example/imputation_1modality.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "id": "6ac3ae1c-badf-4444-834b-672c9475353c",
      "metadata": {},
      "outputs": [],
      "source": [
        "import os\n",
        "import sys\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import scipy as sp\n",
        "import scipy.sparse\n",
        "import h5py\n",
        "\n",
        "import tensorflow as tf\n",
        "import matplotlib.pyplot as plt\n",
        "import scanpy as sc\n",
        "\n",
        "physical_devices = tf.config.list_physical_devices('GPU')\n",
        "try:\n",
        "    tf.config.experimental.set_memory_growth(physical_devices[0], True)\n",
        "except:\n",
        "    # Invalid device or cannot modify virtual devices once initialized.\n",
        "    pass"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "1aa92cf3-dd71-4a00-a3f5-091eb82bb927",
      "metadata": {},
      "source": [
        "# Load data"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "id": "71dcf55d-5daa-418d-a1fa-c7003cc523de",
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<KeysViewHDF5 ['ADT', 'ADT_names', 'RNA.data', 'RNA.indices', 'RNA.indptr', 'RNA.shape', 'batches', 'cell_ids', 'cell_types', 'gene_names', 'id_X_cite', 'id_X_dogma', 'id_Y_asap', 'id_Y_cite', 'id_Y_dogma', 'id_Z_asap', 'id_Z_dogma', 'peak_names', 'peaks.data', 'peaks.indices', 'peaks.indptr', 'peaks.shape', 'sample_sizes']>\n",
            "Sample sizes: [8689 8535]. Feature sizes: [227].\n"
          ]
        }
      ],
      "source": [
        "with h5py.File('dogma_cite_asap.h5', 'r') as f:\n",
        "    print(f.keys())\n",
        "    ADT_names = np.array(f['ADT_names'], dtype='S32').astype(str)\n",
        "    gene_names = np.array(f['gene_names'], dtype='S32').astype(str)\n",
        "    \n",
        "    # Count matrices of three modalities: X for genes, Y for ADT, and Z for peaks.\n",
        "    X = sp.sparse.csc_matrix(\n",
        "        (np.array(f['RNA.data'], dtype=np.float32), \n",
        "         np.array(f['RNA.indices'], dtype=np.int32),\n",
        "         np.array(f['RNA.indptr'], dtype=np.int32)\n",
        "        ), \n",
        "        shape = np.array(f['RNA.shape'], dtype=np.int32)).tocsc().astype(np.float32)\n",
        "    Y = np.array(f['ADT'], dtype=np.float32)\n",
        "    cell_types = np.array(f['cell_types'], dtype='S32').astype(str)\n",
        "    \n",
        "    # The covariate matrix, the last column is the ids of datasets\n",
        "    batches = np.array(f['batches'], dtype=np.float32)\n",
        "    \n",
        "    # The indices of features that are observed in each dataset.\n",
        "    id_X_cite = np.array(f['id_X_cite'], dtype=np.int32)\n",
        "    id_Y_cite = np.array(f['id_Y_cite'], dtype=np.int32)\n",
        "    id_Y_asap = np.array(f['id_Y_asap'], dtype=np.int32)\n",
        "    \n",
        "    # The sample size\n",
        "    sample_sizes = np.array(f['sample_sizes'], dtype=np.int32)\n",
        "\n",
        "# Preprocess the ADT data\n",
        "Y = np.log(Y/np.sum(Y, axis=1, keepdims=True)*1e4+1.)\n",
        "data = Y\n",
        "\n",
        "# Subset the data, only use CITE-seq and ASAP-seq datasets for illustration.\n",
        "batches = batches[sample_sizes[0]:]\n",
        "data = data[sample_sizes[0]:]\n",
        "data_raw = data.copy()\n",
        "sample_sizes = sample_sizes[1:]\n",
        "\n",
        "\n",
        "# Here all the proteins are measured for the two datasets\n",
        "# so id_Y_cite = id_Y_asap with length 227\n",
        "dim_input_arr = np.array([len(ADT_names)])\n",
        "\n",
        "# The masks indicate missing pattern for the whole datasets.\n",
        "masks_raw = np.zeros_like(data, dtype=np.float32)\n",
        "\n",
        "\n",
        "# We randomly remove 10% of entries in each dataset to be missing\n",
        "# Each row corresponds to a cell, and each column corresponds to a feature.\n",
        "# If the dataset is large, one can also provide a mask matrix with shape (num_datasets, num_features)\n",
        "# along with a variable `id_datasets` to reduce memory usages.\n",
        "# -1 means missing and 0 means observed.\n",
        "id_Y_cite_val = np.random.choice(id_Y_cite, int(0.1*len(id_Y_cite)), replace=False)\n",
        "id_Y_asap_val = np.random.choice(id_Y_asap, int(0.1*len(id_Y_asap)), replace=False)\n",
        "\n",
        "masks = masks_raw.copy()\n",
        "masks[:sample_sizes[0], id_Y_cite_val] = -1\n",
        "masks[sample_sizes[0]:, id_Y_asap_val] = -1\n",
        "data = data * (masks != -1)\n",
        "masks = tf.convert_to_tensor(masks, dtype=tf.float32)\n",
        "\n",
        "# data and masks have shape [sum(sample_sizes), sum(dim_input_arr)]\n",
        "print('Sample sizes: {}. Feature sizes: {}.'.format(sample_sizes, dim_input_arr))"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "e9b91486-54c6-4c30-bc50-61c712a7c125",
      "metadata": {},
      "source": [
        "# Set up config for model training"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "id": "78cdecac-c61a-460a-902b-a8b484f530db",
      "metadata": {},
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "1.0.0\n",
            "namespace(beta_kl=2.0, beta_reverse=0.0, beta_modal=array([1.], dtype=float32), p_modal=None, uni_block_names=array(['M-0'], dtype='<U23'), block_names=array(['M-0'], dtype='<U23'), dist_block=array(['NB'], dtype='<U2'), dim_block=array([227]), dim_block_enc=array([64], dtype=int32), dim_block_dec=array([64], dtype=int32), skip_conn=False, max_vals=<tf.Tensor: shape=(1,), dtype=float32, numpy=array([9.0543], dtype=float32)>, gamma=0.0, dim_input_arr=array([227]), dim_block_embed=array([64]), dimensions=array([32], dtype=int32), dim_latent=8, beta_unobs=0.9, p_feat=0.5) (17224, 227) (17224, 227) (17224, 2)\n"
          ]
        }
      ],
      "source": [
        "import sys; sys.path.insert(0, '..') # add parent folder path if not installed via PyPI\n",
        "import scVAEIT\n",
        "print(scVAEIT.__version__)\n",
        "\n",
        "from scVAEIT import VAEIT\n",
        "\n",
        "path_root = 'example/result/1/'\n",
        "\n",
        "config = {\n",
        "    # Dimension of input features for [ADT]\n",
        "    'dim_input_arr': dim_input_arr,\n",
        "\n",
        "    # Blocks for [ADT]\n",
        "    'dim_block': dim_input_arr, # input dimension of blocks\n",
        "    'dist_block':['NB'], # distributions of blocks\n",
        "    'dim_block_enc':np.array([64]), # dimension of first layer of the encoder\n",
        "    'dim_block_dec':np.array([64]), # dimension of first layer of the decoder\n",
        "    'dim_block_embed':np.array([64]), # mask embedding dimension\n",
        "\n",
        "    # Internal network structure\n",
        "    'dimensions':[32], # dimension of latent layers of encoder; the reversed is used for decoder\n",
        "    'dim_latent':8, # the latent dimension bewteen the encoder and decoder\n",
        "\n",
        "    # Weights\n",
        "    'beta_unobs':.9, # weight for masked out observation; weight for observerd values will be 1-beta_unobs.\n",
        "\n",
        "    # Masking probability\n",
        "    'p_feat' : 0.5, # probablity of randomly masking out an entry\n",
        "}\n",
        "\n",
        "batches_cate = batches\n",
        "batches_cont = None\n",
        "\n",
        "model = VAEIT(config, data, masks, None, batches_cate, batches_cont)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "id": "19d70764-04eb-494f-a6c7-94d4571c8c18",
      "metadata": {
        "scrolled": true
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Deleting old log directory at example/result/1/checkpoint/\n",
            "Train - Start of epoch 1\n",
            "33/33 [==============================] - 3s 79ms/step - Reconstructed Loss: 135.8875\n",
            "Epoch 1, Time elapsed: 0.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  143.75,   42.13,   92.60,    9.02,    0.00\n",
            "Train - Start of epoch 2\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 125.0797\n",
            "Epoch 2, Time elapsed: 0.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  123.83,   38.38,   82.83,    2.62,    0.00\n",
            "Train - Start of epoch 3\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 121.5453\n",
            "Epoch 3, Time elapsed: 0.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  119.62,   36.35,   81.60,    1.68,    0.00\n",
            "Train - Start of epoch 4\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 118.9115\n",
            "Epoch 4, Time elapsed: 0.07 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  118.52,   35.98,   81.36,    1.17,    0.00\n",
            "Train - Start of epoch 5\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 118.6656\n",
            "Epoch 5, Time elapsed: 0.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  117.11,   35.79,   80.43,    0.89,    0.00\n",
            "Train - Start of epoch 6\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 117.6111\n",
            "Epoch 6, Time elapsed: 0.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  116.52,   35.60,   80.24,    0.69,    0.00\n",
            "Train - Start of epoch 7\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 114.0384\n",
            "Epoch 7, Time elapsed: 0.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  115.48,   35.42,   79.50,    0.55,    0.00\n",
            "Train - Start of epoch 8\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 115.8321\n",
            "Epoch 8, Time elapsed: 0.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  114.20,   35.26,   78.48,    0.45,    0.00\n",
            "Train - Start of epoch 9\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 115.5469\n",
            "Epoch 9, Time elapsed: 0.12 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  114.92,   35.09,   79.44,    0.38,    0.00\n",
            "Train - Start of epoch 10\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 112.7665\n",
            "Epoch 10, Time elapsed: 0.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  114.11,   34.94,   78.84,    0.33,    0.00\n",
            "Train - Start of epoch 11\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 110.7614\n",
            "Epoch 11, Time elapsed: 0.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  113.05,   34.80,   77.98,    0.28,    0.00\n",
            "Train - Start of epoch 12\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 112.0330\n",
            "Epoch 12, Time elapsed: 0.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  113.91,   34.67,   79.00,    0.25,    0.00\n",
            "Train - Start of epoch 13\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 109.9766\n",
            "Epoch 13, Time elapsed: 0.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  112.67,   34.54,   77.91,    0.22,    0.00\n",
            "Train - Start of epoch 14\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 111.2190\n",
            "Epoch 14, Time elapsed: 0.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  112.13,   34.43,   77.51,    0.19,    0.00\n",
            "Train - Start of epoch 15\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 111.8559\n",
            "Epoch 15, Time elapsed: 0.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  112.28,   34.30,   77.80,    0.18,    0.00\n",
            "Train - Start of epoch 16\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 110.8818\n",
            "Epoch 16, Time elapsed: 0.18 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  111.67,   34.19,   77.32,    0.16,    0.00\n",
            "Train - Start of epoch 17\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 109.5532\n",
            "Epoch 17, Time elapsed: 0.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  111.31,   34.12,   77.04,    0.15,    0.00\n",
            "Train - Start of epoch 18\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 111.8287\n",
            "Epoch 18, Time elapsed: 0.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  111.51,   34.00,   77.37,    0.14,    0.00\n",
            "Train - Start of epoch 19\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 110.8297\n",
            "Epoch 19, Time elapsed: 0.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  109.51,   33.90,   75.49,    0.13,    0.00\n",
            "Train - Start of epoch 20\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 108.6241\n",
            "Epoch 20, Time elapsed: 0.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  109.86,   33.80,   75.94,    0.12,    0.00\n",
            "Train - Start of epoch 21\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 111.1726\n",
            "Epoch 21, Time elapsed: 0.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  110.02,   33.73,   76.18,    0.11,    0.00\n",
            "Train - Start of epoch 22\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 109.6743\n",
            "Epoch 22, Time elapsed: 0.24 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  109.50,   33.66,   75.74,    0.11,    0.00\n",
            "Train - Start of epoch 23\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 107.6827\n",
            "Epoch 23, Time elapsed: 0.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  109.11,   33.56,   75.45,    0.10,    0.00\n",
            "Train - Start of epoch 24\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 109.6301\n",
            "Epoch 24, Time elapsed: 0.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  110.52,   33.48,   76.94,    0.10,    0.00\n",
            "Train - Start of epoch 25\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 107.8580\n",
            "Epoch 25, Time elapsed: 0.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.84,   33.41,   75.34,    0.09,    0.00\n",
            "Train - Start of epoch 26\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.5579\n",
            "Epoch 26, Time elapsed: 0.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.31,   33.36,   74.87,    0.09,    0.00\n",
            "Train - Start of epoch 27\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 107.0618\n",
            "Epoch 27, Time elapsed: 0.29 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.83,   33.26,   75.49,    0.09,    0.00\n",
            "Train - Start of epoch 28\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 107.5824\n",
            "Epoch 28, Time elapsed: 0.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  107.13,   33.20,   73.85,    0.08,    0.00\n",
            "Train - Start of epoch 29\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 110.0125\n",
            "Epoch 29, Time elapsed: 0.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.15,   33.14,   74.92,    0.08,    0.00\n",
            "Train - Start of epoch 30\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 111.9350\n",
            "Epoch 30, Time elapsed: 0.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.34,   33.08,   75.18,    0.08,    0.00\n",
            "Train - Start of epoch 31\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.8684\n",
            "Epoch 31, Time elapsed: 0.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  107.48,   33.02,   74.39,    0.08,    0.00\n",
            "Train - Start of epoch 32\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 109.1703\n",
            "Epoch 32, Time elapsed: 0.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  108.13,   32.97,   75.09,    0.08,    0.00\n",
            "Train - Start of epoch 33\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.4099\n",
            "Epoch 33, Time elapsed: 0.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  107.27,   32.90,   74.29,    0.08,    0.00\n",
            "Train - Start of epoch 34\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 106.5948\n",
            "Epoch 34, Time elapsed: 0.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  106.10,   32.86,   73.16,    0.08,    0.00\n",
            "Train - Start of epoch 35\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 109.1612\n",
            "Epoch 35, Time elapsed: 0.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  107.45,   32.82,   74.55,    0.08,    0.00\n",
            "Train - Start of epoch 36\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.3915\n",
            "Epoch 36, Time elapsed: 0.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  107.41,   32.75,   74.58,    0.09,    0.00\n",
            "Train - Start of epoch 37\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 108.3731\n",
            "Epoch 37, Time elapsed: 0.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  106.27,   32.69,   73.49,    0.09,    0.00\n",
            "Train - Start of epoch 38\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 108.3708\n",
            "Epoch 38, Time elapsed: 0.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  106.16,   32.64,   73.42,    0.10,    0.00\n",
            "Train - Start of epoch 39\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 106.2587\n",
            "Epoch 39, Time elapsed: 0.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.91,   32.58,   73.22,    0.11,    0.00\n",
            "Train - Start of epoch 40\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 105.7582\n",
            "Epoch 40, Time elapsed: 0.41 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.71,   32.54,   73.04,    0.12,    0.00\n",
            "Train - Start of epoch 41\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.9973\n",
            "Epoch 41, Time elapsed: 0.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.68,   32.49,   72.06,    0.13,    0.00\n",
            "Train - Start of epoch 42\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 103.9343\n",
            "Epoch 42, Time elapsed: 0.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.54,   32.44,   72.96,    0.13,    0.00\n",
            "Train - Start of epoch 43\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.2437\n",
            "Epoch 43, Time elapsed: 0.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.75,   32.39,   72.22,    0.14,    0.00\n",
            "Train - Start of epoch 44\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.4741\n",
            "Epoch 44, Time elapsed: 0.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.88,   32.36,   72.39,    0.14,    0.00\n",
            "Train - Start of epoch 45\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.0863\n",
            "Epoch 45, Time elapsed: 0.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.96,   32.32,   73.50,    0.14,    0.00\n",
            "Train - Start of epoch 46\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6775\n",
            "Epoch 46, Time elapsed: 0.47 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.85,   32.27,   73.44,    0.14,    0.00\n",
            "Train - Start of epoch 47\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 107.3482\n",
            "Epoch 47, Time elapsed: 0.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  106.07,   32.25,   73.68,    0.14,    0.00\n",
            "Train - Start of epoch 48\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.2818\n",
            "Epoch 48, Time elapsed: 0.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.50,   32.20,   71.17,    0.14,    0.00\n",
            "Train - Start of epoch 49\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.7405\n",
            "Epoch 49, Time elapsed: 0.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.35,   32.16,   72.05,    0.14,    0.00\n",
            "Train - Start of epoch 50\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6315\n",
            "Epoch 50, Time elapsed: 0.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.34,   32.12,   72.07,    0.14,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=50>\n",
            "Saved checkpoint for epoch 50: example/result/1/checkpoint/ckpt-1\n",
            "Train - Start of epoch 51\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.2025\n",
            "Epoch 51, Time elapsed: 0.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  105.10,   32.10,   72.86,    0.14,    0.00\n",
            "Train - Start of epoch 52\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.4523\n",
            "Epoch 52, Time elapsed: 0.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.11,   32.10,   71.87,    0.14,    0.00\n",
            "Train - Start of epoch 53\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.3870\n",
            "Epoch 53, Time elapsed: 0.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.86,   32.05,   71.67,    0.15,    0.00\n",
            "Train - Start of epoch 54\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.7187\n",
            "Epoch 54, Time elapsed: 0.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.77,   31.99,   71.63,    0.15,    0.00\n",
            "Train - Start of epoch 55\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 105.0993\n",
            "Epoch 55, Time elapsed: 0.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.40,   31.98,   71.28,    0.15,    0.00\n",
            "Train - Start of epoch 56\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.2798\n",
            "Epoch 56, Time elapsed: 0.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.90,   31.93,   72.82,    0.15,    0.00\n",
            "Train - Start of epoch 57\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 104.1571\n",
            "Epoch 57, Time elapsed: 0.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.46,   31.89,   71.41,    0.16,    0.00\n",
            "Train - Start of epoch 58\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.8130\n",
            "Epoch 58, Time elapsed: 0.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.31,   31.86,   71.29,    0.17,    0.00\n",
            "Train - Start of epoch 59\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.1802\n",
            "Epoch 59, Time elapsed: 0.59 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.22,   31.84,   71.19,    0.18,    0.00\n",
            "Train - Start of epoch 60\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 105.3487\n",
            "Epoch 60, Time elapsed: 0.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  104.71,   31.80,   72.74,    0.18,    0.00\n",
            "Train - Start of epoch 61\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.3929\n",
            "Epoch 61, Time elapsed: 0.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.21,   31.76,   71.26,    0.19,    0.00\n",
            "Train - Start of epoch 62\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.9994\n",
            "Epoch 62, Time elapsed: 0.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.33,   31.73,   71.41,    0.18,    0.00\n",
            "Train - Start of epoch 63\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.6897\n",
            "Epoch 63, Time elapsed: 0.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.52,   31.71,   71.62,    0.18,    0.00\n",
            "Train - Start of epoch 64\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.8239\n",
            "Epoch 64, Time elapsed: 0.64 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.87,   31.69,   70.99,    0.19,    0.00\n",
            "Train - Start of epoch 65\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 101.5958\n",
            "Epoch 65, Time elapsed: 0.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.53,   31.67,   70.68,    0.19,    0.00\n",
            "Train - Start of epoch 66\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 104.2596\n",
            "Epoch 66, Time elapsed: 0.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.40,   31.64,   71.58,    0.19,    0.00\n",
            "Train - Start of epoch 67\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.7088\n",
            "Epoch 67, Time elapsed: 0.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.66,   31.62,   70.86,    0.18,    0.00\n",
            "Train - Start of epoch 68\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6033\n",
            "Epoch 68, Time elapsed: 0.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  103.13,   31.61,   71.32,    0.20,    0.00\n",
            "Train - Start of epoch 69\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8412\n",
            "Epoch 69, Time elapsed: 0.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.91,   31.57,   71.16,    0.18,    0.00\n",
            "Train - Start of epoch 70\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 106.5336\n",
            "Epoch 70, Time elapsed: 0.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.88,   31.55,   71.14,    0.19,    0.00\n",
            "Train - Start of epoch 71\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.7006\n",
            "Epoch 71, Time elapsed: 0.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.90,   31.53,   71.18,    0.19,    0.00\n",
            "Train - Start of epoch 72\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.8535\n",
            "Epoch 72, Time elapsed: 0.72 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.42,   31.51,   70.71,    0.19,    0.00\n",
            "Train - Start of epoch 73\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 105.5440\n",
            "Epoch 73, Time elapsed: 0.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.44,   31.49,   70.76,    0.18,    0.00\n",
            "Train - Start of epoch 74\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2270\n",
            "Epoch 74, Time elapsed: 0.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.69,   31.47,   71.03,    0.19,    0.00\n",
            "Train - Start of epoch 75\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.3252\n",
            "Epoch 75, Time elapsed: 0.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.47,   31.47,   70.82,    0.19,    0.00\n",
            "Train - Start of epoch 76\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6819\n",
            "Epoch 76, Time elapsed: 0.75 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.77,   31.43,   71.16,    0.18,    0.00\n",
            "Train - Start of epoch 77\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.0675\n",
            "Epoch 77, Time elapsed: 0.76 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.97,   31.43,   70.36,    0.19,    0.00\n",
            "Train - Start of epoch 78\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 101.9381\n",
            "Epoch 78, Time elapsed: 0.77 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.62,   31.40,   71.03,    0.19,    0.00\n",
            "Train - Start of epoch 79\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.7521\n",
            "Epoch 79, Time elapsed: 0.78 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.53,   31.38,   70.95,    0.19,    0.00\n",
            "Train - Start of epoch 80\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.2606\n",
            "Epoch 80, Time elapsed: 0.79 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.90,   31.40,   70.31,    0.19,    0.00\n",
            "Train - Start of epoch 81\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.1762\n",
            "Epoch 81, Time elapsed: 0.8 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.49,   31.37,   70.93,    0.19,    0.00\n",
            "Train - Start of epoch 82\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.4417\n",
            "Epoch 82, Time elapsed: 0.81 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.21,   31.34,   70.68,    0.19,    0.00\n",
            "Train - Start of epoch 83\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6874\n",
            "Epoch 83, Time elapsed: 0.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.66,   31.34,   71.13,    0.19,    0.00\n",
            "Train - Start of epoch 84\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.9292\n",
            "Epoch 84, Time elapsed: 0.83 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.16,   31.31,   70.66,    0.19,    0.00\n",
            "Train - Start of epoch 85\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6433\n",
            "Epoch 85, Time elapsed: 0.84 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.17,   31.30,   70.68,    0.19,    0.00\n",
            "Train - Start of epoch 86\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.7867\n",
            "Epoch 86, Time elapsed: 0.85 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.07,   31.28,   69.61,    0.19,    0.00\n",
            "Train - Start of epoch 87\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 103.1067\n",
            "Epoch 87, Time elapsed: 0.86 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.25,   31.28,   70.79,    0.19,    0.00\n",
            "Train - Start of epoch 88\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.6727\n",
            "Epoch 88, Time elapsed: 0.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.60,   31.25,   70.16,    0.19,    0.00\n",
            "Train - Start of epoch 89\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2781\n",
            "Epoch 89, Time elapsed: 0.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.60,   31.25,   70.17,    0.18,    0.00\n",
            "Train - Start of epoch 90\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0898\n",
            "Epoch 90, Time elapsed: 0.88 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.58,   31.24,   71.16,    0.19,    0.00\n",
            "Train - Start of epoch 91\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.9524\n",
            "Epoch 91, Time elapsed: 0.89 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.93,   31.22,   70.53,    0.19,    0.00\n",
            "Train - Start of epoch 92\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.7827\n",
            "Epoch 92, Time elapsed: 0.9 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.38,   31.21,   70.99,    0.19,    0.00\n",
            "Train - Start of epoch 93\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.7460\n",
            "Epoch 93, Time elapsed: 0.91 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.87,   31.20,   70.49,    0.19,    0.00\n",
            "Train - Start of epoch 94\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.5473\n",
            "Epoch 94, Time elapsed: 0.92 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.22,   31.19,   70.83,    0.19,    0.00\n",
            "Train - Start of epoch 95\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 106.1342\n",
            "Epoch 95, Time elapsed: 0.93 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.44,   31.17,   71.08,    0.19,    0.00\n",
            "Train - Start of epoch 96\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.4864\n",
            "Epoch 96, Time elapsed: 0.94 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.20,   31.16,   70.85,    0.19,    0.00\n",
            "Train - Start of epoch 97\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.4624\n",
            "Epoch 97, Time elapsed: 0.95 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.28,   31.15,   68.94,    0.19,    0.00\n",
            "Train - Start of epoch 98\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4984\n",
            "Epoch 98, Time elapsed: 0.96 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.17,   31.14,   69.84,    0.19,    0.00\n",
            "Train - Start of epoch 99\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.7014\n",
            "Epoch 99, Time elapsed: 0.97 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.65,   31.13,   70.33,    0.20,    0.00\n",
            "Train - Start of epoch 100\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8755\n",
            "Epoch 100, Time elapsed: 0.98 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.15,   31.12,   70.84,    0.19,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=100>\n",
            "Saved checkpoint for epoch 100: example/result/1/checkpoint/ckpt-2\n",
            "Train - Start of epoch 101\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.6379\n",
            "Epoch 101, Time elapsed: 0.99 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.88,   31.11,   70.58,    0.19,    0.00\n",
            "Train - Start of epoch 102\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9867\n",
            "Epoch 102, Time elapsed: 1.0 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.02,   31.09,   69.74,    0.19,    0.00\n",
            "Train - Start of epoch 103\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 101.1965\n",
            "Epoch 103, Time elapsed: 1.01 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.72,   31.08,   69.45,    0.19,    0.00\n",
            "Train - Start of epoch 104\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9337\n",
            "Epoch 104, Time elapsed: 1.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.06,   31.09,   69.76,    0.21,    0.00\n",
            "Train - Start of epoch 105\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0028\n",
            "Epoch 105, Time elapsed: 1.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.71,   31.07,   69.45,    0.20,    0.00\n",
            "Train - Start of epoch 106\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.5306\n",
            "Epoch 106, Time elapsed: 1.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.57,   31.06,   70.30,    0.20,    0.00\n",
            "Train - Start of epoch 107\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.0951\n",
            "Epoch 107, Time elapsed: 1.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.58,   31.05,   68.33,    0.20,    0.00\n",
            "Train - Start of epoch 108\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.0513\n",
            "Epoch 108, Time elapsed: 1.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.77,   31.03,   70.54,    0.20,    0.00\n",
            "Train - Start of epoch 109\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 101.3800\n",
            "Epoch 109, Time elapsed: 1.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.35,   31.03,   70.11,    0.21,    0.00\n",
            "Train - Start of epoch 110\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.0007\n",
            "Epoch 110, Time elapsed: 1.07 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.72,   31.01,   70.50,    0.21,    0.00\n",
            "Train - Start of epoch 111\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7354\n",
            "Epoch 111, Time elapsed: 1.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.77,   31.00,   69.56,    0.21,    0.00\n",
            "Train - Start of epoch 112\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 102.3419\n",
            "Epoch 112, Time elapsed: 1.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.76,   30.99,   69.55,    0.22,    0.00\n",
            "Train - Start of epoch 113\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4548\n",
            "Epoch 113, Time elapsed: 1.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.87,   30.97,   69.66,    0.24,    0.00\n",
            "Train - Start of epoch 114\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.0591\n",
            "Epoch 114, Time elapsed: 1.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.06,   30.95,   69.87,    0.24,    0.00\n",
            "Train - Start of epoch 115\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5204\n",
            "Epoch 115, Time elapsed: 1.12 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.07,   30.94,   68.89,    0.24,    0.00\n",
            "Train - Start of epoch 116\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.4376\n",
            "Epoch 116, Time elapsed: 1.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.74,   30.92,   69.57,    0.25,    0.00\n",
            "Train - Start of epoch 117\n",
            "33/33 [==============================] - 1s 19ms/step - Reconstructed Loss: 100.4808\n",
            "Epoch 117, Time elapsed: 1.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.17,   30.91,   69.01,    0.25,    0.00\n",
            "Train - Start of epoch 118\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.0076\n",
            "Epoch 118, Time elapsed: 1.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.13,   30.89,   68.98,    0.26,    0.00\n",
            "Train - Start of epoch 119\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.3519\n",
            "Epoch 119, Time elapsed: 1.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.59,   30.90,   69.42,    0.26,    0.00\n",
            "Train - Start of epoch 120\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.3886\n",
            "Epoch 120, Time elapsed: 1.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.40,   30.88,   69.27,    0.26,    0.00\n",
            "Train - Start of epoch 121\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.0472\n",
            "Epoch 121, Time elapsed: 1.18 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.76,   30.86,   69.63,    0.27,    0.00\n",
            "Train - Start of epoch 122\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.8297\n",
            "Epoch 122, Time elapsed: 1.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.04,   30.85,   69.92,    0.27,    0.00\n",
            "Train - Start of epoch 123\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0354\n",
            "Epoch 123, Time elapsed: 1.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  102.00,   30.85,   70.88,    0.27,    0.00\n",
            "Train - Start of epoch 124\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.6159\n",
            "Epoch 124, Time elapsed: 1.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.44,   30.85,   69.32,    0.27,    0.00\n",
            "Train - Start of epoch 125\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4026\n",
            "Epoch 125, Time elapsed: 1.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.96,   30.82,   68.86,    0.27,    0.00\n",
            "Train - Start of epoch 126\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 102.5656\n",
            "Epoch 126, Time elapsed: 1.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.78,   30.81,   70.70,    0.28,    0.00\n",
            "Train - Start of epoch 127\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5581\n",
            "Epoch 127, Time elapsed: 1.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.73,   30.81,   69.65,    0.27,    0.00\n",
            "Train - Start of epoch 128\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.6379\n",
            "Epoch 128, Time elapsed: 1.24 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.01,   30.80,   69.93,    0.28,    0.00\n",
            "Train - Start of epoch 129\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.6324\n",
            "Epoch 129, Time elapsed: 1.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.77,   30.78,   69.71,    0.28,    0.00\n",
            "Train - Start of epoch 130\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.4002\n",
            "Epoch 130, Time elapsed: 1.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.45,   30.78,   69.39,    0.28,    0.00\n",
            "Train - Start of epoch 131\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.6895\n",
            "Epoch 131, Time elapsed: 1.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.39,   30.77,   69.34,    0.28,    0.00\n",
            "Train - Start of epoch 132\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.1308\n",
            "Epoch 132, Time elapsed: 1.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.15,   30.76,   69.12,    0.27,    0.00\n",
            "Train - Start of epoch 133\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.4379\n",
            "Epoch 133, Time elapsed: 1.29 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.88,   30.75,   68.86,    0.28,    0.00\n",
            "Train - Start of epoch 134\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2827\n",
            "Epoch 134, Time elapsed: 1.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.67,   30.73,   68.67,    0.27,    0.00\n",
            "Train - Start of epoch 135\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.1596\n",
            "Epoch 135, Time elapsed: 1.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.19,   30.73,   69.18,    0.28,    0.00\n",
            "Train - Start of epoch 136\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.5787\n",
            "Epoch 136, Time elapsed: 1.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.12,   30.72,   70.11,    0.28,    0.00\n",
            "Train - Start of epoch 137\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1029\n",
            "Epoch 137, Time elapsed: 1.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.32,   30.71,   69.32,    0.29,    0.00\n",
            "Train - Start of epoch 138\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0812\n",
            "Epoch 138, Time elapsed: 1.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.17,   30.71,   69.18,    0.28,    0.00\n",
            "Train - Start of epoch 139\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.5380\n",
            "Epoch 139, Time elapsed: 1.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.89,   30.70,   69.91,    0.28,    0.00\n",
            "Train - Start of epoch 140\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.4919\n",
            "Epoch 140, Time elapsed: 1.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.29,   30.69,   69.33,    0.28,    0.00\n",
            "Train - Start of epoch 141\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6925\n",
            "Epoch 141, Time elapsed: 1.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.75,   30.69,   68.80,    0.27,    0.00\n",
            "Train - Start of epoch 142\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.7804\n",
            "Epoch 142, Time elapsed: 1.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.15,   30.69,   68.18,    0.28,    0.00\n",
            "Train - Start of epoch 143\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.3742\n",
            "Epoch 143, Time elapsed: 1.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.57,   30.67,   69.61,    0.29,    0.00\n",
            "Train - Start of epoch 144\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.1569\n",
            "Epoch 144, Time elapsed: 1.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.99,   30.66,   69.04,    0.29,    0.00\n",
            "Train - Start of epoch 145\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.9880\n",
            "Epoch 145, Time elapsed: 1.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.05,   30.65,   70.12,    0.29,    0.00\n",
            "Train - Start of epoch 146\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.9334\n",
            "Epoch 146, Time elapsed: 1.41 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.89,   30.65,   69.95,    0.29,    0.00\n",
            "Train - Start of epoch 147\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4948\n",
            "Epoch 147, Time elapsed: 1.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.62,   30.65,   68.69,    0.29,    0.00\n",
            "Train - Start of epoch 148\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.2320\n",
            "Epoch 148, Time elapsed: 1.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.29,   30.63,   69.37,    0.28,    0.00\n",
            "Train - Start of epoch 149\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8720\n",
            "Epoch 149, Time elapsed: 1.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.77,   30.63,   68.85,    0.29,    0.00\n",
            "Train - Start of epoch 150\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.5736\n",
            "Epoch 150, Time elapsed: 1.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.85,   30.63,   68.92,    0.30,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=150>\n",
            "Saved checkpoint for epoch 150: example/result/1/checkpoint/ckpt-3\n",
            "Train - Start of epoch 151\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.0939\n",
            "Epoch 151, Time elapsed: 1.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.89,   30.62,   67.98,    0.28,    0.00\n",
            "Train - Start of epoch 152\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 104.2095\n",
            "Epoch 152, Time elapsed: 1.47 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.24,   30.61,   69.34,    0.29,    0.00\n",
            "Train - Start of epoch 153\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5285\n",
            "Epoch 153, Time elapsed: 1.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.20,   30.61,   68.31,    0.28,    0.00\n",
            "Train - Start of epoch 154\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8145\n",
            "Epoch 154, Time elapsed: 1.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.43,   30.61,   70.51,    0.31,    0.00\n",
            "Train - Start of epoch 155\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6788\n",
            "Epoch 155, Time elapsed: 1.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.41,   30.60,   69.51,    0.30,    0.00\n",
            "Train - Start of epoch 156\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0478\n",
            "Epoch 156, Time elapsed: 1.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.74,   30.59,   68.86,    0.29,    0.00\n",
            "Train - Start of epoch 157\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1028\n",
            "Epoch 157, Time elapsed: 1.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.85,   30.59,   68.96,    0.30,    0.00\n",
            "Train - Start of epoch 158\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1133\n",
            "Epoch 158, Time elapsed: 1.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.95,   30.58,   69.07,    0.30,    0.00\n",
            "Train - Start of epoch 159\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3902\n",
            "Epoch 159, Time elapsed: 1.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.22,   30.58,   69.35,    0.30,    0.00\n",
            "Train - Start of epoch 160\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.5717\n",
            "Epoch 160, Time elapsed: 1.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.18,   30.56,   69.32,    0.30,    0.00\n",
            "Train - Start of epoch 161\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.1114\n",
            "Epoch 161, Time elapsed: 1.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.26,   30.56,   69.39,    0.31,    0.00\n",
            "Train - Start of epoch 162\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.7163\n",
            "Epoch 162, Time elapsed: 1.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.26,   30.55,   69.40,    0.31,    0.00\n",
            "Train - Start of epoch 163\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.9551\n",
            "Epoch 163, Time elapsed: 1.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.51,   30.55,   68.65,    0.31,    0.00\n",
            "Train - Start of epoch 164\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0411\n",
            "Epoch 164, Time elapsed: 1.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.28,   30.55,   68.42,    0.31,    0.00\n",
            "Train - Start of epoch 165\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.0652\n",
            "Epoch 165, Time elapsed: 1.59 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.70,   30.54,   67.85,    0.31,    0.00\n",
            "Train - Start of epoch 166\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5760\n",
            "Epoch 166, Time elapsed: 1.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.46,   30.53,   69.62,    0.32,    0.00\n",
            "Train - Start of epoch 167\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1443\n",
            "Epoch 167, Time elapsed: 1.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.18,   30.52,   69.34,    0.33,    0.00\n",
            "Train - Start of epoch 168\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.3287\n",
            "Epoch 168, Time elapsed: 1.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.85,   30.51,   68.01,    0.32,    0.00\n",
            "Train - Start of epoch 169\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3302\n",
            "Epoch 169, Time elapsed: 1.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.63,   30.51,   68.81,    0.31,    0.00\n",
            "Train - Start of epoch 170\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2640\n",
            "Epoch 170, Time elapsed: 1.64 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.45,   30.50,   68.62,    0.33,    0.00\n",
            "Train - Start of epoch 171\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 104.6868\n",
            "Epoch 171, Time elapsed: 1.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  101.05,   30.49,   70.23,    0.33,    0.00\n",
            "Train - Start of epoch 172\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3809\n",
            "Epoch 172, Time elapsed: 1.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.18,   30.49,   68.35,    0.33,    0.00\n",
            "Train - Start of epoch 173\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9010\n",
            "Epoch 173, Time elapsed: 1.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.52,   30.49,   68.69,    0.34,    0.00\n",
            "Train - Start of epoch 174\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1006\n",
            "Epoch 174, Time elapsed: 1.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.19,   30.48,   68.38,    0.34,    0.00\n",
            "Train - Start of epoch 175\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8906\n",
            "Epoch 175, Time elapsed: 1.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.00,   30.47,   69.18,    0.34,    0.00\n",
            "Train - Start of epoch 176\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7960\n",
            "Epoch 176, Time elapsed: 1.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.47,   30.47,   69.65,    0.35,    0.00\n",
            "Train - Start of epoch 177\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.4335\n",
            "Epoch 177, Time elapsed: 1.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.74,   30.46,   68.93,    0.35,    0.00\n",
            "Train - Start of epoch 178\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.7782\n",
            "Epoch 178, Time elapsed: 1.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.54,   30.46,   68.73,    0.34,    0.00\n",
            "Train - Start of epoch 179\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 99.3053\n",
            "Epoch 179, Time elapsed: 1.72 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.25,   30.44,   68.46,    0.35,    0.00\n",
            "Train - Start of epoch 180\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.5001\n",
            "Epoch 180, Time elapsed: 1.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.98,   30.44,   69.18,    0.35,    0.00\n",
            "Train - Start of epoch 181\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.1817\n",
            "Epoch 181, Time elapsed: 1.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.33,   30.44,   68.55,    0.34,    0.00\n",
            "Train - Start of epoch 182\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0266\n",
            "Epoch 182, Time elapsed: 1.75 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.97,   30.44,   69.19,    0.34,    0.00\n",
            "Train - Start of epoch 183\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1699\n",
            "Epoch 183, Time elapsed: 1.76 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.20,   30.43,   68.42,    0.35,    0.00\n",
            "Train - Start of epoch 184\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3613\n",
            "Epoch 184, Time elapsed: 1.77 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.04,   30.43,   69.26,    0.35,    0.00\n",
            "Train - Start of epoch 185\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.3390\n",
            "Epoch 185, Time elapsed: 1.78 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.18,   30.42,   68.41,    0.34,    0.00\n",
            "Train - Start of epoch 186\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7933\n",
            "Epoch 186, Time elapsed: 1.79 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.71,   30.42,   68.94,    0.35,    0.00\n",
            "Train - Start of epoch 187\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 99.1605\n",
            "Epoch 187, Time elapsed: 1.8 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.06,   30.42,   68.30,    0.34,    0.00\n",
            "Train - Start of epoch 188\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4937\n",
            "Epoch 188, Time elapsed: 1.81 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.40,   30.42,   69.62,    0.36,    0.00\n",
            "Train - Start of epoch 189\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6812\n",
            "Epoch 189, Time elapsed: 1.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.31,   30.41,   68.55,    0.35,    0.00\n",
            "Train - Start of epoch 190\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9271\n",
            "Epoch 190, Time elapsed: 1.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.26,   30.40,   69.50,    0.36,    0.00\n",
            "Train - Start of epoch 191\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8027\n",
            "Epoch 191, Time elapsed: 1.83 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.89,   30.40,   69.13,    0.36,    0.00\n",
            "Train - Start of epoch 192\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3676\n",
            "Epoch 192, Time elapsed: 1.84 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.24,   30.39,   68.49,    0.35,    0.00\n",
            "Train - Start of epoch 193\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2870\n",
            "Epoch 193, Time elapsed: 1.85 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.09,   30.40,   68.33,    0.36,    0.00\n",
            "Train - Start of epoch 194\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2228\n",
            "Epoch 194, Time elapsed: 1.86 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.62,   30.40,   68.86,    0.36,    0.00\n",
            "Train - Start of epoch 195\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5841\n",
            "Epoch 195, Time elapsed: 1.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.99,   30.39,   68.25,    0.35,    0.00\n",
            "Train - Start of epoch 196\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.0075\n",
            "Epoch 196, Time elapsed: 1.88 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.41,   30.38,   67.69,    0.34,    0.00\n",
            "Train - Start of epoch 197\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7830\n",
            "Epoch 197, Time elapsed: 1.89 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.07,   30.39,   69.32,    0.37,    0.00\n",
            "Train - Start of epoch 198\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8702\n",
            "Epoch 198, Time elapsed: 1.9 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.48,   30.38,   67.77,    0.34,    0.00\n",
            "Train - Start of epoch 199\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.0308\n",
            "Epoch 199, Time elapsed: 1.91 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.12,   30.37,   69.39,    0.36,    0.00\n",
            "Train - Start of epoch 200\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1156\n",
            "Epoch 200, Time elapsed: 1.92 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.21,   30.37,   68.49,    0.35,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=200>\n",
            "Saved checkpoint for epoch 200: example/result/1/checkpoint/ckpt-4\n",
            "Train - Start of epoch 201\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 100.7282\n",
            "Epoch 201, Time elapsed: 1.93 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.40,   30.37,   67.68,    0.35,    0.00\n",
            "Train - Start of epoch 202\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 99.6742\n",
            "Epoch 202, Time elapsed: 1.94 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.37,   67.84,    0.35,    0.00\n",
            "Train - Start of epoch 203\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0751\n",
            "Epoch 203, Time elapsed: 1.95 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.46,   30.36,   67.76,    0.33,    0.00\n",
            "Train - Start of epoch 204\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9726\n",
            "Epoch 204, Time elapsed: 1.96 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.14,   30.36,   68.42,    0.36,    0.00\n",
            "Train - Start of epoch 205\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0688\n",
            "Epoch 205, Time elapsed: 1.97 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.20,   30.36,   68.49,    0.35,    0.00\n",
            "Train - Start of epoch 206\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 103.7049\n",
            "Epoch 206, Time elapsed: 1.98 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.68,   30.36,   68.97,    0.35,    0.00\n",
            "Train - Start of epoch 207\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.9750\n",
            "Epoch 207, Time elapsed: 1.99 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.69,   30.35,   67.98,    0.36,    0.00\n",
            "Train - Start of epoch 208\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.8764\n",
            "Epoch 208, Time elapsed: 2.0 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.04,   30.35,   68.33,    0.35,    0.00\n",
            "Train - Start of epoch 209\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5656\n",
            "Epoch 209, Time elapsed: 2.01 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.95,   30.34,   68.25,    0.36,    0.00\n",
            "Train - Start of epoch 210\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7496\n",
            "Epoch 210, Time elapsed: 2.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.48,   30.34,   67.80,    0.35,    0.00\n",
            "Train - Start of epoch 211\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8703\n",
            "Epoch 211, Time elapsed: 2.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.89,   30.34,   68.20,    0.35,    0.00\n",
            "Train - Start of epoch 212\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8821\n",
            "Epoch 212, Time elapsed: 2.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.09,   30.33,   68.40,    0.36,    0.00\n",
            "Train - Start of epoch 213\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8146\n",
            "Epoch 213, Time elapsed: 2.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.23,   30.33,   68.54,    0.36,    0.00\n",
            "Train - Start of epoch 214\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2625\n",
            "Epoch 214, Time elapsed: 2.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.29,   30.33,   68.60,    0.36,    0.00\n",
            "Train - Start of epoch 215\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6505\n",
            "Epoch 215, Time elapsed: 2.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.00,   30.33,   68.32,    0.36,    0.00\n",
            "Train - Start of epoch 216\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.7387\n",
            "Epoch 216, Time elapsed: 2.07 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.60,   30.33,   67.92,    0.35,    0.00\n",
            "Train - Start of epoch 217\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6993\n",
            "Epoch 217, Time elapsed: 2.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.21,   30.32,   68.54,    0.35,    0.00\n",
            "Train - Start of epoch 218\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9847\n",
            "Epoch 218, Time elapsed: 2.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.83,   30.32,   69.17,    0.34,    0.00\n",
            "Train - Start of epoch 219\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7957\n",
            "Epoch 219, Time elapsed: 2.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.28,   30.32,   68.60,    0.36,    0.00\n",
            "Train - Start of epoch 220\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5386\n",
            "Epoch 220, Time elapsed: 2.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.07,   30.32,   68.40,    0.35,    0.00\n",
            "Train - Start of epoch 221\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.4315\n",
            "Epoch 221, Time elapsed: 2.12 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.31,   67.97,    0.36,    0.00\n",
            "Train - Start of epoch 222\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.6275\n",
            "Epoch 222, Time elapsed: 2.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.98,   30.31,   68.32,    0.35,    0.00\n",
            "Train - Start of epoch 223\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.3160\n",
            "Epoch 223, Time elapsed: 2.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.05,   30.30,   68.37,    0.37,    0.00\n",
            "Train - Start of epoch 224\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5450\n",
            "Epoch 224, Time elapsed: 2.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.76,   30.31,   69.08,    0.37,    0.00\n",
            "Train - Start of epoch 225\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.4452\n",
            "Epoch 225, Time elapsed: 2.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.59,   30.30,   66.94,    0.35,    0.00\n",
            "Train - Start of epoch 226\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.5246\n",
            "Epoch 226, Time elapsed: 2.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.04,   30.31,   68.39,    0.34,    0.00\n",
            "Train - Start of epoch 227\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 100.3692\n",
            "Epoch 227, Time elapsed: 2.18 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.87,   30.30,   69.22,    0.36,    0.00\n",
            "Train - Start of epoch 228\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.3353\n",
            "Epoch 228, Time elapsed: 2.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :  100.47,   30.30,   69.81,    0.36,    0.00\n",
            "Train - Start of epoch 229\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8861\n",
            "Epoch 229, Time elapsed: 2.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.80,   30.29,   68.18,    0.33,    0.00\n",
            "Train - Start of epoch 230\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2203\n",
            "Epoch 230, Time elapsed: 2.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.73,   30.29,   68.09,    0.35,    0.00\n",
            "Train - Start of epoch 231\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.7987\n",
            "Epoch 231, Time elapsed: 2.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.86,   30.30,   68.20,    0.35,    0.00\n",
            "Train - Start of epoch 232\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1902\n",
            "Epoch 232, Time elapsed: 2.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.18,   30.28,   67.54,    0.35,    0.00\n",
            "Train - Start of epoch 233\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1694\n",
            "Epoch 233, Time elapsed: 2.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.93,   30.29,   68.29,    0.35,    0.00\n",
            "Train - Start of epoch 234\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 97.5522\n",
            "Epoch 234, Time elapsed: 2.24 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.73,   30.29,   69.09,    0.35,    0.00\n",
            "Train - Start of epoch 235\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3437\n",
            "Epoch 235, Time elapsed: 2.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.47,   30.28,   68.83,    0.36,    0.00\n",
            "Train - Start of epoch 236\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7628\n",
            "Epoch 236, Time elapsed: 2.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.70,   30.27,   67.08,    0.35,    0.00\n",
            "Train - Start of epoch 237\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8335\n",
            "Epoch 237, Time elapsed: 2.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.89,   30.28,   68.26,    0.35,    0.00\n",
            "Train - Start of epoch 238\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6956\n",
            "Epoch 238, Time elapsed: 2.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.60,   30.27,   67.98,    0.36,    0.00\n",
            "Train - Start of epoch 239\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.5271\n",
            "Epoch 239, Time elapsed: 2.29 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.70,   30.27,   69.07,    0.37,    0.00\n",
            "Train - Start of epoch 240\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8242\n",
            "Epoch 240, Time elapsed: 2.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.28,   30.27,   68.65,    0.36,    0.00\n",
            "Train - Start of epoch 241\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2778\n",
            "Epoch 241, Time elapsed: 2.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.84,   30.26,   67.23,    0.35,    0.00\n",
            "Train - Start of epoch 242\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8611\n",
            "Epoch 242, Time elapsed: 2.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.93,   30.26,   68.31,    0.35,    0.00\n",
            "Train - Start of epoch 243\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.2284\n",
            "Epoch 243, Time elapsed: 2.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.74,   30.25,   69.11,    0.37,    0.00\n",
            "Train - Start of epoch 244\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.8861\n",
            "Epoch 244, Time elapsed: 2.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.73,   30.25,   68.11,    0.37,    0.00\n",
            "Train - Start of epoch 245\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.9649\n",
            "Epoch 245, Time elapsed: 2.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.59,   30.24,   67.97,    0.37,    0.00\n",
            "Train - Start of epoch 246\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6259\n",
            "Epoch 246, Time elapsed: 2.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.48,   30.25,   68.86,    0.37,    0.00\n",
            "Train - Start of epoch 247\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.1411\n",
            "Epoch 247, Time elapsed: 2.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.90,   30.24,   68.29,    0.37,    0.00\n",
            "Train - Start of epoch 248\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.8911\n",
            "Epoch 248, Time elapsed: 2.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.78,   30.23,   68.17,    0.37,    0.00\n",
            "Train - Start of epoch 249\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.5823\n",
            "Epoch 249, Time elapsed: 2.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.08,   30.24,   68.48,    0.37,    0.00\n",
            "Train - Start of epoch 250\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8642\n",
            "Epoch 250, Time elapsed: 2.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.22,   30.23,   68.62,    0.37,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=250>\n",
            "Saved checkpoint for epoch 250: example/result/1/checkpoint/ckpt-5\n",
            "Train - Start of epoch 251\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.4429\n",
            "Epoch 251, Time elapsed: 2.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.94,   30.22,   67.35,    0.37,    0.00\n",
            "Train - Start of epoch 252\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 99.2348\n",
            "Epoch 252, Time elapsed: 2.41 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.37,   30.22,   68.76,    0.38,    0.00\n",
            "Train - Start of epoch 253\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.8128\n",
            "Epoch 253, Time elapsed: 2.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.22,   67.97,    0.37,    0.00\n",
            "Train - Start of epoch 254\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.7516\n",
            "Epoch 254, Time elapsed: 2.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.89,   30.21,   67.29,    0.38,    0.00\n",
            "Train - Start of epoch 255\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.4878\n",
            "Epoch 255, Time elapsed: 2.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.21,   30.22,   67.62,    0.37,    0.00\n",
            "Train - Start of epoch 256\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1917\n",
            "Epoch 256, Time elapsed: 2.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.44,   30.22,   67.85,    0.37,    0.00\n",
            "Train - Start of epoch 257\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.8811\n",
            "Epoch 257, Time elapsed: 2.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.65,   30.22,   69.06,    0.38,    0.00\n",
            "Train - Start of epoch 258\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5815\n",
            "Epoch 258, Time elapsed: 2.47 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.86,   30.21,   67.28,    0.37,    0.00\n",
            "Train - Start of epoch 259\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3789\n",
            "Epoch 259, Time elapsed: 2.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.30,   30.22,   67.72,    0.37,    0.00\n",
            "Train - Start of epoch 260\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9111\n",
            "Epoch 260, Time elapsed: 2.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.19,   30.21,   68.60,    0.38,    0.00\n",
            "Train - Start of epoch 261\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2450\n",
            "Epoch 261, Time elapsed: 2.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.20,   30.20,   67.62,    0.38,    0.00\n",
            "Train - Start of epoch 262\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4205\n",
            "Epoch 262, Time elapsed: 2.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.59,   30.21,   68.00,    0.38,    0.00\n",
            "Train - Start of epoch 263\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7803\n",
            "Epoch 263, Time elapsed: 2.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.01,   30.21,   67.43,    0.37,    0.00\n",
            "Train - Start of epoch 264\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.3765\n",
            "Epoch 264, Time elapsed: 2.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.97,   30.20,   68.40,    0.37,    0.00\n",
            "Train - Start of epoch 265\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.1040\n",
            "Epoch 265, Time elapsed: 2.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.42,   30.20,   68.83,    0.39,    0.00\n",
            "Train - Start of epoch 266\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.1184\n",
            "Epoch 266, Time elapsed: 2.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.20,   68.07,    0.37,    0.00\n",
            "Train - Start of epoch 267\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 95.9277\n",
            "Epoch 267, Time elapsed: 2.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.24,   30.20,   67.68,    0.37,    0.00\n",
            "Train - Start of epoch 268\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3473\n",
            "Epoch 268, Time elapsed: 2.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.15,   30.20,   67.59,    0.37,    0.00\n",
            "Train - Start of epoch 269\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.8352\n",
            "Epoch 269, Time elapsed: 2.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.89,   30.19,   68.33,    0.37,    0.00\n",
            "Train - Start of epoch 270\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.2195\n",
            "Epoch 270, Time elapsed: 2.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.12,   30.19,   67.55,    0.37,    0.00\n",
            "Train - Start of epoch 271\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.4149\n",
            "Epoch 271, Time elapsed: 2.59 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.96,   30.19,   67.41,    0.36,    0.00\n",
            "Train - Start of epoch 272\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1656\n",
            "Epoch 272, Time elapsed: 2.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.18,   30.18,   68.63,    0.37,    0.00\n",
            "Train - Start of epoch 273\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.4826\n",
            "Epoch 273, Time elapsed: 2.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.39,   30.18,   66.83,    0.37,    0.00\n",
            "Train - Start of epoch 274\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.6792\n",
            "Epoch 274, Time elapsed: 2.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.53,   30.18,   67.98,    0.37,    0.00\n",
            "Train - Start of epoch 275\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.0226\n",
            "Epoch 275, Time elapsed: 2.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.84,   30.18,   69.28,    0.38,    0.00\n",
            "Train - Start of epoch 276\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1889\n",
            "Epoch 276, Time elapsed: 2.64 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.15,   30.18,   67.59,    0.37,    0.00\n",
            "Train - Start of epoch 277\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7404\n",
            "Epoch 277, Time elapsed: 2.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.19,   30.19,   67.63,    0.37,    0.00\n",
            "Train - Start of epoch 278\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6026\n",
            "Epoch 278, Time elapsed: 2.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.22,   30.18,   67.67,    0.36,    0.00\n",
            "Train - Start of epoch 279\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2007\n",
            "Epoch 279, Time elapsed: 2.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.36,   30.18,   68.82,    0.37,    0.00\n",
            "Train - Start of epoch 280\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1945\n",
            "Epoch 280, Time elapsed: 2.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.24,   30.18,   67.69,    0.37,    0.00\n",
            "Train - Start of epoch 281\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6285\n",
            "Epoch 281, Time elapsed: 2.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.17,   68.01,    0.38,    0.00\n",
            "Train - Start of epoch 282\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9740\n",
            "Epoch 282, Time elapsed: 2.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.84,   30.17,   67.30,    0.37,    0.00\n",
            "Train - Start of epoch 283\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.8776\n",
            "Epoch 283, Time elapsed: 2.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.82,   30.17,   68.28,    0.38,    0.00\n",
            "Train - Start of epoch 284\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6517\n",
            "Epoch 284, Time elapsed: 2.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.58,   30.17,   67.04,    0.37,    0.00\n",
            "Train - Start of epoch 285\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.4093\n",
            "Epoch 285, Time elapsed: 2.72 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.63,   30.17,   68.10,    0.37,    0.00\n",
            "Train - Start of epoch 286\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.7385\n",
            "Epoch 286, Time elapsed: 2.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.94,   30.16,   68.40,    0.37,    0.00\n",
            "Train - Start of epoch 287\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4514\n",
            "Epoch 287, Time elapsed: 2.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.08,   30.17,   68.54,    0.37,    0.00\n",
            "Train - Start of epoch 288\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8955\n",
            "Epoch 288, Time elapsed: 2.75 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.33,   30.17,   67.79,    0.37,    0.00\n",
            "Train - Start of epoch 289\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8323\n",
            "Epoch 289, Time elapsed: 2.76 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.94,   30.17,   67.41,    0.35,    0.00\n",
            "Train - Start of epoch 290\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.7735\n",
            "Epoch 290, Time elapsed: 2.77 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.58,   30.16,   68.06,    0.36,    0.00\n",
            "Train - Start of epoch 291\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.8284\n",
            "Epoch 291, Time elapsed: 2.78 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.79,   30.16,   68.26,    0.37,    0.00\n",
            "Train - Start of epoch 292\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 97.0507\n",
            "Epoch 292, Time elapsed: 2.79 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.89,   30.16,   68.36,    0.37,    0.00\n",
            "Train - Start of epoch 293\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6075\n",
            "Epoch 293, Time elapsed: 2.8 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.22,   30.16,   67.71,    0.36,    0.00\n",
            "Train - Start of epoch 294\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.8069\n",
            "Epoch 294, Time elapsed: 2.81 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.94,   30.16,   68.42,    0.37,    0.00\n",
            "Train - Start of epoch 295\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.9955\n",
            "Epoch 295, Time elapsed: 2.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.42,   30.16,   67.89,    0.36,    0.00\n",
            "Train - Start of epoch 296\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.0833\n",
            "Epoch 296, Time elapsed: 2.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.95,   30.16,   68.42,    0.37,    0.00\n",
            "Train - Start of epoch 297\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7946\n",
            "Epoch 297, Time elapsed: 2.83 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.84,   30.15,   68.33,    0.36,    0.00\n",
            "Train - Start of epoch 298\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2683\n",
            "Epoch 298, Time elapsed: 2.84 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.32,   30.16,   67.80,    0.36,    0.00\n",
            "Train - Start of epoch 299\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4957\n",
            "Epoch 299, Time elapsed: 2.85 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.63,   30.15,   68.11,    0.36,    0.00\n",
            "Train - Start of epoch 300\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2931\n",
            "Epoch 300, Time elapsed: 2.86 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.92,   30.15,   67.40,    0.37,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=300>\n",
            "Saved checkpoint for epoch 300: example/result/1/checkpoint/ckpt-6\n",
            "Train - Start of epoch 301\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.0514\n",
            "Epoch 301, Time elapsed: 2.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.12,   30.15,   68.61,    0.36,    0.00\n",
            "Train - Start of epoch 302\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6368\n",
            "Epoch 302, Time elapsed: 2.88 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.73,   30.15,   69.21,    0.37,    0.00\n",
            "Train - Start of epoch 303\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3117\n",
            "Epoch 303, Time elapsed: 2.89 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.60,   30.15,   69.09,    0.37,    0.00\n",
            "Train - Start of epoch 304\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 96.5963\n",
            "Epoch 304, Time elapsed: 2.9 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.02,   30.15,   68.49,    0.37,    0.00\n",
            "Train - Start of epoch 305\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9866\n",
            "Epoch 305, Time elapsed: 2.91 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.98,   30.15,   67.46,    0.37,    0.00\n",
            "Train - Start of epoch 306\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2842\n",
            "Epoch 306, Time elapsed: 2.92 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.11,   30.15,   68.60,    0.36,    0.00\n",
            "Train - Start of epoch 307\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.2372\n",
            "Epoch 307, Time elapsed: 2.93 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.83,   30.15,   67.33,    0.35,    0.00\n",
            "Train - Start of epoch 308\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1596\n",
            "Epoch 308, Time elapsed: 2.94 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   96.77,   30.15,   66.27,    0.35,    0.00\n",
            "Train - Start of epoch 309\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.7002\n",
            "Epoch 309, Time elapsed: 2.95 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.58,   30.14,   67.08,    0.36,    0.00\n",
            "Train - Start of epoch 310\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5447\n",
            "Epoch 310, Time elapsed: 2.96 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.57,   30.14,   68.07,    0.36,    0.00\n",
            "Train - Start of epoch 311\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.7575\n",
            "Epoch 311, Time elapsed: 2.97 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.99,   30.15,   67.49,    0.36,    0.00\n",
            "Train - Start of epoch 312\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 102.1304\n",
            "Epoch 312, Time elapsed: 2.98 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.62,   30.14,   69.11,    0.36,    0.00\n",
            "Train - Start of epoch 313\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3797\n",
            "Epoch 313, Time elapsed: 2.99 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.94,   30.14,   68.45,    0.36,    0.00\n",
            "Train - Start of epoch 314\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7843\n",
            "Epoch 314, Time elapsed: 3.0 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.86,   30.14,   67.37,    0.35,    0.00\n",
            "Train - Start of epoch 315\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.0427\n",
            "Epoch 315, Time elapsed: 3.01 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.52,   30.14,   68.01,    0.37,    0.00\n",
            "Train - Start of epoch 316\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.6499\n",
            "Epoch 316, Time elapsed: 3.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.45,   30.15,   67.94,    0.36,    0.00\n",
            "Train - Start of epoch 317\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6659\n",
            "Epoch 317, Time elapsed: 3.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.64,   30.14,   67.15,    0.35,    0.00\n",
            "Train - Start of epoch 318\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.3696\n",
            "Epoch 318, Time elapsed: 3.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.52,   30.14,   68.03,    0.36,    0.00\n",
            "Train - Start of epoch 319\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5709\n",
            "Epoch 319, Time elapsed: 3.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.96,   30.13,   68.47,    0.36,    0.00\n",
            "Train - Start of epoch 320\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1398\n",
            "Epoch 320, Time elapsed: 3.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.68,   30.13,   67.19,    0.36,    0.00\n",
            "Train - Start of epoch 321\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.9405\n",
            "Epoch 321, Time elapsed: 3.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.16,   30.13,   67.68,    0.35,    0.00\n",
            "Train - Start of epoch 322\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9840\n",
            "Epoch 322, Time elapsed: 3.07 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.47,   30.14,   67.96,    0.37,    0.00\n",
            "Train - Start of epoch 323\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7505\n",
            "Epoch 323, Time elapsed: 3.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.29,   30.13,   67.80,    0.35,    0.00\n",
            "Train - Start of epoch 324\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6245\n",
            "Epoch 324, Time elapsed: 3.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.60,   30.13,   68.11,    0.36,    0.00\n",
            "Train - Start of epoch 325\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3391\n",
            "Epoch 325, Time elapsed: 3.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.95,   30.13,   68.46,    0.35,    0.00\n",
            "Train - Start of epoch 326\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.0484\n",
            "Epoch 326, Time elapsed: 3.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.15,   30.13,   67.65,    0.37,    0.00\n",
            "Train - Start of epoch 327\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.1668\n",
            "Epoch 327, Time elapsed: 3.12 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.58,   30.13,   68.10,    0.35,    0.00\n",
            "Train - Start of epoch 328\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3674\n",
            "Epoch 328, Time elapsed: 3.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.03,   30.12,   68.55,    0.36,    0.00\n",
            "Train - Start of epoch 329\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7302\n",
            "Epoch 329, Time elapsed: 3.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.23,   30.12,   67.75,    0.36,    0.00\n",
            "Train - Start of epoch 330\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5449\n",
            "Epoch 330, Time elapsed: 3.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.74,   30.13,   67.26,    0.35,    0.00\n",
            "Train - Start of epoch 331\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 99.1587\n",
            "Epoch 331, Time elapsed: 3.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.70,   30.12,   68.21,    0.37,    0.00\n",
            "Train - Start of epoch 332\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.9982\n",
            "Epoch 332, Time elapsed: 3.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.17,   30.12,   67.70,    0.34,    0.00\n",
            "Train - Start of epoch 333\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.6396\n",
            "Epoch 333, Time elapsed: 3.18 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.24,   30.12,   67.76,    0.36,    0.00\n",
            "Train - Start of epoch 334\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.9400\n",
            "Epoch 334, Time elapsed: 3.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.13,   30.13,   68.63,    0.37,    0.00\n",
            "Train - Start of epoch 335\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.5131\n",
            "Epoch 335, Time elapsed: 3.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.23,   30.13,   67.75,    0.36,    0.00\n",
            "Train - Start of epoch 336\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8961\n",
            "Epoch 336, Time elapsed: 3.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.05,   30.12,   68.56,    0.37,    0.00\n",
            "Train - Start of epoch 337\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9158\n",
            "Epoch 337, Time elapsed: 3.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.16,   30.12,   67.68,    0.35,    0.00\n",
            "Train - Start of epoch 338\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.7098\n",
            "Epoch 338, Time elapsed: 3.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.43,   30.11,   67.96,    0.36,    0.00\n",
            "Train - Start of epoch 339\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1312\n",
            "Epoch 339, Time elapsed: 3.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.12,   68.16,    0.35,    0.00\n",
            "Train - Start of epoch 340\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.0809\n",
            "Epoch 340, Time elapsed: 3.24 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.19,   30.12,   67.71,    0.35,    0.00\n",
            "Train - Start of epoch 341\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.9373\n",
            "Epoch 341, Time elapsed: 3.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.31,   30.11,   67.84,    0.36,    0.00\n",
            "Train - Start of epoch 342\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2752\n",
            "Epoch 342, Time elapsed: 3.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.52,   30.11,   69.05,    0.36,    0.00\n",
            "Train - Start of epoch 343\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1346\n",
            "Epoch 343, Time elapsed: 3.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.02,   30.11,   67.55,    0.36,    0.00\n",
            "Train - Start of epoch 344\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9999\n",
            "Epoch 344, Time elapsed: 3.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.05,   30.12,   67.58,    0.35,    0.00\n",
            "Train - Start of epoch 345\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3533\n",
            "Epoch 345, Time elapsed: 3.29 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.52,   30.12,   68.04,    0.36,    0.00\n",
            "Train - Start of epoch 346\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1633\n",
            "Epoch 346, Time elapsed: 3.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.26,   30.11,   67.80,    0.36,    0.00\n",
            "Train - Start of epoch 347\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.3326\n",
            "Epoch 347, Time elapsed: 3.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.63,   30.11,   68.15,    0.36,    0.00\n",
            "Train - Start of epoch 348\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4633\n",
            "Epoch 348, Time elapsed: 3.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.12,   68.09,    0.35,    0.00\n",
            "Train - Start of epoch 349\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.7627\n",
            "Epoch 349, Time elapsed: 3.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.87,   30.11,   68.41,    0.36,    0.00\n",
            "Train - Start of epoch 350\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2417\n",
            "Epoch 350, Time elapsed: 3.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.58,   30.11,   68.10,    0.37,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=350>\n",
            "Saved checkpoint for epoch 350: example/result/1/checkpoint/ckpt-7\n",
            "Train - Start of epoch 351\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 94.2639\n",
            "Epoch 351, Time elapsed: 3.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.28,   30.11,   67.82,    0.36,    0.00\n",
            "Train - Start of epoch 352\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9052\n",
            "Epoch 352, Time elapsed: 3.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.09,   30.10,   68.63,    0.36,    0.00\n",
            "Train - Start of epoch 353\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 102.1590\n",
            "Epoch 353, Time elapsed: 3.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.07,   30.10,   67.61,    0.35,    0.00\n",
            "Train - Start of epoch 354\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8952\n",
            "Epoch 354, Time elapsed: 3.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.63,   30.11,   67.17,    0.35,    0.00\n",
            "Train - Start of epoch 355\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.9887\n",
            "Epoch 355, Time elapsed: 3.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.80,   30.10,   67.34,    0.36,    0.00\n",
            "Train - Start of epoch 356\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8534\n",
            "Epoch 356, Time elapsed: 3.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.75,   30.11,   68.29,    0.35,    0.00\n",
            "Train - Start of epoch 357\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.9279\n",
            "Epoch 357, Time elapsed: 3.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.35,   30.10,   67.89,    0.35,    0.00\n",
            "Train - Start of epoch 358\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.0502\n",
            "Epoch 358, Time elapsed: 3.41 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.04,   30.11,   68.58,    0.36,    0.00\n",
            "Train - Start of epoch 359\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.9955\n",
            "Epoch 359, Time elapsed: 3.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.23,   30.10,   66.78,    0.35,    0.00\n",
            "Train - Start of epoch 360\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.0991\n",
            "Epoch 360, Time elapsed: 3.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.66,   30.10,   68.21,    0.35,    0.00\n",
            "Train - Start of epoch 361\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6751\n",
            "Epoch 361, Time elapsed: 3.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.35,   30.10,   67.89,    0.36,    0.00\n",
            "Train - Start of epoch 362\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.0304\n",
            "Epoch 362, Time elapsed: 3.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.05,   30.10,   68.58,    0.37,    0.00\n",
            "Train - Start of epoch 363\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 99.8547\n",
            "Epoch 363, Time elapsed: 3.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.46,   30.10,   68.01,    0.36,    0.00\n",
            "Train - Start of epoch 364\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1774\n",
            "Epoch 364, Time elapsed: 3.47 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.61,   30.10,   67.15,    0.36,    0.00\n",
            "Train - Start of epoch 365\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2176\n",
            "Epoch 365, Time elapsed: 3.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.69,   30.10,   68.23,    0.36,    0.00\n",
            "Train - Start of epoch 366\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6230\n",
            "Epoch 366, Time elapsed: 3.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.57,   30.09,   68.12,    0.35,    0.00\n",
            "Train - Start of epoch 367\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.4495\n",
            "Epoch 367, Time elapsed: 3.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.10,   30.10,   67.64,    0.36,    0.00\n",
            "Train - Start of epoch 368\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2721\n",
            "Epoch 368, Time elapsed: 3.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.32,   30.10,   68.86,    0.36,    0.00\n",
            "Train - Start of epoch 369\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.1307\n",
            "Epoch 369, Time elapsed: 3.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.22,   30.10,   67.77,    0.36,    0.00\n",
            "Train - Start of epoch 370\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8213\n",
            "Epoch 370, Time elapsed: 3.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.16,   30.10,   68.70,    0.37,    0.00\n",
            "Train - Start of epoch 371\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9730\n",
            "Epoch 371, Time elapsed: 3.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.98,   30.09,   67.53,    0.36,    0.00\n",
            "Train - Start of epoch 372\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8306\n",
            "Epoch 372, Time elapsed: 3.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.92,   30.09,   68.46,    0.37,    0.00\n",
            "Train - Start of epoch 373\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6665\n",
            "Epoch 373, Time elapsed: 3.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.14,   30.09,   67.68,    0.37,    0.00\n",
            "Train - Start of epoch 374\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2132\n",
            "Epoch 374, Time elapsed: 3.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.49,   30.09,   68.05,    0.35,    0.00\n",
            "Train - Start of epoch 375\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2680\n",
            "Epoch 375, Time elapsed: 3.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.94,   30.08,   67.51,    0.35,    0.00\n",
            "Train - Start of epoch 376\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9768\n",
            "Epoch 376, Time elapsed: 3.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.97,   30.08,   67.53,    0.36,    0.00\n",
            "Train - Start of epoch 377\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.6309\n",
            "Epoch 377, Time elapsed: 3.59 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.80,   30.09,   67.36,    0.35,    0.00\n",
            "Train - Start of epoch 378\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6736\n",
            "Epoch 378, Time elapsed: 3.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.98,   30.08,   68.53,    0.37,    0.00\n",
            "Train - Start of epoch 379\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1572\n",
            "Epoch 379, Time elapsed: 3.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.91,   30.09,   67.47,    0.36,    0.00\n",
            "Train - Start of epoch 380\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3858\n",
            "Epoch 380, Time elapsed: 3.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.42,   30.08,   66.98,    0.36,    0.00\n",
            "Train - Start of epoch 381\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9507\n",
            "Epoch 381, Time elapsed: 3.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.34,   30.08,   67.89,    0.37,    0.00\n",
            "Train - Start of epoch 382\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.3812\n",
            "Epoch 382, Time elapsed: 3.64 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.46,   30.09,   68.01,    0.36,    0.00\n",
            "Train - Start of epoch 383\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5014\n",
            "Epoch 383, Time elapsed: 3.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.42,   30.09,   67.97,    0.37,    0.00\n",
            "Train - Start of epoch 384\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4381\n",
            "Epoch 384, Time elapsed: 3.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.24,   30.08,   68.80,    0.36,    0.00\n",
            "Train - Start of epoch 385\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.1827\n",
            "Epoch 385, Time elapsed: 3.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.12,   30.08,   67.68,    0.36,    0.00\n",
            "Train - Start of epoch 386\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.6774\n",
            "Epoch 386, Time elapsed: 3.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.54,   30.08,   68.10,    0.36,    0.00\n",
            "Train - Start of epoch 387\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.5745\n",
            "Epoch 387, Time elapsed: 3.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.57,   30.07,   67.15,    0.36,    0.00\n",
            "Train - Start of epoch 388\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.2184\n",
            "Epoch 388, Time elapsed: 3.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.46,   30.08,   68.03,    0.35,    0.00\n",
            "Train - Start of epoch 389\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5651\n",
            "Epoch 389, Time elapsed: 3.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.79,   30.08,   67.35,    0.36,    0.00\n",
            "Train - Start of epoch 390\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.9393\n",
            "Epoch 390, Time elapsed: 3.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.43,   30.08,   67.98,    0.37,    0.00\n",
            "Train - Start of epoch 391\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8266\n",
            "Epoch 391, Time elapsed: 3.72 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.50,   30.07,   68.06,    0.36,    0.00\n",
            "Train - Start of epoch 392\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.6980\n",
            "Epoch 392, Time elapsed: 3.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.96,   30.08,   67.54,    0.34,    0.00\n",
            "Train - Start of epoch 393\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2644\n",
            "Epoch 393, Time elapsed: 3.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.62,   30.07,   67.20,    0.35,    0.00\n",
            "Train - Start of epoch 394\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2857\n",
            "Epoch 394, Time elapsed: 3.75 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.13,   30.07,   68.69,    0.37,    0.00\n",
            "Train - Start of epoch 395\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.3599\n",
            "Epoch 395, Time elapsed: 3.76 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.93,   30.07,   68.48,    0.38,    0.00\n",
            "Train - Start of epoch 396\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4200\n",
            "Epoch 396, Time elapsed: 3.77 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.93,   30.07,   67.49,    0.36,    0.00\n",
            "Train - Start of epoch 397\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3532\n",
            "Epoch 397, Time elapsed: 3.78 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.60,   30.07,   68.16,    0.37,    0.00\n",
            "Train - Start of epoch 398\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8111\n",
            "Epoch 398, Time elapsed: 3.79 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.03,   30.08,   67.59,    0.36,    0.00\n",
            "Train - Start of epoch 399\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6713\n",
            "Epoch 399, Time elapsed: 3.8 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.64,   30.07,   67.22,    0.36,    0.00\n",
            "Train - Start of epoch 400\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1106\n",
            "Epoch 400, Time elapsed: 3.81 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.07,   68.20,    0.37,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=400>\n",
            "Saved checkpoint for epoch 400: example/result/1/checkpoint/ckpt-8\n",
            "Train - Start of epoch 401\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1843\n",
            "Epoch 401, Time elapsed: 3.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.80,   30.07,   67.38,    0.35,    0.00\n",
            "Train - Start of epoch 402\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.5026\n",
            "Epoch 402, Time elapsed: 3.83 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.53,   30.07,   67.10,    0.36,    0.00\n",
            "Train - Start of epoch 403\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7912\n",
            "Epoch 403, Time elapsed: 3.84 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.20,   30.07,   68.75,    0.37,    0.00\n",
            "Train - Start of epoch 404\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5432\n",
            "Epoch 404, Time elapsed: 3.85 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.70,   30.07,   68.26,    0.37,    0.00\n",
            "Train - Start of epoch 405\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 100.7715\n",
            "Epoch 405, Time elapsed: 3.86 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.26,   30.07,   67.83,    0.36,    0.00\n",
            "Train - Start of epoch 406\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.1262\n",
            "Epoch 406, Time elapsed: 3.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.42,   30.06,   68.01,    0.35,    0.00\n",
            "Train - Start of epoch 407\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.0942\n",
            "Epoch 407, Time elapsed: 3.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.24,   30.07,   67.80,    0.37,    0.00\n",
            "Train - Start of epoch 408\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.3449\n",
            "Epoch 408, Time elapsed: 3.88 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.04,   30.06,   67.61,    0.37,    0.00\n",
            "Train - Start of epoch 409\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.8740\n",
            "Epoch 409, Time elapsed: 3.89 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.74,   30.07,   68.30,    0.37,    0.00\n",
            "Train - Start of epoch 410\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.7758\n",
            "Epoch 410, Time elapsed: 3.9 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.18,   30.06,   67.76,    0.36,    0.00\n",
            "Train - Start of epoch 411\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8494\n",
            "Epoch 411, Time elapsed: 3.91 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.45,   30.06,   67.04,    0.35,    0.00\n",
            "Train - Start of epoch 412\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.2382\n",
            "Epoch 412, Time elapsed: 3.92 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.21,   30.07,   67.79,    0.35,    0.00\n",
            "Train - Start of epoch 413\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.2838\n",
            "Epoch 413, Time elapsed: 3.93 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   96.93,   30.06,   66.50,    0.36,    0.00\n",
            "Train - Start of epoch 414\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5441\n",
            "Epoch 414, Time elapsed: 3.94 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.39,   30.06,   67.96,    0.37,    0.00\n",
            "Train - Start of epoch 415\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9197\n",
            "Epoch 415, Time elapsed: 3.95 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.40,   30.07,   67.96,    0.37,    0.00\n",
            "Train - Start of epoch 416\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 98.5067\n",
            "Epoch 416, Time elapsed: 3.96 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.85,   30.05,   68.44,    0.35,    0.00\n",
            "Train - Start of epoch 417\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.3523\n",
            "Epoch 417, Time elapsed: 3.97 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.48,   30.06,   68.06,    0.37,    0.00\n",
            "Train - Start of epoch 418\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8662\n",
            "Epoch 418, Time elapsed: 3.98 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.01,   30.05,   67.59,    0.37,    0.00\n",
            "Train - Start of epoch 419\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 102.3527\n",
            "Epoch 419, Time elapsed: 3.99 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.33,   30.06,   68.89,    0.38,    0.00\n",
            "Train - Start of epoch 420\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3321\n",
            "Epoch 420, Time elapsed: 4.0 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.74,   30.06,   67.32,    0.36,    0.00\n",
            "Train - Start of epoch 421\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.7337\n",
            "Epoch 421, Time elapsed: 4.01 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.00,   30.06,   68.57,    0.37,    0.00\n",
            "Train - Start of epoch 422\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2048\n",
            "Epoch 422, Time elapsed: 4.01 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.86,   30.06,   68.43,    0.37,    0.00\n",
            "Train - Start of epoch 423\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1521\n",
            "Epoch 423, Time elapsed: 4.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.06,   68.21,    0.37,    0.00\n",
            "Train - Start of epoch 424\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.2252\n",
            "Epoch 424, Time elapsed: 4.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.26,   30.05,   67.85,    0.36,    0.00\n",
            "Train - Start of epoch 425\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.7814\n",
            "Epoch 425, Time elapsed: 4.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.17,   30.05,   66.76,    0.36,    0.00\n",
            "Train - Start of epoch 426\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6508\n",
            "Epoch 426, Time elapsed: 4.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.77,   30.05,   68.36,    0.36,    0.00\n",
            "Train - Start of epoch 427\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7592\n",
            "Epoch 427, Time elapsed: 4.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.47,   30.05,   68.06,    0.37,    0.00\n",
            "Train - Start of epoch 428\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5556\n",
            "Epoch 428, Time elapsed: 4.07 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.73,   30.06,   68.31,    0.36,    0.00\n",
            "Train - Start of epoch 429\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.9654\n",
            "Epoch 429, Time elapsed: 4.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.22,   30.05,   66.81,    0.35,    0.00\n",
            "Train - Start of epoch 430\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.6959\n",
            "Epoch 430, Time elapsed: 4.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.43,   30.05,   69.01,    0.37,    0.00\n",
            "Train - Start of epoch 431\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.7773\n",
            "Epoch 431, Time elapsed: 4.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.26,   30.05,   67.85,    0.36,    0.00\n",
            "Train - Start of epoch 432\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.7434\n",
            "Epoch 432, Time elapsed: 4.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.85,   30.05,   67.44,    0.36,    0.00\n",
            "Train - Start of epoch 433\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3451\n",
            "Epoch 433, Time elapsed: 4.12 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.69,   30.06,   68.27,    0.37,    0.00\n",
            "Train - Start of epoch 434\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.2879\n",
            "Epoch 434, Time elapsed: 4.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.10,   30.05,   67.70,    0.36,    0.00\n",
            "Train - Start of epoch 435\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 99.2798\n",
            "Epoch 435, Time elapsed: 4.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.96,   30.05,   68.52,    0.39,    0.00\n",
            "Train - Start of epoch 436\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.7206\n",
            "Epoch 436, Time elapsed: 4.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.59,   30.05,   68.17,    0.37,    0.00\n",
            "Train - Start of epoch 437\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.9936\n",
            "Epoch 437, Time elapsed: 4.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.69,   30.05,   67.28,    0.36,    0.00\n",
            "Train - Start of epoch 438\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6107\n",
            "Epoch 438, Time elapsed: 4.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.16,   30.05,   67.75,    0.35,    0.00\n",
            "Train - Start of epoch 439\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3881\n",
            "Epoch 439, Time elapsed: 4.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.31,   30.05,   67.90,    0.36,    0.00\n",
            "Train - Start of epoch 440\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.6819\n",
            "Epoch 440, Time elapsed: 4.18 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.62,   30.04,   67.20,    0.37,    0.00\n",
            "Train - Start of epoch 441\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 96.1603\n",
            "Epoch 441, Time elapsed: 4.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.04,   30.04,   67.63,    0.36,    0.00\n",
            "Train - Start of epoch 442\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.7551\n",
            "Epoch 442, Time elapsed: 4.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.30,   30.05,   67.88,    0.37,    0.00\n",
            "Train - Start of epoch 443\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8764\n",
            "Epoch 443, Time elapsed: 4.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.08,   30.05,   67.66,    0.38,    0.00\n",
            "Train - Start of epoch 444\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3800\n",
            "Epoch 444, Time elapsed: 4.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.66,   30.05,   68.23,    0.38,    0.00\n",
            "Train - Start of epoch 445\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 97.8141\n",
            "Epoch 445, Time elapsed: 4.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.75,   30.04,   68.34,    0.37,    0.00\n",
            "Train - Start of epoch 446\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.8101\n",
            "Epoch 446, Time elapsed: 4.24 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.24,   30.05,   66.82,    0.36,    0.00\n",
            "Train - Start of epoch 447\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 96.6986\n",
            "Epoch 447, Time elapsed: 4.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.51,   30.04,   68.09,    0.37,    0.00\n",
            "Train - Start of epoch 448\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.6331\n",
            "Epoch 448, Time elapsed: 4.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.41,   30.04,   68.00,    0.37,    0.00\n",
            "Train - Start of epoch 449\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.4182\n",
            "Epoch 449, Time elapsed: 4.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.52,   30.04,   68.12,    0.36,    0.00\n",
            "Train - Start of epoch 450\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.4710\n",
            "Epoch 450, Time elapsed: 4.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.47,   30.04,   68.06,    0.36,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=450>\n",
            "Saved checkpoint for epoch 450: example/result/1/checkpoint/ckpt-9\n",
            "Train - Start of epoch 451\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.2519\n",
            "Epoch 451, Time elapsed: 4.29 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.62,   30.04,   67.21,    0.37,    0.00\n",
            "Train - Start of epoch 452\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.0697\n",
            "Epoch 452, Time elapsed: 4.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.03,   30.04,   67.62,    0.37,    0.00\n",
            "Train - Start of epoch 453\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.6318\n",
            "Epoch 453, Time elapsed: 4.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.00,   30.04,   67.60,    0.36,    0.00\n",
            "Train - Start of epoch 454\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0519\n",
            "Epoch 454, Time elapsed: 4.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.15,   30.04,   67.75,    0.36,    0.00\n",
            "Train - Start of epoch 455\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 97.0466\n",
            "Epoch 455, Time elapsed: 4.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.86,   30.04,   67.47,    0.36,    0.00\n",
            "Train - Start of epoch 456\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8416\n",
            "Epoch 456, Time elapsed: 4.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.27,   30.04,   67.86,    0.37,    0.00\n",
            "Train - Start of epoch 457\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.9909\n",
            "Epoch 457, Time elapsed: 4.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.17,   30.04,   67.76,    0.36,    0.00\n",
            "Train - Start of epoch 458\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.4751\n",
            "Epoch 458, Time elapsed: 4.35 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.79,   30.04,   68.39,    0.36,    0.00\n",
            "Train - Start of epoch 459\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 100.7769\n",
            "Epoch 459, Time elapsed: 4.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.62,   30.04,   68.22,    0.36,    0.00\n",
            "Train - Start of epoch 460\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.2088\n",
            "Epoch 460, Time elapsed: 4.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.41,   30.04,   68.01,    0.36,    0.00\n",
            "Train - Start of epoch 461\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1048\n",
            "Epoch 461, Time elapsed: 4.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.68,   30.04,   68.27,    0.37,    0.00\n",
            "Train - Start of epoch 462\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2166\n",
            "Epoch 462, Time elapsed: 4.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.83,   30.04,   67.42,    0.37,    0.00\n",
            "Train - Start of epoch 463\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.0105\n",
            "Epoch 463, Time elapsed: 4.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.61,   30.04,   68.20,    0.37,    0.00\n",
            "Train - Start of epoch 464\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 99.9430\n",
            "Epoch 464, Time elapsed: 4.41 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.91,   30.04,   67.51,    0.36,    0.00\n",
            "Train - Start of epoch 465\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 101.1766\n",
            "Epoch 465, Time elapsed: 4.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.63,   30.04,   69.22,    0.37,    0.00\n",
            "Train - Start of epoch 466\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.2414\n",
            "Epoch 466, Time elapsed: 4.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.09,   30.04,   67.70,    0.35,    0.00\n",
            "Train - Start of epoch 467\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.3121\n",
            "Epoch 467, Time elapsed: 4.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.35,   30.03,   67.95,    0.37,    0.00\n",
            "Train - Start of epoch 468\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.3240\n",
            "Epoch 468, Time elapsed: 4.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.08,   30.03,   68.68,    0.37,    0.00\n",
            "Train - Start of epoch 469\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 94.3569\n",
            "Epoch 469, Time elapsed: 4.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.53,   30.04,   67.12,    0.38,    0.00\n",
            "Train - Start of epoch 470\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.2533\n",
            "Epoch 470, Time elapsed: 4.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.78,   30.04,   67.38,    0.37,    0.00\n",
            "Train - Start of epoch 471\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.5032\n",
            "Epoch 471, Time elapsed: 4.47 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.04,   30.03,   67.65,    0.36,    0.00\n",
            "Train - Start of epoch 472\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.2491\n",
            "Epoch 472, Time elapsed: 4.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.19,   30.03,   67.79,    0.36,    0.00\n",
            "Train - Start of epoch 473\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.9122\n",
            "Epoch 473, Time elapsed: 4.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.70,   30.04,   68.29,    0.37,    0.00\n",
            "Train - Start of epoch 474\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.1756\n",
            "Epoch 474, Time elapsed: 4.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.58,   30.04,   68.17,    0.37,    0.00\n",
            "Train - Start of epoch 475\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.5800\n",
            "Epoch 475, Time elapsed: 4.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.15,   30.03,   67.76,    0.36,    0.00\n",
            "Train - Start of epoch 476\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.7935\n",
            "Epoch 476, Time elapsed: 4.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.10,   30.03,   67.69,    0.37,    0.00\n",
            "Train - Start of epoch 477\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.7989\n",
            "Epoch 477, Time elapsed: 4.53 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.92,   30.04,   67.51,    0.37,    0.00\n",
            "Train - Start of epoch 478\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.8116\n",
            "Epoch 478, Time elapsed: 4.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.85,   30.03,   67.45,    0.36,    0.00\n",
            "Train - Start of epoch 479\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4129\n",
            "Epoch 479, Time elapsed: 4.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.79,   30.04,   67.40,    0.36,    0.00\n",
            "Train - Start of epoch 480\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.5917\n",
            "Epoch 480, Time elapsed: 4.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.29,   30.03,   66.91,    0.36,    0.00\n",
            "Train - Start of epoch 481\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.7395\n",
            "Epoch 481, Time elapsed: 4.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.94,   30.04,   67.55,    0.35,    0.00\n",
            "Train - Start of epoch 482\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 96.5407\n",
            "Epoch 482, Time elapsed: 4.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.34,   30.03,   67.95,    0.36,    0.00\n",
            "Train - Start of epoch 483\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.4088\n",
            "Epoch 483, Time elapsed: 4.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.29,   30.03,   67.89,    0.36,    0.00\n",
            "Train - Start of epoch 484\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.0523\n",
            "Epoch 484, Time elapsed: 4.59 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.42,   30.03,   67.04,    0.36,    0.00\n",
            "Train - Start of epoch 485\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.8467\n",
            "Epoch 485, Time elapsed: 4.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.17,   30.03,   67.78,    0.36,    0.00\n",
            "Train - Start of epoch 486\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8014\n",
            "Epoch 486, Time elapsed: 4.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.86,   30.03,   67.48,    0.36,    0.00\n",
            "Train - Start of epoch 487\n",
            "33/33 [==============================] - 1s 18ms/step - Reconstructed Loss: 100.0009\n",
            "Epoch 487, Time elapsed: 4.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.04,   30.03,   67.64,    0.36,    0.00\n",
            "Train - Start of epoch 488\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.3367\n",
            "Epoch 488, Time elapsed: 4.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.85,   30.03,   67.46,    0.36,    0.00\n",
            "Train - Start of epoch 489\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.1317\n",
            "Epoch 489, Time elapsed: 4.64 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.16,   30.03,   67.76,    0.37,    0.00\n",
            "Train - Start of epoch 490\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 97.1536\n",
            "Epoch 490, Time elapsed: 4.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.73,   30.03,   68.34,    0.36,    0.00\n",
            "Train - Start of epoch 491\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 99.6983\n",
            "Epoch 491, Time elapsed: 4.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.10,   30.03,   68.69,    0.38,    0.00\n",
            "Train - Start of epoch 492\n",
            "33/33 [==============================] - 1s 16ms/step - Reconstructed Loss: 98.4834\n",
            "Epoch 492, Time elapsed: 4.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.02,   30.03,   67.61,    0.37,    0.00\n",
            "Train - Start of epoch 493\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0966\n",
            "Epoch 493, Time elapsed: 4.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.60,   30.03,   67.20,    0.36,    0.00\n",
            "Train - Start of epoch 494\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.6433\n",
            "Epoch 494, Time elapsed: 4.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.35,   30.03,   67.97,    0.36,    0.00\n",
            "Train - Start of epoch 495\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.0467\n",
            "Epoch 495, Time elapsed: 4.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.38,   30.03,   68.00,    0.35,    0.00\n",
            "Train - Start of epoch 496\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.8130\n",
            "Epoch 496, Time elapsed: 4.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.10,   30.03,   67.71,    0.36,    0.00\n",
            "Train - Start of epoch 497\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 100.0758\n",
            "Epoch 497, Time elapsed: 4.72 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.01,   30.03,   68.62,    0.36,    0.00\n",
            "Train - Start of epoch 498\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 96.4648\n",
            "Epoch 498, Time elapsed: 4.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.82,   30.03,   68.43,    0.36,    0.00\n",
            "Train - Start of epoch 499\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 95.8670\n",
            "Epoch 499, Time elapsed: 4.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.03,   68.18,    0.36,    0.00\n",
            "Train - Start of epoch 500\n",
            "33/33 [==============================] - 1s 17ms/step - Reconstructed Loss: 98.8424\n",
            "Epoch 500, Time elapsed: 4.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.64,   30.03,   67.25,    0.36,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=500>\n",
            "Saved checkpoint for epoch 500: example/result/1/checkpoint/ckpt-10\n",
            "Train Done.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "train :   98.08,   30.23,   67.53,    0.32,    0.00\n",
            "Train - Start of epoch 219\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.3915\n",
            "Epoch 219, Time elapsed: 2.69 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.69,   30.23,   68.13,    0.33,    0.00\n",
            "Train - Start of epoch 220\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.2566\n",
            "Epoch 220, Time elapsed: 2.7 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.88,   30.22,   67.33,    0.33,    0.00\n",
            "Train - Start of epoch 221\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.9395\n",
            "Epoch 221, Time elapsed: 2.71 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.22,   68.08,    0.34,    0.00\n",
            "Train - Start of epoch 222\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.1320\n",
            "Epoch 222, Time elapsed: 2.73 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.55,   30.22,   68.00,    0.33,    0.00\n",
            "Train - Start of epoch 223\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.9590\n",
            "Epoch 223, Time elapsed: 2.74 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.63,   30.21,   67.08,    0.34,    0.00\n",
            "Train - Start of epoch 224\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.7926\n",
            "Epoch 224, Time elapsed: 2.75 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.03,   30.22,   67.47,    0.34,    0.00\n",
            "Train - Start of epoch 225\n",
            "33/33 [==============================] - 1s 24ms/step - Reconstructed Loss: 100.7138\n",
            "Epoch 225, Time elapsed: 2.76 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.65,   30.21,   68.10,    0.34,    0.00\n",
            "Train - Start of epoch 226\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 98.3948\n",
            "Epoch 226, Time elapsed: 2.77 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.58,   30.21,   69.03,    0.34,    0.00\n",
            "Train - Start of epoch 227\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.2843\n",
            "Epoch 227, Time elapsed: 2.79 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.90,   30.20,   68.36,    0.34,    0.00\n",
            "Train - Start of epoch 228\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 99.3481\n",
            "Epoch 228, Time elapsed: 2.8 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.79,   30.19,   68.25,    0.35,    0.00\n",
            "Train - Start of epoch 229\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.4978\n",
            "Epoch 229, Time elapsed: 2.81 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.19,   30.19,   67.66,    0.34,    0.00\n",
            "Train - Start of epoch 230\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.4057\n",
            "Epoch 230, Time elapsed: 2.82 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.72,   30.19,   68.19,    0.34,    0.00\n",
            "Train - Start of epoch 231\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.9794\n",
            "Epoch 231, Time elapsed: 2.84 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.07,   30.19,   67.54,    0.34,    0.00\n",
            "Train - Start of epoch 232\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 99.1199\n",
            "Epoch 232, Time elapsed: 2.86 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.53,   30.18,   67.00,    0.35,    0.00\n",
            "Train - Start of epoch 233\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 98.8604\n",
            "Epoch 233, Time elapsed: 2.87 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.17,   68.11,    0.35,    0.00\n",
            "Train - Start of epoch 234\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.8515\n",
            "Epoch 234, Time elapsed: 2.88 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.04,   30.18,   68.49,    0.36,    0.00\n",
            "Train - Start of epoch 235\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 101.1100\n",
            "Epoch 235, Time elapsed: 2.89 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.91,   30.18,   69.38,    0.35,    0.00\n",
            "Train - Start of epoch 236\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.8090\n",
            "Epoch 236, Time elapsed: 2.91 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.27,   30.17,   67.75,    0.35,    0.00\n",
            "Train - Start of epoch 237\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 101.1705\n",
            "Epoch 237, Time elapsed: 2.92 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.88,   30.17,   68.35,    0.36,    0.00\n",
            "Train - Start of epoch 238\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 98.4892\n",
            "Epoch 238, Time elapsed: 2.93 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.10,   30.16,   68.58,    0.36,    0.00\n",
            "Train - Start of epoch 239\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 99.2315\n",
            "Epoch 239, Time elapsed: 2.94 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.60,   30.16,   69.07,    0.36,    0.00\n",
            "Train - Start of epoch 240\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.4523\n",
            "Epoch 240, Time elapsed: 2.95 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.45,   30.16,   67.93,    0.35,    0.00\n",
            "Train - Start of epoch 241\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.5423\n",
            "Epoch 241, Time elapsed: 2.97 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.10,   30.16,   68.57,    0.36,    0.00\n",
            "Train - Start of epoch 242\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.8598\n",
            "Epoch 242, Time elapsed: 2.98 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.52,   30.16,   68.00,    0.37,    0.00\n",
            "Train - Start of epoch 243\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.4771\n",
            "Epoch 243, Time elapsed: 2.99 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.57,   30.16,   68.06,    0.35,    0.00\n",
            "Train - Start of epoch 244\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 96.1163\n",
            "Epoch 244, Time elapsed: 3.0 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.42,   30.15,   68.91,    0.36,    0.00\n",
            "Train - Start of epoch 245\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.8857\n",
            "Epoch 245, Time elapsed: 3.02 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.20,   30.15,   67.71,    0.35,    0.00\n",
            "Train - Start of epoch 246\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 97.4152\n",
            "Epoch 246, Time elapsed: 3.03 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.55,   30.15,   67.04,    0.36,    0.00\n",
            "Train - Start of epoch 247\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.0712\n",
            "Epoch 247, Time elapsed: 3.04 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.34,   30.14,   68.84,    0.36,    0.00\n",
            "Train - Start of epoch 248\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 99.6517\n",
            "Epoch 248, Time elapsed: 3.05 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.75,   30.14,   68.24,    0.37,    0.00\n",
            "Train - Start of epoch 249\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.6091\n",
            "Epoch 249, Time elapsed: 3.06 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.24,   30.15,   68.73,    0.36,    0.00\n",
            "Train - Start of epoch 250\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 98.3359\n",
            "Epoch 250, Time elapsed: 3.08 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.64,   30.13,   68.13,    0.38,    0.00\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=250>\n",
            "Saved checkpoint for epoch 250: example/result/1/checkpoint/ckpt-5\n",
            "Train - Start of epoch 251\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 98.1671\n",
            "Epoch 251, Time elapsed: 3.09 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.19,   30.14,   67.69,    0.35,    0.00\n",
            "Train - Start of epoch 252\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 98.9395\n",
            "Epoch 252, Time elapsed: 3.1 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.02,   30.13,   68.52,    0.37,    0.00\n",
            "Train - Start of epoch 253\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 101.8144\n",
            "Epoch 253, Time elapsed: 3.11 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.87,   30.14,   68.38,    0.36,    0.00\n",
            "Train - Start of epoch 254\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.8955\n",
            "Epoch 254, Time elapsed: 3.13 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.75,   30.13,   68.26,    0.36,    0.00\n",
            "Train - Start of epoch 255\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.3747\n",
            "Epoch 255, Time elapsed: 3.14 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.33,   30.13,   67.83,    0.37,    0.00\n",
            "Train - Start of epoch 256\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.1588\n",
            "Epoch 256, Time elapsed: 3.15 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.49,   30.13,   68.00,    0.36,    0.00\n",
            "Train - Start of epoch 257\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.3990\n",
            "Epoch 257, Time elapsed: 3.16 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.12,   30.13,   67.62,    0.37,    0.00\n",
            "Train - Start of epoch 258\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 94.8721\n",
            "Epoch 258, Time elapsed: 3.17 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.40,   30.12,   67.91,    0.36,    0.00\n",
            "Train - Start of epoch 259\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 97.7527\n",
            "Epoch 259, Time elapsed: 3.19 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.29,   30.13,   67.80,    0.36,    0.00\n",
            "Train - Start of epoch 260\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.5462\n",
            "Epoch 260, Time elapsed: 3.2 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.39,   30.12,   67.90,    0.36,    0.00\n",
            "Train - Start of epoch 261\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 101.7936\n",
            "Epoch 261, Time elapsed: 3.21 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.87,   30.12,   68.40,    0.36,    0.00\n",
            "Train - Start of epoch 262\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 99.3164\n",
            "Epoch 262, Time elapsed: 3.22 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.10,   30.12,   68.61,    0.37,    0.00\n",
            "Train - Start of epoch 263\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.8952\n",
            "Epoch 263, Time elapsed: 3.23 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.22,   30.12,   67.73,    0.36,    0.00\n",
            "Train - Start of epoch 264\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.3369\n",
            "Epoch 264, Time elapsed: 3.25 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.74,   30.11,   69.26,    0.37,    0.00\n",
            "Train - Start of epoch 265\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.1140\n",
            "Epoch 265, Time elapsed: 3.26 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.42,   30.11,   67.96,    0.35,    0.00\n",
            "Train - Start of epoch 266\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.1058\n",
            "Epoch 266, Time elapsed: 3.27 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.87,   30.11,   67.39,    0.37,    0.00\n",
            "Train - Start of epoch 267\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 99.4951\n",
            "Epoch 267, Time elapsed: 3.28 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.90,   30.11,   68.42,    0.37,    0.00\n",
            "Train - Start of epoch 268\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 101.2513\n",
            "Epoch 268, Time elapsed: 3.3 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.39,   30.11,   67.92,    0.36,    0.00\n",
            "Train - Start of epoch 269\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 101.6505\n",
            "Epoch 269, Time elapsed: 3.31 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.59,   30.11,   68.11,    0.37,    0.00\n",
            "Train - Start of epoch 270\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.6282\n",
            "Epoch 270, Time elapsed: 3.32 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.33,   30.10,   67.87,    0.36,    0.00\n",
            "Train - Start of epoch 271\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.5223\n",
            "Epoch 271, Time elapsed: 3.33 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.73,   30.10,   67.26,    0.36,    0.00\n",
            "Train - Start of epoch 272\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.8434\n",
            "Epoch 272, Time elapsed: 3.34 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.74,   30.10,   67.28,    0.36,    0.00\n",
            "Train - Start of epoch 273\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.6557\n",
            "Epoch 273, Time elapsed: 3.36 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.21,   30.10,   67.75,    0.36,    0.00\n",
            "Train - Start of epoch 274\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.8116\n",
            "Epoch 274, Time elapsed: 3.37 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.67,   30.10,   68.21,    0.36,    0.00\n",
            "Train - Start of epoch 275\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 99.8781\n",
            "Epoch 275, Time elapsed: 3.38 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.40,   30.10,   68.94,    0.36,    0.00\n",
            "Train - Start of epoch 276\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 99.8836\n",
            "Epoch 276, Time elapsed: 3.39 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.74,   30.10,   68.27,    0.37,    0.00\n",
            "Train - Start of epoch 277\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.8108\n",
            "Epoch 277, Time elapsed: 3.4 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.91,   30.10,   67.45,    0.37,    0.00\n",
            "Train - Start of epoch 278\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 96.4779\n",
            "Epoch 278, Time elapsed: 3.42 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.16,   30.09,   67.71,    0.36,    0.00\n",
            "Train - Start of epoch 279\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.2151\n",
            "Epoch 279, Time elapsed: 3.43 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.35,   30.09,   67.90,    0.36,    0.00\n",
            "Train - Start of epoch 280\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.5321\n",
            "Epoch 280, Time elapsed: 3.44 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.99,   30.09,   68.54,    0.37,    0.00\n",
            "Train - Start of epoch 281\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 99.2965\n",
            "Epoch 281, Time elapsed: 3.45 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.87,   30.09,   67.43,    0.35,    0.00\n",
            "Train - Start of epoch 282\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 103.3090\n",
            "Epoch 282, Time elapsed: 3.46 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "train :   98.61,   30.09,   68.15,    0.37,    0.00\n",
            "Train - Start of epoch 283\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 100.6914\n",
            "Epoch 283, Time elapsed: 3.48 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.70,   30.08,   68.25,    0.37,    0.00\n",
            "Train - Start of epoch 284\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 95.9664\n",
            "Epoch 284, Time elapsed: 3.49 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.23,   30.08,   67.79,    0.36,    0.00\n",
            "Train - Start of epoch 285\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.8914\n",
            "Epoch 285, Time elapsed: 3.5 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.56,   30.08,   68.13,    0.36,    0.00\n",
            "Train - Start of epoch 286\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 96.1879\n",
            "Epoch 286, Time elapsed: 3.51 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.02,   30.07,   67.60,    0.35,    0.00\n",
            "Train - Start of epoch 287\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 98.6323\n",
            "Epoch 287, Time elapsed: 3.52 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   99.36,   30.08,   68.91,    0.37,    0.00\n",
            "Train - Start of epoch 288\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 99.6579\n",
            "Epoch 288, Time elapsed: 3.54 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.98,   30.08,   67.54,    0.36,    0.00\n",
            "Train - Start of epoch 289\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.6438\n",
            "Epoch 289, Time elapsed: 3.55 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.16,   30.08,   66.73,    0.36,    0.00\n",
            "Train - Start of epoch 290\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 100.5297\n",
            "Epoch 290, Time elapsed: 3.56 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.22,   30.08,   67.79,    0.36,    0.00\n",
            "Train - Start of epoch 291\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 101.0737\n",
            "Epoch 291, Time elapsed: 3.57 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.25,   30.07,   67.81,    0.36,    0.00\n",
            "Train - Start of epoch 292\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 100.6270\n",
            "Epoch 292, Time elapsed: 3.58 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.05,   30.07,   67.62,    0.36,    0.00\n",
            "Train - Start of epoch 293\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 97.7465\n",
            "Epoch 293, Time elapsed: 3.6 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.36,   30.07,   66.94,    0.35,    0.00\n",
            "Train - Start of epoch 294\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 101.3460\n",
            "Epoch 294, Time elapsed: 3.61 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.10,   30.07,   67.66,    0.37,    0.00\n",
            "Train - Start of epoch 295\n",
            "33/33 [==============================] - 1s 20ms/step - Reconstructed Loss: 98.0927\n",
            "Epoch 295, Time elapsed: 3.62 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   97.73,   30.07,   67.29,    0.37,    0.00\n",
            "Train - Start of epoch 296\n",
            "33/33 [==============================] - 1s 23ms/step - Reconstructed Loss: 99.3127\n",
            "Epoch 296, Time elapsed: 3.63 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.48,   30.07,   68.04,    0.36,    0.00\n",
            "Train - Start of epoch 297\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.0296\n",
            "Epoch 297, Time elapsed: 3.65 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.93,   30.07,   68.51,    0.36,    0.00\n",
            "Train - Start of epoch 298\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 99.3677\n",
            "Epoch 298, Time elapsed: 3.66 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.85,   30.07,   68.41,    0.37,    0.00\n",
            "Train - Start of epoch 299\n",
            "33/33 [==============================] - 1s 21ms/step - Reconstructed Loss: 98.2522\n",
            "Epoch 299, Time elapsed: 3.67 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.39,   30.07,   67.96,    0.36,    0.00\n",
            "Train - Start of epoch 300\n",
            "33/33 [==============================] - 1s 22ms/step - Reconstructed Loss: 98.9130\n",
            "Epoch 300, Time elapsed: 3.68 minutes\n",
            "          total, obs_M-0, unobs_M-0,      kl,     mmd\n",
            "train :   98.63,   30.06,   68.21,    0.35,    0.00\n",
            "<tf.Variable 'Variable:0' shape=() dtype=int32, numpy=300>\n",
            "Saved checkpoint for epoch 300: example/result/1/checkpoint/ckpt-6\n",
            "Train Done.\n"
          ]
        }
      ],
      "source": [
        "# Here we only train for 500 epochs for illustration purpose\n",
        "hist = model.train(\n",
        "        num_epoch=500, batch_size=512, save_every_epoch=50,\n",
        "        verbose=True, checkpoint_dir=path_root+'checkpoint/')"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "9c16c9d7-6d1f-4818-af96-a64f21383cf8",
      "metadata": {},
      "source": [
        "We can visualize the training loss. It flucuates since we argument the data with random masking, which also provides regularization against overfitting.\n",
        "One may continue training until the loss gets stablized."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "id": "bc76140a-c000-4593-a8a3-57145cd8e09c",
      "metadata": {},
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'Trianing loss')"
            ]
          },
          "execution_count": 5,
          "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": [
        "plt.plot(hist['train']['total'])\n",
        "plt.xlabel('Epoch')\n",
        "plt.ylabel('Trianing loss')"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "0569a846-54d7-4537-9929-03665b7de039",
      "metadata": {},
      "source": [
        "We can load the trained model:\n",
        "\n",
        "```python\n",
        "# load the model and ensure it is loaded successfully\n",
        "checkpoint = tf.train.Checkpoint(net=model.vae)\n",
        "n_ckpt = 10\n",
        "status = checkpoint.restore(path_root+'checkpoint/ckpt-{}'.format(n_ckpt))\n",
        "model.vae(tf.zeros((1,np.sum(model.vae.config.dim_input_arr))),\n",
        "          tf.zeros((1,np.sum(model.vae.config.dim_input_arr))),\n",
        "          tf.zeros((1,np.sum(model.batches.shape[1]))), L=1, training=False)\n",
        "print(status)\n",
        "```"
      ]
    },
    {
      "cell_type": "markdown",
      "id": "f7743e04-4cf3-44f1-b19f-0787e0f261f4",
      "metadata": {},
      "source": [
        "# Inspect imputation results"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "id": "db681b86-f796-4c35-91f6-96f4145cfa8a",
      "metadata": {
        "scrolled": true
      },
      "outputs": [],
      "source": [
        "denoised_data = model.get_denoised_data()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "id": "5fa12909",
      "metadata": {},
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 1000x400 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "fig, axes = plt.subplots(1,2,figsize=(10,4))\n",
        "axes[0].scatter(\n",
        "    data_raw[masks==0],\n",
        "    denoised_data[masks==0], s=0.1, alpha=0.1\n",
        "    )\n",
        "axes[0].set_xlabel('Observed')\n",
        "axes[0].set_ylabel('Denoised')\n",
        "\n",
        "axes[1].scatter(\n",
        "    data_raw[(masks_raw==0)&(masks==-1)],\n",
        "    denoised_data[(masks_raw==0)&(masks==-1)], s=0.1, alpha=0.1\n",
        "    )\n",
        "axes[1].set_xlabel('Observed')\n",
        "axes[1].set_ylabel('Imputed')\n",
        "plt.show()"
      ]
    }
  ],
  "metadata": {
    "kernelspec": {
      "display_name": "tf",
      "language": "python",
      "name": "tf"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.11.5"
    },
    "toc": {
      "base_numbering": 1,
      "nav_menu": {},
      "number_sections": true,
      "sideBar": true,
      "skip_h1_title": false,
      "title_cell": "Table of Contents",
      "title_sidebar": "Contents",
      "toc_cell": false,
      "toc_position": {},
      "toc_section_display": true,
      "toc_window_display": false
    },
    "varInspector": {
      "cols": {
        "lenName": 16,
        "lenType": 16,
        "lenVar": 40
      },
      "kernels_config": {
        "python": {
          "delete_cmd_postfix": "",
          "delete_cmd_prefix": "del ",
          "library": "var_list.py",
          "varRefreshCmd": "print(var_dic_list())"
        },
        "r": {
          "delete_cmd_postfix": ") ",
          "delete_cmd_prefix": "rm(",
          "library": "var_list.r",
          "varRefreshCmd": "cat(var_dic_list()) "
        }
      },
      "types_to_exclude": [
        "module",
        "function",
        "builtin_function_or_method",
        "instance",
        "_Feature"
      ],
      "window_display": false
    }
  },
  "nbformat": 4,
  "nbformat_minor": 5
}