pytorch_pennylane_hybrid_binary_classifier.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pennylane as qml\n",
    "from pennylane import numpy as np\n",
    "\n",
    "import torch\n",
    "from torch.autograd import Function\n",
    "from torchvision import datasets, transforms\n",
    "import torch.optim as optim\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#make a 7qubit device\n",
    "dev = qml.device(\"default.qubit\", wires=4, shots=100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#state prepration template\n",
    "def data_prepration(d):\n",
    "\n",
    "    for i,j in enumerate(d):\n",
    "        r = j/510\n",
    "        qml.RY(r*np.pi, wires=i)\n",
    "    \n",
    "    qml.Barrier(wires=(len(d)-1,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#inverce MERA and fully entangler layer's template\n",
    "def inverce_MERA(p):\n",
    "\n",
    "    qml.Hadamard(wires=0)\n",
    "    qml.Hadamard(wires=1)\n",
    "    qml.Hadamard(wires=2)\n",
    "    qml.Hadamard(wires=3)\n",
    "\n",
    "    qml.CNOT(wires=(0,1))\n",
    "    qml.CNOT(wires=(1,2))\n",
    "    qml.CNOT(wires=(2,3))\n",
    "    qml.CNOT(wires=(3,0))\n",
    "\n",
    "    qml.Barrier(wires=(0,3))\n",
    "    \n",
    "    qml.CRZ(p[0],wires=(1,0))\n",
    "    qml.CRX(p[1],wires=(1,0))\n",
    "\n",
    "    qml.CRZ(p[2],wires=(3,2))\n",
    "    qml.CRX(p[3],wires=(3,2))\n",
    "\n",
    "    qml.CRZ(p[4],wires=(2,0))\n",
    "    qml.CRX(p[5],wires=(2,0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#configure circuit in a qnode using dev\n",
    "@qml.qnode(dev)\n",
    "def inverce_MERA_layer(inputs, mera_parameters):\n",
    "\n",
    "    data_prepration(inputs)\n",
    "\n",
    "    inverce_MERA(mera_parameters)\n",
    "\n",
    "    return qml.expval(qml.PauliZ(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "#define hyper_parameters and weights-init\n",
    "np.random.seed(0)\n",
    "random_trash_data = np.random.rand(4, requires_grad=True)\n",
    "weights_init = 0.01 * np.random.randn(6, requires_grad=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#needs weights_inits hyperparameter\n",
    "qml.drawer.use_style(style=\"solarized_dark\")\n",
    "fig, ax = qml.draw_mpl(inverce_MERA_layer)(random_trash_data ,weights_init)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# create pytorch layer from Qnode "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "weight_shapes = {\"mera_parameters\": 6}\n",
    "qlayer = qml.qnn.TorchLayer(inverce_MERA_layer, weight_shapes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# working with data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Concentrating on the first 100 samples\n",
    "n_samples = 100\n",
    "\n",
    "X_train = datasets.MNIST(root='./data', train=True, download=True,\n",
    "                         transform=transforms.Compose([transforms.ToTensor()]))\n",
    "\n",
    "# Leaving only labels 0 and 1 \n",
    "idx = np.append(np.where(X_train.targets == 0)[0][:n_samples], \n",
    "                np.where(X_train.targets == 1)[0][:n_samples])\n",
    "\n",
    "X_train.data = X_train.data[idx]\n",
    "X_train.targets = X_train.targets[idx]\n",
    "# X_train.targets = (X_train.targets[idx]*2) + (np.ones(200)*(-1))\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(X_train, batch_size=1, shuffle=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x216 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_samples_show = 6\n",
    "\n",
    "data_iter = iter(train_loader)\n",
    "fig, axes = plt.subplots(nrows=1, ncols=n_samples_show, figsize=(10, 3))\n",
    "\n",
    "while n_samples_show > 0:\n",
    "    images, targets = data_iter.__next__()\n",
    "\n",
    "    axes[n_samples_show - 1].imshow(images[0].numpy().squeeze(), cmap='gray')\n",
    "    axes[n_samples_show - 1].set_xticks([])\n",
    "    axes[n_samples_show - 1].set_yticks([])\n",
    "    axes[n_samples_show - 1].set_title(\"Labeled: {}\".format(targets.item()))\n",
    "    \n",
    "    n_samples_show -= 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "n_samples = 50\n",
    "\n",
    "X_test = datasets.MNIST(root='./data', train=False, download=True,\n",
    "                        transform=transforms.Compose([transforms.ToTensor()]))\n",
    "\n",
    "idx = np.append(np.where(X_test.targets == 0)[0][:n_samples], \n",
    "                np.where(X_test.targets == 1)[0][:n_samples])\n",
    "\n",
    "X_test.data = X_test.data[idx]\n",
    "X_test.targets = X_test.targets[idx]\n",
    "# X_test.targets = (X_test.targets[idx]*2) + (np.ones(100)*(-1))\n",
    "\n",
    "test_loader = torch.utils.data.DataLoader(X_test, batch_size=1, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Creating the Hybrid Neural Network "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Net(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Net, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(1, 6, kernel_size=5)\n",
    "        self.conv2 = nn.Conv2d(6, 16, kernel_size=5)\n",
    "        # self.dropout = nn.Dropout2d()\n",
    "        self.fc1 = nn.Linear(256, 64)\n",
    "        self.fc2 = nn.Linear(64, 4)\n",
    "        self.qnode = qlayer\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.conv1(x))\n",
    "        x = F.max_pool2d(x, 2)\n",
    "        x = F.relu(self.conv2(x))\n",
    "        x = F.max_pool2d(x, 2)\n",
    "        # x = self.dropout(x)\n",
    "        x = x.view(1, -1)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = self.fc2(x)\n",
    "        x = self.qnode(x)\n",
    "        x = x.view(1, -1)\n",
    "        return torch.cat((x, 1 - x), -1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Training [5%]\tLoss: -0.8031\n",
      "Training [10%]\tLoss: -1.3218\n",
      "Training [15%]\tLoss: -1.3730\n",
      "Training [20%]\tLoss: -1.3993\n",
      "Training [25%]\tLoss: -1.4221\n",
      "Training [30%]\tLoss: -1.4412\n",
      "Training [35%]\tLoss: -1.4457\n",
      "Training [40%]\tLoss: -1.4551\n",
      "Training [45%]\tLoss: -1.4637\n",
      "Training [50%]\tLoss: -1.4667\n",
      "Training [55%]\tLoss: -1.4693\n",
      "Training [60%]\tLoss: -1.4775\n",
      "Training [65%]\tLoss: -1.4792\n",
      "Training [70%]\tLoss: -1.4797\n",
      "Training [75%]\tLoss: -1.4807\n",
      "Training [80%]\tLoss: -1.4800\n",
      "Training [85%]\tLoss: -1.4826\n",
      "Training [90%]\tLoss: -1.4863\n",
      "Training [95%]\tLoss: -1.4880\n",
      "Training [100%]\tLoss: -1.4881\n"
     ]
    }
   ],
   "source": [
    "model = Net()\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.001)\n",
    "loss_func = nn.NLLLoss()\n",
    "\n",
    "epochs = 20\n",
    "loss_list = []\n",
    "\n",
    "model.train()\n",
    "for epoch in range(epochs):\n",
    "    total_loss = []\n",
    "    for batch_idx, (data, target) in enumerate(train_loader):\n",
    "        optimizer.zero_grad()\n",
    "        # Forward pass\n",
    "        output = model(data)\n",
    "        # Calculating loss\n",
    "        loss = loss_func(output, target)\n",
    "        # Backward pass\n",
    "        loss.backward()\n",
    "        # Optimize the weights\n",
    "        optimizer.step()\n",
    "        \n",
    "        total_loss.append(loss.item())\n",
    "    loss_list.append(sum(total_loss)/len(total_loss))\n",
    "    print('Training [{:.0f}%]\\tLoss: {:.4f}'.format(\n",
    "        100. * (epoch + 1) / epochs, loss_list[-1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Neg Log Likelihood Loss')"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYoAAAEWCAYAAAB42tAoAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAAsTAAALEwEAmpwYAAAvMklEQVR4nO3deZhjZZn38e+d1NLVVb3Qi0ADzS5CcAERWRUHVEQFYRQdxeA2xHHXcUZmeEfRGWZcxn3USYsKEXAdQWSRxRGREWy6kaXDYiN70/RG07WmqpLc7x/nVHWqOkmluypLJb/PdeVKcs6Tc+6kljvPc57F3B0REZFSIvUOQEREGpsShYiIlKVEISIiZSlRiIhIWUoUIiJSlhKFiIiUpUQhFYulkrfEUsn37UT55bFUsj+WSkZL7L8wlkpeNnMR1k4slfznWCp58UyXFWlEbfUOQGonlko+BrwvHU/cXLDtXeG2E2b6fOl44gmgZ1deG0slTwJ+C3wnHU98oGD7bcDF6XjikjD2HwCfSscTXywo8xRwTjqeuGXSMa8HTgyfdgIOjITPL0vHE++vNL50PPHv1Si7s2KppAEfBs4D9ge2ArcDn0vHE/dV67zSWlSjkKqIpZIz8SVkAHhnLJXcr0yZZ4F/jKWS86Y6WDqeeF06nuhJxxM9wOXAF8eeFyaJGYq9Vr4OfBT4CLAIeD5wFfD6OsY0wSz7PKUI/QBlXCyV/AfgmHQ88dcF274BeDqe+Gi46cBYKrkSeAHBN/53p+OJZ8N/5o8C7wM+AzwWSyXj4bb2dDyRjaWS+wOXAEcCdwAPTRHSc8CV4fHeXaLMAwTfoj8BfHan3nCBWCrpwIeAjxH8XewfSyW/DpwFLADWAh9LxxO/D8tfCByUjifOKXjv7wL+FZgLfDUdT1y0C2W7gP8GTgeeIagxfSQdT+xdJOaDgQ8Cx6bjiZUFuy4vKLMA+CbwOmAQ+C7w7+l4Ij9WmyT4WbyX4PP+QDqeuD6WSr4V+Id0PHFUwbE+DrwqHU+cHkslO4GLgLMJamdXAh9PxxNDYW3wsvC8HwduiqWS55V7X7FUcllY/hVAf/iZfKPg8zsMyABnAk8A56bjiVXh/n0IEuaJBF9+f5SOJz4U7nsP8A/AHsBK4Lx0PPH45M9SylONQgpdBpwaSyUXwvg3wbcBqYIyceA9wJ5AFvjGpGO8EjgUeG2R418BrAaWEPyTPLeCmC4C/jqWSh5Spsy/AB+LpZKLKjheOW8CXk7wTwngTuAlBN/UrwB+Fksl55R5/QnAIcDJwKdjqeShu1D2M8B+wAHAq4FzyhzjZOCpSUlism8SJLoDCH42cSYm3ZcTJOwlwBeB74XNWb8CDgmT0Zi3E3wOAJ8nqL28BDgI2Av4dEHZPQg+t30JmsVKvq9YKhkJz3dPeJyTCX6ehb9DpwM/BhYCVwP/Fb42ClwDPB4ef6+wHLFU8gzgnwmS/VLg98CPynxWUoISReu5KpZKPjd2A749tiMdT6wHbgXeEm46FdicjidWF7z+h+l4Yk06nhgg+Ad99qSL1Rem44mBdDwxVHjSWCq5HHgZ8C/peGI4HU/cSvDPoax0PPEMwTfRz5UpczdwE/CpqY43hf9IxxPPjsWejicuS8cTW9LxRDYdT3yZ4JtzuYT12XQ8MZSOJ+4h+Kf34l0oezbBN/6t6XjiKXZMxIUWA+tL7Qx/Lm8D/ikdT/Sl44nHgC8D7ywo9ng6nvhuOp7IAZcSfAHYPR1PDAK/BP4mPNbBBLXIq8NEch5BDeLZdDzRB/x7eK4xeeAz4c96aIr39TJgaTqe+Fw6nhhJxxOPENR8Co93WzqeuC6M84cFn9fRwDKC2s9AOp7IpOOJ28J97yf4mT6QjieyYYwviaWS+5b5TKUINT21njcVu5hdsP9S4O8I/lDPIfijLPRkwePHgXaCb6PF9hdaBmwNE0zh6/epIOYvAH+JpZLl/vF+GlgZSyW/UsHxSpkQeyyV/CRBk8wyggvf85n4Xid7puDxIOUv5Jcqu2xSHKU+T4AtBP/YS1lC8PMpbGp5nOBb9w5xpOOJwVgqSUEsVxAkls8R1CauCss8j6DJbHVYHsCAwi8Mm9LxRKbgebn3tS+wLPziMiZKUAPYIU6Cz2tOWOPdhyDZZdnRvsDXY6nklwu2GcH7V/PTTlCikMmuAr4TSyUPB94A/OOk/YX/2JcDo8Dmgu2lpiNeD+wWSyW7C5LF8jLlx6XjiS2xVPJrBM1Vpco8GEslfwFcMNXxyhiPJZZKnkjw3k8G0mGb/laCfzTVtB7YG7g/fF4ukf4G+FYslTxqrL1+ks0EP599C463HFhXYSw3AUtjqeRLCGoWHy847hAQS8cTpY41+eda7n09CTyajicOZuc9CSyPpZJtRZLFk8BF6Xji8iKvk52gRCETpOOJTCyV/DnBt8mVYRfXQufEUskU8BjBN82fp+OJXME3y1LHfTyWSq4CPhtLJf+ZoMngjQTtzZX4CvAI5f9Rfxa4d4oylZpHcA1mE9AWSyXPJ6hRVNtPgX+KpZJ3Enxr/1Cpgul4Ym0slfw28KNYKvm3wB8ImpPfBOyXjic+H0slfwpcFHYsWERw0f8/KwkkHU+MxlLJnwFfCl97U7g9H0slvwt8NZZKfigdT2yMpZJ7AYen44kbduF9rQT6YqnkpwiapEYIrnN1peOJO6cIcyVBEvp8LJX8DJADXpqOJ/6PoMnyX2Op5N3peCIdXth/TTqe+Fkl71+20zUKKeZS4IXs2OxEuO0SgqaAOQTdMiv1doKLp88SXNxMlS++XTqe6CW42FrygnU6nng0jK97J2Iq5Qbg18CfCZopMpRvBpopnwOeIugZdTPwc2C4TPmPEFzY/RZBr6W/EPQMGrv+82GCbsaPALcRfAH4/k7EcwVwCvCzSd/YPwU8DNwRSyV7w1jLXb8p+b7C6w5vILgw/ihBjeVigovwZYWvfSPBBfUnwnO8Ndx3JUGz5Y/DGNcQ9P6SnWRauEgmCy88PwjsEf6DljqJpZJ/B7wtHU+8st6xzKRmfV/NSk1PMkHYVfETwI+VJGovlkruSdCF9HbgYODvCbuCzmbN+r5ahRKFjIulkt3ABoKmllPrHE6r6gCSBNNxPEcwJuDb5V4wSzTr+2oJanoSEZGydDFbRETKasqmJ2trd9o76h2GiMjskRnc7O5Li+1qykRBewcceHi9oxARmT3SK0uOVlfTk4iIlKVEISIiZSlRiIhIWUoUIiJSVt0ShZktMrObzGxteL9biXJfNLO0mT1gZt8ws2rP3ikiIgXqWaM4H/iNux9MMF3y+ZMLmNlxwPHAi4DDCRY40dwwIiI1VM9EcQbBLKWE928qUsYJZijtIFhdrJ1gigkREamReiaK3d19bBnHZ4DdJxdw99uB3xLMN78euMHdHyh2MDM7z8xWmdkqcsUWu5ra/INOYM7SA3fptSIizaqqA+7M7GaCRdYnm7AKmbu7me0w6ZSZHUSwgMne4aabzOxEd//95LLuvgJYAWBd3bs0gdWCA4+j/6l7yWz6y668XESkKVU1Ubj7KaX2mdkGM9vT3deb2Z7AxiLFzgTucPf+8DXXA8cycS3dGZMd6qWtqxaLmImIzB71bHq6Gjg3fHwu8MsiZZ4AXmlmbWbWTnAhu2jT00zIZfqIzplXrcOLiMxK9UwUnwdebWZrCZZa/DyAmR1lZheHZX5OsLTjfcA9wD3u/qtiB5sJuUwv0TmqUYiIFKrbpIDuvgU4ucj2VcD7wsc5IFGrmLKZPqKd3WAGWqdDRATQyOwJcplezCJEO3vqHYqISMNQoiiQy/QBqPlJRKSAEkWBbKYXgDZd0BYRGadEUSA3FCSKqLrIioiMU6IokB8dwnNZNT2JiBRQopgkm+lT05OISAElikk0lkJEZCIlikk0OltEZCIlikmymV7aVKMQERmnRDFJLtOHRduItHfVOxQRkYagRDFJLhxLoeYnEZGAEsUk2aFwdLbGUoiIAEoUO8hpdLaIyARKFJPkhvtxd3WRFREJKVFM5nlyw/26RiEiElKiKCKX6VMXWRGRkBJFEcHobNUoRERAiaKoYHS2ahQiIqBEUVR2qJdoRxcWqdtKsSIiDUOJoojxle40lkJEpD6JwswWmdlNZrY2vN+tRLkvmNma8PbWWsW3fUlUXacQEalXjeJ84DfufjDwm/D5BGb2euBI4CXAy4FPmllNvuJvXxJVNQoRkXolijOAS8PHlwJvKlLmMOBWd8+6+wBwL3BqLYLTfE8iItvVK1Hs7u7rw8fPALsXKXMPcKqZzTWzJcCrgH1KHdDMzjOzVWa2ilx2WsF5bpT8aEaJQkQEqFq3HjO7GdijyK4LCp+4u5uZTy7k7jea2cuAPwCbgNuBXKnzufsKYAWAdXXvcLydpXUpREQCVUsU7n5KqX1mtsHM9nT39Wa2J7CxxDEuAi4KX3MF8OeqBFuExlKIiATq1fR0NXBu+Phc4JeTC5hZ1MwWh49fBLwIuLFWAeaGeol2qelJRKReI8o+D/zUzN4LPA6cDWBmRwHvd/f3Ae3A780MoBc4x92nd/FhJ2QzfUQ7e8AMfNotWSIis1ZdEoW7bwFOLrJ9FfC+8HGGoOdTXeQyfZhFiHb2jI+rEBFpRRqZXcL2LrK6TiEirU2JooSsRmeLiABKFCXlNDpbRARQoigpPzKI57KqUYhIy1OiKCM33KdEISItT4mijOxQH22aalxEWpwSRRlaElVERImirKym8RARmTpRmNlHzWy+Bb5nZneZ2WtqEVy95TK9RKLtRNrn1DsUEZG6qaRG8R537wVeA+wGvJNgCo6mt32lO9UqRKR1VZIoLLw/Dfihu6cLtjU1jc4WEaksUaw2sxsJEsUNZjYPyFc3rMYwNjq7TRe0RaSFVTIp4HsJ1q1+xN0HzWwR8O6qRtUgcpk+3J2ousiKSAurpEZxLPCQuz9nZucA/w/YVt2wGoTnyQ8PqIusiLS0ShLFd4BBM3sx8PfAX4BUVaNqIMGSqEoUItK6KkkUWXd34Azgv9z9W0DL/OfUkqgi0uoqSRR9ZvZPBN1irzWzCMHqcy1Bo7NFpNVVkijeCgwTjKd4Btgb+FJVo2og2Uwf0Y65WKReq8aKiNTXlIkiTA6XAwvM7A1Axt1b5hrF9rEUqlWISGuqZAqPs4GVwFuAs4E/mtmbqx1Yo9DobBFpdZU0PV0AvMzdz3X3OHA08C/TOamZvcXM0maWN7OjypQ71cweMrOHzez86ZxzV2WHwpXuulSjEJHWVEmiiLj7xoLnWyp8XTlrgLOAW0sVMLMo8C3gdcBhwN+Y2WHTPO9OU41CRFpdJVdof21mNwA/Cp+/Fbh+Oid19wcAzMpOGXU08LC7PxKW/TFBF937p3PuneW5EfKjGV2jEJGWNWWicPd/MLOzgBPCTSvc/crqhgXAXsCTBc+fAl5eqrCZnQecB0B7x4wGonUpRKSVVdTn091/Afxi7LmZPeHuy8u9xsxuBvYosusCd//lTkVZAXdfAawAsK5un8lj5zJ9Gp0tIi1rVwcHTDnNuLufsovHHrMO2Kfg+d7htprLZXppX3JAPU4tIlJ3u3pReka/sZdwJ3Cwme1vZh3A24Cra3DeHeQyfUQ7e2iRZThERCYoWaMws0+U2gX0TOekZnYm8E1gKcG0IHe7+2vNbBlwsbuf5u5ZM/sQcAMQBb4fLppUc9mhXiwSIdrZTW64vx4hiIjUTbmmp3KN8l+fzknDi+E7XBB396cJFkgae34dcN10zjUTxrvIds1XohCRllMyUbj7Z2sZSCPTNB4i0sqmO3CuJWTDRNGmLrIi0oKUKCqQHxnE8zmNpRCRlqREUSGtSyEirWpXej0B4O5fmflwGlc206emJxFpSZX0ejoEeBnbxzC8kWDa8ZaSy/TRMb/YQHMRkeY2Za8nM7sVONLd+8LnFwLX1iS6BpIb6iW6+/PrHYaISM1Vco1id2Ck4PlIuK2lZDN9RKLtRNrn1DsUEZGaqmSupxSw0syuJBiVfQZwSTWDakSFYynyo5k6RyMiUjuVTDN+kZldD5xIMMfTu939T1WPrMEULmA02repztGIiNROpbPH5oA8QaLIVy+cxrV90J26yIpIa5nyGoWZfRS4HFgCPA+4zMw+XO3AGo2WRBWRVlVJjeK9wMvdfQDAzL4A3E4w+2vr8Dy54X4NuhORllNJrycjaHoak6NFF2bQkqgi0ooqqVH8APjjpF5P36tqVA0qN9RL29wF9Q5DRKSmKun19BUzuwU4gRbu9QTBdYrORftMXVBEpIlUOilgjiBJtGyvJwh6PkU75kIkWu9QRERqRr2edsJYzydNDigirUS9nnbC9i6y88gObq1zNCIitaFeTzshOz6Nh2oUItI6KkkUY72eLgxnjr2DafZ6MrO3mFnazPJmdlSZct83s41mtmY655spOY3OFpEWNGWiCBcoeg/wbHh7t7t/bZrnXQOcBdw6RblLgFOnea4Z49kR8qPDGnQnIi2l0rme7gbWj5U3s+Xu/sSuntTdHwiPM1W5W81sv109TzXkMr1Eu9T0JCKtY8pEEfZw+gywge3XJxx4UXVD2zlmdh5wHgDtHVU7j5ZEFZFWU0mN4qPAIe6+ZWcObGY3A8XWDr3A3X+5M8eqhLuvAFYAWFe3z/Txx+QyvbQv2b9ahxcRaTiVJIongW07e2B3P2Xnw2l8uUwf0c55bK9YiYg0t5KJwsw+ET58BLjFzK4Fhsf2hxe5W04204tFIkQ7u8kN99c7HBGRqivX62leeHsCuAnoKNg2rW4/ZnammT0FHAtca2Y3hNuXmdl1BeV+RDC47xAze8rM3jud884ErUshIq2mZI3C3T9brZO6+5XAlUW2Pw2cVvD8b6oVw64qHJ298w1yIiKzT7mmp6+5+8fM7FcUaYx399OrGlmDyg6Fg+7URVZEWkS5i9k/DO//sxaBzBb5kQE8n9OgOxFpGeWanlaH97+rXTizQy7Tp0QhIi2jXNPTfRTv/2mAu3tDDbirJQ26E5FWUq7p6Q01i2KWyWV66Zi/e73DEBGpiZLdY9398bFbuOng8PFGgskBW1bQ9KQahYi0hkpWuPtb4OdAMty0N3BVFWNqeNlML5G2Dqyts96hiIhUXSXrUXwQOB7oBXD3tQRLoras3FgXWdUqRKQFVJIoht19ZOyJmbXR4pMcjQ+601gKEWkBlSSK35nZPwNdZvZq4GfAr6obVmObMDpbRKTJVZIozgc2AfcBCeA6d7+gqlE1uOxwkCjU9CQiraCSacYvdPdPA98FMLOomV3u7u+obmgNLJ8jNzygGoWItIRKahT7mNk/AZhZB/A/wNqqRjULZDO9ShQi0hIqSRTvAV4YJotrgN+5+4VVjWoWyGl0toi0iJKJwsyONLMjgSOArwNvJahJ/C7c3tJyqlGISIsod43iy5OebwUOC7c78FfVCmo2yA31Ee3shkgU8rl6hyMiUjXlZo99VS0DmW2ymbFBd/PIDj5X32BERKqo3Oyx57j7ZQVrZ0/QqmtmjylcElWJQkSaWbmmp+7wvlhDfEuPzIbgGgVo0J2INL9yTU/J8H6HtbPN7GNVjGlWyGY06E5EWkMl3WOLKdocVSkze4uZpc0sb2ZHlSizj5n91szuD8t+dDrnnGmeHSafHVaNQkSa3q4mCpvmedcAZwG3limTBf7e3Q8DjgE+aGaHTfO8M0rrUohIK6hkCo9ipnWNwt0fADArnW/cfT2wPnzcZ2YPAHsB90/n3DMpq7WzRaQFlOv11EfpNbO7qhZR8Vj2Ixj498cyZc4DzgOgvaMmceWGepmzZL+anEtEpF7KXcye1ldlM7sZ2KPIrgvc/Zc7cZwegvmlPubuvaXKufsKYAWAdXXXpFdWLtNHtLOnFqcSEambXW16mpK7nzLdY5hZO0GSuNzdfzH9qGZWNtOLRaJEOrvJDw/UOxwRkarY1YvZVWfBBYzvAQ806uC+XEZLoopI86tLojCzM83sKeBY4FozuyHcvszMrguLHQ+8E/grM7s7vJ1Wj3hL0Up3ItIKqtb0VI67XwlcWWT708Bp4ePbmH433KrKjo/OVo1CRJrXlImiRO+nbcAqgnEOj1QjsNkgPzyA53NqehKRplZJjeJrwFPAFQTf8N8GHAjcBXwfOKlKsc0KOY2lEJEmV8k1itPdPenufe7eG3ZDfa27/wTYrcrxNbxspo9ol2oUItK8KkkUg2Z2tplFwtvZQCbcp1lkM320qUYhIk2skkTxDoLeRxvD2zuBc8ysC/hQFWObFYIlUVWjEJHmNeU1ivBi9RtL7L5tZsOZfXKZPiJtHVhbJ54drnc4IiIzbsoahZntbWZXmtnG8PY/ZrZ3LYKbDQqXRBURaUaVND39ALgaWBbefhVuEyYuiSoi0owqSRRL3f0H7p4Nb5cAS6sc16yhJVFFpNlVkii2mNk5ZhYNb+cAW6od2GwxviSqusiKSJOqJFG8BzgbeIZgIaE3A++qYkyzSz5HbnhANQoRaVqV9Hp6HDi9cJuZ/SfwyWoFNduoi6yINLNdnT327BmNYpbLatCdiDSxXU0UDT2ra60F8z2pRiEizancmtmLSu1CiWKCXKaXaGc3RKKQz9U7HBGRGVXuGsVqgrmciiWFkeqEMzuN93zqnEd26Ln6BiMiMsNKJgp337+WgcxmuaHtYymUKESk2TTsmtmzyfjobI2lEJEmpEQxA7IanS0iTawuicLM3mJmaTPLm9lRJcrMMbOVZnZPWPaztY6zUp4dJp8d0ZKoItKUKlkzu1jvpz53H53GedcAZwHJMmWGgb9y934zawduM7Pr3f2OaZy3aoJBd6pRiEjzqWTN7LuAfYCtBD2gFgLPmNkG4G/dffXOntTdHwAwK93L1t0d6A+ftoe3hl1RL6uxFCLSpCpperoJOM3dl7j7YuB1wDXAB4BvVzO4cBLCuwlW1rvJ3f9YzfNNRy7Tq9HZItKUKkkUx7j7DWNP3P1G4NiwCaiz1IvM7GYzW1Pkdkalwbl7zt1fAuwNHG1mh5c533lmtsrMVpHLVnqKGROMzlaiEJHmU0nT03oz+xTw4/D5W4ENZhYF8qVe5O6nzEB8Y8d6zsx+C5xKcH2jWJkVwAoA6+queRNVdqgXi0SJdHSTHxmo9elFRKqmkhrF2wm+0V8FXElwveLtQJQqTg5oZkvNbGH4uAt4NfBgtc43XTmtSyEiTaqSacY3Ax82s253n/xV+eFdOamZnQl8k2ClvGvN7G53f62ZLQMudvfTgD2BS8OaSwT4qbtfsyvnq4XtS6LOg23r6xyNiMjMqaR77HHAxUAPsNzMXgwk3P0Du3pSd7+SoHYyefvTwGnh43uBI3b1HLW2fUlU1ShEpLlU0vT0VeC1hMufuvs9wCuqGdRslBsewPN5OhfsWe9QRERmVEUjs939yUmbNJf2Dpz+p+6mZ/kRLDx0xq7ji4jUXSW9np4Mm588HCH9UeCB6oY1Oz177zWQz7HgwOOItM/h2fuuBW/YMYIiIhWpJFG8H/g6sBewDrgR+GA1g5rNnl1zPbmRQRY+/5VE27vY9KdfaDEjEZnVKu319I4axNI0tv35d+RHhlh0+Kns3v52Nq76CZ7VWk8iMjuVWwr102Ve5+7+r1WIp2n0PbaS/Oggi198BrsfE2fjyivIjwzWOywRkZ1W7mL2QJEbwHuBT1U5rqYwsG4Nm1b9lPZ5S9njuHep66yIzEolE4W7f3nsRjA1RhfwboKpPA6oUXyz3tDGtWy84zKinT3scfy7aetZUu+QRER2StnusWa2yMz+DbiXoJnqSHf/lLtvrEl0TWJ465M884dLMIuyx3HvomPhsnqHJCJSsZKJwsy+BNwJ9AEvdPcL3X1rzSJrMqN9G3nmDz8gPzrM7sfEmbNk/3qHJCJSEfMS/fzNLE+wylyWiQsGGcHF7IZtcLeubufAkjOS11W0s4fnvfwdtHcvZvPdVzK4XkNSRKQBpFeudveiS1OXu0YRcfcud5/n7vMLbvMaOUk0utxwPxtuv5ThbetZcuSb6Vk+a6azEpEWVdEUHjKz8qMZNt7xQzKbHmbxi97I/AOPr3dIIiIlKVHUieezbLzzJwysu4/dDj2ZhYe+ut4hiYgUVckUHlItnmfzn64kNzLEggOPJdrRxZZ7f6X5oUSkoShRNICt6V+THxlk4SEn0TF/D7Y9fFt4kVsJQ0TqT4miQWxbeyuj/ZtZ+PxXsvSlb2Z0YAu9D/+B/nX3alJBEamrkt1jZ7NG7h5bia49XsCCg06gc+Eyspleev9yB/1PrMZzo/UOTUSaVZnusUoUDWzOkgNYcNDxzFmyP7mRQfoeXRlONpipd2gi0mzKJAo1PTWwzOZHyGx+hI6Fe7HgoBNYeMhJzD/wOPofX03vo3eQy/TVO0QRaQF16R5rZm8xs7SZ5c2saAYrKBs1sz+Z2TW1iq/RjDy3jk2rfsLTv/sOg888yLz9X85er/owi174etrm7lbv8ESkydWrRrEGOAtIVlB2bOnVlh8NPtq3iS13X8W2h25h/oHH0rPPEfQsP4LBp+9n28P/x2jfhnqHKCJNqC6Jwt0fADCzsuXMbG/g9cBFwCeqH9nskB16jmfXXM9za29l/v7HMG/fo+je63AGN6yl95HbGd7yWL1DFJEm0ujXKL4G/CMwb6qCZnYecB4A7R1VDapR5IcHeO7B37Dt4duYt9/LmL//y9nj2DijfZvoe2I1/U/eg2eH6x2miMxyVUsUZnYzsEeRXRe4+y8reP0bgI3uvtrMTpqqvLuvIFhgKej11EI8O0zvw7fR98gdzN3zMObtdxSLYqey8AUnM7huDX2Pr2Jk2/p6hykis1TVEoW7nzLNQxwPnG5mpwFzgPlmdpm7nzP96JqT57MMrLuXgXX30j5/D+bt+1K693ohPcuPYPi5dfQ9torBp9N4PlvvUEVkFqnrOAozuwX4pLuvmqLcSWG5N1R03CYZRzETrK2Tnr1fRM++R9Exbym5kSEGnrqHvsdXkx3YUu/wRKRRNNo4CjM7E/gmsBS41szudvfXmtky4GJ3P60ecTUjzw7T99id9D12J52L9mXevi8NrmcccAxDmx+l/7E7GdzwZ/B8vUMVkQalkdktKNLRTc/yI5i3/Eja5i4km+mj/4m76H/iLg3iE2lVmsJDijO6nncQ8/Y7ijlLDwJ3hjb9hcGn1zC44SE8O1LvAEWkVhqt6UkahTO0cS1DG9fS1rWQnn2PpHvZ4cw94kw8l2Vw41oGn17D0Ia1ugAu0sJUo5AddO62N3OXxejeM0Z0Tg/57AhDGx5iYF2aoU0P63qGSDNSjUJ2xvDWpxje+hRb0zcyZ/G+zF0WY+6eh9K91wvJjQwx+MwDDD6dJrPlMa3GJ9ICVKOQyliEOUsPoHtZjLm7v4BIeye54X4G1t/P4Lo0w1ufrHeEIjIdupgtM8kibcx53kF0L4vRtfvziUTbyQ5tY+Dp+xnZ+iS5kSHyIwPkRgbJjwyhJV1FZgE1PclM8nyWoWceZOiZB7FoB127P5/uZTHm7380duCxE8u6kx8ZDJPG2P0AueHBSdsHgvtMP0osIo1FiUKmxXMjQXfap9dgbZ20dS0k2jmXSMdcoh1ziXR0h/dziXbOpb1nMdGO5UQ6ujDbcTkUz2UZ7d/MSN9GRvs2Mdq/idG+TWQHt9bh3YkIKFHIDPLsMKN9GxitcMxepL0rTCDd44mlrXsRHfOWMmfRcnr2ftF42Xx2hNH+zUHy6NvISJhAckPbqvRuRGSMEoXUTX50iPzoUMk5p6ytg/aepXTMW0r7vOfRPm8pc5buT88+Ly44xvB4rWNkrAbSv1kJRGQGKVFIw/LsCCPPrWPkuXUTtkfa59Des5T2MIF0zFtK1+4H07P8iPEy+dwo2YEtjPZvCWoi/VvI9m9mdGALnhut9VsRmdWUKGTWyY9mGN765A5dciMdc2nvWUJ7z2Lau5fQ1rOYjgV7MnfPQydcD8kObStIIJvDhLJZ81yJlKBEIU0jPzLI8LNPMPzsExN3RKK0z11Ee89i2sYSSc8SevZ+MZH2zu2vz46Qy/SSzw6Tz47g2eHw8TA+OrL98fj2gjKjwXZNdSLNSIlCml8+F1672LTDrmhnz4TkEe3oJtLeiUU7ic5dSKStE2vrDO4j0SlP5Z7HsyPkcyPBfXYEz42G9yNF9hXej5Ib7ier5jFpMEoU0tJyw/3khvsZ3vLY1IUjUSJh0oi0dWLtnUSinUTaO7Ynk2gHkbYOrK2DSHT7fbSzm0jbbtv3RzuwyI7dg8dkM33j11iyA88yOvZ4cKvm2pKaU6IQqVQ+Rz4cIDgjItHtyaStA4t20janh7buxbR3L6KtZzFz93gB0c7u8Ze458kOPjcxeYSPc0O9aLCiVIMShUi95HPk80MwOkQu3FRsBZBI+xzauhcFyaN7Me3di2nrXkTPon2ItG2/xuK5LLnhfjyfxfO54D6XnXhfYjtj2/M5cMfxYMJHzwOOe/h8/HE+zEkFz8ce53O458Hz4fHyQZPc2ON8uM9zkM+Pl92BGVgk6IhgFnZICLdFgsdj+8bKuecnNOdp0sqZoUQh0uDyoxlGnnuakeee3mFfpLOb9rHk0bOYaMdcLNKGRaJYtC14HO0gMr69DYtGCx63Y2Z1eFc78nyYLi0yYzHlc6P4hGtBoxOvDRVeMxq7Hx0iP5oJb9sftzIlCpFZLD88wPDwwI49vXaGRcKk0bb9G/qEb+tjz8cej33Dn7QPg0gEs+j2b/3hN32LFG6LBseOhPssOv4Ys6DGgW+vfeBhrcPHayeTH49vswiRtvYJ14LGrwkVXDeKds7doUw57r5D4ij6ODtcUHvL4rnc+GMmbA9rb7OkqVCJQqTVhc01nmvlpW8Ni7YTae8MppZpnxPeuibdb3/cMWfe+GOL7tq/Us/nJiaWsSQ5sdSEux32h81rTtBFfMPtl+5SLOXUJVGY2VuAC4FDgaPdfVWJco8BfUAOyJaaAldEZHocz42Qy43s0sBLi7QFE12G3ai3N+21bX8ebZt4X6QMk7pg79gEN+n5pP357PBOx16JetUo1gBnAckKyr7K3TdXOR4RkV3m+WyYYJpzdH9dEoW7PwDFsqWIiDSa0iN+GoMDN5rZajM7r1xBMzvPzFaZ2SpymkZBRGSmVK1GYWY3A3sU2XWBu/+ywsOc4O7rzOx5wE1m9qC731qsoLuvAFZAuBSqiIjMiKolCnc/ZQaOsS6832hmVwJHA0UThYiIVEfDNj2ZWbeZzRt7DLyG4CK4iIjUUF0ShZmdaWZPAccC15rZDeH2ZWZ2XVhsd+A2M7sHWAlc6+6/rke8IiKtzLwJ50Kxrm7nwMPrHYaIyOyRXrm61Fi1hm16EhGRxtCcNQqzTcDju/jyJUAjD/BTfNOj+KZH8U1PI8e3r7svLbajKRPFdJjZqkaeKkTxTY/imx7FNz2NHl8panoSEZGylChERKQsJYodrah3AFNQfNOj+KZH8U1Po8dXlK5RiIhIWapRiIhIWUoUIiJSVssmCjM71cweMrOHzez8Ivs7zewn4f4/mtl+NYxtHzP7rZndb2ZpM/tokTInmdk2M7s7vH26VvGF53/MzO4Lz73DCoUW+Eb4+d1rZkfWMLZDCj6Xu82s18w+NqlMTT8/M/u+mW00szUF2xaZ2U1mtja8363Ea88Ny6w1s3NrGN+XzOzB8Od3pZktLPHasr8LVYzvQjNbV/AzPK3Ea8v+rVcxvp8UxPaYmd1d4rVV//ymzd1b7gZEgb8ABwAdwD3AYZPKfAD47/Dx24Cf1DC+PYEjw8fzgD8Xie8k4Jo6foaPAUvK7D8NuJ5g7cZjgD/W8Wf9DMFgorp9fsArgCOBNQXbvgicHz4+H/hCkdctAh4J73cLH+9Wo/heA7SFj79QLL5KfheqGN+FwCcr+PmX/VuvVnyT9n8Z+HS9Pr/p3lq1RnE08LC7P+LuI8CPgTMmlTkDGFul/OfAyVajJfncfb273xU+7gMeAPaqxbln0BlAygN3AAvNbM86xHEy8Bd339WR+jPCg3VUnp20ufB37FLgTUVe+lrgJnd/1t23AjcBp9YiPne/0d3HVgG7A9h7ps9bqRKfXyUq+VuftnLxhf83zgZ+NNPnrZVWTRR7AU8WPH+KHf8Rj5cJ/1i2AYtrEl2BsMnrCOCPRXYfa2b3mNn1ZharbWRTrj5YyWdcC2+j9B9oPT8/gN3dfX34+BmCGZMna5TP8T0ENcRiKl6Jsgo+FDaNfb9E010jfH4nAhvcfW2J/fX8/CrSqoliVjCzHuB/gI+5e++k3XcRNKe8GPgmcFWNwzvB3Y8EXgd80MxeUePzT8nMOoDTgZ8V2V3vz28CD9ogGrKvupldAGSBy0sUqdfvwneAA4GXAOsJmnca0d9QvjbR8H9LrZoo1gH7FDzfO9xWtIyZtQELgC01iS44ZztBkrjc3X8xeb+797p7f/j4OqDdzJbUKj4vWH0QGFt9sFAln3G1vQ64y903TN5R788vtGGsOS6831ikTF0/RzN7F/AG4B1hMttBBb8LVeHuG9w95+554Lslzlvvz68NOAv4Saky9fr8dkarJoo7gYPNbP/wW+fbgKsnlbkaGOth8mbgf0v9ocy0sE3ze8AD7v6VEmX2GLtmYmZHE/wsa5LIrLLVB68G4mHvp2OAbQXNLLVS8ptcPT+/AoW/Y+cCxdaSvwF4jZntFjatvCbcVnVmdirwj8Dp7j5YokzdVqKcdM3rzBLnreRvvZpOAR5096eK7azn57dT6n01vV43gl45fyboEXFBuO1zBH8UAHMImiweJlhh74AaxnYCQTPEvcDd4e004P3A+8MyHwLSBL047gCOq2F8B4TnvSeMYezzK4zPgG+Fn+99wFE1/vl2E/zjX1CwrW6fH0HCWg+MErSTv5fgmtdvgLXAzcCisOxRwMUFr31P+Hv4MPDuGsb3MEH7/tjv4FgvwGXAdeV+F2oU3w/D3617Cf757zk5vvD5Dn/rtYgv3H7J2O9cQdmaf37TvWkKDxERKatVm55ERKRCShQiIlKWEoWIiJSlRCEiImUpUYiISFlKFDJrmdnigtk5n5k0k2jHFK89ysy+UcE5/jBDsZ5kZtcUPD5uJo4bHm8/M3t7wfOK3ptIpdrqHYDIrnL3LQTTN2BmFwL97v6fY/vNrM23T2o3+bWrgCmndHb3GfuHXuAkoB+oOAmVey/AfsDbgSug8vcmUinVKKSpmNklZvbfZvZH4ItmdrSZ3W5mfzKzP5jZIWG5wm/4F4aTyt1iZo+Y2UcKjtdfUP4WM/u5BWs0XF4wsvu0cNtqC9bguKZMfPsRDPz7eFjzOdHMlprZ/5jZneHt+IK4fmhm/wf8MKw5/N7M7gpvY0ns88CJ4fE+Pum9LTKzq8KJ8+4wsxeVe8/hSOFrLZgscY2ZvXUGfzwyS6lGIc1ob4KR1jkzmw+c6O5ZMzsF+Hfgr4u85gXAqwjW/3jIzL7j7qOTyhwBxICngf8DjrdgoZkk8Ap3f9TMyk4l7e6Pmdl/U1D7MbMrgK+6+21mtpxgio5Dw5ccRjBp3JCZzQVe7e4ZMzuYYDTwUQRrWXzS3d8QHu+kglN+FviTu7/JzP4KSBHWwoq9Z4IpzJ9299eHx1pQ7v1Ia1CikGb0M3fPhY8XAJeG/1gdaC/xmmvdfRgYNrONBFN+T56fZ6WHc/ZYsFrZfgRNSI+4+6NhmR8BOztV9CnAYbZ9uZP5FswcDHC1uw+Fj9uB/zKzlwA54PkVHPsEwsTo7v8bXteZH+4r9p7vA75sZl8gWNjp9zv5XqQJKVFIMxooePyvwG/d/cyw2eeWEq8ZLnico/jfRiVldkUEOMbdM4Ubw8RR+F4+DmwAXhy+ZkL5XbDD+3H3P1uwbO1pwL+Z2W/c/XPTPI/McrpGIc1uAdunlX5XFY7/EHCAbV9TvZI2/T6C5p4xNwIfHnsS1hiKWQCs92Ba7XcSLPNZ7HiFfg+8IzzuScBm33Ftk3FmtgwYdPfLgC8RLO8pLU6JQprdF4H/MLM/UYUadNgs9AHg12a2muCf9rYpXvYr4Myxi9nAR4CjwgvO9xNc7C7m28C5ZnYPwfWFsdrGvUAuvAD98UmvuRB4qZndS3DR+1zKeyGwMmxa+wzwb1OUlxag2WNFpsnMety9P+wF9S1grbt/td5xicwU1ShEpu9vw2/gaYLmoWR9wxGZWapRiIhIWapRiIhIWUoUIiJSlhKFiIiUpUQhIiJlKVGIiEhZ/x/9YExUrtLuhQAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(loss_list)\n",
    "plt.title('Hybrid NN Training Convergence')\n",
    "plt.xlabel('Training Iterations')\n",
    "plt.ylabel('Neg Log Likelihood Loss')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# test and rsualt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Performance on test data:\n",
      "\tLoss: -1.4861\n",
      "\tAccuracy: 100.0%\n"
     ]
    }
   ],
   "source": [
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    \n",
    "    correct = 0\n",
    "    for batch_idx, (data, target) in enumerate(test_loader):\n",
    "        output = model(data)\n",
    "        \n",
    "        pred = output.argmax(dim=-1, keepdim=True) \n",
    "        correct += pred.eq(target.view_as(pred)).sum().item()\n",
    "        \n",
    "        loss = loss_func(output, target)\n",
    "        total_loss.append(loss.item())\n",
    "        \n",
    "    print('Performance on test data:\\n\\tLoss: {:.4f}\\n\\tAccuracy: {:.1f}%'.format(\n",
    "        sum(total_loss) / len(total_loss),\n",
    "        correct / len(test_loader) * 100)\n",
    "        )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x216 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "n_samples_show = 6\n",
    "count = 0\n",
    "fig, axes = plt.subplots(nrows=1, ncols=n_samples_show, figsize=(10, 3))\n",
    "\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    for batch_idx, (data, target) in enumerate(test_loader):\n",
    "        if count == n_samples_show:\n",
    "            break\n",
    "        output = model(data)\n",
    "        \n",
    "        pred = output.argmax(dim=1, keepdim=True) \n",
    "\n",
    "        axes[count].imshow(data[0].numpy().squeeze(), cmap='gray')\n",
    "\n",
    "        axes[count].set_xticks([])\n",
    "        axes[count].set_yticks([])\n",
    "        axes[count].set_title('Predicted {}'.format(pred.item()))\n",
    "        \n",
    "        count += 1"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.10.8 ('pytorch')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.8"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "52010cec013b31ef191393eb6bfa9555cfa9414d7025b933f1b59fa991772dbb"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}