{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CIFAR 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "%reload_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.conv_learner import *\n",
    "from fastai.models.cifar10.wideresnet import wrn_22\n",
    "torch.backends.cudnn.benchmark = True\n",
    "PATH = Path(\"data/cifar10/\")\n",
    "os.makedirs(PATH,exist_ok=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "classes = ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck')\n",
    "stats = (np.array([ 0.4914 ,  0.48216,  0.44653]), np.array([ 0.24703,  0.24349,  0.26159]))\n",
    "\n",
    "bs=512\n",
    "sz=32"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "tfms = tfms_from_stats(stats, sz, aug_tfms=[RandomCrop(sz), RandomFlip()], pad=sz//8)\n",
    "data = ImageClassifierData.from_paths(PATH, val_name='test', tfms=tfms, bs=bs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "m = wrn_22()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = ConvLearner.from_model_data(m, data)\n",
    "learn.crit = nn.CrossEntropyLoss()\n",
    "learn.metrics = [accuracy]\n",
    "wd=1e-4\n",
    "lr=1.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0f6fa780bc0b4b1e978fb7572e718ce0",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=30), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch      trn_loss   val_loss   accuracy                 \n",
      "    0      1.456755   1.499619   0.5062    \n",
      "    1      1.057333   1.157792   0.6116                   \n",
      "    2      0.829041   0.783326   0.723                     \n",
      "    3      0.66619    0.808943   0.7358                    \n",
      "    4      0.570876   0.748631   0.7361                    \n",
      "    5      0.495598   1.038086   0.6717                    \n",
      "    6      0.448354   0.533581   0.8181                    \n",
      "    7      0.415957   0.546836   0.816                     \n",
      "    8      0.390528   0.61025    0.7827                    \n",
      "    9      0.36144    0.751214   0.764                     \n",
      "    10     0.351138   0.756213   0.7769                    \n",
      "    11     0.33065    0.872244   0.7549                    \n",
      "    12     0.323868   0.530568   0.8215                    \n",
      "    13     0.301522   0.633277   0.8                       \n",
      "    14     0.281426   0.609825   0.8141                    \n",
      "    15     0.261843   0.792786   0.7706                    \n",
      "    16     0.243936   0.727103   0.797                     \n",
      "    17     0.233351   0.481732   0.8525                    \n",
      "    18     0.219056   0.522896   0.8375                    \n",
      "    19     0.196971   0.350686   0.8835                    \n",
      "    20     0.180855   0.389286   0.8754                    \n",
      "    21     0.150032   0.372619   0.883                     \n",
      "    22     0.118364   0.255543   0.9182                    \n",
      "    23     0.080524   0.22061    0.9311                     \n",
      "    24     0.051989   0.207242   0.9347                     \n",
      "    25     0.03802    0.21347    0.9368                     \n",
      "    26     0.030564   0.211374   0.9381                     \n",
      "    27     0.023117   0.214783   0.9398                     \n",
      "    28     0.020133   0.21228    0.9421                     \n",
      "    29     0.017761   0.212101   0.9423                     \n",
      "\n",
      "CPU times: user 34min 14s, sys: 54min 24s, total: 1h 28min 38s\n",
      "Wall time: 17min 16s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[array([0.2121]), 0.9423000004768372]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "%time learn.fit(lr, 1, wds=wd, cycle_len=30, use_clr_beta=(20,20,0.95,0.85))"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  },
  "toc": {
   "colors": {
    "hover_highlight": "#DAA520",
    "navigate_num": "#000000",
    "navigate_text": "#333333",
    "running_highlight": "#FF0000",
    "selected_highlight": "#FFD700",
    "sidebar_border": "#EEEEEE",
    "wrapper_background": "#FFFFFF"
   },
   "moveMenuLeft": true,
   "nav_menu": {
    "height": "266px",
    "width": "252px"
   },
   "navigate_menu": true,
   "number_sections": true,
   "sideBar": true,
   "threshold": 4,
   "toc_cell": false,
   "toc_section_display": "block",
   "toc_window_display": false,
   "widenNotebook": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}