test_ntk.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-18 17:57:17.416180: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n",
      "2023-03-18 17:57:17.589873: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory\n",
      "2023-03-18 17:57:17.589893: I tensorflow/compiler/xla/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.\n",
      "2023-03-18 17:57:18.361110: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer.so.7'; dlerror: libnvinfer.so.7: cannot open shared object file: No such file or directory\n",
      "2023-03-18 17:57:18.361165: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libnvinfer_plugin.so.7'; dlerror: libnvinfer_plugin.so.7: cannot open shared object file: No such file or directory\n",
      "2023-03-18 17:57:18.361172: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "\n",
    "from sklearn.gaussian_process.kernels import ConstantKernel, RBF\n",
    "from sklearn.gaussian_process import GaussianProcessRegressor\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "dense_config = dict(\n",
    "    use_bias = True,\n",
    "    kernel_initializer = \"glorot_normal\",\n",
    "    bias_initializer = \"glorot_normal\",\n",
    "    activation = \"tanh\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "x = np.linspace(-10,10,500)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_model():\n",
    "    inputs = tf.keras.Input(shape=(1,))\n",
    "\n",
    "    hidden_layers = [\n",
    "        tf.keras.layers.Dense(2000, **dense_config),\n",
    "    ]\n",
    "\n",
    "    output_layer = tf.keras.layers.Dense(1, **dense_config)\n",
    "\n",
    "    layer_output = hidden_layers[0](inputs)\n",
    "    for layer in hidden_layers[1:]:\n",
    "        layer_output = layer(layer_output)\n",
    "    outputs = output_layer(layer_output)\n",
    "\n",
    "    return tf.keras.Model(inputs,outputs)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16/16 [==============================] - 0s 1ms/step\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023-03-18 17:57:19.981473: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory\n",
      "2023-03-18 17:57:19.981507: W tensorflow/compiler/xla/stream_executor/cuda/cuda_driver.cc:265] failed call to cuInit: UNKNOWN ERROR (303)\n",
      "2023-03-18 17:57:19.981538: I tensorflow/compiler/xla/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (roy-XPS-13-7390): /proc/driver/nvidia/version does not exist\n",
      "2023-03-18 17:57:19.981834: I tensorflow/core/platform/cpu_feature_guard.cc:193] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\n",
      "To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 955us/step\n",
      "16/16 [==============================] - 0s 874us/step\n",
      "16/16 [==============================] - 0s 992us/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 887us/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 998us/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 958us/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 929us/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 6ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "16/16 [==============================] - 0s 2ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(100):\n",
    "    model = make_model()\n",
    "    model.compile(loss=\"mse\")\n",
    "    y = model.predict(x)\n",
    "    plt.plot(x,y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 430ms/step - loss: 0.4014\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3481\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3194\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2641\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2302\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2006\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1615\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1209\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0911\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0699\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0465\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0241\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0111\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0066\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0044\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0024\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0034\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0089\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0149\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0178\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0197\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0229\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0252\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0250\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0230\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0213\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0198\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0170\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0132\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0101\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0078\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0055\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0030\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0013\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.6936e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.3364e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.4205e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.0449e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0031\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0033\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0031\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0028\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0026\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0016\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 698ms/step - loss: 0.3668\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2673\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2792\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2186\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1662\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1435\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1220\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0904\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0565\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0338\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0265\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0211\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0098\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.9740e-04\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0015\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0075\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0118\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0125\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0124\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0151\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0199\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0221\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0203\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0182\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0179\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0177\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0157\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0122\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0089\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0072\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0061\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.4920e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.7474e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 9.5735e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.9997e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 4.7131e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.1493e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0018\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0022\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0022\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0025\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0026\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0023\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0014\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 426ms/step - loss: 0.0550\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0700\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0196\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0293\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0203\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0011\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0064\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0139\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0106\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0044\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0068\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0135\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0117\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0063\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0062\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0087\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0079\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0039\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0013\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0034\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.9741e-05\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0026\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0022\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.2574e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0012\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.1958e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.4022e-05\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.8586e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.8456e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 2.1603e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.0367e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.7126e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.3940e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.6714e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.0589e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.7954e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.0465e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.4440e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.1415e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.1205e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.3259e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 452ms/step - loss: 0.0602\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0599\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0600\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0273\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0060\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0190\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0222\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0066\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.9426e-04\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0079\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0146\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0134\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0074\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0042\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0078\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0116\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0091\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0041\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0028\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0047\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0057\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0038\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.6607e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8450e-04\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0018\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0029\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0017\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.0005e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.9939e-04\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.3384e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.1084e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.9743e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.0047e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.7355e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.7277e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.4956e-05\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.0072e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.7760e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.2933e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.9006e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 2.6275e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 4.4347e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.7498e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.4868e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 460ms/step - loss: 0.7954\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.5204\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1938\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0821\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1367\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1786\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1974\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2042\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2050\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2027\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1985\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1927\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1850\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1745\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1598\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1389\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1089\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0689\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0252\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 8.6571e-04\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0209\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0615\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0743\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0517\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0181\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0014\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0077\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0232\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0350\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0378\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0318\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0197\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0071\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0049\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0137\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0181\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0141\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0059\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.4060e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.3200e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0046\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0081\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0089\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0032\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.2066e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.0032e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0046\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 441ms/step - loss: 0.3206\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2945\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2430\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1921\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1656\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1358\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0980\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0651\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0466\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0338\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0176\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0048\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0020\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0047\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0066\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0071\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0096\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0152\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0199\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0207\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0205\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0215\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0221\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0204\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0171\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0144\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0126\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0102\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0069\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0043\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0029\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0020\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.7830e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.2379e-05\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 2.1127e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.8317e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0013\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0020\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0027\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0018\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0013\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0011\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 7.1500e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 3.3268e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 411ms/step - loss: 0.8108\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.7194\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.6003\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.6150\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6257\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5712\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5302\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5226\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5200\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5035\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4719\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4324\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3967\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3724\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3500\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3155\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2694\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2224\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1824\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1478\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.1138\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0790\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0469\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0232\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0111\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0071\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0053\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0046\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0083\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0164\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0251\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0310\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0337\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0351\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0369\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0379\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0362\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0320\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0274\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0236\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0197\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0150\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0101\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0061\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0036\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.9292e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.2612e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.5533e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 423ms/step - loss: 0.4431\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3347\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3588\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2984\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2377\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2104\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1887\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1551\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1137\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0775\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0574\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0469\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0320\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0138\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0031\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0072\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0096\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0094\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0099\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0141\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0202\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0231\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0222\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0210\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0214\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0219\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0203\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0168\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0132\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0110\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0095\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0073\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0043\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.5888e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.8019e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.0509e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0011\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0028\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0029\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0025\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 429ms/step - loss: 0.6170\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5011\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3633\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3679\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3614\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2898\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2248\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1949\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1787\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1549\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1204\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0811\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0481\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0320\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0296\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0256\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0138\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0025\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.6373e-04\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0073\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0149\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0188\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0185\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0171\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0182\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0223\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0256\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0246\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0207\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0177\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0167\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0162\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0142\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0107\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0069\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0046\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0039\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0034\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.9496e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.0031e-06\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.4370e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0014\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0017\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0027\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0029\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0025\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 425ms/step - loss: 0.5858\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2410\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0657\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1067\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1509\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1708\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1770\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1757\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1692\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1579\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1407\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1159\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0819\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0412\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0080\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0070\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0388\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0625\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0539\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0257\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0041\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0033\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0158\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0284\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0338\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0305\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0206\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0091\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0024\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0120\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0165\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0135\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0062\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 9.0717e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.9730e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0039\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0070\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0078\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0058\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0027\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.7178e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.5536e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0043\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0040\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.8093e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.5206e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0013\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 430ms/step - loss: 0.8003\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.5336\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2083\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0830\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1345\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1782\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1981\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2053\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2061\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2036\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1989\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1926\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1842\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1731\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1579\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1364\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1061\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0661\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0234\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.8958e-04\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0212\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0606\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0730\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0511\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0184\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0019\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0075\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0226\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0343\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0374\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0318\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0202\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0078\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0013\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0043\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0127\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0175\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0144\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0066\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.5880e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.9607e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0039\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0075\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0087\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0070\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0037\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.4533e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.5563e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0022\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0042\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 432ms/step - loss: 0.0912\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0937\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0422\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0365\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0287\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0065\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0047\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0105\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0086\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0038\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0064\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0135\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0134\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0098\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0105\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0127\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0115\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0076\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0050\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0052\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0047\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0020\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.7923e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.4755e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0015\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0026\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0020\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0023\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.6391e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.1352e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.7857e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.5463e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.8451e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 1.7061e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.9227e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.8421e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.0266e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.3755e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.9779e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 4.6204e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 3.8901e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 4.7545e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.2110e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.2191e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 400ms/step - loss: 0.9217\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.6868\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3664\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1957\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2362\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2875\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3122\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3210\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3212\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3164\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3083\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2976\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2847\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2694\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2515\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2303\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2045\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1727\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1332\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0871\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0445\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0271\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0456\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0679\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0616\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0320\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0056\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0142\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0297\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0379\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0360\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0262\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0145\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0090\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0137\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0220\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0240\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0170\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0077\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0031\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 28ms/step - loss: 0.0044\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0077\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0093\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0078\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.7575e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.3130e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0046\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 433ms/step - loss: 0.7527\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4526\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1556\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1015\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1573\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1916\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2055\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2089\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2068\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2014\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1938\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1838\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1709\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1537\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1302\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0982\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0577\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0177\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0260\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0610\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0671\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0433\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0139\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0102\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0249\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0349\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0363\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0293\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0173\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0062\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0020\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0067\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0143\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0169\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0121\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0046\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.7210e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0047\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0076\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0080\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0059\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0028\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.5072e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.7502e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0045\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0043\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 882ms/step - loss: 1.0867\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.9936\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.7969\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5014\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2825\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2957\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3619\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4003\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.4191\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4277\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4306\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4301\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4271\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4221\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4152\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4063\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3950\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3805\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3618\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3373\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3050\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.2638\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2180\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1847\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1865\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2123\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.2187\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1905\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1475\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1171\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1072\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1067\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1042\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0954\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0796\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0583\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0349\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0157\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0076\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0110\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0161\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0146\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0090\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0079\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0142\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0226\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0269\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0251\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0198\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0153\n",
      "16/16 [==============================] - 0s 890us/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 475ms/step - loss: 0.8148\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5405\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2189\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1118\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1660\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2065\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2236\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2282\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2264\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2213\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2141\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2054\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1950\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1824\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1664\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1451\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1162\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0778\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0342\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0036\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0131\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0518\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0717\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0558\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0236\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0036\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0218\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0347\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0389\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0338\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0221\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0093\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0057\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0139\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0182\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0144\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0063\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.1422e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 7.7224e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0044\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0078\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0086\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0066\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0032\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.7994e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.8109e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0028\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0047\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 536ms/step - loss: 0.5214\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1883\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0274\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0708\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1181\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1416\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1509\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1512\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1445\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1303\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1076\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0753\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0367\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0060\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0057\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0360\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0598\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0538\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0277\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0052\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.5646e-04\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0107\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0228\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0294\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0284\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0209\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0104\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0131\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0140\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0090\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0028\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.3464e-05\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.0015\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0048\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0070\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0044\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0016\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.2363e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.4922e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0038\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0031\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0014\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.2947e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.6976e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 557ms/step - loss: 0.2392\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2246\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1616\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1462\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1108\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0718\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0530\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0391\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0210\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0066\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0041\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0064\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0043\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0035\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0086\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0142\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0159\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0157\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0173\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0196\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0188\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0159\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0142\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0133\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.0110\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0077\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0052\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0041\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0029\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.4729e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.1713e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.0177e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.8556e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.9388e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0024\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0026\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0017\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0015\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.7825e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.0018e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.1309e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.5782e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.5573e-06\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 537ms/step - loss: 0.5046\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1810\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0953\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1454\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1763\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1862\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1842\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1746\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1585\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1353\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1039\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0650\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0269\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0096\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0262\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0514\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0524\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0299\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0068\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0110\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0240\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0311\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0297\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0213\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0108\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0050\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0075\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0144\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0176\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0134\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0061\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0016\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0020\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0050\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0072\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0068\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0043\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.2144e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0013\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0034\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0043\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0032\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.2046e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0022\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0026\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0021\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 463ms/step - loss: 0.8452\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.7758\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.6630\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5959\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.6644\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6306\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.5714\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5490\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5525\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5497\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.5305\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4971\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4584\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4271\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4080\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3878\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3527\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3057\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2587\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2186\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1827\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1455\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1061\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0681\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0382\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0204\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0120\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0068\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0034\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0050\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0126\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0223\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0298\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0338\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0360\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0383\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0404\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0401\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0365\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0317\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0276\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0236\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0190\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0136\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0086\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0051\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0031\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.6802e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.2837e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 482ms/step - loss: 0.5286\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1920\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0310\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0758\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1226\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1454\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1542\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1543\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1476\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1339\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1118\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0801\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0413\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0083\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0044\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0338\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0598\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0556\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0295\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0060\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.6325e-04\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0106\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0230\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0300\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0292\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0215\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0108\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0072\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0137\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0144\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0089\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0026\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.5449e-05\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0052\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0073\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0042\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0013\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.1725e-05\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0030\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0039\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0012\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8211e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.3256e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 495ms/step - loss: 0.6390\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2939\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.1093\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1477\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1913\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2096\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2137\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2101\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2019\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1898\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1735\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1520\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1233\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0861\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0437\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0114\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0133\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0443\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0623\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0485\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0195\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0021\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0061\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0207\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0327\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0359\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0300\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0185\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0079\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0050\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0108\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0177\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0179\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0110\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0035\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 8.5519e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0031\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0066\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0082\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0068\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0036\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.2170e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.1869e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0023\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0044\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0047\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.9071e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 49ms/step - loss: 0.0016\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 0s 448ms/step - loss: 0.6448\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5120\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5161\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5027\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4399\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3996\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3837\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3632\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3278\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2820\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2379\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2057\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1804\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1487\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1098\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0743\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0498\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0341\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0218\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0105\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.0831e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0051\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0116\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0154\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0177\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0215\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0263\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 38ms/step - loss: 0.0296\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0299\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0282\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0264\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0252\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0232\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0195\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0152\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0117\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0092\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0041\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 7.1385e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.7384e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.3932e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.0248e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.5608e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0025\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0031\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0032\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 884ms/step - loss: 0.4729\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4792\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.4164\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.3760\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3568\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3241\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.2782\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2346\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.2034\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1733\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1344\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0953\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0661\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0458\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0285\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0129\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0027\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0050\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0081\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.0105\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0153\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0217\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0262\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0276\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0277\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0283\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0284\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0260\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0222\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0190\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0162\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0128\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0089\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0057\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0037\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0023\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.0656e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.3952e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 2.1559e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.5245e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0014\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0018\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0025\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0033\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0036\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0036\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0036\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0034\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0031\n",
      "16/16 [==============================] - 0s 6ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 663ms/step - loss: 0.4612\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1570\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.1102\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.1590\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1837\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1890\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1830\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1688\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1470\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 36ms/step - loss: 0.1168\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0787\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0390\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0158\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0256\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0494\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0522\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0308\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0072\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 3.8282e-04\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.0097\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0226\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0297\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0283\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0200\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.0102\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0057\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0096\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0166\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0183\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0127\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0054\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0021\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0036\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0065\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0077\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0061\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.0030\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 4.3049e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 3.1080e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0022\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0040\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0039\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0023\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.3868e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 9.9062e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0020\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0028\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0026\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0017\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.2634e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 753ms/step - loss: 0.8026\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.6960\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.5996\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6732\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.6434\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5823\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.5609\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.5666\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.5654\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5473\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5153\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.4786\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4498\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4334\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4152\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3818\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3367\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2916\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 29ms/step - loss: 0.2527\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2170\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1789\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1372\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0956\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0604\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0365\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0219\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0113\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0033\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0013\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0068\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0160\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0243\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0296\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0330\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0365\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0400\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0413\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0390\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0350\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0312\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0276\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0231\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0176\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0121\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0078\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0050\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0029\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0010\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.7317e-05\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.0491e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 679ms/step - loss: 1.0985\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.0170\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.8402\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.5574\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.3108\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2941\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.3634\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4073\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.4294\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4401\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4446\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4454\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4436\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.4398\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.4340\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.4261\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4156\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4017\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.3831\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3579\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3241\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2816\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.2390\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2193\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.2388\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.2668\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.2636\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2295\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.1912\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1715\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1699\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1724\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1699\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1593\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1409\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1165\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0898\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0667\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0525\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0458\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0382\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0244\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0090\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 7.7447e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0026\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0100\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 0.0165\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0186\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0178\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.0180\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 966ms/step - loss: 0.6274\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4919\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4534\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.4623\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4027\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3476\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3211\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2998\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2665\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2219\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.1757\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1398\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1164\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0935\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 35ms/step - loss: 0.0642\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.0359\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0181\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0112\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0089\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0066\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0044\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0051\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0107\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0182\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0228\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0239\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0247\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0269\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0286\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0280\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0250\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0215\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0189\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0167\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0136\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0097\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0063\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0043\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0031\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0018\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 26ms/step - loss: 5.8839e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 3.2970e-06\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 3.6859e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0011\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0030\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0034\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0034\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.5194\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1844\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0760\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 30ms/step - loss: 0.1263\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1611\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1742\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1753\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1691\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.1568\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.1381\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1114\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0761\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0361\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0075\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0124\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0425\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0582\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0440\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0171\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0017\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0057\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0190\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0297\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0323\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0266\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0160\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0063\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0036\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0088\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.0154\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0160\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0100\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0032\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.4355e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0025\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0058\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0075\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0064\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0035\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.7584e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 3.0269e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0039\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0043\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0028\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0011\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 5.6085e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0012\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0024\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.4844\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1568\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0620\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1136\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1480\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1612\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1625\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1559\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1424\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1213\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0913\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0535\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0173\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0044\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0262\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0533\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0534\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0300\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0067\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0012\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0110\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0237\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0305\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0290\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0207\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0100\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0031\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0044\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0112\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0158\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.0133\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0065\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.0013\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 8.6071e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0038\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0066\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0072\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0052\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.7211e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.7392e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0027\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0041\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0036\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0019\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.7238e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.9803e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.1181\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1103\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0575\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0611\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0302\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0125\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0144\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0110\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0026\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0019\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0100\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0120\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0090\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0108\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0150\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0153\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0120\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0097\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0103\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0095\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0059\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0034\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0032\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.6243e-05\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.2841e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0024\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0020\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0014\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.2397e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.6007e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.7487e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.5801e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.2667e-05\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.0256e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.9816e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.7310e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.3964e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.6818e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.9627e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.9147e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.2640e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 695ms/step - loss: 0.3796\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0720\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0189\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0849\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1196\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1271\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1172\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0938\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0594\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0217\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.8963e-04\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0150\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0459\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0548\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0364\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0119\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0079\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0192\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0267\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0271\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0209\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0112\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0029\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.9228e-04\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0060\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0121\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0129\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0079\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0020\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.6032e-05\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0053\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0069\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0059\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0032\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 7.0782e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.0905e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0036\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0036\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0020\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.4591e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8164e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0010\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0019\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0019\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0012\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.9050e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 1.1126e-05\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 690ms/step - loss: 0.8133\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.7264\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.6034\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.5936\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.6171\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.5649\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5184\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.5068\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5029\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4853\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4521\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4104\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3729\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.3472\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.3231\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2862\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2384\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1918\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1534\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1208\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0889\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0572\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0298\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0122\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0060\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0058\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0060\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0074\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0130\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0218\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0294\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0335\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0348\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0357\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0370\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0368\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0335\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0287\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0244\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0207\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0166\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0118\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0073\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0042\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 3.9897e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.5476e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 35ms/step - loss: 9.0531e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.0021\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 640ms/step - loss: 0.1557\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.1403e-04\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0451\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0752\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0738\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0514\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0198\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.3625e-04\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0125\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0350\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0338\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0147\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0011\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0032\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0126\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0190\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0183\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0116\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0036\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.6093e-05\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0036\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0092\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0100\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0054\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.5808e-04\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 3.1714e-04\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0030\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0054\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0054\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0032\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.8771e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.2471e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0032\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.4394e-05\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.2414e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 3.3749e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.0370e-05\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.8282e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0010\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 9.3138e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.5668e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.0566e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.9450e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.5104e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 633ms/step - loss: 0.9243\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.7372\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4300\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1645\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1426\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2008\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2336\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2462\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2485\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2457\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2401\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2331\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2254\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.2172\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 25ms/step - loss: 0.2086\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1993\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1886\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1753\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1576\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1327\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0978\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0530\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0110\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0036\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0423\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0773\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0716\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0374\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0080\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0043\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0188\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0344\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0416\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0386\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0274\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0130\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0027\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0113\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0186\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0173\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0091\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.1828e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0036\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0079\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0098\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0084\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0049\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0016\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 618ms/step - loss: 0.0858\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0465\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0486\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0211\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0040\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0152\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0149\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0026\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0107\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0136\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0097\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0055\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0072\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0109\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0091\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0043\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0046\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0048\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.7105e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.6044e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0023\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0022\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.3463e-04\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.4526e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0021\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0025\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 9.6303e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0017\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0010\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.8563e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 3.4658e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 6.8478e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.0341e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.1495e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.3715e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.1781e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.3268e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.7054e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.8821e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 5.0055e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.6010e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.1544e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.5378e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.5711e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.9869e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 599ms/step - loss: 0.9828\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.7965\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4917\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2385\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2298\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2915\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.3261\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.3404\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.3439\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.3414\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3352\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3264\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3155\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3028\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2883\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2720\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2535\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2321\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2069\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1761\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1378\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0915\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0442\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0181\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0336\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0629\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0635\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0352\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0074\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0037\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0195\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0369\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0449\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0411\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0287\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0155\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0107\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0166\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0239\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0228\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0134\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0042\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0082\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0093\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0071\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0034\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.9855e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0016\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 607ms/step - loss: 0.1887\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0256\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0637\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0873\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0785\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0505\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0174\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 9.7459e-04\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0150\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0339\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0303\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0124\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0054\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0147\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0205\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0196\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0133\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0059\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0028\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0058\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0105\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0107\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0062\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0016\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.5121e-04\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0029\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0050\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0050\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.1345e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.6465e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0016\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0031\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0015\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.7428e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.3562e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0020\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0015\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.7413e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.4858e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.8090e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.6140e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 9.0061e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.7499e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.2801e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 1.4493e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.9019e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 663ms/step - loss: 0.4699\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3524\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3848\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3278\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2606\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2294\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2102\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1802\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1393\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0980\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0695\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0565\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0449\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0264\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0091\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0047\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0088\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0103\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0098\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0107\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0155\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0214\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0238\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0224\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0211\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0214\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0219\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0203\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0168\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0130\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0106\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0092\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0073\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0022\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0012\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0010\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.2091e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 3.2218e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.6480e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0017\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0022\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0023\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0026\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0022\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 544ms/step - loss: 0.1191\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1567\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0698\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0609\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0608\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0413\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0149\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0120\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0187\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0107\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0038\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0080\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0160\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0192\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0162\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0114\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0100\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0129\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0141\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0103\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0056\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0051\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0052\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0032\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.9558e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.1735e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0025\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0019\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.8829e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0014\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0029\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0015\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0015\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0016\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.8116e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.4363e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.7602e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 6.0929e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.6468e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.7639e-05\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 9.8484e-05\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 3.9455e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 4.7920e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.1792e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 636ms/step - loss: 0.8782\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.6004\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2836\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1960\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2558\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2962\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3127\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3166\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3135\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3062\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2959\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2832\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2680\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2502\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2291\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2036\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1720\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1326\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0861\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0422\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0240\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0434\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0667\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0599\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0296\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0043\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0024\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 31ms/step - loss: 0.0173\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0329\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0399\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0362\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0251\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0134\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0098\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0162\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0238\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0232\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0144\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0056\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0028\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0053\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0085\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0092\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0068\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0030\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.1212e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.5149e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0035\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0055\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0051\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 607ms/step - loss: 0.6565\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5248\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4379\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4824\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4326\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3612\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3227\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3054\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2815\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2429\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.1949\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1492\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1181\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0999\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0795\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0516\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0255\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0112\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0081\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0091\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0091\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0077\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0078\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0123\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0200\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0255\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0264\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0254\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0259\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0273\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0272\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0246\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0205\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0167\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0142\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0122\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0092\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0057\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0031\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0020\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0016\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.5396e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.7708e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.0998e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0019\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0025\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0028\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0031\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 635ms/step - loss: 0.4286\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.3295\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3453\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2856\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2309\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2056\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1818\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1463\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1059\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0741\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0576\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0452\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0273\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0104\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0032\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0045\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0072\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0075\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0072\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0096\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0155\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0203\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0210\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0202\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0207\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0219\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0214\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0187\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0153\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0130\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0115\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0092\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0062\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0037\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0025\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.5305e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 4.9324e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 6.3531e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0013\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0015\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 26ms/step - loss: 0.0021\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0026\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0027\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0025\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0023\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0022\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0020\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 653ms/step - loss: 0.8908\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.8442\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.7555\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.6502\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6898\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.7011\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.6489\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 32ms/step - loss: 0.6162\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6215\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.6317\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6279\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.6089\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.5813\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.5569\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.5459\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.5413\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.5256\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.4950\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4600\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4296\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4025\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3716\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3327\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2866\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2382\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1931\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1521\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1120\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0721\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0377\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0150\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0050\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0036\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0068\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0130\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0219\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0325\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0415\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0468\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0490\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0499\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0495\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0466\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0414\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0349\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0284\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0222\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0159\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0099\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0052\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 647ms/step - loss: 0.0860\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0103\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0521\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0551\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0316\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0049\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0041\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0249\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0269\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0104\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.6331e-04\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0052\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0136\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0162\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0116\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0041\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.3632e-04\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0036\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0085\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0080\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0031\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 6.9746e-05\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0014\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0042\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0051\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.0033\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.7762e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.6290e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0018\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0031\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.0407e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.0072e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.9703e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0015\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.1023e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.4488e-06\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.6772e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.0323e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.5567e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.5160e-05\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.7529e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.9215e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.3127e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.3740e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 2.9399e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 28ms/step - loss: 2.7313e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 4.2957e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 668ms/step - loss: 0.7051\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3688\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1500\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1738\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2222\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2439\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2495\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2465\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2384\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2266\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2113\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1922\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1682\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1374\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0988\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0553\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0207\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0174\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0440\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0619\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0497\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0212\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0022\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0045\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0189\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0319\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0364\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0316\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0207\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0102\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0071\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0126\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0195\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0195\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0125\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0046\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0036\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0068\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0082\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0067\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0035\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 7.4544e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.8830e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0023\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0044\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0047\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0032\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0015\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 672ms/step - loss: 0.9734\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.7844\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4766\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2244\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2177\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2784\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.3117\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3250\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3277\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3245\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3178\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.3087\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.2977\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2852\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2714\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2562\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2395\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2208\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1992\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1734\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1411\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1004\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0531\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0145\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0133\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0482\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0692\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0532\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0210\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0132\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0318\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0443\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0451\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0352\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0204\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0094\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0094\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0174\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0222\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0174\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0076\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0011\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0014\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0055\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0089\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0091\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0064\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 38ms/step - loss: 0.0029\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0012\n",
      "16/16 [==============================] - 1s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 727ms/step - loss: 0.3124\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0368\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0202\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0796\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1086\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1134\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1016\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0762\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0414\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0092\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0019\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0260\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0492\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0453\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0227\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0039\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0016\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0108\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0206\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0248\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0221\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0142\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0051\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.4287e-04\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0087\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0124\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0099\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0041\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.2166e-04\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.2937e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0036\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0061\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0062\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0041\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0014\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.0622e-05\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.6250e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0027\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0035\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0025\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.3676e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.7792e-05\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 4.6105e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0014\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.8553e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.0652e-05\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8573e-04\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 824ms/step - loss: 0.2694\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0560\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0742\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1099\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1204\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1137\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0939\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0633\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0285\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0065\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0150\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0376\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0406\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0224\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0042\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0017\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0113\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0214\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0248\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0205\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 26ms/step - loss: 0.0118\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0047\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0043\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0097\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 30ms/step - loss: 0.0138\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0116\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0055\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0012\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0014\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0041\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0060\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0056\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0031\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 7.3809e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 2.6098e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0018\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0035\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0035\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0021\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 7.6994e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 7.4585e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0016\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0023\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0020\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0011\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.9613e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 2.6455e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 7.7522e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0011\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 7.7244e-04\n",
      "16/16 [==============================] - 0s 5ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 818ms/step - loss: 0.7372\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4063\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1746\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1909\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2418\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2655\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2722\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.2697\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2618\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2499\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.2345\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2154\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1915\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1615\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1239\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0799\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0391\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0225\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0401\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0622\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0575\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0304\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0056\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0010\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0130\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0275\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0352\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0335\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0244\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0134\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0080\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0120\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0198\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0220\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0160\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0074\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0029\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0038\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0069\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0086\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0074\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0041\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0010\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.4599e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0041\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0048\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0035\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0012\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 729ms/step - loss: 0.0669\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0742\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0301\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0234\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0238\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0047\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0027\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0099\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0096\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0045\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0052\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0115\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0120\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0078\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0072\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0093\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0088\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0053\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0027\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0031\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0035\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.4462e-05\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.0884e-04\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0016\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.5977e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0024\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0021\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0012\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.8688e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.6919e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.9207e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.6035e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.7308e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.7044e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.5192e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.7954e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.4441e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 4.0483e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.4020e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.7737e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 2.4765e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 3.1086e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.8232e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.0723e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 625ms/step - loss: 0.8285\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.5222\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2309\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1982\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2578\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2908\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.3031\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.3044\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2996\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2909\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2792\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2648\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2474\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2265\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2010\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1690\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1292\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0829\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0412\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0270\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0479\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0682\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0583\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0274\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0034\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0023\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0166\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0313\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0377\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0341\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0234\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0123\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0091\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0156\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0234\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0231\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0148\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0062\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0033\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0056\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0087\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0094\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0069\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.2534e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.1279e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0031\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0051\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0048\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0029\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 734ms/step - loss: 0.0193\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0399\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0089\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0085\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0174\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0024\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0040\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0118\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0095\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0026\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0026\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0076\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0056\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.2092e-04\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0013\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0041\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0037\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.1611e-04\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.8444e-04\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0025\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0029\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 9.3467e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.5028e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0021\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.2724e-04\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.2675e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.9821e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 4.0356e-04\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.0123e-05\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 6.2611e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.7215e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.5119e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.0123e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.1487e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.2088e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8903e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.8371e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 28ms/step - loss: 3.6757e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 3.4094e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.8282e-05\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.3286e-05\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.7141e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.8427e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.1857e-05\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.2402e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.1641e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.1423e-05\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.4828e-05\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 749ms/step - loss: 0.0430\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0601\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0517\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0156\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0071\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0252\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0161\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0042\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0134\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0152\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0093\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0037\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0056\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0109\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0098\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0042\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0020\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0041\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0057\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.0683e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0020\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0032\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0019\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.0553e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0024\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0024\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0013\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.4532e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0016\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.3579e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.3986e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.3488e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 7.3231e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 6.9680e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 2.1657e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 8.9967e-05\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 4.7059e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 6.4893e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 3.5547e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.5465e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.2532e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.5854e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.7285e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.8035e-05\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.1473e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 716ms/step - loss: 0.2042\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0075\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0412\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0818\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0891\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0739\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0441\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0123\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.3136e-04\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0192\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0399\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0356\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0156\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0019\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0035\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0129\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0202\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0209\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0153\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0070\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0011\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0072\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0109\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0088\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0035\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.7472e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 9.3812e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0038\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0057\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0052\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0028\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.7284e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.5756e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0018\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0032\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0029\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0013\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.6062e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.3613e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0013\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0018\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0014\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.6317e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.6970e-05\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.3365e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.1018e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0010\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.1356e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 1.2706e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 630ms/step - loss: 0.7519\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.4426\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1510\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1125\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1685\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2005\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2129\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2152\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2124\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2066\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1986\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1885\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1758\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1592\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1366\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1058\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0659\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0239\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0018\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0214\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0584\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0686\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0464\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0157\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0022\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0097\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0250\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0357\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0373\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0301\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0178\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0063\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0024\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0076\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0155\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0174\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0118\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0040\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.5988e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0016\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0054\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0081\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0081\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0055\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0023\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.9568e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0012\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0034\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0048\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0041\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 666ms/step - loss: 0.5589\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4262\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4555\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4153\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3431\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3036\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2852\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2591\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2188\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1712\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1296\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1037\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0867\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0645\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0374\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0166\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0081\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0078\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0086\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0077\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0067\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0089\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0154\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0221\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0248\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0242\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0242\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0256\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0264\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0248\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0212\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0174\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0148\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0129\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0103\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0068\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0039\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0025\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0019\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.8898e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.1726e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.9157e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0015\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0023\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0025\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0029\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0033\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0032\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 695ms/step - loss: 0.6049\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4676\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4004\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4261\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3697\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3044\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2709\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2515\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2232\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1826\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1373\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1001\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0790\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0651\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0456\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0228\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0075\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0041\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0069\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0095\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0096\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0094\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0124\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0189\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0245\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0257\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0242\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0238\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0247\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0246\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0221\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0181\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0143\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0120\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0102\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0076\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0045\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0015\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0013\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 9.3085e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.1874e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.9453e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0011\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0020\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0025\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0025\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0027\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0030\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0032\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 725ms/step - loss: 0.5583\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4194\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4338\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4027\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3302\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2853\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2647\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2393\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2004\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1541\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1128\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0871\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0720\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0531\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0293\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0110\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0046\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0062\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0088\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0092\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0088\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0110\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0170\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0233\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0256\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0246\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0239\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0247\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0251\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0233\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0196\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0156\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0130\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0112\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0087\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0056\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0029\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0014\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.9756e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.7229e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.1994e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0018\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0025\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0029\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0032\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0030\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 737ms/step - loss: 0.6022\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2652\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0465\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0723\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1253\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1527\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1645\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1678\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1653\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1578\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1445\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1237\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0935\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0546\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0162\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0013\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0253\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0587\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0641\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0408\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0122\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 27ms/step - loss: 4.8699e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0075\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0211\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0307\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0324\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0264\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0156\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0050\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.1816e-04\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0049\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0126\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0159\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0120\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0049\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.0327e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.5979e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0072\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0077\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0057\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0025\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 3.5384e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.1088e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0024\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0040\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0038\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0021\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.0675e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 1.9511e-04\n",
      "16/16 [==============================] - 0s 7ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 724ms/step - loss: 0.5622\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2216\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0923\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1403\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1769\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1907\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1920\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1860\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.1746\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1579\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1347\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1035\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0645\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0253\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0063\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0233\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0522\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0565\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0346\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0094\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0014\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0108\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0246\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0329\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0324\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0243\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0131\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0053\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0058\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0126\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0174\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0149\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0075\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 0.0018\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 26ms/step - loss: 0.0010\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0039\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0068\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0075\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0055\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0024\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.0720e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.8981e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0029\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0044\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0040\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0023\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.9337e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.4851e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0017\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0025\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 705ms/step - loss: 0.0634\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0527\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0340\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0075\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0151\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0137\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0012\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0033\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0107\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0107\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0060\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0056\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0101\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0103\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0057\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0040\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0054\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0055\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0028\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.1501e-04\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0011\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0024\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0015\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.3439e-04\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0023\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0022\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0013\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0013\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0019\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0016\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.7916e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.1455e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.3109e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.2476e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.0786e-05\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.4634e-05\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.8403e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.5921e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.0636e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 3.3139e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.7861e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.7338e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 2.4865e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.9131e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.6288e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.8446e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 4.2634e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.1666e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 1.4401e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.0731\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0373\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0588\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0408\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0113\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0036\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0205\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0215\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0067\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.7019e-04\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0075\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0145\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0145\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0087\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0034\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0041\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0087\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.0094\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0049\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0013\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0018\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0041\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0046\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0027\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.8999e-04\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.2670e-04\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0021\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0018\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.4981e-04\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.5215e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0019\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0021\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0012\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 3.6791e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 4.9764e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0011\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0010\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.2428e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.5249e-06\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 3.4109e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.7250e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 5.1028e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.4927e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.3508e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.2906e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.1526e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.6813e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 9.2999e-05\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 1.9733e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 782ms/step - loss: 0.8730\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.7983\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6874\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6702\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.7108\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6703\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.6365\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6405\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.6516\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6491\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.6324\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6094\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.5927\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5893\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.5870\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5714\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5458\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5218\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5032\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4846\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4597\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.4268\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.3886\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3496\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3113\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2696\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2213\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1696\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1210\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0796\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0460\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0203\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0045\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.2346e-05\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0054\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0151\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0250\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0346\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0438\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0506\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0538\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0537\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0519\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0486\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0433\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0361\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0285\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0215\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0149\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0089\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 812ms/step - loss: 0.0766\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0597\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0418\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0133\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0150\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0155\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0028\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0013\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0084\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0105\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0070\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0056\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0096\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0117\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0083\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0056\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0062\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0068\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0047\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0011\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0018\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.1043e-04\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.3806e-04\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0016\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0021\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0015\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0012\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0018\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0021\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.3884e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.5681e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.8236e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.6935e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.3772e-05\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.3362e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.1994e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 2.1936e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 1.2166e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 3.7204e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.2471e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.8079e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.9645e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.1104e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 3.9792e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.0203e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.2777e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.7767e-04\n",
      "16/16 [==============================] - 0s 1ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 701ms/step - loss: 1.0076\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.8007\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4851\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2796\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3143\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3771\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4108\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4267\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4334\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4348\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4325\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4271\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.4185\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4062\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3889\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3649\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3323\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2912\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2502\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2321\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2523\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2808\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2796\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2485\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.2130\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1951\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1950\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1996\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1995\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1916\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1760\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1544\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1303\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1092\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0955\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0875\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0775\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0601\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0389\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0216\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0126\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0103\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.0103\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0095\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0076\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0068\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0103\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0175\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0232\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0235\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 783ms/step - loss: 1.0690\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.9414\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.6896\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3841\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2726\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3375\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3913\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4185\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4313\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4366\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 31ms/step - loss: 0.4375\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4354\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 26ms/step - loss: 0.4309\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.4242\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4150\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.4026\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3861\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3639\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3337\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2942\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2489\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2147\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2171\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2471\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2585\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2337\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1924\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1634\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1557\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1575\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1568\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1487\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.1325\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1096\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0832\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0588\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0428\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0364\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0322\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0225\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0091\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 6.3819e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0020\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0097\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0172\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0202\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0191\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0179\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0197\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0225\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 852ms/step - loss: 0.6119\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5016\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.5503\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4956\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4365\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.4163\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.4027\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3732\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3287\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2809\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.2438\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2178\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.1869\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1453\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1034\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0719\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0509\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0342\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0186\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0059\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.4249e-04\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0029\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0093\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0136\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0161\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0202\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0260\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0307\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0320\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0308\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0292\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0284\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0268\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0231\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0184\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0144\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0116\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0087\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0056\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0012\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.0251e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.0680e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.2855e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 7.1214e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 27ms/step - loss: 0.0015\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0026\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0032\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0035\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0037\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 783ms/step - loss: 0.0270\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0367\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0421\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0132\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0019\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0231\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0181\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0018\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0029\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0120\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0137\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0071\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.6944e-04\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0024\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0081\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0072\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0018\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.2214e-04\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0031\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0052\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0037\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.1940e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 1.9926e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0023\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0033\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0016\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.0298e-05\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.9556e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.8401e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.3216e-05\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 7.5271e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.1948e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.1016e-05\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 2.6289e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.9703e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 6.9819e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.6452e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.8509e-05\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.2528e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.2942e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.7946e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.6019e-06\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.2562e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.6026e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.7046e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 1.2494e-06\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 1.1640e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 668ms/step - loss: 0.2250\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0118\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0419\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0857\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0955\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0822\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0534\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0192\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 1.6026e-04\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0138\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0389\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0411\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0220\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0043\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0022\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0111\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0199\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0227\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0186\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0103\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0027\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.2178e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0055\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0107\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0106\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0056\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.1671e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.8597e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0029\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0055\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0059\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0040\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0013\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 1.3327e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0011\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 28ms/step - loss: 0.0029\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0034\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0021\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 5.2615e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.4558e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 9.2184e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0018\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0018\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 9.8143e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 1.7160e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 5.7728e-05\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.8859e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0010\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.6589e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.2468e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 716ms/step - loss: 0.6248\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2996\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0460\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0498\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1093\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1430\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1592\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1657\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1659\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1607\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1496\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1313\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1038\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0667\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0260\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0012\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 37ms/step - loss: 0.0148\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0512\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0680\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0523\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0220\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0027\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0151\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0270\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0323\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0298\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0209\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0096\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0014\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0011\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0078\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0145\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0149\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0092\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0026\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.6110e-06\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0019\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0054\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0076\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0072\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0047\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.9337e-05\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 8.3933e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0028\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0040\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0033\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0015\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 2.0970e-04\n",
      "16/16 [==============================] - 0s 5ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 714ms/step - loss: 0.1844\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0995\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1145\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0671\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0372\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0346\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0295\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0154\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0024\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0020\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0102\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0115\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0070\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0068\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0121\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0162\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0155\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0122\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0103\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0115\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0120\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0091\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0054\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0040\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0042\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0037\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0019\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.7753e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 3.4837e-04\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0014\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0016\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 9.6750e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 9.5313e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0018\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0024\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0022\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0016\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0019\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0016\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 9.5894e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.9179e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.2034e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.6186e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 2.3747e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 8.6764e-06\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.2133e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.9479e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.4915e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.7056e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 667ms/step - loss: 0.4952\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3797\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4123\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3585\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2923\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2605\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2412\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2111\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1694\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1257\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0933\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0760\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0608\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0391\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0176\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0064\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0054\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0077\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0084\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0075\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0083\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0131\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0198\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0233\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0229\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0222\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0230\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0240\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0228\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0196\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0158\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0132\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0115\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0093\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0061\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0034\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 0.0017\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0012\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 4.7801e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 1.3715e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 6.0399e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0015\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0020\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0021\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0023\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0027\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0029\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0025\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 699ms/step - loss: 0.7588\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4519\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1635\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1235\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1787\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2105\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2226\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2243\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2206\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2137\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2045\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.1932\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1791\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1612\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1377\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1063\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0665\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0250\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0029\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0207\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0559\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0664\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0455\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0160\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0096\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0246\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0354\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0372\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0304\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0183\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0071\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0033\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0082\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0156\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0173\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0118\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0041\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 2.5777e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0015\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0052\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0078\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0078\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0054\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0023\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 5.5155e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0034\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0048\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0042\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 780ms/step - loss: 1.0194\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.8691\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5949\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3006\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2243\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2870\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3314\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3516\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3587\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3587\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.3544\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3473\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3380\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3268\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3139\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.2994\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2829\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2643\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2430\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.2178\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1874\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.1496\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.1034\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0542\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0223\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0319\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0620\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0659\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0385\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0087\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0029\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0185\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0368\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0457\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0423\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0298\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0163\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0116\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0179\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0254\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0235\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0134\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0042\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0020\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0052\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0088\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0094\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0068\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0030\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.7374e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 759ms/step - loss: 0.1465\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0018\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0469\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0734\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0681\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0430\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0125\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.5691e-04\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0181\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0355\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0280\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0092\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.9886e-04\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0058\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0148\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0189\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0161\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0086\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0018\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.9896e-04\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0056\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0098\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0084\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0034\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.4134e-04\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0010\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0039\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0055\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0020\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.7771e-04\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 5.9853e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0024\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0032\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0021\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.1434e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 8.5498e-05\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.7402e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0017\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.8657e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.9830e-05\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 29ms/step - loss: 1.5756e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 7.6539e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0010\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 6.4494e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.2026e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 5.2008e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.8372e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.2325e-04\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 811ms/step - loss: 0.3123\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2737\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.2330\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1820\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1543\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1271\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0924\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0610\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0426\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0308\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0162\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0044\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0017\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0042\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0060\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0065\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0091\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0145\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0186\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0192\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0193\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0205\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0208\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0189\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0159\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0136\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0119\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0094\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0062\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0040\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0029\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0018\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.2325e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.8204e-05\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.4220e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 8.3795e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 3ms/step - loss: 0.0010\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0013\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0020\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0026\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0027\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0026\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0021\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0013\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0010\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.2159e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 2.9910e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 733ms/step - loss: 1.0163\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.8442\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.5493\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2822\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2589\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3249\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3647\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3829\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3898\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3906\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3875\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3816\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3736\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3633\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3507\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3355\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.3168\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2937\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2645\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2274\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1820\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1338\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1008\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1027\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1227\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1205\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0875\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0468\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0238\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0225\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0302\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0361\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0355\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0275\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0149\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0046\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0051\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0166\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0273\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0267\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0181\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0116\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0121\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0160\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0177\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0150\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0091\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0037\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0033\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 2s 2s/step - loss: 0.2579\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0220\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0349\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0825\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1023\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1018\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0859\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0574\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0230\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0120\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0389\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0458\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0280\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0068\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.7535e-04\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0076\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0181\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0234\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0213\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0136\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0047\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 2.7732e-04\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0032\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0095\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0121\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0086\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0029\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.0384e-05\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0014\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0045\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0063\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0055\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0029\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.6704e-04\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.5737e-04\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0017\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0033\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0032\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0016\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.6302e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 1.8972e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0018\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 9.0903e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 1.4568e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.5753e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 6.0305e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0010\n",
      "16/16 [==============================] - 0s 5ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 792ms/step - loss: 0.7972\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.4991\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1989\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.1432\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1999\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2342\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.2473\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2493\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2454\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2380\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2282\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2164\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.2025\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1857\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1649\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.1380\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1031\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0608\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0206\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0063\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0311\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0618\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0623\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0359\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0090\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0025\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0142\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0294\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0380\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0368\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0276\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0149\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0060\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0063\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0136\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0191\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0167\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0087\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.4231e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0038\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0072\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0083\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0066\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0033\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 7.7592e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.3422e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0026\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0046\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0048\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 776ms/step - loss: 0.0845\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0083\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0501\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0558\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0348\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0074\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0231\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0284\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0125\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.9010e-04\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0037\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0124\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0161\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0122\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0046\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 6.7345e-05\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0027\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0081\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0085\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0038\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 1.8712e-04\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0011\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0040\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0052\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0037\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0010\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 1.9767e-05\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0014\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0030\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0025\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 7.9868e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.3816e-06\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 7.5434e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.0742e-04\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 2.0755e-05\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 3.2188e-04\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.9945e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0010\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 3.7991e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.2020e-07\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 2.5835e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 6.4180e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 5.9192e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 2.0348e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 9.6418e-07\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 1.9553e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.1856e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 855ms/step - loss: 0.0921\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.1371\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0510\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0511\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0549\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0371\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0117\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0013\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0135\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0204\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0112\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0038\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0080\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0159\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0185\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0149\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0093\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0078\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0109\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0124\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0087\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0040\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0028\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0042\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0048\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0031\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.2989e-04\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.3718e-05\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0016\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0027\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0021\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 9.8920e-04\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0012\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0023\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0027\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0020\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0011\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0011\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0015\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0014\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 6.1613e-04\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 1.8799e-04\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 27ms/step - loss: 3.8229e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 6.0766e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.1208e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 7.9347e-05\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 1.0364e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 4.0973e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.0947e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 3.2009e-04\n",
      "16/16 [==============================] - 0s 2ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 870ms/step - loss: 0.1973\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0046\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0410\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0808\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0878\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0722\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0421\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0107\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 8.0476e-04\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0215\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0408\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0341\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 20ms/step - loss: 0.0134\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0011\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0043\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0140\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0206\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0202\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0139\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0054\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 4.6113e-04\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0024\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0082\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0110\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0078\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0025\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.2701e-06\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0016\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0045\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 4ms/step - loss: 0.0059\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0048\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 2.3915e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.2123e-04\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0023\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0033\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0025\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.3232e-04\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 3.9726e-05\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 5.8120e-04\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 21ms/step - loss: 0.0015\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0018\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.0012\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 3.1338e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 8.7221e-06\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.4644e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 9.8083e-04\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 39ms/step - loss: 9.4503e-04\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 4.2391e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 3.5878e-05\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 969ms/step - loss: 0.1418\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1874\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0934\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0745\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0747\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0564\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0275\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0068\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0092\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0190\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0147\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0046\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0040\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0118\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0183\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0186\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0144\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0111\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0124\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 24ms/step - loss: 0.0156\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0149\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0102\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0064\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0060\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0067\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0058\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0032\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 8.0564e-04\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 4.9258e-04\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0022\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0013\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 6.1675e-04\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0012\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0024\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0027\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0021\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0016\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0017\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0021\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0018\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0011\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 6.2316e-04\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 6.9870e-04\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 7.5104e-04\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 4.4205e-04\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 6.7141e-05\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 5.0140e-05\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 2.9801e-04\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 3.8276e-04\n",
      "16/16 [==============================] - 0s 4ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 1.0988\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 1.0377\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 23ms/step - loss: 0.8991\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.6481\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3565\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2492\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3099\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.3628\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.3894\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.4013\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.4053\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.4049\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4015\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3961\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3889\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.3802\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3700\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.3583\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.3449\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.3297\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3121\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2918\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.2677\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2386\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.2024\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.1571\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1044\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0577\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0452\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0691\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0806\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0553\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0170\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 5.0398e-05\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0109\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0309\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 5ms/step - loss: 0.0437\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0438\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0327\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0189\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.0142\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0227\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0320\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0292\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0168\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.0066\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0053\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.0090\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0117\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 22ms/step - loss: 0.0103\n",
      "16/16 [==============================] - 0s 3ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.8887\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.8423\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.7540\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.6473\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.6796\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.6946\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.6435\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.6083\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.6108\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.6198\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.6152\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.5954\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5665\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.5401\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5262\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.5190\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.5010\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.4680\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.4299\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3958\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.3651\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.3311\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.2897\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.2420\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.1930\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1490\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1113\n",
      "Epoch 28/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0766\n",
      "Epoch 29/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0440\n",
      "Epoch 30/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0182\n",
      "Epoch 31/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0044\n",
      "Epoch 32/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0022\n",
      "Epoch 33/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0064\n",
      "Epoch 34/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0127\n",
      "Epoch 35/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0198\n",
      "Epoch 36/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0281\n",
      "Epoch 37/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0372\n",
      "Epoch 38/50\n",
      "1/1 [==============================] - 0s 7ms/step - loss: 0.0444\n",
      "Epoch 39/50\n",
      "1/1 [==============================] - 0s 6ms/step - loss: 0.0476\n",
      "Epoch 40/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0478\n",
      "Epoch 41/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0470\n",
      "Epoch 42/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.0452\n",
      "Epoch 43/50\n",
      "1/1 [==============================] - 0s 52ms/step - loss: 0.0415\n",
      "Epoch 44/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0356\n",
      "Epoch 45/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0289\n",
      "Epoch 46/50\n",
      "1/1 [==============================] - 0s 29ms/step - loss: 0.0225\n",
      "Epoch 47/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0168\n",
      "Epoch 48/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.0114\n",
      "Epoch 49/50\n",
      "1/1 [==============================] - 0s 32ms/step - loss: 0.0064\n",
      "Epoch 50/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0028\n",
      "16/16 [==============================] - 0s 9ms/step\n",
      "Epoch 1/50\n",
      "1/1 [==============================] - 1s 1s/step - loss: 0.7626\n",
      "Epoch 2/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.4543\n",
      "Epoch 3/50\n",
      "1/1 [==============================] - 0s 8ms/step - loss: 0.1588\n",
      "Epoch 4/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1183\n",
      "Epoch 5/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1749\n",
      "Epoch 6/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.2068\n",
      "Epoch 7/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.2187\n",
      "Epoch 8/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.2205\n",
      "Epoch 9/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.2171\n",
      "Epoch 10/50\n",
      "1/1 [==============================] - 0s 13ms/step - loss: 0.2106\n",
      "Epoch 11/50\n",
      "1/1 [==============================] - 0s 9ms/step - loss: 0.2021\n",
      "Epoch 12/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1915\n",
      "Epoch 13/50\n",
      "1/1 [==============================] - 0s 12ms/step - loss: 0.1783\n",
      "Epoch 14/50\n",
      "1/1 [==============================] - 0s 11ms/step - loss: 0.1614\n",
      "Epoch 15/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.1387\n",
      "Epoch 16/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.1079\n",
      "Epoch 17/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0677\n",
      "Epoch 18/50\n",
      "1/1 [==============================] - 0s 10ms/step - loss: 0.0249\n",
      "Epoch 19/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0017\n",
      "Epoch 20/50\n",
      "1/1 [==============================] - 0s 30ms/step - loss: 0.0213\n",
      "Epoch 21/50\n",
      "1/1 [==============================] - 0s 15ms/step - loss: 0.0591\n",
      "Epoch 22/50\n",
      "1/1 [==============================] - 0s 16ms/step - loss: 0.0694\n",
      "Epoch 23/50\n",
      "1/1 [==============================] - 0s 17ms/step - loss: 0.0465\n",
      "Epoch 24/50\n",
      "1/1 [==============================] - 0s 18ms/step - loss: 0.0155\n",
      "Epoch 25/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0026\n",
      "Epoch 26/50\n",
      "1/1 [==============================] - 0s 14ms/step - loss: 0.0111\n",
      "Epoch 27/50\n",
      "1/1 [==============================] - 0s 19ms/step - loss: 0.0267\n"
     ]
    }
   ],
   "source": [
    "x_train = np.array([0.1,5])\n",
    "y_train = np.array([0.2,-0.3])\n",
    "for i in range(100):\n",
    "    model = make_model()\n",
    "    model.compile(loss=\"mse\",optimizer=tf.keras.optimizers.Adam(learning_rate=1e-3))\n",
    "    model.fit(x_train,y_train,epochs=50)\n",
    "    y = model.predict(x)\n",
    "    plt.plot(x,y)\n",
    "plt.plot(x_train,y_train,'.r',lw=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/roy/github/RoyStegeman/dl_theory/venv/lib/python3.10/site-packages/sklearn/gaussian_process/kernels.py:420: ConvergenceWarning: The optimal value found for dimension 0 of parameter k2__length_scale is close to the specified lower bound 10. Decreasing the bound and calling fit again may find a better value.\n",
      "  warnings.warn(\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.744**2 * RBF(length_scale=10)"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "kernel = 1 * RBF(length_scale=1.0, length_scale_bounds=(1e-2, 100))\n",
    "gaussian_process = GaussianProcessRegressor(kernel=kernel, n_restarts_optimizer=9)\n",
    "gaussian_process.fit(x_train.reshape(-1, 1), y_train)\n",
    "gaussian_process.kernel_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PolyCollection at 0x7f14a19a7550>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean_prediction, std_prediction = gaussian_process.predict(x.reshape(-1, 1), return_std=True)\n",
    "plt.plot(x, mean_prediction, label=\"Mean prediction\")\n",
    "plt.fill_between(\n",
    "    x.ravel(),\n",
    "    mean_prediction - 1.96 * std_prediction,\n",
    "    mean_prediction + 1.96 * std_prediction,\n",
    "    alpha=0.5,\n",
    "    label=r\"95% confidence interval\",\n",
    ")\n",
    "plt.plot(x_train,y_train,'.r',lw=2)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "venv",
   "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": "3.10.7"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}