fastbook/clean/05_pet_breeds.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#hide\n",
    "!pip install -Uqq fastbook\n",
    "import fastbook\n",
    "fastbook.setup_book()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#hide\n",
    "from fastbook import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Image Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## From Dogs and Cats to Pet Breeds"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.vision.all import *\n",
    "path = untar_data(URLs.PETS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#hide\n",
    "Path.BASE_PATH = path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(#3) [Path('annotations'),Path('images'),Path('models')]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "path.ls()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(#7394) [Path('images/great_pyrenees_173.jpg'),Path('images/wheaten_terrier_46.jpg'),Path('images/Ragdoll_262.jpg'),Path('images/german_shorthaired_3.jpg'),Path('images/american_bulldog_196.jpg'),Path('images/boxer_188.jpg'),Path('images/staffordshire_bull_terrier_173.jpg'),Path('images/basset_hound_71.jpg'),Path('images/staffordshire_bull_terrier_37.jpg'),Path('images/yorkshire_terrier_18.jpg')...]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(path/\"images\").ls()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fname = (path/\"images\").ls()[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['great_pyrenees']"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "re.findall(r'(.+)_\\d+.jpg$', fname.name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "pets = DataBlock(blocks = (ImageBlock, CategoryBlock),\n",
    "                 get_items=get_image_files, \n",
    "                 splitter=RandomSplitter(seed=42),\n",
    "                 get_y=using_attr(RegexLabeller(r'(.+)_\\d+.jpg$'), 'name'),\n",
    "                 item_tfms=Resize(460),\n",
    "                 batch_tfms=aug_transforms(size=224, min_scale=0.75))\n",
    "dls = pets.dataloaders(path/\"images\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Presizing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hide_input": false
   },
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dblock1 = DataBlock(blocks=(ImageBlock(), CategoryBlock()),\n",
    "                   get_y=parent_label,\n",
    "                   item_tfms=Resize(460))\n",
    "dls1 = dblock1.dataloaders([(Path.cwd()/'images'/'grizzly.jpg')]*100, bs=8)\n",
    "dls1.train.get_idxs = lambda: Inf.ones\n",
    "x,y = dls1.valid.one_batch()\n",
    "_,axs = subplots(1, 2)\n",
    "\n",
    "x1 = TensorImage(x.clone())\n",
    "x1 = x1.affine_coord(sz=224)\n",
    "x1 = x1.rotate(draw=30, p=1.)\n",
    "x1 = x1.zoom(draw=1.2, p=1.)\n",
    "x1 = x1.warp(draw_x=-0.2, draw_y=0.2, p=1.)\n",
    "\n",
    "tfms = setup_aug_tfms([Rotate(draw=30, p=1, size=224), Zoom(draw=1.2, p=1., size=224),\n",
    "                       Warp(draw_x=-0.2, draw_y=0.2, p=1., size=224)])\n",
    "x = Pipeline(tfms)(x)\n",
    "#x.affine_coord(coord_tfm=coord_tfm, sz=size, mode=mode, pad_mode=pad_mode)\n",
    "TensorImage(x[0]).show(ctx=axs[0])\n",
    "TensorImage(x1[0]).show(ctx=axs[1]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking and Debugging a DataBlock"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 648x216 with 3 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dls.show_batch(nrows=1, ncols=3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Setting-up type transforms pipelines\n",
      "Collecting items from /home/jhoward/.fastai/data/oxford-iiit-pet/images\n",
      "Found 7390 items\n",
      "2 datasets of sizes 5912,1478\n",
      "Setting up Pipeline: PILBase.create\n",
      "Setting up Pipeline: partial -> Categorize\n",
      "\n",
      "Building one sample\n",
      "  Pipeline: PILBase.create\n",
      "    starting from\n",
      "      /home/jhoward/.fastai/data/oxford-iiit-pet/images/american_pit_bull_terrier_31.jpg\n",
      "    applying PILBase.create gives\n",
      "      PILImage mode=RGB size=500x414\n",
      "  Pipeline: partial -> Categorize\n",
      "    starting from\n",
      "      /home/jhoward/.fastai/data/oxford-iiit-pet/images/american_pit_bull_terrier_31.jpg\n",
      "    applying partial gives\n",
      "      american_pit_bull_terrier\n",
      "    applying Categorize gives\n",
      "      TensorCategory(13)\n",
      "\n",
      "Final sample: (PILImage mode=RGB size=500x414, TensorCategory(13))\n",
      "\n",
      "\n",
      "Setting up after_item: Pipeline: ToTensor\n",
      "Setting up before_batch: Pipeline: \n",
      "Setting up after_batch: Pipeline: IntToFloatTensor\n",
      "\n",
      "Building one batch\n",
      "Applying item_tfms to the first sample:\n",
      "  Pipeline: ToTensor\n",
      "    starting from\n",
      "      (PILImage mode=RGB size=500x414, TensorCategory(13))\n",
      "    applying ToTensor gives\n",
      "      (TensorImage of size 3x414x500, TensorCategory(13))\n",
      "\n",
      "Adding the next 3 samples\n",
      "\n",
      "No before_batch transform to apply\n",
      "\n",
      "Collating items in a batch\n",
      "Error! It's not possible to collate your items in a batch\n",
      "Could not collate the 0-th members of your tuples because got the following shapes\n",
      "torch.Size([3, 414, 500]),torch.Size([3, 375, 500]),torch.Size([3, 500, 281]),torch.Size([3, 203, 300])\n"
     ]
    },
    {
     "ename": "RuntimeError",
     "evalue": "invalid argument 0: Sizes of tensors must match except in dimension 0. Got 414 and 375 in dimension 2 at /opt/conda/conda-bld/pytorch_1579022060824/work/aten/src/TH/generic/THTensor.cpp:612",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-11-8c0a3d421ca2>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      4\u001b[0m                  \u001b[0msplitter\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mRandomSplitter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mseed\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m42\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m                  get_y=using_attr(RegexLabeller(r'(.+)_\\d+.jpg$'), 'name'))\n\u001b[0;32m----> 6\u001b[0;31m \u001b[0mpets1\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msummary\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpath\u001b[0m\u001b[0;34m/\u001b[0m\u001b[0;34m\"images\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/git/fastai/fastai/data/block.py\u001b[0m in \u001b[0;36msummary\u001b[0;34m(self, source, bs, show_batch, **kwargs)\u001b[0m\n\u001b[1;32m    182\u001b[0m         \u001b[0mwhy\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0m_find_fail_collate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    183\u001b[0m         \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Make sure all parts of your samples are tensors of the same size\"\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mwhy\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mwhy\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 184\u001b[0;31m         \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    185\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    186\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mf\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mf\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mdls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mafter_batch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfs\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mname\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34m'noop'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m!=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/fastai/fastai/data/block.py\u001b[0m in \u001b[0;36msummary\u001b[0;34m(self, source, bs, show_batch, **kwargs)\u001b[0m\n\u001b[1;32m    176\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"\\nCollating items in a batch\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    177\u001b[0m     \u001b[0;32mtry\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 178\u001b[0;31m         \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdls\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    179\u001b[0m         \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mretain_types\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mis_listy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    180\u001b[0m     \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/fastai/fastai/data/load.py\u001b[0m in \u001b[0;36mcreate_batch\u001b[0;34m(self, b)\u001b[0m\n\u001b[1;32m    125\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mretain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mres\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;32mreturn\u001b[0m \u001b[0mretain_types\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mres\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mis_listy\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    126\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mcreate_item\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;32mreturn\u001b[0m \u001b[0mnext\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mit\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdataset\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 127\u001b[0;31m     \u001b[0;32mdef\u001b[0m \u001b[0mcreate_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mfa_collate\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mfa_convert\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprebatched\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    128\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mdo_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mretain\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcreate_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbefore_batch\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mb\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    129\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mto\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdevice\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdevice\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/fastai/fastai/data/load.py\u001b[0m in \u001b[0;36mfa_collate\u001b[0;34m(t)\u001b[0m\n\u001b[1;32m     44\u001b[0m     \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     45\u001b[0m     return (default_collate(t) if isinstance(b, _collate_types)\n\u001b[0;32m---> 46\u001b[0;31m             \u001b[0;32melse\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mfa_collate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mSequence\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     47\u001b[0m             else default_collate(t))\n\u001b[1;32m     48\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/fastai/fastai/data/load.py\u001b[0m in \u001b[0;36m<listcomp>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m     44\u001b[0m     \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     45\u001b[0m     return (default_collate(t) if isinstance(b, _collate_types)\n\u001b[0;32m---> 46\u001b[0;31m             \u001b[0;32melse\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mfa_collate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mSequence\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     47\u001b[0m             else default_collate(t))\n\u001b[1;32m     48\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/git/fastai/fastai/data/load.py\u001b[0m in \u001b[0;36mfa_collate\u001b[0;34m(t)\u001b[0m\n\u001b[1;32m     43\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mfa_collate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     44\u001b[0m     \u001b[0mb\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 45\u001b[0;31m     return (default_collate(t) if isinstance(b, _collate_types)\n\u001b[0m\u001b[1;32m     46\u001b[0m             \u001b[0;32melse\u001b[0m \u001b[0mtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mfa_collate\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mt\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0misinstance\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mb\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mSequence\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     47\u001b[0m             else default_collate(t))\n",
      "\u001b[0;32m~/anaconda3/lib/python3.7/site-packages/torch/utils/data/_utils/collate.py\u001b[0m in \u001b[0;36mdefault_collate\u001b[0;34m(batch)\u001b[0m\n\u001b[1;32m     53\u001b[0m             \u001b[0mstorage\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0melem\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstorage\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_new_shared\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnumel\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     54\u001b[0m             \u001b[0mout\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0melem\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnew\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstorage\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 55\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstack\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mbatch\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mout\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     56\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0melem_type\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__module__\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;34m'numpy'\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0melem_type\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34m'str_'\u001b[0m\u001b[0;31m \u001b[0m\u001b[0;31m\\\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     57\u001b[0m             \u001b[0;32mand\u001b[0m \u001b[0melem_type\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__name__\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0;34m'string_'\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mRuntimeError\u001b[0m: invalid argument 0: Sizes of tensors must match except in dimension 0. Got 414 and 375 in dimension 2 at /opt/conda/conda-bld/pytorch_1579022060824/work/aten/src/TH/generic/THTensor.cpp:612"
     ]
    }
   ],
   "source": [
    "pets1 = DataBlock(blocks = (ImageBlock, CategoryBlock),\n",
    "                 get_items=get_image_files, \n",
    "                 splitter=RandomSplitter(seed=42),\n",
    "                 get_y=using_attr(RegexLabeller(r'(.+)_\\d+.jpg$'), 'name'))\n",
    "pets1.summary(path/\"images\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.551305</td>\n",
       "      <td>0.322132</td>\n",
       "      <td>0.106225</td>\n",
       "      <td>00:19</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.529473</td>\n",
       "      <td>0.312148</td>\n",
       "      <td>0.095399</td>\n",
       "      <td>00:23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.330207</td>\n",
       "      <td>0.245883</td>\n",
       "      <td>0.080514</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "learn.fine_tune(2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Cross-Entropy Loss"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Viewing Activations and Labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "x,y = dls.one_batch()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "TensorCategory([ 0,  5, 23, 36,  5, 20, 29, 34, 33, 32, 31, 24, 12, 36,  8, 26, 30,  2, 12, 17,  7, 23, 12, 29, 21,  4, 35, 33,  0, 20, 26, 30,  3,  6, 36,  2, 17, 32, 11,  6,  3, 30,  5, 26, 26, 29,  7, 36,\n",
       "        31, 26, 26,  8, 13, 30, 11, 12, 36, 31, 34, 20, 15,  8,  8, 23], device='cuda:5')"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "tensor([9.9911e-01, 5.0433e-05, 3.7515e-07, 8.8590e-07, 8.1794e-05, 1.8991e-05, 9.9280e-06, 5.4656e-07, 6.7920e-06, 2.3486e-04, 3.7872e-04, 2.0796e-05, 4.0443e-07, 1.6933e-07, 2.0502e-07, 3.1354e-08,\n",
       "        9.4115e-08, 2.9782e-06, 2.0243e-07, 8.5262e-08, 1.0900e-07, 1.0175e-07, 4.4780e-09, 1.4285e-07, 1.0718e-07, 8.1411e-07, 3.6618e-07, 4.0950e-07, 3.8525e-08, 2.3660e-07, 5.3747e-08, 2.5448e-07,\n",
       "        6.5860e-08, 8.0937e-05, 2.7464e-07, 5.6760e-07, 1.5462e-08])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds,_ = learn.get_preds(dl=[(x,y)])\n",
    "preds[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(37, tensor(1.0000))"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(preds[0]),preds[0].sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Softmax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_function(torch.sigmoid, min=-4,max=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#hide\n",
    "torch.random.manual_seed(42);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 0.6734,  0.2576],\n",
       "        [ 0.4689,  0.4607],\n",
       "        [-2.2457, -0.3727],\n",
       "        [ 4.4164, -1.2760],\n",
       "        [ 0.9233,  0.5347],\n",
       "        [ 1.0698,  1.6187]])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts = torch.randn((6,2))*2\n",
    "acts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.6623, 0.5641],\n",
       "        [0.6151, 0.6132],\n",
       "        [0.0957, 0.4079],\n",
       "        [0.9881, 0.2182],\n",
       "        [0.7157, 0.6306],\n",
       "        [0.7446, 0.8346]])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "acts.sigmoid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([0.6025, 0.5021, 0.1332, 0.9966, 0.5959, 0.3661])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(acts[:,0]-acts[:,1]).sigmoid()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.6025, 0.3975],\n",
       "        [0.5021, 0.4979],\n",
       "        [0.1332, 0.8668],\n",
       "        [0.9966, 0.0034],\n",
       "        [0.5959, 0.4041],\n",
       "        [0.3661, 0.6339]])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm_acts = torch.softmax(acts, dim=1)\n",
    "sm_acts"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Log Likelihood"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "targ = tensor([0,1,0,1,1,0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[0.6025, 0.3975],\n",
       "        [0.5021, 0.4979],\n",
       "        [0.1332, 0.8668],\n",
       "        [0.9966, 0.0034],\n",
       "        [0.5959, 0.4041],\n",
       "        [0.3661, 0.6339]])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sm_acts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([0.6025, 0.4979, 0.1332, 0.0034, 0.4041, 0.3661])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "idx = range(6)\n",
    "sm_acts[idx, targ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table ><thead>    <tr>        <th class=\"col_heading level0 col0\" >3</th>        <th class=\"col_heading level0 col1\" >7</th>        <th class=\"col_heading level0 col2\" >targ</th>        <th class=\"col_heading level0 col3\" >idx</th>        <th class=\"col_heading level0 col4\" >loss</th>    </tr></thead><tbody>\n",
       "                <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row0_col0\" class=\"data row0 col0\" >0.602469</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row0_col1\" class=\"data row0 col1\" >0.397531</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row0_col2\" class=\"data row0 col2\" >0</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row0_col3\" class=\"data row0 col3\" >0</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row0_col4\" class=\"data row0 col4\" >0.602469</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row1_col0\" class=\"data row1 col0\" >0.502065</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row1_col1\" class=\"data row1 col1\" >0.497935</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row1_col2\" class=\"data row1 col2\" >1</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row1_col3\" class=\"data row1 col3\" >1</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row1_col4\" class=\"data row1 col4\" >0.497935</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row2_col0\" class=\"data row2 col0\" >0.133188</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row2_col1\" class=\"data row2 col1\" >0.866811</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row2_col2\" class=\"data row2 col2\" >0</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row2_col3\" class=\"data row2 col3\" >2</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row2_col4\" class=\"data row2 col4\" >0.133188</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row3_col0\" class=\"data row3 col0\" >0.99664</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row3_col1\" class=\"data row3 col1\" >0.00336017</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row3_col2\" class=\"data row3 col2\" >1</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row3_col3\" class=\"data row3 col3\" >3</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row3_col4\" class=\"data row3 col4\" >0.00336017</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row4_col0\" class=\"data row4 col0\" >0.595949</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row4_col1\" class=\"data row4 col1\" >0.404051</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row4_col2\" class=\"data row4 col2\" >1</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row4_col3\" class=\"data row4 col3\" >4</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row4_col4\" class=\"data row4 col4\" >0.404051</td>\n",
       "            </tr>\n",
       "            <tr>\n",
       "                                <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row5_col0\" class=\"data row5 col0\" >0.366118</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row5_col1\" class=\"data row5 col1\" >0.633882</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row5_col2\" class=\"data row5 col2\" >0</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row5_col3\" class=\"data row5 col3\" >5</td>\n",
       "                        <td id=\"T_6b1e5e2c_8421_11ea_8806_bdb274b1aa15row5_col4\" class=\"data row5 col4\" >0.366118</td>\n",
       "            </tr>\n",
       "    </tbody></table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from IPython.display import HTML\n",
    "df = pd.DataFrame(sm_acts, columns=[\"3\",\"7\"])\n",
    "df['targ'] = targ\n",
    "df['idx'] = idx\n",
    "df['loss'] = sm_acts[range(6), targ]\n",
    "t = df.style.hide_index()\n",
    "#To have html code compatible with our script\n",
    "html = t._repr_html_().split('</style>')[1]\n",
    "html = re.sub(r'<table id=\"([^\"]+)\"\\s*>', r'<table >', html)\n",
    "display(HTML(html))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([-0.6025, -0.4979, -0.1332, -0.0034, -0.4041, -0.3661])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "-sm_acts[idx, targ]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([-0.6025, -0.4979, -0.1332, -0.0034, -0.4041, -0.3661])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "F.nll_loss(sm_acts, targ, reduction='none')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Taking the Log"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_function(torch.log, min=0,max=4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "loss_func = nn.CrossEntropyLoss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(1.8045)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "loss_func(acts, targ)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor(1.8045)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "F.cross_entropy(acts, targ)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([0.5067, 0.6973, 2.0160, 5.6958, 0.9062, 1.0048])"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "nn.CrossEntropyLoss(reduction='none')(acts, targ)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Model Interpretation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "interp = ClassificationInterpretation.from_learner(learn)\n",
    "interp.plot_confusion_matrix(figsize=(12,12), dpi=60)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('american_pit_bull_terrier', 'staffordshire_bull_terrier', 10),\n",
       " ('Ragdoll', 'Birman', 8),\n",
       " ('Siamese', 'Birman', 6),\n",
       " ('Bengal', 'Egyptian_Mau', 5),\n",
       " ('american_pit_bull_terrier', 'american_bulldog', 5)]"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interp.most_confused(min_val=5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Improving Our Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The Learning Rate Finder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>2.778816</td>\n",
       "      <td>5.150732</td>\n",
       "      <td>0.504060</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>4.354680</td>\n",
       "      <td>3.003533</td>\n",
       "      <td>0.834235</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "learn.fine_tune(1, base_lr=0.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "lr_min,lr_steep = learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Minimum/10: 1.00e-02, steepest point: 5.25e-03\n"
     ]
    }
   ],
   "source": [
    "print(f\"Minimum/10: {lr_min:.2e}, steepest point: {lr_steep:.2e}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.328591</td>\n",
       "      <td>0.344678</td>\n",
       "      <td>0.114344</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.540180</td>\n",
       "      <td>0.420945</td>\n",
       "      <td>0.127876</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.329827</td>\n",
       "      <td>0.248813</td>\n",
       "      <td>0.083221</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "learn.fine_tune(2, base_lr=3e-3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unfreezing and Transfer Learning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.fine_tune??"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.188042</td>\n",
       "      <td>0.355024</td>\n",
       "      <td>0.102842</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.534234</td>\n",
       "      <td>0.302453</td>\n",
       "      <td>0.094723</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.325031</td>\n",
       "      <td>0.222268</td>\n",
       "      <td>0.074425</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "learn.fit_one_cycle(3, 3e-3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(1.0964782268274575e-05, 1.5848931980144698e-06)"
      ]
     },
     "execution_count": null,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.263579</td>\n",
       "      <td>0.217419</td>\n",
       "      <td>0.069012</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.253060</td>\n",
       "      <td>0.210346</td>\n",
       "      <td>0.062923</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.224340</td>\n",
       "      <td>0.207357</td>\n",
       "      <td>0.060217</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.200195</td>\n",
       "      <td>0.207244</td>\n",
       "      <td>0.061570</td>\n",
       "      <td>00:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.194269</td>\n",
       "      <td>0.200149</td>\n",
       "      <td>0.059540</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>0.173164</td>\n",
       "      <td>0.202301</td>\n",
       "      <td>0.059540</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(6, lr_max=1e-5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Discriminative Learning Rates"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.145300</td>\n",
       "      <td>0.345568</td>\n",
       "      <td>0.119756</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.533986</td>\n",
       "      <td>0.251944</td>\n",
       "      <td>0.077131</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.317696</td>\n",
       "      <td>0.208371</td>\n",
       "      <td>0.069012</td>\n",
       "      <td>00:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.257977</td>\n",
       "      <td>0.205400</td>\n",
       "      <td>0.067659</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.246763</td>\n",
       "      <td>0.205107</td>\n",
       "      <td>0.066306</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.240595</td>\n",
       "      <td>0.193848</td>\n",
       "      <td>0.062246</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.209988</td>\n",
       "      <td>0.198061</td>\n",
       "      <td>0.062923</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.194756</td>\n",
       "      <td>0.193130</td>\n",
       "      <td>0.064276</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>0.169985</td>\n",
       "      <td>0.187885</td>\n",
       "      <td>0.056157</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>0.153205</td>\n",
       "      <td>0.186145</td>\n",
       "      <td>0.058863</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>0.141480</td>\n",
       "      <td>0.185316</td>\n",
       "      <td>0.053451</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>0.128564</td>\n",
       "      <td>0.180999</td>\n",
       "      <td>0.051421</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>0.126941</td>\n",
       "      <td>0.186288</td>\n",
       "      <td>0.054127</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>0.130064</td>\n",
       "      <td>0.181764</td>\n",
       "      <td>0.054127</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>0.124281</td>\n",
       "      <td>0.181855</td>\n",
       "      <td>0.054127</td>\n",
       "      <td>00:25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn = cnn_learner(dls, resnet34, metrics=error_rate)\n",
    "learn.fit_one_cycle(3, 3e-3)\n",
    "learn.unfreeze()\n",
    "learn.fit_one_cycle(12, lr_max=slice(1e-6,1e-4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.recorder.plot_loss()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Selecting the Number of Epochs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Deeper Architectures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>1.427505</td>\n",
       "      <td>0.310554</td>\n",
       "      <td>0.098782</td>\n",
       "      <td>00:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.606785</td>\n",
       "      <td>0.302325</td>\n",
       "      <td>0.094723</td>\n",
       "      <td>00:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.409267</td>\n",
       "      <td>0.294803</td>\n",
       "      <td>0.091340</td>\n",
       "      <td>00:21</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>error_rate</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.261121</td>\n",
       "      <td>0.274507</td>\n",
       "      <td>0.083897</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.296653</td>\n",
       "      <td>0.318649</td>\n",
       "      <td>0.084574</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.242356</td>\n",
       "      <td>0.253677</td>\n",
       "      <td>0.069012</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.150684</td>\n",
       "      <td>0.251438</td>\n",
       "      <td>0.065629</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.094997</td>\n",
       "      <td>0.239772</td>\n",
       "      <td>0.064276</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>0.061144</td>\n",
       "      <td>0.228082</td>\n",
       "      <td>0.054804</td>\n",
       "      <td>00:26</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from fastai.callback.fp16 import *\n",
    "learn = cnn_learner(dls, resnet50, metrics=error_rate).to_fp16()\n",
    "learn.fine_tune(6, freeze_epochs=3)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Conclusion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Questionnaire"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Why do we first resize to a large size on the CPU, and then to a smaller size on the GPU?\n",
    "1. If you are not familiar with regular expressions, find a regular expression tutorial, and some problem sets, and complete them. Have a look on the book's website for suggestions.\n",
    "1. What are the two ways in which data is most commonly provided, for most deep learning datasets?\n",
    "1. Look up the documentation for `L` and try using a few of the new methods is that it adds.\n",
    "1. Look up the documentation for the Python `pathlib` module and try using a few methods of the `Path` class.\n",
    "1. Give two examples of ways that image transformations can degrade the quality of the data.\n",
    "1. What method does fastai provide to view the data in a `DataLoaders`?\n",
    "1. What method does fastai provide to help you debug a `DataBlock`?\n",
    "1. Should you hold off on training a model until you have thoroughly cleaned your data?\n",
    "1. What are the two pieces that are combined into cross-entropy loss in PyTorch?\n",
    "1. What are the two properties of activations that softmax ensures? Why is this important?\n",
    "1. When might you want your activations to not have these two properties?\n",
    "1. Calculate the `exp` and `softmax` columns of <<bear_softmax>> yourself (i.e., in a spreadsheet, with a calculator, or in a notebook).\n",
    "1. Why can't we use `torch.where` to create a loss function for datasets where our label can have more than two categories?\n",
    "1. What is the value of log(-2)? Why?\n",
    "1. What are two good rules of thumb for picking a learning rate from the learning rate finder?\n",
    "1. What two steps does the `fine_tune` method do?\n",
    "1. In Jupyter Notebook, how do you get the source code for a method or function?\n",
    "1. What are discriminative learning rates?\n",
    "1. How is a Python `slice` object interpreted when passed as a learning rate to fastai?\n",
    "1. Why is early stopping a poor choice when using 1cycle training?\n",
    "1. What is the difference between `resnet50` and `resnet101`?\n",
    "1. What does `to_fp16` do?"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Further Research"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Find the paper by Leslie Smith that introduced the learning rate finder, and read it.\n",
    "1. See if you can improve the accuracy of the classifier in this chapter. What's the best accuracy you can achieve? Look on the forums and the book's website to see what other students have achieved with this dataset, and how they did it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "jupytext": {
   "split_at_heading": true
  },
  "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": "3.7.7"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": false,
   "sideBar": true,
   "skip_h1_title": true,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}