuNIsKwf-a",
    "outputId": "ef464090-ee50-409d-d2b4-0035cd34f3e2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-labels-idx1-ubyte.gz\n",
      "32768/29515 [=================================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/train-images-idx3-ubyte.gz\n",
      "26427392/26421880 [==============================] - 1s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-labels-idx1-ubyte.gz\n",
      "8192/5148 [===============================================] - 0s 0us/step\n",
      "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/t10k-images-idx3-ubyte.gz\n",
      "4423680/4422102 [==============================] - 0s 0us/step\n"
     ]
    }
   ],
   "source": [
    "(X_train_full, y_train_full), (X_test, y_test) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "g9J70Cgvwf-h",
    "outputId": "558d0b2f-9103-4133-a290-a6dc3f296c91"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28)"
      ]
     },
     "execution_count": 4,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_full.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "FjrvMlfXwf-q",
    "outputId": "47ca83a0-c183-49bc-b974-da8e28a2253d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dtype('uint8')"
      ]
     },
     "execution_count": 5,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train_full.dtype"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ssmkiGVNwf-v"
   },
   "outputs": [],
   "source": [
    "X_valid, X_train = X_train_full[:5000] / 255., X_train_full[5000:] / 255.\n",
    "y_valid, y_train = y_train_full[:5000], y_train_full[5000:]\n",
    "X_test = X_test / 255."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 247
    },
    "colab_type": "code",
    "id": "K2Zh8fVYwf-3",
    "outputId": "bb676875-5573-4ee0-849b-afb59f4a0d04"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(X_train[0], cmap=\"binary\")\n",
    "plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "o81-RUKjwf--",
    "outputId": "284c7a06-d575-4f6e-bfe3-b6e573859234"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "4"
      ]
     },
     "execution_count": 8,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "LI9R1Ifgwf_F",
    "outputId": "07f8699f-fe9c-401d-8c5b-269c28aaf26e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4, 0, 7, ..., 3, 0, 5], dtype=uint8)"
      ]
     },
     "execution_count": 9,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "pExzkI2-wf_L"
   },
   "outputs": [],
   "source": [
    "class_names = [\"T-shirt/top\", \"Trouser\", \"Pullover\", \"Dress\", \"Coat\",\n",
    "               \"Sandal\", \"Shirt\", \"Sneaker\", \"Bag\", \"Ankle boot\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "_FgBnFs6wf_P",
    "outputId": "7af72924-2ae5-4841-bb74-bc39d9219821"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Coat'"
      ]
     },
     "execution_count": 11,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class_names[y_train[0]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "30nB_-dHwf_W",
    "outputId": "3cd9dcd8-d642-4d6d-9cd7-ad43575e2453"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(5000, 28, 28)"
      ]
     },
     "execution_count": 12,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_valid.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "JNwkZ8dwwf_c",
    "outputId": "e517bd5f-6049-40a6-e2a8-26d7e12c9310"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 28, 28)"
      ]
     },
     "execution_count": 13,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_test.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "eEm_nk9hwf_h"
   },
   "source": [
    "Let's take a look at a sample of the images in the dataset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 441
    },
    "colab_type": "code",
    "id": "7wG2ii14wf_i",
    "outputId": "af279970-a4c0-434a-b208-41ae315d1294"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x432 with 40 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_rows = 4\n",
    "n_cols = 10\n",
    "\n",
    "plt.figure(figsize = (n_cols * 1.5, n_rows * 1.5))\n",
    "\n",
    "for row in range(n_rows):\n",
    "    for col in range(n_cols):\n",
    "        index = n_cols * row + col\n",
    "        plt.subplot(n_rows, n_cols, index + 1)\n",
    "        plt.imshow(X_train[index], cmap=\"binary\", interpolation=\"nearest\")\n",
    "        plt.axis('off')\n",
    "        plt.title(class_names[y_train[index]])\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4yUKSM42wf_o"
   },
   "source": [
    "## CREATING THE MODEL USING THE SEQUENTIAL API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3QkNDhXhwf_p"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential()\n",
    "model.add(keras.layers.Flatten(input_shape=[28, 28]))\n",
    "model.add(keras.layers.Dense(300, activation = 'relu'))\n",
    "model.add(keras.layers.Dense(100, activation = 'relu'))\n",
    "model.add(keras.layers.Dense(10, activation = 'softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "HjrdGnHwwf_v"
   },
   "outputs": [],
   "source": [
    "keras.backend.clear_session()\n",
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "t3k_1Mx8wf_2"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Flatten(input_shape=[28, 28]),\n",
    "    keras.layers.Dense(300, activation=\"relu\"),\n",
    "    keras.layers.Dense(100, activation=\"relu\"),\n",
    "    keras.layers.Dense(10, activation=\"softmax\")\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "XtaYcR2dwf_7",
    "outputId": "714d8ce4-cd17-4953-8a87-00e6f84e056a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "flatten (Flatten)            (None, 784)               0         \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 300)               235500    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 100)               30100     \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 10)                1010      \n",
      "=================================================================\n",
      "Total params: 266,610\n",
      "Trainable params: 266,610\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "wwnpxsRXwgAA",
    "outputId": "9d54e85b-b0b5-4a7d-eade-85d1d86f5a52"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<tensorflow.python.keras.layers.core.Flatten at 0x7f8de6beb908>,\n",
       " <tensorflow.python.keras.layers.core.Dense at 0x7f8de6bebac8>,\n",
       " <tensorflow.python.keras.layers.core.Dense at 0x7f8de6bebd30>,\n",
       " <tensorflow.python.keras.layers.core.Dense at 0x7f8de6bebf98>]"
      ]
     },
     "execution_count": 19,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "43AurhP2wgAE"
   },
   "outputs": [],
   "source": [
    "hidden1 = model.layers[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "sixQ3HxqwgAL",
    "outputId": "cf0c710a-86a3-4a81-afd6-05399133521d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'dense'"
      ]
     },
     "execution_count": 21,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hidden1.name"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "73YbdsBDwgAP",
    "outputId": "33ba1e68-7440-44a6-ce33-eaa6d8f0da85"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 22,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.get_layer('dense') is hidden1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "iHCKf6JOwgAT",
    "outputId": "4d9c4bd9-4ba5-49ed-f961-c7a185bde4ef"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 23,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keras.utils.plot_model(model, \"my_fashion_mnist_model.png\", show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "LX-ACMBpwgAZ"
   },
   "outputs": [],
   "source": [
    "weights, biases = hidden1.get_weights()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "yOCVomWKwgAe",
    "outputId": "3dfc05e0-8c9c-4432-f213-9e0583346f9c"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.00315184,  0.04630247, -0.00316229, ...,  0.01646411,\n",
       "        -0.02042289, -0.03946767],\n",
       "       [ 0.05136564, -0.02859101,  0.04602057, ...,  0.00757318,\n",
       "        -0.06618038, -0.05471007],\n",
       "       [ 0.04691272, -0.0026296 ,  0.07020119, ..., -0.04660381,\n",
       "        -0.03666015,  0.02501245],\n",
       "       ...,\n",
       "       [ 0.012565  ,  0.05742019, -0.06992737, ...,  0.04477017,\n",
       "         0.03706373,  0.05789621],\n",
       "       [-0.06575304,  0.07304707,  0.05090609, ...,  0.03085814,\n",
       "        -0.00268235, -0.05790088],\n",
       "       [ 0.02920057, -0.00429318,  0.01269243, ...,  0.01845723,\n",
       "        -0.05683304,  0.01923399]], dtype=float32)"
      ]
     },
     "execution_count": 25,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "3_DJr6xwwgAj",
    "outputId": "1cb32887-194b-463c-bcec-4c9b6b8341bb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(784, 300)"
      ]
     },
     "execution_count": 26,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weights.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "ZR3cjJG9wgAu",
    "outputId": "3dafc55f-e30b-4464-98ab-db2a4b5ca531"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,\n",
       "       0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], dtype=float32)"
      ]
     },
     "execution_count": 27,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "biases"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "YOdtfedowgAz",
    "outputId": "2de4abef-2118-4a61-ddc7-535f95591936"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(300,)"
      ]
     },
     "execution_count": 28,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "biases.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1Cr9IBz-wgA5"
   },
   "outputs": [],
   "source": [
    "model.compile(loss = 'sparse_categorical_crossentropy',\n",
    "             optimizer='sgd',\n",
    "             metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "l5wTAYf2wgA_"
   },
   "source": [
    "### TRAINING AND EVALUATING THE MODEL"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "e-PSgnsUwgBA",
    "outputId": "6534c1dd-3e5a-443e-e999-053ed86124b1"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.7215 - accuracy: 0.7639 - val_loss: 0.5147 - val_accuracy: 0.8252\n",
      "Epoch 2/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.4901 - accuracy: 0.8297 - val_loss: 0.5010 - val_accuracy: 0.8214\n",
      "Epoch 3/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.4452 - accuracy: 0.8431 - val_loss: 0.4206 - val_accuracy: 0.8534\n",
      "Epoch 4/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.4176 - accuracy: 0.8529 - val_loss: 0.4022 - val_accuracy: 0.8632\n",
      "Epoch 5/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3979 - accuracy: 0.8594 - val_loss: 0.3955 - val_accuracy: 0.8594\n",
      "Epoch 6/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3822 - accuracy: 0.8658 - val_loss: 0.3791 - val_accuracy: 0.8680\n",
      "Epoch 7/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3680 - accuracy: 0.8699 - val_loss: 0.3813 - val_accuracy: 0.8652\n",
      "Epoch 8/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3578 - accuracy: 0.8727 - val_loss: 0.3622 - val_accuracy: 0.8720\n",
      "Epoch 9/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3475 - accuracy: 0.8757 - val_loss: 0.3599 - val_accuracy: 0.8674\n",
      "Epoch 10/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3394 - accuracy: 0.8784 - val_loss: 0.3416 - val_accuracy: 0.8798\n",
      "Epoch 11/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3296 - accuracy: 0.8820 - val_loss: 0.3392 - val_accuracy: 0.8806\n",
      "Epoch 12/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3221 - accuracy: 0.8854 - val_loss: 0.3350 - val_accuracy: 0.8834\n",
      "Epoch 13/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3152 - accuracy: 0.8875 - val_loss: 0.3393 - val_accuracy: 0.8806\n",
      "Epoch 14/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3079 - accuracy: 0.8887 - val_loss: 0.3435 - val_accuracy: 0.8794\n",
      "Epoch 15/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.3012 - accuracy: 0.8911 - val_loss: 0.3329 - val_accuracy: 0.8790\n",
      "Epoch 16/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2949 - accuracy: 0.8930 - val_loss: 0.3247 - val_accuracy: 0.8824\n",
      "Epoch 17/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2897 - accuracy: 0.8952 - val_loss: 0.3321 - val_accuracy: 0.8814\n",
      "Epoch 18/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2839 - accuracy: 0.8974 - val_loss: 0.3304 - val_accuracy: 0.8820\n",
      "Epoch 19/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2789 - accuracy: 0.8992 - val_loss: 0.3159 - val_accuracy: 0.8866\n",
      "Epoch 20/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2744 - accuracy: 0.8998 - val_loss: 0.3092 - val_accuracy: 0.8912\n",
      "Epoch 21/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2695 - accuracy: 0.9029 - val_loss: 0.3112 - val_accuracy: 0.8880\n",
      "Epoch 22/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2641 - accuracy: 0.9050 - val_loss: 0.3204 - val_accuracy: 0.8856\n",
      "Epoch 23/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2592 - accuracy: 0.9065 - val_loss: 0.3052 - val_accuracy: 0.8890\n",
      "Epoch 24/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2555 - accuracy: 0.9074 - val_loss: 0.3079 - val_accuracy: 0.8886\n",
      "Epoch 25/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2506 - accuracy: 0.9088 - val_loss: 0.3134 - val_accuracy: 0.8850\n",
      "Epoch 26/30\n",
      "1719/1719 [==============================] - 6s 4ms/step - loss: 0.2463 - accuracy: 0.9104 - val_loss: 0.3154 - val_accuracy: 0.8832\n",
      "Epoch 27/30\n",
      "1719/1719 [==============================] - 6s 3ms/step - loss: 0.2420 - accuracy: 0.9122 - val_loss: 0.3142 - val_accuracy: 0.8858\n",
      "Epoch 28/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2387 - accuracy: 0.9133 - val_loss: 0.3005 - val_accuracy: 0.8902\n",
      "Epoch 29/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2344 - accuracy: 0.9147 - val_loss: 0.3136 - val_accuracy: 0.8858\n",
      "Epoch 30/30\n",
      "1719/1719 [==============================] - 5s 3ms/step - loss: 0.2316 - accuracy: 0.9162 - val_loss: 0.3003 - val_accuracy: 0.8914\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs=30, validation_data=(X_valid, y_valid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "iVEd_zH1wgBE",
    "outputId": "ce1d46b2-d8d4-4b44-8840-bcc293b1cc05"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])"
      ]
     },
     "execution_count": 31,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "history.history.keys()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "vlc0vNycwgBI",
    "outputId": "46113e3c-d63d-49c0-cc23-a30469c4b14c"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(history.history).plot(figsize=(10, 6))\n",
    "plt.grid(True)\n",
    "plt.ylim(0,1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "16soTo-FwgBN",
    "outputId": "0197c013-2466-4c96-c7e7-fc757405686f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 [==============================] - 1s 3ms/step - loss: 0.3343 - accuracy: 0.8789\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.33433637022972107, 0.8788999915122986]"
      ]
     },
     "execution_count": 33,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "L75vBkGkwgBR"
   },
   "source": [
    "### USING THE MODEL TO MAKE PREDICTIONS"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "LlS4rsLEwgBS",
    "outputId": "b89c48c4-b21a-46dd-ec47-c214c8dede6d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.  , 0.  , 0.  , 0.  , 0.  , 0.01, 0.  , 0.07, 0.  , 0.92],\n",
       "       [0.  , 0.  , 1.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ],\n",
       "       [0.  , 1.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  , 0.  ]],\n",
       "      dtype=float32)"
      ]
     },
     "execution_count": 34,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_new = X_test[:3]\n",
    "y_proba = model.predict(X_new)\n",
    "y_proba.round(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9wLSaEDBwgBX"
   },
   "outputs": [],
   "source": [
    "# y_pred = model.predict_classes(X_new)\n",
    "y_pred = np.argmax(model.predict(X_new), axis=-1) # Equivalent to code above as it is deprocated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "nslzbs6QwgBd",
    "outputId": "3b82b453-3f8e-4710-a70b-bef76e0d0fcc"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 2, 1])"
      ]
     },
     "execution_count": 36,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "0mzrtaOXwgBk",
    "outputId": "61e00535-2084-412c-e58f-9f358c8d06b9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['Ankle boot', 'Pullover', 'Trouser'], dtype='<U11')"
      ]
     },
     "execution_count": 37,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.array(class_names)[y_pred]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "kacBbcJPwgBp",
    "outputId": "590f2e86-dc1c-4b01-ec97-98a5a95772de"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 2, 1], dtype=uint8)"
      ]
     },
     "execution_count": 38,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_new = y_test[:3]\n",
    "y_new"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "MRvaNFJzwgBy",
    "outputId": "bc1250df-21b7-4a8e-805c-ee155c0f4a75"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 518.4x172.8 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7.2, 2.4))\n",
    "for index, image in enumerate(X_new):\n",
    "    plt.subplot(1, 3, index + 1)\n",
    "    plt.imshow(image, cmap=\"binary\", interpolation=\"nearest\")\n",
    "    plt.axis('off')\n",
    "    plt.title(class_names[y_test[index]], fontsize=12)\n",
    "plt.subplots_adjust(wspace=0.2, hspace=0.5)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "a9fWQmqzwgB4"
   },
   "source": [
    "## Building a Regression MLP Using the Sequential API"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "JIdafMLTwgB5",
    "outputId": "ae8da988-a891-43ca-bf8c-23f2c49f057a"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Downloading Cal. housing from https://ndownloader.figshare.com/files/5976036 to /root/scikit_learn_data\n"
     ]
    }
   ],
   "source": [
    "from sklearn.datasets import fetch_california_housing\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "\n",
    "housing = fetch_california_housing()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "sDND8p6ewgB_"
   },
   "outputs": [],
   "source": [
    "X_train_full, X_test, y_train_full, y_test = train_test_split(housing.data, housing.target)\n",
    "X_train, X_valid, y_train, y_valid = train_test_split(X_train_full, y_train_full)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "gkdr00f2wgCD"
   },
   "outputs": [],
   "source": [
    "scaler = StandardScaler()\n",
    "X_train = scaler.fit_transform(X_train)\n",
    "X_valid = scaler.transform(X_valid)\n",
    "X_test = scaler.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "YY9HiB8QwgCH",
    "outputId": "e2b72149-c3af-44d5-908d-c6957a68450a"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(11610, 8)"
      ]
     },
     "execution_count": 43,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "OizwcJFDwgCO",
    "outputId": "10384c3a-c7fb-40d1-c601-294a51916c72"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(8,)"
      ]
     },
     "execution_count": 44,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train.shape[1:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "rimM829xwgCS"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, 'relu', input_shape = X_train.shape[1:]),\n",
    "    keras.layers.Dense(1)\n",
    "])\n",
    "\n",
    "model.compile(loss = 'mean_squared_error', optimizer = 'sgd')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 312
    },
    "colab_type": "code",
    "id": "1_HYyVQ3wgCY",
    "outputId": "c808ef7e-a386-4ce9-aa6d-945b482770fd"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 46,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "keras.utils.plot_model(model, 'california_housing_regression.png', show_shapes=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 717
    },
    "colab_type": "code",
    "id": "t7HO8GlQwgCc",
    "outputId": "715e455f-2a61-4ca1-82fa-1357ea73eea3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 2.1019 - val_loss: 0.9200\n",
      "Epoch 2/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.8252 - val_loss: 0.7068\n",
      "Epoch 3/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6903 - val_loss: 0.6480\n",
      "Epoch 4/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6425 - val_loss: 0.6086\n",
      "Epoch 5/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6068 - val_loss: 0.5776\n",
      "Epoch 6/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5797 - val_loss: 0.5530\n",
      "Epoch 7/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5583 - val_loss: 0.5345\n",
      "Epoch 8/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5415 - val_loss: 0.5199\n",
      "Epoch 9/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5281 - val_loss: 0.5093\n",
      "Epoch 10/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5176 - val_loss: 0.4999\n",
      "Epoch 11/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5087 - val_loss: 0.4922\n",
      "Epoch 12/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5014 - val_loss: 0.4866\n",
      "Epoch 13/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4955 - val_loss: 0.4817\n",
      "Epoch 14/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4907 - val_loss: 0.4768\n",
      "Epoch 15/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4865 - val_loss: 0.4733\n",
      "Epoch 16/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4833 - val_loss: 0.4702\n",
      "Epoch 17/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4801 - val_loss: 0.4681\n",
      "Epoch 18/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4774 - val_loss: 0.4654\n",
      "Epoch 19/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4748 - val_loss: 0.4632\n",
      "Epoch 20/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4726 - val_loss: 0.4611\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss=\"mean_squared_error\", optimizer=keras.optimizers.SGD(lr=1e-3))\n",
    "history = model.fit(X_train, y_train, epochs=20, validation_data=(X_valid, y_valid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "gMjhKTGJwgCg",
    "outputId": "9ee1beb9-deb7-481e-e8c5-5ed8ed40f5d8"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 2ms/step - loss: 0.4667\n"
     ]
    }
   ],
   "source": [
    "mse_test = model.evaluate(X_test, y_test)\n",
    "X_new = X_test[:3]\n",
    "y_pred = model.predict(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "L5lZRFNKwgCk"
   },
   "outputs": [],
   "source": [
    "X_new = X_test[:3] # pretend these are new instances\n",
    "y_pred = model.predict(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 70
    },
    "colab_type": "code",
    "id": "ANnpAXs4wgCn",
    "outputId": "d43670e7-7c9b-4644-f835-0bed93c09dc6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.0032455],\n",
       "       [1.379019 ],\n",
       "       [2.3937652]], dtype=float32)"
      ]
     },
     "execution_count": 50,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 377
    },
    "colab_type": "code",
    "id": "SWd024ytwgCt",
    "outputId": "ce279c60-a991-48a9-873c-f76db8f31d76"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light",
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "pd.DataFrame(history.history).plot(figsize=(10, 6))\n",
    "plt.grid(True)\n",
    "plt.ylim(0,1)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZA0YwZtEwgCx"
   },
   "source": [
    "# Functional API"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZhTvpFaqwgC1"
   },
   "source": [
    "Not all neural network models are simply sequential. Some may have complex topologies. Some may have multiple inputs and/or multiple outputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "DGAGZ6EWwgC2"
   },
   "outputs": [],
   "source": [
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "mjGiH92fwgC9"
   },
   "outputs": [],
   "source": [
    "input_ = keras.layers.Input(shape = X_train.shape[1:])\n",
    "hidden1 = keras.layers.Dense(30, activation = 'relu')(input_)\n",
    "hidden2 = keras.layers.Dense(30, activation = 'relu')(hidden1)\n",
    "concat = keras.layers.concatenate([input_, hidden2])\n",
    "output = keras.layers.Dense(1)(concat)\n",
    "\n",
    "model = keras.models.Model(inputs = [input_], outputs = [output])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 350
    },
    "colab_type": "code",
    "id": "WO_06xcawgDC",
    "outputId": "572283ec-9b30-4a0e-8b00-b9faf9f869ce"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model\"\n",
      "__________________________________________________________________________________________________\n",
      "Layer (type)                    Output Shape         Param #     Connected to                     \n",
      "==================================================================================================\n",
      "input_1 (InputLayer)            [(None, 8)]          0                                            \n",
      "__________________________________________________________________________________________________\n",
      "dense_5 (Dense)                 (None, 30)           270         input_1[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_6 (Dense)                 (None, 30)           930         dense_5[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "concatenate (Concatenate)       (None, 38)           0           input_1[0][0]                    \n",
      "                                                                 dense_6[0][0]                    \n",
      "__________________________________________________________________________________________________\n",
      "dense_7 (Dense)                 (None, 1)            39          concatenate[0][0]                \n",
      "==================================================================================================\n",
      "Total params: 1,239\n",
      "Trainable params: 1,239\n",
      "Non-trainable params: 0\n",
      "__________________________________________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "2S8cx9r1wgDR",
    "outputId": "5b68ac86-626b-42d8-f808-e8b798d4d916"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 2.0043 - val_loss: 0.7419\n",
      "Epoch 2/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.7490 - val_loss: 0.6072\n",
      "Epoch 3/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.6397 - val_loss: 0.5725\n",
      "Epoch 4/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5981 - val_loss: 0.5460\n",
      "Epoch 5/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5647 - val_loss: 0.5311\n",
      "Epoch 6/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5486 - val_loss: 0.5208\n",
      "Epoch 7/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5386 - val_loss: 0.5104\n",
      "Epoch 8/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5271 - val_loss: 0.5028\n",
      "Epoch 9/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5182 - val_loss: 0.4970\n",
      "Epoch 10/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5110 - val_loss: 0.4897\n",
      "Epoch 11/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5040 - val_loss: 0.4835\n",
      "Epoch 12/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4990 - val_loss: 0.4792\n",
      "Epoch 13/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4925 - val_loss: 0.4757\n",
      "Epoch 14/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4878 - val_loss: 0.4700\n",
      "Epoch 15/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4830 - val_loss: 0.4665\n",
      "Epoch 16/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4798 - val_loss: 0.4627\n",
      "Epoch 17/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4754 - val_loss: 0.4604\n",
      "Epoch 18/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4720 - val_loss: 0.4562\n",
      "Epoch 19/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4676 - val_loss: 0.4535\n",
      "Epoch 20/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4643 - val_loss: 0.4503\n",
      "Epoch 21/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4607 - val_loss: 0.4480\n",
      "Epoch 22/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4580 - val_loss: 0.4475\n",
      "Epoch 23/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4553 - val_loss: 0.4426\n",
      "Epoch 24/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4523 - val_loss: 0.4412\n",
      "Epoch 25/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4494 - val_loss: 0.4394\n",
      "Epoch 26/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4472 - val_loss: 0.4384\n",
      "Epoch 27/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4447 - val_loss: 0.4348\n",
      "Epoch 28/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4421 - val_loss: 0.4330\n",
      "Epoch 29/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4397 - val_loss: 0.4315\n",
      "Epoch 30/30\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4378 - val_loss: 0.4296\n",
      "162/162 [==============================] - 0s 2ms/step - loss: 0.4366\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss = 'mean_squared_error', optimizer=keras.optimizers.SGD(lr = 1e-3))\n",
    "history = model.fit(X_train, y_train, epochs=30, validation_data = (X_valid, y_valid))\n",
    "mse_test = model.evaluate(X_test, y_test)\n",
    "y_pred = model.predict(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "oHojWjQlwgDV",
    "outputId": "06833325-b287-4315-adad-f112223ba06f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.4365706443786621"
      ]
     },
     "execution_count": 56,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mse_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 70
    },
    "colab_type": "code",
    "id": "FjbXRtg5wgDe",
    "outputId": "88129129-29e7-48d6-cfe2-9015f86debbb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.978745 ],\n",
       "       [1.201359 ],\n",
       "       [2.3135884]], dtype=float32)"
      ]
     },
     "execution_count": 57,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "NW6tq_v2wgDk",
    "outputId": "bd7b8be8-c89a-4c60-c6d3-8bd060c70948"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.888, 1.064, 2.098])"
      ]
     },
     "execution_count": 58,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WDILvxXpwgDp"
   },
   "source": [
    "What if you want to send different subsets of input features through the wide or deep paths? We will send 5 features (features 0 to 4), and 6 through the deep path (features 2 to 7). Note that 3 features will go through both (features 2, 3 and 4)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "9S4pa6GwwgDp"
   },
   "outputs": [],
   "source": [
    "inputA = keras.layers.Input(shape = [5], name='wide_input')\n",
    "inputB = keras.layers.Input(shape = [6], name='deep_input')\n",
    "hidden1 = keras.layers.Dense(30, activation = 'relu')(inputB)\n",
    "hidden2 = keras.layers.Dense(30, activation = 'relu')(hidden1)\n",
    "concat = keras.layers.concatenate([inputA, hidden2])\n",
    "output = keras.layers.Dense(1, name = 'Output')(concat)\n",
    "\n",
    "model = keras.models.Model(inputs = [inputA, inputB], outputs = [output])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "tez686IZwgDu"
   },
   "outputs": [],
   "source": [
    "model.compile(loss = 'mse', optimizer=keras.optimizers.SGD(lr = 1e-3))\n",
    "\n",
    "X_train_A, X_train_B = X_train[:, :5], X_train[:,2:]\n",
    "X_valid_A, X_valid_B = X_valid[:, :5], X_valid[:,2:]\n",
    "X_test_A, X_test_B = X_test[:, :5], X_test[:,2:]\n",
    "X_new_A, X_new_B = X_new[:, :5], X_new[:,2:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 717
    },
    "colab_type": "code",
    "id": "CQgM11H6wgD4",
    "outputId": "25c66df8-14c9-4ccc-f6cf-a77608d05f3d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 1.9979 - val_loss: 0.9279\n",
      "Epoch 2/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.8090 - val_loss: 0.6992\n",
      "Epoch 3/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6807 - val_loss: 0.6305\n",
      "Epoch 4/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6327 - val_loss: 0.5963\n",
      "Epoch 5/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6039 - val_loss: 0.5736\n",
      "Epoch 6/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5831 - val_loss: 0.5557\n",
      "Epoch 7/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5666 - val_loss: 0.5422\n",
      "Epoch 8/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5532 - val_loss: 0.5310\n",
      "Epoch 9/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5421 - val_loss: 0.5230\n",
      "Epoch 10/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5335 - val_loss: 0.5144\n",
      "Epoch 11/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5255 - val_loss: 0.5081\n",
      "Epoch 12/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5197 - val_loss: 0.5027\n",
      "Epoch 13/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5138 - val_loss: 0.4987\n",
      "Epoch 14/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5096 - val_loss: 0.4940\n",
      "Epoch 15/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5054 - val_loss: 0.4904\n",
      "Epoch 16/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5020 - val_loss: 0.4869\n",
      "Epoch 17/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4988 - val_loss: 0.4843\n",
      "Epoch 18/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4958 - val_loss: 0.4816\n",
      "Epoch 19/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4931 - val_loss: 0.4789\n",
      "Epoch 20/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4905 - val_loss: 0.4765\n"
     ]
    }
   ],
   "source": [
    "history = model.fit((X_train_A, X_train_B), y_train, epochs = 20, validation_data = ((X_valid_A, X_valid_B), y_valid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "7GqlOcnUwgD7",
    "outputId": "2079b410-925b-420f-c76c-e49424be3a22"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 2ms/step - loss: 0.4847\n"
     ]
    }
   ],
   "source": [
    "mse_test = model.evaluate((X_test_A, X_test_B), y_test)\n",
    "y_pred = model.predict((X_new_A, X_new_B))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 70
    },
    "colab_type": "code",
    "id": "CP16Ix4cwgD_",
    "outputId": "b2e70076-ae7c-44d8-efca-2f0c4652d57b"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.89252585],\n",
       "       [1.2798747 ],\n",
       "       [2.5379155 ]], dtype=float32)"
      ]
     },
     "execution_count": 63,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "w7JEvRcswgEE",
    "outputId": "b1cc5079-6f1b-4e05-cc00-14aa54d427ac"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.888, 1.064, 2.098])"
      ]
     },
     "execution_count": 64,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "i8y6VKbPwgEI"
   },
   "source": [
    "\n",
    "Adding an auxiliary output for regularization:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "mmgUgI-xwgEI"
   },
   "outputs": [],
   "source": [
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ubI_BcNEwgEM"
   },
   "outputs": [],
   "source": [
    "input_A = keras.layers.Input(shape=[5], name=\"wide_input\")\n",
    "input_B = keras.layers.Input(shape=[6], name=\"deep_input\")\n",
    "hidden1 = keras.layers.Dense(30, activation=\"relu\")(input_B)\n",
    "hidden2 = keras.layers.Dense(30, activation=\"relu\")(hidden1)\n",
    "concat = keras.layers.concatenate([input_A, hidden2])\n",
    "output = keras.layers.Dense(1, name=\"main_output\")(concat)\n",
    "aux_output = keras.layers.Dense(1, name=\"aux_output\")(hidden2)\n",
    "model = keras.models.Model(inputs=[input_A, input_B], outputs=[output, aux_output])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "xGODQUczwgEO"
   },
   "outputs": [],
   "source": [
    "model.compile(loss = ['mse','mse'], loss_weights = [0.9, 0.1], optimizer = keras.optimizers.SGD(lr = 1e-3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "jnj1dgtwwgEQ",
    "outputId": "50404e19-bf04-4594-d099-8ed2a5522978"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 2.7548 - main_output_loss: 2.5201 - aux_output_loss: 4.8672 - val_loss: 1.4389 - val_main_output_loss: 1.1401 - val_aux_output_loss: 4.1287\n",
      "Epoch 2/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 1.2315 - main_output_loss: 0.9874 - aux_output_loss: 3.4287 - val_loss: 0.9747 - val_main_output_loss: 0.7658 - val_aux_output_loss: 2.8543\n",
      "Epoch 3/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.9175 - main_output_loss: 0.7458 - aux_output_loss: 2.4626 - val_loss: 0.8354 - val_main_output_loss: 0.6906 - val_aux_output_loss: 2.1386\n",
      "Epoch 4/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.8183 - main_output_loss: 0.6944 - aux_output_loss: 1.9327 - val_loss: 0.7677 - val_main_output_loss: 0.6575 - val_aux_output_loss: 1.7590\n",
      "Epoch 5/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7637 - main_output_loss: 0.6642 - aux_output_loss: 1.6588 - val_loss: 0.7238 - val_main_output_loss: 0.6308 - val_aux_output_loss: 1.5610\n",
      "Epoch 6/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7298 - main_output_loss: 0.6424 - aux_output_loss: 1.5166 - val_loss: 0.6949 - val_main_output_loss: 0.6104 - val_aux_output_loss: 1.4556\n",
      "Epoch 7/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7044 - main_output_loss: 0.6230 - aux_output_loss: 1.4372 - val_loss: 0.6735 - val_main_output_loss: 0.5933 - val_aux_output_loss: 1.3949\n",
      "Epoch 8/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6840 - main_output_loss: 0.6060 - aux_output_loss: 1.3859 - val_loss: 0.6566 - val_main_output_loss: 0.5794 - val_aux_output_loss: 1.3514\n",
      "Epoch 9/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6673 - main_output_loss: 0.5918 - aux_output_loss: 1.3473 - val_loss: 0.6424 - val_main_output_loss: 0.5673 - val_aux_output_loss: 1.3185\n",
      "Epoch 10/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6534 - main_output_loss: 0.5798 - aux_output_loss: 1.3162 - val_loss: 0.6287 - val_main_output_loss: 0.5553 - val_aux_output_loss: 1.2897\n",
      "Epoch 11/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6404 - main_output_loss: 0.5683 - aux_output_loss: 1.2895 - val_loss: 0.6172 - val_main_output_loss: 0.5453 - val_aux_output_loss: 1.2636\n",
      "Epoch 12/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6296 - main_output_loss: 0.5591 - aux_output_loss: 1.2641 - val_loss: 0.6077 - val_main_output_loss: 0.5375 - val_aux_output_loss: 1.2399\n",
      "Epoch 13/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6202 - main_output_loss: 0.5511 - aux_output_loss: 1.2416 - val_loss: 0.5995 - val_main_output_loss: 0.5308 - val_aux_output_loss: 1.2174\n",
      "Epoch 14/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6121 - main_output_loss: 0.5446 - aux_output_loss: 1.2200 - val_loss: 0.5921 - val_main_output_loss: 0.5250 - val_aux_output_loss: 1.1964\n",
      "Epoch 15/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6051 - main_output_loss: 0.5391 - aux_output_loss: 1.1994 - val_loss: 0.5861 - val_main_output_loss: 0.5205 - val_aux_output_loss: 1.1763\n",
      "Epoch 16/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5988 - main_output_loss: 0.5342 - aux_output_loss: 1.1799 - val_loss: 0.5798 - val_main_output_loss: 0.5157 - val_aux_output_loss: 1.1570\n",
      "Epoch 17/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5934 - main_output_loss: 0.5304 - aux_output_loss: 1.1612 - val_loss: 0.5754 - val_main_output_loss: 0.5128 - val_aux_output_loss: 1.1391\n",
      "Epoch 18/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5885 - main_output_loss: 0.5269 - aux_output_loss: 1.1437 - val_loss: 0.5708 - val_main_output_loss: 0.5096 - val_aux_output_loss: 1.1213\n",
      "Epoch 19/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5844 - main_output_loss: 0.5242 - aux_output_loss: 1.1263 - val_loss: 0.5669 - val_main_output_loss: 0.5072 - val_aux_output_loss: 1.1044\n",
      "Epoch 20/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5800 - main_output_loss: 0.5211 - aux_output_loss: 1.1096 - val_loss: 0.5631 - val_main_output_loss: 0.5048 - val_aux_output_loss: 1.0878\n",
      "Epoch 21/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5761 - main_output_loss: 0.5187 - aux_output_loss: 1.0932 - val_loss: 0.5593 - val_main_output_loss: 0.5024 - val_aux_output_loss: 1.0714\n",
      "Epoch 22/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5725 - main_output_loss: 0.5164 - aux_output_loss: 1.0772 - val_loss: 0.5567 - val_main_output_loss: 0.5013 - val_aux_output_loss: 1.0552\n",
      "Epoch 23/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5692 - main_output_loss: 0.5145 - aux_output_loss: 1.0617 - val_loss: 0.5525 - val_main_output_loss: 0.4984 - val_aux_output_loss: 1.0396\n",
      "Epoch 24/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5657 - main_output_loss: 0.5123 - aux_output_loss: 1.0459 - val_loss: 0.5496 - val_main_output_loss: 0.4968 - val_aux_output_loss: 1.0248\n",
      "Epoch 25/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5620 - main_output_loss: 0.5098 - aux_output_loss: 1.0312 - val_loss: 0.5477 - val_main_output_loss: 0.4963 - val_aux_output_loss: 1.0100\n",
      "Epoch 26/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5594 - main_output_loss: 0.5087 - aux_output_loss: 1.0162 - val_loss: 0.5435 - val_main_output_loss: 0.4933 - val_aux_output_loss: 0.9953\n",
      "Epoch 27/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5563 - main_output_loss: 0.5067 - aux_output_loss: 1.0026 - val_loss: 0.5407 - val_main_output_loss: 0.4917 - val_aux_output_loss: 0.9809\n",
      "Epoch 28/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5528 - main_output_loss: 0.5044 - aux_output_loss: 0.9884 - val_loss: 0.5377 - val_main_output_loss: 0.4900 - val_aux_output_loss: 0.9667\n",
      "Epoch 29/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5497 - main_output_loss: 0.5025 - aux_output_loss: 0.9743 - val_loss: 0.5348 - val_main_output_loss: 0.4884 - val_aux_output_loss: 0.9529\n",
      "Epoch 30/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5468 - main_output_loss: 0.5009 - aux_output_loss: 0.9600 - val_loss: 0.5320 - val_main_output_loss: 0.4867 - val_aux_output_loss: 0.9395\n"
     ]
    }
   ],
   "source": [
    "history = model.fit([X_train_A, X_train_B], [y_train, y_train], epochs = 30, validation_data=([X_valid_A, X_valid_B], [y_valid, y_valid]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "xtXQSE0LwgEX",
    "outputId": "fc75a6f6-5173-4fdd-f560-c1bc876d325d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 3ms/step - loss: 0.5359 - main_output_loss: 0.4907 - aux_output_loss: 0.9422\n"
     ]
    }
   ],
   "source": [
    "total_loss, main_loss, aux_loss = model.evaluate(\n",
    "    [X_test_A, X_test_B], [y_test, y_test])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "lO8MZTgCwgEZ",
    "outputId": "bb406c67-d491-4b41-96c0-0355d8ac5b80"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.535856306552887, 0.49070999026298523, 0.942171037197113)"
      ]
     },
     "execution_count": 70,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_loss, main_loss, aux_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "T7nkzyFOwgEf"
   },
   "outputs": [],
   "source": [
    "y_pred_main, y_pred_aux = model.predict([X_new_A, X_new_B])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 105
    },
    "colab_type": "code",
    "id": "NbRRIkbiwgEh",
    "outputId": "f0c05f6c-00ee-4e43-aacd-32f93abc9989"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0.9418996],\n",
       "        [1.3402183],\n",
       "        [2.5042229]], dtype=float32), array([[1.4926925],\n",
       "        [1.9279394],\n",
       "        [2.5046647]], dtype=float32))"
      ]
     },
     "execution_count": 72,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_main, y_pred_aux"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "Npp1d8FfwgEm",
    "outputId": "e786a763-1b65-47e6-e1f9-11c2e81edc79"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.888, 1.064, 2.098])"
      ]
     },
     "execution_count": 73,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "WieQBcmq6BVJ"
   },
   "source": [
    "## Using the Subclassing API to Build Dynamic Models"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "vnYmoDBpwgEq"
   },
   "outputs": [],
   "source": [
    "class WideAndDeepModel(keras.Model):\n",
    "    def __init__(self, units = 30, activation = 'relu', **kwargs):\n",
    "        super().__init__(**kwargs)\n",
    "        self.hidden1 = keras.layers.Dense(units, activation=activation)\n",
    "        self.hidden2 = keras.layers.Dense(units, activation=activation)\n",
    "        self.main_output = keras.layers.Dense(1)\n",
    "        self.aux_output = keras.layers.Dense(1)\n",
    "\n",
    "    def call(self, inputs):\n",
    "        input_A, input_B = inputs\n",
    "        hidden1 = self.hidden1(input_B)\n",
    "        hidden2 = self.hidden2(hidden1)\n",
    "        concat = keras.layers.concatenate([input_A, hidden2])\n",
    "        main_output = self.main_output(concat)\n",
    "        aux_output = self.aux_output(hidden2)\n",
    "        return main_output, aux_output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "fW-AWvGy7m1L"
   },
   "outputs": [],
   "source": [
    "model = WideAndDeepModel()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0IQP9hFv7oxx"
   },
   "outputs": [],
   "source": [
    "model.compile(loss='mse', loss_weights=[.9, .1], optimizer=keras.optimizers.SGD(lr = 1e-3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "sCqlv7u28Hnj",
    "outputId": "f2a96b7e-dd69-46aa-8913-d20e831646e9"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 2.5845 - output_1_loss: 2.3068 - output_2_loss: 5.0842 - val_loss: 1.3696 - val_output_1_loss: 1.0528 - val_output_2_loss: 4.2206\n",
      "Epoch 2/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 1.1704 - output_1_loss: 0.9245 - output_2_loss: 3.3843 - val_loss: 0.9420 - val_output_1_loss: 0.7365 - val_output_2_loss: 2.7920\n",
      "Epoch 3/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.8734 - output_1_loss: 0.7029 - output_2_loss: 2.4080 - val_loss: 0.8074 - val_output_1_loss: 0.6626 - val_output_2_loss: 2.1101\n",
      "Epoch 4/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7823 - output_1_loss: 0.6538 - output_2_loss: 1.9380 - val_loss: 0.7420 - val_output_1_loss: 0.6261 - val_output_2_loss: 1.7847\n",
      "Epoch 5/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7330 - output_1_loss: 0.6251 - output_2_loss: 1.7038 - val_loss: 0.7034 - val_output_1_loss: 0.6019 - val_output_2_loss: 1.6166\n",
      "Epoch 6/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.7012 - output_1_loss: 0.6041 - output_2_loss: 1.5754 - val_loss: 0.6767 - val_output_1_loss: 0.5832 - val_output_2_loss: 1.5175\n",
      "Epoch 7/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6777 - output_1_loss: 0.5868 - output_2_loss: 1.4956 - val_loss: 0.6564 - val_output_1_loss: 0.5680 - val_output_2_loss: 1.4521\n",
      "Epoch 8/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6592 - output_1_loss: 0.5727 - output_2_loss: 1.4373 - val_loss: 0.6399 - val_output_1_loss: 0.5553 - val_output_2_loss: 1.4017\n",
      "Epoch 9/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6435 - output_1_loss: 0.5606 - output_2_loss: 1.3900 - val_loss: 0.6270 - val_output_1_loss: 0.5454 - val_output_2_loss: 1.3615\n",
      "Epoch 10/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6308 - output_1_loss: 0.5507 - output_2_loss: 1.3512 - val_loss: 0.6136 - val_output_1_loss: 0.5346 - val_output_2_loss: 1.3246\n",
      "Epoch 11/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6186 - output_1_loss: 0.5410 - output_2_loss: 1.3169 - val_loss: 0.6022 - val_output_1_loss: 0.5257 - val_output_2_loss: 1.2906\n",
      "Epoch 12/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.6079 - output_1_loss: 0.5329 - output_2_loss: 1.2837 - val_loss: 0.5919 - val_output_1_loss: 0.5177 - val_output_2_loss: 1.2600\n",
      "Epoch 13/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5979 - output_1_loss: 0.5251 - output_2_loss: 1.2535 - val_loss: 0.5827 - val_output_1_loss: 0.5107 - val_output_2_loss: 1.2310\n",
      "Epoch 14/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5891 - output_1_loss: 0.5184 - output_2_loss: 1.2256 - val_loss: 0.5742 - val_output_1_loss: 0.5043 - val_output_2_loss: 1.2036\n",
      "Epoch 15/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5807 - output_1_loss: 0.5120 - output_2_loss: 1.1989 - val_loss: 0.5664 - val_output_1_loss: 0.4986 - val_output_2_loss: 1.1765\n",
      "Epoch 16/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5730 - output_1_loss: 0.5064 - output_2_loss: 1.1727 - val_loss: 0.5583 - val_output_1_loss: 0.4925 - val_output_2_loss: 1.1508\n",
      "Epoch 17/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5658 - output_1_loss: 0.5012 - output_2_loss: 1.1474 - val_loss: 0.5520 - val_output_1_loss: 0.4881 - val_output_2_loss: 1.1265\n",
      "Epoch 18/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5592 - output_1_loss: 0.4964 - output_2_loss: 1.1241 - val_loss: 0.5453 - val_output_1_loss: 0.4833 - val_output_2_loss: 1.1032\n",
      "Epoch 19/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5531 - output_1_loss: 0.4922 - output_2_loss: 1.1008 - val_loss: 0.5396 - val_output_1_loss: 0.4795 - val_output_2_loss: 1.0804\n",
      "Epoch 20/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5466 - output_1_loss: 0.4875 - output_2_loss: 1.0790 - val_loss: 0.5343 - val_output_1_loss: 0.4761 - val_output_2_loss: 1.0583\n",
      "Epoch 21/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5413 - output_1_loss: 0.4840 - output_2_loss: 1.0577 - val_loss: 0.5285 - val_output_1_loss: 0.4720 - val_output_2_loss: 1.0374\n",
      "Epoch 22/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5363 - output_1_loss: 0.4806 - output_2_loss: 1.0374 - val_loss: 0.5240 - val_output_1_loss: 0.4691 - val_output_2_loss: 1.0181\n",
      "Epoch 23/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5317 - output_1_loss: 0.4776 - output_2_loss: 1.0182 - val_loss: 0.5188 - val_output_1_loss: 0.4655 - val_output_2_loss: 0.9984\n",
      "Epoch 24/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5271 - output_1_loss: 0.4746 - output_2_loss: 0.9993 - val_loss: 0.5149 - val_output_1_loss: 0.4632 - val_output_2_loss: 0.9806\n",
      "Epoch 25/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5224 - output_1_loss: 0.4714 - output_2_loss: 0.9818 - val_loss: 0.5135 - val_output_1_loss: 0.4634 - val_output_2_loss: 0.9639\n",
      "Epoch 26/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5192 - output_1_loss: 0.4697 - output_2_loss: 0.9653 - val_loss: 0.5081 - val_output_1_loss: 0.4594 - val_output_2_loss: 0.9464\n",
      "Epoch 27/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5156 - output_1_loss: 0.4674 - output_2_loss: 0.9497 - val_loss: 0.5041 - val_output_1_loss: 0.4568 - val_output_2_loss: 0.9303\n",
      "Epoch 28/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5113 - output_1_loss: 0.4644 - output_2_loss: 0.9339 - val_loss: 0.5003 - val_output_1_loss: 0.4542 - val_output_2_loss: 0.9153\n",
      "Epoch 29/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5081 - output_1_loss: 0.4624 - output_2_loss: 0.9192 - val_loss: 0.4973 - val_output_1_loss: 0.4525 - val_output_2_loss: 0.9007\n",
      "Epoch 30/30\n",
      "363/363 [==============================] - 2s 5ms/step - loss: 0.5049 - output_1_loss: 0.4605 - output_2_loss: 0.9047 - val_loss: 0.4941 - val_output_1_loss: 0.4505 - val_output_2_loss: 0.8868\n"
     ]
    }
   ],
   "source": [
    "history = model.fit([X_train_A, X_train_B], [y_train, y_train], epochs = 30, validation_data=([X_valid_A, X_valid_B], [y_valid, y_valid]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "KjXtrgir8QQu",
    "outputId": "7e620608-c49a-4028-9485-539a7bff9eca"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 3ms/step - loss: 0.4966 - output_1_loss: 0.4538 - output_2_loss: 0.8815\n"
     ]
    }
   ],
   "source": [
    "total_loss, main_loss, aux_loss = model.evaluate((X_test_A, X_test_B), (y_test, y_test))\n",
    "y_pred_main, y_pred_aux = model.predict((X_new_A, X_new_B))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "lw8CpNpm8agu",
    "outputId": "d9b203c8-db80-4535-8afc-05806f779c25"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(array([[0.9041923],\n",
       "        [1.0551423],\n",
       "        [2.2951694]], dtype=float32), array([[1.4478953],\n",
       "        [1.7210677],\n",
       "        [2.5156014]], dtype=float32))"
      ]
     },
     "execution_count": 79,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred_main, y_pred_aux"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "0XZdIy_X8oqI"
   },
   "outputs": [],
   "source": [
    "## Saving and Restoring"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZJK_O708aYaP"
   },
   "outputs": [],
   "source": [
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "yyG1oru1aYaS"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, activation = 'relu', input_shape=[8]),\n",
    "    keras.layers.Dense(30, activation = 'relu'),\n",
    "    keras.layers.Dense(1)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "Y8plSJ8-aYaU",
    "outputId": "3175e3ac-f5dc-48a4-9f5b-76f0faadc558"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 2.0913 - val_loss: 0.9353\n",
      "Epoch 2/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.8252 - val_loss: 0.7311\n",
      "Epoch 3/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.7204 - val_loss: 0.6738\n",
      "Epoch 4/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6631 - val_loss: 0.6278\n",
      "Epoch 5/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6241 - val_loss: 0.5984\n",
      "Epoch 6/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5951 - val_loss: 0.5725\n",
      "Epoch 7/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5729 - val_loss: 0.5528\n",
      "Epoch 8/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5520 - val_loss: 0.5351\n",
      "Epoch 9/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5356 - val_loss: 0.5214\n",
      "Epoch 10/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5219 - val_loss: 0.5090\n",
      "Epoch 11/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5097 - val_loss: 0.4986\n",
      "Epoch 12/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4999 - val_loss: 0.4898\n",
      "Epoch 13/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4907 - val_loss: 0.4820\n",
      "Epoch 14/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4830 - val_loss: 0.4758\n",
      "Epoch 15/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4762 - val_loss: 0.4696\n",
      "Epoch 16/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4714 - val_loss: 0.4657\n",
      "Epoch 17/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4661 - val_loss: 0.4617\n",
      "Epoch 18/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4630 - val_loss: 0.4584\n",
      "Epoch 19/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4578 - val_loss: 0.4541\n",
      "Epoch 20/20\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4542 - val_loss: 0.4514\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss = 'mse', optimizer = keras.optimizers.SGD(lr = 1e-3))\n",
    "history = model.fit(X_train, y_train, epochs = 20, validation_data = (X_valid, y_valid))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "QqSfq1E8aYaW",
    "outputId": "e23eaa2d-5241-4841-d2fe-9d3c932207f4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 2ms/step - loss: 0.4592\n"
     ]
    }
   ],
   "source": [
    "mse_test = model.evaluate(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "o8xnDHexaYaY"
   },
   "outputs": [],
   "source": [
    "model.save('my_keras_model.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "g5CC07amaYab",
    "outputId": "ab8c993c-cd7e-4277-8c6b-54c23f0c554f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.1143327],\n",
       "       [1.2733324],\n",
       "       [2.4446313]], dtype=float32)"
      ]
     },
     "execution_count": 86,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.predict(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "f4TVWBwjaYac",
    "outputId": "9e2e9706-dda0-47c4-f64c-2371301330d3"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([0.888, 1.064, 2.098])"
      ]
     },
     "execution_count": 87,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_test[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "M7j7a0LbaYae"
   },
   "outputs": [],
   "source": [
    "model.save_weights(\"my_keras_weights.ckpt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 89,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "3keiBLfVaYag",
    "outputId": "ab1c1b5e-20af-4d2d-aad9-67338a552911"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x7f8deb5ccda0>"
      ]
     },
     "execution_count": 89,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_weights(\"my_keras_weights.ckpt\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "CVbcOaXUaYah"
   },
   "source": [
    "Using Callbacks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NwSpVdVTaYah"
   },
   "outputs": [],
   "source": [
    "keras.backend.clear_session()\n",
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zJGNsn5PaYak"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, activation=\"relu\", input_shape=[8]),\n",
    "    keras.layers.Dense(30, activation=\"relu\"),\n",
    "    keras.layers.Dense(1)\n",
    "])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "E1ogXvDTaYal",
    "outputId": "093040f5-c9cd-443c-9ef1-275e6bb9f3b2"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 2.0913 - val_loss: 0.9353\n",
      "Epoch 2/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.8252 - val_loss: 0.7311\n",
      "Epoch 3/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.7204 - val_loss: 0.6738\n",
      "Epoch 4/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6631 - val_loss: 0.6278\n",
      "Epoch 5/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.6241 - val_loss: 0.5984\n",
      "Epoch 6/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5951 - val_loss: 0.5725\n",
      "Epoch 7/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5729 - val_loss: 0.5528\n",
      "Epoch 8/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5520 - val_loss: 0.5351\n",
      "Epoch 9/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5356 - val_loss: 0.5214\n",
      "Epoch 10/10\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5219 - val_loss: 0.5090\n",
      "162/162 [==============================] - 0s 2ms/step - loss: 0.5175\n"
     ]
    }
   ],
   "source": [
    "model.compile(loss=\"mse\", optimizer=keras.optimizers.SGD(lr=1e-3))\n",
    "checkpoint_cb = keras.callbacks.ModelCheckpoint(\"my_keras_model.h5\", save_best_only=True)\n",
    "history = model.fit(X_train, y_train, epochs=10,\n",
    "                    validation_data=(X_valid, y_valid),\n",
    "                    callbacks=[checkpoint_cb])\n",
    "model = keras.models.load_model(\"my_keras_model.h5\") # rollback to best model\n",
    "mse_test = model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "auunnRg4aYan"
   },
   "outputs": [],
   "source": [
    "model.compile(loss = 'mse', optimizer=keras.optimizers.SGD(lr = 1e-3))\n",
    "early_stopping_cb = keras.callbacks.EarlyStopping(patience= 10, restore_best_weights=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "ZLtnPF7maYap",
    "outputId": "91e5e9fe-08ea-4c44-8034-da9bc487fc72"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5094 - val_loss: 0.4982\n",
      "Epoch 2/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5022 - val_loss: 0.4905\n",
      "Epoch 3/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4911 - val_loss: 0.4818\n",
      "Epoch 4/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4832 - val_loss: 0.4756\n",
      "Epoch 5/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4766 - val_loss: 0.4701\n",
      "Epoch 6/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4709 - val_loss: 0.4660\n",
      "Epoch 7/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4679 - val_loss: 0.4622\n",
      "Epoch 8/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4616 - val_loss: 0.4580\n",
      "Epoch 9/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4579 - val_loss: 0.4552\n",
      "Epoch 10/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4544 - val_loss: 0.4516\n",
      "Epoch 11/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4513 - val_loss: 0.4490\n",
      "Epoch 12/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4491 - val_loss: 0.4461\n",
      "Epoch 13/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4451 - val_loss: 0.4435\n",
      "Epoch 14/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4423 - val_loss: 0.4409\n",
      "Epoch 15/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4395 - val_loss: 0.4382\n",
      "Epoch 16/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4384 - val_loss: 0.4368\n",
      "Epoch 17/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4356 - val_loss: 0.4347\n",
      "Epoch 18/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4350 - val_loss: 0.4327\n",
      "Epoch 19/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4311 - val_loss: 0.4301\n",
      "Epoch 20/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4292 - val_loss: 0.4286\n",
      "Epoch 21/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4271 - val_loss: 0.4275\n",
      "Epoch 22/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4257 - val_loss: 0.4261\n",
      "Epoch 23/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4242 - val_loss: 0.4230\n",
      "Epoch 24/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4226 - val_loss: 0.4223\n",
      "Epoch 25/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4211 - val_loss: 0.4222\n",
      "Epoch 26/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4198 - val_loss: 0.4209\n",
      "Epoch 27/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4185 - val_loss: 0.4182\n",
      "Epoch 28/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4167 - val_loss: 0.4169\n",
      "Epoch 29/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4153 - val_loss: 0.4159\n",
      "Epoch 30/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4146 - val_loss: 0.4146\n",
      "Epoch 31/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4135 - val_loss: 0.4133\n",
      "Epoch 32/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4119 - val_loss: 0.4123\n",
      "Epoch 33/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4110 - val_loss: 0.4115\n",
      "Epoch 34/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4097 - val_loss: 0.4127\n",
      "Epoch 35/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4103 - val_loss: 0.4103\n",
      "Epoch 36/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4078 - val_loss: 0.4108\n",
      "Epoch 37/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4072 - val_loss: 0.4078\n",
      "Epoch 38/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4056 - val_loss: 0.4062\n",
      "Epoch 39/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4043 - val_loss: 0.4071\n",
      "Epoch 40/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4103 - val_loss: 0.4074\n",
      "Epoch 41/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4048 - val_loss: 0.4135\n",
      "Epoch 42/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4048 - val_loss: 0.4036\n",
      "Epoch 43/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4013 - val_loss: 0.4059\n",
      "Epoch 44/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4008 - val_loss: 0.4023\n",
      "Epoch 45/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3990 - val_loss: 0.4014\n",
      "Epoch 46/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3986 - val_loss: 0.4022\n",
      "Epoch 47/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3999 - val_loss: 0.3998\n",
      "Epoch 48/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3967 - val_loss: 0.3988\n",
      "Epoch 49/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3966 - val_loss: 0.3972\n",
      "Epoch 50/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3948 - val_loss: 0.3978\n",
      "Epoch 51/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3936 - val_loss: 0.3971\n",
      "Epoch 52/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3931 - val_loss: 0.3960\n",
      "Epoch 53/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3920 - val_loss: 0.3957\n",
      "Epoch 54/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3949 - val_loss: 0.3956\n",
      "Epoch 55/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3928 - val_loss: 0.4004\n",
      "Epoch 56/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3928 - val_loss: 0.3939\n",
      "Epoch 57/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3933 - val_loss: 0.3928\n",
      "Epoch 58/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3882 - val_loss: 0.3909\n",
      "Epoch 59/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3875 - val_loss: 0.3907\n",
      "Epoch 60/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3869 - val_loss: 0.3916\n",
      "Epoch 61/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3884 - val_loss: 0.3907\n",
      "Epoch 62/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3894 - val_loss: 0.3884\n",
      "Epoch 63/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3843 - val_loss: 0.3876\n",
      "Epoch 64/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3832 - val_loss: 0.3876\n",
      "Epoch 65/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3825 - val_loss: 0.3862\n",
      "Epoch 66/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3815 - val_loss: 0.3858\n",
      "Epoch 67/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3809 - val_loss: 0.3849\n",
      "Epoch 68/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3798 - val_loss: 0.3852\n",
      "Epoch 69/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3795 - val_loss: 0.3838\n",
      "Epoch 70/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3790 - val_loss: 0.3825\n",
      "Epoch 71/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3775 - val_loss: 0.3830\n",
      "Epoch 72/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3775 - val_loss: 0.3820\n",
      "Epoch 73/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3783 - val_loss: 0.3815\n",
      "Epoch 74/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3762 - val_loss: 0.3813\n",
      "Epoch 75/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3773 - val_loss: 0.3815\n",
      "Epoch 76/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3743 - val_loss: 0.3798\n",
      "Epoch 77/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3730 - val_loss: 0.3789\n",
      "Epoch 78/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3725 - val_loss: 0.3782\n",
      "Epoch 79/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3723 - val_loss: 0.3777\n",
      "Epoch 80/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3709 - val_loss: 0.3770\n",
      "Epoch 81/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3705 - val_loss: 0.3775\n",
      "Epoch 82/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3705 - val_loss: 0.3747\n",
      "Epoch 83/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3690 - val_loss: 0.3750\n",
      "Epoch 84/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3690 - val_loss: 0.3745\n",
      "Epoch 85/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3684 - val_loss: 0.3746\n",
      "Epoch 86/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3675 - val_loss: 0.3726\n",
      "Epoch 87/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3662 - val_loss: 0.3736\n",
      "Epoch 88/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3675 - val_loss: 0.3713\n",
      "Epoch 89/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3652 - val_loss: 0.3721\n",
      "Epoch 90/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3654 - val_loss: 0.3710\n",
      "Epoch 91/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3639 - val_loss: 0.3744\n",
      "Epoch 92/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3648 - val_loss: 0.3737\n",
      "Epoch 93/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3804 - val_loss: 0.3736\n",
      "Epoch 94/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3655 - val_loss: 0.3734\n",
      "Epoch 95/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3645 - val_loss: 0.3754\n",
      "Epoch 96/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3652 - val_loss: 0.3664\n",
      "Epoch 97/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3613 - val_loss: 0.3674\n",
      "Epoch 98/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3623 - val_loss: 0.3653\n",
      "Epoch 99/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3590 - val_loss: 0.3659\n",
      "Epoch 100/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3586 - val_loss: 0.3649\n",
      "162/162 [==============================] - 0s 2ms/step - loss: 0.3650\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(X_train, y_train, epochs = 100, validation_data=(X_valid, y_valid),\n",
    "                   callbacks=[checkpoint_cb, early_stopping_cb])\n",
    "mse_test = model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "zRVcX9tVaYar"
   },
   "outputs": [],
   "source": [
    "class PrintValTrainRatioCallback(keras.callbacks.Callback):\n",
    "    def on_epoch_end(self, epoch, logs):\n",
    "        print(\"\\nval/train: {:.8f}\".format(logs[\"val_loss\"] / logs[\"loss\"]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 96,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 0
    },
    "colab_type": "code",
    "id": "2kmG0LZUaYav",
    "outputId": "91222e97-c85d-405a-c99f-99a587684cf4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/20\n",
      "347/363 [===========================>..] - ETA: 0s - loss: 0.3578\n",
      "val/train: 1.01771465\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3580 - val_loss: 0.3644\n",
      "Epoch 2/20\n",
      "352/363 [============================>.] - ETA: 0s - loss: 0.3572\n",
      "val/train: 1.01663544\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3584 - val_loss: 0.3644\n",
      "Epoch 3/20\n",
      "361/363 [============================>.] - ETA: 0s - loss: 0.3573\n",
      "val/train: 1.01399864\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3586 - val_loss: 0.3636\n",
      "Epoch 4/20\n",
      "349/363 [===========================>..] - ETA: 0s - loss: 0.3582\n",
      "val/train: 1.02037507\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3565 - val_loss: 0.3637\n",
      "Epoch 5/20\n",
      "352/363 [============================>.] - ETA: 0s - loss: 0.3567\n",
      "val/train: 1.01573936\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3560 - val_loss: 0.3616\n",
      "Epoch 6/20\n",
      "348/363 [===========================>..] - ETA: 0s - loss: 0.3545\n",
      "val/train: 1.01985401\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3545 - val_loss: 0.3615\n",
      "Epoch 7/20\n",
      "346/363 [===========================>..] - ETA: 0s - loss: 0.3565\n",
      "val/train: 1.01919398\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3550 - val_loss: 0.3618\n",
      "Epoch 8/20\n",
      "349/363 [===========================>..] - ETA: 0s - loss: 0.3494\n",
      "val/train: 1.02520534\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3536 - val_loss: 0.3625\n",
      "Epoch 9/20\n",
      "345/363 [===========================>..] - ETA: 0s - loss: 0.3550\n",
      "val/train: 1.02003557\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3535 - val_loss: 0.3606\n",
      "Epoch 10/20\n",
      "346/363 [===========================>..] - ETA: 0s - loss: 0.3511\n",
      "val/train: 1.02079806\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3523 - val_loss: 0.3596\n",
      "Epoch 11/20\n",
      "347/363 [===========================>..] - ETA: 0s - loss: 0.3512\n",
      "val/train: 1.01960887\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3539 - val_loss: 0.3608\n",
      "Epoch 12/20\n",
      "348/363 [===========================>..] - ETA: 0s - loss: 0.3559\n",
      "val/train: 1.00812803\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3548 - val_loss: 0.3577\n",
      "Epoch 13/20\n",
      "355/363 [============================>.] - ETA: 0s - loss: 0.3498\n",
      "val/train: 1.02027386\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3506 - val_loss: 0.3577\n",
      "Epoch 14/20\n",
      "348/363 [===========================>..] - ETA: 0s - loss: 0.3488\n",
      "val/train: 1.02349571\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3497 - val_loss: 0.3579\n",
      "Epoch 15/20\n",
      "362/363 [============================>.] - ETA: 0s - loss: 0.3493\n",
      "val/train: 1.01802185\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3493 - val_loss: 0.3556\n",
      "Epoch 16/20\n",
      "344/363 [===========================>..] - ETA: 0s - loss: 0.3493\n",
      "val/train: 1.02205378\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3495 - val_loss: 0.3572\n",
      "Epoch 17/20\n",
      "352/363 [============================>.] - ETA: 0s - loss: 0.3477\n",
      "val/train: 1.01681623\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3493 - val_loss: 0.3552\n",
      "Epoch 18/20\n",
      "351/363 [============================>.] - ETA: 0s - loss: 0.3517\n",
      "val/train: 1.01502900\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3520 - val_loss: 0.3573\n",
      "Epoch 19/20\n",
      "345/363 [===========================>..] - ETA: 0s - loss: 0.3473\n",
      "val/train: 1.01943558\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3485 - val_loss: 0.3553\n",
      "Epoch 20/20\n",
      "355/363 [============================>.] - ETA: 0s - loss: 0.3474\n",
      "val/train: 1.02865854\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3484 - val_loss: 0.3584\n"
     ]
    }
   ],
   "source": [
    "val_train_ratio_cb = PrintValTrainRatioCallback()\n",
    "history = model.fit(X_train, y_train, epochs=20,\n",
    "                    validation_data=(X_valid, y_valid),\n",
    "                    callbacks=[val_train_ratio_cb])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "srJ1unTgaYax"
   },
   "source": [
    "# TensorBoard"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "owAZsZ_waYax"
   },
   "outputs": [],
   "source": [
    "import os\n",
    "root_logdir = os.path.join(os.curdir, \"my_logs\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4CH3O0F0aYaz"
   },
   "outputs": [],
   "source": [
    "def get_run_logdir():\n",
    "    import time\n",
    "    run_id = time.strftime(\"run_%Y_%m_%d-%H_%M_%S\")\n",
    "    return os.path.join(root_logdir, run_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "hglb0V88aYa0"
   },
   "outputs": [],
   "source": [
    "run_logdir = get_run_logdir()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 100,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "cn7Xn-hMaYa2",
    "outputId": "39a5ca11-f43f-4742-e346-3cacfdafc4f2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./my_logs/run_2020_06_05-12_27_07'"
      ]
     },
     "execution_count": 100,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "run_logdir"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "bspPk43IaYa4"
   },
   "outputs": [],
   "source": [
    "keras.backend.clear_session()\n",
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ZJYn5DYXaYa5"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, activation=\"relu\", input_shape=[8]),\n",
    "    keras.layers.Dense(30, activation=\"relu\"),\n",
    "    keras.layers.Dense(1)\n",
    "])    \n",
    "model.compile(loss=\"mse\", optimizer=keras.optimizers.SGD(lr=1e-3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "dYA35FmPaYa7",
    "outputId": "025bbf5a-56ae-43ea-960e-6fa0ec0f9a7f"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 2.0913 - val_loss: 0.9353\n",
      "Epoch 2/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.8252 - val_loss: 0.7311\n",
      "Epoch 3/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.7204 - val_loss: 0.6738\n",
      "Epoch 4/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.6631 - val_loss: 0.6278\n",
      "Epoch 5/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.6241 - val_loss: 0.5984\n",
      "Epoch 6/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.5951 - val_loss: 0.5725\n",
      "Epoch 7/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.5729 - val_loss: 0.5528\n",
      "Epoch 8/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.5520 - val_loss: 0.5351\n",
      "Epoch 9/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.5356 - val_loss: 0.5214\n",
      "Epoch 10/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.5219 - val_loss: 0.5090\n",
      "Epoch 11/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.5097 - val_loss: 0.4986\n",
      "Epoch 12/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4999 - val_loss: 0.4898\n",
      "Epoch 13/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4907 - val_loss: 0.4820\n",
      "Epoch 14/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4830 - val_loss: 0.4758\n",
      "Epoch 15/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4762 - val_loss: 0.4696\n",
      "Epoch 16/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4714 - val_loss: 0.4657\n",
      "Epoch 17/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4661 - val_loss: 0.4617\n",
      "Epoch 18/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4630 - val_loss: 0.4584\n",
      "Epoch 19/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4578 - val_loss: 0.4541\n",
      "Epoch 20/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4542 - val_loss: 0.4514\n",
      "Epoch 21/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4507 - val_loss: 0.4494\n",
      "Epoch 22/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4479 - val_loss: 0.4470\n",
      "Epoch 23/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4451 - val_loss: 0.4426\n",
      "Epoch 24/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4424 - val_loss: 0.4409\n",
      "Epoch 25/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4398 - val_loss: 0.4399\n",
      "Epoch 26/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4376 - val_loss: 0.4372\n",
      "Epoch 27/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4354 - val_loss: 0.4344\n",
      "Epoch 28/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4330 - val_loss: 0.4322\n",
      "Epoch 29/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.4308 - val_loss: 0.4304\n",
      "Epoch 30/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4294 - val_loss: 0.4286\n"
     ]
    }
   ],
   "source": [
    "tensorboard_cb = keras.callbacks.TensorBoard(run_logdir)\n",
    "history = model.fit(X_train, y_train, epochs = 30,\n",
    "                   validation_data = (X_valid, y_valid),\n",
    "                   callbacks=[tensorboard_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 104,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 820
    },
    "colab_type": "code",
    "id": "BkoQW-CqaYa9",
    "outputId": "21f4116b-a7ee-4088-acfd-769de6af99ae"
   },
   "outputs": [
    {
     "data": {
      "application/javascript": [
       "\n",
       "        (async () => {\n",
       "            const url = await google.colab.kernel.proxyPort(6006, {\"cache\": true});\n",
       "            const iframe = document.createElement('iframe');\n",
       "            iframe.src = url;\n",
       "            iframe.setAttribute('width', '100%');\n",
       "            iframe.setAttribute('height', '800');\n",
       "            iframe.setAttribute('frameborder', 0);\n",
       "            document.body.appendChild(iframe);\n",
       "        })();\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.Javascript object>"
      ]
     },
     "metadata": {
      "tags": []
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%load_ext tensorboard\n",
    "%tensorboard --logdir=./my_logs --port=6006"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "JxkbPx3RaYa_",
    "outputId": "f7fc81ab-d625-48e2-90a1-a39bb4a2535f"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./my_logs/run_2020_06_05-12_27_55'"
      ]
     },
     "execution_count": 105,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "run_logdir2 = get_run_logdir()\n",
    "run_logdir2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "lTKoiLiUaYbA"
   },
   "outputs": [],
   "source": [
    "keras.backend.clear_session()\n",
    "np.random.seed(42)\n",
    "tf.random.set_seed(42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "Fn3aLf9eaYbC"
   },
   "outputs": [],
   "source": [
    "model = keras.models.Sequential([\n",
    "    keras.layers.Dense(30, activation=\"relu\", input_shape=[8]),\n",
    "    keras.layers.Dense(30, activation=\"relu\"),\n",
    "    keras.layers.Dense(1)\n",
    "])    \n",
    "model.compile(loss=\"mse\", optimizer=keras.optimizers.SGD(lr=0.05))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "aQ3-GQyUaYbF",
    "outputId": "da796ed1-6061-4735-a3c1-ccec1fb22182"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.6990 - val_loss: 0.4774\n",
      "Epoch 2/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4820 - val_loss: 0.4097\n",
      "Epoch 3/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4450 - val_loss: 0.4178\n",
      "Epoch 4/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.4022 - val_loss: 0.4002\n",
      "Epoch 5/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.3928 - val_loss: 0.3712\n",
      "Epoch 6/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3733 - val_loss: 0.3935\n",
      "Epoch 7/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3785 - val_loss: 0.3501\n",
      "Epoch 8/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.3645 - val_loss: 0.3375\n",
      "Epoch 9/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3536 - val_loss: 0.3342\n",
      "Epoch 10/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3477 - val_loss: 0.3435\n",
      "Epoch 11/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.3421 - val_loss: 0.3391\n",
      "Epoch 12/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3320 - val_loss: 0.3113\n",
      "Epoch 13/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3268 - val_loss: 0.3251\n",
      "Epoch 14/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3227 - val_loss: 0.3273\n",
      "Epoch 15/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3223 - val_loss: 0.3145\n",
      "Epoch 16/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.3129 - val_loss: 0.3049\n",
      "Epoch 17/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3133 - val_loss: 0.3144\n",
      "Epoch 18/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3101 - val_loss: 0.2979\n",
      "Epoch 19/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3064 - val_loss: 0.2989\n",
      "Epoch 20/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3073 - val_loss: 0.3114\n",
      "Epoch 21/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3021 - val_loss: 0.3114\n",
      "Epoch 22/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3017 - val_loss: 0.3072\n",
      "Epoch 23/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.3010 - val_loss: 0.2949\n",
      "Epoch 24/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3048 - val_loss: 0.3533\n",
      "Epoch 25/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3121 - val_loss: 0.3238\n",
      "Epoch 26/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3036 - val_loss: 0.3086\n",
      "Epoch 27/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.2957 - val_loss: 0.3074\n",
      "Epoch 28/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.2950 - val_loss: 0.2933\n",
      "Epoch 29/30\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.2962 - val_loss: 0.2866\n",
      "Epoch 30/30\n",
      "363/363 [==============================] - 2s 4ms/step - loss: 0.2924 - val_loss: 0.3447\n"
     ]
    }
   ],
   "source": [
    "tensorboard_cb = keras.callbacks.TensorBoard(run_logdir2)\n",
    "history = model.fit(X_train, y_train, epochs=30,\n",
    "                    validation_data=(X_valid, y_valid),\n",
    "                    callbacks=[checkpoint_cb, tensorboard_cb])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "GrLiHwaPaYbH"
   },
   "outputs": [],
   "source": [
    "test_logdir = get_run_logdir()\n",
    "writer = tf.summary.create_file_writer(test_logdir)\n",
    "with writer.as_default():\n",
    "    for step in range(1, 1000 + 1):\n",
    "        tf.summary.scalar(\"my_scalar\", np.sin(step / 10), step=step)\n",
    "        data = (np.random.randn(100) + 2) * step / 100 # some random data\n",
    "        tf.summary.histogram(\"my_hist\", data, buckets=50, step=step)\n",
    "        images = np.random.rand(2, 32, 32, 3) # random 32×32 RGB images\n",
    "        tf.summary.image(\"my_images\", images * step / 1000, step=step)\n",
    "        texts = [\"The step is \" + str(step), \"Its square is \" + str(step**2)]\n",
    "        tf.summary.text(\"my_text\", texts, step=step)\n",
    "        sine_wave = tf.math.sin(tf.range(12000) / 48000 * 2 * np.pi * step)\n",
    "        audio = tf.reshape(tf.cast(sine_wave, tf.float32), [1, -1, 1])\n",
    "        tf.summary.audio(\"my_audio\", audio, sample_rate=48000, step=step)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "HxahnzLNaYbJ"
   },
   "source": [
    "## Fine-Tuning Neural Network Hyperparameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "E8cwhYeCaYbJ"
   },
   "outputs": [],
   "source": [
    "def build_model(n_hidden = 1, n_neurons = 30, learning_rate = 3e-3, input_shape=[8]):\n",
    "    model = keras.models.Sequential()\n",
    "    model.add(keras.layers.InputLayer(input_shape = input_shape))\n",
    "    for layer in range(n_hidden):\n",
    "        model.add(keras.layers.Dense(n_neurons, activation='relu'))\n",
    "    model.add(keras.layers.Dense(1))\n",
    "    optimizer = keras.optimizers.SGD(lr = learning_rate)\n",
    "    model.compile(loss = 'mse', optimizer=optimizer)\n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "1-G3KNTNaYbM"
   },
   "outputs": [],
   "source": [
    "keras_reg = keras.wrappers.scikit_learn.KerasRegressor(build_model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "YiN2y1kraYbO",
    "outputId": "0ef904b1-8c7d-49fa-e4d8-1eed810483ee"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 1.1234 - val_loss: 0.6957\n",
      "Epoch 2/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.7247 - val_loss: 0.5923\n",
      "Epoch 3/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5896 - val_loss: 0.5439\n",
      "Epoch 4/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5488 - val_loss: 0.5201\n",
      "Epoch 5/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5245 - val_loss: 0.4988\n",
      "Epoch 6/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.5071 - val_loss: 0.4873\n",
      "Epoch 7/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4941 - val_loss: 0.4757\n",
      "Epoch 8/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4833 - val_loss: 0.4666\n",
      "Epoch 9/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4752 - val_loss: 0.4604\n",
      "Epoch 10/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4681 - val_loss: 0.4544\n",
      "Epoch 11/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4626 - val_loss: 0.4505\n",
      "Epoch 12/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4569 - val_loss: 0.4491\n",
      "Epoch 13/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4535 - val_loss: 0.4423\n",
      "Epoch 14/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4493 - val_loss: 0.4392\n",
      "Epoch 15/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4463 - val_loss: 0.4373\n",
      "Epoch 16/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4425 - val_loss: 0.4339\n",
      "Epoch 17/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4400 - val_loss: 0.4329\n",
      "Epoch 18/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4366 - val_loss: 0.4301\n",
      "Epoch 19/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4335 - val_loss: 0.4288\n",
      "Epoch 20/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4309 - val_loss: 0.4250\n",
      "Epoch 21/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4287 - val_loss: 0.4227\n",
      "Epoch 22/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4263 - val_loss: 0.4218\n",
      "Epoch 23/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4237 - val_loss: 0.4212\n",
      "Epoch 24/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4219 - val_loss: 0.4189\n",
      "Epoch 25/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4196 - val_loss: 0.4177\n",
      "Epoch 26/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4176 - val_loss: 0.4155\n",
      "Epoch 27/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4156 - val_loss: 0.4150\n",
      "Epoch 28/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4131 - val_loss: 0.4138\n",
      "Epoch 29/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4115 - val_loss: 0.4109\n",
      "Epoch 30/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4099 - val_loss: 0.4105\n",
      "Epoch 31/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4078 - val_loss: 0.4073\n",
      "Epoch 32/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4066 - val_loss: 0.4071\n",
      "Epoch 33/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4044 - val_loss: 0.4049\n",
      "Epoch 34/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4033 - val_loss: 0.4038\n",
      "Epoch 35/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4014 - val_loss: 0.4034\n",
      "Epoch 36/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.4001 - val_loss: 0.4024\n",
      "Epoch 37/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3987 - val_loss: 0.3999\n",
      "Epoch 38/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3971 - val_loss: 0.3986\n",
      "Epoch 39/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3962 - val_loss: 0.3990\n",
      "Epoch 40/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3948 - val_loss: 0.3976\n",
      "Epoch 41/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3934 - val_loss: 0.3978\n",
      "Epoch 42/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3923 - val_loss: 0.3953\n",
      "Epoch 43/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3906 - val_loss: 0.3963\n",
      "Epoch 44/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3898 - val_loss: 0.3929\n",
      "Epoch 45/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3883 - val_loss: 0.3925\n",
      "Epoch 46/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3873 - val_loss: 0.3909\n",
      "Epoch 47/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3864 - val_loss: 0.3901\n",
      "Epoch 48/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3855 - val_loss: 0.3913\n",
      "Epoch 49/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3848 - val_loss: 0.3891\n",
      "Epoch 50/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3833 - val_loss: 0.3882\n",
      "Epoch 51/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3829 - val_loss: 0.3870\n",
      "Epoch 52/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3812 - val_loss: 0.3856\n",
      "Epoch 53/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3808 - val_loss: 0.3868\n",
      "Epoch 54/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3800 - val_loss: 0.3861\n",
      "Epoch 55/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3814 - val_loss: 0.3838\n",
      "Epoch 56/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3785 - val_loss: 0.3837\n",
      "Epoch 57/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3786 - val_loss: 0.3837\n",
      "Epoch 58/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3770 - val_loss: 0.3823\n",
      "Epoch 59/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3764 - val_loss: 0.3824\n",
      "Epoch 60/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3758 - val_loss: 0.3825\n",
      "Epoch 61/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3755 - val_loss: 0.3799\n",
      "Epoch 62/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3748 - val_loss: 0.3799\n",
      "Epoch 63/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3739 - val_loss: 0.3793\n",
      "Epoch 64/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3734 - val_loss: 0.3789\n",
      "Epoch 65/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3722 - val_loss: 0.3815\n",
      "Epoch 66/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3719 - val_loss: 0.3782\n",
      "Epoch 67/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3714 - val_loss: 0.3795\n",
      "Epoch 68/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3709 - val_loss: 0.3777\n",
      "Epoch 69/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3704 - val_loss: 0.3783\n",
      "Epoch 70/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3694 - val_loss: 0.3765\n",
      "Epoch 71/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3687 - val_loss: 0.3772\n",
      "Epoch 72/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3673 - val_loss: 0.3788\n",
      "Epoch 73/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3677 - val_loss: 0.3744\n",
      "Epoch 74/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3672 - val_loss: 0.3742\n",
      "Epoch 75/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3662 - val_loss: 0.3746\n",
      "Epoch 76/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3654 - val_loss: 0.3745\n",
      "Epoch 77/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3661 - val_loss: 0.3731\n",
      "Epoch 78/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3652 - val_loss: 0.3730\n",
      "Epoch 79/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3641 - val_loss: 0.3715\n",
      "Epoch 80/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3630 - val_loss: 0.3720\n",
      "Epoch 81/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3630 - val_loss: 0.3708\n",
      "Epoch 82/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3622 - val_loss: 0.3711\n",
      "Epoch 83/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3614 - val_loss: 0.3698\n",
      "Epoch 84/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3608 - val_loss: 0.3707\n",
      "Epoch 85/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3611 - val_loss: 0.3703\n",
      "Epoch 86/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3603 - val_loss: 0.3702\n",
      "Epoch 87/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3617 - val_loss: 0.3695\n",
      "Epoch 88/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3599 - val_loss: 0.3716\n",
      "Epoch 89/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3582 - val_loss: 0.3692\n",
      "Epoch 90/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3576 - val_loss: 0.3698\n",
      "Epoch 91/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3594 - val_loss: 0.3691\n",
      "Epoch 92/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3579 - val_loss: 0.3673\n",
      "Epoch 93/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3569 - val_loss: 0.3661\n",
      "Epoch 94/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3559 - val_loss: 0.3664\n",
      "Epoch 95/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3555 - val_loss: 0.3676\n",
      "Epoch 96/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3540 - val_loss: 0.3643\n",
      "Epoch 97/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3535 - val_loss: 0.3659\n",
      "Epoch 98/100\n",
      "363/363 [==============================] - 1s 4ms/step - loss: 0.3533 - val_loss: 0.3663\n",
      "Epoch 99/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3531 - val_loss: 0.3651\n",
      "Epoch 100/100\n",
      "363/363 [==============================] - 1s 3ms/step - loss: 0.3519 - val_loss: 0.3670\n",
      "162/162 [==============================] - 0s 2ms/step - loss: 0.3705\n"
     ]
    }
   ],
   "source": [
    "keras_reg.fit(X_train, y_train, epochs = 100,\n",
    "             validation_data = (X_valid, y_valid),\n",
    "             callbacks = [keras.callbacks.EarlyStopping(patience=10)])\n",
    "mse_test = keras_reg.score(X_test, y_test)\n",
    "y_pred =  keras_reg.predict(X_new)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "eGJIwLMWaYbQ"
   },
   "outputs": [],
   "source": [
    "np.random.seed(259)\n",
    "tf.random.set_seed(259)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "UBXx467naYbR"
   },
   "outputs": [],
   "source": [
    "from scipy.stats import reciprocal\n",
    "from sklearn.model_selection import RandomizedSearchCV"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "VB3BmnavaYbT"
   },
   "outputs": [],
   "source": [
    "param_distribs = {\n",
    "    'n_hidden' : [0,1,2,3],\n",
    "    'n_neurons': np.arange(1,100),\n",
    "    'learning_rate': reciprocal(3e-4, 3e-2)\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 116,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1000
    },
    "colab_type": "code",
    "id": "lu18vPShaYbU",
    "outputId": "774682da-2087-4090-eb7f-d7fd550251df"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 10 candidates, totalling 30 fits\n",
      "[CV] learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28 ....\n",
      "Epoch 1/100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "242/242 [==============================] - 1s 4ms/step - loss: 1.2629 - val_loss: 0.6580\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5163 - val_loss: 0.5086\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4714 - val_loss: 0.4663\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4529 - val_loss: 0.4447\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4267 - val_loss: 0.4325\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4145 - val_loss: 0.4211\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4016 - val_loss: 0.4087\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3932 - val_loss: 0.4030\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3852 - val_loss: 0.4012\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3803 - val_loss: 0.3914\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3772 - val_loss: 0.3855\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3701 - val_loss: 0.3827\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3665 - val_loss: 0.3866\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3637 - val_loss: 0.3743\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3591 - val_loss: 0.3689\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3569 - val_loss: 0.3634\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3536 - val_loss: 0.3629\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3504 - val_loss: 0.3660\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3475 - val_loss: 0.3645\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3460 - val_loss: 0.3532\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3432 - val_loss: 0.3563\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3403 - val_loss: 0.3549\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3407 - val_loss: 0.3464\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3370 - val_loss: 0.3499\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3354 - val_loss: 0.3503\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3334 - val_loss: 0.3466\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3328 - val_loss: 0.3714\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3323 - val_loss: 0.3511\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3296 - val_loss: 0.3366\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3287 - val_loss: 0.3386\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3251 - val_loss: 0.3650\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3242 - val_loss: 0.3307\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3226 - val_loss: 0.3272\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3227 - val_loss: 0.3278\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3228 - val_loss: 0.3258\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3222 - val_loss: 0.3338\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3180 - val_loss: 0.3231\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3161 - val_loss: 0.3363\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3172 - val_loss: 0.3244\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3164 - val_loss: 0.3386\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3174 - val_loss: 0.3228\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3150 - val_loss: 0.3348\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3135 - val_loss: 0.3146\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3115 - val_loss: 0.3413\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3143 - val_loss: 0.3155\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3110 - val_loss: 0.3468\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3111 - val_loss: 0.3144\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3086 - val_loss: 0.3111\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3060 - val_loss: 0.3144\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3078 - val_loss: 0.3119\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3052 - val_loss: 0.3171\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3043 - val_loss: 0.3561\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3046 - val_loss: 0.3166\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3039 - val_loss: 0.3211\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3023 - val_loss: 0.3143\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3022 - val_loss: 0.3114\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3024 - val_loss: 0.3302\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3002 - val_loss: 0.3226\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3470\n",
      "[CV]  learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28, total=  57.5s\n",
      "[CV] learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28 ....\n",
      "Epoch 1/100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:   57.5s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "242/242 [==============================] - 1s 4ms/step - loss: 1.2126 - val_loss: 0.6250\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5717 - val_loss: 0.5022\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4991 - val_loss: 0.4611\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4674 - val_loss: 0.4459\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4486 - val_loss: 0.4297\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4343 - val_loss: 0.4215\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 5ms/step - loss: 0.4211 - val_loss: 0.4095\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4127 - val_loss: 0.4027\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4023 - val_loss: 0.3974\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3963 - val_loss: 0.3926\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3880 - val_loss: 0.3837\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3815 - val_loss: 0.3933\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3762 - val_loss: 0.3866\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3716 - val_loss: 0.3715\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3668 - val_loss: 0.3680\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3648 - val_loss: 0.3725\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3591 - val_loss: 0.3664\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3567 - val_loss: 0.3672\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3563 - val_loss: 0.3570\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3548 - val_loss: 0.4072\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4481 - val_loss: 0.3694\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3551 - val_loss: 0.3526\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3480 - val_loss: 0.3479\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3446 - val_loss: 0.3553\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3409 - val_loss: 0.3488\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3372 - val_loss: 0.3454\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3350 - val_loss: 0.3398\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3329 - val_loss: 0.3570\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3315 - val_loss: 0.3466\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3298 - val_loss: 0.3317\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3263 - val_loss: 0.3327\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3253 - val_loss: 0.3327\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3223 - val_loss: 0.3280\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3202 - val_loss: 0.3342\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3206 - val_loss: 0.3308\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3197 - val_loss: 0.3275\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3166 - val_loss: 0.3256\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3142 - val_loss: 0.3282\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3119 - val_loss: 0.3291\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3152 - val_loss: 0.3213\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3113 - val_loss: 0.3179\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3105 - val_loss: 0.3147\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3091 - val_loss: 0.3190\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3093 - val_loss: 0.3229\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3080 - val_loss: 0.3124\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3079 - val_loss: 0.3295\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3054 - val_loss: 0.3187\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3059 - val_loss: 0.3185\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3027 - val_loss: 0.3144\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3039 - val_loss: 0.3153\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3019 - val_loss: 0.3184\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3026 - val_loss: 0.3310\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3001 - val_loss: 0.3136\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3006 - val_loss: 0.3110\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2990 - val_loss: 0.3068\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2982 - val_loss: 0.3261\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2974 - val_loss: 0.3149\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2973 - val_loss: 0.3024\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2964 - val_loss: 0.3121\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2960 - val_loss: 0.3010\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2931 - val_loss: 0.3652\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2959 - val_loss: 0.3067\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2928 - val_loss: 0.3032\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2938 - val_loss: 0.3150\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2922 - val_loss: 0.3027\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2930 - val_loss: 0.2990\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2896 - val_loss: 0.3064\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2909 - val_loss: 0.2978\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2900 - val_loss: 0.3038\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2870 - val_loss: 0.3047\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2895 - val_loss: 0.2947\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2892 - val_loss: 0.2990\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2872 - val_loss: 0.3021\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2882 - val_loss: 0.2984\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2853 - val_loss: 0.3097\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2872 - val_loss: 0.2935\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2838 - val_loss: 0.2979\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2847 - val_loss: 0.2965\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2809 - val_loss: 0.2956\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2826 - val_loss: 0.3244\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2835 - val_loss: 0.3082\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2835 - val_loss: 0.2895\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2804 - val_loss: 0.3379\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2837 - val_loss: 0.2900\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2801 - val_loss: 0.3050\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2796 - val_loss: 0.2890\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2788 - val_loss: 0.2972\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2801 - val_loss: 0.3036\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2778 - val_loss: 0.3234\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2791 - val_loss: 0.3021\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2798 - val_loss: 0.2912\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2796 - val_loss: 0.2856\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2782 - val_loss: 0.2961\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2771 - val_loss: 0.2936\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2780 - val_loss: 0.2888\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2778 - val_loss: 0.3028\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2750 - val_loss: 0.2886\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2749 - val_loss: 0.2895\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2750 - val_loss: 0.2885\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2751 - val_loss: 0.2928\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3070\n",
      "[CV]  learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28, total= 1.6min\n",
      "[CV] learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9556 - val_loss: 0.6703\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5977 - val_loss: 0.5402\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5321 - val_loss: 0.4982\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4940 - val_loss: 0.4720\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4701 - val_loss: 0.4516\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4502 - val_loss: 0.4339\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4384 - val_loss: 0.4269\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4259 - val_loss: 0.4219\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4167 - val_loss: 0.4094\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4098 - val_loss: 0.4105\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3992 - val_loss: 0.3995\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3933 - val_loss: 0.3836\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3897 - val_loss: 0.3985\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3817 - val_loss: 0.3715\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3745 - val_loss: 0.3664\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3693 - val_loss: 0.3718\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3645 - val_loss: 0.3593\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3602 - val_loss: 0.3658\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3581 - val_loss: 0.3550\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3539 - val_loss: 0.3531\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3499 - val_loss: 0.3580\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3460 - val_loss: 0.3621\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3430 - val_loss: 0.3413\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3429 - val_loss: 0.3419\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3382 - val_loss: 0.3402\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3333 - val_loss: 0.3337\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3314 - val_loss: 0.3356\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3303 - val_loss: 0.3362\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3264 - val_loss: 0.3363\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3258 - val_loss: 0.3294\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3244 - val_loss: 0.3259\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3223 - val_loss: 0.3248\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3250 - val_loss: 0.3338\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3200 - val_loss: 0.3274\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3195 - val_loss: 0.3300\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3171 - val_loss: 0.3210\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3148 - val_loss: 0.3253\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3123 - val_loss: 0.3375\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3124 - val_loss: 0.3283\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3106 - val_loss: 0.3168\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3126 - val_loss: 0.3265\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3084 - val_loss: 0.3198\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3088 - val_loss: 0.3376\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3048 - val_loss: 0.3295\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3063 - val_loss: 0.3236\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3043 - val_loss: 0.3275\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3050 - val_loss: 0.3187\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3028 - val_loss: 0.3105\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3057 - val_loss: 0.3261\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2986 - val_loss: 0.3561\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3027 - val_loss: 0.3558\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2999 - val_loss: 0.3168\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3001 - val_loss: 0.3157\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2986 - val_loss: 0.3130\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2976 - val_loss: 0.3258\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2982 - val_loss: 0.3252\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2945 - val_loss: 0.3086\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2952 - val_loss: 0.3286\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2938 - val_loss: 0.3207\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2953 - val_loss: 0.3147\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2910 - val_loss: 0.3486\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2926 - val_loss: 0.3186\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2919 - val_loss: 0.3010\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2902 - val_loss: 0.3314\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2906 - val_loss: 0.3144\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2906 - val_loss: 0.3033\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2888 - val_loss: 0.3170\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2873 - val_loss: 0.3075\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2898 - val_loss: 0.3149\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2883 - val_loss: 0.3183\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2885 - val_loss: 0.3078\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2886 - val_loss: 0.3053\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2859 - val_loss: 0.3248\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3388\n",
      "[CV]  learning_rate=0.007203790442032935, n_hidden=3, n_neurons=28, total= 1.2min\n",
      "[CV] learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 3.4485 - val_loss: 1.9271\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.3972 - val_loss: 1.1131\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9779 - val_loss: 0.8794\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8039 - val_loss: 0.7475\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7050 - val_loss: 0.6745\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6491 - val_loss: 0.6314\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6167 - val_loss: 0.6054\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5953 - val_loss: 0.5870\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5788 - val_loss: 0.5726\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5652 - val_loss: 0.5612\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5538 - val_loss: 0.5493\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5436 - val_loss: 0.5388\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5343 - val_loss: 0.5296\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5256 - val_loss: 0.5219\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5180 - val_loss: 0.5141\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5110 - val_loss: 0.5074\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5044 - val_loss: 0.5006\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4985 - val_loss: 0.4951\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4931 - val_loss: 0.4897\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4881 - val_loss: 0.4850\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4838 - val_loss: 0.4812\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4800 - val_loss: 0.4772\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4763 - val_loss: 0.4736\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4731 - val_loss: 0.4708\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4701 - val_loss: 0.4682\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4674 - val_loss: 0.4657\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4648 - val_loss: 0.4628\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4625 - val_loss: 0.4607\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4605 - val_loss: 0.4590\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4586 - val_loss: 0.4580\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4569 - val_loss: 0.4561\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4553 - val_loss: 0.4548\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4538 - val_loss: 0.4532\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4523 - val_loss: 0.4523\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4509 - val_loss: 0.4514\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4497 - val_loss: 0.4500\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4485 - val_loss: 0.4494\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4475 - val_loss: 0.4483\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4465 - val_loss: 0.4470\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4454 - val_loss: 0.4461\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4444 - val_loss: 0.4461\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4437 - val_loss: 0.4455\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4427 - val_loss: 0.4439\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4420 - val_loss: 0.4441\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4414 - val_loss: 0.4432\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4406 - val_loss: 0.4426\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4399 - val_loss: 0.4419\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4393 - val_loss: 0.4408\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4385 - val_loss: 0.4406\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4379 - val_loss: 0.4399\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4371 - val_loss: 0.4395\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4365 - val_loss: 0.4389\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4358 - val_loss: 0.4376\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4353 - val_loss: 0.4373\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4348 - val_loss: 0.4372\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4342 - val_loss: 0.4364\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4336 - val_loss: 0.4356\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4330 - val_loss: 0.4357\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4325 - val_loss: 0.4353\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4320 - val_loss: 0.4340\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4316 - val_loss: 0.4336\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4310 - val_loss: 0.4334\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4306 - val_loss: 0.4328\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4302 - val_loss: 0.4323\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4296 - val_loss: 0.4323\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4292 - val_loss: 0.4319\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4289 - val_loss: 0.4314\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4283 - val_loss: 0.4308\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4278 - val_loss: 0.4303\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4275 - val_loss: 0.4305\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4271 - val_loss: 0.4297\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4266 - val_loss: 0.4294\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4262 - val_loss: 0.4286\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4258 - val_loss: 0.4283\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4254 - val_loss: 0.4283\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4249 - val_loss: 0.4278\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4246 - val_loss: 0.4273\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4240 - val_loss: 0.4272\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4235 - val_loss: 0.4268\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4233 - val_loss: 0.4272\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4231 - val_loss: 0.4257\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4226 - val_loss: 0.4255\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4221 - val_loss: 0.4259\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4220 - val_loss: 0.4245\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4215 - val_loss: 0.4243\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4212 - val_loss: 0.4242\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4208 - val_loss: 0.4239\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4205 - val_loss: 0.4233\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4201 - val_loss: 0.4229\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4197 - val_loss: 0.4227\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4192 - val_loss: 0.4220\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4191 - val_loss: 0.4220\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4186 - val_loss: 0.4221\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4182 - val_loss: 0.4215\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4179 - val_loss: 0.4205\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4176 - val_loss: 0.4204\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4171 - val_loss: 0.4205\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4168 - val_loss: 0.4201\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4164 - val_loss: 0.4195\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4161 - val_loss: 0.4186\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.4380\n",
      "[CV]  learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6, total= 1.6min\n",
      "[CV] learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 3.6981 - val_loss: 2.1644\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.5269 - val_loss: 1.1730\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.0266 - val_loss: 0.9495\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8883 - val_loss: 0.8472\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8172 - val_loss: 0.7855\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7736 - val_loss: 0.7459\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7441 - val_loss: 0.7174\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7214 - val_loss: 0.6956\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7031 - val_loss: 0.6773\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6868 - val_loss: 0.6611\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6723 - val_loss: 0.6471\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6594 - val_loss: 0.6343\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6476 - val_loss: 0.6229\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6367 - val_loss: 0.6130\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6271 - val_loss: 0.6038\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6187 - val_loss: 0.5961\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6108 - val_loss: 0.5889\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6033 - val_loss: 0.5822\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5964 - val_loss: 0.5759\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5895 - val_loss: 0.5703\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5834 - val_loss: 0.5653\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5782 - val_loss: 0.5596\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5727 - val_loss: 0.5554\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5680 - val_loss: 0.5501\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5633 - val_loss: 0.5461\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5590 - val_loss: 0.5419\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5546 - val_loss: 0.5378\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5503 - val_loss: 0.5338\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5462 - val_loss: 0.5297\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5422 - val_loss: 0.5260\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5384 - val_loss: 0.5220\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5350 - val_loss: 0.5183\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5314 - val_loss: 0.5147\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5279 - val_loss: 0.5115\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5246 - val_loss: 0.5086\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5216 - val_loss: 0.5049\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5187 - val_loss: 0.5022\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5158 - val_loss: 0.4994\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5131 - val_loss: 0.4965\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5105 - val_loss: 0.4940\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5080 - val_loss: 0.4914\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5055 - val_loss: 0.4892\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5034 - val_loss: 0.4869\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5014 - val_loss: 0.4848\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4993 - val_loss: 0.4829\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4972 - val_loss: 0.4810\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4953 - val_loss: 0.4793\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4936 - val_loss: 0.4776\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4919 - val_loss: 0.4761\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4906 - val_loss: 0.4752\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4890 - val_loss: 0.4736\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4874 - val_loss: 0.4726\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4862 - val_loss: 0.4709\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4849 - val_loss: 0.4698\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4836 - val_loss: 0.4686\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4825 - val_loss: 0.4674\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4810 - val_loss: 0.4666\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4803 - val_loss: 0.4655\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4792 - val_loss: 0.4643\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4780 - val_loss: 0.4630\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4770 - val_loss: 0.4625\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4761 - val_loss: 0.4615\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4751 - val_loss: 0.4604\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4742 - val_loss: 0.4597\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4733 - val_loss: 0.4585\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4722 - val_loss: 0.4575\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4710 - val_loss: 0.4579\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4706 - val_loss: 0.4562\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4698 - val_loss: 0.4552\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4689 - val_loss: 0.4543\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4681 - val_loss: 0.4533\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4672 - val_loss: 0.4527\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4663 - val_loss: 0.4517\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4655 - val_loss: 0.4508\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4647 - val_loss: 0.4506\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4639 - val_loss: 0.4498\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4631 - val_loss: 0.4489\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4623 - val_loss: 0.4483\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4614 - val_loss: 0.4478\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4610 - val_loss: 0.4465\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4602 - val_loss: 0.4459\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4595 - val_loss: 0.4454\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4588 - val_loss: 0.4445\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4580 - val_loss: 0.4442\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4573 - val_loss: 0.4433\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4565 - val_loss: 0.4426\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4559 - val_loss: 0.4421\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4551 - val_loss: 0.4409\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4544 - val_loss: 0.4411\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4539 - val_loss: 0.4397\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4530 - val_loss: 0.4386\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4525 - val_loss: 0.4384\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4519 - val_loss: 0.4376\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4512 - val_loss: 0.4371\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4502 - val_loss: 0.4359\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4496 - val_loss: 0.4363\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4491 - val_loss: 0.4354\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4484 - val_loss: 0.4349\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4478 - val_loss: 0.4334\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4472 - val_loss: 0.4329\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.4358\n",
      "[CV]  learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6, total= 1.6min\n",
      "[CV] learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 2.2016 - val_loss: 1.0244\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9549 - val_loss: 0.8981\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8718 - val_loss: 0.8512\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8234 - val_loss: 0.8168\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7898 - val_loss: 0.7884\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7621 - val_loss: 0.7650\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7397 - val_loss: 0.7448\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7195 - val_loss: 0.7278\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7011 - val_loss: 0.7116\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6846 - val_loss: 0.6974\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6698 - val_loss: 0.6855\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6564 - val_loss: 0.6733\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6437 - val_loss: 0.6629\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6320 - val_loss: 0.6535\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6208 - val_loss: 0.6421\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6097 - val_loss: 0.6316\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5996 - val_loss: 0.6262\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5897 - val_loss: 0.6183\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5798 - val_loss: 0.6101\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5705 - val_loss: 0.6033\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5617 - val_loss: 0.5956\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5530 - val_loss: 0.5913\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5452 - val_loss: 0.5837\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5380 - val_loss: 0.5759\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5307 - val_loss: 0.5715\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5250 - val_loss: 0.5676\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5193 - val_loss: 0.5646\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5142 - val_loss: 0.5607\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5095 - val_loss: 0.5576\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5050 - val_loss: 0.5523\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5012 - val_loss: 0.5498\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4975 - val_loss: 0.5473\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4940 - val_loss: 0.5451\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4904 - val_loss: 0.5395\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4873 - val_loss: 0.5390\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4840 - val_loss: 0.5330\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4818 - val_loss: 0.5308\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4786 - val_loss: 0.5325\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4769 - val_loss: 0.5296\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4741 - val_loss: 0.5258\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4719 - val_loss: 0.5268\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4701 - val_loss: 0.5224\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4678 - val_loss: 0.5182\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4660 - val_loss: 0.5178\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4642 - val_loss: 0.5160\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4626 - val_loss: 0.5138\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4606 - val_loss: 0.5199\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4600 - val_loss: 0.5135\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4585 - val_loss: 0.5099\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4571 - val_loss: 0.5125\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4561 - val_loss: 0.5053\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4547 - val_loss: 0.5039\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4538 - val_loss: 0.5021\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4530 - val_loss: 0.5000\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4517 - val_loss: 0.4992\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4508 - val_loss: 0.4983\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4497 - val_loss: 0.4948\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4492 - val_loss: 0.4971\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4482 - val_loss: 0.4912\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4475 - val_loss: 0.4903\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4466 - val_loss: 0.4913\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4457 - val_loss: 0.4879\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4450 - val_loss: 0.4862\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4446 - val_loss: 0.4881\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 5ms/step - loss: 0.4439 - val_loss: 0.4860\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4431 - val_loss: 0.4833\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4424 - val_loss: 0.4814\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4415 - val_loss: 0.4784\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4414 - val_loss: 0.4767\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4407 - val_loss: 0.4755\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4401 - val_loss: 0.4752\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4394 - val_loss: 0.4770\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4389 - val_loss: 0.4765\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4387 - val_loss: 0.4726\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4379 - val_loss: 0.4735\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4374 - val_loss: 0.4733\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4368 - val_loss: 0.4726\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4360 - val_loss: 0.4708\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4359 - val_loss: 0.4717\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4355 - val_loss: 0.4712\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4348 - val_loss: 0.4705\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4342 - val_loss: 0.4681\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4340 - val_loss: 0.4701\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4335 - val_loss: 0.4690\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4328 - val_loss: 0.4704\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4325 - val_loss: 0.4678\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4322 - val_loss: 0.4643\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4314 - val_loss: 0.4663\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4312 - val_loss: 0.4653\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4308 - val_loss: 0.4654\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4305 - val_loss: 0.4642\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4298 - val_loss: 0.4628\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4293 - val_loss: 0.4610\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4291 - val_loss: 0.4623\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4286 - val_loss: 0.4599\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4281 - val_loss: 0.4616\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4280 - val_loss: 0.4595\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4275 - val_loss: 0.4576\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4270 - val_loss: 0.4592\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4268 - val_loss: 0.4578\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.6047\n",
      "[CV]  learning_rate=0.0010218620660022116, n_hidden=2, n_neurons=6, total= 1.6min\n",
      "[CV] learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 2.7959 - val_loss: 0.7111\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 1.3506 - val_loss: 0.6175\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7912 - val_loss: 0.5527\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6694 - val_loss: 0.5335\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5715 - val_loss: 0.5228\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5424 - val_loss: 0.5210\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5228 - val_loss: 0.5181\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5222 - val_loss: 0.5161\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5203 - val_loss: 0.5164\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5209 - val_loss: 0.5131\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5295 - val_loss: 0.5157\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5247 - val_loss: 0.5315\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5183 - val_loss: 0.5134\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5224 - val_loss: 0.5198\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5476 - val_loss: 0.5195\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5434 - val_loss: 0.5128\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5298 - val_loss: 0.5147\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5193 - val_loss: 0.5155\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5486 - val_loss: 0.5178\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5221 - val_loss: 0.5281\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6415 - val_loss: 0.5294\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5202 - val_loss: 0.5211\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5250 - val_loss: 0.5133\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5180 - val_loss: 0.5223\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5186 - val_loss: 0.5135\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5191 - val_loss: 0.5136\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.5561\n",
      "[CV]  learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82, total=  22.7s\n",
      "[CV] learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.4779 - val_loss: 0.5332\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5712 - val_loss: 0.5260\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5601 - val_loss: 0.5184\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5460 - val_loss: 0.5161\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5432 - val_loss: 0.5148\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5419 - val_loss: 0.5144\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5415 - val_loss: 0.5137\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5434 - val_loss: 0.5136\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5416 - val_loss: 0.5137\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5417 - val_loss: 0.5176\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5780 - val_loss: 0.5221\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5832 - val_loss: 0.5121\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5401 - val_loss: 0.5149\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5419 - val_loss: 0.5234\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6136 - val_loss: 0.5282\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6302 - val_loss: 0.5183\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5958 - val_loss: 0.5231\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5426 - val_loss: 0.5273\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6741 - val_loss: 0.5331\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5451 - val_loss: 0.5587\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8706 - val_loss: 0.5564\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5546 - val_loss: 0.5611\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.5168\n",
      "[CV]  learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82, total=  19.3s\n",
      "[CV] learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.3445 - val_loss: 0.7443\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6108 - val_loss: 0.6843\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5745 - val_loss: 0.6692\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5721 - val_loss: 0.6342\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5568 - val_loss: 0.6168\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5467 - val_loss: 0.6412\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5509 - val_loss: 0.5956\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5425 - val_loss: 0.5887\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5466 - val_loss: 0.5864\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5379 - val_loss: 0.5811\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5370 - val_loss: 0.5806\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5399 - val_loss: 0.5766\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5392 - val_loss: 0.5739\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5411 - val_loss: 0.5704\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5358 - val_loss: 0.5740\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5441 - val_loss: 0.5668\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5351 - val_loss: 0.5909\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5378 - val_loss: 0.5688\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5402 - val_loss: 0.5687\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5521 - val_loss: 0.5715\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5397 - val_loss: 0.5707\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5366 - val_loss: 0.5795\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5389 - val_loss: 0.5679\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5366 - val_loss: 0.5649\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5493 - val_loss: 0.5656\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5408 - val_loss: 0.5670\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5490 - val_loss: 0.5730\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5368 - val_loss: 0.5690\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5467 - val_loss: 0.5716\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5338 - val_loss: 0.5679\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5452 - val_loss: 0.5694\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5425 - val_loss: 0.5689\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5350 - val_loss: 0.5729\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5415 - val_loss: 0.5643\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5355 - val_loss: 0.5649\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5415 - val_loss: 0.5658\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5382 - val_loss: 0.5655\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5426 - val_loss: 0.5670\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5386 - val_loss: 0.5649\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5431 - val_loss: 0.5703\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5330 - val_loss: 0.5689\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5417 - val_loss: 0.5672\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5416 - val_loss: 0.5643\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5383 - val_loss: 0.5742\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5413 - val_loss: 0.5773\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5379 - val_loss: 0.5726\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5393 - val_loss: 0.5707\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5406 - val_loss: 0.5658\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5367 - val_loss: 0.6068\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5434 - val_loss: 0.5662\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5357 - val_loss: 0.5681\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5391 - val_loss: 0.5679\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5490 - val_loss: 0.5643\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.7766\n",
      "[CV]  learning_rate=0.007837571489520578, n_hidden=0, n_neurons=82, total=  45.7s\n",
      "[CV] learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7895 - val_loss: 0.6331\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6132 - val_loss: 0.5751\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4894 - val_loss: 0.4594\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4250 - val_loss: 0.4128\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4008 - val_loss: 0.4111\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3937 - val_loss: 0.3984\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3850 - val_loss: 0.3898\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3810 - val_loss: 0.3851\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3828 - val_loss: 0.3902\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3700 - val_loss: 0.3797\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3651 - val_loss: 0.3713\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3545 - val_loss: 0.3678\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3528 - val_loss: 0.3778\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3540 - val_loss: 0.3571\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3429 - val_loss: 0.3533\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3531 - val_loss: 0.3502\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3394 - val_loss: 0.3566\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3380 - val_loss: 0.3594\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3469 - val_loss: 0.3553\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3435 - val_loss: 0.3436\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3357 - val_loss: 0.3440\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3275 - val_loss: 0.3463\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3263 - val_loss: 0.3329\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3221 - val_loss: 0.3433\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3199 - val_loss: 0.3403\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3189 - val_loss: 0.3321\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3192 - val_loss: 0.3572\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3142 - val_loss: 0.3447\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3114 - val_loss: 0.3197\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3123 - val_loss: 0.3247\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3072 - val_loss: 0.3578\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3060 - val_loss: 0.3158\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3054 - val_loss: 0.3135\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3051 - val_loss: 0.3129\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3019 - val_loss: 0.3151\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3026 - val_loss: 0.3209\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2989 - val_loss: 0.3135\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3006 - val_loss: 0.3274\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2987 - val_loss: 0.3088\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2949 - val_loss: 0.3314\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2954 - val_loss: 0.3035\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2949 - val_loss: 0.3250\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2935 - val_loss: 0.3000\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2917 - val_loss: 0.3219\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2962 - val_loss: 0.3027\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2943 - val_loss: 0.5157\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3150 - val_loss: 0.3022\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2999 - val_loss: 0.2953\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2881 - val_loss: 0.2999\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2893 - val_loss: 0.3012\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2858 - val_loss: 0.3039\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2859 - val_loss: 0.3535\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2850 - val_loss: 0.3224\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2838 - val_loss: 0.3046\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2844 - val_loss: 0.3004\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2813 - val_loss: 0.2987\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2828 - val_loss: 0.3268\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2814 - val_loss: 0.3114\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3285\n",
      "[CV]  learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77, total=  55.8s\n",
      "[CV] learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9083 - val_loss: 0.6158\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5582 - val_loss: 0.4536\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4583 - val_loss: 0.4289\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4367 - val_loss: 0.4157\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4178 - val_loss: 0.4040\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4066 - val_loss: 0.3924\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3966 - val_loss: 0.3886\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3913 - val_loss: 0.3791\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3818 - val_loss: 0.3835\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3779 - val_loss: 0.3759\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3704 - val_loss: 0.3696\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3656 - val_loss: 0.3753\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3601 - val_loss: 0.3647\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3561 - val_loss: 0.3514\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3514 - val_loss: 0.3485\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3489 - val_loss: 0.3497\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3442 - val_loss: 0.3482\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3422 - val_loss: 0.3508\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3400 - val_loss: 0.3404\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3360 - val_loss: 0.3516\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3324 - val_loss: 0.3537\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3337 - val_loss: 0.3421\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3286 - val_loss: 0.3263\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3275 - val_loss: 0.3464\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3317\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3196 - val_loss: 0.3380\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3187 - val_loss: 0.3244\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3185 - val_loss: 0.3432\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3187 - val_loss: 0.3406\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3174 - val_loss: 0.3169\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3125 - val_loss: 0.3269\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3127 - val_loss: 0.3198\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3082 - val_loss: 0.3139\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3052 - val_loss: 0.3200\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3082 - val_loss: 0.3271\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3059 - val_loss: 0.3226\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3034 - val_loss: 0.3095\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3022 - val_loss: 0.3179\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2997 - val_loss: 0.3254\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3029 - val_loss: 0.3112\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2976 - val_loss: 0.3090\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2978 - val_loss: 0.3084\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2962 - val_loss: 0.3154\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2981 - val_loss: 0.3121\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2942 - val_loss: 0.3022\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2953 - val_loss: 0.3156\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2917 - val_loss: 0.3098\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2914 - val_loss: 0.3081\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2896 - val_loss: 0.3054\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2904 - val_loss: 0.3105\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2890 - val_loss: 0.3120\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2909 - val_loss: 0.3227\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2869 - val_loss: 0.3023\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2878 - val_loss: 0.2982\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2872 - val_loss: 0.3039\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2851 - val_loss: 0.3220\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2844 - val_loss: 0.3148\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2841 - val_loss: 0.2963\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2841 - val_loss: 0.3098\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2839 - val_loss: 0.2966\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2793 - val_loss: 0.3578\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2826 - val_loss: 0.3047\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2810 - val_loss: 0.2954\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2812 - val_loss: 0.3084\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2791 - val_loss: 0.2975\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2797 - val_loss: 0.2921\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2780 - val_loss: 0.2962\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2784 - val_loss: 0.2950\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2774 - val_loss: 0.3066\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2723 - val_loss: 0.2955\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2772 - val_loss: 0.2881\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2763 - val_loss: 0.2919\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2735 - val_loss: 0.2993\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2752 - val_loss: 0.2916\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2727 - val_loss: 0.3013\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2729 - val_loss: 0.2903\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2723 - val_loss: 0.2915\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2723 - val_loss: 0.2891\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2692 - val_loss: 0.2921\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2688 - val_loss: 0.3361\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2709 - val_loss: 0.3134\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3330\n",
      "[CV]  learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77, total= 1.3min\n",
      "[CV] learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7801 - val_loss: 0.5242\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5214 - val_loss: 0.4594\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4692 - val_loss: 0.4473\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4498 - val_loss: 0.4319\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4348 - val_loss: 0.4162\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4174 - val_loss: 0.4302\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4068 - val_loss: 0.4019\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4004 - val_loss: 0.3949\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3853 - val_loss: 0.3837\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3805 - val_loss: 0.3794\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3701 - val_loss: 0.3753\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3691 - val_loss: 0.3593\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3612 - val_loss: 0.3776\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3551 - val_loss: 0.3460\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3483 - val_loss: 0.3448\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3448 - val_loss: 0.3479\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3428 - val_loss: 0.3593\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3406 - val_loss: 0.3443\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3392 - val_loss: 0.3351\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3330 - val_loss: 0.3443\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3322 - val_loss: 0.3363\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3265 - val_loss: 0.3443\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3247 - val_loss: 0.3192\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3269 - val_loss: 0.3231\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3206 - val_loss: 0.3200\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3166 - val_loss: 0.3165\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3136 - val_loss: 0.3190\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3126 - val_loss: 0.3229\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3101 - val_loss: 0.3162\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3095 - val_loss: 0.3091\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3089 - val_loss: 0.3112\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3072 - val_loss: 0.3140\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3115 - val_loss: 0.3145\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3019 - val_loss: 0.3182\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3127 - val_loss: 0.3192\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3060 - val_loss: 0.3049\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3029 - val_loss: 0.3118\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2973 - val_loss: 0.3193\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2980 - val_loss: 0.3140\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2980 - val_loss: 0.2993\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2968 - val_loss: 0.3174\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2950 - val_loss: 0.3056\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2946 - val_loss: 0.3129\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2927 - val_loss: 0.3188\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2930 - val_loss: 0.3097\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2890 - val_loss: 0.3084\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3033 - val_loss: 0.3061\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2913 - val_loss: 0.2923\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2923 - val_loss: 0.3090\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2834 - val_loss: 0.3283\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2895 - val_loss: 0.3451\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2878 - val_loss: 0.3007\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2873 - val_loss: 0.2956\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2854 - val_loss: 0.2981\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2844 - val_loss: 0.3088\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2844 - val_loss: 0.3219\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2810 - val_loss: 0.2960\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.2820 - val_loss: 0.3264\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3297\n",
      "[CV]  learning_rate=0.012559844265179787, n_hidden=2, n_neurons=77, total=  55.1s\n",
      "[CV] learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6 .....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 5ms/step - loss: 2.1136 - val_loss: 2.1036\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 26.7874 - val_loss: 19.4340\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 239.5283 - val_loss: 312.5493\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 9810.7871 - val_loss: 4291.4624\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 88458.0391 - val_loss: 61870.2773\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 800278.2500 - val_loss: 964000.9375\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 713220.6875 - val_loss: 14747722.0000\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 55924332.0000 - val_loss: 220120864.0000\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 130081120.0000 - val_loss: 3407904512.0000\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 13510848512.0000 - val_loss: 53029810176.0000\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 1070578335744.0000 - val_loss: 947485999104.0000\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 103379976192.0000\n",
      "[CV]  learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6, total=  10.1s\n",
      "[CV] learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6 .....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.2420 - val_loss: 0.9160\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 8.7736 - val_loss: 9.9497\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 273.4003 - val_loss: 329.5512\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 10378.8135 - val_loss: 11749.0146\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 368846.0625 - val_loss: 425852.1250\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 13021054.0000 - val_loss: 15633050.0000\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 19003412.0000 - val_loss: 558797504.0000\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 682675968.0000 - val_loss: 20061749248.0000\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 24276584448.0000 - val_loss: 726943596544.0000\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 1002550984704.0000 - val_loss: 26249900589056.0000\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 787276028182528.0000 - val_loss: 1089506233024512.0000\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 292835265871872.0000\n",
      "[CV]  learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6, total=  10.0s\n",
      "[CV] learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6 .....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.0994 - val_loss: 0.6367\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6522 - val_loss: 0.5789\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5609 - val_loss: 0.6028\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6200 - val_loss: 0.5692\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5714 - val_loss: 0.5540\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5597 - val_loss: 0.7452\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5827 - val_loss: 0.5519\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5509 - val_loss: 0.5514\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5752 - val_loss: 0.5610\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5420 - val_loss: 0.5509\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5436 - val_loss: 0.5607\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5641 - val_loss: 0.5515\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5461 - val_loss: 0.5560\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5531 - val_loss: 0.5494\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5408 - val_loss: 0.5460\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5469 - val_loss: 0.5452\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5418 - val_loss: 0.6187\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5599 - val_loss: 0.5584\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5358 - val_loss: 0.5597\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5841 - val_loss: 0.5648\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5400 - val_loss: 0.5656\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5515 - val_loss: 0.5729\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5437 - val_loss: 0.5589\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5414 - val_loss: 0.5506\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5996 - val_loss: 0.5533\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5488 - val_loss: 0.5612\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.7528\n",
      "[CV]  learning_rate=0.017076040781989342, n_hidden=0, n_neurons=6, total=  22.4s\n",
      "[CV] learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 2.9732 - val_loss: 2.7657\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8999 - val_loss: 0.4930\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4490 - val_loss: 1.0958\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.4616 - val_loss: 0.4187\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4032 - val_loss: 0.4154\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4032 - val_loss: 0.4013\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3861 - val_loss: 0.3914\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3811 - val_loss: 0.3876\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3780 - val_loss: 0.3922\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3760 - val_loss: 0.3847\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3719 - val_loss: 0.3823\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3690 - val_loss: 0.3769\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3666 - val_loss: 0.3772\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3706 - val_loss: 0.3778\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3628 - val_loss: 0.3690\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3610 - val_loss: 0.3711\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3583 - val_loss: 0.3676\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3577 - val_loss: 0.3654\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3574 - val_loss: 0.3686\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3589 - val_loss: 0.3630\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3497 - val_loss: 0.3643\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3512 - val_loss: 0.3609\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3461 - val_loss: 0.3561\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3449 - val_loss: 0.3736\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3450 - val_loss: 0.3585\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3435 - val_loss: 0.3540\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3442 - val_loss: 0.3584\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3398 - val_loss: 0.3541\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3370 - val_loss: 0.3465\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3362 - val_loss: 0.3567\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3342 - val_loss: 0.3568\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3344 - val_loss: 0.3450\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3320 - val_loss: 0.3439\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3346 - val_loss: 0.3435\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3312 - val_loss: 0.3437\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3332 - val_loss: 0.3426\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3276 - val_loss: 0.3417\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3303 - val_loss: 0.3518\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3282 - val_loss: 0.3528\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3260 - val_loss: 0.3510\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3248 - val_loss: 0.3427\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3502\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3243 - val_loss: 0.3406\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3365 - val_loss: 0.4070\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4136 - val_loss: 0.3456\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3312 - val_loss: 0.4792\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3318 - val_loss: 0.3388\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3255 - val_loss: 0.3369\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3229 - val_loss: 0.3356\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3223 - val_loss: 0.3349\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3202 - val_loss: 0.3346\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3217 - val_loss: 0.4535\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3507 - val_loss: 0.3359\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3207 - val_loss: 0.3388\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3230 - val_loss: 0.3372\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3200 - val_loss: 0.3320\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3190 - val_loss: 0.4304\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3270 - val_loss: 0.3431\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3154 - val_loss: 0.3343\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3158 - val_loss: 0.3302\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3190 - val_loss: 0.3339\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3315 - val_loss: 0.3386\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3319\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3261 - val_loss: 0.3310\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3229 - val_loss: 0.3316\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3211 - val_loss: 0.3291\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3134 - val_loss: 0.3279\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3125 - val_loss: 0.3275\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3198 - val_loss: 0.4283\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3173 - val_loss: 0.3274\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3103 - val_loss: 0.3331\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3108 - val_loss: 0.3308\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3093 - val_loss: 0.3255\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3081 - val_loss: 0.3283\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3130 - val_loss: 0.3446\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3114 - val_loss: 0.3233\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3062 - val_loss: 0.3214\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3060 - val_loss: 0.3242\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3052 - val_loss: 0.3229\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3050 - val_loss: 0.3374\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3047 - val_loss: 0.3417\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3098 - val_loss: 0.3232\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3115 - val_loss: 0.3272\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3061 - val_loss: 0.3239\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3037 - val_loss: 0.3232\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3075 - val_loss: 0.3232\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3049 - val_loss: 0.3209\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3031 - val_loss: 0.3192\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3041 - val_loss: 0.3272\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3085 - val_loss: 0.3226\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3023 - val_loss: 0.3222\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3101 - val_loss: 0.3494\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3078 - val_loss: 0.3246\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3045 - val_loss: 0.3200\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3023 - val_loss: 0.3216\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3026 - val_loss: 0.3255\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3010 - val_loss: 0.3220\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3034 - val_loss: 0.3222\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3345\n",
      "[CV]  learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98, total= 1.5min\n",
      "[CV] learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.0321 - val_loss: 0.7667\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.1022 - val_loss: 0.5107\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5065 - val_loss: 0.4681\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4765 - val_loss: 0.4538\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4604 - val_loss: 0.4391\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4476 - val_loss: 0.4312\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4393 - val_loss: 0.4237\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4328 - val_loss: 0.4172\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4269 - val_loss: 0.4141\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4219 - val_loss: 0.4118\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4149 - val_loss: 0.4046\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4107 - val_loss: 0.4022\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4070 - val_loss: 0.4029\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4031 - val_loss: 0.3960\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3994 - val_loss: 0.3920\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3977 - val_loss: 0.3922\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3935 - val_loss: 0.3934\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3913 - val_loss: 0.3880\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3909 - val_loss: 0.3856\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3888 - val_loss: 0.3904\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3842 - val_loss: 0.3884\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3826 - val_loss: 0.3796\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3790 - val_loss: 0.3778\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3779 - val_loss: 0.3778\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3753 - val_loss: 0.3824\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3747 - val_loss: 0.3791\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3725 - val_loss: 0.3752\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3708 - val_loss: 0.3782\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3696 - val_loss: 0.3756\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3696 - val_loss: 0.3725\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3708 - val_loss: 0.3729\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3719 - val_loss: 0.3694\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3649 - val_loss: 0.3686\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3636 - val_loss: 0.3715\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3628 - val_loss: 0.3693\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3622 - val_loss: 0.3657\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3600 - val_loss: 0.3702\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3603 - val_loss: 0.3678\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3582 - val_loss: 0.3690\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3575 - val_loss: 0.3731\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3547 - val_loss: 0.3615\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3551 - val_loss: 0.3596\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3549 - val_loss: 0.3611\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3554 - val_loss: 0.3601\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3560 - val_loss: 0.3613\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3556 - val_loss: 0.3845\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3522 - val_loss: 0.3600\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3527 - val_loss: 0.3637\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3510 - val_loss: 0.3606\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3506 - val_loss: 0.3571\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3490 - val_loss: 0.3548\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3473 - val_loss: 0.3596\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3460 - val_loss: 0.3527\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3454 - val_loss: 0.3543\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3452 - val_loss: 0.3559\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3442 - val_loss: 0.3507\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3448 - val_loss: 0.3533\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3424 - val_loss: 0.3556\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3438 - val_loss: 0.3524\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3460 - val_loss: 0.3506\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3708\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3419 - val_loss: 0.3489\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3442 - val_loss: 0.3504\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3401 - val_loss: 0.3595\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3402 - val_loss: 0.3460\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3527\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3385 - val_loss: 0.3523\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3380 - val_loss: 0.3507\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3399 - val_loss: 0.3484\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3370 - val_loss: 0.3485\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3360 - val_loss: 0.3444\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3372 - val_loss: 0.3478\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3375 - val_loss: 0.3447\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3340 - val_loss: 0.3433\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3389 - val_loss: 0.3493\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3359 - val_loss: 0.3429\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3362 - val_loss: 0.3495\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3324 - val_loss: 0.3448\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3298 - val_loss: 0.3484\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3325 - val_loss: 0.3529\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3316 - val_loss: 0.4553\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3614 - val_loss: 0.3425\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3289 - val_loss: 0.3463\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3288 - val_loss: 0.3409\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3289 - val_loss: 0.3443\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3282 - val_loss: 0.3391\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3273 - val_loss: 0.3389\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3271 - val_loss: 0.3418\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3278 - val_loss: 0.3402\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3274 - val_loss: 0.3394\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3293 - val_loss: 0.3372\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3325 - val_loss: 0.3357\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3249 - val_loss: 0.4130\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3347 - val_loss: 0.3482\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3294 - val_loss: 0.3346\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3245 - val_loss: 0.3380\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3248 - val_loss: 0.3358\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3241 - val_loss: 0.3762\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3240 - val_loss: 0.3329\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3220 - val_loss: 0.3354\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3451\n",
      "[CV]  learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98, total= 1.5min\n",
      "[CV] learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9462 - val_loss: 0.6339\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5854 - val_loss: 0.5098\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5135 - val_loss: 0.4761\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5042 - val_loss: 0.4680\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4801 - val_loss: 0.4551\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4660 - val_loss: 0.4775\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4564 - val_loss: 0.4484\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4590 - val_loss: 0.4410\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4548 - val_loss: 0.4458\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4397 - val_loss: 0.4387\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4329 - val_loss: 0.4334\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4282 - val_loss: 0.4229\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4294 - val_loss: 0.4266\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4241 - val_loss: 0.4172\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4191 - val_loss: 0.4112\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4152 - val_loss: 0.4118\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4180 - val_loss: 0.5665\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4266 - val_loss: 0.4089\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4054 - val_loss: 0.4051\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4099 - val_loss: 0.4096\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4025 - val_loss: 0.4075\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4030 - val_loss: 0.4008\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3999 - val_loss: 0.3900\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4106 - val_loss: 0.3883\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3988 - val_loss: 0.3885\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3911 - val_loss: 0.3922\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3842 - val_loss: 0.3878\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3828 - val_loss: 0.3884\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3809 - val_loss: 0.3843\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.3833 - val_loss: 0.3777\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3754 - val_loss: 0.3804\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3751 - val_loss: 0.3852\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4143 - val_loss: 0.3927\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3715 - val_loss: 0.3707\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3808 - val_loss: 0.3721\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3682 - val_loss: 0.3699\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3859 - val_loss: 0.3769\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3645 - val_loss: 0.3714\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3645 - val_loss: 0.3671\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3610 - val_loss: 0.3646\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3588 - val_loss: 0.3669\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3691 - val_loss: 0.3659\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3567 - val_loss: 0.3668\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3555 - val_loss: 0.3640\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3631 - val_loss: 0.3641\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3545 - val_loss: 0.3648\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4958 - val_loss: 0.3799\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3626 - val_loss: 0.3614\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3559 - val_loss: 0.3624\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3529 - val_loss: 0.3662\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3559 - val_loss: 0.3559\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3543 - val_loss: 0.3562\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3537 - val_loss: 0.3529\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3483 - val_loss: 0.3545\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3448 - val_loss: 0.3572\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3444 - val_loss: 0.3524\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3469 - val_loss: 0.3487\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3435 - val_loss: 0.3499\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3425 - val_loss: 0.3508\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3514\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3386 - val_loss: 0.3543\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3376 - val_loss: 0.3456\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3404 - val_loss: 0.3450\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3376 - val_loss: 0.3592\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.3351 - val_loss: 0.3423\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3429 - val_loss: 0.3474\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3344 - val_loss: 0.3470\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3312 - val_loss: 0.3394\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3314 - val_loss: 0.3395\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3387 - val_loss: 0.3374\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3300 - val_loss: 0.3426\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3291 - val_loss: 0.3416\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3284 - val_loss: 0.3416\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3278 - val_loss: 0.3422\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3301 - val_loss: 0.3374\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3257 - val_loss: 0.3375\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3378 - val_loss: 0.3379\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3296 - val_loss: 0.3353\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3247 - val_loss: 0.3377\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3300 - val_loss: 0.3416\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3260 - val_loss: 0.4072\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3342 - val_loss: 0.3340\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3230 - val_loss: 0.3365\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3274 - val_loss: 0.3667\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3254 - val_loss: 0.3453\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3224 - val_loss: 0.3326\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3226 - val_loss: 0.3291\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3186 - val_loss: 0.3303\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3227 - val_loss: 0.3377\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3213 - val_loss: 0.3277\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3207 - val_loss: 0.3294\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3335 - val_loss: 0.3270\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3176 - val_loss: 0.3333\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3357 - val_loss: 0.3311\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3168 - val_loss: 0.3250\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3166 - val_loss: 0.3238\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3257 - val_loss: 0.3267\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3173 - val_loss: 0.3282\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3280 - val_loss: 0.3288\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3154 - val_loss: 0.3279\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3476\n",
      "[CV]  learning_rate=0.008199846927131736, n_hidden=1, n_neurons=98, total= 1.5min\n",
      "[CV] learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 2.0020 - val_loss: 2.0492\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.8307 - val_loss: 0.6047\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5828 - val_loss: 0.5574\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5399 - val_loss: 0.5253\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5105 - val_loss: 0.5017\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4890 - val_loss: 0.4838\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4725 - val_loss: 0.4704\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4593 - val_loss: 0.4586\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4488 - val_loss: 0.4493\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4407 - val_loss: 0.4434\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4337 - val_loss: 0.4371\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4277 - val_loss: 0.4321\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4227 - val_loss: 0.4278\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4182 - val_loss: 0.4232\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4141 - val_loss: 0.4195\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4110 - val_loss: 0.4172\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4074 - val_loss: 0.4133\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4042 - val_loss: 0.4113\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4009 - val_loss: 0.4095\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3994 - val_loss: 0.4056\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3961 - val_loss: 0.4059\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3942 - val_loss: 0.4029\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3923 - val_loss: 0.4004\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3902 - val_loss: 0.3999\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3879 - val_loss: 0.3988\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3861 - val_loss: 0.3971\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 5ms/step - loss: 0.3844 - val_loss: 0.3945\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3823 - val_loss: 0.3920\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3799 - val_loss: 0.3902\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3796 - val_loss: 0.3921\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3770 - val_loss: 0.3895\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3762 - val_loss: 0.3853\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3736 - val_loss: 0.3851\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3726 - val_loss: 0.3841\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3702 - val_loss: 0.3827\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3695 - val_loss: 0.3813\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3667 - val_loss: 0.3803\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3680 - val_loss: 0.3810\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3659 - val_loss: 0.3793\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3640 - val_loss: 0.3795\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3622 - val_loss: 0.3763\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3623 - val_loss: 0.3786\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3603 - val_loss: 0.3753\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3607 - val_loss: 0.3778\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3598 - val_loss: 0.3732\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3575 - val_loss: 0.3737\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3562 - val_loss: 0.3715\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3563 - val_loss: 0.3691\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3539 - val_loss: 0.3693\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3539 - val_loss: 0.3697\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3510 - val_loss: 0.3708\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3521 - val_loss: 0.3735\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3509 - val_loss: 0.3669\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3490 - val_loss: 0.3667\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3523 - val_loss: 0.3653\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3470 - val_loss: 0.3632\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3479 - val_loss: 0.3649\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3479 - val_loss: 0.3688\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3470 - val_loss: 0.3659\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3474 - val_loss: 0.3629\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3450 - val_loss: 0.3652\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3468 - val_loss: 0.3631\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3492 - val_loss: 0.3592\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3418 - val_loss: 0.3583\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3576\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3403 - val_loss: 0.3571\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3412 - val_loss: 0.3569\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3406 - val_loss: 0.3566\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3434 - val_loss: 0.3572\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3366 - val_loss: 0.3564\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3381 - val_loss: 0.3535\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3369 - val_loss: 0.3532\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3357 - val_loss: 0.3520\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3352 - val_loss: 0.3519\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3373 - val_loss: 0.3564\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3337 - val_loss: 0.3497\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3338 - val_loss: 0.3511\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3332 - val_loss: 0.3524\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3318 - val_loss: 0.3492\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3320 - val_loss: 0.3565\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3325 - val_loss: 0.3505\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3321 - val_loss: 0.3468\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3303 - val_loss: 0.3504\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3297 - val_loss: 0.3461\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3327 - val_loss: 0.3450\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3282 - val_loss: 0.3451\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3278 - val_loss: 0.3445\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3278 - val_loss: 0.3449\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3286 - val_loss: 0.3446\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3266 - val_loss: 0.3447\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3262 - val_loss: 0.3423\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3305 - val_loss: 0.3444\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3252 - val_loss: 0.3414\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3255 - val_loss: 0.3464\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3262 - val_loss: 0.3399\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3243 - val_loss: 0.3386\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3236 - val_loss: 0.3442\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3237 - val_loss: 0.3383\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3220 - val_loss: 0.3390\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3218 - val_loss: 0.3403\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3608\n",
      "[CV]  learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53, total= 1.6min\n",
      "[CV] learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.6347 - val_loss: 0.7254\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7206 - val_loss: 0.6069\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6093 - val_loss: 0.5446\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5519 - val_loss: 0.5139\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5210 - val_loss: 0.4886\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4988 - val_loss: 0.4766\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4862 - val_loss: 0.4630\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4756 - val_loss: 0.4542\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4667 - val_loss: 0.4477\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4595 - val_loss: 0.4452\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4542 - val_loss: 0.4374\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4506 - val_loss: 0.4319\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4426 - val_loss: 0.4285\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4380 - val_loss: 0.4270\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4336 - val_loss: 0.4222\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4320 - val_loss: 0.4190\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4263 - val_loss: 0.4161\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4233 - val_loss: 0.4151\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4225 - val_loss: 0.4141\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4196 - val_loss: 0.4153\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4297 - val_loss: 0.4195\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4221 - val_loss: 0.4250\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4221 - val_loss: 0.4086\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4114 - val_loss: 0.4138\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4110 - val_loss: 0.4047\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4054 - val_loss: 0.4127\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4106 - val_loss: 0.4045\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4071 - val_loss: 0.4065\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4203 - val_loss: 0.4042\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4067 - val_loss: 0.4014\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4114 - val_loss: 0.3981\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4091 - val_loss: 0.3926\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3929 - val_loss: 0.3916\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3917 - val_loss: 0.3924\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3862 - val_loss: 0.3880\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3859 - val_loss: 0.3864\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3839 - val_loss: 0.3873\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3845 - val_loss: 0.3853\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3798 - val_loss: 0.3859\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3788 - val_loss: 0.3847\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3763 - val_loss: 0.3790\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3760 - val_loss: 0.3781\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3746 - val_loss: 0.3780\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3766 - val_loss: 0.3747\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3712 - val_loss: 0.3754\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3716 - val_loss: 0.3733\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3677 - val_loss: 0.3739\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3677 - val_loss: 0.3719\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3657 - val_loss: 0.3709\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3647 - val_loss: 0.3722\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3640 - val_loss: 0.3696\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3670 - val_loss: 0.3721\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3610 - val_loss: 0.3672\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3595 - val_loss: 0.3680\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3624 - val_loss: 0.3639\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3574 - val_loss: 0.3635\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3574 - val_loss: 0.3657\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3602 - val_loss: 0.3684\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3591 - val_loss: 0.3679\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3590 - val_loss: 0.3620\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3546 - val_loss: 0.3673\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3625 - val_loss: 0.3627\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3680 - val_loss: 0.3607\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3516 - val_loss: 0.3595\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3491 - val_loss: 0.3549\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3485 - val_loss: 0.3544\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3476 - val_loss: 0.3574\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3478 - val_loss: 0.3561\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3482 - val_loss: 0.3513\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3419 - val_loss: 0.3512\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3431 - val_loss: 0.3504\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3417 - val_loss: 0.3503\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3402 - val_loss: 0.3485\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3398 - val_loss: 0.3469\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3412 - val_loss: 0.3496\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3369 - val_loss: 0.3451\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3361 - val_loss: 0.3476\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3358 - val_loss: 0.3465\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3331 - val_loss: 0.3448\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3338 - val_loss: 0.3481\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3341 - val_loss: 0.3454\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3335 - val_loss: 0.3397\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3302 - val_loss: 0.3451\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3304 - val_loss: 0.3391\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3305 - val_loss: 0.3415\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3286 - val_loss: 0.3391\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3280 - val_loss: 0.3385\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3282 - val_loss: 0.3401\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3279 - val_loss: 0.3410\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3266 - val_loss: 0.3382\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3276 - val_loss: 0.3361\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3313 - val_loss: 0.3336\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3248 - val_loss: 0.3336\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3250 - val_loss: 0.3356\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3243 - val_loss: 0.3322\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3232 - val_loss: 0.3322\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3211 - val_loss: 0.3323\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3208 - val_loss: 0.3304\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3193 - val_loss: 0.3292\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3201 - val_loss: 0.3345\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3383\n",
      "[CV]  learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53, total= 1.6min\n",
      "[CV] learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.3437 - val_loss: 0.7095\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6725 - val_loss: 0.6302\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6173 - val_loss: 0.5813\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5794 - val_loss: 0.5471\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5499 - val_loss: 0.5190\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5260 - val_loss: 0.5008\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5095 - val_loss: 0.4861\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4965 - val_loss: 0.4749\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4849 - val_loss: 0.4652\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4759 - val_loss: 0.4620\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4673 - val_loss: 0.4531\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4604 - val_loss: 0.4462\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4546 - val_loss: 0.4411\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4486 - val_loss: 0.4371\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4432 - val_loss: 0.4317\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4383 - val_loss: 0.4329\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4347 - val_loss: 0.4262\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4311 - val_loss: 0.4233\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4278 - val_loss: 0.4207\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4252 - val_loss: 0.4195\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4213 - val_loss: 0.4183\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4181 - val_loss: 0.4141\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4155 - val_loss: 0.4116\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4133 - val_loss: 0.4089\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4099 - val_loss: 0.4083\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4075 - val_loss: 0.4068\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4054 - val_loss: 0.4042\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4028 - val_loss: 0.4050\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4009 - val_loss: 0.4020\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3984 - val_loss: 0.3993\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3967 - val_loss: 0.3984\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3944 - val_loss: 0.3972\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3924 - val_loss: 0.3959\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3903 - val_loss: 0.3926\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3886 - val_loss: 0.3914\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3875 - val_loss: 0.3902\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3855 - val_loss: 0.3888\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3828 - val_loss: 0.3899\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3819 - val_loss: 0.3887\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3807 - val_loss: 0.3857\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3789 - val_loss: 0.3864\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3777 - val_loss: 0.3840\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3760 - val_loss: 0.3827\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3742 - val_loss: 0.3797\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3729 - val_loss: 0.3832\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3713 - val_loss: 0.3798\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3701 - val_loss: 0.3832\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3692 - val_loss: 0.3764\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3682 - val_loss: 0.3745\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3660 - val_loss: 0.3798\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3659 - val_loss: 0.3711\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3637 - val_loss: 0.3710\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3625 - val_loss: 0.3703\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3614 - val_loss: 0.3708\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3596 - val_loss: 0.3699\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3588 - val_loss: 0.3676\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3571 - val_loss: 0.3649\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3561 - val_loss: 0.3675\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3545 - val_loss: 0.3645\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3536 - val_loss: 0.3639\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3525 - val_loss: 0.3612\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3514 - val_loss: 0.3598\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3504 - val_loss: 0.3603\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3494 - val_loss: 0.3616\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3485 - val_loss: 0.3570\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3476 - val_loss: 0.3561\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3464 - val_loss: 0.3554\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3440 - val_loss: 0.3539\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3449 - val_loss: 0.3514\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3440 - val_loss: 0.3506\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3426 - val_loss: 0.3516\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3420 - val_loss: 0.3507\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3407 - val_loss: 0.3502\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3402 - val_loss: 0.3504\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3395 - val_loss: 0.3473\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3385 - val_loss: 0.3483\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3373 - val_loss: 0.3462\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3363 - val_loss: 0.3453\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3358 - val_loss: 0.3443\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3346 - val_loss: 0.3446\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3339 - val_loss: 0.3456\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3335 - val_loss: 0.3416\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3327 - val_loss: 0.3409\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3323 - val_loss: 0.3412\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3306 - val_loss: 0.3477\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3310 - val_loss: 0.3389\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3297 - val_loss: 0.3393\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3283 - val_loss: 0.3383\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3279 - val_loss: 0.3427\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3281 - val_loss: 0.3371\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3275 - val_loss: 0.3350\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3260 - val_loss: 0.3347\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3254 - val_loss: 0.3380\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3266 - val_loss: 0.3342\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3244 - val_loss: 0.3328\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3231 - val_loss: 0.3317\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3234 - val_loss: 0.3324\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3234 - val_loss: 0.3315\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3228 - val_loss: 0.3315\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3218 - val_loss: 0.3347\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3415\n",
      "[CV]  learning_rate=0.0026782150761593213, n_hidden=2, n_neurons=53, total= 1.6min\n",
      "[CV] learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 3.8325 - val_loss: 1.9082\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.3768 - val_loss: 1.0652\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.9198 - val_loss: 0.8434\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7912 - val_loss: 0.7665\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7411 - val_loss: 0.7299\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7115 - val_loss: 0.7030\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6901 - val_loss: 0.6819\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6710 - val_loss: 0.6639\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6544 - val_loss: 0.6473\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6396 - val_loss: 0.6326\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6264 - val_loss: 0.6198\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6145 - val_loss: 0.6095\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6036 - val_loss: 0.5978\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5948 - val_loss: 0.5894\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5863 - val_loss: 0.5808\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5789 - val_loss: 0.5736\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5721 - val_loss: 0.5666\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5663 - val_loss: 0.5610\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5608 - val_loss: 0.5558\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5560 - val_loss: 0.5510\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5519 - val_loss: 0.5478\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5477 - val_loss: 0.5432\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5442 - val_loss: 0.5396\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5416 - val_loss: 0.5382\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5387 - val_loss: 0.5348\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5363 - val_loss: 0.5322\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5343 - val_loss: 0.5304\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5320 - val_loss: 0.5285\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5304 - val_loss: 0.5270\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5287 - val_loss: 0.5249\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5272 - val_loss: 0.5234\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5262 - val_loss: 0.5224\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5248 - val_loss: 0.5209\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5241 - val_loss: 0.5205\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5231 - val_loss: 0.5199\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5217 - val_loss: 0.5182\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5215 - val_loss: 0.5175\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5208 - val_loss: 0.5174\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5202 - val_loss: 0.5181\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5195 - val_loss: 0.5163\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5191 - val_loss: 0.5156\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5189 - val_loss: 0.5157\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5181 - val_loss: 0.5148\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5181 - val_loss: 0.5158\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5173 - val_loss: 0.5141\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5172 - val_loss: 0.5153\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5168 - val_loss: 0.5139\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5167 - val_loss: 0.5134\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5165 - val_loss: 0.5146\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5165 - val_loss: 0.5141\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5162 - val_loss: 0.5133\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5159 - val_loss: 0.5142\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5159 - val_loss: 0.5130\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5158 - val_loss: 0.5135\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5154 - val_loss: 0.5125\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5157 - val_loss: 0.5131\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5154 - val_loss: 0.5135\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5153 - val_loss: 0.5139\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5150 - val_loss: 0.5123\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5151 - val_loss: 0.5131\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5152 - val_loss: 0.5123\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5152 - val_loss: 0.5128\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5151 - val_loss: 0.5126\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5147 - val_loss: 0.5116\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5150 - val_loss: 0.5123\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5149 - val_loss: 0.5120\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5149 - val_loss: 0.5126\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5148 - val_loss: 0.5119\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5148 - val_loss: 0.5128\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5148 - val_loss: 0.5124\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5147 - val_loss: 0.5120\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5144 - val_loss: 0.5115\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5146 - val_loss: 0.5112\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5149 - val_loss: 0.5120\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5147 - val_loss: 0.5127\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5146 - val_loss: 0.5120\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5147 - val_loss: 0.5126\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5146 - val_loss: 0.5119\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5146 - val_loss: 0.5123\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5144 - val_loss: 0.5118\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5146 - val_loss: 0.5126\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5145 - val_loss: 0.5120\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5149 - val_loss: 0.5119\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.5609\n",
      "[CV]  learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56, total= 1.2min\n",
      "[CV] learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 4.4935 - val_loss: 1.9720\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 1.3211 - val_loss: 0.9188\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7917 - val_loss: 0.6798\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6680 - val_loss: 0.6168\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6320 - val_loss: 0.5939\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6165 - val_loss: 0.5819\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6068 - val_loss: 0.5730\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5989 - val_loss: 0.5657\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5922 - val_loss: 0.5593\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5863 - val_loss: 0.5537\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5810 - val_loss: 0.5489\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5762 - val_loss: 0.5446\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5720 - val_loss: 0.5405\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5682 - val_loss: 0.5372\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5648 - val_loss: 0.5340\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5618 - val_loss: 0.5315\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5591 - val_loss: 0.5291\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5567 - val_loss: 0.5270\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5545 - val_loss: 0.5251\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5526 - val_loss: 0.5233\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5510 - val_loss: 0.5220\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5494 - val_loss: 0.5207\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5479 - val_loss: 0.5196\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5469 - val_loss: 0.5184\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5457 - val_loss: 0.5175\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5448 - val_loss: 0.5167\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5439 - val_loss: 0.5160\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5430 - val_loss: 0.5155\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5424 - val_loss: 0.5149\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5416 - val_loss: 0.5144\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5412 - val_loss: 0.5139\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5408 - val_loss: 0.5135\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5402 - val_loss: 0.5131\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5399 - val_loss: 0.5128\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5395 - val_loss: 0.5128\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5392 - val_loss: 0.5125\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5389 - val_loss: 0.5124\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5386 - val_loss: 0.5123\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5384 - val_loss: 0.5120\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5382 - val_loss: 0.5119\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5380 - val_loss: 0.5117\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5379 - val_loss: 0.5115\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5377 - val_loss: 0.5115\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5375 - val_loss: 0.5115\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5375 - val_loss: 0.5113\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5373 - val_loss: 0.5113\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5372 - val_loss: 0.5114\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5370 - val_loss: 0.5111\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5370 - val_loss: 0.5111\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5370 - val_loss: 0.5112\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5370 - val_loss: 0.5112\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5368 - val_loss: 0.5111\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5369 - val_loss: 0.5110\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5368 - val_loss: 0.5111\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5368 - val_loss: 0.5111\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5367 - val_loss: 0.5110\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5366 - val_loss: 0.5110\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5367 - val_loss: 0.5110\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5365 - val_loss: 0.5111\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5366 - val_loss: 0.5109\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5365 - val_loss: 0.5110\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5366 - val_loss: 0.5109\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5366 - val_loss: 0.5110\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5365 - val_loss: 0.5110\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5365 - val_loss: 0.5110\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5365 - val_loss: 0.5109\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5364 - val_loss: 0.5111\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5364 - val_loss: 0.5111\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5364 - val_loss: 0.5110\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5365 - val_loss: 0.5110\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.5166\n",
      "[CV]  learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56, total= 1.0min\n",
      "[CV] learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56 ...\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 5.0243 - val_loss: 2.8136\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 1.6959 - val_loss: 1.4295\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.9748 - val_loss: 1.0858\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7952 - val_loss: 0.9816\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7394 - val_loss: 0.9387\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.7129 - val_loss: 0.9123\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6953 - val_loss: 0.8875\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6784 - val_loss: 0.8657\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6677 - val_loss: 0.8495\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6545 - val_loss: 0.8327\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6444 - val_loss: 0.8191\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6329 - val_loss: 0.8038\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6263 - val_loss: 0.7912\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6183 - val_loss: 0.7791\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6103 - val_loss: 0.7714\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.6045 - val_loss: 0.7595\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5976 - val_loss: 0.7527\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5922 - val_loss: 0.7399\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5881 - val_loss: 0.7319\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5838 - val_loss: 0.7246\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5790 - val_loss: 0.7170\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5750 - val_loss: 0.7124\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5717 - val_loss: 0.7040\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5678 - val_loss: 0.6964\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5665 - val_loss: 0.6915\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5636 - val_loss: 0.6861\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5612 - val_loss: 0.6822\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5581 - val_loss: 0.6760\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5570 - val_loss: 0.6722\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5540 - val_loss: 0.6665\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5529 - val_loss: 0.6622\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5522 - val_loss: 0.6596\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5496 - val_loss: 0.6578\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5485 - val_loss: 0.6520\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5469 - val_loss: 0.6481\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5473 - val_loss: 0.6456\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5455 - val_loss: 0.6423\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5447 - val_loss: 0.6396\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5435 - val_loss: 0.6366\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5432 - val_loss: 0.6355\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5415 - val_loss: 0.6314\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5418 - val_loss: 0.6300\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5411 - val_loss: 0.6274\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5398 - val_loss: 0.6249\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5399 - val_loss: 0.6240\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5391 - val_loss: 0.6210\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5387 - val_loss: 0.6193\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5386 - val_loss: 0.6180\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5377 - val_loss: 0.6174\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5379 - val_loss: 0.6151\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5364 - val_loss: 0.6119\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5372 - val_loss: 0.6104\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5371 - val_loss: 0.6094\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5361 - val_loss: 0.6076\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5366 - val_loss: 0.6072\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5356 - val_loss: 0.6051\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5361 - val_loss: 0.6040\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5353 - val_loss: 0.6027\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5351 - val_loss: 0.6013\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5353 - val_loss: 0.6005\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5348 - val_loss: 0.5993\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5335 - val_loss: 0.5990\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5345 - val_loss: 0.5972\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5349 - val_loss: 0.5969\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5340 - val_loss: 0.5956\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5343 - val_loss: 0.5954\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5334 - val_loss: 0.5937\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5335 - val_loss: 0.5928\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5346 - val_loss: 0.5925\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5343 - val_loss: 0.5917\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5340 - val_loss: 0.5911\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5338 - val_loss: 0.5902\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5334 - val_loss: 0.5895\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5336 - val_loss: 0.5900\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5333 - val_loss: 0.5884\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5333 - val_loss: 0.5876\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5340 - val_loss: 0.5876\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5332 - val_loss: 0.5867\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5334 - val_loss: 0.5859\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5337 - val_loss: 0.5855\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5334 - val_loss: 0.5854\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5335 - val_loss: 0.5850\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5333 - val_loss: 0.5853\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5335 - val_loss: 0.5844\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5332 - val_loss: 0.5846\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5333 - val_loss: 0.5838\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5327 - val_loss: 0.5824\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5330 - val_loss: 0.5832\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5336 - val_loss: 0.5824\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5327 - val_loss: 0.5815\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5332 - val_loss: 0.5817\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5326 - val_loss: 0.5809\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5328 - val_loss: 0.5804\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5330 - val_loss: 0.5801\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5334 - val_loss: 0.5800\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5331 - val_loss: 0.5802\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5333 - val_loss: 0.5799\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5328 - val_loss: 0.5792\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5330 - val_loss: 0.5791\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.5332 - val_loss: 0.5789\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.8555\n",
      "[CV]  learning_rate=0.0014957422387183704, n_hidden=0, n_neurons=56, total= 1.4min\n",
      "[CV] learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.2895 - val_loss: 0.7425\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.7053 - val_loss: 0.6589\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6254 - val_loss: 0.6077\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5921 - val_loss: 0.5528\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5423 - val_loss: 0.5288\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5357 - val_loss: 0.5094\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5145 - val_loss: 0.5458\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5339 - val_loss: 0.5234\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5133 - val_loss: 0.4947\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4941 - val_loss: 0.4893\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5499 - val_loss: 0.4964\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5606 - val_loss: 0.4787\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4750 - val_loss: 0.4668\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4809 - val_loss: 0.4687\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4798 - val_loss: 0.4649\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4839 - val_loss: 0.4574\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4682 - val_loss: 0.4549\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4540 - val_loss: 0.4547\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4596 - val_loss: 0.4516\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4545 - val_loss: 0.4487\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4632 - val_loss: 0.4473\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4459 - val_loss: 0.4399\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4383 - val_loss: 0.4323\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4304 - val_loss: 0.4400\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4290 - val_loss: 0.4299\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4238 - val_loss: 0.4281\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4268 - val_loss: 0.4244\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4205 - val_loss: 0.4235\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4250 - val_loss: 0.4224\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4264 - val_loss: 0.4249\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4224 - val_loss: 0.4194\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4210 - val_loss: 0.4136\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4095 - val_loss: 0.4121\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4113 - val_loss: 0.4105\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4099 - val_loss: 0.4092\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4032 - val_loss: 0.4079\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4038 - val_loss: 0.4088\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4064 - val_loss: 0.4084\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3981 - val_loss: 0.4736\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3984 - val_loss: 0.4009\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3939 - val_loss: 0.4009\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3941 - val_loss: 0.4037\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3941 - val_loss: 0.3990\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4102 - val_loss: 0.4520\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4581 - val_loss: 0.3974\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3887 - val_loss: 0.3963\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3848 - val_loss: 0.3926\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3851 - val_loss: 0.3922\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3842 - val_loss: 0.3894\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3834 - val_loss: 0.3901\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3791 - val_loss: 0.3898\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3836 - val_loss: 0.4511\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3889 - val_loss: 0.3865\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3770 - val_loss: 0.3848\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3845 - val_loss: 0.3855\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3731 - val_loss: 0.3827\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3752 - val_loss: 0.3927\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3798 - val_loss: 0.3844\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3744 - val_loss: 0.3851\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3741 - val_loss: 0.3774\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3707 - val_loss: 0.3782\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3861 - val_loss: 0.3780\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3742 - val_loss: 0.3760\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3663 - val_loss: 0.3733\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3661 - val_loss: 0.3761\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4077 - val_loss: 0.3764\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3800 - val_loss: 0.3741\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3667 - val_loss: 0.3735\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4134 - val_loss: 0.8511\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4200 - val_loss: 0.3746\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3670 - val_loss: 0.3706\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3637 - val_loss: 0.3701\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3609 - val_loss: 0.3763\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3618 - val_loss: 0.3709\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3633 - val_loss: 0.3734\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3613 - val_loss: 0.3666\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3634 - val_loss: 0.3673\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3618 - val_loss: 0.3691\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3564 - val_loss: 0.3663\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3577 - val_loss: 0.3688\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3616 - val_loss: 0.4033\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3616 - val_loss: 0.3620\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3542 - val_loss: 0.3660\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3570 - val_loss: 0.3618\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3583 - val_loss: 0.3636\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3594 - val_loss: 0.3619\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3650 - val_loss: 0.3597\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3515 - val_loss: 0.3606\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3551 - val_loss: 0.3907\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3659 - val_loss: 0.3623\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3520 - val_loss: 0.3590\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3783 - val_loss: 0.4083\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3558 - val_loss: 0.3570\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3509 - val_loss: 0.3619\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3522 - val_loss: 0.3563\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3485 - val_loss: 0.3577\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3574 - val_loss: 0.3589\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3498 - val_loss: 0.3548\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3475 - val_loss: 0.3545\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3466 - val_loss: 0.3573\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3791\n",
      "[CV]  learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25, total= 1.5min\n",
      "[CV] learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.1332 - val_loss: 0.6372\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6340 - val_loss: 0.5630\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5670 - val_loss: 0.5196\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5330 - val_loss: 0.4978\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5112 - val_loss: 0.4814\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4971 - val_loss: 0.4704\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4847 - val_loss: 0.4605\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4765 - val_loss: 0.4518\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.4679 - val_loss: 0.4472\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4617 - val_loss: 0.4438\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4585 - val_loss: 0.4402\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4642 - val_loss: 0.4319\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4460 - val_loss: 0.4287\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4413 - val_loss: 0.4313\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4447 - val_loss: 0.4297\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4537 - val_loss: 0.4238\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4393 - val_loss: 0.4251\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4334 - val_loss: 0.4298\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4611 - val_loss: 0.4398\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4523 - val_loss: 0.4194\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4275 - val_loss: 0.4183\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4210 - val_loss: 0.4137\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4177 - val_loss: 0.4143\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4160 - val_loss: 0.4100\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4136 - val_loss: 0.4112\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4123 - val_loss: 0.4077\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.4102 - val_loss: 0.4073\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4087 - val_loss: 0.4069\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4068 - val_loss: 0.4053\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4059 - val_loss: 0.4030\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4038 - val_loss: 0.4023\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4030 - val_loss: 0.4011\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4006 - val_loss: 0.4002\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3990 - val_loss: 0.4015\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3968 - val_loss: 0.3998\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3969 - val_loss: 0.3970\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3954 - val_loss: 0.3985\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3947 - val_loss: 0.3976\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3919 - val_loss: 0.3972\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3915 - val_loss: 0.3955\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3895 - val_loss: 0.3920\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3886 - val_loss: 0.3904\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3875 - val_loss: 0.3902\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3874 - val_loss: 0.3899\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3859 - val_loss: 0.3895\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3863 - val_loss: 0.3891\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3839 - val_loss: 0.3907\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3838 - val_loss: 0.3867\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3822 - val_loss: 0.3866\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3816 - val_loss: 0.3871\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3856 - val_loss: 0.3895\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3898 - val_loss: 0.3871\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3796 - val_loss: 0.3824\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3774 - val_loss: 0.3831\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3773 - val_loss: 0.3806\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3759 - val_loss: 0.3816\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3755 - val_loss: 0.3798\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3742 - val_loss: 0.3824\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3753 - val_loss: 0.3794\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3750 - val_loss: 0.3778\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3729 - val_loss: 0.3817\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3786 - val_loss: 0.3794\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3826 - val_loss: 0.3774\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3723 - val_loss: 0.3789\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3711 - val_loss: 0.3743\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3694 - val_loss: 0.3740\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3687 - val_loss: 0.3753\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3693 - val_loss: 0.3750\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3683 - val_loss: 0.3730\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3664 - val_loss: 0.3717\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3665 - val_loss: 0.3708\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3659 - val_loss: 0.3703\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3646 - val_loss: 0.3692\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3649 - val_loss: 0.3682\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3647 - val_loss: 0.3708\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3633 - val_loss: 0.3676\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3618 - val_loss: 0.3691\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3617 - val_loss: 0.3691\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3604 - val_loss: 0.3683\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3609 - val_loss: 0.3678\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3603 - val_loss: 0.3676\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3604 - val_loss: 0.3642\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3586 - val_loss: 0.3708\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3654 - val_loss: 0.3662\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3724 - val_loss: 0.3690\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3611 - val_loss: 0.3633\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3620 - val_loss: 0.3641\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3569 - val_loss: 0.3613\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3565 - val_loss: 0.3649\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3569 - val_loss: 0.3623\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3610 - val_loss: 0.3610\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3571 - val_loss: 0.3621\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3622 - val_loss: 0.3582\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3541 - val_loss: 0.3621\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3524 - val_loss: 0.3593\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3574 - val_loss: 0.3636\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3565 - val_loss: 0.3611\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3534 - val_loss: 0.3579\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3506 - val_loss: 0.3556\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3480 - val_loss: 0.3546\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3521\n",
      "[CV]  learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25, total= 1.5min\n",
      "[CV] learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.4586 - val_loss: 0.7462\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6781 - val_loss: 0.6110\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5976 - val_loss: 0.5520\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5497 - val_loss: 0.5097\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5230 - val_loss: 0.4907\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5064 - val_loss: 0.4793\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4975 - val_loss: 0.4739\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4903 - val_loss: 0.4663\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4814 - val_loss: 0.4652\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4776 - val_loss: 0.4621\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4728 - val_loss: 0.4613\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4693 - val_loss: 0.4560\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4678 - val_loss: 0.4556\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4640 - val_loss: 0.4542\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4614 - val_loss: 0.4505\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4582 - val_loss: 0.4508\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4571 - val_loss: 0.4583\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4563 - val_loss: 0.4501\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4525 - val_loss: 0.4483\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4535 - val_loss: 0.4509\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4489 - val_loss: 0.4489\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4478 - val_loss: 0.4458\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4443 - val_loss: 0.4422\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4447 - val_loss: 0.4383\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4396 - val_loss: 0.4394\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4399 - val_loss: 0.4374\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4366 - val_loss: 0.4339\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4353 - val_loss: 0.4348\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4331 - val_loss: 0.4333\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.4298 - val_loss: 0.4279\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4287 - val_loss: 0.4295\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4265 - val_loss: 0.4257\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4260 - val_loss: 0.4244\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4233 - val_loss: 0.4203\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4213 - val_loss: 0.4210\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4197 - val_loss: 0.4177\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4187 - val_loss: 0.4155\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4150 - val_loss: 0.4170\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4135 - val_loss: 0.4139\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4120 - val_loss: 0.4108\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4115 - val_loss: 0.4113\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4099 - val_loss: 0.4103\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4089 - val_loss: 0.4060\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4068 - val_loss: 0.4056\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4072 - val_loss: 0.4101\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4045 - val_loss: 0.4013\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4058 - val_loss: 0.4153\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4036 - val_loss: 0.4028\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4004 - val_loss: 0.3981\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3990 - val_loss: 0.4035\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4007 - val_loss: 0.3947\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4002 - val_loss: 0.3952\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3968 - val_loss: 0.3925\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3945 - val_loss: 0.3935\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3931 - val_loss: 0.3919\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3935 - val_loss: 0.3902\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3937 - val_loss: 0.3892\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 3ms/step - loss: 0.3905 - val_loss: 0.3919\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3915 - val_loss: 0.3886\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3888 - val_loss: 0.3876\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3868 - val_loss: 0.3895\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3861 - val_loss: 0.3850\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3883 - val_loss: 0.3869\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3853 - val_loss: 0.3887\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3855 - val_loss: 0.3854\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3886 - val_loss: 0.3821\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3833 - val_loss: 0.3848\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3809 - val_loss: 0.3829\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3822 - val_loss: 0.3796\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3838 - val_loss: 0.3783\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3827 - val_loss: 0.3793\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3817 - val_loss: 0.3814\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3798 - val_loss: 0.3809\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3795 - val_loss: 0.3804\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3791 - val_loss: 0.3784\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3776 - val_loss: 0.3800\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3766 - val_loss: 0.3777\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3777 - val_loss: 0.3775\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3755 - val_loss: 0.3780\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3757 - val_loss: 0.3786\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3736 - val_loss: 0.3806\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3756 - val_loss: 0.3729\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3741 - val_loss: 0.3750\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3749 - val_loss: 0.3744\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3722 - val_loss: 0.3791\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3727 - val_loss: 0.3722\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3719 - val_loss: 0.3722\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3709 - val_loss: 0.3706\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3706 - val_loss: 0.3761\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3717 - val_loss: 0.3701\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3702 - val_loss: 0.3690\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3714 - val_loss: 0.3683\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3691 - val_loss: 0.3703\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3703 - val_loss: 0.3713\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3698 - val_loss: 0.3682\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3662 - val_loss: 0.3706\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3695 - val_loss: 0.3680\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3674 - val_loss: 0.3680\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3671 - val_loss: 0.3693\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3641 - val_loss: 0.3665\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.4164\n",
      "[CV]  learning_rate=0.005136788539140083, n_hidden=1, n_neurons=25, total= 1.5min\n",
      "[CV] learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.0492 - val_loss: 0.6343\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6008 - val_loss: 0.5482\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5169 - val_loss: 0.4964\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4815 - val_loss: 0.4737\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4618 - val_loss: 0.4618\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4525 - val_loss: 0.4554\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4464 - val_loss: 0.4474\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4388 - val_loss: 0.4402\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4327 - val_loss: 0.4361\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4280 - val_loss: 0.4356\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4252 - val_loss: 0.4271\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4182 - val_loss: 0.4242\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4151 - val_loss: 0.4226\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4111 - val_loss: 0.4163\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4067 - val_loss: 0.4135\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4038 - val_loss: 0.4116\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4012 - val_loss: 0.4072\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3981 - val_loss: 0.4067\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3934 - val_loss: 0.4138\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3925 - val_loss: 0.4004\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3897 - val_loss: 0.4025\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3871 - val_loss: 0.3989\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3856 - val_loss: 0.3944\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3830 - val_loss: 0.4008\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3817 - val_loss: 0.3955\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3801 - val_loss: 0.3913\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3780 - val_loss: 0.3997\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3757 - val_loss: 0.3895\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3743 - val_loss: 0.3842\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3727 - val_loss: 0.3911\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3706 - val_loss: 0.3965\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3702 - val_loss: 0.3807\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3680 - val_loss: 0.3816\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3665 - val_loss: 0.3770\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3658 - val_loss: 0.3764\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3668 - val_loss: 0.3786\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3622 - val_loss: 0.3770\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3624 - val_loss: 0.3791\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3626 - val_loss: 0.3736\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3603 - val_loss: 0.3906\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3610 - val_loss: 0.3701\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3578 - val_loss: 0.3728\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3577 - val_loss: 0.3682\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3569 - val_loss: 0.3741\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3649 - val_loss: 0.3652\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3568 - val_loss: 0.3691\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3546 - val_loss: 0.3653\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3523 - val_loss: 0.3626\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3517 - val_loss: 0.3606\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3528 - val_loss: 0.3605\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3590 - val_loss: 0.3641\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3651 - val_loss: 0.3832\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3522 - val_loss: 0.3608\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3491 - val_loss: 0.3633\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3479 - val_loss: 0.3640\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3468 - val_loss: 0.3559\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 5ms/step - loss: 0.3462 - val_loss: 0.3640\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3439 - val_loss: 0.3656\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3446 - val_loss: 0.3531\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3441 - val_loss: 0.3520\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3428 - val_loss: 0.3560\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3437 - val_loss: 0.3558\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3475\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3409 - val_loss: 0.3466\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3373 - val_loss: 0.3446\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3402 - val_loss: 0.3459\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3396 - val_loss: 0.3452\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3376 - val_loss: 0.3474\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3366 - val_loss: 0.3612\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3356 - val_loss: 0.3413\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3346 - val_loss: 0.3538\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3351 - val_loss: 0.3452\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3344 - val_loss: 0.3423\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3344 - val_loss: 0.3403\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3333 - val_loss: 0.3493\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3326 - val_loss: 0.3424\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3298 - val_loss: 0.3349\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3299 - val_loss: 0.3380\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3291 - val_loss: 0.3355\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3297 - val_loss: 0.3493\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3286 - val_loss: 0.3440\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3276 - val_loss: 0.3354\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3284 - val_loss: 0.3456\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3299 - val_loss: 0.3420\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3273 - val_loss: 0.3336\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3277 - val_loss: 0.3310\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3266 - val_loss: 0.3317\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3256 - val_loss: 0.3306\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3268 - val_loss: 0.3412\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3255 - val_loss: 0.3308\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3245 - val_loss: 0.3316\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3277 - val_loss: 0.3455\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3261 - val_loss: 0.3302\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3227 - val_loss: 0.3274\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3247 - val_loss: 0.3369\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3248 - val_loss: 0.3255\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3231 - val_loss: 0.3335\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3225 - val_loss: 0.3335\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3227 - val_loss: 0.3294\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3201 - val_loss: 0.3339\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3598\n",
      "[CV]  learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14, total= 1.6min\n",
      "[CV] learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.8357 - val_loss: 0.9095\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6678 - val_loss: 0.5917\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5953 - val_loss: 0.5560\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5625 - val_loss: 0.5343\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5385 - val_loss: 0.5102\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5168 - val_loss: 0.4977\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4996 - val_loss: 0.4768\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4832 - val_loss: 0.4639\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4687 - val_loss: 0.4535\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4589 - val_loss: 0.4496\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4487 - val_loss: 0.4357\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4403 - val_loss: 0.4292\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4331 - val_loss: 0.4240\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4262 - val_loss: 0.4173\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4194 - val_loss: 0.4106\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4138 - val_loss: 0.4078\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4090 - val_loss: 0.4045\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4046 - val_loss: 0.4012\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4011 - val_loss: 0.3972\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3982 - val_loss: 0.3952\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3944 - val_loss: 0.3969\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3920 - val_loss: 0.3895\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3884 - val_loss: 0.3870\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3859 - val_loss: 0.3837\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3830 - val_loss: 0.3856\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3812 - val_loss: 0.3814\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3788 - val_loss: 0.3776\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3762 - val_loss: 0.3765\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3747 - val_loss: 0.3754\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3731 - val_loss: 0.3698\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3704 - val_loss: 0.3684\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3690 - val_loss: 0.3655\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3666 - val_loss: 0.3625\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3647 - val_loss: 0.3639\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3624 - val_loss: 0.3612\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3617 - val_loss: 0.3596\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3596 - val_loss: 0.3616\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3579 - val_loss: 0.3578\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3560 - val_loss: 0.3588\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3559 - val_loss: 0.3573\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3530 - val_loss: 0.3515\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3523 - val_loss: 0.3471\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3506 - val_loss: 0.3465\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3503 - val_loss: 0.3472\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3487 - val_loss: 0.3471\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3477 - val_loss: 0.3450\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3458 - val_loss: 0.3450\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3461 - val_loss: 0.3422\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3442 - val_loss: 0.3418\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3438 - val_loss: 0.3420\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3435 - val_loss: 0.3415\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3414 - val_loss: 0.3495\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3410 - val_loss: 0.3385\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3403 - val_loss: 0.3383\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3389 - val_loss: 0.3370\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3389 - val_loss: 0.3371\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3382 - val_loss: 0.3336\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3352 - val_loss: 0.3374\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3356 - val_loss: 0.3331\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3356 - val_loss: 0.3310\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3343 - val_loss: 0.3411\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3347 - val_loss: 0.3313\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3338 - val_loss: 0.3342\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3329 - val_loss: 0.3319\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3333 - val_loss: 0.3260\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3318 - val_loss: 0.3287\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3307 - val_loss: 0.3319\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3310 - val_loss: 0.3303\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3296 - val_loss: 0.3273\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3291 - val_loss: 0.3272\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3301 - val_loss: 0.3230\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3303 - val_loss: 0.3264\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3293 - val_loss: 0.3247\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3284 - val_loss: 0.3230\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3268 - val_loss: 0.3323\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3287 - val_loss: 0.3225\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3259 - val_loss: 0.3264\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3269 - val_loss: 0.3268\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3256 - val_loss: 0.3250\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3257 - val_loss: 0.3302\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3255 - val_loss: 0.3363\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3265 - val_loss: 0.3179\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3233 - val_loss: 0.3323\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3257 - val_loss: 0.3174\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3236 - val_loss: 0.3231\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3248 - val_loss: 0.3186\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3210\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3233 - val_loss: 0.3221\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3247\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3226 - val_loss: 0.3194\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3235 - val_loss: 0.3177\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3229 - val_loss: 0.3173\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3231 - val_loss: 0.3278\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3230 - val_loss: 0.3181\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3217 - val_loss: 0.3169\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3223 - val_loss: 0.3226\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3217 - val_loss: 0.3173\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3214 - val_loss: 0.3196\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3209 - val_loss: 0.3150\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3203 - val_loss: 0.3152\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3359\n",
      "[CV]  learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14, total= 1.6min\n",
      "[CV] learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14 ....\n",
      "Epoch 1/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 1.1393 - val_loss: 0.6747\n",
      "Epoch 2/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.6455 - val_loss: 0.5714\n",
      "Epoch 3/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5615 - val_loss: 0.5162\n",
      "Epoch 4/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.5124 - val_loss: 0.4837\n",
      "Epoch 5/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4871 - val_loss: 0.4695\n",
      "Epoch 6/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4734 - val_loss: 0.4638\n",
      "Epoch 7/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4643 - val_loss: 0.4587\n",
      "Epoch 8/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4575 - val_loss: 0.4543\n",
      "Epoch 9/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4510 - val_loss: 0.4499\n",
      "Epoch 10/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4462 - val_loss: 0.4540\n",
      "Epoch 11/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4401 - val_loss: 0.4409\n",
      "Epoch 12/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4356 - val_loss: 0.4338\n",
      "Epoch 13/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4317 - val_loss: 0.4323\n",
      "Epoch 14/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4277 - val_loss: 0.4290\n",
      "Epoch 15/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4227 - val_loss: 0.4214\n",
      "Epoch 16/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4188 - val_loss: 0.4247\n",
      "Epoch 17/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4149 - val_loss: 0.4168\n",
      "Epoch 18/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4117 - val_loss: 0.4149\n",
      "Epoch 19/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4083 - val_loss: 0.4107\n",
      "Epoch 20/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4055 - val_loss: 0.4206\n",
      "Epoch 21/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.4026 - val_loss: 0.4081\n",
      "Epoch 22/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3995 - val_loss: 0.4026\n",
      "Epoch 23/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3958 - val_loss: 0.4003\n",
      "Epoch 24/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3936 - val_loss: 0.3999\n",
      "Epoch 25/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3898 - val_loss: 0.3996\n",
      "Epoch 26/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3869 - val_loss: 0.3937\n",
      "Epoch 27/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3850 - val_loss: 0.3931\n",
      "Epoch 28/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3829 - val_loss: 0.3932\n",
      "Epoch 29/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3804 - val_loss: 0.3887\n",
      "Epoch 30/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3778 - val_loss: 0.3849\n",
      "Epoch 31/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3766 - val_loss: 0.3885\n",
      "Epoch 32/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3751 - val_loss: 0.3831\n",
      "Epoch 33/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3732 - val_loss: 0.3816\n",
      "Epoch 34/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3708 - val_loss: 0.3794\n",
      "Epoch 35/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3698 - val_loss: 0.3806\n",
      "Epoch 36/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3703 - val_loss: 0.3759\n",
      "Epoch 37/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3673 - val_loss: 0.3766\n",
      "Epoch 38/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3641 - val_loss: 0.3783\n",
      "Epoch 39/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3645 - val_loss: 0.3719\n",
      "Epoch 40/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3634 - val_loss: 0.3694\n",
      "Epoch 41/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3619 - val_loss: 0.3730\n",
      "Epoch 42/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3609 - val_loss: 0.3758\n",
      "Epoch 43/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3593 - val_loss: 0.3716\n",
      "Epoch 44/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3577 - val_loss: 0.3666\n",
      "Epoch 45/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3575 - val_loss: 0.3667\n",
      "Epoch 46/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3555 - val_loss: 0.3656\n",
      "Epoch 47/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3538 - val_loss: 0.3698\n",
      "Epoch 48/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3540 - val_loss: 0.3578\n",
      "Epoch 49/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3537 - val_loss: 0.3559\n",
      "Epoch 50/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3508 - val_loss: 0.3617\n",
      "Epoch 51/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3515 - val_loss: 0.3565\n",
      "Epoch 52/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3506 - val_loss: 0.3522\n",
      "Epoch 53/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3496 - val_loss: 0.3504\n",
      "Epoch 54/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3474 - val_loss: 0.3520\n",
      "Epoch 55/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3461 - val_loss: 0.3514\n",
      "Epoch 56/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3463 - val_loss: 0.3520\n",
      "Epoch 57/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3445 - val_loss: 0.3443\n",
      "Epoch 58/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3437 - val_loss: 0.3519\n",
      "Epoch 59/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3428 - val_loss: 0.3480\n",
      "Epoch 60/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3429 - val_loss: 0.3460\n",
      "Epoch 61/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3408 - val_loss: 0.3785\n",
      "Epoch 62/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3405 - val_loss: 0.3413\n",
      "Epoch 63/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3390 - val_loss: 0.3462\n",
      "Epoch 64/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3381 - val_loss: 0.3696\n",
      "Epoch 65/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3390 - val_loss: 0.3414\n",
      "Epoch 66/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3371 - val_loss: 0.3414\n",
      "Epoch 67/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3371 - val_loss: 0.3426\n",
      "Epoch 68/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3345 - val_loss: 0.3368\n",
      "Epoch 69/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3359 - val_loss: 0.3370\n",
      "Epoch 70/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3352 - val_loss: 0.3336\n",
      "Epoch 71/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3339 - val_loss: 0.3424\n",
      "Epoch 72/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3333 - val_loss: 0.3385\n",
      "Epoch 73/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3318 - val_loss: 0.3375\n",
      "Epoch 74/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3323 - val_loss: 0.3350\n",
      "Epoch 75/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3305 - val_loss: 0.3330\n",
      "Epoch 76/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3301 - val_loss: 0.3338\n",
      "Epoch 77/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3294 - val_loss: 0.3346\n",
      "Epoch 78/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3281 - val_loss: 0.3313\n",
      "Epoch 79/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3278 - val_loss: 0.3399\n",
      "Epoch 80/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3271 - val_loss: 0.3364\n",
      "Epoch 81/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3270 - val_loss: 0.3351\n",
      "Epoch 82/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3281 - val_loss: 0.3287\n",
      "Epoch 83/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3256 - val_loss: 0.3312\n",
      "Epoch 84/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3259 - val_loss: 0.3356\n",
      "Epoch 85/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3240 - val_loss: 0.3471\n",
      "Epoch 86/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3250 - val_loss: 0.3231\n",
      "Epoch 87/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3233 - val_loss: 0.3225\n",
      "Epoch 88/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3208 - val_loss: 0.3264\n",
      "Epoch 89/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3219 - val_loss: 0.3369\n",
      "Epoch 90/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3213 - val_loss: 0.3224\n",
      "Epoch 91/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3214 - val_loss: 0.3218\n",
      "Epoch 92/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3190 - val_loss: 0.3224\n",
      "Epoch 93/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3178 - val_loss: 0.3401\n",
      "Epoch 94/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3177 - val_loss: 0.3199\n",
      "Epoch 95/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3166 - val_loss: 0.3165\n",
      "Epoch 96/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3153 - val_loss: 0.3182\n",
      "Epoch 97/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3157 - val_loss: 0.3205\n",
      "Epoch 98/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3157 - val_loss: 0.3190\n",
      "Epoch 99/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3140 - val_loss: 0.3192\n",
      "Epoch 100/100\n",
      "242/242 [==============================] - 1s 4ms/step - loss: 0.3132 - val_loss: 0.3172\n",
      "121/121 [==============================] - 0s 2ms/step - loss: 0.3752\n",
      "[CV]  learning_rate=0.006254354988900308, n_hidden=2, n_neurons=14, total= 1.6min\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  30 out of  30 | elapsed: 35.9min finished\n"
     ]
    },
    {
     "ename": "RuntimeError",
     "evalue": "ignored",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-116-b5a8bf7e14c8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m rnd_search_cv.fit(X_train, y_train, epochs = 100, \n\u001b[1;32m      4\u001b[0m                   \u001b[0mvalidation_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mX_valid\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_valid\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m                  callbacks = keras.callbacks.EarlyStopping(patience=10))\n\u001b[0m",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/sklearn/model_selection/_search.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, X, y, groups, **fit_params)\u001b[0m\n\u001b[1;32m    734\u001b[0m             \u001b[0;31m# of the params are estimators as well.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    735\u001b[0m             self.best_estimator_ = clone(clone(base_estimator).set_params(\n\u001b[0;32m--> 736\u001b[0;31m                 **self.best_params_))\n\u001b[0m\u001b[1;32m    737\u001b[0m             \u001b[0mrefit_start_time\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtime\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtime\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    738\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0my\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/sklearn/base.py\u001b[0m in \u001b[0;36mclone\u001b[0;34m(estimator, safe)\u001b[0m\n\u001b[1;32m     80\u001b[0m             raise RuntimeError('Cannot clone object %s, as the constructor '\n\u001b[1;32m     81\u001b[0m                                \u001b[0;34m'either does not set or modifies parameter %s'\u001b[0m \u001b[0;34m%\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 82\u001b[0;31m                                (estimator, name))\n\u001b[0m\u001b[1;32m     83\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mnew_object\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     84\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mRuntimeError\u001b[0m: Cannot clone object <tensorflow.python.keras.wrappers.scikit_learn.KerasRegressor object at 0x7f8debd0ff60>, as the constructor either does not set or modifies parameter learning_rate"
     ]
    }
   ],
   "source": [
    "rnd_search_cv = RandomizedSearchCV(keras_reg, param_distribs, \n",
    "                                   n_iter = 10, cv = 3, verbose=2)\n",
    "rnd_search_cv.fit(X_train, y_train, epochs = 100, \n",
    "                  validation_data = (X_valid, y_valid),\n",
    "                 callbacks = keras.callbacks.EarlyStopping(patience=10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "s9AOTKutaYbW",
    "outputId": "e87ecf32-dedb-46cb-9093-c1000040cfb9"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'learning_rate': 0.012559844265179787, 'n_hidden': 2, 'n_neurons': 77}"
      ]
     },
     "execution_count": 117,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 118,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 35
    },
    "colab_type": "code",
    "id": "bL31Z36baYbX",
    "outputId": "85420539-d049-43c8-8b8b-1e81b3e7dd9d"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "-0.3304374615351359"
      ]
     },
     "execution_count": 118,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rnd_search_cv.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 52
    },
    "colab_type": "code",
    "id": "azD9BDPIaYbe",
    "outputId": "a4203139-a775-4699-c570-f63c110ad2fe"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "162/162 [==============================] - 0s 2ms/step - loss: 0.3508\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.350837767124176"
      ]
     },
     "execution_count": 121,
     "metadata": {
      "tags": []
     },
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(X_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "g_1AvsSynCjd"
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "name": "Artificial_Neural_Networks_with_Keras.ipynb",
   "provenance": []
  },
  "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}