model_based_optimization_witth_vae.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Model-based Optimization using Variational Auto-Encoders\n",
    "\n",
    "This notebook contains the source code for the Medium article _Optimize What You Can Predict: Model-Based Optimization Using Variational Auto-Encoders_. We first build and train a multi-head Variational Autoencoder. Then we implement methods to navigate in the latent space such that the output of certain predictive models is optimized."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import math\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import Dataset, DataLoader\n",
    "from torch.optim import Adam\n",
    "from torch.optim.lr_scheduler import StepLR\n",
    "from torchvision import datasets, transforms\n",
    "from sklearn.metrics import accuracy_score, mean_squared_error, mean_absolute_error\n",
    "from skimage import morphology, transform\n",
    "from tqdm.notebook import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Hyper-parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# helper function\n",
    "def dim_after_convs():\n",
    "    \"\"\"\n",
    "    Compute the (flattened) dimensionality of the tensor after the convolutional part of the encoder,\n",
    "    as well as the single-axis dimensionality.\n",
    "    \"\"\"\n",
    "    dim = INP_HEIGHT\n",
    "    for _ in ENC_CONV_LAYER_CHANNELS[1:]:\n",
    "        dim = math.floor(1 + (dim - KERNEL_SIZE) / STRIDE)\n",
    "    flattened_dim = (dim ** 2) * ENC_CONV_LAYER_CHANNELS[-1]\n",
    "    return flattened_dim, dim"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# folders\n",
    "DATA_DIR = \"./data\"\n",
    "\n",
    "# data\n",
    "NUM_CLASSES = 10\n",
    "NUM_CHANNELS = 1\n",
    "INP_HEIGHT = 28\n",
    "INP_WIDTH = 28\n",
    "INP_DIM = INP_WIDTH * INP_HEIGHT\n",
    "\n",
    "# training\n",
    "BATCH_SIZE = 512\n",
    "INIT_LR = 1e-3\n",
    "LR_DECAY_GAMMA = 0.5\n",
    "LR_DECAY_STEPS = 5\n",
    "KLD_BETA = 1e-4\n",
    "NUM_EPOCHS = 60\n",
    "OPTIMIZER = Adam\n",
    "\n",
    "# architecture\n",
    "EMBEDDING_DIM = 16\n",
    "KERNEL_SIZE = 4\n",
    "STRIDE = 2\n",
    "ENC_CONV_LAYER_CHANNELS = [NUM_CHANNELS, 64, 128]\n",
    "FL_DIM_AFTER_CONV, SINGLE_DIM_AFTER_CONV = dim_after_convs()\n",
    "ENC_DENSE_LAYER_SIZES = [FL_DIM_AFTER_CONV, 200, EMBEDDING_DIM]\n",
    "DEC_DENSE_LAYER_SIZES = [EMBEDDING_DIM, 200, FL_DIM_AFTER_CONV]\n",
    "DEC_CONV_LAYER_CHANNELS = [ENC_CONV_LAYER_CHANNELS[-1], 64, NUM_CHANNELS]\n",
    "DIGIT_LAYER_SIZES = [EMBEDDING_DIM, 64, NUM_CLASSES]\n",
    "THICKNESS_LAYER_SIZES = [EMBEDDING_DIM, 64, 1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DigitThicknessSet(Dataset):\n",
    "    \n",
    "    def __init__(self, train: bool):\n",
    "        self.data_set = datasets.MNIST(DATA_DIR, train=train, download=True)\n",
    "        self.to_tensor = transforms.ToTensor()\n",
    "        self.normalize = transforms.Normalize((0.1307,), (0.3081,))\n",
    "        \n",
    "        # pre-compute thickness for each digit\n",
    "        self.thickness_map = {}\n",
    "        for i, (img, _) in enumerate(tqdm(self.data_set)):\n",
    "            self.thickness_map[i] = self._mean_thickness(img)\n",
    "        \n",
    "    def __getitem__(self, index):\n",
    "        img, digit = self.data_set[index]\n",
    "        thickness = self.thickness_map[index]\n",
    "        img = self.normalize(self.to_tensor(img))\n",
    "        return img, digit, thickness\n",
    "    \n",
    "    def __len__(self):\n",
    "        return len(self.data_set)\n",
    "        \n",
    "    def _mean_thickness(self, img, upscale=3, thresh=0.5):\n",
    "        img = np.array(img)\n",
    "        img = transform.pyramid_expand(img, upscale=upscale, order=3, multichannel=False)\n",
    "        img_min, img_max = img.min(), img.max()\n",
    "        bin_img = (img >= img_min + (img_max - img_min) * thresh)\n",
    "        skeleton, dist_map = morphology.medial_axis(bin_img, return_distance=True)\n",
    "        return 2. * np.mean(dist_map[skeleton]) / upscale"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_data():\n",
    "    train_set = DataLoader(DigitThicknessSet(train=True), batch_size=BATCH_SIZE, shuffle=True)\n",
    "    test_set = DataLoader(DigitThicknessSet(train=False), batch_size=BATCH_SIZE)\n",
    "    return train_set, test_set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model definition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "class VariationalEncoder(nn.Module):\n",
    "    \n",
    "    def __init__(self):\n",
    "        super(VariationalEncoder, self).__init__()\n",
    "        self.create_network()\n",
    "\n",
    "    def create_network(self):\n",
    "        \n",
    "        # convolutional part\n",
    "        for i in range(1, len(ENC_CONV_LAYER_CHANNELS)):\n",
    "            self.add_module(f\"enc_conv_{i}\", nn.Conv2d(in_channels=ENC_CONV_LAYER_CHANNELS[i - 1],\n",
    "                                                       out_channels=ENC_CONV_LAYER_CHANNELS[i],\n",
    "                                                       kernel_size=KERNEL_SIZE, stride=STRIDE))\n",
    "            self.add_module(f\"enc_conv_activation_{i}\", nn.ReLU())\n",
    "            \n",
    "        # fully-connected part\n",
    "        for i in range(1, len(ENC_DENSE_LAYER_SIZES)):\n",
    "            if i == len(ENC_DENSE_LAYER_SIZES) - 1:\n",
    "                self.add_module('mu', nn.Linear(ENC_DENSE_LAYER_SIZES[i - 1],\n",
    "                                                ENC_DENSE_LAYER_SIZES[i], bias=True))\n",
    "                self.add_module('logvar', nn.Linear(ENC_DENSE_LAYER_SIZES[i - 1],\n",
    "                                                    ENC_DENSE_LAYER_SIZES[i], bias=True))\n",
    "            else:\n",
    "                self.add_module(f\"enc_linear_{i}\", nn.Linear(ENC_DENSE_LAYER_SIZES[i - 1],\n",
    "                                                             ENC_DENSE_LAYER_SIZES[i], bias=True))\n",
    "                self.add_module(f\"enc_linear_activation_{i}\", nn.ReLU())\n",
    "\n",
    "    def forward(self, inp):\n",
    "        tmp = inp\n",
    "        for mod_name, module in self._modules.items():\n",
    "            if mod_name == 'mu':\n",
    "                mu = module(tmp)\n",
    "            elif mod_name == 'logvar':\n",
    "                logvar = module(tmp)\n",
    "            elif mod_name == 'enc_linear_1':\n",
    "                tmp = module(torch.flatten(tmp, start_dim=1))\n",
    "            else:\n",
    "                tmp = module(tmp)\n",
    "        return mu, logvar"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Decoder(nn.Module):\n",
    "\n",
    "    def __init__(self):\n",
    "        super(Decoder, self).__init__()\n",
    "        self.create_network()\n",
    "\n",
    "    def create_network(self):\n",
    "        \n",
    "        # fully-connected part\n",
    "        for i in range(1, len(DEC_DENSE_LAYER_SIZES)):\n",
    "            self.add_module(f\"dec_linear_{i}\", nn.Linear(DEC_DENSE_LAYER_SIZES[i - 1],\n",
    "                                                         DEC_DENSE_LAYER_SIZES[i], bias=True))\n",
    "            self.add_module(f\"dec_linear_activitation_{i}\", nn.ReLU())\n",
    "            \n",
    "        # convolutional part\n",
    "        for i in range(1, len(DEC_CONV_LAYER_CHANNELS)):\n",
    "            if i == len(DEC_CONV_LAYER_CHANNELS) - 1:\n",
    "                self.add_module(f\"dec_conv_{i}\", nn.ConvTranspose2d(DEC_CONV_LAYER_CHANNELS[i - 1],\n",
    "                                                                    DEC_CONV_LAYER_CHANNELS[i],\n",
    "                                                                    KERNEL_SIZE, STRIDE))\n",
    "            else:\n",
    "                self.add_module(f\"dec_conv_{i}\", nn.ConvTranspose2d(DEC_CONV_LAYER_CHANNELS[i - 1],\n",
    "                                                                    DEC_CONV_LAYER_CHANNELS[i],\n",
    "                                                                    KERNEL_SIZE, STRIDE, output_padding=1))\n",
    "                self.add_module(f\"dec_conv_activitation_{i}\", nn.ReLU())\n",
    "            \n",
    "    def forward(self, inp):\n",
    "        tmp = inp\n",
    "        for mod_name, module in self._modules.items():\n",
    "            if mod_name == 'dec_conv_1':\n",
    "                tmp = tmp.view([-1, ENC_CONV_LAYER_CHANNELS[-1], SINGLE_DIM_AFTER_CONV, SINGLE_DIM_AFTER_CONV])\n",
    "            tmp = module(tmp)\n",
    "        return tmp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DigitClassifier(nn.Module):\n",
    "\n",
    "    def __init__(self):\n",
    "        super(DigitClassifier, self).__init__()\n",
    "        self.create_network()\n",
    "\n",
    "    def create_network(self):\n",
    "        for i in range(1, len(DIGIT_LAYER_SIZES)):\n",
    "            if i == len(DIGIT_LAYER_SIZES) - 1:\n",
    "                self.add_module(f\"clf_linear_{i}\", nn.Linear(DIGIT_LAYER_SIZES[i - 1],\n",
    "                                                             DIGIT_LAYER_SIZES[i], bias=True))\n",
    "                self.add_module(\"clf_out\", nn.LogSoftmax(dim=1))\n",
    "            else:\n",
    "                self.add_module(f\"clf_linear_{i}\", nn.Linear(DIGIT_LAYER_SIZES[i - 1],\n",
    "                                                             DIGIT_LAYER_SIZES[i], bias=True))\n",
    "                self.add_module(f\"clf_activitation_{i}\", nn.ReLU())\n",
    "\n",
    "    def forward(self, inp):\n",
    "        tmp = inp\n",
    "        for mod_name, module in self._modules.items():\n",
    "            tmp = module(tmp)\n",
    "        return tmp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "class ThicknessModel(nn.Module):\n",
    "\n",
    "    def __init__(self):\n",
    "        super(ThicknessModel, self).__init__()\n",
    "        self.create_network()\n",
    "\n",
    "    def create_network(self):\n",
    "        for i in range(1, len(THICKNESS_LAYER_SIZES)):\n",
    "            if i == len(THICKNESS_LAYER_SIZES) - 1:\n",
    "                self.add_module(f\"clf_t_linear_{i}\", nn.Linear(THICKNESS_LAYER_SIZES[i - 1],\n",
    "                                                               THICKNESS_LAYER_SIZES[i], bias=True))\n",
    "            else:\n",
    "                self.add_module(f\"clf_t_linear_{i}\", nn.Linear(THICKNESS_LAYER_SIZES[i - 1],\n",
    "                                                               THICKNESS_LAYER_SIZES[i], bias=True))\n",
    "                self.add_module(f\"clf_t_activitation_{i}\", nn.ReLU())\n",
    "\n",
    "    def forward(self, inp):\n",
    "        tmp = inp\n",
    "        for mod_name, module in self._modules.items():\n",
    "            tmp = module(tmp)\n",
    "        return tmp"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class BetaKLDLoss:\n",
    "    \n",
    "    def __init__(self, beta):\n",
    "        self.beta = beta    \n",
    "        \n",
    "    def __call__(self, mu, logvar):\n",
    "        kld = - 0.5 * torch.sum(1 + logvar - mu.pow(2) - logvar.exp())\n",
    "        return self.beta * kld"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def accuracy(ref, pred):\n",
    "    return accuracy_score(ref, torch.argmax(pred, 1))\n",
    "\n",
    "def l2_reconstruction_error(ref, pred):\n",
    "    return mean_squared_error(torch.flatten(ref), torch.flatten(pred))    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class VAE(nn.Module):\n",
    "    \n",
    "    def __init__(self):\n",
    "        super(VAE, self).__init__()\n",
    "        self.encoder = VariationalEncoder()\n",
    "        self.decoder = Decoder()\n",
    "        self.classifier = DigitClassifier()\n",
    "        self.thickness_model = ThicknessModel()\n",
    "        \n",
    "        self.clf_loss = nn.NLLLoss()\n",
    "        self.t_loss = nn.MSELoss()\n",
    "        self.dec_loss = nn.MSELoss()\n",
    "        self.kld_loss = BetaKLDLoss(beta=KLD_BETA)\n",
    "\n",
    "    def forward(self, inp):\n",
    "        mu, logvar = self.encoder(inp)\n",
    "        emb = self._reparameterize(mu, logvar)\n",
    "        inp_hat = self.decoder(emb)\n",
    "        t_hat = self.thickness_model(emb)\n",
    "        y_hat = self.classifier(emb)\n",
    "        return inp_hat, y_hat, t_hat, mu, logvar\n",
    "        \n",
    "    def _reparameterize(self, mu, logvar):\n",
    "        std = torch.exp(0.5 * logvar)\n",
    "        eps = torch.randn_like(std)\n",
    "        return mu + eps * std\n",
    "    \n",
    "    def train_(self, train_data_loader):\n",
    "        \n",
    "        trainable_params = list(self.encoder.parameters()) + list(self.decoder.parameters()) \\\n",
    "                            + list(self.classifier.parameters()) + list(self.thickness_model.parameters())\n",
    "        optimizer = OPTIMIZER(trainable_params, INIT_LR)\n",
    "        scheduler = StepLR(optimizer, step_size=LR_DECAY_STEPS, gamma=LR_DECAY_GAMMA)\n",
    "        \n",
    "        for epoch in range(1, NUM_EPOCHS + 1):\n",
    "            for inp_batch, y_batch, t_batch in tqdm(train_data_loader):\n",
    "                self.train()\n",
    "                optimizer.zero_grad()\n",
    "                inp_hat_batch, y_hat_batch, t_hat_batch, mu, logvar = self.forward(inp_batch)\n",
    "                total_loss = self.clf_loss(y_hat_batch, y_batch) + \\\n",
    "                             self.dec_loss(inp_hat_batch, inp_batch) + \\\n",
    "                             self.t_loss(torch.squeeze(t_hat_batch), t_batch) + \\\n",
    "                             self.kld_loss(mu, logvar)\n",
    "                total_loss.backward()\n",
    "                optimizer.step()\n",
    "            scheduler.step()\n",
    "\n",
    "    def eval_(self, test_data_loader):\n",
    "        with torch.no_grad():\n",
    "            inp_batch, y_batch, t_batch = next(iter(test_data_loader))            \n",
    "            inp_hat_batch, y_hat_batch, t_hat_batch,  _, _ = self.forward(inp_batch)\n",
    "            t_error = mean_absolute_error(t_batch, t_hat_batch)\n",
    "            acc = accuracy(y_batch, y_hat_batch)\n",
    "            rec_err = l2_reconstruction_error(inp_batch, inp_hat_batch)\n",
    "            print(f\"The mean classification accuracy is {acc}.\")\n",
    "            print(f\"The mean reconstruction error is {rec_err}.\")\n",
    "            print(f\"The mean absolute thickness error is {t_error}.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0fee4f1132ab496398ce501661007f77",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=60000.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dbfa8058c05542e686f320f08cf6de17",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=10000.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "train_set, test_set = load_data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d7fd427854b5473bb13b0cdbdd94fd96",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2b5f0c3031f438cb35471c1dc79d053",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e374e32292a547e5a949e0a5e6e14ffa",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "31055badd1b14c32b3a825c059bb7b61",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e79c61ece2ef40a0994584ada6a38ba6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1bb000104c004c6f8f2d926bdc91dff4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7b0e7e65fffe4fd08c57e45c8f578d1b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "563ccde7c1384e62ba16b8f20a61e6c8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "92e054f3c5fa447fb9a6970351ec11d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "acd6a76500b743b48a174d170246f4c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8dfeacc4d9f5488c9c57736923a3d582",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "277baabc869c46dc81bcbea4d1149d24",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b8b3ed39839d454f8841fd6ffef1f6f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7fd1657c2de44b64bae37958b2ef6f30",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "57c2d30ca0a347728e6ab0e0c0406c67",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f4e632d6bfd34f73ba91c72e49ecfc21",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5ac2dd7b5a0a434bbe6f1bd0b4b5d466",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "726b745851c14cd88d8bc864f7b0a1e8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "926b20f286f04fc59f907891fb718590",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "67ddaf1e6373420ea3cafc7bc19009a1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "001d11f074444f30b99b3ea248a8236e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "447d012cffa34f64b91f6720892431fe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d8f7087d44db4fba8a4dcbdee3223233",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f01d1ef3536f4a80a851232bb605664c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "027caa2ea8984934b0a9db6ec9bee53a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "aea69b3bda4944e9857ec23088981534",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "647d01d803844c0cb81fb400231d3296",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b8f8228c8bc34590a415c945aad4450b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "53821626f10c49b38fa82d03c50fd35b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7c8f5ccd7aaa4cacbb6784061e0a6dd9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "8681b5d0d22148fcb91f7a6f2aea57bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1cb4717951c84db1b9574558256ec88e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2b371c47d78f4e32a9909e2c591c0ed0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a2bee9494c74744b7b088d525f19c2e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2fed8b6c99b94ebeb5ea518acb60e74b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1b700e15565c4361abd3b27d7d72e129",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "924746d93cb54b33976ea08116712ee4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4433079af001467b9a25eab858fbfb4c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "95dd9a2059b84220a920d7625271cfb4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "42267a1c3a8b40eaab18d4e46d9d1009",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "63f0cb619f5c4e2b98d872c4af1efca2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "22c4244aa0aa4e048e7e9a91f8e9be35",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "354d2b741aaa461689233b5b21ac958f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "366fe9349f354f22b6a5a430711b210d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "463be90344e146ab9fc80aeb265c7452",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0a92315f6dc2445f83f45f5b54c15fe4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7315133546fd437eadc97742350ca39a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "999cd1579ab14b4480ba5bf2a484c700",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "dae7bdf235004cf5b8fd694ee345e5fd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "914eabf20fd84113af80e13f9272e2f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "002f853a223343119e8b14d9edc6d5e8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ef9db25b89a34ca2960dc288331769b2",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1f116d64bf4640d69ce6c97cf6db2ef5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "e2f0ca8460de4cbaa1b55d1d1c9a009e",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9911bc9819c940a3bd54330827cf1ae3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7934a6b179374a2b8c1672820286d916",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "61a06a008916444683c7a36a1e3397f9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fb0aefd7a4fa4f0ebb37263acfa9766b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c89c751175b54fb580acceabc6b1658a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "355ae5c5610f4a968227f69ddeff57d1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=118.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "The mean classification accuracy is 0.9765625.\n",
      "The mean reconstruction error is 0.34404683113098145.\n",
      "The mean absolute thickness error is 0.17375248670578003.\n"
     ]
    }
   ],
   "source": [
    "model = VAE()\n",
    "model.train_(train_set)\n",
    "model.eval_(test_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13ec35a10>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# plot some images\n",
    "num_samples = 8\n",
    "x = list(test_set)[0][0][:num_samples]\n",
    "with torch.no_grad():\n",
    "    x_hat = model(x)[0]\n",
    "x_row = torch.squeeze(torch.cat(torch.split(x, 1), 3))\n",
    "x_hat_row = torch.squeeze(torch.cat(torch.split(x_hat, 1), 3))\n",
    "img = torch.cat([x_row, x_hat_row], 0)\n",
    "plt.figure(figsize=(16,4))\n",
    "plt.imshow(img, cmap='Greys')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Optimization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "def latent_thickness_step(z, t_model, step_size):\n",
    "    z = z.clone().detach().requires_grad_(True)\n",
    "    t_model.zero_grad()\n",
    "    t_hat = t_model(z)\n",
    "    t_hat[0].backward(retain_graph=True)\n",
    "    return z + step_size * z.grad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def latent_thickness_optimization(z, model, step_size, n_steps=20):\n",
    "    x_hats = [model.decoder(z)]\n",
    "    for step in range(n_steps):\n",
    "        z = latent_thickness_step(z, model.thickness_model, step_size)\n",
    "        x_hats.append(model.decoder(z))\n",
    "    return x_hats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def latent_digit_step(z, clf, target, step_size):\n",
    "    z = z.clone().detach().requires_grad_(True)\n",
    "    clf.zero_grad()\n",
    "    y_hat = clf(z)\n",
    "    y_hat[0][target].backward(retain_graph=True)\n",
    "    return z + step_size * z.grad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "def latent_digit_optimization(z, model, target, step_size, n_steps=20):\n",
    "    x_hats = [model.decoder(z)]\n",
    "    for step in range(n_steps):\n",
    "        z = latent_digit_step(z, model.classifier, target, step_size)\n",
    "        x_hats.append(model.decoder(z))\n",
    "    return x_hats"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First, let's pick a random image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13ec1e190>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "z_start = torch.rand(1, EMBEDDING_DIM)\n",
    "plt.imshow(model.decoder(z_start).squeeze().detach().numpy(), cmap='Greys')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now, let's increase the thickness."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13e6a3710>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "x_hats = latent_thickness_optimization(z_start, model, 0.2)\n",
    "plt.figure(figsize=(16,4))\n",
    "plt.imshow(torch.squeeze(torch.cat(x_hats, 3)).detach().numpy(), cmap='Greys')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can also gradually transform it into another digit."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x13e8da990>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x2880 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "c_x_hats = []\n",
    "for target_digit in range(NUM_CLASSES):\n",
    "    x_hats = latent_digit_optimization(z_start, model, target_digit, 0.03)\n",
    "    c_x_hats.append(x_hats)\n",
    "imgs = []\n",
    "for i in range(NUM_CLASSES):\n",
    "    imgs.append(torch.squeeze(torch.cat(c_x_hats[i], 3)).detach().numpy())\n",
    "plt.figure(figsize=(16,40))\n",
    "plt.imshow(np.concatenate(imgs), cmap='Greys')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}