Notebook20 - Sparse Variational Dropout.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "etUt8t6gv8HK"
   },
   "source": [
    "# Sparse Variational Dropout\n",
    "\n",
    "- Variational Dropout Sparsifies Deep Neural Networks https://arxiv.org/abs/1701.05369\n",
    "- Cheating link https://github.com/ars-ashuha/sparse-vd-pytorch/blob/master/svdo-solution.ipynb"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "iiCiVLaJv8HV"
   },
   "source": [
    "# Implementation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "ICMEDWnov8HW"
   },
   "outputs": [],
   "source": [
    "import math\n",
    "import torch\n",
    "import numpy as np\n",
    "\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "\n",
    "from logger import Logger\n",
    "from torch.nn import Parameter\n",
    "from torchvision import datasets, transforms"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "KcEIF5_kv8HY"
   },
   "outputs": [],
   "source": [
    "class LinearSVDO(nn.Module):\n",
    "    def __init__(self, in_features, out_features, threshold, bias=True):\n",
    "        super(LinearSVDO, self).__init__()\n",
    "        self.in_features = in_features\n",
    "        self.out_features = out_features\n",
    "        self.threshold = threshold\n",
    "\n",
    "        self.W = Parameter(torch.Tensor(out_features, in_features))\n",
    "                ###########################################################\n",
    "        ########         You Code should be here         ##########\n",
    "        # Create a Parameter to store log sigma\n",
    "        self.log_sigma = Parameter(torch.Tensor(out_features, in_features))\n",
    "        self.bias = Parameter(torch.Tensor(1, out_features))\n",
    "        \n",
    "        self.reset_parameters()\n",
    "\n",
    "    def reset_parameters(self):\n",
    "        self.bias.data.zero_()\n",
    "        self.W.data.normal_(0, 0.02)\n",
    "        self.log_sigma.data.fill_(-5)        \n",
    "        \n",
    "    def forward(self, x):\n",
    "        self.log_alpha = self.log_sigma * 2.0 - torch.log(1e-16 + 2.0 * torch.abs(self.W))\n",
    "        self.log_alpha = torch.clamp(self.log_alpha, -10, 10) \n",
    "        \n",
    "        if self.training:\n",
    "            # Compute activation's mean e.g x.dot(W) + b\n",
    "            lrt_mean =  F.linear(x, self.W) + self.bias\n",
    "            # Compute activation's var e.g sqrt((x*x).dot(sigma * sigma) + 1e-8)\n",
    "            lrt_std = torch.sqrt(F.linear(x * x, torch.exp(self.log_sigma * 2.0)) + 1e-8)\n",
    "            eps = lrt_std.data.new(lrt_std.size()).normal_()\n",
    "            return lrt_mean + lrt_std * eps\n",
    "        # Prune out redundant wights e.g. W * mask(log_alpha < 3) \n",
    "        return F.linear(x, self.W * (self.log_alpha < 3).float()) + self.bias\n",
    "        \n",
    "    def kl_reg(self):\n",
    "        ###########################################################\n",
    "        ########         You Code should be here         ##########\n",
    "        ########  Eval Approximation of KL Divergence    ##########\n",
    "        # use torch.log1p for numerical stability\n",
    "        # Return KL here -- a scalar \n",
    "        k1, k2, k3 = torch.Tensor([0.63576]), torch.Tensor([1.8732]), torch.Tensor([1.48695])\n",
    "        kl = k1 * torch.sigmoid(k2 + k3 * self.log_alpha) - 0.5 * torch.log1p(torch.exp(-self.log_alpha))\n",
    "        a = - torch.sum(kl)\n",
    "        return a"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "7mGfHQ4Nv8Ha"
   },
   "outputs": [],
   "source": [
    "# Define a simple 2 layer Network\n",
    "class Net(nn.Module):\n",
    "    def __init__(self, threshold):\n",
    "        super(Net, self).__init__()\n",
    "        self.fc1 = LinearSVDO(28*28, 300, threshold)\n",
    "        self.fc2 = LinearSVDO(300,  10, threshold)\n",
    "        self.threshold = threshold\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = F.log_softmax(self.fc2(x), dim=1)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3fi-O-SFv8Hc"
   },
   "outputs": [],
   "source": [
    "# Load a dataset\n",
    "def get_mnist(batch_size):\n",
    "    trsnform = transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))])\n",
    "    train_loader = torch.utils.data.DataLoader(\n",
    "        datasets.MNIST('../data', train=True, download=True,\n",
    "        transform=trsnform), batch_size=batch_size, shuffle=True)\n",
    "    test_loader = torch.utils.data.DataLoader(\n",
    "        datasets.MNIST('../data', train=False, download=True,\n",
    "        transform=trsnform), batch_size=batch_size, shuffle=True)\n",
    "\n",
    "    return train_loader, test_loader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "4oGOpuEsv8He"
   },
   "outputs": [],
   "source": [
    "# Define New Loss Function -- SGVLB \n",
    "class SGVLB(nn.Module):\n",
    "    def __init__(self, net, train_size):\n",
    "        super(SGVLB, self).__init__()\n",
    "        self.train_size = train_size\n",
    "        self.net = net\n",
    "\n",
    "    def forward(self, input, target, kl_weight=1.0):\n",
    "        assert not target.requires_grad\n",
    "        kl = 0.0\n",
    "        for module in self.net.children():\n",
    "            if hasattr(module, 'kl_reg'):\n",
    "                kl = kl + module.kl_reg()\n",
    "        ###########################################################\n",
    "        ########         You Code should be here         ##########    \n",
    "        # Compute Stochastic Gradient Variational Lower Bound\n",
    "        # Do not forget to scale up Data term to N/M,\n",
    "        # where N is a size of the dataset and M is a size of minibatch\n",
    "        return F.cross_entropy(input, target) * self.train_size + kl_weight * kl"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "S7HkpvRVv8Hh"
   },
   "outputs": [],
   "source": [
    "model = Net(threshold=3)\n",
    "optimizer = optim.Adam(model.parameters(), lr=1e-3)\n",
    "scheduler = torch.optim.lr_scheduler.MultiStepLR(optimizer, milestones=[50,60,70,80], gamma=0.2)\n",
    "\n",
    "fmt = {'tr_los': '3.1e', 'te_loss': '3.1e', 'sp_0': '.3f', 'sp_1': '.3f', 'lr': '3.1e', 'kl': '.2f'}\n",
    "logger = Logger('sparse_vd', fmt=fmt)\n",
    "\n",
    "train_loader, test_loader = get_mnist(batch_size=100)\n",
    "sgvlb = SGVLB(model, len(train_loader.dataset))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 1730
    },
    "colab_type": "code",
    "id": "v6mf7WjhJIqA",
    "outputId": "20da9957-22f0-45d9-faa8-705ca4b58a20"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "  epoch    kl       lr    tr_los    tr_acc    te_loss    te_acc    sp_0    sp_1\n",
      "-------  ----  -------  --------  --------  ---------  --------  ------  ------\n",
      "      1  0.04  1.0e-03   4.1e+02      92.3    2.9e+02      96.2   0.001   0.000\n",
      "      2  0.06  1.0e-03   2.8e+02      96.9    1.7e+02      97.5   0.018   0.000\n",
      "      3  0.08  1.0e-03   1.4e+02      97.9    8.7e+01      97.9   0.074   0.000\n",
      "      4  0.10  1.0e-03   5.6e+01      98.2    3.5e+01      98.1   0.131   0.000\n",
      "      5  0.12  1.0e-03   7.1e+00      98.2   -2.5e+00      98.2   0.190   0.001\n",
      "      6  0.14  1.0e-03  -3.1e+01      98.2   -3.4e+01      98.2   0.245   0.002\n",
      "      7  0.16  1.0e-03  -5.9e+01      98.2   -5.8e+01      98.2   0.266   0.003\n",
      "      8  0.18  1.0e-03  -9.1e+01      98.1   -9.5e+01      98.3   0.326   0.001\n",
      "      9  0.20  1.0e-03  -1.2e+02      98.0   -1.2e+02      98.3   0.379   0.002\n",
      "     10  0.22  1.0e-03  -1.5e+02      98.0   -1.5e+02      98.2   0.365   0.002\n",
      "     11  0.24  1.0e-03  -1.8e+02      98.0   -1.8e+02      98.2   0.423   0.003\n",
      "     12  0.26  1.0e-03  -2.0e+02      97.9   -2.1e+02      98.2   0.451   0.001\n",
      "     13  0.28  1.0e-03  -2.3e+02      97.9   -2.4e+02      98.2   0.478   0.002\n",
      "     14  0.30  1.0e-03  -2.6e+02      97.9   -2.6e+02      98.2   0.475   0.006\n",
      "     15  0.32  1.0e-03  -2.9e+02      97.8   -2.9e+02      98.3   0.491   0.006\n",
      "     16  0.34  1.0e-03  -3.2e+02      97.9   -3.2e+02      98.2   0.508   0.004\n",
      "     17  0.36  1.0e-03  -3.5e+02      97.8   -3.5e+02      98.3   0.549   0.004\n",
      "     18  0.38  1.0e-03  -3.8e+02      97.8   -3.8e+02      98.2   0.577   0.005\n",
      "     19  0.40  1.0e-03  -4.1e+02      97.8   -4.2e+02      98.3   0.652   0.009\n",
      "     20  0.42  1.0e-03  -4.3e+02      97.7   -4.5e+02      98.2   0.659   0.009\n",
      "     21  0.44  1.0e-03  -4.7e+02      97.8   -4.7e+02      98.4   0.641   0.006\n",
      "     22  0.46  1.0e-03  -4.9e+02      97.7   -5.0e+02      98.3   0.637   0.011\n",
      "     23  0.48  1.0e-03  -5.2e+02      97.7   -5.3e+02      98.3   0.677   0.011\n",
      "     24  0.50  1.0e-03  -5.6e+02      97.7   -5.6e+02      98.3   0.655   0.010\n",
      "     25  0.52  1.0e-03  -5.9e+02      97.7   -6.0e+02      98.3   0.732   0.012\n",
      "     26  0.54  1.0e-03  -6.1e+02      97.6   -6.2e+02      98.3   0.738   0.015\n",
      "     27  0.56  1.0e-03  -6.5e+02      97.8   -6.6e+02      98.4   0.758   0.019\n",
      "     28  0.58  1.0e-03  -6.7e+02      97.6   -6.8e+02      98.4   0.749   0.014\n",
      "     29  0.60  1.0e-03  -7.0e+02      97.6   -7.1e+02      98.3   0.748   0.015\n",
      "     30  0.62  1.0e-03  -7.4e+02      97.7   -7.4e+02      98.2   0.740   0.014\n",
      "     31  0.64  1.0e-03  -7.7e+02      97.7   -7.8e+02      98.3   0.774   0.021\n",
      "     32  0.66  1.0e-03  -7.9e+02      97.6   -8.0e+02      98.3   0.766   0.014\n",
      "     33  0.68  1.0e-03  -8.2e+02      97.6   -8.3e+02      98.4   0.778   0.021\n",
      "     34  0.70  1.0e-03  -8.5e+02      97.6   -8.6e+02      98.5   0.787   0.018\n",
      "     35  0.72  1.0e-03  -8.9e+02      97.5   -9.0e+02      98.4   0.837   0.025\n",
      "     36  0.74  1.0e-03  -9.2e+02      97.6   -9.3e+02      98.3   0.817   0.029\n",
      "     37  0.76  1.0e-03  -9.4e+02      97.5   -9.6e+02      98.4   0.829   0.040\n",
      "     38  0.78  1.0e-03  -9.8e+02      97.5   -1.0e+03      98.4   0.858   0.039\n",
      "     39  0.80  1.0e-03  -1.0e+03      97.5   -1.0e+03      98.3   0.873   0.051\n",
      "     40  0.82  1.0e-03  -1.0e+03      97.5   -1.0e+03      98.3   0.829   0.036\n",
      "     41  0.84  1.0e-03  -1.1e+03      97.6   -1.1e+03      98.4   0.845   0.033\n",
      "     42  0.86  1.0e-03  -1.1e+03      97.5   -1.1e+03      98.3   0.846   0.047\n",
      "     43  0.88  1.0e-03  -1.1e+03      97.5   -1.1e+03      98.3   0.829   0.049\n",
      "     44  0.90  1.0e-03  -1.2e+03      97.6   -1.2e+03      98.3   0.854   0.046\n",
      "     45  0.92  1.0e-03  -1.2e+03      97.5   -1.2e+03      98.4   0.852   0.045\n",
      "     46  0.94  1.0e-03  -1.2e+03      97.5   -1.2e+03      98.4   0.863   0.062\n",
      "     47  0.96  1.0e-03  -1.3e+03      97.5   -1.3e+03      98.3   0.880   0.063\n",
      "     48  0.98  1.0e-03  -1.3e+03      97.5   -1.3e+03      98.3   0.872   0.074\n",
      "     49     1  1.0e-03  -1.3e+03      97.4   -1.3e+03      98.2   0.884   0.070\n",
      "     50     1  1.0e-03  -1.3e+03      97.5   -1.3e+03      98.3   0.887   0.085\n",
      "     51     1  2.0e-04  -1.3e+03      97.6   -1.4e+03      98.3   0.964   0.339\n",
      "     52     1  2.0e-04  -1.4e+03      97.6   -1.4e+03      98.3   0.967   0.370\n",
      "     53     1  2.0e-04  -1.4e+03      97.6   -1.4e+03      98.3   0.962   0.330\n",
      "     54     1  2.0e-04  -1.4e+03      97.7   -1.4e+03      98.4   0.963   0.363\n",
      "     55     1  2.0e-04  -1.4e+03      97.7   -1.4e+03      98.4   0.967   0.384\n",
      "     56     1  2.0e-04  -1.4e+03      97.7   -1.4e+03      98.3   0.968   0.406\n",
      "     57     1  2.0e-04  -1.4e+03      97.7   -1.4e+03      98.3   0.966   0.374\n",
      "     58     1  2.0e-04  -1.4e+03      97.8   -1.4e+03      98.3   0.965   0.387\n",
      "     59     1  2.0e-04  -1.4e+03      97.8   -1.4e+03      98.4   0.968   0.403\n",
      "     60     1  2.0e-04  -1.4e+03      97.8   -1.4e+03      98.4   0.965   0.387\n",
      "     61     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.695\n",
      "     62     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.692\n",
      "     63     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.692\n",
      "     64     1  4.0e-05  -1.4e+03      97.9   -1.4e+03      98.3   0.974   0.692\n",
      "     65     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.691\n",
      "     66     1  4.0e-05  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.697\n",
      "     67     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.707\n",
      "     68     1  4.0e-05  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.697\n",
      "     69     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.704\n",
      "     70     1  4.0e-05  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.701\n",
      "     71     1  8.0e-06  -1.4e+03      98.0   -1.4e+03      98.4   0.974   0.777\n",
      "     72     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.781\n",
      "     73     1  8.0e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.780\n",
      "     74     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.780\n",
      "     75     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.780\n",
      "     76     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.778\n",
      "     77     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.779\n",
      "     78     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.776\n",
      "     79     1  8.0e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.778\n",
      "     80     1  8.0e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.777\n",
      "     81     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     82     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     83     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     84     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     85     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     86     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     87     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     88     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     89     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     90     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     91     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     92     1  1.6e-06  -1.4e+03      97.8   -1.4e+03      98.4   0.974   0.784\n",
      "     93     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     94     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     95     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     96     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     97     1  1.6e-06  -1.4e+03      98.0   -1.4e+03      98.4   0.974   0.784\n",
      "     98     1  1.6e-06  -1.4e+03      97.9   -1.4e+03      98.4   0.974   0.784\n",
      "     99     1  1.6e-06  -1.4e+03      98.0   -1.4e+03      98.4   0.974   0.784\n",
      "    100     1  1.6e-06  -1.4e+03      98.0   -1.4e+03      98.4   0.974   0.784\n"
     ]
    }
   ],
   "source": [
    "kl_weight = 0.02\n",
    "epochs = 100\n",
    "\n",
    "for epoch in range(1, epochs + 1):\n",
    "    scheduler.step()\n",
    "    model.train()\n",
    "    train_loss, train_acc = 0, 0 \n",
    "    kl_weight = min(kl_weight+0.02, 1)\n",
    "    logger.add_scalar(epoch, 'kl', kl_weight)\n",
    "    logger.add_scalar(epoch, 'lr', scheduler.get_lr()[0])\n",
    "    for batch_idx, (data, target) in enumerate(train_loader):\n",
    "        data = data.view(-1, 28*28)\n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        output = model(data)\n",
    "        pred = output.data.max(1)[1] \n",
    "        loss = sgvlb(output, target, kl_weight)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        train_loss += loss \n",
    "        train_acc += np.sum(pred.numpy() == target.data.numpy())\n",
    "\n",
    "    logger.add_scalar(epoch, 'tr_los', train_loss / len(train_loader.dataset))\n",
    "    logger.add_scalar(epoch, 'tr_acc', train_acc / len(train_loader.dataset) * 100)\n",
    "    \n",
    "    \n",
    "    model.eval()\n",
    "    test_loss, test_acc = 0, 0\n",
    "    for batch_idx, (data, target) in enumerate(test_loader):\n",
    "        data = data.view(-1, 28*28)\n",
    "        output = model(data)\n",
    "        test_loss += float(sgvlb(output, target, kl_weight))\n",
    "        pred = output.data.max(1)[1] \n",
    "        test_acc += np.sum(pred.numpy() == target.data.numpy())\n",
    "        \n",
    "    logger.add_scalar(epoch, 'te_loss', test_loss / len(test_loader.dataset))\n",
    "    logger.add_scalar(epoch, 'te_acc', test_acc / len(test_loader.dataset) * 100)\n",
    "    \n",
    "    for i, c in enumerate(model.children()):\n",
    "        if hasattr(c, 'kl_reg'):\n",
    "            logger.add_scalar(epoch, 'sp_%s' % i, (c.log_alpha.data.numpy() > model.threshold).mean())\n",
    "            \n",
    "    logger.iter_info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 34
    },
    "colab_type": "code",
    "id": "UQVizuLov8Hu",
    "outputId": "2af8aae0-dae3-4a80-9c97-349316e19ad4"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "keept weight ratio = 35.64267544515936\n"
     ]
    }
   ],
   "source": [
    "all_w, kep_w = 0, 0\n",
    "\n",
    "for c in model.children():\n",
    "    kep_w += (c.log_alpha.data.numpy() < model.threshold).sum()\n",
    "    all_w += c.log_alpha.data.numpy().size\n",
    "\n",
    "print('keept weight ratio =', all_w/kep_w)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "FNMUWDjtJgd-"
   },
   "source": [
    "    # Good result should be like \n",
    "    #   epoch    kl       lr    tr_los    tr_acc    te_loss    te_acc    sp_0    sp_1\n",
    "    #  -------  ----  -------  --------  --------  ---------  --------  ------  ------\n",
    "    #      100     1  1.6e-06  -1.4e+03      98.0   -1.4e+03      98.3   0.969   0.760\n",
    "    # keept weight ratio = 30.109973454683352"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.colorbar.Colorbar at 0x1279dcbe0>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1277f94e0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "from matplotlib import rcParams\n",
    "rcParams['figure.figsize'] = 16, 3\n",
    "rcParams['figure.dpi'] = 300\n",
    "\n",
    "\n",
    "log_alpha = (model.fc1.log_alpha.detach().numpy() < 3).astype(np.float)\n",
    "W = model.fc1.W.detach().numpy()\n",
    "\n",
    "plt.imshow(log_alpha * W, cmap='hot', interpolation=None)\n",
    "plt.colorbar()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(-0.5, 279.5, 279.5, -0.5)"
      ]
     },
     "execution_count": 80,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x125ba6400>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "s = 0\n",
    "from matplotlib import rcParams\n",
    "rcParams['figure.figsize'] = 8, 5\n",
    "\n",
    "z = np.zeros((28*10, 28*10))\n",
    "\n",
    "for i in range(10):\n",
    "    for j in range(10):\n",
    "        s += 1\n",
    "        z[i*28:(i+1)*28, j*28:(j+1)*28] =  np.abs((log_alpha * W)[s].reshape(28, 28))\n",
    "        \n",
    "z[::28] = 2\n",
    "z[:, ::28] = 2\n",
    "z[-1] = 2\n",
    "z[:, -1] = 2\n",
    "        \n",
    "plt.imshow(z, cmap='hot_r')\n",
    "plt.colorbar()\n",
    "plt.axis('off')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.],\n",
       "       [0., 0., 0., ..., 0., 0., 0.]])"
      ]
     },
     "execution_count": 78,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "z[::28]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Compression with Sparse Matrixes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "3wnL7Hp9v8Hy"
   },
   "outputs": [],
   "source": [
    "import scipy\n",
    "import numpy as np\n",
    "from scipy.sparse import csc_matrix, csc_matrix, coo_matrix, dok_matrix\n",
    "\n",
    "row, col, data = [], [], []\n",
    "M = list(model.children())[0].W.data.numpy()\n",
    "LA = list(model.children())[0].log_alpha.data.numpy()\n",
    "\n",
    "for i in range(300):\n",
    "    for j in range(28*28):\n",
    "        if LA[i, j] < 3:\n",
    "            row += [i]\n",
    "            col += [j]\n",
    "            data += [M[i, j]]\n",
    "\n",
    "Mcsr = csc_matrix((data, (row, col)), shape=(300, 28*28))\n",
    "Mcsc = csc_matrix((data, (row, col)), shape=(300, 28*28))\n",
    "Mcoo = coo_matrix((data, (row, col)), shape=(300, 28*28))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "e4T8E4Miv8H0"
   },
   "outputs": [],
   "source": [
    "np.savez_compressed('M_w', M)\n",
    "scipy.sparse.save_npz('Mcsr_w', Mcsr)\n",
    "scipy.sparse.save_npz('Mcsc_w', Mcsc)\n",
    "scipy.sparse.save_npz('Mcoo_w', Mcoo)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "colab": {
     "base_uri": "https://localhost:8080/",
     "height": 84
    },
    "colab_type": "code",
    "id": "QebP7_Emv8H2",
    "outputId": "c9c70b01-9c69-40d8-879e-fd93bd4cda7c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-rw-r--r--   1 ars  staff   856K Aug 24 18:50 M_w.npz\n",
      "-rw-r--r--   1 ars  staff    30K Aug 24 18:50 Mcoo_w.npz\n",
      "-rw-r--r--   1 ars  staff    30K Aug 24 18:50 Mcsc_w.npz\n",
      "-rw-r--r--   1 ars  staff    30K Aug 24 18:50 Mcsr_w.npz\n"
     ]
    }
   ],
   "source": [
    "ls -lah | grep _w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "3. Sparse Variational Dropout.ipynb",
   "provenance": [],
   "version": "0.3.2"
  },
  "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": "3.9.5"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {},
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}