ch13_part1.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "tags": []
   },
   "source": [
    "# Machine Learning with PyTorch and Scikit-Learn  \n",
    "# -- Code Examples"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Package version checks"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Add folder to path in order to load from the check_packages.py script:"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Chapter 13: Going Deeper -- the Mechanics of PyTorch (Part 1/3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Outline**\n",
    "\n",
    "- [The key features of PyTorch](#The-key-features-of-PyTorch)\n",
    "- [PyTorch's computation graphs](#PyTorchs-computation-graphs)\n",
    "  - [Understanding computation graphs](#Understanding-computation-graphs)\n",
    "  - [Creating a graph in PyTorch](#Creating-a-graph-in-PyTorch)\n",
    "- [PyTorch tensor objects for storing and updating model parameters](#PyTorch-tensor-objects-for-storing-and-updating-model-parameters)\n",
    "- [Computing gradients via automatic differentiation](#Computing-gradients-via-automatic-differentiation)\n",
    "  - [Computing the gradients of the loss with respect to trainable variables](#Computing-the-gradients-of-the-loss-with-respect-to-trainable-variables)\n",
    "  - [Understanding automatic differentiation](#Understanding-automatic-differentiation)\n",
    "  - [Adversarial examples](#Adversarial-examples)\n",
    "- [Simplifying implementations of common architectures via the torch.nn module](#Simplifying-implementations-of-common-architectures-via-the-torch.nn-module)\n",
    "  - [Implementing models based on nn.Sequential](#Implementing-models-based-on-nn-Sequential)\n",
    "  - [Choosing a loss function](#Choosing-a-loss-function)\n",
    "  - [Solving an XOR classification problem](#Solving-an-XOR-classification-problem)\n",
    "  - [Making model building more flexible with nn.Module](#Making-model-building-more-flexible-with-nn.Module)\n",
    "  - [Writing custom layers in PyTorch](#Writing-custom-layers-in-PyTorch)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from IPython.display import Image\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## The key features of PyTorch\n",
    "\n",
    "## PyTorch's computation graphs\n",
    "\n",
    "### Understanding computation graphs\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## PyTorch Tensor objects for storing and updating model parameters"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a = torch.tensor(3.14, requires_grad=True)\n",
    "b = torch.tensor([1.0, 2.0, 3.0], requires_grad=True) \n",
    "print(a)\n",
    "print(b)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "a.requires_grad"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "w = torch.tensor([1.0, 2.0, 3.0])\n",
    "\n",
    "print(w.requires_grad)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "w.requires_grad_()\n",
    "\n",
    "print(w.requires_grad)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "\n",
    "\n",
    "torch.manual_seed(1)\n",
    "w = torch.empty(2, 3)\n",
    "nn.init.xavier_normal_(w)\n",
    "print(w)\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "class MyModule(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.w1 = torch.empty(2, 3, requires_grad=True)\n",
    "        nn.init.xavier_normal_(self.w1)\n",
    "        self.w2 = torch.empty(1, 2, requires_grad=True)\n",
    "        nn.init.xavier_normal_(self.w2)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Computing gradients via automatic differentiation and GradientTape\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Computing the gradients of the loss with respect to trainable variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "w = torch.tensor(1.0, requires_grad=True)\n",
    "b = torch.tensor(0.5, requires_grad=True) \n",
    "\n",
    "x = torch.tensor([1.4])\n",
    "y = torch.tensor([2.1])\n",
    "\n",
    "\n",
    "z = torch.add(torch.mul(w, x), b)\n",
    " \n",
    "loss = (y-z).pow(2).sum()\n",
    "loss.backward()\n",
    "\n",
    "print('dL/dw : ', w.grad)\n",
    "print('dL/db : ', b.grad)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# verifying the computed gradient dL/dw\n",
    "print(2 * x * ((w * x + b) - y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Simplifying implementations of common architectures via the torch.nn module\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Implementing models based on nn.Sequential"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = nn.Sequential(\n",
    "    nn.Linear(4, 16),\n",
    "    nn.ReLU(),\n",
    "    nn.Linear(16, 32),\n",
    "    nn.ReLU()\n",
    ")\n",
    "\n",
    "model\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Configuring layers\n",
    "\n",
    " * Initializers `nn.init`: https://pytorch.org/docs/stable/nn.init.html \n",
    " * L1 Regularizers `nn.L1Loss`: https://pytorch.org/docs/stable/generated/torch.nn.L1Loss.html#torch.nn.L1Loss\n",
    " * L2 Regularizers `weight_decay`: https://pytorch.org/docs/stable/optim.html\n",
    " * Activations: https://pytorch.org/docs/stable/nn.html#non-linear-activations-weighted-sum-nonlinearity  \n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "nn.init.xavier_uniform_(model[0].weight)\n",
    " \n",
    "l1_weight = 0.01\n",
    "l1_penalty = l1_weight * model[2].weight.abs().sum()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Compiling a model\n",
    "\n",
    " * Optimizers `torch.optim`:  https://pytorch.org/docs/stable/optim.html#algorithms\n",
    " * Loss Functions `tf.keras.losses`: https://pytorch.org/docs/stable/nn.html#loss-functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_fn = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.001)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Solving an XOR classification problem"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "np.random.seed(1)\n",
    "torch.manual_seed(1)\n",
    "x = np.random.uniform(low=-1, high=1, size=(200, 2))\n",
    "y = np.ones(len(x))\n",
    "y[x[:, 0] * x[:, 1]<0] = 0\n",
    "\n",
    "n_train = 100\n",
    "x_train = torch.tensor(x[:n_train, :], dtype=torch.float32)\n",
    "y_train = torch.tensor(y[:n_train], dtype=torch.float32)\n",
    "x_valid = torch.tensor(x[n_train:, :], dtype=torch.float32)\n",
    "y_valid = torch.tensor(y[n_train:], dtype=torch.float32)\n",
    "\n",
    "fig = plt.figure(figsize=(6, 6))\n",
    "plt.plot(x[y==0, 0], \n",
    "         x[y==0, 1], 'o', alpha=0.75, markersize=10)\n",
    "plt.plot(x[y==1, 0], \n",
    "         x[y==1, 1], '<', alpha=0.75, markersize=10)\n",
    "plt.xlabel(r'$x_1$', size=15)\n",
    "plt.ylabel(r'$x_2$', size=15)\n",
    "\n",
    "#plt.savefig('figures/13_02.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torch.utils.data import DataLoader, TensorDataset\n",
    " \n",
    "    \n",
    "train_ds = TensorDataset(x_train, y_train)\n",
    "batch_size = 2\n",
    "torch.manual_seed(1)\n",
    "train_dl = DataLoader(train_ds, batch_size, shuffle=True)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sequential(\n",
       "  (0): Linear(in_features=2, out_features=1, bias=True)\n",
       "  (1): Sigmoid()\n",
       ")"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = nn.Sequential(\n",
    "    nn.Linear(2, 1),\n",
    "    nn.Sigmoid()\n",
    ")\n",
    "\n",
    "model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_fn = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.001)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.manual_seed(1)\n",
    "num_epochs = 200\n",
    "def train(model, num_epochs, train_dl, x_valid, y_valid):\n",
    "    loss_hist_train = [0] * num_epochs\n",
    "    accuracy_hist_train = [0] * num_epochs\n",
    "    loss_hist_valid = [0] * num_epochs\n",
    "    accuracy_hist_valid = [0] * num_epochs\n",
    "    for epoch in range(num_epochs):\n",
    "        for x_batch, y_batch in train_dl:\n",
    "            pred = model(x_batch)[:, 0]\n",
    "            loss = loss_fn(pred, y_batch)\n",
    "            loss.backward()\n",
    "            optimizer.step()\n",
    "            optimizer.zero_grad()\n",
    "            loss_hist_train[epoch] += loss.item()\n",
    "            is_correct = ((pred>=0.5).float() == y_batch).float()\n",
    "            accuracy_hist_train[epoch] += is_correct.mean()\n",
    "\n",
    "        loss_hist_train[epoch] /= n_train/batch_size\n",
    "        accuracy_hist_train[epoch] /= n_train/batch_size\n",
    "\n",
    "        pred = model(x_valid)[:, 0]\n",
    "        loss = loss_fn(pred, y_valid)\n",
    "        loss_hist_valid[epoch] = loss.item()\n",
    "        is_correct = ((pred>=0.5).float() == y_valid).float()\n",
    "        accuracy_hist_valid[epoch] += is_correct.mean()\n",
    "    return loss_hist_train, loss_hist_valid, accuracy_hist_train, accuracy_hist_valid\n",
    "\n",
    "history = train(model, num_epochs, train_dl, x_valid, y_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "([0.6849763441085815,\n",
       "  0.6849789452552796,\n",
       "  0.6849628388881683,\n",
       "  0.6849853837490082,\n",
       "  0.6850148952007293,\n",
       "  0.6849688553810119,\n",
       "  0.6849533462524414,\n",
       "  0.6849581480026246,\n",
       "  0.6849741876125336,\n",
       "  0.6849447691440582,\n",
       "  0.6849531865119934,\n",
       "  0.6849647295475007,\n",
       "  0.6849541962146759,\n",
       "  0.6849592018127442,\n",
       "  0.6849512434005738,\n",
       "  0.6849564325809478,\n",
       "  0.6849433445930481,\n",
       "  0.6849400901794434,\n",
       "  0.6849469745159149,\n",
       "  0.684949471950531,\n",
       "  0.6849367511272431,\n",
       "  0.6849219703674316,\n",
       "  0.6849459660053253,\n",
       "  0.684915566444397,\n",
       "  0.684931390285492,\n",
       "  0.6849301409721374,\n",
       "  0.6849223649501801,\n",
       "  0.6849297702312469,\n",
       "  0.6849168133735657,\n",
       "  0.6849211382865906,\n",
       "  0.6849245977401733,\n",
       "  0.6849147009849549,\n",
       "  0.6849098777770997,\n",
       "  0.6849123799800872,\n",
       "  0.6849095737934112,\n",
       "  0.6849171638488769,\n",
       "  0.6849183428287506,\n",
       "  0.6849106550216675,\n",
       "  0.6848993134498597,\n",
       "  0.6849026441574096,\n",
       "  0.6849029278755188,\n",
       "  0.6848958551883697,\n",
       "  0.6849106335639954,\n",
       "  0.6848908841609955,\n",
       "  0.6849149322509765,\n",
       "  0.6849101161956788,\n",
       "  0.684895327091217,\n",
       "  0.6848850524425507,\n",
       "  0.6848851573467255,\n",
       "  0.6849034547805786,\n",
       "  0.6849061930179596,\n",
       "  0.6848968005180359,\n",
       "  0.6848818051815033,\n",
       "  0.6849090588092804,\n",
       "  0.6848935568332672,\n",
       "  0.6848984801769257,\n",
       "  0.684898453950882,\n",
       "  0.6848932361602783,\n",
       "  0.6849045920372009,\n",
       "  0.6849009037017822,\n",
       "  0.6849037551879883,\n",
       "  0.6849017035961151,\n",
       "  0.684877359867096,\n",
       "  0.6848781549930573,\n",
       "  0.6848915266990662,\n",
       "  0.6849019610881806,\n",
       "  0.6848920202255249,\n",
       "  0.6848845875263214,\n",
       "  0.6848898983001709,\n",
       "  0.6848746287822723,\n",
       "  0.6848703038692474,\n",
       "  0.6848812437057495,\n",
       "  0.684890810251236,\n",
       "  0.6848704981803894,\n",
       "  0.6848904597759247,\n",
       "  0.6848882877826691,\n",
       "  0.6848724853992462,\n",
       "  0.6848876893520355,\n",
       "  0.6848758840560913,\n",
       "  0.6848761427402497,\n",
       "  0.6848804354667664,\n",
       "  0.6848655939102173,\n",
       "  0.6848663461208343,\n",
       "  0.6848678326606751,\n",
       "  0.6848726785182953,\n",
       "  0.6848569989204407,\n",
       "  0.6848764276504516,\n",
       "  0.6848705697059632,\n",
       "  0.6848860657215119,\n",
       "  0.6848923194408417,\n",
       "  0.6848782229423523,\n",
       "  0.684868848323822,\n",
       "  0.6848760914802551,\n",
       "  0.6848637783527374,\n",
       "  0.6848778688907623,\n",
       "  0.6848737168312072,\n",
       "  0.6848660576343536,\n",
       "  0.6848806071281434,\n",
       "  0.6848744833469391,\n",
       "  0.6848702692985534,\n",
       "  0.6848756611347199,\n",
       "  0.6848722326755524,\n",
       "  0.6848532402515412,\n",
       "  0.6848624885082245,\n",
       "  0.6848690295219422,\n",
       "  0.6848692667484283,\n",
       "  0.6848544645309448,\n",
       "  0.6848604393005371,\n",
       "  0.6848657131195068,\n",
       "  0.684878202676773,\n",
       "  0.6848615479469299,\n",
       "  0.6848555862903595,\n",
       "  0.684865517616272,\n",
       "  0.6848697662353516,\n",
       "  0.6848483037948608,\n",
       "  0.6848589563369751,\n",
       "  0.6848675644397736,\n",
       "  0.6848504388332367,\n",
       "  0.6848464059829712,\n",
       "  0.6848581516742707,\n",
       "  0.6848609375953675,\n",
       "  0.684887627363205,\n",
       "  0.6848741471767426,\n",
       "  0.6848730015754699,\n",
       "  0.6848525857925415,\n",
       "  0.6848613202571869,\n",
       "  0.6848695683479309,\n",
       "  0.684844799041748,\n",
       "  0.6848634088039398,\n",
       "  0.6848493218421936,\n",
       "  0.6848493254184723,\n",
       "  0.6848443472385406,\n",
       "  0.684851942062378,\n",
       "  0.684858511686325,\n",
       "  0.6848705339431763,\n",
       "  0.6848435330390931,\n",
       "  0.6848683774471283,\n",
       "  0.6848504972457886,\n",
       "  0.6848558282852173,\n",
       "  0.6848508703708649,\n",
       "  0.6848643851280213,\n",
       "  0.6848468959331513,\n",
       "  0.684864684343338,\n",
       "  0.6848644495010376,\n",
       "  0.684860713481903,\n",
       "  0.6848455834388733,\n",
       "  0.6848563647270203,\n",
       "  0.6848507797718049,\n",
       "  0.6848541402816772,\n",
       "  0.6848582720756531,\n",
       "  0.6848487675189971,\n",
       "  0.6848657298088073,\n",
       "  0.6848556578159333,\n",
       "  0.6848618352413177,\n",
       "  0.684849294424057,\n",
       "  0.6848419773578643,\n",
       "  0.6848416316509247,\n",
       "  0.6848632633686066,\n",
       "  0.6848629987239838,\n",
       "  0.6848483252525329,\n",
       "  0.6848569643497467,\n",
       "  0.6848445892333984,\n",
       "  0.6848549151420593,\n",
       "  0.684846179485321,\n",
       "  0.6848542153835296,\n",
       "  0.6848616290092469,\n",
       "  0.6848366212844849,\n",
       "  0.6848563063144684,\n",
       "  0.684869145154953,\n",
       "  0.6848474776744843,\n",
       "  0.6848471450805664,\n",
       "  0.6848555076122284,\n",
       "  0.6848633825778961,\n",
       "  0.6848599410057068,\n",
       "  0.6848612344264984,\n",
       "  0.6848606193065643,\n",
       "  0.6848476588726043,\n",
       "  0.684861478805542,\n",
       "  0.6848581683635712,\n",
       "  0.6848566997051239,\n",
       "  0.6848756110668183,\n",
       "  0.6848662066459655,\n",
       "  0.6848685467243194,\n",
       "  0.6848386549949645,\n",
       "  0.6848382174968719,\n",
       "  0.6848460006713867,\n",
       "  0.68486483335495,\n",
       "  0.6848359775543212,\n",
       "  0.6848607301712036,\n",
       "  0.6848610138893128,\n",
       "  0.6848300492763519,\n",
       "  0.6848530161380768,\n",
       "  0.68485675573349,\n",
       "  0.6848490381240845,\n",
       "  0.6848722875118256,\n",
       "  0.6848633575439453,\n",
       "  0.684848073720932,\n",
       "  0.6848435950279236,\n",
       "  0.6848557484149933,\n",
       "  0.68484947681427],\n",
       " [0.6797455549240112,\n",
       "  0.6797419786453247,\n",
       "  0.6797382831573486,\n",
       "  0.6797345876693726,\n",
       "  0.6797310709953308,\n",
       "  0.6797273755073547,\n",
       "  0.6797239780426025,\n",
       "  0.6797202229499817,\n",
       "  0.679716944694519,\n",
       "  0.6797134876251221,\n",
       "  0.6797101497650146,\n",
       "  0.6797068119049072,\n",
       "  0.6797033548355103,\n",
       "  0.6796999573707581,\n",
       "  0.6796965003013611,\n",
       "  0.6796930432319641,\n",
       "  0.6796897053718567,\n",
       "  0.6796864867210388,\n",
       "  0.6796832084655762,\n",
       "  0.679679811000824,\n",
       "  0.6796765923500061,\n",
       "  0.679673433303833,\n",
       "  0.6796701550483704,\n",
       "  0.6796672344207764,\n",
       "  0.6796642541885376,\n",
       "  0.6796608567237854,\n",
       "  0.6796578168869019,\n",
       "  0.6796545386314392,\n",
       "  0.6796517372131348,\n",
       "  0.6796484589576721,\n",
       "  0.679645299911499,\n",
       "  0.6796424388885498,\n",
       "  0.679639458656311,\n",
       "  0.6796364784240723,\n",
       "  0.6796332597732544,\n",
       "  0.6796303391456604,\n",
       "  0.6796273589134216,\n",
       "  0.6796243190765381,\n",
       "  0.6796214580535889,\n",
       "  0.6796184778213501,\n",
       "  0.6796156167984009,\n",
       "  0.6796128749847412,\n",
       "  0.6796103119850159,\n",
       "  0.6796074509620667,\n",
       "  0.6796047687530518,\n",
       "  0.6796018481254578,\n",
       "  0.6795990467071533,\n",
       "  0.6795963048934937,\n",
       "  0.6795936822891235,\n",
       "  0.6795909404754639,\n",
       "  0.6795881390571594,\n",
       "  0.679585337638855,\n",
       "  0.6795828342437744,\n",
       "  0.6795803308486938,\n",
       "  0.6795775890350342,\n",
       "  0.6795747876167297,\n",
       "  0.6795721650123596,\n",
       "  0.6795695424079895,\n",
       "  0.6795670390129089,\n",
       "  0.6795645356178284,\n",
       "  0.679561972618103,\n",
       "  0.6795592308044434,\n",
       "  0.6795567274093628,\n",
       "  0.6795539259910583,\n",
       "  0.6795514822006226,\n",
       "  0.6795491576194763,\n",
       "  0.6795464158058167,\n",
       "  0.6795438528060913,\n",
       "  0.6795412302017212,\n",
       "  0.6795385479927063,\n",
       "  0.6795361042022705,\n",
       "  0.6795334815979004,\n",
       "  0.679530918598175,\n",
       "  0.6795284748077393,\n",
       "  0.6795262098312378,\n",
       "  0.6795237064361572,\n",
       "  0.6795213222503662,\n",
       "  0.6795190572738647,\n",
       "  0.6795165538787842,\n",
       "  0.6795142292976379,\n",
       "  0.6795118451118469,\n",
       "  0.6795094013214111,\n",
       "  0.6795070767402649,\n",
       "  0.6795045733451843,\n",
       "  0.6795022487640381,\n",
       "  0.6794999837875366,\n",
       "  0.6794975399971008,\n",
       "  0.6794949173927307,\n",
       "  0.6794924736022949,\n",
       "  0.6794901490211487,\n",
       "  0.6794877648353577,\n",
       "  0.6794853806495667,\n",
       "  0.6794832348823547,\n",
       "  0.6794807314872742,\n",
       "  0.6794787049293518,\n",
       "  0.6794763803482056,\n",
       "  0.6794741153717041,\n",
       "  0.6794719099998474,\n",
       "  0.679469883441925,\n",
       "  0.6794677972793579,\n",
       "  0.679465651512146,\n",
       "  0.6794632077217102,\n",
       "  0.6794610023498535,\n",
       "  0.6794586777687073,\n",
       "  0.6794567704200745,\n",
       "  0.6794545650482178,\n",
       "  0.6794525980949402,\n",
       "  0.6794503927230835,\n",
       "  0.6794483661651611,\n",
       "  0.6794461011886597,\n",
       "  0.679443895816803,\n",
       "  0.6794416904449463,\n",
       "  0.6794397234916687,\n",
       "  0.6794374585151672,\n",
       "  0.6794354319572449,\n",
       "  0.6794333457946777,\n",
       "  0.679431140422821,\n",
       "  0.6794289946556091,\n",
       "  0.6794270277023315,\n",
       "  0.6794247627258301,\n",
       "  0.6794226169586182,\n",
       "  0.6794206500053406,\n",
       "  0.6794183254241943,\n",
       "  0.6794163584709167,\n",
       "  0.6794143915176392,\n",
       "  0.679412305355072,\n",
       "  0.6794101595878601,\n",
       "  0.6794079542160034,\n",
       "  0.679405927658081,\n",
       "  0.6794039011001587,\n",
       "  0.679401695728302,\n",
       "  0.6793997287750244,\n",
       "  0.6793975830078125,\n",
       "  0.6793956160545349,\n",
       "  0.679393470287323,\n",
       "  0.6793915629386902,\n",
       "  0.679389476776123,\n",
       "  0.6793873310089111,\n",
       "  0.6793854236602783,\n",
       "  0.679383397102356,\n",
       "  0.679381251335144,\n",
       "  0.6793793439865112,\n",
       "  0.6793774962425232,\n",
       "  0.679375410079956,\n",
       "  0.679373562335968,\n",
       "  0.6793716549873352,\n",
       "  0.6793699860572815,\n",
       "  0.6793680787086487,\n",
       "  0.6793661713600159,\n",
       "  0.6793643832206726,\n",
       "  0.6793625354766846,\n",
       "  0.6793608069419861,\n",
       "  0.6793588399887085,\n",
       "  0.6793566346168518,\n",
       "  0.6793548464775085,\n",
       "  0.6793529987335205,\n",
       "  0.6793512105941772,\n",
       "  0.6793493628501892,\n",
       "  0.6793473958969116,\n",
       "  0.6793454885482788,\n",
       "  0.6793434619903564,\n",
       "  0.6793416738510132,\n",
       "  0.6793398857116699,\n",
       "  0.6793379783630371,\n",
       "  0.6793361902236938,\n",
       "  0.6793345808982849,\n",
       "  0.6793325543403625,\n",
       "  0.6793306469917297,\n",
       "  0.6793287396430969,\n",
       "  0.6793267130851746,\n",
       "  0.6793251037597656,\n",
       "  0.6793233752250671,\n",
       "  0.6793214678764343,\n",
       "  0.6793196797370911,\n",
       "  0.6793180704116821,\n",
       "  0.6793163418769836,\n",
       "  0.6793146729469299,\n",
       "  0.6793131828308105,\n",
       "  0.6793116927146912,\n",
       "  0.679309606552124,\n",
       "  0.6793081760406494,\n",
       "  0.6793067455291748,\n",
       "  0.6793049573898315,\n",
       "  0.6793034076690674,\n",
       "  0.679301917552948,\n",
       "  0.6793001294136047,\n",
       "  0.679298460483551,\n",
       "  0.6792969703674316,\n",
       "  0.6792954802513123,\n",
       "  0.6792937517166138,\n",
       "  0.6792920827865601,\n",
       "  0.6792902946472168,\n",
       "  0.6792888045310974,\n",
       "  0.6792870163917542,\n",
       "  0.6792855858802795,\n",
       "  0.6792840361595154,\n",
       "  0.6792824268341064,\n",
       "  0.6792811751365662,\n",
       "  0.679279625415802,\n",
       "  0.6792781352996826],\n",
       " [tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6400),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6600),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500),\n",
       "  tensor(0.6500)],\n",
       " [tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100),\n",
       "  tensor(0.7100)])"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train(model, num_epochs, train_dl, x_valid, y_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Epochs')"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(16, 4))\n",
    "ax = fig.add_subplot(1, 2, 1)\n",
    "plt.plot(history[0], lw=4)\n",
    "plt.plot(history[1], lw=4)\n",
    "plt.legend(['Train loss', 'Validation loss'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 2, 2)\n",
    "plt.plot(history[2], lw=4)\n",
    "plt.plot(history[3], lw=4)\n",
    "plt.legend(['Train acc.', 'Validation acc.'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "#plt.savefig('figures/13_03.png', dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sequential(\n",
       "  (0): Linear(in_features=2, out_features=4, bias=True)\n",
       "  (1): ReLU()\n",
       "  (2): Linear(in_features=4, out_features=4, bias=True)\n",
       "  (3): ReLU()\n",
       "  (4): Linear(in_features=4, out_features=1, bias=True)\n",
       "  (5): Sigmoid()\n",
       ")"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = nn.Sequential(\n",
    "    nn.Linear(2, 4),\n",
    "    nn.ReLU(),\n",
    "    nn.Linear(4, 4),\n",
    "    nn.ReLU(),\n",
    "    nn.Linear(4, 1),\n",
    "    nn.Sigmoid()\n",
    ")\n",
    " \n",
    "loss_fn = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.015)\n",
    "\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "history = train(model, num_epochs, train_dl, x_valid, y_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 0, 'Epochs')"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(16, 4))\n",
    "ax = fig.add_subplot(1, 2, 1)\n",
    "plt.plot(history[0], lw=4)\n",
    "plt.plot(history[1], lw=4)\n",
    "plt.legend(['Train loss', 'Validation loss'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 2, 2)\n",
    "plt.plot(history[2], lw=4)\n",
    "plt.plot(history[3], lw=4)\n",
    "plt.legend(['Train acc.', 'Validation acc.'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "#plt.savefig('figures/13_04.png', dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Making model building more flexible with nn.Module\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "MyModule(\n",
       "  (module_list): ModuleList(\n",
       "    (0): Linear(in_features=2, out_features=4, bias=True)\n",
       "    (1): ReLU()\n",
       "    (2): Linear(in_features=4, out_features=4, bias=True)\n",
       "    (3): ReLU()\n",
       "    (4): Linear(in_features=4, out_features=1, bias=True)\n",
       "    (5): Sigmoid()\n",
       "  )\n",
       ")"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "class MyModule(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        l1 = nn.Linear(2, 4)\n",
    "        a1 = nn.ReLU()\n",
    "        l2 = nn.Linear(4, 4)\n",
    "        a2 = nn.ReLU()\n",
    "        l3 = nn.Linear(4, 1)\n",
    "        a3 = nn.Sigmoid()\n",
    "        l = [l1, a1, l2, a2, l3, a3]\n",
    "        self.module_list = nn.ModuleList(l)\n",
    "\n",
    "    def forward(self, x):\n",
    "        for f in self.module_list:\n",
    "            x = f(x)\n",
    "        return x\n",
    "    \n",
    "    def predict(self, x):\n",
    "        x = torch.tensor(x, dtype=torch.float32)\n",
    "        pred = self.forward(x)[:, 0]\n",
    "        return (pred>=0.5).float()\n",
    "            \n",
    "model = MyModule()\n",
    "model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_fn = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.015)\n",
    "    \n",
    "# torch.manual_seed(1)\n",
    "history = train(model, num_epochs, train_dl, x_valid, y_valid)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install mlxtend"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1600x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from mlxtend.plotting import plot_decision_regions\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize=(16, 4))\n",
    "ax = fig.add_subplot(1, 3, 1)\n",
    "plt.plot(history[0], lw=4)\n",
    "plt.plot(history[1], lw=4)\n",
    "plt.legend(['Train loss', 'Validation loss'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 3, 2)\n",
    "plt.plot(history[2], lw=4)\n",
    "plt.plot(history[3], lw=4)\n",
    "plt.legend(['Train acc.', 'Validation acc.'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 3, 3)\n",
    "plot_decision_regions(X=x_valid.numpy(), \n",
    "                      y=y_valid.numpy().astype(np.int64),\n",
    "                      clf=model)\n",
    "ax.set_xlabel(r'$x_1$', size=15)\n",
    "ax.xaxis.set_label_coords(1, -0.025)\n",
    "ax.set_ylabel(r'$x_2$', size=15)\n",
    "ax.yaxis.set_label_coords(-0.025, 1)\n",
    "\n",
    "#plt.savefig('figures/13_05.png', dpi=300)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Writing custom layers in PyTorch\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class NoisyLinear(nn.Module):\n",
    "    def __init__(self, input_size, output_size, noise_stddev=0.1):\n",
    "        super().__init__()\n",
    "        w = torch.Tensor(input_size, output_size)\n",
    "        self.w = nn.Parameter(w)  # nn.Parameter is a Tensor that's a module parameter.\n",
    "        nn.init.xavier_uniform_(self.w)\n",
    "        b = torch.Tensor(output_size).fill_(0)\n",
    "        self.b = nn.Parameter(b)\n",
    "        self.noise_stddev = noise_stddev\n",
    "\n",
    "    def forward(self, x, training=False):\n",
    "        if training:\n",
    "            noise = torch.normal(0.0, self.noise_stddev, x.shape)\n",
    "            x_new = torch.add(x, noise)\n",
    "        else:\n",
    "            x_new = x\n",
    "        return torch.add(torch.mm(x_new, self.w), self.b)   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "## testing:\n",
    "\n",
    "torch.manual_seed(1)\n",
    "\n",
    "noisy_layer = NoisyLinear(4, 2)\n",
    " \n",
    "x = torch.zeros((1, 4))\n",
    "print(noisy_layer(x, training=True))\n",
    "\n",
    "print(noisy_layer(x, training=True))\n",
    " \n",
    "print(noisy_layer(x, training=False))\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "class MyNoisyModule(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.l1 = NoisyLinear(2, 4, 0.07)\n",
    "        self.a1 = nn.ReLU()\n",
    "        self.l2 = nn.Linear(4, 4)\n",
    "        self.a2 = nn.ReLU()\n",
    "        self.l3 = nn.Linear(4, 1)\n",
    "        self.a3 = nn.Sigmoid()\n",
    "        \n",
    "    def forward(self, x, training=False):\n",
    "        x = self.l1(x, training)\n",
    "        x = self.a1(x)\n",
    "        x = self.l2(x)\n",
    "        x = self.a2(x)\n",
    "        x = self.l3(x)\n",
    "        x = self.a3(x)\n",
    "        return x\n",
    "    \n",
    "    def predict(self, x):\n",
    "        x = torch.tensor(x, dtype=torch.float32)\n",
    "        pred = self.forward(x)[:, 0]\n",
    "        return (pred>=0.5).float()\n",
    "\n",
    "torch.manual_seed(1)\n",
    "model = MyNoisyModule()\n",
    "model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_fn = nn.BCELoss()\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.015)\n",
    "    \n",
    "torch.manual_seed(1)\n",
    "\n",
    "loss_hist_train = [0] * num_epochs\n",
    "accuracy_hist_train = [0] * num_epochs\n",
    "loss_hist_valid = [0] * num_epochs\n",
    "accuracy_hist_valid = [0] * num_epochs\n",
    "for epoch in range(num_epochs):\n",
    "    for x_batch, y_batch in train_dl:\n",
    "        pred = model(x_batch, True)[:, 0]\n",
    "        loss = loss_fn(pred, y_batch)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        optimizer.zero_grad()\n",
    "        loss_hist_train[epoch] += loss.item()\n",
    "        is_correct = ((pred>=0.5).float() == y_batch).float()\n",
    "        accuracy_hist_train[epoch] += is_correct.mean()\n",
    "\n",
    "    loss_hist_train[epoch] /= n_train/batch_size\n",
    "    accuracy_hist_train[epoch] /= n_train/batch_size\n",
    "\n",
    "    pred = model(x_valid)[:, 0]\n",
    "    loss = loss_fn(pred, y_valid)\n",
    "    loss_hist_valid[epoch] = loss.item()\n",
    "    is_correct = ((pred>=0.5).float() == y_valid).float()\n",
    "    accuracy_hist_valid[epoch] += is_correct.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from mlxtend.plotting import plot_decision_regions\n",
    "\n",
    "\n",
    "fig = plt.figure(figsize=(16, 4))\n",
    "ax = fig.add_subplot(1, 3, 1)\n",
    "plt.plot(loss_hist_train, lw=4)\n",
    "plt.plot(loss_hist_valid, lw=4)\n",
    "plt.legend(['Train loss', 'Validation loss'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 3, 2)\n",
    "plt.plot(accuracy_hist_train, lw=4)\n",
    "plt.plot(accuracy_hist_valid, lw=4)\n",
    "plt.legend(['Train acc.', 'Validation acc.'], fontsize=15)\n",
    "ax.set_xlabel('Epochs', size=15)\n",
    "\n",
    "ax = fig.add_subplot(1, 3, 3)\n",
    "plot_decision_regions(X=x_valid.numpy(), \n",
    "                      y=y_valid.numpy().astype(np.int64),\n",
    "                      clf=model)\n",
    "ax.set_xlabel(r'$x_1$', size=15)\n",
    "ax.xaxis.set_label_coords(1, -0.025)\n",
    "ax.set_ylabel(r'$x_2$', size=15)\n",
    "ax.yaxis.set_label_coords(-0.025, 1)\n",
    "\n",
    "#plt.savefig('figures/13_06.png', dpi=300)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.12.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}