experiment_alex_cifar10.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:59:36.847663",
     "start_time": "2016-08-30T17:59:36.371472"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# %load_ext autoreload\n",
    "# %autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:59:37.224813",
     "start_time": "2016-08-30T17:59:36.851604"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from branchynet.net import BranchyNet\n",
    "from branchynet.links import *\n",
    "import chainer.functions as F\n",
    "import chainer.links as L\n",
    "from branchynet import utils, visualize\n",
    "from chainer import cuda"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Define Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-09-01T10:38:51.206772",
     "start_time": "2016-09-01T10:38:50.827907"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from networks import alex_cifar10\n",
    "\n",
    "branchyNet = alex_cifar10.get_network()\n",
    "if cuda.available:\n",
    "    branchyNet.to_gpu()\n",
    "branchyNet.training()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-09-01T10:38:53.317517",
     "start_time": "2016-09-01T10:38:52.619292"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from datasets import pcifar10\n",
    "\n",
    "x_train,y_train,x_test,y_test = pcifar10.get_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:04:36.762787",
     "start_time": "2016-08-30T17:04:36.743275"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "TRAIN_BATCHSIZE = 512\n",
    "TEST_BATCHSIZE = 1\n",
    "TRAIN_NUM_EPOCHS = 50"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train Main Network"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:06:01.258431",
     "start_time": "2016-08-30T17:04:36.830718"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "main_loss, main_acc, main_time = utils.train(branchyNet, x_train, y_train, main=True, batchsize=TRAIN_BATCHSIZE,\n",
    "                                             num_epoch=TRAIN_NUM_EPOCHS)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Train BranchyNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:11:06.345265",
     "start_time": "2016-08-30T17:06:01.268848"
    },
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "TRAIN_NUM_EPOCHS = 100\n",
    "branch_loss, branch_acc, branch_time = utils.train(branchyNet, x_train, y_train, batchsize=TRAIN_BATCHSIZE,\n",
    "                                             num_epoch=TRAIN_NUM_EPOCHS)\n",
    "\n",
    "#set network to inference mode\n",
    "branchyNet.testing()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Visualizing Network Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:11:06.835335",
     "start_time": "2016-08-30T17:11:06.348432"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAu8AAAJyCAYAAAB0ey7MAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nkiSpThjeJUmSpDpheJckSZLqxP8Hm7xMxzIXUiAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa1c7550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAu8AAAJyCAYAAAB0ey7MAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nktQnDO+SJElSnzC8S5IkSX3i/wPgwhdQOh/4sAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa1859e8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualize.plot_layers(main_loss, xlabel='Epochs', ylabel='Training Loss')\n",
    "visualize.plot_layers(main_acc, xlabel='Epochs', ylabel='Training Accuracy')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:11:07.341112",
     "start_time": "2016-08-30T17:11:06.837397"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvQAAAJyCAYAAACrNVTTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xae8a630>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAvQAAAJyCAYAAACrNVTTAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nEZEYpoReRERERCSG/X8raDysfn9mYwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xae8a5f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualize.plot_layers(zip(*branch_loss), xlabel='Epochs', ylabel='Training Loss')\n",
    "visualize.plot_layers(zip(*branch_acc), xlabel='Epochs', ylabel='Training Accuracy')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Run test suite and visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:12:25.444980",
     "start_time": "2016-08-30T17:11:24.248544"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#set network to inference mode\n",
    "branchyNet.testing()\n",
    "branchyNet.verbose = False\n",
    "if cuda.available:\n",
    "    branchyNet.to_gpu()\n",
    "g_baseacc, g_basediff, _, _ = utils.test(branchyNet,x_test,y_test,main=True,batchsize=TEST_BATCHSIZE)\n",
    "g_basediff = (g_basediff / float(len(y_test))) * 1000.\n",
    "\n",
    "#branchyNet.to_cpu()\n",
    "c_baseacc, c_basediff, _, _ = utils.test(branchyNet,x_test,y_test,main=True,batchsize=TEST_BATCHSIZE)\n",
    "c_basediff = (c_basediff / float(len(y_test))) * 1000."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:46:39.620682",
     "start_time": "2016-08-30T17:46:39.589059"
    },
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Specify thresholds\n",
    "thresholds = [0.0001, 0.001, 0.005, 0.01, 0.05, 0.1, 0.5, 0.75, 1., 5., 10.]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:33:15.836122",
     "start_time": "2016-08-30T17:21:55.266512"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#GPU\n",
    "if cuda.available:\n",
    "    branchyNet.to_gpu()\n",
    "g_ts, g_accs, g_diffs, g_exits = utils.screen_branchy(branchyNet, x_test, y_test, thresholds,\n",
    "                                                    batchsize=TEST_BATCHSIZE, verbose=True)\n",
    "\n",
    "#convert to ms\n",
    "g_diffs *= 1000."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Kung Lab\\Miniconda2\\lib\\site-packages\\matplotlib\\__init__.py:872: UserWarning: text.fontsize is deprecated and replaced with font.size; please use the latter.\n",
      "  warnings.warn(self.msg_depr % (key, alt_key))\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT8AAAEOCAYAAAD/gBCjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xacd2ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualize.plot_line_tradeoff(g_accs, g_diffs, g_ts, g_exits, g_baseacc, g_basediff, all_samples=False, inc_amt=-0.0001000,\n",
    "                             our_label='BranchyAlexNet', orig_label='AlexNet', xlabel='Runtime (ms)', \n",
    "                             title='AlexNet GPU', output_path='_figs/alexnet_gpu.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2016-08-30T17:46:57.326Z"
    },
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#CPU\n",
    "branchyNet.to_cpu()\n",
    "c_ts, c_accs, c_diffs, c_exits  = utils.screen_branchy(branchyNet, x_test, y_test, thresholds,\n",
    "                                                     batchsize=TEST_BATCHSIZE, verbose=True)\n",
    "\n",
    "#convert to ms\n",
    "c_diffs *= 1000."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:35:52.802997",
     "start_time": "2016-08-30T17:35:52.193698"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAT8AAAEOCAYAAAD/gBCjAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0xa02abe0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "visualize.plot_line_tradeoff(c_accs, c_diffs, c_ts, c_exits, c_baseacc, c_basediff, all_samples=False, inc_amt=-0.0001000,\n",
    "                             our_label='BranchyAlexNet', orig_label='AlexNet', xlabel='Runtime (ms)',\n",
    "                             title='AlexNet CPU', output_path='_figs/alexnet_cpu.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2016-08-30T17:11:07.655674",
     "start_time": "2016-08-30T17:04:45.607Z"
    },
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU Results:\n",
      "        Network        Acc.(%)       Time(ms)           Gain      Thrshld.T        Exit(%)\n",
      "        AlexNet          77.76           5.26              -              -              -\n",
      "      B-AlexNet          78.94           4.55           1.16{0.0001, 0.0001}{43.61, 17.41, 38.98}\n",
      "      B-AlexNet          79.16           4.48           1.17{0.0001, 0.001}{43.61, 23.21, 33.18}\n",
      "      B-AlexNet          79.27           4.39           1.20{0.0001, 0.005}{43.61, 27.82, 28.57}\n",
      "      B-AlexNet          79.37           4.46           1.18 {0.0001, 0.01}{43.61, 29.96, 26.43}\n",
      "      B-AlexNet          79.36           4.11           1.28 {0.0001, 0.05}{43.61, 35.79, 20.6}\n",
      "      B-AlexNet          79.18           4.03           1.30  {0.0001, 0.1}{43.61, 38.98, 17.41}\n",
      "      B-AlexNet          78.38           3.83           1.37  {0.0001, 0.5}{43.61, 48.45, 7.94}\n",
      "      B-AlexNet          77.53           3.67           1.43 {0.0001, 0.75}{43.61, 54.43, 1.96}\n",
      "      B-AlexNet          77.33           3.67           1.43  {0.0001, 1.0}{43.61, 55.83, 0.56}\n",
      "      B-AlexNet          77.30           3.65           1.44  {0.0001, 5.0}{43.61, 56.39, 0.0}\n",
      "      B-AlexNet          77.30           3.68           1.43 {0.0001, 10.0}{43.61, 56.39, 0.0}\n",
      "      B-AlexNet          79.12           4.36           1.21{0.001, 0.0001}{52.62, 13.1, 34.28}\n",
      "      B-AlexNet          79.35           4.29           1.22 {0.001, 0.001}{52.62, 17.96, 29.42}\n",
      "      B-AlexNet          79.42           4.29           1.23 {0.001, 0.005}{52.62, 21.88, 25.5}\n",
      "      B-AlexNet          79.54           4.08           1.29  {0.001, 0.01}{52.62, 23.76, 23.62}\n",
      "      B-AlexNet          79.53           3.95           1.33  {0.001, 0.05}{52.62, 28.86, 18.52}\n",
      "      B-AlexNet          79.39           3.93           1.34   {0.001, 0.1}{52.62, 31.75, 15.63}\n",
      "      B-AlexNet          78.66           3.66           1.44   {0.001, 0.5}{52.62, 40.25, 7.13}\n",
      "      B-AlexNet          77.90           3.61           1.46  {0.001, 0.75}{52.62, 45.58, 1.8}\n",
      "      B-AlexNet          77.70           3.54           1.49   {0.001, 1.0}{52.62, 46.85, 0.53}\n",
      "      B-AlexNet          77.66           3.54           1.49   {0.001, 5.0}{52.62, 47.38, 0.0}\n",
      "      B-AlexNet          77.66           3.48           1.51  {0.001, 10.0}{52.62, 47.38, 0.0}\n",
      "      B-AlexNet          79.17           4.11           1.28{0.005, 0.0001}{59.94, 9.93, 30.13}\n",
      "      B-AlexNet          79.38           4.01           1.31 {0.005, 0.001}{59.94, 14.08, 25.98}\n",
      "      B-AlexNet          79.47           4.02           1.31 {0.005, 0.005}{59.94, 17.41, 22.65}\n",
      "      B-AlexNet          79.56           3.82           1.38  {0.005, 0.01}{59.94, 19.05, 21.01}\n",
      "      B-AlexNet          79.64           3.75           1.40  {0.005, 0.05}{59.94, 23.56, 16.5}\n",
      "      B-AlexNet          79.53           3.74           1.41   {0.005, 0.1}{59.94, 26.03, 14.03}\n",
      "      B-AlexNet          78.88           3.49           1.51   {0.005, 0.5}{59.94, 33.54, 6.52}\n",
      "      B-AlexNet          78.28           3.48           1.51  {0.005, 0.75}{59.94, 38.36, 1.7}\n",
      "      B-AlexNet          78.09           3.46           1.52   {0.005, 1.0}{59.94, 39.54, 0.52}\n",
      "      B-AlexNet          78.05           3.40           1.55   {0.005, 5.0}{59.94, 40.06, 0.0}\n",
      "      B-AlexNet          78.05           3.35           1.57  {0.005, 10.0}{59.94, 40.06, 0.0}\n",
      "      B-AlexNet          79.14           4.02           1.31 {0.01, 0.0001}{63.42, 8.67, 27.91}\n",
      "      B-AlexNet          79.35           3.95           1.33  {0.01, 0.001}{63.42, 12.57, 24.01}\n",
      "      B-AlexNet          79.44           3.99           1.32  {0.01, 0.005}{63.42, 15.59, 20.99}\n",
      "      B-AlexNet          79.54           3.86           1.36   {0.01, 0.01}{63.42, 17.14, 19.44}\n",
      "      B-AlexNet          79.66           3.63           1.45   {0.01, 0.05}{63.42, 21.36, 15.22}\n",
      "      B-AlexNet          79.54           3.69           1.43    {0.01, 0.1}{63.42, 23.64, 12.94}\n",
      "      B-AlexNet          78.91           3.42           1.54    {0.01, 0.5}{63.42, 30.55, 6.03}\n",
      "      B-AlexNet          78.40           3.39           1.55   {0.01, 0.75}{63.42, 34.97, 1.61}\n",
      "      B-AlexNet          78.23           3.42           1.54    {0.01, 1.0}{63.42, 36.08, 0.5}\n",
      "      B-AlexNet          78.20           3.47           1.51    {0.01, 5.0}{63.42, 36.58, 0.0}\n",
      "      B-AlexNet          78.20           3.36           1.57   {0.01, 10.0}{63.42, 36.58, 0.0}\n",
      "      B-AlexNet          78.92           3.67           1.43 {0.05, 0.0001}{72.19, 6.08, 21.73}\n",
      "      B-AlexNet          79.05           3.65           1.44  {0.05, 0.001}{72.19, 9.01, 18.8}\n",
      "      B-AlexNet          79.15           3.52           1.49  {0.05, 0.005}{72.19, 11.44, 16.37}\n",
      "      B-AlexNet          79.28           3.55           1.48   {0.05, 0.01}{72.19, 12.56, 15.25}\n",
      "      B-AlexNet          79.28           3.56           1.48   {0.05, 0.05}{72.19, 15.73, 12.08}\n",
      "      B-AlexNet          79.23           3.56           1.48    {0.05, 0.1}{72.19, 17.52, 10.29}\n",
      "      B-AlexNet          78.68           3.26           1.61    {0.05, 0.5}{72.19, 23.04, 4.77}\n",
      "      B-AlexNet          78.38           3.27           1.61   {0.05, 0.75}{72.19, 26.5, 1.31}\n",
      "      B-AlexNet          78.26           3.31           1.59    {0.05, 1.0}{72.19, 27.38, 0.43}\n",
      "      B-AlexNet          78.25           3.23           1.63    {0.05, 5.0}{72.19, 27.81, 0.0}\n",
      "      B-AlexNet          78.25           3.32           1.58   {0.05, 10.0}{72.19, 27.81, 0.0}\n",
      "      B-AlexNet          78.57           3.62           1.45  {0.1, 0.0001}{76.54, 4.86, 18.6}\n",
      "      B-AlexNet          78.70           3.63           1.45   {0.1, 0.001}{76.54, 7.34, 16.12}\n",
      "      B-AlexNet          78.77           3.43           1.53   {0.1, 0.005}{76.54, 9.32, 14.14}\n",
      "      B-AlexNet          78.86           3.49           1.51    {0.1, 0.01}{76.54, 10.22, 13.24}\n",
      "      B-AlexNet          78.87           3.51           1.50    {0.1, 0.05}{76.54, 12.98, 10.48}\n",
      "      B-AlexNet          78.85           3.35           1.57     {0.1, 0.1}{76.54, 14.55, 8.91}\n",
      "      B-AlexNet          78.35           3.21           1.64     {0.1, 0.5}{76.54, 19.34, 4.12}\n",
      "      B-AlexNet          78.13           3.22           1.63    {0.1, 0.75}{76.54, 22.35, 1.11}\n",
      "      B-AlexNet          78.04           3.14           1.67     {0.1, 1.0}{76.54, 23.07, 0.39}\n",
      "      B-AlexNet          78.04           3.19           1.65     {0.1, 5.0}{76.54, 23.46, 0.0}\n",
      "      B-AlexNet          78.04           3.18           1.66    {0.1, 10.0}{76.54, 23.46, 0.0}\n",
      "      B-AlexNet          77.08           3.04           1.73  {0.5, 0.0001}{89.61, 1.85, 8.54}\n",
      "      B-AlexNet          77.15           3.06           1.72   {0.5, 0.001}{89.61, 3.06, 7.33}\n",
      "      B-AlexNet          77.21           3.19           1.65   {0.5, 0.005}{89.61, 3.9, 6.49}\n",
      "      B-AlexNet          77.26           3.14           1.68    {0.5, 0.01}{89.61, 4.27, 6.12}\n",
      "      B-AlexNet          77.26           3.19           1.65    {0.5, 0.05}{89.61, 5.44, 4.95}\n",
      "      B-AlexNet          77.26           3.05           1.72     {0.5, 0.1}{89.61, 6.11, 4.28}\n",
      "      B-AlexNet          77.03           2.88           1.83     {0.5, 0.5}{89.61, 8.26, 2.13}\n",
      "      B-AlexNet          76.83           2.94           1.79    {0.5, 0.75}{89.61, 9.82, 0.57}\n",
      "      B-AlexNet          76.78           2.88           1.82     {0.5, 1.0}{89.61, 10.21, 0.18}\n",
      "      B-AlexNet          76.78           2.93           1.80     {0.5, 5.0}{89.61, 10.39, 0.0}\n",
      "      B-AlexNet          76.78           2.86           1.84    {0.5, 10.0}{89.61, 10.39, 0.0}\n",
      "      B-AlexNet          75.99           2.66           1.98 {0.75, 0.0001}{96.76, 0.37, 2.87}\n",
      "      B-AlexNet          75.96           2.83           1.86  {0.75, 0.001}{96.76, 0.79, 2.45}\n",
      "      B-AlexNet          75.99           2.91           1.81  {0.75, 0.005}{96.76, 1.1, 2.14}\n",
      "      B-AlexNet          76.00           2.88           1.83   {0.75, 0.01}{96.76, 1.2, 2.04}\n",
      "      B-AlexNet          75.97           3.03           1.73   {0.75, 0.05}{96.76, 1.53, 1.71}\n",
      "      B-AlexNet          75.99           2.84           1.85    {0.75, 0.1}{96.76, 1.78, 1.46}\n",
      "      B-AlexNet          75.94           2.83           1.85    {0.75, 0.5}{96.76, 2.48, 0.76}\n",
      "      B-AlexNet          75.93           2.94           1.79   {0.75, 0.75}{96.76, 2.93, 0.31}\n",
      "      B-AlexNet          75.90           2.93           1.79    {0.75, 1.0}{96.76, 3.16, 0.08}\n",
      "      B-AlexNet          75.93           2.90           1.81    {0.75, 5.0}{96.76, 3.24, 0.0}\n",
      "      B-AlexNet          75.93           2.92           1.80   {0.75, 10.0}{96.76, 3.24, 0.0}\n",
      "      B-AlexNet          75.54           2.75           1.91  {1.0, 0.0001}{98.97, 0.09, 0.94}\n",
      "      B-AlexNet          75.52           2.95           1.78   {1.0, 0.001}{98.97, 0.19, 0.84}\n",
      "      B-AlexNet          75.54           2.77           1.90   {1.0, 0.005}{98.97, 0.29, 0.74}\n",
      "      B-AlexNet          75.54           2.69           1.95    {1.0, 0.01}{98.97, 0.33, 0.7}\n",
      "      B-AlexNet          75.52           3.01           1.75    {1.0, 0.05}{98.97, 0.52, 0.51}\n",
      "      B-AlexNet          75.51           2.89           1.82     {1.0, 0.1}{98.97, 0.61, 0.42}\n",
      "      B-AlexNet          75.51           2.74           1.92     {1.0, 0.5}{98.97, 0.78, 0.25}\n",
      "      B-AlexNet          75.50           2.78           1.89    {1.0, 0.75}{98.97, 0.92, 0.11}\n",
      "      B-AlexNet          75.49           2.80           1.88     {1.0, 1.0}{98.97, 0.99, 0.04}\n",
      "      B-AlexNet          75.51           2.79           1.88     {1.0, 5.0}{98.97, 1.03, 0.0}\n",
      "      B-AlexNet          75.51           2.87           1.83    {1.0, 10.0}{98.97, 1.03, 0.0}\n",
      "      B-AlexNet          75.29           2.58           2.04  {5.0, 0.0001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.85   {5.0, 0.001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.74           1.92   {5.0, 0.005}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.85    {5.0, 0.01}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.85    {5.0, 0.05}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.77           1.90     {5.0, 0.1}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.73           1.92     {5.0, 0.5}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.76           1.90    {5.0, 0.75}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.66           1.98     {5.0, 1.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.86           1.84     {5.0, 5.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.72           1.93    {5.0, 10.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.79           1.88 {10.0, 0.0001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.88           1.83  {10.0, 0.001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.86           1.84  {10.0, 0.005}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.75           1.91   {10.0, 0.01}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.80           1.88   {10.0, 0.05}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.76           1.91    {10.0, 0.1}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.71           1.94    {10.0, 0.5}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.68           1.96   {10.0, 0.75}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.80           1.88    {10.0, 1.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.89           1.82    {10.0, 5.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.82           1.86   {10.0, 10.0}{100.0, 0.0, 0.0}\n",
      "\n",
      "\n",
      "\n",
      "GPU Results:\n",
      "        Network        Acc.(%)       Time(ms)           Gain      Thrshld.T        Exit(%)\n",
      "        AlexNet          77.76           5.30              -              -              -\n",
      "      B-AlexNet          78.94           4.55           1.17{0.0001, 0.0001}{43.61, 17.41, 38.98}\n",
      "      B-AlexNet          79.16           4.48           1.18{0.0001, 0.001}{43.61, 23.21, 33.18}\n",
      "      B-AlexNet          79.27           4.39           1.21{0.0001, 0.005}{43.61, 27.82, 28.57}\n",
      "      B-AlexNet          79.37           4.46           1.19 {0.0001, 0.01}{43.61, 29.96, 26.43}\n",
      "      B-AlexNet          79.36           4.11           1.29 {0.0001, 0.05}{43.61, 35.79, 20.6}\n",
      "      B-AlexNet          79.18           4.03           1.31  {0.0001, 0.1}{43.61, 38.98, 17.41}\n",
      "      B-AlexNet          78.38           3.83           1.38  {0.0001, 0.5}{43.61, 48.45, 7.94}\n",
      "      B-AlexNet          77.53           3.67           1.44 {0.0001, 0.75}{43.61, 54.43, 1.96}\n",
      "      B-AlexNet          77.33           3.67           1.45  {0.0001, 1.0}{43.61, 55.83, 0.56}\n",
      "      B-AlexNet          77.30           3.65           1.45  {0.0001, 5.0}{43.61, 56.39, 0.0}\n",
      "      B-AlexNet          77.30           3.68           1.44 {0.0001, 10.0}{43.61, 56.39, 0.0}\n",
      "      B-AlexNet          79.12           4.36           1.21{0.001, 0.0001}{52.62, 13.1, 34.28}\n",
      "      B-AlexNet          79.35           4.29           1.23 {0.001, 0.001}{52.62, 17.96, 29.42}\n",
      "      B-AlexNet          79.42           4.29           1.23 {0.001, 0.005}{52.62, 21.88, 25.5}\n",
      "      B-AlexNet          79.54           4.08           1.30  {0.001, 0.01}{52.62, 23.76, 23.62}\n",
      "      B-AlexNet          79.53           3.95           1.34  {0.001, 0.05}{52.62, 28.86, 18.52}\n",
      "      B-AlexNet          79.39           3.93           1.35   {0.001, 0.1}{52.62, 31.75, 15.63}\n",
      "      B-AlexNet          78.66           3.66           1.45   {0.001, 0.5}{52.62, 40.25, 7.13}\n",
      "      B-AlexNet          77.90           3.61           1.47  {0.001, 0.75}{52.62, 45.58, 1.8}\n",
      "      B-AlexNet          77.70           3.54           1.50   {0.001, 1.0}{52.62, 46.85, 0.53}\n",
      "      B-AlexNet          77.66           3.54           1.50   {0.001, 5.0}{52.62, 47.38, 0.0}\n",
      "      B-AlexNet          77.66           3.48           1.52  {0.001, 10.0}{52.62, 47.38, 0.0}\n",
      "      B-AlexNet          79.17           4.11           1.29{0.005, 0.0001}{59.94, 9.93, 30.13}\n",
      "      B-AlexNet          79.38           4.01           1.32 {0.005, 0.001}{59.94, 14.08, 25.98}\n",
      "      B-AlexNet          79.47           4.02           1.32 {0.005, 0.005}{59.94, 17.41, 22.65}\n",
      "      B-AlexNet          79.56           3.82           1.39  {0.005, 0.01}{59.94, 19.05, 21.01}\n",
      "      B-AlexNet          79.64           3.75           1.41  {0.005, 0.05}{59.94, 23.56, 16.5}\n",
      "      B-AlexNet          79.53           3.74           1.42   {0.005, 0.1}{59.94, 26.03, 14.03}\n",
      "      B-AlexNet          78.88           3.49           1.52   {0.005, 0.5}{59.94, 33.54, 6.52}\n",
      "      B-AlexNet          78.28           3.48           1.52  {0.005, 0.75}{59.94, 38.36, 1.7}\n",
      "      B-AlexNet          78.09           3.46           1.53   {0.005, 1.0}{59.94, 39.54, 0.52}\n",
      "      B-AlexNet          78.05           3.40           1.56   {0.005, 5.0}{59.94, 40.06, 0.0}\n",
      "      B-AlexNet          78.05           3.35           1.58  {0.005, 10.0}{59.94, 40.06, 0.0}\n",
      "      B-AlexNet          79.14           4.02           1.32 {0.01, 0.0001}{63.42, 8.67, 27.91}\n",
      "      B-AlexNet          79.35           3.95           1.34  {0.01, 0.001}{63.42, 12.57, 24.01}\n",
      "      B-AlexNet          79.44           3.99           1.33  {0.01, 0.005}{63.42, 15.59, 20.99}\n",
      "      B-AlexNet          79.54           3.86           1.37   {0.01, 0.01}{63.42, 17.14, 19.44}\n",
      "      B-AlexNet          79.66           3.63           1.46   {0.01, 0.05}{63.42, 21.36, 15.22}\n",
      "      B-AlexNet          79.54           3.69           1.44    {0.01, 0.1}{63.42, 23.64, 12.94}\n",
      "      B-AlexNet          78.91           3.42           1.55    {0.01, 0.5}{63.42, 30.55, 6.03}\n",
      "      B-AlexNet          78.40           3.39           1.56   {0.01, 0.75}{63.42, 34.97, 1.61}\n",
      "      B-AlexNet          78.23           3.42           1.55    {0.01, 1.0}{63.42, 36.08, 0.5}\n",
      "      B-AlexNet          78.20           3.47           1.52    {0.01, 5.0}{63.42, 36.58, 0.0}\n",
      "      B-AlexNet          78.20           3.36           1.58   {0.01, 10.0}{63.42, 36.58, 0.0}\n",
      "      B-AlexNet          78.92           3.67           1.44 {0.05, 0.0001}{72.19, 6.08, 21.73}\n",
      "      B-AlexNet          79.05           3.65           1.45  {0.05, 0.001}{72.19, 9.01, 18.8}\n",
      "      B-AlexNet          79.15           3.52           1.50  {0.05, 0.005}{72.19, 11.44, 16.37}\n",
      "      B-AlexNet          79.28           3.55           1.49   {0.05, 0.01}{72.19, 12.56, 15.25}\n",
      "      B-AlexNet          79.28           3.56           1.49   {0.05, 0.05}{72.19, 15.73, 12.08}\n",
      "      B-AlexNet          79.23           3.56           1.49    {0.05, 0.1}{72.19, 17.52, 10.29}\n",
      "      B-AlexNet          78.68           3.26           1.62    {0.05, 0.5}{72.19, 23.04, 4.77}\n",
      "      B-AlexNet          78.38           3.27           1.62   {0.05, 0.75}{72.19, 26.5, 1.31}\n",
      "      B-AlexNet          78.26           3.31           1.60    {0.05, 1.0}{72.19, 27.38, 0.43}\n",
      "      B-AlexNet          78.25           3.23           1.64    {0.05, 5.0}{72.19, 27.81, 0.0}\n",
      "      B-AlexNet          78.25           3.32           1.60   {0.05, 10.0}{72.19, 27.81, 0.0}\n",
      "      B-AlexNet          78.57           3.62           1.46  {0.1, 0.0001}{76.54, 4.86, 18.6}\n",
      "      B-AlexNet          78.70           3.63           1.46   {0.1, 0.001}{76.54, 7.34, 16.12}\n",
      "      B-AlexNet          78.77           3.43           1.54   {0.1, 0.005}{76.54, 9.32, 14.14}\n",
      "      B-AlexNet          78.86           3.49           1.52    {0.1, 0.01}{76.54, 10.22, 13.24}\n",
      "      B-AlexNet          78.87           3.51           1.51    {0.1, 0.05}{76.54, 12.98, 10.48}\n",
      "      B-AlexNet          78.85           3.35           1.58     {0.1, 0.1}{76.54, 14.55, 8.91}\n",
      "      B-AlexNet          78.35           3.21           1.65     {0.1, 0.5}{76.54, 19.34, 4.12}\n",
      "      B-AlexNet          78.13           3.22           1.65    {0.1, 0.75}{76.54, 22.35, 1.11}\n",
      "      B-AlexNet          78.04           3.14           1.69     {0.1, 1.0}{76.54, 23.07, 0.39}\n",
      "      B-AlexNet          78.04           3.19           1.66     {0.1, 5.0}{76.54, 23.46, 0.0}\n",
      "      B-AlexNet          78.04           3.18           1.67    {0.1, 10.0}{76.54, 23.46, 0.0}\n",
      "      B-AlexNet          77.08           3.04           1.74  {0.5, 0.0001}{89.61, 1.85, 8.54}\n",
      "      B-AlexNet          77.15           3.06           1.73   {0.5, 0.001}{89.61, 3.06, 7.33}\n",
      "      B-AlexNet          77.21           3.19           1.66   {0.5, 0.005}{89.61, 3.9, 6.49}\n",
      "      B-AlexNet          77.26           3.14           1.69    {0.5, 0.01}{89.61, 4.27, 6.12}\n",
      "      B-AlexNet          77.26           3.19           1.66    {0.5, 0.05}{89.61, 5.44, 4.95}\n",
      "      B-AlexNet          77.26           3.05           1.74     {0.5, 0.1}{89.61, 6.11, 4.28}\n",
      "      B-AlexNet          77.03           2.88           1.84     {0.5, 0.5}{89.61, 8.26, 2.13}\n",
      "      B-AlexNet          76.83           2.94           1.80    {0.5, 0.75}{89.61, 9.82, 0.57}\n",
      "      B-AlexNet          76.78           2.88           1.84     {0.5, 1.0}{89.61, 10.21, 0.18}\n",
      "      B-AlexNet          76.78           2.93           1.81     {0.5, 5.0}{89.61, 10.39, 0.0}\n",
      "      B-AlexNet          76.78           2.86           1.85    {0.5, 10.0}{89.61, 10.39, 0.0}\n",
      "      B-AlexNet          75.99           2.66           1.99 {0.75, 0.0001}{96.76, 0.37, 2.87}\n",
      "      B-AlexNet          75.96           2.83           1.87  {0.75, 0.001}{96.76, 0.79, 2.45}\n",
      "      B-AlexNet          75.99           2.91           1.82  {0.75, 0.005}{96.76, 1.1, 2.14}\n",
      "      B-AlexNet          76.00           2.88           1.84   {0.75, 0.01}{96.76, 1.2, 2.04}\n",
      "      B-AlexNet          75.97           3.03           1.75   {0.75, 0.05}{96.76, 1.53, 1.71}\n",
      "      B-AlexNet          75.99           2.84           1.87    {0.75, 0.1}{96.76, 1.78, 1.46}\n",
      "      B-AlexNet          75.94           2.83           1.87    {0.75, 0.5}{96.76, 2.48, 0.76}\n",
      "      B-AlexNet          75.93           2.94           1.80   {0.75, 0.75}{96.76, 2.93, 0.31}\n",
      "      B-AlexNet          75.90           2.93           1.81    {0.75, 1.0}{96.76, 3.16, 0.08}\n",
      "      B-AlexNet          75.93           2.90           1.82    {0.75, 5.0}{96.76, 3.24, 0.0}\n",
      "      B-AlexNet          75.93           2.92           1.81   {0.75, 10.0}{96.76, 3.24, 0.0}\n",
      "      B-AlexNet          75.54           2.75           1.93  {1.0, 0.0001}{98.97, 0.09, 0.94}\n",
      "      B-AlexNet          75.52           2.95           1.79   {1.0, 0.001}{98.97, 0.19, 0.84}\n",
      "      B-AlexNet          75.54           2.77           1.91   {1.0, 0.005}{98.97, 0.29, 0.74}\n",
      "      B-AlexNet          75.54           2.69           1.97    {1.0, 0.01}{98.97, 0.33, 0.7}\n",
      "      B-AlexNet          75.52           3.01           1.76    {1.0, 0.05}{98.97, 0.52, 0.51}\n",
      "      B-AlexNet          75.51           2.89           1.84     {1.0, 0.1}{98.97, 0.61, 0.42}\n",
      "      B-AlexNet          75.51           2.74           1.94     {1.0, 0.5}{98.97, 0.78, 0.25}\n",
      "      B-AlexNet          75.50           2.78           1.90    {1.0, 0.75}{98.97, 0.92, 0.11}\n",
      "      B-AlexNet          75.49           2.80           1.89     {1.0, 1.0}{98.97, 0.99, 0.04}\n",
      "      B-AlexNet          75.51           2.79           1.90     {1.0, 5.0}{98.97, 1.03, 0.0}\n",
      "      B-AlexNet          75.51           2.87           1.84    {1.0, 10.0}{98.97, 1.03, 0.0}\n",
      "      B-AlexNet          75.29           2.58           2.05  {5.0, 0.0001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.86   {5.0, 0.001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.74           1.94   {5.0, 0.005}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.86    {5.0, 0.01}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.85           1.86    {5.0, 0.05}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.77           1.91     {5.0, 0.1}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.73           1.94     {5.0, 0.5}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.76           1.92    {5.0, 0.75}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.66           1.99     {5.0, 1.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.86           1.85     {5.0, 5.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.72           1.95    {5.0, 10.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.79           1.90 {10.0, 0.0001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.88           1.84  {10.0, 0.001}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.86           1.85  {10.0, 0.005}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.75           1.93   {10.0, 0.01}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.80           1.89   {10.0, 0.05}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.76           1.92    {10.0, 0.1}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.71           1.95    {10.0, 0.5}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.68           1.98   {10.0, 0.75}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.80           1.89    {10.0, 1.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.89           1.83    {10.0, 5.0}{100.0, 0.0, 0.0}\n",
      "      B-AlexNet          75.29           2.82           1.88   {10.0, 10.0}{100.0, 0.0, 0.0}\n"
     ]
    }
   ],
   "source": [
    "reload(utils)\n",
    "print 'CPU Results:'\n",
    "utils.branchy_table_results('AlexNet', c_baseacc, c_basediff, c_accs, c_diffs, c_exits, c_ts)\n",
    "\n",
    "print '\\n\\n'\n",
    "\n",
    "print 'GPU Results:'\n",
    "utils.branchy_table_results('AlexNet', g_baseacc, g_basediff, g_accs, c_diffs, g_exits, g_ts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Save model/data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "import dill\n",
    "branchyNet.to_cpu()\n",
    "with open(\"_models/alexnet_cifar10.bn\", \"wb\") as f:\n",
    "    dill.dump(branchyNet, f)\n",
    "with open(\"_models/alexnet_cifar10_gpu_results.pkl\", \"wb\") as f:\n",
    "    dill.dump({'accs': g_accs, 'rt': g_diffs, 'exits': g_exits, 'ts': g_ts, 'baseacc': g_baseacc, 'basediff': g_basediff}, f)\n",
    "with open(\"_models/alexnet_cifar10_cpu_results.pkl\", \"wb\") as f:\n",
    "    dill.dump({'accs': c_accs, 'rt': c_diffs, 'exits': c_exits, 'ts': c_ts, 'baseacc': c_baseacc, 'basediff': c_basediff}, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import dill\n",
    "with open(\"_models/lenet_mnist_gpu_results.pkl\", \"rb\") as f:\n",
    "    gpu_res = dill.load(f)\n",
    "with open(\"_models/lenet_mnist_cpu_results.pkl\", \"rb\") as f:\n",
    "    cpu_res = dill.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [default]",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}